diff --git "a/codet5_qlora_official_0.001/checkpoint-117744/trainer_state.json" "b/codet5_qlora_official_0.001/checkpoint-117744/trainer_state.json" new file mode 100644--- /dev/null +++ "b/codet5_qlora_official_0.001/checkpoint-117744/trainer_state.json" @@ -0,0 +1,165045 @@ +{ + "best_metric": 0.021544713242252534, + "best_model_checkpoint": "./results-cc/code-t5/codet5_qlora_official_0.001/checkpoint-88308", + "epoch": 8.0, + "eval_steps": 500, + "global_step": 117744, + "is_hyper_param_search": false, + "is_local_process_zero": true, + "is_world_process_zero": true, + "log_history": [ + { + "epoch": 0.0003397200706617747, + "grad_norm": 2.207773447036743, + "learning_rate": 0.0009999575349911674, + "loss": 9.2175, + "step": 5 + }, + { + "epoch": 0.0006794401413235494, + "grad_norm": 0.8029590249061584, + "learning_rate": 0.0009999150699823347, + "loss": 6.2854, + "step": 10 + }, + { + "epoch": 0.0010191602119853241, + "grad_norm": 1.3657054901123047, + "learning_rate": 0.0009998726049735018, + "loss": 5.4114, + "step": 15 + }, + { + "epoch": 0.001358880282647099, + "grad_norm": 0.7428395748138428, + "learning_rate": 0.0009998301399646691, + "loss": 4.4378, + "step": 20 + }, + { + "epoch": 0.0016986003533088735, + "grad_norm": 0.5788111090660095, + "learning_rate": 0.0009997876749558365, + "loss": 4.3423, + "step": 25 + }, + { + "epoch": 0.0020383204239706482, + "grad_norm": 0.5947818160057068, + "learning_rate": 0.0009997452099470036, + "loss": 4.1356, + "step": 30 + }, + { + "epoch": 0.002378040494632423, + "grad_norm": 0.7070780992507935, + "learning_rate": 0.000999702744938171, + "loss": 3.9679, + "step": 35 + }, + { + "epoch": 0.002717760565294198, + "grad_norm": 1.4084078073501587, + "learning_rate": 0.0009996602799293383, + "loss": 4.2031, + "step": 40 + }, + { + "epoch": 0.0030574806359559724, + "grad_norm": 0.570355236530304, + "learning_rate": 0.0009996178149205056, + "loss": 4.3601, + "step": 45 + }, + { + "epoch": 0.003397200706617747, + "grad_norm": 9.142182350158691, + "learning_rate": 0.0009995753499116727, + "loss": 3.9385, + "step": 50 + }, + { + "epoch": 0.0037369207772795215, + "grad_norm": 0.5780104398727417, + "learning_rate": 0.00099953288490284, + "loss": 4.0844, + "step": 55 + }, + { + "epoch": 0.0040766408479412965, + "grad_norm": 1.0214359760284424, + "learning_rate": 0.0009994904198940074, + "loss": 3.9468, + "step": 60 + }, + { + "epoch": 0.0044163609186030715, + "grad_norm": 0.4513009488582611, + "learning_rate": 0.0009994479548851745, + "loss": 4.0269, + "step": 65 + }, + { + "epoch": 0.004756080989264846, + "grad_norm": 0.4692077040672302, + "learning_rate": 0.000999405489876342, + "loss": 4.2709, + "step": 70 + }, + { + "epoch": 0.005095801059926621, + "grad_norm": 0.6555360555648804, + "learning_rate": 0.0009993630248675092, + "loss": 4.1327, + "step": 75 + }, + { + "epoch": 0.005435521130588396, + "grad_norm": 1.6717242002487183, + "learning_rate": 0.0009993205598586765, + "loss": 3.8472, + "step": 80 + }, + { + "epoch": 0.00577524120125017, + "grad_norm": 0.559617817401886, + "learning_rate": 0.0009992780948498438, + "loss": 3.9607, + "step": 85 + }, + { + "epoch": 0.006114961271911945, + "grad_norm": 2.2418723106384277, + "learning_rate": 0.000999235629841011, + "loss": 4.0371, + "step": 90 + }, + { + "epoch": 0.006454681342573719, + "grad_norm": 0.4393124282360077, + "learning_rate": 0.0009991931648321783, + "loss": 4.2313, + "step": 95 + }, + { + "epoch": 0.006794401413235494, + "grad_norm": 0.7975866794586182, + "learning_rate": 0.0009991506998233456, + "loss": 4.1234, + "step": 100 + }, + { + "epoch": 0.007134121483897269, + "grad_norm": 0.7241436243057251, + "learning_rate": 0.000999108234814513, + "loss": 3.655, + "step": 105 + }, + { + "epoch": 0.007473841554559043, + "grad_norm": 0.45974311232566833, + "learning_rate": 0.00099906576980568, + "loss": 4.0754, + "step": 110 + }, + { + "epoch": 0.007813561625220818, + "grad_norm": 0.42426973581314087, + "learning_rate": 0.0009990233047968474, + "loss": 4.0885, + "step": 115 + }, + { + "epoch": 0.008153281695882593, + "grad_norm": 0.7785698771476746, + "learning_rate": 0.0009989808397880147, + "loss": 4.2543, + "step": 120 + }, + { + "epoch": 0.008493001766544368, + "grad_norm": 1.1263967752456665, + "learning_rate": 0.0009989383747791819, + "loss": 4.0638, + "step": 125 + }, + { + "epoch": 0.008832721837206143, + "grad_norm": 0.5730156302452087, + "learning_rate": 0.0009988959097703492, + "loss": 3.8552, + "step": 130 + }, + { + "epoch": 0.009172441907867916, + "grad_norm": 0.5796052813529968, + "learning_rate": 0.0009988534447615165, + "loss": 3.8604, + "step": 135 + }, + { + "epoch": 0.009512161978529691, + "grad_norm": 0.6022884249687195, + "learning_rate": 0.0009988109797526839, + "loss": 3.8499, + "step": 140 + }, + { + "epoch": 0.009851882049191466, + "grad_norm": 0.8763172030448914, + "learning_rate": 0.0009987685147438512, + "loss": 3.8815, + "step": 145 + }, + { + "epoch": 0.010191602119853241, + "grad_norm": 0.7262821197509766, + "learning_rate": 0.0009987260497350183, + "loss": 4.0481, + "step": 150 + }, + { + "epoch": 0.010531322190515016, + "grad_norm": 1.2450162172317505, + "learning_rate": 0.0009986835847261856, + "loss": 3.8825, + "step": 155 + }, + { + "epoch": 0.010871042261176791, + "grad_norm": 1.4473909139633179, + "learning_rate": 0.000998641119717353, + "loss": 3.9704, + "step": 160 + }, + { + "epoch": 0.011210762331838564, + "grad_norm": 0.9646778106689453, + "learning_rate": 0.00099859865470852, + "loss": 4.0052, + "step": 165 + }, + { + "epoch": 0.01155048240250034, + "grad_norm": 0.5259531140327454, + "learning_rate": 0.0009985561896996874, + "loss": 3.9015, + "step": 170 + }, + { + "epoch": 0.011890202473162114, + "grad_norm": 0.5808979272842407, + "learning_rate": 0.0009985137246908548, + "loss": 3.9114, + "step": 175 + }, + { + "epoch": 0.01222992254382389, + "grad_norm": 0.6354779005050659, + "learning_rate": 0.000998471259682022, + "loss": 3.9704, + "step": 180 + }, + { + "epoch": 0.012569642614485664, + "grad_norm": 0.5465064644813538, + "learning_rate": 0.0009984287946731894, + "loss": 3.9919, + "step": 185 + }, + { + "epoch": 0.012909362685147438, + "grad_norm": 0.7034687399864197, + "learning_rate": 0.0009983863296643566, + "loss": 3.89, + "step": 190 + }, + { + "epoch": 0.013249082755809213, + "grad_norm": 0.5259936451911926, + "learning_rate": 0.0009983438646555239, + "loss": 4.1512, + "step": 195 + }, + { + "epoch": 0.013588802826470988, + "grad_norm": 0.6321644186973572, + "learning_rate": 0.0009983013996466912, + "loss": 3.9967, + "step": 200 + }, + { + "epoch": 0.013928522897132763, + "grad_norm": 0.4751060903072357, + "learning_rate": 0.0009982589346378583, + "loss": 4.1144, + "step": 205 + }, + { + "epoch": 0.014268242967794538, + "grad_norm": 2.088681936264038, + "learning_rate": 0.0009982164696290257, + "loss": 3.8195, + "step": 210 + }, + { + "epoch": 0.014607963038456313, + "grad_norm": 0.4030695855617523, + "learning_rate": 0.000998174004620193, + "loss": 3.8732, + "step": 215 + }, + { + "epoch": 0.014947683109118086, + "grad_norm": 0.492465615272522, + "learning_rate": 0.0009981315396113603, + "loss": 4.0153, + "step": 220 + }, + { + "epoch": 0.015287403179779861, + "grad_norm": 0.4808414876461029, + "learning_rate": 0.0009980890746025275, + "loss": 3.9177, + "step": 225 + }, + { + "epoch": 0.015627123250441636, + "grad_norm": 0.4167867600917816, + "learning_rate": 0.0009980466095936948, + "loss": 3.8987, + "step": 230 + }, + { + "epoch": 0.01596684332110341, + "grad_norm": 0.9164467453956604, + "learning_rate": 0.0009980041445848621, + "loss": 3.9862, + "step": 235 + }, + { + "epoch": 0.016306563391765186, + "grad_norm": 0.6926063299179077, + "learning_rate": 0.0009979616795760292, + "loss": 3.8676, + "step": 240 + }, + { + "epoch": 0.01664628346242696, + "grad_norm": 0.5375890135765076, + "learning_rate": 0.0009979192145671968, + "loss": 3.9015, + "step": 245 + }, + { + "epoch": 0.016986003533088736, + "grad_norm": 0.5443463325500488, + "learning_rate": 0.000997876749558364, + "loss": 4.0356, + "step": 250 + }, + { + "epoch": 0.01732572360375051, + "grad_norm": 0.6313775777816772, + "learning_rate": 0.0009978342845495313, + "loss": 3.8561, + "step": 255 + }, + { + "epoch": 0.017665443674412286, + "grad_norm": 0.8847453594207764, + "learning_rate": 0.0009977918195406986, + "loss": 4.0132, + "step": 260 + }, + { + "epoch": 0.01800516374507406, + "grad_norm": 0.48864248394966125, + "learning_rate": 0.0009977493545318657, + "loss": 4.0664, + "step": 265 + }, + { + "epoch": 0.018344883815735832, + "grad_norm": 0.6228508949279785, + "learning_rate": 0.000997706889523033, + "loss": 3.8738, + "step": 270 + }, + { + "epoch": 0.01868460388639761, + "grad_norm": 0.7404700517654419, + "learning_rate": 0.0009976644245142004, + "loss": 3.7775, + "step": 275 + }, + { + "epoch": 0.019024323957059382, + "grad_norm": 0.47269967198371887, + "learning_rate": 0.0009976219595053677, + "loss": 3.8748, + "step": 280 + }, + { + "epoch": 0.01936404402772116, + "grad_norm": 5.355066299438477, + "learning_rate": 0.0009975794944965348, + "loss": 3.9724, + "step": 285 + }, + { + "epoch": 0.019703764098382932, + "grad_norm": 0.7765783667564392, + "learning_rate": 0.0009975370294877022, + "loss": 3.992, + "step": 290 + }, + { + "epoch": 0.020043484169044706, + "grad_norm": 0.5294786691665649, + "learning_rate": 0.0009974945644788695, + "loss": 3.9407, + "step": 295 + }, + { + "epoch": 0.020383204239706482, + "grad_norm": 0.5246041417121887, + "learning_rate": 0.0009974520994700366, + "loss": 3.6985, + "step": 300 + }, + { + "epoch": 0.020722924310368256, + "grad_norm": 0.8263882994651794, + "learning_rate": 0.000997409634461204, + "loss": 3.949, + "step": 305 + }, + { + "epoch": 0.021062644381030032, + "grad_norm": 0.4506300091743469, + "learning_rate": 0.0009973671694523713, + "loss": 3.718, + "step": 310 + }, + { + "epoch": 0.021402364451691806, + "grad_norm": 0.4988076686859131, + "learning_rate": 0.0009973247044435386, + "loss": 4.163, + "step": 315 + }, + { + "epoch": 0.021742084522353582, + "grad_norm": 0.5724675059318542, + "learning_rate": 0.000997282239434706, + "loss": 3.9906, + "step": 320 + }, + { + "epoch": 0.022081804593015356, + "grad_norm": 0.5702639222145081, + "learning_rate": 0.000997239774425873, + "loss": 3.8087, + "step": 325 + }, + { + "epoch": 0.02242152466367713, + "grad_norm": 0.45527079701423645, + "learning_rate": 0.0009971973094170404, + "loss": 3.6432, + "step": 330 + }, + { + "epoch": 0.022761244734338906, + "grad_norm": 0.5885451436042786, + "learning_rate": 0.0009971548444082077, + "loss": 4.0753, + "step": 335 + }, + { + "epoch": 0.02310096480500068, + "grad_norm": 0.5633436441421509, + "learning_rate": 0.0009971123793993748, + "loss": 3.9803, + "step": 340 + }, + { + "epoch": 0.023440684875662456, + "grad_norm": 0.6600042581558228, + "learning_rate": 0.0009970699143905422, + "loss": 3.9767, + "step": 345 + }, + { + "epoch": 0.02378040494632423, + "grad_norm": 0.5405495166778564, + "learning_rate": 0.0009970274493817095, + "loss": 3.75, + "step": 350 + }, + { + "epoch": 0.024120125016986002, + "grad_norm": 0.6013660430908203, + "learning_rate": 0.0009969849843728769, + "loss": 4.1565, + "step": 355 + }, + { + "epoch": 0.02445984508764778, + "grad_norm": 0.6407164335250854, + "learning_rate": 0.000996942519364044, + "loss": 4.0018, + "step": 360 + }, + { + "epoch": 0.024799565158309552, + "grad_norm": 0.5490573644638062, + "learning_rate": 0.0009969000543552113, + "loss": 3.9018, + "step": 365 + }, + { + "epoch": 0.02513928522897133, + "grad_norm": 0.5638839602470398, + "learning_rate": 0.0009968575893463786, + "loss": 4.0383, + "step": 370 + }, + { + "epoch": 0.025479005299633102, + "grad_norm": 1.9583743810653687, + "learning_rate": 0.0009968151243375458, + "loss": 3.8543, + "step": 375 + }, + { + "epoch": 0.025818725370294875, + "grad_norm": 1.9940112829208374, + "learning_rate": 0.000996772659328713, + "loss": 4.0845, + "step": 380 + }, + { + "epoch": 0.026158445440956652, + "grad_norm": 0.5722443461418152, + "learning_rate": 0.0009967301943198804, + "loss": 4.0051, + "step": 385 + }, + { + "epoch": 0.026498165511618425, + "grad_norm": 0.6977593898773193, + "learning_rate": 0.0009966877293110478, + "loss": 3.7642, + "step": 390 + }, + { + "epoch": 0.026837885582280202, + "grad_norm": 0.54580157995224, + "learning_rate": 0.000996645264302215, + "loss": 3.5164, + "step": 395 + }, + { + "epoch": 0.027177605652941975, + "grad_norm": 8.007853507995605, + "learning_rate": 0.0009966027992933822, + "loss": 3.8012, + "step": 400 + }, + { + "epoch": 0.027517325723603752, + "grad_norm": 0.5816832780838013, + "learning_rate": 0.0009965603342845495, + "loss": 3.6734, + "step": 405 + }, + { + "epoch": 0.027857045794265525, + "grad_norm": 0.5928305387496948, + "learning_rate": 0.0009965178692757169, + "loss": 3.9956, + "step": 410 + }, + { + "epoch": 0.0281967658649273, + "grad_norm": 0.8828011155128479, + "learning_rate": 0.000996475404266884, + "loss": 4.1078, + "step": 415 + }, + { + "epoch": 0.028536485935589075, + "grad_norm": 0.6904022693634033, + "learning_rate": 0.0009964329392580513, + "loss": 3.8418, + "step": 420 + }, + { + "epoch": 0.02887620600625085, + "grad_norm": 0.5612075328826904, + "learning_rate": 0.0009963904742492187, + "loss": 4.1817, + "step": 425 + }, + { + "epoch": 0.029215926076912625, + "grad_norm": 0.5470679402351379, + "learning_rate": 0.000996348009240386, + "loss": 3.9428, + "step": 430 + }, + { + "epoch": 0.0295556461475744, + "grad_norm": 0.8141860961914062, + "learning_rate": 0.0009963055442315531, + "loss": 3.9853, + "step": 435 + }, + { + "epoch": 0.029895366218236172, + "grad_norm": 0.5441884994506836, + "learning_rate": 0.0009962630792227205, + "loss": 3.8899, + "step": 440 + }, + { + "epoch": 0.03023508628889795, + "grad_norm": 0.8587472438812256, + "learning_rate": 0.0009962206142138878, + "loss": 3.7975, + "step": 445 + }, + { + "epoch": 0.030574806359559722, + "grad_norm": 0.5279538035392761, + "learning_rate": 0.000996178149205055, + "loss": 3.5993, + "step": 450 + }, + { + "epoch": 0.0309145264302215, + "grad_norm": 0.504540205001831, + "learning_rate": 0.0009961356841962225, + "loss": 3.8558, + "step": 455 + }, + { + "epoch": 0.03125424650088327, + "grad_norm": 0.5363671183586121, + "learning_rate": 0.0009960932191873896, + "loss": 4.1573, + "step": 460 + }, + { + "epoch": 0.03159396657154505, + "grad_norm": 2.302508592605591, + "learning_rate": 0.000996050754178557, + "loss": 3.8831, + "step": 465 + }, + { + "epoch": 0.03193368664220682, + "grad_norm": 0.7786630392074585, + "learning_rate": 0.0009960082891697242, + "loss": 3.885, + "step": 470 + }, + { + "epoch": 0.032273406712868595, + "grad_norm": 0.5628072023391724, + "learning_rate": 0.0009959658241608914, + "loss": 4.0699, + "step": 475 + }, + { + "epoch": 0.03261312678353037, + "grad_norm": 0.8729355931282043, + "learning_rate": 0.0009959233591520587, + "loss": 3.7228, + "step": 480 + }, + { + "epoch": 0.03295284685419215, + "grad_norm": 0.5414428114891052, + "learning_rate": 0.000995880894143226, + "loss": 3.6357, + "step": 485 + }, + { + "epoch": 0.03329256692485392, + "grad_norm": 0.6227617263793945, + "learning_rate": 0.0009958384291343934, + "loss": 3.8537, + "step": 490 + }, + { + "epoch": 0.033632286995515695, + "grad_norm": 0.6045665144920349, + "learning_rate": 0.0009957959641255605, + "loss": 3.9296, + "step": 495 + }, + { + "epoch": 0.03397200706617747, + "grad_norm": 1.6914665699005127, + "learning_rate": 0.0009957534991167278, + "loss": 3.8143, + "step": 500 + }, + { + "epoch": 0.03431172713683924, + "grad_norm": 0.559988260269165, + "learning_rate": 0.0009957110341078951, + "loss": 3.829, + "step": 505 + }, + { + "epoch": 0.03465144720750102, + "grad_norm": 0.6654518842697144, + "learning_rate": 0.0009956685690990623, + "loss": 3.851, + "step": 510 + }, + { + "epoch": 0.034991167278162795, + "grad_norm": 0.5775548219680786, + "learning_rate": 0.0009956261040902296, + "loss": 3.7705, + "step": 515 + }, + { + "epoch": 0.03533088734882457, + "grad_norm": 0.5970734357833862, + "learning_rate": 0.000995583639081397, + "loss": 3.9808, + "step": 520 + }, + { + "epoch": 0.03567060741948634, + "grad_norm": 0.6709187030792236, + "learning_rate": 0.0009955411740725643, + "loss": 3.7881, + "step": 525 + }, + { + "epoch": 0.03601032749014812, + "grad_norm": 1.4802911281585693, + "learning_rate": 0.0009954987090637316, + "loss": 3.6446, + "step": 530 + }, + { + "epoch": 0.036350047560809895, + "grad_norm": 0.6298632621765137, + "learning_rate": 0.0009954562440548987, + "loss": 3.84, + "step": 535 + }, + { + "epoch": 0.036689767631471665, + "grad_norm": 1.1104154586791992, + "learning_rate": 0.000995413779046066, + "loss": 3.8281, + "step": 540 + }, + { + "epoch": 0.03702948770213344, + "grad_norm": 3.1669414043426514, + "learning_rate": 0.0009953713140372334, + "loss": 3.8184, + "step": 545 + }, + { + "epoch": 0.03736920777279522, + "grad_norm": 0.643075704574585, + "learning_rate": 0.0009953288490284005, + "loss": 4.1517, + "step": 550 + }, + { + "epoch": 0.03770892784345699, + "grad_norm": 0.5925756096839905, + "learning_rate": 0.000995286384019568, + "loss": 3.8738, + "step": 555 + }, + { + "epoch": 0.038048647914118765, + "grad_norm": 0.5645868182182312, + "learning_rate": 0.0009952439190107352, + "loss": 3.9242, + "step": 560 + }, + { + "epoch": 0.03838836798478054, + "grad_norm": 2.423346996307373, + "learning_rate": 0.0009952014540019025, + "loss": 3.9295, + "step": 565 + }, + { + "epoch": 0.03872808805544232, + "grad_norm": 0.5646160840988159, + "learning_rate": 0.0009951589889930698, + "loss": 4.0769, + "step": 570 + }, + { + "epoch": 0.03906780812610409, + "grad_norm": 0.6235633492469788, + "learning_rate": 0.000995116523984237, + "loss": 4.0743, + "step": 575 + }, + { + "epoch": 0.039407528196765865, + "grad_norm": 0.9926804900169373, + "learning_rate": 0.0009950740589754043, + "loss": 3.7669, + "step": 580 + }, + { + "epoch": 0.03974724826742764, + "grad_norm": 0.5877676606178284, + "learning_rate": 0.0009950315939665716, + "loss": 3.7585, + "step": 585 + }, + { + "epoch": 0.04008696833808941, + "grad_norm": 0.6339868307113647, + "learning_rate": 0.000994989128957739, + "loss": 3.6846, + "step": 590 + }, + { + "epoch": 0.04042668840875119, + "grad_norm": 0.7069829106330872, + "learning_rate": 0.000994946663948906, + "loss": 3.725, + "step": 595 + }, + { + "epoch": 0.040766408479412965, + "grad_norm": 0.575941264629364, + "learning_rate": 0.0009949041989400734, + "loss": 3.9028, + "step": 600 + }, + { + "epoch": 0.04110612855007474, + "grad_norm": 0.7453792095184326, + "learning_rate": 0.0009948617339312407, + "loss": 3.7812, + "step": 605 + }, + { + "epoch": 0.04144584862073651, + "grad_norm": 0.5961312055587769, + "learning_rate": 0.0009948192689224079, + "loss": 3.9243, + "step": 610 + }, + { + "epoch": 0.04178556869139829, + "grad_norm": 0.5527133941650391, + "learning_rate": 0.0009947768039135752, + "loss": 3.8414, + "step": 615 + }, + { + "epoch": 0.042125288762060065, + "grad_norm": 0.4514964818954468, + "learning_rate": 0.0009947343389047425, + "loss": 3.6732, + "step": 620 + }, + { + "epoch": 0.042465008832721834, + "grad_norm": 0.42869192361831665, + "learning_rate": 0.0009946918738959099, + "loss": 3.4291, + "step": 625 + }, + { + "epoch": 0.04280472890338361, + "grad_norm": 0.5432568788528442, + "learning_rate": 0.0009946494088870772, + "loss": 3.6035, + "step": 630 + }, + { + "epoch": 0.04314444897404539, + "grad_norm": 0.5461887121200562, + "learning_rate": 0.0009946069438782443, + "loss": 3.8103, + "step": 635 + }, + { + "epoch": 0.043484169044707165, + "grad_norm": 0.6913933157920837, + "learning_rate": 0.0009945644788694117, + "loss": 3.6833, + "step": 640 + }, + { + "epoch": 0.043823889115368934, + "grad_norm": 0.5428763628005981, + "learning_rate": 0.000994522013860579, + "loss": 4.0695, + "step": 645 + }, + { + "epoch": 0.04416360918603071, + "grad_norm": 0.4568764865398407, + "learning_rate": 0.000994479548851746, + "loss": 3.9515, + "step": 650 + }, + { + "epoch": 0.04450332925669249, + "grad_norm": 0.4596848785877228, + "learning_rate": 0.0009944370838429134, + "loss": 3.8036, + "step": 655 + }, + { + "epoch": 0.04484304932735426, + "grad_norm": 0.5661909580230713, + "learning_rate": 0.0009943946188340808, + "loss": 3.7296, + "step": 660 + }, + { + "epoch": 0.045182769398016034, + "grad_norm": 0.5530750751495361, + "learning_rate": 0.0009943521538252481, + "loss": 3.7028, + "step": 665 + }, + { + "epoch": 0.04552248946867781, + "grad_norm": 0.5476675629615784, + "learning_rate": 0.0009943096888164152, + "loss": 3.9626, + "step": 670 + }, + { + "epoch": 0.04586220953933958, + "grad_norm": 1.141539216041565, + "learning_rate": 0.0009942672238075826, + "loss": 3.7415, + "step": 675 + }, + { + "epoch": 0.04620192961000136, + "grad_norm": 0.5888988971710205, + "learning_rate": 0.00099422475879875, + "loss": 3.8676, + "step": 680 + }, + { + "epoch": 0.046541649680663134, + "grad_norm": 0.5741140246391296, + "learning_rate": 0.000994182293789917, + "loss": 3.9568, + "step": 685 + }, + { + "epoch": 0.04688136975132491, + "grad_norm": 1.4437280893325806, + "learning_rate": 0.0009941398287810843, + "loss": 3.9491, + "step": 690 + }, + { + "epoch": 0.04722108982198668, + "grad_norm": 0.5097208619117737, + "learning_rate": 0.0009940973637722517, + "loss": 3.8049, + "step": 695 + }, + { + "epoch": 0.04756080989264846, + "grad_norm": 0.43953296542167664, + "learning_rate": 0.000994054898763419, + "loss": 3.969, + "step": 700 + }, + { + "epoch": 0.047900529963310234, + "grad_norm": 1.2057006359100342, + "learning_rate": 0.0009940124337545863, + "loss": 3.9099, + "step": 705 + }, + { + "epoch": 0.048240250033972004, + "grad_norm": 0.5880506038665771, + "learning_rate": 0.0009939699687457535, + "loss": 3.8779, + "step": 710 + }, + { + "epoch": 0.04857997010463378, + "grad_norm": 0.47119975090026855, + "learning_rate": 0.0009939275037369208, + "loss": 3.7105, + "step": 715 + }, + { + "epoch": 0.04891969017529556, + "grad_norm": 0.549010694026947, + "learning_rate": 0.0009938850387280881, + "loss": 3.9291, + "step": 720 + }, + { + "epoch": 0.049259410245957334, + "grad_norm": 0.5610355138778687, + "learning_rate": 0.0009938425737192553, + "loss": 3.7261, + "step": 725 + }, + { + "epoch": 0.049599130316619104, + "grad_norm": 1.5965944528579712, + "learning_rate": 0.0009938001087104226, + "loss": 3.824, + "step": 730 + }, + { + "epoch": 0.04993885038728088, + "grad_norm": 0.6607491970062256, + "learning_rate": 0.00099375764370159, + "loss": 3.7982, + "step": 735 + }, + { + "epoch": 0.05027857045794266, + "grad_norm": 1.4106281995773315, + "learning_rate": 0.0009937151786927573, + "loss": 3.9767, + "step": 740 + }, + { + "epoch": 0.05061829052860443, + "grad_norm": 0.5237374305725098, + "learning_rate": 0.0009936727136839244, + "loss": 3.6383, + "step": 745 + }, + { + "epoch": 0.050958010599266204, + "grad_norm": 0.6122420430183411, + "learning_rate": 0.0009936302486750917, + "loss": 3.8833, + "step": 750 + }, + { + "epoch": 0.05129773066992798, + "grad_norm": 0.5208083987236023, + "learning_rate": 0.000993587783666259, + "loss": 3.8577, + "step": 755 + }, + { + "epoch": 0.05163745074058975, + "grad_norm": 0.6950027346611023, + "learning_rate": 0.0009935453186574262, + "loss": 4.0022, + "step": 760 + }, + { + "epoch": 0.05197717081125153, + "grad_norm": 0.5676352381706238, + "learning_rate": 0.0009935028536485937, + "loss": 3.7205, + "step": 765 + }, + { + "epoch": 0.052316890881913304, + "grad_norm": 0.5784592032432556, + "learning_rate": 0.0009934603886397608, + "loss": 3.762, + "step": 770 + }, + { + "epoch": 0.05265661095257508, + "grad_norm": 1.3643052577972412, + "learning_rate": 0.0009934179236309282, + "loss": 3.9225, + "step": 775 + }, + { + "epoch": 0.05299633102323685, + "grad_norm": 0.552307665348053, + "learning_rate": 0.0009933754586220955, + "loss": 3.9288, + "step": 780 + }, + { + "epoch": 0.05333605109389863, + "grad_norm": 0.6428114175796509, + "learning_rate": 0.0009933329936132626, + "loss": 3.6841, + "step": 785 + }, + { + "epoch": 0.053675771164560404, + "grad_norm": 0.6585874557495117, + "learning_rate": 0.00099329052860443, + "loss": 3.9182, + "step": 790 + }, + { + "epoch": 0.054015491235222174, + "grad_norm": 0.6242063045501709, + "learning_rate": 0.0009932480635955973, + "loss": 3.7315, + "step": 795 + }, + { + "epoch": 0.05435521130588395, + "grad_norm": 1.172575831413269, + "learning_rate": 0.0009932055985867646, + "loss": 4.0225, + "step": 800 + }, + { + "epoch": 0.05469493137654573, + "grad_norm": 0.5162668228149414, + "learning_rate": 0.0009931631335779317, + "loss": 4.0547, + "step": 805 + }, + { + "epoch": 0.055034651447207504, + "grad_norm": 0.5157238841056824, + "learning_rate": 0.000993120668569099, + "loss": 3.9932, + "step": 810 + }, + { + "epoch": 0.055374371517869274, + "grad_norm": 0.7152529954910278, + "learning_rate": 0.0009930782035602664, + "loss": 3.8642, + "step": 815 + }, + { + "epoch": 0.05571409158853105, + "grad_norm": 0.6313608288764954, + "learning_rate": 0.0009930357385514335, + "loss": 3.7289, + "step": 820 + }, + { + "epoch": 0.05605381165919283, + "grad_norm": 0.5713692903518677, + "learning_rate": 0.0009929932735426009, + "loss": 3.6429, + "step": 825 + }, + { + "epoch": 0.0563935317298546, + "grad_norm": 0.6273418068885803, + "learning_rate": 0.0009929508085337682, + "loss": 3.695, + "step": 830 + }, + { + "epoch": 0.056733251800516374, + "grad_norm": 1.2013376951217651, + "learning_rate": 0.0009929083435249355, + "loss": 3.896, + "step": 835 + }, + { + "epoch": 0.05707297187117815, + "grad_norm": 0.6832719445228577, + "learning_rate": 0.0009928658785161029, + "loss": 4.0995, + "step": 840 + }, + { + "epoch": 0.05741269194183993, + "grad_norm": 0.6568312048912048, + "learning_rate": 0.00099282341350727, + "loss": 3.9602, + "step": 845 + }, + { + "epoch": 0.0577524120125017, + "grad_norm": 0.7554712891578674, + "learning_rate": 0.0009927809484984373, + "loss": 3.8707, + "step": 850 + }, + { + "epoch": 0.058092132083163474, + "grad_norm": 0.5548936128616333, + "learning_rate": 0.0009927384834896046, + "loss": 3.708, + "step": 855 + }, + { + "epoch": 0.05843185215382525, + "grad_norm": 0.5560023188591003, + "learning_rate": 0.0009926960184807718, + "loss": 3.5404, + "step": 860 + }, + { + "epoch": 0.05877157222448702, + "grad_norm": 0.7071141004562378, + "learning_rate": 0.000992653553471939, + "loss": 3.8469, + "step": 865 + }, + { + "epoch": 0.0591112922951488, + "grad_norm": 0.5891933441162109, + "learning_rate": 0.0009926110884631064, + "loss": 3.9003, + "step": 870 + }, + { + "epoch": 0.059451012365810574, + "grad_norm": 0.6138566732406616, + "learning_rate": 0.0009925686234542738, + "loss": 3.6245, + "step": 875 + }, + { + "epoch": 0.059790732436472344, + "grad_norm": 0.6222836971282959, + "learning_rate": 0.000992526158445441, + "loss": 3.733, + "step": 880 + }, + { + "epoch": 0.06013045250713412, + "grad_norm": 0.6209458708763123, + "learning_rate": 0.0009924836934366082, + "loss": 3.8125, + "step": 885 + }, + { + "epoch": 0.0604701725777959, + "grad_norm": 0.8434022068977356, + "learning_rate": 0.0009924412284277755, + "loss": 3.9784, + "step": 890 + }, + { + "epoch": 0.060809892648457674, + "grad_norm": 0.5856556296348572, + "learning_rate": 0.0009923987634189429, + "loss": 3.8822, + "step": 895 + }, + { + "epoch": 0.061149612719119444, + "grad_norm": 0.6009406447410583, + "learning_rate": 0.00099235629841011, + "loss": 3.8985, + "step": 900 + }, + { + "epoch": 0.06148933278978122, + "grad_norm": 0.6219965815544128, + "learning_rate": 0.0009923138334012773, + "loss": 3.7295, + "step": 905 + }, + { + "epoch": 0.061829052860443, + "grad_norm": 0.7353580594062805, + "learning_rate": 0.0009922713683924447, + "loss": 3.6831, + "step": 910 + }, + { + "epoch": 0.06216877293110477, + "grad_norm": 0.7379398345947266, + "learning_rate": 0.000992228903383612, + "loss": 4.0579, + "step": 915 + }, + { + "epoch": 0.06250849300176654, + "grad_norm": 0.6934113502502441, + "learning_rate": 0.0009921864383747791, + "loss": 3.8067, + "step": 920 + }, + { + "epoch": 0.06284821307242831, + "grad_norm": 0.6656597852706909, + "learning_rate": 0.0009921439733659465, + "loss": 3.908, + "step": 925 + }, + { + "epoch": 0.0631879331430901, + "grad_norm": 2.1865758895874023, + "learning_rate": 0.0009921015083571138, + "loss": 3.9016, + "step": 930 + }, + { + "epoch": 0.06352765321375187, + "grad_norm": 0.6319058537483215, + "learning_rate": 0.000992059043348281, + "loss": 4.0054, + "step": 935 + }, + { + "epoch": 0.06386737328441364, + "grad_norm": 0.5362616777420044, + "learning_rate": 0.0009920165783394485, + "loss": 3.6477, + "step": 940 + }, + { + "epoch": 0.06420709335507542, + "grad_norm": 0.5525217652320862, + "learning_rate": 0.0009919741133306156, + "loss": 3.7794, + "step": 945 + }, + { + "epoch": 0.06454681342573719, + "grad_norm": 0.7375221252441406, + "learning_rate": 0.000991931648321783, + "loss": 3.8622, + "step": 950 + }, + { + "epoch": 0.06488653349639897, + "grad_norm": 0.6783074140548706, + "learning_rate": 0.0009918891833129502, + "loss": 3.7648, + "step": 955 + }, + { + "epoch": 0.06522625356706074, + "grad_norm": 0.7984088063240051, + "learning_rate": 0.0009918467183041174, + "loss": 3.894, + "step": 960 + }, + { + "epoch": 0.06556597363772251, + "grad_norm": 0.48116791248321533, + "learning_rate": 0.0009918042532952847, + "loss": 3.7698, + "step": 965 + }, + { + "epoch": 0.0659056937083843, + "grad_norm": 0.9424999356269836, + "learning_rate": 0.000991761788286452, + "loss": 4.0673, + "step": 970 + }, + { + "epoch": 0.06624541377904607, + "grad_norm": 0.5739089250564575, + "learning_rate": 0.0009917193232776194, + "loss": 4.0236, + "step": 975 + }, + { + "epoch": 0.06658513384970784, + "grad_norm": 0.5319967269897461, + "learning_rate": 0.0009916768582687865, + "loss": 3.8909, + "step": 980 + }, + { + "epoch": 0.06692485392036962, + "grad_norm": 0.6642907857894897, + "learning_rate": 0.0009916343932599538, + "loss": 3.8548, + "step": 985 + }, + { + "epoch": 0.06726457399103139, + "grad_norm": 0.5319867134094238, + "learning_rate": 0.0009915919282511212, + "loss": 3.8652, + "step": 990 + }, + { + "epoch": 0.06760429406169316, + "grad_norm": 0.5827558636665344, + "learning_rate": 0.0009915494632422883, + "loss": 3.7097, + "step": 995 + }, + { + "epoch": 0.06794401413235494, + "grad_norm": 0.6750186681747437, + "learning_rate": 0.0009915069982334556, + "loss": 3.8518, + "step": 1000 + }, + { + "epoch": 0.06828373420301671, + "grad_norm": 1.1676689386367798, + "learning_rate": 0.000991464533224623, + "loss": 3.7971, + "step": 1005 + }, + { + "epoch": 0.06862345427367848, + "grad_norm": 0.5687863826751709, + "learning_rate": 0.0009914220682157903, + "loss": 3.6428, + "step": 1010 + }, + { + "epoch": 0.06896317434434027, + "grad_norm": 0.6025897860527039, + "learning_rate": 0.0009913796032069576, + "loss": 3.9043, + "step": 1015 + }, + { + "epoch": 0.06930289441500204, + "grad_norm": 0.5458235740661621, + "learning_rate": 0.0009913371381981247, + "loss": 3.7681, + "step": 1020 + }, + { + "epoch": 0.0696426144856638, + "grad_norm": 0.505095899105072, + "learning_rate": 0.000991294673189292, + "loss": 3.5802, + "step": 1025 + }, + { + "epoch": 0.06998233455632559, + "grad_norm": 0.7556877136230469, + "learning_rate": 0.0009912522081804594, + "loss": 4.0027, + "step": 1030 + }, + { + "epoch": 0.07032205462698736, + "grad_norm": 0.7739489674568176, + "learning_rate": 0.0009912097431716265, + "loss": 3.7581, + "step": 1035 + }, + { + "epoch": 0.07066177469764914, + "grad_norm": 0.7745015621185303, + "learning_rate": 0.0009911672781627938, + "loss": 3.7831, + "step": 1040 + }, + { + "epoch": 0.07100149476831091, + "grad_norm": 1.3215758800506592, + "learning_rate": 0.0009911248131539612, + "loss": 3.7766, + "step": 1045 + }, + { + "epoch": 0.07134121483897268, + "grad_norm": 0.6260070204734802, + "learning_rate": 0.0009910823481451285, + "loss": 4.033, + "step": 1050 + }, + { + "epoch": 0.07168093490963447, + "grad_norm": 0.9822431802749634, + "learning_rate": 0.0009910398831362956, + "loss": 3.6278, + "step": 1055 + }, + { + "epoch": 0.07202065498029624, + "grad_norm": 0.6256703734397888, + "learning_rate": 0.000990997418127463, + "loss": 3.6758, + "step": 1060 + }, + { + "epoch": 0.072360375050958, + "grad_norm": 0.5993952751159668, + "learning_rate": 0.0009909549531186303, + "loss": 3.5871, + "step": 1065 + }, + { + "epoch": 0.07270009512161979, + "grad_norm": 0.6881741881370544, + "learning_rate": 0.0009909124881097974, + "loss": 3.9959, + "step": 1070 + }, + { + "epoch": 0.07303981519228156, + "grad_norm": 2.6121256351470947, + "learning_rate": 0.0009908700231009647, + "loss": 3.6342, + "step": 1075 + }, + { + "epoch": 0.07337953526294333, + "grad_norm": 0.6979096531867981, + "learning_rate": 0.000990827558092132, + "loss": 4.0179, + "step": 1080 + }, + { + "epoch": 0.07371925533360511, + "grad_norm": 0.5363898873329163, + "learning_rate": 0.0009907850930832994, + "loss": 3.7014, + "step": 1085 + }, + { + "epoch": 0.07405897540426688, + "grad_norm": 0.6762584447860718, + "learning_rate": 0.0009907426280744668, + "loss": 3.8049, + "step": 1090 + }, + { + "epoch": 0.07439869547492865, + "grad_norm": 0.7676011919975281, + "learning_rate": 0.0009907001630656339, + "loss": 3.8933, + "step": 1095 + }, + { + "epoch": 0.07473841554559044, + "grad_norm": 0.6409669518470764, + "learning_rate": 0.0009906576980568012, + "loss": 3.378, + "step": 1100 + }, + { + "epoch": 0.0750781356162522, + "grad_norm": 0.8464885354042053, + "learning_rate": 0.0009906152330479685, + "loss": 4.0465, + "step": 1105 + }, + { + "epoch": 0.07541785568691398, + "grad_norm": 0.9416410326957703, + "learning_rate": 0.0009905727680391357, + "loss": 3.7612, + "step": 1110 + }, + { + "epoch": 0.07575757575757576, + "grad_norm": 1.8878381252288818, + "learning_rate": 0.000990530303030303, + "loss": 3.8833, + "step": 1115 + }, + { + "epoch": 0.07609729582823753, + "grad_norm": 0.6034110188484192, + "learning_rate": 0.0009904878380214703, + "loss": 3.7747, + "step": 1120 + }, + { + "epoch": 0.07643701589889931, + "grad_norm": 0.848849356174469, + "learning_rate": 0.0009904453730126377, + "loss": 3.9122, + "step": 1125 + }, + { + "epoch": 0.07677673596956108, + "grad_norm": 0.7755124568939209, + "learning_rate": 0.0009904029080038048, + "loss": 3.776, + "step": 1130 + }, + { + "epoch": 0.07711645604022285, + "grad_norm": 1.3329683542251587, + "learning_rate": 0.0009903604429949721, + "loss": 3.9555, + "step": 1135 + }, + { + "epoch": 0.07745617611088464, + "grad_norm": 0.7614222168922424, + "learning_rate": 0.0009903179779861394, + "loss": 3.5661, + "step": 1140 + }, + { + "epoch": 0.0777958961815464, + "grad_norm": 0.8013120889663696, + "learning_rate": 0.0009902755129773066, + "loss": 3.9451, + "step": 1145 + }, + { + "epoch": 0.07813561625220818, + "grad_norm": 0.7201833128929138, + "learning_rate": 0.0009902330479684741, + "loss": 3.7353, + "step": 1150 + }, + { + "epoch": 0.07847533632286996, + "grad_norm": 0.6137555837631226, + "learning_rate": 0.0009901905829596412, + "loss": 3.8554, + "step": 1155 + }, + { + "epoch": 0.07881505639353173, + "grad_norm": 0.5318056344985962, + "learning_rate": 0.0009901481179508086, + "loss": 3.7185, + "step": 1160 + }, + { + "epoch": 0.0791547764641935, + "grad_norm": 0.6779008507728577, + "learning_rate": 0.000990105652941976, + "loss": 3.9978, + "step": 1165 + }, + { + "epoch": 0.07949449653485528, + "grad_norm": 0.5661518573760986, + "learning_rate": 0.000990063187933143, + "loss": 3.8424, + "step": 1170 + }, + { + "epoch": 0.07983421660551705, + "grad_norm": 0.6510219573974609, + "learning_rate": 0.0009900207229243104, + "loss": 3.7111, + "step": 1175 + }, + { + "epoch": 0.08017393667617882, + "grad_norm": 0.6256716847419739, + "learning_rate": 0.0009899782579154777, + "loss": 4.2326, + "step": 1180 + }, + { + "epoch": 0.0805136567468406, + "grad_norm": 0.5679534077644348, + "learning_rate": 0.000989935792906645, + "loss": 3.897, + "step": 1185 + }, + { + "epoch": 0.08085337681750238, + "grad_norm": 0.5974684953689575, + "learning_rate": 0.0009898933278978121, + "loss": 4.1923, + "step": 1190 + }, + { + "epoch": 0.08119309688816416, + "grad_norm": 0.8721858263015747, + "learning_rate": 0.0009898508628889795, + "loss": 4.3006, + "step": 1195 + }, + { + "epoch": 0.08153281695882593, + "grad_norm": 112.88106536865234, + "learning_rate": 0.0009898083978801468, + "loss": 5.4792, + "step": 1200 + }, + { + "epoch": 0.0818725370294877, + "grad_norm": 2.139004707336426, + "learning_rate": 0.0009897659328713141, + "loss": 5.7927, + "step": 1205 + }, + { + "epoch": 0.08221225710014948, + "grad_norm": 2.9734630584716797, + "learning_rate": 0.0009897234678624813, + "loss": 4.1369, + "step": 1210 + }, + { + "epoch": 0.08255197717081125, + "grad_norm": 0.7371381521224976, + "learning_rate": 0.0009896810028536486, + "loss": 3.8499, + "step": 1215 + }, + { + "epoch": 0.08289169724147302, + "grad_norm": 0.7136930823326111, + "learning_rate": 0.000989638537844816, + "loss": 3.7277, + "step": 1220 + }, + { + "epoch": 0.0832314173121348, + "grad_norm": 0.6129528880119324, + "learning_rate": 0.0009895960728359833, + "loss": 3.7181, + "step": 1225 + }, + { + "epoch": 0.08357113738279658, + "grad_norm": 0.7247201204299927, + "learning_rate": 0.0009895536078271504, + "loss": 3.9597, + "step": 1230 + }, + { + "epoch": 0.08391085745345835, + "grad_norm": 0.6700114011764526, + "learning_rate": 0.0009895111428183177, + "loss": 3.7295, + "step": 1235 + }, + { + "epoch": 0.08425057752412013, + "grad_norm": 0.8911071419715881, + "learning_rate": 0.000989468677809485, + "loss": 3.8105, + "step": 1240 + }, + { + "epoch": 0.0845902975947819, + "grad_norm": 0.6714710593223572, + "learning_rate": 0.0009894262128006522, + "loss": 3.5437, + "step": 1245 + }, + { + "epoch": 0.08493001766544367, + "grad_norm": 0.670673668384552, + "learning_rate": 0.0009893837477918197, + "loss": 3.8434, + "step": 1250 + }, + { + "epoch": 0.08526973773610545, + "grad_norm": 0.6780371069908142, + "learning_rate": 0.0009893412827829868, + "loss": 3.984, + "step": 1255 + }, + { + "epoch": 0.08560945780676722, + "grad_norm": 0.6912269592285156, + "learning_rate": 0.0009892988177741542, + "loss": 3.6702, + "step": 1260 + }, + { + "epoch": 0.08594917787742899, + "grad_norm": 0.7322074770927429, + "learning_rate": 0.0009892563527653215, + "loss": 3.7321, + "step": 1265 + }, + { + "epoch": 0.08628889794809078, + "grad_norm": 0.7525180578231812, + "learning_rate": 0.0009892138877564886, + "loss": 3.725, + "step": 1270 + }, + { + "epoch": 0.08662861801875255, + "grad_norm": 0.6995285153388977, + "learning_rate": 0.000989171422747656, + "loss": 3.7849, + "step": 1275 + }, + { + "epoch": 0.08696833808941433, + "grad_norm": 0.514626681804657, + "learning_rate": 0.0009891289577388233, + "loss": 3.9906, + "step": 1280 + }, + { + "epoch": 0.0873080581600761, + "grad_norm": 0.5982587933540344, + "learning_rate": 0.0009890864927299906, + "loss": 4.3071, + "step": 1285 + }, + { + "epoch": 0.08764777823073787, + "grad_norm": 1.2551847696304321, + "learning_rate": 0.0009890440277211577, + "loss": 3.8248, + "step": 1290 + }, + { + "epoch": 0.08798749830139965, + "grad_norm": 0.5657306909561157, + "learning_rate": 0.000989001562712325, + "loss": 3.765, + "step": 1295 + }, + { + "epoch": 0.08832721837206142, + "grad_norm": 0.5734697580337524, + "learning_rate": 0.0009889590977034924, + "loss": 4.0299, + "step": 1300 + }, + { + "epoch": 0.08866693844272319, + "grad_norm": 0.5558915734291077, + "learning_rate": 0.0009889166326946595, + "loss": 3.8244, + "step": 1305 + }, + { + "epoch": 0.08900665851338498, + "grad_norm": 0.556370735168457, + "learning_rate": 0.0009888741676858269, + "loss": 3.8871, + "step": 1310 + }, + { + "epoch": 0.08934637858404675, + "grad_norm": 0.6792877316474915, + "learning_rate": 0.0009888317026769942, + "loss": 3.7253, + "step": 1315 + }, + { + "epoch": 0.08968609865470852, + "grad_norm": 0.6492648720741272, + "learning_rate": 0.0009887892376681615, + "loss": 3.4992, + "step": 1320 + }, + { + "epoch": 0.0900258187253703, + "grad_norm": 0.652964174747467, + "learning_rate": 0.0009887467726593289, + "loss": 3.5441, + "step": 1325 + }, + { + "epoch": 0.09036553879603207, + "grad_norm": 0.6389694213867188, + "learning_rate": 0.000988704307650496, + "loss": 3.815, + "step": 1330 + }, + { + "epoch": 0.09070525886669384, + "grad_norm": 0.7050741910934448, + "learning_rate": 0.0009886618426416633, + "loss": 3.8492, + "step": 1335 + }, + { + "epoch": 0.09104497893735562, + "grad_norm": 0.5823991894721985, + "learning_rate": 0.0009886193776328306, + "loss": 3.9839, + "step": 1340 + }, + { + "epoch": 0.09138469900801739, + "grad_norm": 0.7883529663085938, + "learning_rate": 0.0009885769126239978, + "loss": 3.5452, + "step": 1345 + }, + { + "epoch": 0.09172441907867916, + "grad_norm": 1.3914811611175537, + "learning_rate": 0.000988534447615165, + "loss": 3.6251, + "step": 1350 + }, + { + "epoch": 0.09206413914934095, + "grad_norm": 0.6352505087852478, + "learning_rate": 0.0009884919826063324, + "loss": 3.8847, + "step": 1355 + }, + { + "epoch": 0.09240385922000272, + "grad_norm": 1.3413283824920654, + "learning_rate": 0.0009884495175974998, + "loss": 3.7842, + "step": 1360 + }, + { + "epoch": 0.0927435792906645, + "grad_norm": 0.5921688079833984, + "learning_rate": 0.0009884070525886669, + "loss": 3.7323, + "step": 1365 + }, + { + "epoch": 0.09308329936132627, + "grad_norm": 0.5520976185798645, + "learning_rate": 0.0009883645875798342, + "loss": 3.828, + "step": 1370 + }, + { + "epoch": 0.09342301943198804, + "grad_norm": 0.5818116068840027, + "learning_rate": 0.0009883221225710016, + "loss": 3.8957, + "step": 1375 + }, + { + "epoch": 0.09376273950264982, + "grad_norm": 0.6413286328315735, + "learning_rate": 0.0009882796575621687, + "loss": 3.6669, + "step": 1380 + }, + { + "epoch": 0.09410245957331159, + "grad_norm": 0.7118954658508301, + "learning_rate": 0.000988237192553336, + "loss": 3.7792, + "step": 1385 + }, + { + "epoch": 0.09444217964397336, + "grad_norm": 0.6792705059051514, + "learning_rate": 0.0009881947275445033, + "loss": 3.7641, + "step": 1390 + }, + { + "epoch": 0.09478189971463515, + "grad_norm": 0.7626094222068787, + "learning_rate": 0.0009881522625356707, + "loss": 3.7633, + "step": 1395 + }, + { + "epoch": 0.09512161978529692, + "grad_norm": 0.5472246408462524, + "learning_rate": 0.000988109797526838, + "loss": 3.8334, + "step": 1400 + }, + { + "epoch": 0.09546133985595869, + "grad_norm": 0.5271095037460327, + "learning_rate": 0.0009880673325180051, + "loss": 3.9326, + "step": 1405 + }, + { + "epoch": 0.09580105992662047, + "grad_norm": 0.8590952157974243, + "learning_rate": 0.0009880248675091725, + "loss": 3.8794, + "step": 1410 + }, + { + "epoch": 0.09614077999728224, + "grad_norm": 0.5687350630760193, + "learning_rate": 0.0009879824025003398, + "loss": 3.8372, + "step": 1415 + }, + { + "epoch": 0.09648050006794401, + "grad_norm": 0.9411936402320862, + "learning_rate": 0.000987939937491507, + "loss": 3.9607, + "step": 1420 + }, + { + "epoch": 0.09682022013860579, + "grad_norm": 0.7167156934738159, + "learning_rate": 0.0009878974724826742, + "loss": 3.8414, + "step": 1425 + }, + { + "epoch": 0.09715994020926756, + "grad_norm": 0.5776757001876831, + "learning_rate": 0.0009878550074738416, + "loss": 3.7853, + "step": 1430 + }, + { + "epoch": 0.09749966027992933, + "grad_norm": 1.1498574018478394, + "learning_rate": 0.000987812542465009, + "loss": 3.7495, + "step": 1435 + }, + { + "epoch": 0.09783938035059112, + "grad_norm": 0.8664551377296448, + "learning_rate": 0.000987770077456176, + "loss": 3.8079, + "step": 1440 + }, + { + "epoch": 0.09817910042125289, + "grad_norm": 0.5510807633399963, + "learning_rate": 0.0009877276124473434, + "loss": 3.6465, + "step": 1445 + }, + { + "epoch": 0.09851882049191467, + "grad_norm": 0.5712562203407288, + "learning_rate": 0.0009876851474385107, + "loss": 3.7579, + "step": 1450 + }, + { + "epoch": 0.09885854056257644, + "grad_norm": 0.4920804798603058, + "learning_rate": 0.0009876426824296778, + "loss": 3.7946, + "step": 1455 + }, + { + "epoch": 0.09919826063323821, + "grad_norm": 0.5596398115158081, + "learning_rate": 0.0009876002174208454, + "loss": 3.8949, + "step": 1460 + }, + { + "epoch": 0.09953798070389999, + "grad_norm": 0.9163932204246521, + "learning_rate": 0.0009875577524120125, + "loss": 3.789, + "step": 1465 + }, + { + "epoch": 0.09987770077456176, + "grad_norm": 0.6597535610198975, + "learning_rate": 0.0009875152874031798, + "loss": 3.9565, + "step": 1470 + }, + { + "epoch": 0.10021742084522353, + "grad_norm": 0.6468210220336914, + "learning_rate": 0.0009874728223943472, + "loss": 3.5579, + "step": 1475 + }, + { + "epoch": 0.10055714091588532, + "grad_norm": 0.7370440363883972, + "learning_rate": 0.0009874303573855143, + "loss": 3.8719, + "step": 1480 + }, + { + "epoch": 0.10089686098654709, + "grad_norm": 0.7131441831588745, + "learning_rate": 0.0009873878923766816, + "loss": 3.6775, + "step": 1485 + }, + { + "epoch": 0.10123658105720885, + "grad_norm": 0.6849778294563293, + "learning_rate": 0.000987345427367849, + "loss": 3.7835, + "step": 1490 + }, + { + "epoch": 0.10157630112787064, + "grad_norm": 0.5652458071708679, + "learning_rate": 0.0009873029623590163, + "loss": 3.7406, + "step": 1495 + }, + { + "epoch": 0.10191602119853241, + "grad_norm": 0.6087997555732727, + "learning_rate": 0.0009872604973501834, + "loss": 3.9017, + "step": 1500 + }, + { + "epoch": 0.10225574126919418, + "grad_norm": 0.6391319632530212, + "learning_rate": 0.0009872180323413507, + "loss": 3.8589, + "step": 1505 + }, + { + "epoch": 0.10259546133985596, + "grad_norm": 0.7061350345611572, + "learning_rate": 0.000987175567332518, + "loss": 3.7486, + "step": 1510 + }, + { + "epoch": 0.10293518141051773, + "grad_norm": 0.5502421259880066, + "learning_rate": 0.0009871331023236852, + "loss": 3.8157, + "step": 1515 + }, + { + "epoch": 0.1032749014811795, + "grad_norm": 0.5989504456520081, + "learning_rate": 0.0009870906373148525, + "loss": 3.6464, + "step": 1520 + }, + { + "epoch": 0.10361462155184128, + "grad_norm": 0.8456800580024719, + "learning_rate": 0.0009870481723060198, + "loss": 3.8163, + "step": 1525 + }, + { + "epoch": 0.10395434162250305, + "grad_norm": 0.6394452452659607, + "learning_rate": 0.0009870057072971872, + "loss": 3.8105, + "step": 1530 + }, + { + "epoch": 0.10429406169316484, + "grad_norm": 0.8114145994186401, + "learning_rate": 0.0009869632422883545, + "loss": 3.9587, + "step": 1535 + }, + { + "epoch": 0.10463378176382661, + "grad_norm": 0.6726270914077759, + "learning_rate": 0.0009869207772795216, + "loss": 3.7907, + "step": 1540 + }, + { + "epoch": 0.10497350183448838, + "grad_norm": 0.6676947474479675, + "learning_rate": 0.000986878312270689, + "loss": 3.8472, + "step": 1545 + }, + { + "epoch": 0.10531322190515016, + "grad_norm": 0.8118389248847961, + "learning_rate": 0.0009868358472618563, + "loss": 3.7511, + "step": 1550 + }, + { + "epoch": 0.10565294197581193, + "grad_norm": 0.6756294369697571, + "learning_rate": 0.0009867933822530234, + "loss": 3.6921, + "step": 1555 + }, + { + "epoch": 0.1059926620464737, + "grad_norm": 0.6851464509963989, + "learning_rate": 0.0009867509172441908, + "loss": 3.7618, + "step": 1560 + }, + { + "epoch": 0.10633238211713548, + "grad_norm": 0.7778677344322205, + "learning_rate": 0.000986708452235358, + "loss": 3.896, + "step": 1565 + }, + { + "epoch": 0.10667210218779725, + "grad_norm": 0.6365407705307007, + "learning_rate": 0.0009866659872265254, + "loss": 3.8857, + "step": 1570 + }, + { + "epoch": 0.10701182225845902, + "grad_norm": 0.6037127375602722, + "learning_rate": 0.0009866235222176928, + "loss": 3.967, + "step": 1575 + }, + { + "epoch": 0.10735154232912081, + "grad_norm": 0.5145928263664246, + "learning_rate": 0.0009865810572088599, + "loss": 3.8159, + "step": 1580 + }, + { + "epoch": 0.10769126239978258, + "grad_norm": 0.5643739700317383, + "learning_rate": 0.0009865385922000272, + "loss": 3.7708, + "step": 1585 + }, + { + "epoch": 0.10803098247044435, + "grad_norm": 0.6068370938301086, + "learning_rate": 0.0009864961271911945, + "loss": 3.7037, + "step": 1590 + }, + { + "epoch": 0.10837070254110613, + "grad_norm": 0.500476598739624, + "learning_rate": 0.0009864536621823617, + "loss": 3.8879, + "step": 1595 + }, + { + "epoch": 0.1087104226117679, + "grad_norm": 0.488675981760025, + "learning_rate": 0.000986411197173529, + "loss": 3.7898, + "step": 1600 + }, + { + "epoch": 0.10905014268242967, + "grad_norm": 0.5935916304588318, + "learning_rate": 0.0009863687321646963, + "loss": 3.7842, + "step": 1605 + }, + { + "epoch": 0.10938986275309145, + "grad_norm": 0.7996106147766113, + "learning_rate": 0.0009863262671558637, + "loss": 3.9069, + "step": 1610 + }, + { + "epoch": 0.10972958282375322, + "grad_norm": 0.6345498561859131, + "learning_rate": 0.0009862838021470308, + "loss": 3.7557, + "step": 1615 + }, + { + "epoch": 0.11006930289441501, + "grad_norm": 0.592496395111084, + "learning_rate": 0.0009862413371381981, + "loss": 3.6109, + "step": 1620 + }, + { + "epoch": 0.11040902296507678, + "grad_norm": 0.5481966137886047, + "learning_rate": 0.0009861988721293654, + "loss": 3.7252, + "step": 1625 + }, + { + "epoch": 0.11074874303573855, + "grad_norm": 0.6437157392501831, + "learning_rate": 0.0009861564071205326, + "loss": 3.8171, + "step": 1630 + }, + { + "epoch": 0.11108846310640033, + "grad_norm": 0.6112599968910217, + "learning_rate": 0.0009861139421117001, + "loss": 3.788, + "step": 1635 + }, + { + "epoch": 0.1114281831770621, + "grad_norm": 0.595553457736969, + "learning_rate": 0.0009860714771028672, + "loss": 3.8753, + "step": 1640 + }, + { + "epoch": 0.11176790324772387, + "grad_norm": 0.6088777780532837, + "learning_rate": 0.0009860290120940346, + "loss": 3.9003, + "step": 1645 + }, + { + "epoch": 0.11210762331838565, + "grad_norm": 0.5482221841812134, + "learning_rate": 0.000985986547085202, + "loss": 3.6909, + "step": 1650 + }, + { + "epoch": 0.11244734338904742, + "grad_norm": 0.6865453720092773, + "learning_rate": 0.000985944082076369, + "loss": 3.8776, + "step": 1655 + }, + { + "epoch": 0.1127870634597092, + "grad_norm": 0.7188596129417419, + "learning_rate": 0.0009859016170675364, + "loss": 3.8061, + "step": 1660 + }, + { + "epoch": 0.11312678353037098, + "grad_norm": 0.8115466237068176, + "learning_rate": 0.0009858591520587037, + "loss": 3.8013, + "step": 1665 + }, + { + "epoch": 0.11346650360103275, + "grad_norm": 0.647230863571167, + "learning_rate": 0.000985816687049871, + "loss": 3.7771, + "step": 1670 + }, + { + "epoch": 0.11380622367169452, + "grad_norm": 0.5701841711997986, + "learning_rate": 0.0009857742220410381, + "loss": 3.8349, + "step": 1675 + }, + { + "epoch": 0.1141459437423563, + "grad_norm": 0.6241211891174316, + "learning_rate": 0.0009857317570322055, + "loss": 3.8027, + "step": 1680 + }, + { + "epoch": 0.11448566381301807, + "grad_norm": 0.521457314491272, + "learning_rate": 0.0009856892920233728, + "loss": 3.645, + "step": 1685 + }, + { + "epoch": 0.11482538388367985, + "grad_norm": 0.5666539072990417, + "learning_rate": 0.00098564682701454, + "loss": 3.4762, + "step": 1690 + }, + { + "epoch": 0.11516510395434162, + "grad_norm": 0.5097023844718933, + "learning_rate": 0.0009856043620057073, + "loss": 3.6349, + "step": 1695 + }, + { + "epoch": 0.1155048240250034, + "grad_norm": 0.6535791754722595, + "learning_rate": 0.0009855618969968746, + "loss": 3.8198, + "step": 1700 + }, + { + "epoch": 0.11584454409566518, + "grad_norm": 0.5113033056259155, + "learning_rate": 0.000985519431988042, + "loss": 3.5389, + "step": 1705 + }, + { + "epoch": 0.11618426416632695, + "grad_norm": 0.4774572253227234, + "learning_rate": 0.0009854769669792093, + "loss": 3.983, + "step": 1710 + }, + { + "epoch": 0.11652398423698872, + "grad_norm": 0.6864311695098877, + "learning_rate": 0.0009854345019703764, + "loss": 3.5762, + "step": 1715 + }, + { + "epoch": 0.1168637043076505, + "grad_norm": 0.6275101900100708, + "learning_rate": 0.0009853920369615437, + "loss": 3.8248, + "step": 1720 + }, + { + "epoch": 0.11720342437831227, + "grad_norm": 0.673237144947052, + "learning_rate": 0.000985349571952711, + "loss": 3.593, + "step": 1725 + }, + { + "epoch": 0.11754314444897404, + "grad_norm": 0.6461753249168396, + "learning_rate": 0.0009853071069438782, + "loss": 3.6515, + "step": 1730 + }, + { + "epoch": 0.11788286451963582, + "grad_norm": 0.9219866991043091, + "learning_rate": 0.0009852646419350455, + "loss": 3.6874, + "step": 1735 + }, + { + "epoch": 0.1182225845902976, + "grad_norm": 0.6653991937637329, + "learning_rate": 0.0009852221769262128, + "loss": 3.8494, + "step": 1740 + }, + { + "epoch": 0.11856230466095936, + "grad_norm": 0.6028766632080078, + "learning_rate": 0.0009851797119173802, + "loss": 3.9718, + "step": 1745 + }, + { + "epoch": 0.11890202473162115, + "grad_norm": 0.5921105146408081, + "learning_rate": 0.0009851372469085473, + "loss": 3.8311, + "step": 1750 + }, + { + "epoch": 0.11924174480228292, + "grad_norm": 0.6015687584877014, + "learning_rate": 0.0009850947818997146, + "loss": 3.8485, + "step": 1755 + }, + { + "epoch": 0.11958146487294469, + "grad_norm": 0.5676931738853455, + "learning_rate": 0.000985052316890882, + "loss": 3.9177, + "step": 1760 + }, + { + "epoch": 0.11992118494360647, + "grad_norm": 0.6627026200294495, + "learning_rate": 0.000985009851882049, + "loss": 3.84, + "step": 1765 + }, + { + "epoch": 0.12026090501426824, + "grad_norm": 0.6077857613563538, + "learning_rate": 0.0009849673868732166, + "loss": 3.8357, + "step": 1770 + }, + { + "epoch": 0.12060062508493002, + "grad_norm": 0.7264236807823181, + "learning_rate": 0.0009849249218643837, + "loss": 3.749, + "step": 1775 + }, + { + "epoch": 0.1209403451555918, + "grad_norm": 0.5790697336196899, + "learning_rate": 0.000984882456855551, + "loss": 3.8881, + "step": 1780 + }, + { + "epoch": 0.12128006522625356, + "grad_norm": 0.6634170413017273, + "learning_rate": 0.0009848399918467184, + "loss": 3.9171, + "step": 1785 + }, + { + "epoch": 0.12161978529691535, + "grad_norm": 0.7707322835922241, + "learning_rate": 0.0009847975268378855, + "loss": 3.8622, + "step": 1790 + }, + { + "epoch": 0.12195950536757712, + "grad_norm": 0.6475215554237366, + "learning_rate": 0.0009847550618290529, + "loss": 3.6067, + "step": 1795 + }, + { + "epoch": 0.12229922543823889, + "grad_norm": 0.6176405549049377, + "learning_rate": 0.0009847125968202202, + "loss": 3.6783, + "step": 1800 + }, + { + "epoch": 0.12263894550890067, + "grad_norm": 0.6158005595207214, + "learning_rate": 0.0009846701318113875, + "loss": 3.7764, + "step": 1805 + }, + { + "epoch": 0.12297866557956244, + "grad_norm": 0.5201879739761353, + "learning_rate": 0.0009846276668025546, + "loss": 3.6984, + "step": 1810 + }, + { + "epoch": 0.12331838565022421, + "grad_norm": 0.551987886428833, + "learning_rate": 0.000984585201793722, + "loss": 3.7062, + "step": 1815 + }, + { + "epoch": 0.123658105720886, + "grad_norm": 0.6607746481895447, + "learning_rate": 0.0009845427367848893, + "loss": 3.7366, + "step": 1820 + }, + { + "epoch": 0.12399782579154776, + "grad_norm": 0.963157057762146, + "learning_rate": 0.0009845002717760564, + "loss": 3.6742, + "step": 1825 + }, + { + "epoch": 0.12433754586220953, + "grad_norm": 0.5994268655776978, + "learning_rate": 0.0009844578067672238, + "loss": 3.846, + "step": 1830 + }, + { + "epoch": 0.12467726593287132, + "grad_norm": 0.6295331120491028, + "learning_rate": 0.000984415341758391, + "loss": 3.7987, + "step": 1835 + }, + { + "epoch": 0.1250169860035331, + "grad_norm": 0.6552512645721436, + "learning_rate": 0.0009843728767495584, + "loss": 3.8192, + "step": 1840 + }, + { + "epoch": 0.12535670607419486, + "grad_norm": 0.8361065983772278, + "learning_rate": 0.0009843304117407258, + "loss": 3.7101, + "step": 1845 + }, + { + "epoch": 0.12569642614485663, + "grad_norm": 1.1641591787338257, + "learning_rate": 0.0009842879467318929, + "loss": 3.8828, + "step": 1850 + }, + { + "epoch": 0.12603614621551842, + "grad_norm": 0.6541483402252197, + "learning_rate": 0.0009842454817230602, + "loss": 3.8318, + "step": 1855 + }, + { + "epoch": 0.1263758662861802, + "grad_norm": 0.6286070346832275, + "learning_rate": 0.0009842030167142276, + "loss": 3.7528, + "step": 1860 + }, + { + "epoch": 0.12671558635684196, + "grad_norm": 0.6533440947532654, + "learning_rate": 0.0009841605517053947, + "loss": 3.5678, + "step": 1865 + }, + { + "epoch": 0.12705530642750373, + "grad_norm": 0.6015164256095886, + "learning_rate": 0.000984118086696562, + "loss": 3.8394, + "step": 1870 + }, + { + "epoch": 0.1273950264981655, + "grad_norm": 0.5898406505584717, + "learning_rate": 0.0009840756216877293, + "loss": 3.6599, + "step": 1875 + }, + { + "epoch": 0.12773474656882727, + "grad_norm": 0.7372815012931824, + "learning_rate": 0.0009840331566788967, + "loss": 3.7581, + "step": 1880 + }, + { + "epoch": 0.12807446663948907, + "grad_norm": 0.5894709825515747, + "learning_rate": 0.000983990691670064, + "loss": 3.8598, + "step": 1885 + }, + { + "epoch": 0.12841418671015084, + "grad_norm": 0.7415690422058105, + "learning_rate": 0.0009839482266612311, + "loss": 3.3445, + "step": 1890 + }, + { + "epoch": 0.1287539067808126, + "grad_norm": 0.669512152671814, + "learning_rate": 0.0009839057616523985, + "loss": 3.8245, + "step": 1895 + }, + { + "epoch": 0.12909362685147438, + "grad_norm": 0.6851697564125061, + "learning_rate": 0.0009838632966435658, + "loss": 3.6915, + "step": 1900 + }, + { + "epoch": 0.12943334692213615, + "grad_norm": 0.6504551768302917, + "learning_rate": 0.000983820831634733, + "loss": 3.5653, + "step": 1905 + }, + { + "epoch": 0.12977306699279795, + "grad_norm": 0.6741543412208557, + "learning_rate": 0.0009837783666259003, + "loss": 3.844, + "step": 1910 + }, + { + "epoch": 0.13011278706345972, + "grad_norm": 0.6617727875709534, + "learning_rate": 0.0009837359016170676, + "loss": 3.9249, + "step": 1915 + }, + { + "epoch": 0.1304525071341215, + "grad_norm": 0.6049027442932129, + "learning_rate": 0.000983693436608235, + "loss": 3.4405, + "step": 1920 + }, + { + "epoch": 0.13079222720478326, + "grad_norm": 0.6029608845710754, + "learning_rate": 0.000983650971599402, + "loss": 3.823, + "step": 1925 + }, + { + "epoch": 0.13113194727544503, + "grad_norm": 0.6659353375434875, + "learning_rate": 0.0009836085065905694, + "loss": 3.7454, + "step": 1930 + }, + { + "epoch": 0.1314716673461068, + "grad_norm": 0.8009145855903625, + "learning_rate": 0.0009835660415817367, + "loss": 3.7289, + "step": 1935 + }, + { + "epoch": 0.1318113874167686, + "grad_norm": 0.5986393094062805, + "learning_rate": 0.0009835235765729038, + "loss": 3.3837, + "step": 1940 + }, + { + "epoch": 0.13215110748743036, + "grad_norm": 0.8296811580657959, + "learning_rate": 0.0009834811115640714, + "loss": 3.6203, + "step": 1945 + }, + { + "epoch": 0.13249082755809213, + "grad_norm": 0.5641279816627502, + "learning_rate": 0.0009834386465552385, + "loss": 3.8938, + "step": 1950 + }, + { + "epoch": 0.1328305476287539, + "grad_norm": 0.6389486789703369, + "learning_rate": 0.0009833961815464058, + "loss": 3.7309, + "step": 1955 + }, + { + "epoch": 0.13317026769941567, + "grad_norm": 0.6988972425460815, + "learning_rate": 0.0009833537165375732, + "loss": 3.7553, + "step": 1960 + }, + { + "epoch": 0.13350998777007744, + "grad_norm": 2.3557167053222656, + "learning_rate": 0.0009833112515287403, + "loss": 3.4593, + "step": 1965 + }, + { + "epoch": 0.13384970784073924, + "grad_norm": 0.8683882355690002, + "learning_rate": 0.0009832687865199076, + "loss": 3.6475, + "step": 1970 + }, + { + "epoch": 0.134189427911401, + "grad_norm": 0.6037503480911255, + "learning_rate": 0.000983226321511075, + "loss": 3.9733, + "step": 1975 + }, + { + "epoch": 0.13452914798206278, + "grad_norm": 0.6086533069610596, + "learning_rate": 0.0009831838565022423, + "loss": 3.6101, + "step": 1980 + }, + { + "epoch": 0.13486886805272455, + "grad_norm": 0.5632864236831665, + "learning_rate": 0.0009831413914934094, + "loss": 3.7772, + "step": 1985 + }, + { + "epoch": 0.13520858812338632, + "grad_norm": 0.621701717376709, + "learning_rate": 0.0009830989264845767, + "loss": 3.7126, + "step": 1990 + }, + { + "epoch": 0.13554830819404812, + "grad_norm": 0.6236476302146912, + "learning_rate": 0.000983056461475744, + "loss": 3.8092, + "step": 1995 + }, + { + "epoch": 0.1358880282647099, + "grad_norm": 0.6200979351997375, + "learning_rate": 0.0009830139964669112, + "loss": 3.7179, + "step": 2000 + }, + { + "epoch": 0.13622774833537166, + "grad_norm": 0.778945803642273, + "learning_rate": 0.0009829715314580785, + "loss": 3.6423, + "step": 2005 + }, + { + "epoch": 0.13656746840603343, + "grad_norm": 0.583135724067688, + "learning_rate": 0.0009829290664492459, + "loss": 4.018, + "step": 2010 + }, + { + "epoch": 0.1369071884766952, + "grad_norm": 0.48970645666122437, + "learning_rate": 0.0009828866014404132, + "loss": 3.799, + "step": 2015 + }, + { + "epoch": 0.13724690854735697, + "grad_norm": 0.7138967514038086, + "learning_rate": 0.0009828441364315805, + "loss": 3.9335, + "step": 2020 + }, + { + "epoch": 0.13758662861801876, + "grad_norm": 0.6615206599235535, + "learning_rate": 0.0009828016714227476, + "loss": 3.645, + "step": 2025 + }, + { + "epoch": 0.13792634868868053, + "grad_norm": 0.7684316635131836, + "learning_rate": 0.000982759206413915, + "loss": 3.7033, + "step": 2030 + }, + { + "epoch": 0.1382660687593423, + "grad_norm": 0.6158739924430847, + "learning_rate": 0.0009827167414050823, + "loss": 3.9084, + "step": 2035 + }, + { + "epoch": 0.13860578883000407, + "grad_norm": 0.5423030257225037, + "learning_rate": 0.0009826742763962494, + "loss": 3.7914, + "step": 2040 + }, + { + "epoch": 0.13894550890066584, + "grad_norm": 0.501213550567627, + "learning_rate": 0.0009826318113874168, + "loss": 3.9085, + "step": 2045 + }, + { + "epoch": 0.1392852289713276, + "grad_norm": 0.6407187581062317, + "learning_rate": 0.000982589346378584, + "loss": 3.7081, + "step": 2050 + }, + { + "epoch": 0.1396249490419894, + "grad_norm": 0.6124802231788635, + "learning_rate": 0.0009825468813697514, + "loss": 3.7414, + "step": 2055 + }, + { + "epoch": 0.13996466911265118, + "grad_norm": 0.543357789516449, + "learning_rate": 0.0009825044163609185, + "loss": 3.8518, + "step": 2060 + }, + { + "epoch": 0.14030438918331295, + "grad_norm": 0.6752829551696777, + "learning_rate": 0.0009824619513520859, + "loss": 3.9744, + "step": 2065 + }, + { + "epoch": 0.14064410925397472, + "grad_norm": 0.6038799285888672, + "learning_rate": 0.0009824194863432532, + "loss": 3.6754, + "step": 2070 + }, + { + "epoch": 0.1409838293246365, + "grad_norm": 0.717085599899292, + "learning_rate": 0.0009823770213344203, + "loss": 3.5664, + "step": 2075 + }, + { + "epoch": 0.1413235493952983, + "grad_norm": 0.4654978811740875, + "learning_rate": 0.0009823345563255877, + "loss": 3.8664, + "step": 2080 + }, + { + "epoch": 0.14166326946596006, + "grad_norm": 0.5906335711479187, + "learning_rate": 0.000982292091316755, + "loss": 3.893, + "step": 2085 + }, + { + "epoch": 0.14200298953662183, + "grad_norm": 0.5542263388633728, + "learning_rate": 0.0009822496263079223, + "loss": 3.8011, + "step": 2090 + }, + { + "epoch": 0.1423427096072836, + "grad_norm": 0.6158429980278015, + "learning_rate": 0.0009822071612990897, + "loss": 3.8606, + "step": 2095 + }, + { + "epoch": 0.14268242967794537, + "grad_norm": 0.5313276052474976, + "learning_rate": 0.0009821646962902568, + "loss": 3.6783, + "step": 2100 + }, + { + "epoch": 0.14302214974860714, + "grad_norm": 0.6322357654571533, + "learning_rate": 0.0009821222312814241, + "loss": 3.8702, + "step": 2105 + }, + { + "epoch": 0.14336186981926893, + "grad_norm": 0.5938016176223755, + "learning_rate": 0.0009820797662725915, + "loss": 3.8795, + "step": 2110 + }, + { + "epoch": 0.1437015898899307, + "grad_norm": 0.7242531776428223, + "learning_rate": 0.0009820373012637586, + "loss": 3.4051, + "step": 2115 + }, + { + "epoch": 0.14404130996059247, + "grad_norm": 0.5325323939323425, + "learning_rate": 0.000981994836254926, + "loss": 3.9024, + "step": 2120 + }, + { + "epoch": 0.14438103003125424, + "grad_norm": 0.6194315552711487, + "learning_rate": 0.0009819523712460932, + "loss": 3.7453, + "step": 2125 + }, + { + "epoch": 0.144720750101916, + "grad_norm": 0.7634878754615784, + "learning_rate": 0.0009819099062372606, + "loss": 3.7032, + "step": 2130 + }, + { + "epoch": 0.14506047017257778, + "grad_norm": 0.8421163558959961, + "learning_rate": 0.0009818674412284277, + "loss": 3.7392, + "step": 2135 + }, + { + "epoch": 0.14540019024323958, + "grad_norm": 0.5848629474639893, + "learning_rate": 0.000981824976219595, + "loss": 3.7481, + "step": 2140 + }, + { + "epoch": 0.14573991031390135, + "grad_norm": 0.640381395816803, + "learning_rate": 0.0009817825112107624, + "loss": 3.8533, + "step": 2145 + }, + { + "epoch": 0.14607963038456312, + "grad_norm": 2.21744441986084, + "learning_rate": 0.0009817400462019295, + "loss": 3.627, + "step": 2150 + }, + { + "epoch": 0.1464193504552249, + "grad_norm": 0.7434030771255493, + "learning_rate": 0.000981697581193097, + "loss": 4.1559, + "step": 2155 + }, + { + "epoch": 0.14675907052588666, + "grad_norm": 0.6058130860328674, + "learning_rate": 0.0009816551161842641, + "loss": 3.7676, + "step": 2160 + }, + { + "epoch": 0.14709879059654846, + "grad_norm": 0.7785553932189941, + "learning_rate": 0.0009816126511754315, + "loss": 3.7069, + "step": 2165 + }, + { + "epoch": 0.14743851066721023, + "grad_norm": 0.664575457572937, + "learning_rate": 0.0009815701861665988, + "loss": 3.4932, + "step": 2170 + }, + { + "epoch": 0.147778230737872, + "grad_norm": 0.6050861477851868, + "learning_rate": 0.000981527721157766, + "loss": 3.9172, + "step": 2175 + }, + { + "epoch": 0.14811795080853377, + "grad_norm": 0.6301407217979431, + "learning_rate": 0.0009814852561489333, + "loss": 3.6907, + "step": 2180 + }, + { + "epoch": 0.14845767087919554, + "grad_norm": 0.6113725304603577, + "learning_rate": 0.0009814427911401006, + "loss": 3.7832, + "step": 2185 + }, + { + "epoch": 0.1487973909498573, + "grad_norm": 0.6586499214172363, + "learning_rate": 0.000981400326131268, + "loss": 3.7312, + "step": 2190 + }, + { + "epoch": 0.1491371110205191, + "grad_norm": 0.6039511561393738, + "learning_rate": 0.000981357861122435, + "loss": 3.7128, + "step": 2195 + }, + { + "epoch": 0.14947683109118087, + "grad_norm": 0.6801586747169495, + "learning_rate": 0.0009813153961136024, + "loss": 3.7563, + "step": 2200 + }, + { + "epoch": 0.14981655116184264, + "grad_norm": 0.5684835314750671, + "learning_rate": 0.0009812729311047697, + "loss": 3.7606, + "step": 2205 + }, + { + "epoch": 0.1501562712325044, + "grad_norm": 0.5985090732574463, + "learning_rate": 0.0009812304660959368, + "loss": 3.8605, + "step": 2210 + }, + { + "epoch": 0.15049599130316618, + "grad_norm": 0.6146273016929626, + "learning_rate": 0.0009811880010871042, + "loss": 3.735, + "step": 2215 + }, + { + "epoch": 0.15083571137382795, + "grad_norm": 0.6372639536857605, + "learning_rate": 0.0009811455360782715, + "loss": 3.8889, + "step": 2220 + }, + { + "epoch": 0.15117543144448975, + "grad_norm": 0.6332539916038513, + "learning_rate": 0.0009811030710694388, + "loss": 3.9223, + "step": 2225 + }, + { + "epoch": 0.15151515151515152, + "grad_norm": 0.6205673217773438, + "learning_rate": 0.0009810606060606062, + "loss": 3.4679, + "step": 2230 + }, + { + "epoch": 0.1518548715858133, + "grad_norm": 0.5693390965461731, + "learning_rate": 0.0009810181410517733, + "loss": 3.8416, + "step": 2235 + }, + { + "epoch": 0.15219459165647506, + "grad_norm": 0.5709364414215088, + "learning_rate": 0.0009809756760429406, + "loss": 3.8709, + "step": 2240 + }, + { + "epoch": 0.15253431172713683, + "grad_norm": 0.47355231642723083, + "learning_rate": 0.000980933211034108, + "loss": 3.8758, + "step": 2245 + }, + { + "epoch": 0.15287403179779863, + "grad_norm": 0.7836024165153503, + "learning_rate": 0.000980890746025275, + "loss": 3.576, + "step": 2250 + }, + { + "epoch": 0.1532137518684604, + "grad_norm": 0.6559258699417114, + "learning_rate": 0.0009808482810164426, + "loss": 3.8208, + "step": 2255 + }, + { + "epoch": 0.15355347193912217, + "grad_norm": 0.7773926854133606, + "learning_rate": 0.0009808058160076097, + "loss": 3.8461, + "step": 2260 + }, + { + "epoch": 0.15389319200978394, + "grad_norm": 0.6133778691291809, + "learning_rate": 0.000980763350998777, + "loss": 3.767, + "step": 2265 + }, + { + "epoch": 0.1542329120804457, + "grad_norm": 0.6255161166191101, + "learning_rate": 0.0009807208859899444, + "loss": 3.8351, + "step": 2270 + }, + { + "epoch": 0.15457263215110748, + "grad_norm": 0.9302341341972351, + "learning_rate": 0.0009806784209811115, + "loss": 3.7139, + "step": 2275 + }, + { + "epoch": 0.15491235222176927, + "grad_norm": 0.7181994915008545, + "learning_rate": 0.0009806359559722789, + "loss": 3.6355, + "step": 2280 + }, + { + "epoch": 0.15525207229243104, + "grad_norm": 0.6309781670570374, + "learning_rate": 0.0009805934909634462, + "loss": 3.6095, + "step": 2285 + }, + { + "epoch": 0.1555917923630928, + "grad_norm": 0.702839195728302, + "learning_rate": 0.0009805510259546135, + "loss": 3.6753, + "step": 2290 + }, + { + "epoch": 0.15593151243375458, + "grad_norm": 0.677828848361969, + "learning_rate": 0.0009805085609457807, + "loss": 3.5938, + "step": 2295 + }, + { + "epoch": 0.15627123250441635, + "grad_norm": 1.157195806503296, + "learning_rate": 0.000980466095936948, + "loss": 3.7628, + "step": 2300 + }, + { + "epoch": 0.15661095257507815, + "grad_norm": 0.6162868142127991, + "learning_rate": 0.0009804236309281153, + "loss": 3.7932, + "step": 2305 + }, + { + "epoch": 0.15695067264573992, + "grad_norm": 0.6598204970359802, + "learning_rate": 0.0009803811659192824, + "loss": 3.6078, + "step": 2310 + }, + { + "epoch": 0.1572903927164017, + "grad_norm": 0.5610998272895813, + "learning_rate": 0.0009803387009104498, + "loss": 3.8062, + "step": 2315 + }, + { + "epoch": 0.15763011278706346, + "grad_norm": 0.6655665636062622, + "learning_rate": 0.0009802962359016171, + "loss": 3.7769, + "step": 2320 + }, + { + "epoch": 0.15796983285772523, + "grad_norm": 0.6042912006378174, + "learning_rate": 0.0009802537708927844, + "loss": 3.9204, + "step": 2325 + }, + { + "epoch": 0.158309552928387, + "grad_norm": 0.5904920697212219, + "learning_rate": 0.0009802113058839518, + "loss": 3.6929, + "step": 2330 + }, + { + "epoch": 0.1586492729990488, + "grad_norm": 0.5653850436210632, + "learning_rate": 0.000980168840875119, + "loss": 3.7477, + "step": 2335 + }, + { + "epoch": 0.15898899306971057, + "grad_norm": 0.5770355463027954, + "learning_rate": 0.0009801263758662862, + "loss": 3.7471, + "step": 2340 + }, + { + "epoch": 0.15932871314037234, + "grad_norm": 0.6086682081222534, + "learning_rate": 0.0009800839108574536, + "loss": 3.8202, + "step": 2345 + }, + { + "epoch": 0.1596684332110341, + "grad_norm": 0.6483132839202881, + "learning_rate": 0.0009800414458486207, + "loss": 3.9549, + "step": 2350 + }, + { + "epoch": 0.16000815328169588, + "grad_norm": 0.5497443675994873, + "learning_rate": 0.000979998980839788, + "loss": 3.7957, + "step": 2355 + }, + { + "epoch": 0.16034787335235764, + "grad_norm": 0.6367260813713074, + "learning_rate": 0.0009799565158309553, + "loss": 3.8473, + "step": 2360 + }, + { + "epoch": 0.16068759342301944, + "grad_norm": 0.6136810183525085, + "learning_rate": 0.0009799140508221227, + "loss": 3.5634, + "step": 2365 + }, + { + "epoch": 0.1610273134936812, + "grad_norm": 0.8666849136352539, + "learning_rate": 0.0009798715858132898, + "loss": 3.7025, + "step": 2370 + }, + { + "epoch": 0.16136703356434298, + "grad_norm": 0.6140803694725037, + "learning_rate": 0.0009798291208044571, + "loss": 3.7509, + "step": 2375 + }, + { + "epoch": 0.16170675363500475, + "grad_norm": 0.8208211660385132, + "learning_rate": 0.0009797866557956245, + "loss": 3.6264, + "step": 2380 + }, + { + "epoch": 0.16204647370566652, + "grad_norm": 0.6705807447433472, + "learning_rate": 0.0009797441907867916, + "loss": 3.7361, + "step": 2385 + }, + { + "epoch": 0.16238619377632832, + "grad_norm": 0.7236992120742798, + "learning_rate": 0.000979701725777959, + "loss": 3.8665, + "step": 2390 + }, + { + "epoch": 0.1627259138469901, + "grad_norm": 1.1335411071777344, + "learning_rate": 0.0009796592607691263, + "loss": 3.5574, + "step": 2395 + }, + { + "epoch": 0.16306563391765186, + "grad_norm": 0.565281331539154, + "learning_rate": 0.0009796167957602936, + "loss": 3.7638, + "step": 2400 + }, + { + "epoch": 0.16340535398831363, + "grad_norm": 0.5212329626083374, + "learning_rate": 0.000979574330751461, + "loss": 3.5795, + "step": 2405 + }, + { + "epoch": 0.1637450740589754, + "grad_norm": 0.5030485391616821, + "learning_rate": 0.000979531865742628, + "loss": 3.8794, + "step": 2410 + }, + { + "epoch": 0.16408479412963717, + "grad_norm": 0.609352171421051, + "learning_rate": 0.0009794894007337954, + "loss": 3.9417, + "step": 2415 + }, + { + "epoch": 0.16442451420029897, + "grad_norm": 0.7634018659591675, + "learning_rate": 0.0009794469357249627, + "loss": 3.8014, + "step": 2420 + }, + { + "epoch": 0.16476423427096074, + "grad_norm": 0.7297620177268982, + "learning_rate": 0.0009794044707161298, + "loss": 3.733, + "step": 2425 + }, + { + "epoch": 0.1651039543416225, + "grad_norm": 0.6963150501251221, + "learning_rate": 0.0009793620057072972, + "loss": 3.7993, + "step": 2430 + }, + { + "epoch": 0.16544367441228428, + "grad_norm": 0.6455408930778503, + "learning_rate": 0.0009793195406984645, + "loss": 3.8651, + "step": 2435 + }, + { + "epoch": 0.16578339448294604, + "grad_norm": 0.587772011756897, + "learning_rate": 0.0009792770756896318, + "loss": 3.7844, + "step": 2440 + }, + { + "epoch": 0.16612311455360781, + "grad_norm": 1.017780065536499, + "learning_rate": 0.000979234610680799, + "loss": 3.8287, + "step": 2445 + }, + { + "epoch": 0.1664628346242696, + "grad_norm": 0.5755481719970703, + "learning_rate": 0.0009791921456719663, + "loss": 3.6867, + "step": 2450 + }, + { + "epoch": 0.16680255469493138, + "grad_norm": 0.8737612366676331, + "learning_rate": 0.0009791496806631336, + "loss": 3.7362, + "step": 2455 + }, + { + "epoch": 0.16714227476559315, + "grad_norm": 0.6732498407363892, + "learning_rate": 0.0009791072156543007, + "loss": 3.8185, + "step": 2460 + }, + { + "epoch": 0.16748199483625492, + "grad_norm": 0.6353683471679688, + "learning_rate": 0.0009790647506454683, + "loss": 3.6812, + "step": 2465 + }, + { + "epoch": 0.1678217149069167, + "grad_norm": 0.7803769111633301, + "learning_rate": 0.0009790222856366354, + "loss": 3.8419, + "step": 2470 + }, + { + "epoch": 0.1681614349775785, + "grad_norm": 0.8309979438781738, + "learning_rate": 0.0009789798206278027, + "loss": 3.8639, + "step": 2475 + }, + { + "epoch": 0.16850115504824026, + "grad_norm": 0.6595731973648071, + "learning_rate": 0.00097893735561897, + "loss": 3.7646, + "step": 2480 + }, + { + "epoch": 0.16884087511890203, + "grad_norm": 0.664367139339447, + "learning_rate": 0.0009788948906101372, + "loss": 3.8229, + "step": 2485 + }, + { + "epoch": 0.1691805951895638, + "grad_norm": 0.482677161693573, + "learning_rate": 0.0009788524256013045, + "loss": 3.8088, + "step": 2490 + }, + { + "epoch": 0.16952031526022557, + "grad_norm": 0.691443681716919, + "learning_rate": 0.0009788099605924719, + "loss": 3.7539, + "step": 2495 + }, + { + "epoch": 0.16986003533088734, + "grad_norm": 0.5566770434379578, + "learning_rate": 0.0009787674955836392, + "loss": 3.8654, + "step": 2500 + }, + { + "epoch": 0.17019975540154914, + "grad_norm": 0.6081881523132324, + "learning_rate": 0.0009787250305748063, + "loss": 3.9523, + "step": 2505 + }, + { + "epoch": 0.1705394754722109, + "grad_norm": 0.5837857723236084, + "learning_rate": 0.0009786825655659736, + "loss": 3.9644, + "step": 2510 + }, + { + "epoch": 0.17087919554287267, + "grad_norm": 0.5104924440383911, + "learning_rate": 0.000978640100557141, + "loss": 3.7676, + "step": 2515 + }, + { + "epoch": 0.17121891561353444, + "grad_norm": 0.511321485042572, + "learning_rate": 0.000978597635548308, + "loss": 3.9301, + "step": 2520 + }, + { + "epoch": 0.17155863568419621, + "grad_norm": 0.5427649021148682, + "learning_rate": 0.0009785551705394754, + "loss": 3.8802, + "step": 2525 + }, + { + "epoch": 0.17189835575485798, + "grad_norm": 0.6896429657936096, + "learning_rate": 0.0009785127055306428, + "loss": 3.7196, + "step": 2530 + }, + { + "epoch": 0.17223807582551978, + "grad_norm": 0.7047936320304871, + "learning_rate": 0.00097847024052181, + "loss": 3.9621, + "step": 2535 + }, + { + "epoch": 0.17257779589618155, + "grad_norm": 0.6143273115158081, + "learning_rate": 0.0009784277755129774, + "loss": 3.6378, + "step": 2540 + }, + { + "epoch": 0.17291751596684332, + "grad_norm": 0.6251726746559143, + "learning_rate": 0.0009783853105041446, + "loss": 3.7208, + "step": 2545 + }, + { + "epoch": 0.1732572360375051, + "grad_norm": 0.5483553409576416, + "learning_rate": 0.0009783428454953119, + "loss": 3.7922, + "step": 2550 + }, + { + "epoch": 0.17359695610816686, + "grad_norm": 0.6072860360145569, + "learning_rate": 0.0009783003804864792, + "loss": 3.7269, + "step": 2555 + }, + { + "epoch": 0.17393667617882866, + "grad_norm": 0.5628990530967712, + "learning_rate": 0.0009782579154776463, + "loss": 3.5317, + "step": 2560 + }, + { + "epoch": 0.17427639624949043, + "grad_norm": 0.8806532621383667, + "learning_rate": 0.0009782154504688137, + "loss": 3.6866, + "step": 2565 + }, + { + "epoch": 0.1746161163201522, + "grad_norm": 0.7904943227767944, + "learning_rate": 0.000978172985459981, + "loss": 3.8671, + "step": 2570 + }, + { + "epoch": 0.17495583639081397, + "grad_norm": 0.9674472808837891, + "learning_rate": 0.0009781305204511483, + "loss": 3.9379, + "step": 2575 + }, + { + "epoch": 0.17529555646147574, + "grad_norm": 0.6988072991371155, + "learning_rate": 0.0009780880554423157, + "loss": 3.9829, + "step": 2580 + }, + { + "epoch": 0.1756352765321375, + "grad_norm": 0.5407254099845886, + "learning_rate": 0.0009780455904334828, + "loss": 3.6546, + "step": 2585 + }, + { + "epoch": 0.1759749966027993, + "grad_norm": 0.5559465885162354, + "learning_rate": 0.0009780031254246501, + "loss": 3.6391, + "step": 2590 + }, + { + "epoch": 0.17631471667346107, + "grad_norm": 0.8155473470687866, + "learning_rate": 0.0009779606604158175, + "loss": 3.7127, + "step": 2595 + }, + { + "epoch": 0.17665443674412284, + "grad_norm": 0.7560978531837463, + "learning_rate": 0.0009779181954069846, + "loss": 3.5354, + "step": 2600 + }, + { + "epoch": 0.17699415681478461, + "grad_norm": 1.3758387565612793, + "learning_rate": 0.000977875730398152, + "loss": 4.0057, + "step": 2605 + }, + { + "epoch": 0.17733387688544638, + "grad_norm": 0.6253563761711121, + "learning_rate": 0.0009778332653893192, + "loss": 3.7131, + "step": 2610 + }, + { + "epoch": 0.17767359695610815, + "grad_norm": 0.5841245055198669, + "learning_rate": 0.0009777908003804866, + "loss": 3.4909, + "step": 2615 + }, + { + "epoch": 0.17801331702676995, + "grad_norm": 0.7301042079925537, + "learning_rate": 0.0009777483353716537, + "loss": 3.7663, + "step": 2620 + }, + { + "epoch": 0.17835303709743172, + "grad_norm": 0.6407013535499573, + "learning_rate": 0.000977705870362821, + "loss": 3.6259, + "step": 2625 + }, + { + "epoch": 0.1786927571680935, + "grad_norm": 0.6893696784973145, + "learning_rate": 0.0009776634053539884, + "loss": 3.633, + "step": 2630 + }, + { + "epoch": 0.17903247723875526, + "grad_norm": 0.5852348208427429, + "learning_rate": 0.0009776209403451555, + "loss": 3.791, + "step": 2635 + }, + { + "epoch": 0.17937219730941703, + "grad_norm": 0.5743594765663147, + "learning_rate": 0.000977578475336323, + "loss": 3.8689, + "step": 2640 + }, + { + "epoch": 0.17971191738007883, + "grad_norm": 0.5142814517021179, + "learning_rate": 0.0009775360103274902, + "loss": 3.7479, + "step": 2645 + }, + { + "epoch": 0.1800516374507406, + "grad_norm": 0.6178168654441833, + "learning_rate": 0.0009774935453186575, + "loss": 3.9161, + "step": 2650 + }, + { + "epoch": 0.18039135752140237, + "grad_norm": 0.6452828645706177, + "learning_rate": 0.0009774510803098248, + "loss": 3.8226, + "step": 2655 + }, + { + "epoch": 0.18073107759206414, + "grad_norm": 0.5602723360061646, + "learning_rate": 0.000977408615300992, + "loss": 3.6176, + "step": 2660 + }, + { + "epoch": 0.1810707976627259, + "grad_norm": 0.6433743834495544, + "learning_rate": 0.0009773661502921593, + "loss": 3.6957, + "step": 2665 + }, + { + "epoch": 0.18141051773338768, + "grad_norm": 0.7300789952278137, + "learning_rate": 0.0009773236852833266, + "loss": 3.7096, + "step": 2670 + }, + { + "epoch": 0.18175023780404947, + "grad_norm": 1.0117199420928955, + "learning_rate": 0.000977281220274494, + "loss": 3.6927, + "step": 2675 + }, + { + "epoch": 0.18208995787471124, + "grad_norm": 0.7773440480232239, + "learning_rate": 0.000977238755265661, + "loss": 3.8849, + "step": 2680 + }, + { + "epoch": 0.18242967794537301, + "grad_norm": 0.48008543252944946, + "learning_rate": 0.0009771962902568284, + "loss": 3.7642, + "step": 2685 + }, + { + "epoch": 0.18276939801603478, + "grad_norm": 0.5967749953269958, + "learning_rate": 0.0009771538252479957, + "loss": 3.7564, + "step": 2690 + }, + { + "epoch": 0.18310911808669655, + "grad_norm": 0.6126387715339661, + "learning_rate": 0.0009771113602391628, + "loss": 3.3185, + "step": 2695 + }, + { + "epoch": 0.18344883815735832, + "grad_norm": 0.6584519147872925, + "learning_rate": 0.0009770688952303302, + "loss": 3.8923, + "step": 2700 + }, + { + "epoch": 0.18378855822802012, + "grad_norm": 0.6974896192550659, + "learning_rate": 0.0009770264302214975, + "loss": 3.6187, + "step": 2705 + }, + { + "epoch": 0.1841282782986819, + "grad_norm": 0.6637903451919556, + "learning_rate": 0.0009769839652126648, + "loss": 3.677, + "step": 2710 + }, + { + "epoch": 0.18446799836934366, + "grad_norm": 0.9200413227081299, + "learning_rate": 0.0009769415002038322, + "loss": 3.7962, + "step": 2715 + }, + { + "epoch": 0.18480771844000543, + "grad_norm": 0.69618159532547, + "learning_rate": 0.0009768990351949993, + "loss": 3.7602, + "step": 2720 + }, + { + "epoch": 0.1851474385106672, + "grad_norm": 0.6195826530456543, + "learning_rate": 0.0009768565701861666, + "loss": 3.8316, + "step": 2725 + }, + { + "epoch": 0.185487158581329, + "grad_norm": 0.6163364052772522, + "learning_rate": 0.000976814105177334, + "loss": 3.9855, + "step": 2730 + }, + { + "epoch": 0.18582687865199077, + "grad_norm": 0.6414899230003357, + "learning_rate": 0.000976771640168501, + "loss": 3.8684, + "step": 2735 + }, + { + "epoch": 0.18616659872265254, + "grad_norm": 0.5263423323631287, + "learning_rate": 0.0009767291751596684, + "loss": 3.5558, + "step": 2740 + }, + { + "epoch": 0.1865063187933143, + "grad_norm": 0.49367937445640564, + "learning_rate": 0.0009766867101508358, + "loss": 3.6349, + "step": 2745 + }, + { + "epoch": 0.18684603886397608, + "grad_norm": 0.691084623336792, + "learning_rate": 0.000976644245142003, + "loss": 3.8187, + "step": 2750 + }, + { + "epoch": 0.18718575893463785, + "grad_norm": 0.6548663377761841, + "learning_rate": 0.0009766017801331702, + "loss": 4.0068, + "step": 2755 + }, + { + "epoch": 0.18752547900529964, + "grad_norm": 0.6146909594535828, + "learning_rate": 0.0009765593151243376, + "loss": 3.7553, + "step": 2760 + }, + { + "epoch": 0.18786519907596141, + "grad_norm": 0.5466817617416382, + "learning_rate": 0.0009765168501155049, + "loss": 3.8668, + "step": 2765 + }, + { + "epoch": 0.18820491914662318, + "grad_norm": 0.4874761402606964, + "learning_rate": 0.0009764743851066721, + "loss": 3.8772, + "step": 2770 + }, + { + "epoch": 0.18854463921728495, + "grad_norm": 0.6687860488891602, + "learning_rate": 0.0009764319200978394, + "loss": 3.9953, + "step": 2775 + }, + { + "epoch": 0.18888435928794672, + "grad_norm": 0.5901328325271606, + "learning_rate": 0.0009763894550890067, + "loss": 3.8041, + "step": 2780 + }, + { + "epoch": 0.1892240793586085, + "grad_norm": 0.5916528701782227, + "learning_rate": 0.0009763469900801739, + "loss": 3.428, + "step": 2785 + }, + { + "epoch": 0.1895637994292703, + "grad_norm": 0.6788733601570129, + "learning_rate": 0.0009763045250713412, + "loss": 3.7871, + "step": 2790 + }, + { + "epoch": 0.18990351949993206, + "grad_norm": 0.6555401086807251, + "learning_rate": 0.0009762620600625086, + "loss": 3.5644, + "step": 2795 + }, + { + "epoch": 0.19024323957059383, + "grad_norm": 0.7433831691741943, + "learning_rate": 0.0009762195950536758, + "loss": 3.718, + "step": 2800 + }, + { + "epoch": 0.1905829596412556, + "grad_norm": 0.7285386919975281, + "learning_rate": 0.0009761771300448431, + "loss": 3.7351, + "step": 2805 + }, + { + "epoch": 0.19092267971191737, + "grad_norm": 0.5669515132904053, + "learning_rate": 0.0009761346650360103, + "loss": 3.2801, + "step": 2810 + }, + { + "epoch": 0.19126239978257917, + "grad_norm": 0.8962259888648987, + "learning_rate": 0.0009760922000271776, + "loss": 3.7908, + "step": 2815 + }, + { + "epoch": 0.19160211985324094, + "grad_norm": 0.6690675020217896, + "learning_rate": 0.0009760497350183449, + "loss": 3.8255, + "step": 2820 + }, + { + "epoch": 0.1919418399239027, + "grad_norm": 0.5441779494285583, + "learning_rate": 0.0009760072700095121, + "loss": 3.6706, + "step": 2825 + }, + { + "epoch": 0.19228155999456448, + "grad_norm": 0.5559229850769043, + "learning_rate": 0.0009759648050006795, + "loss": 3.8617, + "step": 2830 + }, + { + "epoch": 0.19262128006522625, + "grad_norm": 0.73357093334198, + "learning_rate": 0.0009759223399918468, + "loss": 3.7379, + "step": 2835 + }, + { + "epoch": 0.19296100013588802, + "grad_norm": 0.7806567549705505, + "learning_rate": 0.000975879874983014, + "loss": 3.7481, + "step": 2840 + }, + { + "epoch": 0.19330072020654981, + "grad_norm": 0.4998747706413269, + "learning_rate": 0.0009758374099741812, + "loss": 3.8569, + "step": 2845 + }, + { + "epoch": 0.19364044027721158, + "grad_norm": 0.5296024680137634, + "learning_rate": 0.0009757949449653486, + "loss": 3.7117, + "step": 2850 + }, + { + "epoch": 0.19398016034787335, + "grad_norm": 0.867391049861908, + "learning_rate": 0.0009757524799565158, + "loss": 3.8734, + "step": 2855 + }, + { + "epoch": 0.19431988041853512, + "grad_norm": 0.6658104062080383, + "learning_rate": 0.000975710014947683, + "loss": 3.6793, + "step": 2860 + }, + { + "epoch": 0.1946596004891969, + "grad_norm": 0.8781140446662903, + "learning_rate": 0.0009756675499388505, + "loss": 3.7553, + "step": 2865 + }, + { + "epoch": 0.19499932055985866, + "grad_norm": 0.849490761756897, + "learning_rate": 0.0009756250849300177, + "loss": 3.5455, + "step": 2870 + }, + { + "epoch": 0.19533904063052046, + "grad_norm": 0.8140732645988464, + "learning_rate": 0.0009755826199211849, + "loss": 3.7532, + "step": 2875 + }, + { + "epoch": 0.19567876070118223, + "grad_norm": 0.6011113524436951, + "learning_rate": 0.0009755401549123523, + "loss": 3.8898, + "step": 2880 + }, + { + "epoch": 0.196018480771844, + "grad_norm": 0.6402468085289001, + "learning_rate": 0.0009754976899035195, + "loss": 3.7352, + "step": 2885 + }, + { + "epoch": 0.19635820084250577, + "grad_norm": 0.6490626931190491, + "learning_rate": 0.0009754552248946867, + "loss": 3.4159, + "step": 2890 + }, + { + "epoch": 0.19669792091316754, + "grad_norm": 0.6855554580688477, + "learning_rate": 0.000975412759885854, + "loss": 3.5553, + "step": 2895 + }, + { + "epoch": 0.19703764098382934, + "grad_norm": 0.7514801621437073, + "learning_rate": 0.0009753702948770214, + "loss": 3.8005, + "step": 2900 + }, + { + "epoch": 0.1973773610544911, + "grad_norm": 2.0744807720184326, + "learning_rate": 0.0009753278298681887, + "loss": 3.5033, + "step": 2905 + }, + { + "epoch": 0.19771708112515288, + "grad_norm": 1.0016511678695679, + "learning_rate": 0.0009752853648593559, + "loss": 3.6079, + "step": 2910 + }, + { + "epoch": 0.19805680119581465, + "grad_norm": 0.7487754225730896, + "learning_rate": 0.0009752428998505232, + "loss": 3.7548, + "step": 2915 + }, + { + "epoch": 0.19839652126647642, + "grad_norm": 0.6639905571937561, + "learning_rate": 0.0009752004348416905, + "loss": 3.7385, + "step": 2920 + }, + { + "epoch": 0.1987362413371382, + "grad_norm": 0.6206386089324951, + "learning_rate": 0.0009751579698328577, + "loss": 3.7201, + "step": 2925 + }, + { + "epoch": 0.19907596140779998, + "grad_norm": 0.6712602376937866, + "learning_rate": 0.000975115504824025, + "loss": 3.5394, + "step": 2930 + }, + { + "epoch": 0.19941568147846175, + "grad_norm": 0.620564877986908, + "learning_rate": 0.0009750730398151924, + "loss": 3.8665, + "step": 2935 + }, + { + "epoch": 0.19975540154912352, + "grad_norm": 1.075077772140503, + "learning_rate": 0.0009750305748063596, + "loss": 3.5622, + "step": 2940 + }, + { + "epoch": 0.2000951216197853, + "grad_norm": 0.7147380709648132, + "learning_rate": 0.0009749881097975268, + "loss": 3.7676, + "step": 2945 + }, + { + "epoch": 0.20043484169044706, + "grad_norm": 0.7028825283050537, + "learning_rate": 0.0009749456447886942, + "loss": 3.8291, + "step": 2950 + }, + { + "epoch": 0.20077456176110883, + "grad_norm": 0.6004649996757507, + "learning_rate": 0.0009749031797798614, + "loss": 4.0362, + "step": 2955 + }, + { + "epoch": 0.20111428183177063, + "grad_norm": 0.7003999948501587, + "learning_rate": 0.0009748607147710286, + "loss": 3.9086, + "step": 2960 + }, + { + "epoch": 0.2014540019024324, + "grad_norm": 0.483829528093338, + "learning_rate": 0.000974818249762196, + "loss": 3.5937, + "step": 2965 + }, + { + "epoch": 0.20179372197309417, + "grad_norm": 0.5830460786819458, + "learning_rate": 0.0009747757847533633, + "loss": 3.709, + "step": 2970 + }, + { + "epoch": 0.20213344204375594, + "grad_norm": 0.6097733974456787, + "learning_rate": 0.0009747333197445305, + "loss": 3.6495, + "step": 2975 + }, + { + "epoch": 0.2024731621144177, + "grad_norm": 0.6874327659606934, + "learning_rate": 0.0009746908547356979, + "loss": 3.7219, + "step": 2980 + }, + { + "epoch": 0.2028128821850795, + "grad_norm": 0.7092781662940979, + "learning_rate": 0.0009746483897268651, + "loss": 3.6287, + "step": 2985 + }, + { + "epoch": 0.20315260225574128, + "grad_norm": 0.5750575065612793, + "learning_rate": 0.0009746059247180323, + "loss": 3.759, + "step": 2990 + }, + { + "epoch": 0.20349232232640305, + "grad_norm": 0.8717525601387024, + "learning_rate": 0.0009745634597091996, + "loss": 3.7773, + "step": 2995 + }, + { + "epoch": 0.20383204239706482, + "grad_norm": 0.5540760159492493, + "learning_rate": 0.0009745209947003669, + "loss": 3.627, + "step": 3000 + }, + { + "epoch": 0.2041717624677266, + "grad_norm": 0.9130904674530029, + "learning_rate": 0.0009744785296915342, + "loss": 3.6937, + "step": 3005 + }, + { + "epoch": 0.20451148253838836, + "grad_norm": 0.6381457448005676, + "learning_rate": 0.0009744360646827015, + "loss": 3.8555, + "step": 3010 + }, + { + "epoch": 0.20485120260905015, + "grad_norm": 0.6486433744430542, + "learning_rate": 0.0009743935996738688, + "loss": 3.8618, + "step": 3015 + }, + { + "epoch": 0.20519092267971192, + "grad_norm": 0.6014708280563354, + "learning_rate": 0.000974351134665036, + "loss": 3.8432, + "step": 3020 + }, + { + "epoch": 0.2055306427503737, + "grad_norm": 0.5839025378227234, + "learning_rate": 0.0009743086696562033, + "loss": 3.7629, + "step": 3025 + }, + { + "epoch": 0.20587036282103546, + "grad_norm": 0.7098844051361084, + "learning_rate": 0.0009742662046473706, + "loss": 3.75, + "step": 3030 + }, + { + "epoch": 0.20621008289169723, + "grad_norm": 0.5805578827857971, + "learning_rate": 0.0009742237396385378, + "loss": 3.6937, + "step": 3035 + }, + { + "epoch": 0.206549802962359, + "grad_norm": 0.6617678999900818, + "learning_rate": 0.0009741812746297052, + "loss": 3.7325, + "step": 3040 + }, + { + "epoch": 0.2068895230330208, + "grad_norm": 0.7151147127151489, + "learning_rate": 0.0009741388096208724, + "loss": 3.6406, + "step": 3045 + }, + { + "epoch": 0.20722924310368257, + "grad_norm": 0.588606059551239, + "learning_rate": 0.0009740963446120397, + "loss": 3.6841, + "step": 3050 + }, + { + "epoch": 0.20756896317434434, + "grad_norm": 0.6761735081672668, + "learning_rate": 0.000974053879603207, + "loss": 3.5365, + "step": 3055 + }, + { + "epoch": 0.2079086832450061, + "grad_norm": 0.8985639810562134, + "learning_rate": 0.0009740114145943742, + "loss": 3.8136, + "step": 3060 + }, + { + "epoch": 0.20824840331566788, + "grad_norm": 0.5298454761505127, + "learning_rate": 0.0009739689495855415, + "loss": 3.5731, + "step": 3065 + }, + { + "epoch": 0.20858812338632968, + "grad_norm": 0.6606918573379517, + "learning_rate": 0.0009739264845767088, + "loss": 3.858, + "step": 3070 + }, + { + "epoch": 0.20892784345699145, + "grad_norm": 0.5564335584640503, + "learning_rate": 0.0009738840195678761, + "loss": 3.653, + "step": 3075 + }, + { + "epoch": 0.20926756352765322, + "grad_norm": 0.6696929335594177, + "learning_rate": 0.0009738415545590434, + "loss": 3.6613, + "step": 3080 + }, + { + "epoch": 0.209607283598315, + "grad_norm": 2.227457046508789, + "learning_rate": 0.0009737990895502107, + "loss": 3.9475, + "step": 3085 + }, + { + "epoch": 0.20994700366897676, + "grad_norm": 0.795903742313385, + "learning_rate": 0.0009737566245413779, + "loss": 3.7129, + "step": 3090 + }, + { + "epoch": 0.21028672373963853, + "grad_norm": 0.6205352544784546, + "learning_rate": 0.0009737141595325451, + "loss": 3.8524, + "step": 3095 + }, + { + "epoch": 0.21062644381030032, + "grad_norm": 0.6212173104286194, + "learning_rate": 0.0009736716945237125, + "loss": 3.9522, + "step": 3100 + }, + { + "epoch": 0.2109661638809621, + "grad_norm": 1.2397836446762085, + "learning_rate": 0.0009736292295148797, + "loss": 3.6891, + "step": 3105 + }, + { + "epoch": 0.21130588395162386, + "grad_norm": 0.7002935409545898, + "learning_rate": 0.000973586764506047, + "loss": 3.9087, + "step": 3110 + }, + { + "epoch": 0.21164560402228563, + "grad_norm": 0.6223004460334778, + "learning_rate": 0.0009735442994972144, + "loss": 3.8046, + "step": 3115 + }, + { + "epoch": 0.2119853240929474, + "grad_norm": 1.711371898651123, + "learning_rate": 0.0009735018344883816, + "loss": 3.8441, + "step": 3120 + }, + { + "epoch": 0.21232504416360917, + "grad_norm": 0.7471145391464233, + "learning_rate": 0.0009734593694795488, + "loss": 3.4484, + "step": 3125 + }, + { + "epoch": 0.21266476423427097, + "grad_norm": 0.700401246547699, + "learning_rate": 0.0009734169044707162, + "loss": 3.9248, + "step": 3130 + }, + { + "epoch": 0.21300448430493274, + "grad_norm": 3.3507747650146484, + "learning_rate": 0.0009733744394618834, + "loss": 3.7096, + "step": 3135 + }, + { + "epoch": 0.2133442043755945, + "grad_norm": 0.6101709008216858, + "learning_rate": 0.0009733319744530506, + "loss": 3.6657, + "step": 3140 + }, + { + "epoch": 0.21368392444625628, + "grad_norm": 0.6883925795555115, + "learning_rate": 0.000973289509444218, + "loss": 3.9038, + "step": 3145 + }, + { + "epoch": 0.21402364451691805, + "grad_norm": 0.6930251121520996, + "learning_rate": 0.0009732470444353853, + "loss": 3.3827, + "step": 3150 + }, + { + "epoch": 0.21436336458757985, + "grad_norm": 1.6364690065383911, + "learning_rate": 0.0009732045794265525, + "loss": 3.7638, + "step": 3155 + }, + { + "epoch": 0.21470308465824162, + "grad_norm": 0.5960918068885803, + "learning_rate": 0.0009731621144177198, + "loss": 3.7745, + "step": 3160 + }, + { + "epoch": 0.2150428047289034, + "grad_norm": 0.48007920384407043, + "learning_rate": 0.0009731196494088871, + "loss": 3.7569, + "step": 3165 + }, + { + "epoch": 0.21538252479956516, + "grad_norm": 0.8826533555984497, + "learning_rate": 0.0009730771844000543, + "loss": 3.5113, + "step": 3170 + }, + { + "epoch": 0.21572224487022693, + "grad_norm": 0.9135482311248779, + "learning_rate": 0.0009730347193912217, + "loss": 3.7115, + "step": 3175 + }, + { + "epoch": 0.2160619649408887, + "grad_norm": 0.5753097534179688, + "learning_rate": 0.000972992254382389, + "loss": 3.7479, + "step": 3180 + }, + { + "epoch": 0.2164016850115505, + "grad_norm": 0.7939460873603821, + "learning_rate": 0.0009729497893735562, + "loss": 3.7569, + "step": 3185 + }, + { + "epoch": 0.21674140508221226, + "grad_norm": 0.8315974473953247, + "learning_rate": 0.0009729073243647235, + "loss": 4.0186, + "step": 3190 + }, + { + "epoch": 0.21708112515287403, + "grad_norm": 0.613703727722168, + "learning_rate": 0.0009728648593558907, + "loss": 3.7061, + "step": 3195 + }, + { + "epoch": 0.2174208452235358, + "grad_norm": 0.5988765358924866, + "learning_rate": 0.000972822394347058, + "loss": 3.8294, + "step": 3200 + }, + { + "epoch": 0.21776056529419757, + "grad_norm": 0.8050014972686768, + "learning_rate": 0.0009727799293382253, + "loss": 3.9711, + "step": 3205 + }, + { + "epoch": 0.21810028536485934, + "grad_norm": 0.6768515110015869, + "learning_rate": 0.0009727374643293926, + "loss": 3.9341, + "step": 3210 + }, + { + "epoch": 0.21844000543552114, + "grad_norm": 0.6579541563987732, + "learning_rate": 0.0009726949993205599, + "loss": 3.7913, + "step": 3215 + }, + { + "epoch": 0.2187797255061829, + "grad_norm": 0.7019892334938049, + "learning_rate": 0.0009726525343117272, + "loss": 3.6519, + "step": 3220 + }, + { + "epoch": 0.21911944557684468, + "grad_norm": 0.67347651720047, + "learning_rate": 0.0009726100693028944, + "loss": 3.6208, + "step": 3225 + }, + { + "epoch": 0.21945916564750645, + "grad_norm": 0.5936116576194763, + "learning_rate": 0.0009725676042940616, + "loss": 3.7613, + "step": 3230 + }, + { + "epoch": 0.21979888571816822, + "grad_norm": 0.6363143920898438, + "learning_rate": 0.000972525139285229, + "loss": 3.6958, + "step": 3235 + }, + { + "epoch": 0.22013860578883002, + "grad_norm": 0.653922975063324, + "learning_rate": 0.0009724826742763962, + "loss": 3.6718, + "step": 3240 + }, + { + "epoch": 0.22047832585949179, + "grad_norm": 0.5786622166633606, + "learning_rate": 0.0009724402092675637, + "loss": 3.6836, + "step": 3245 + }, + { + "epoch": 0.22081804593015356, + "grad_norm": 0.6533204913139343, + "learning_rate": 0.0009723977442587309, + "loss": 3.7397, + "step": 3250 + }, + { + "epoch": 0.22115776600081533, + "grad_norm": 0.5819668173789978, + "learning_rate": 0.0009723552792498981, + "loss": 3.6297, + "step": 3255 + }, + { + "epoch": 0.2214974860714771, + "grad_norm": 0.6860673427581787, + "learning_rate": 0.0009723128142410654, + "loss": 3.7599, + "step": 3260 + }, + { + "epoch": 0.22183720614213887, + "grad_norm": 0.5516849756240845, + "learning_rate": 0.0009722703492322327, + "loss": 3.7819, + "step": 3265 + }, + { + "epoch": 0.22217692621280066, + "grad_norm": 0.7628173232078552, + "learning_rate": 0.0009722278842233999, + "loss": 3.8201, + "step": 3270 + }, + { + "epoch": 0.22251664628346243, + "grad_norm": 0.6516159176826477, + "learning_rate": 0.0009721854192145672, + "loss": 3.8293, + "step": 3275 + }, + { + "epoch": 0.2228563663541242, + "grad_norm": 0.5998335480690002, + "learning_rate": 0.0009721429542057346, + "loss": 3.6546, + "step": 3280 + }, + { + "epoch": 0.22319608642478597, + "grad_norm": 0.6493220329284668, + "learning_rate": 0.0009721004891969018, + "loss": 3.999, + "step": 3285 + }, + { + "epoch": 0.22353580649544774, + "grad_norm": 0.9815689921379089, + "learning_rate": 0.0009720580241880691, + "loss": 3.7525, + "step": 3290 + }, + { + "epoch": 0.22387552656610954, + "grad_norm": 0.7050333023071289, + "learning_rate": 0.0009720155591792363, + "loss": 3.4007, + "step": 3295 + }, + { + "epoch": 0.2242152466367713, + "grad_norm": 0.8392935395240784, + "learning_rate": 0.0009719730941704036, + "loss": 3.6176, + "step": 3300 + }, + { + "epoch": 0.22455496670743308, + "grad_norm": 0.7056354284286499, + "learning_rate": 0.0009719306291615709, + "loss": 3.8879, + "step": 3305 + }, + { + "epoch": 0.22489468677809485, + "grad_norm": 0.5982909202575684, + "learning_rate": 0.0009718881641527381, + "loss": 3.4792, + "step": 3310 + }, + { + "epoch": 0.22523440684875662, + "grad_norm": 0.7409027218818665, + "learning_rate": 0.0009718456991439055, + "loss": 3.6163, + "step": 3315 + }, + { + "epoch": 0.2255741269194184, + "grad_norm": 0.6308236718177795, + "learning_rate": 0.0009718032341350728, + "loss": 3.6235, + "step": 3320 + }, + { + "epoch": 0.22591384699008019, + "grad_norm": 0.7042643427848816, + "learning_rate": 0.00097176076912624, + "loss": 3.6881, + "step": 3325 + }, + { + "epoch": 0.22625356706074196, + "grad_norm": 0.593842625617981, + "learning_rate": 0.0009717183041174073, + "loss": 3.919, + "step": 3330 + }, + { + "epoch": 0.22659328713140373, + "grad_norm": 0.5943030714988708, + "learning_rate": 0.0009716758391085746, + "loss": 3.6208, + "step": 3335 + }, + { + "epoch": 0.2269330072020655, + "grad_norm": 1.3144179582595825, + "learning_rate": 0.0009716333740997418, + "loss": 3.4317, + "step": 3340 + }, + { + "epoch": 0.22727272727272727, + "grad_norm": 0.6511430740356445, + "learning_rate": 0.000971590909090909, + "loss": 3.7321, + "step": 3345 + }, + { + "epoch": 0.22761244734338903, + "grad_norm": 0.6584237217903137, + "learning_rate": 0.0009715484440820765, + "loss": 3.803, + "step": 3350 + }, + { + "epoch": 0.22795216741405083, + "grad_norm": 0.6386221647262573, + "learning_rate": 0.0009715059790732437, + "loss": 3.6276, + "step": 3355 + }, + { + "epoch": 0.2282918874847126, + "grad_norm": 0.5515703558921814, + "learning_rate": 0.0009714635140644109, + "loss": 3.7117, + "step": 3360 + }, + { + "epoch": 0.22863160755537437, + "grad_norm": 0.6941877007484436, + "learning_rate": 0.0009714210490555783, + "loss": 3.8806, + "step": 3365 + }, + { + "epoch": 0.22897132762603614, + "grad_norm": 1.0481723546981812, + "learning_rate": 0.0009713785840467455, + "loss": 3.771, + "step": 3370 + }, + { + "epoch": 0.2293110476966979, + "grad_norm": 0.9220503568649292, + "learning_rate": 0.0009713361190379127, + "loss": 3.7536, + "step": 3375 + }, + { + "epoch": 0.2296507677673597, + "grad_norm": 0.7632156014442444, + "learning_rate": 0.00097129365402908, + "loss": 3.6445, + "step": 3380 + }, + { + "epoch": 0.22999048783802148, + "grad_norm": 0.7810511589050293, + "learning_rate": 0.0009712511890202474, + "loss": 3.8208, + "step": 3385 + }, + { + "epoch": 0.23033020790868325, + "grad_norm": 0.5559766888618469, + "learning_rate": 0.0009712087240114146, + "loss": 3.6546, + "step": 3390 + }, + { + "epoch": 0.23066992797934502, + "grad_norm": 0.635195791721344, + "learning_rate": 0.000971166259002582, + "loss": 3.8679, + "step": 3395 + }, + { + "epoch": 0.2310096480500068, + "grad_norm": 0.7121466994285583, + "learning_rate": 0.0009711237939937492, + "loss": 3.565, + "step": 3400 + }, + { + "epoch": 0.23134936812066856, + "grad_norm": 2.192315101623535, + "learning_rate": 0.0009710813289849164, + "loss": 3.8571, + "step": 3405 + }, + { + "epoch": 0.23168908819133036, + "grad_norm": 0.643825352191925, + "learning_rate": 0.0009710388639760837, + "loss": 3.7185, + "step": 3410 + }, + { + "epoch": 0.23202880826199213, + "grad_norm": 0.7523700594902039, + "learning_rate": 0.000970996398967251, + "loss": 3.798, + "step": 3415 + }, + { + "epoch": 0.2323685283326539, + "grad_norm": 0.7138873934745789, + "learning_rate": 0.0009709539339584183, + "loss": 3.8475, + "step": 3420 + }, + { + "epoch": 0.23270824840331567, + "grad_norm": 0.7480379939079285, + "learning_rate": 0.0009709114689495856, + "loss": 3.8934, + "step": 3425 + }, + { + "epoch": 0.23304796847397743, + "grad_norm": 0.681869387626648, + "learning_rate": 0.0009708690039407529, + "loss": 3.8174, + "step": 3430 + }, + { + "epoch": 0.2333876885446392, + "grad_norm": 0.5060685873031616, + "learning_rate": 0.0009708265389319201, + "loss": 3.8649, + "step": 3435 + }, + { + "epoch": 0.233727408615301, + "grad_norm": 0.7572314143180847, + "learning_rate": 0.0009707840739230874, + "loss": 3.5466, + "step": 3440 + }, + { + "epoch": 0.23406712868596277, + "grad_norm": 1.0155407190322876, + "learning_rate": 0.0009707416089142546, + "loss": 3.6435, + "step": 3445 + }, + { + "epoch": 0.23440684875662454, + "grad_norm": 0.7484281063079834, + "learning_rate": 0.0009706991439054219, + "loss": 3.718, + "step": 3450 + }, + { + "epoch": 0.2347465688272863, + "grad_norm": 0.711729109287262, + "learning_rate": 0.0009706566788965893, + "loss": 3.749, + "step": 3455 + }, + { + "epoch": 0.23508628889794808, + "grad_norm": 0.887844979763031, + "learning_rate": 0.0009706142138877565, + "loss": 3.6235, + "step": 3460 + }, + { + "epoch": 0.23542600896860988, + "grad_norm": 0.5965444445610046, + "learning_rate": 0.0009705717488789238, + "loss": 3.8136, + "step": 3465 + }, + { + "epoch": 0.23576572903927165, + "grad_norm": 0.7257551550865173, + "learning_rate": 0.0009705292838700911, + "loss": 3.5925, + "step": 3470 + }, + { + "epoch": 0.23610544910993342, + "grad_norm": 0.7158923149108887, + "learning_rate": 0.0009704868188612583, + "loss": 3.6457, + "step": 3475 + }, + { + "epoch": 0.2364451691805952, + "grad_norm": 0.8055842518806458, + "learning_rate": 0.0009704443538524255, + "loss": 3.8343, + "step": 3480 + }, + { + "epoch": 0.23678488925125696, + "grad_norm": 0.6616836786270142, + "learning_rate": 0.0009704018888435929, + "loss": 3.808, + "step": 3485 + }, + { + "epoch": 0.23712460932191873, + "grad_norm": 0.6936202049255371, + "learning_rate": 0.0009703594238347602, + "loss": 3.4899, + "step": 3490 + }, + { + "epoch": 0.23746432939258053, + "grad_norm": 0.604406476020813, + "learning_rate": 0.0009703169588259274, + "loss": 3.6836, + "step": 3495 + }, + { + "epoch": 0.2378040494632423, + "grad_norm": 0.740138053894043, + "learning_rate": 0.0009702744938170948, + "loss": 3.8944, + "step": 3500 + }, + { + "epoch": 0.23814376953390406, + "grad_norm": 0.7477846145629883, + "learning_rate": 0.000970232028808262, + "loss": 3.84, + "step": 3505 + }, + { + "epoch": 0.23848348960456583, + "grad_norm": 0.8168776631355286, + "learning_rate": 0.0009701895637994292, + "loss": 3.6887, + "step": 3510 + }, + { + "epoch": 0.2388232096752276, + "grad_norm": 0.6762919425964355, + "learning_rate": 0.0009701470987905966, + "loss": 3.7436, + "step": 3515 + }, + { + "epoch": 0.23916292974588937, + "grad_norm": 0.6656241416931152, + "learning_rate": 0.0009701046337817638, + "loss": 3.6114, + "step": 3520 + }, + { + "epoch": 0.23950264981655117, + "grad_norm": 0.6061438918113708, + "learning_rate": 0.0009700621687729311, + "loss": 3.8649, + "step": 3525 + }, + { + "epoch": 0.23984236988721294, + "grad_norm": 1.1483242511749268, + "learning_rate": 0.0009700197037640985, + "loss": 3.7656, + "step": 3530 + }, + { + "epoch": 0.2401820899578747, + "grad_norm": 0.6067122220993042, + "learning_rate": 0.0009699772387552657, + "loss": 3.4087, + "step": 3535 + }, + { + "epoch": 0.24052181002853648, + "grad_norm": 0.6878045797348022, + "learning_rate": 0.0009699347737464329, + "loss": 3.7125, + "step": 3540 + }, + { + "epoch": 0.24086153009919825, + "grad_norm": 0.8297713994979858, + "learning_rate": 0.0009698923087376002, + "loss": 3.8188, + "step": 3545 + }, + { + "epoch": 0.24120125016986005, + "grad_norm": 0.7606715559959412, + "learning_rate": 0.0009698498437287675, + "loss": 3.9155, + "step": 3550 + }, + { + "epoch": 0.24154097024052182, + "grad_norm": 0.5454121232032776, + "learning_rate": 0.0009698073787199347, + "loss": 3.8156, + "step": 3555 + }, + { + "epoch": 0.2418806903111836, + "grad_norm": 0.7684235572814941, + "learning_rate": 0.0009697649137111021, + "loss": 3.538, + "step": 3560 + }, + { + "epoch": 0.24222041038184536, + "grad_norm": 0.6121727228164673, + "learning_rate": 0.0009697224487022694, + "loss": 3.6914, + "step": 3565 + }, + { + "epoch": 0.24256013045250713, + "grad_norm": 0.7470285892486572, + "learning_rate": 0.0009696799836934366, + "loss": 3.5258, + "step": 3570 + }, + { + "epoch": 0.2428998505231689, + "grad_norm": 0.6992287039756775, + "learning_rate": 0.0009696375186846039, + "loss": 3.8434, + "step": 3575 + }, + { + "epoch": 0.2432395705938307, + "grad_norm": 0.7074785232543945, + "learning_rate": 0.0009695950536757711, + "loss": 3.9296, + "step": 3580 + }, + { + "epoch": 0.24357929066449246, + "grad_norm": 0.5813422799110413, + "learning_rate": 0.0009695525886669385, + "loss": 3.8766, + "step": 3585 + }, + { + "epoch": 0.24391901073515423, + "grad_norm": 2.3182342052459717, + "learning_rate": 0.0009695101236581057, + "loss": 3.605, + "step": 3590 + }, + { + "epoch": 0.244258730805816, + "grad_norm": 1.234378695487976, + "learning_rate": 0.000969467658649273, + "loss": 3.6938, + "step": 3595 + }, + { + "epoch": 0.24459845087647777, + "grad_norm": 0.6325404047966003, + "learning_rate": 0.0009694251936404404, + "loss": 3.694, + "step": 3600 + }, + { + "epoch": 0.24493817094713954, + "grad_norm": 0.8209719061851501, + "learning_rate": 0.0009693827286316076, + "loss": 3.6918, + "step": 3605 + }, + { + "epoch": 0.24527789101780134, + "grad_norm": 1.4580378532409668, + "learning_rate": 0.0009693402636227748, + "loss": 3.7744, + "step": 3610 + }, + { + "epoch": 0.2456176110884631, + "grad_norm": 0.6046342253684998, + "learning_rate": 0.0009692977986139422, + "loss": 3.8888, + "step": 3615 + }, + { + "epoch": 0.24595733115912488, + "grad_norm": 0.599679708480835, + "learning_rate": 0.0009692553336051094, + "loss": 3.576, + "step": 3620 + }, + { + "epoch": 0.24629705122978665, + "grad_norm": 0.8095775842666626, + "learning_rate": 0.0009692128685962766, + "loss": 3.7028, + "step": 3625 + }, + { + "epoch": 0.24663677130044842, + "grad_norm": 0.6189206838607788, + "learning_rate": 0.0009691704035874441, + "loss": 3.596, + "step": 3630 + }, + { + "epoch": 0.24697649137111022, + "grad_norm": 0.7269278168678284, + "learning_rate": 0.0009691279385786113, + "loss": 3.794, + "step": 3635 + }, + { + "epoch": 0.247316211441772, + "grad_norm": 0.6310652494430542, + "learning_rate": 0.0009690854735697785, + "loss": 3.81, + "step": 3640 + }, + { + "epoch": 0.24765593151243376, + "grad_norm": 0.7167465090751648, + "learning_rate": 0.0009690430085609458, + "loss": 3.9076, + "step": 3645 + }, + { + "epoch": 0.24799565158309553, + "grad_norm": 0.5810459852218628, + "learning_rate": 0.0009690005435521131, + "loss": 3.6931, + "step": 3650 + }, + { + "epoch": 0.2483353716537573, + "grad_norm": 0.730108380317688, + "learning_rate": 0.0009689580785432803, + "loss": 3.5263, + "step": 3655 + }, + { + "epoch": 0.24867509172441907, + "grad_norm": 0.7339727282524109, + "learning_rate": 0.0009689156135344476, + "loss": 3.6601, + "step": 3660 + }, + { + "epoch": 0.24901481179508086, + "grad_norm": 1.1305773258209229, + "learning_rate": 0.000968873148525615, + "loss": 3.6521, + "step": 3665 + }, + { + "epoch": 0.24935453186574263, + "grad_norm": 1.4087789058685303, + "learning_rate": 0.0009688306835167822, + "loss": 3.7739, + "step": 3670 + }, + { + "epoch": 0.2496942519364044, + "grad_norm": 0.9541872143745422, + "learning_rate": 0.0009687882185079495, + "loss": 3.8882, + "step": 3675 + }, + { + "epoch": 0.2500339720070662, + "grad_norm": 0.6053314805030823, + "learning_rate": 0.0009687457534991167, + "loss": 3.7305, + "step": 3680 + }, + { + "epoch": 0.25037369207772797, + "grad_norm": 0.7858479619026184, + "learning_rate": 0.000968703288490284, + "loss": 3.6703, + "step": 3685 + }, + { + "epoch": 0.2507134121483897, + "grad_norm": 0.5707651972770691, + "learning_rate": 0.0009686608234814513, + "loss": 3.5633, + "step": 3690 + }, + { + "epoch": 0.2510531322190515, + "grad_norm": 0.6128956079483032, + "learning_rate": 0.0009686183584726185, + "loss": 3.8556, + "step": 3695 + }, + { + "epoch": 0.25139285228971325, + "grad_norm": 0.753255307674408, + "learning_rate": 0.0009685758934637859, + "loss": 3.8814, + "step": 3700 + }, + { + "epoch": 0.25173257236037505, + "grad_norm": 0.6241249442100525, + "learning_rate": 0.0009685334284549532, + "loss": 3.824, + "step": 3705 + }, + { + "epoch": 0.25207229243103685, + "grad_norm": 0.7240098118782043, + "learning_rate": 0.0009684909634461204, + "loss": 3.6703, + "step": 3710 + }, + { + "epoch": 0.2524120125016986, + "grad_norm": 0.8464183807373047, + "learning_rate": 0.0009684484984372877, + "loss": 3.6528, + "step": 3715 + }, + { + "epoch": 0.2527517325723604, + "grad_norm": 0.5208722352981567, + "learning_rate": 0.000968406033428455, + "loss": 3.6292, + "step": 3720 + }, + { + "epoch": 0.25309145264302213, + "grad_norm": 0.5412998795509338, + "learning_rate": 0.0009683635684196222, + "loss": 3.775, + "step": 3725 + }, + { + "epoch": 0.2534311727136839, + "grad_norm": 0.709400475025177, + "learning_rate": 0.0009683211034107894, + "loss": 3.9485, + "step": 3730 + }, + { + "epoch": 0.2537708927843457, + "grad_norm": 0.7480413913726807, + "learning_rate": 0.0009682786384019569, + "loss": 3.6651, + "step": 3735 + }, + { + "epoch": 0.25411061285500747, + "grad_norm": 0.6629950404167175, + "learning_rate": 0.0009682361733931241, + "loss": 3.7367, + "step": 3740 + }, + { + "epoch": 0.25445033292566926, + "grad_norm": 0.7563631534576416, + "learning_rate": 0.0009681937083842913, + "loss": 3.5439, + "step": 3745 + }, + { + "epoch": 0.254790052996331, + "grad_norm": 0.5678558945655823, + "learning_rate": 0.0009681512433754587, + "loss": 3.5644, + "step": 3750 + }, + { + "epoch": 0.2551297730669928, + "grad_norm": 1.0968127250671387, + "learning_rate": 0.0009681087783666259, + "loss": 3.9246, + "step": 3755 + }, + { + "epoch": 0.25546949313765455, + "grad_norm": 0.7039735913276672, + "learning_rate": 0.0009680663133577931, + "loss": 3.6756, + "step": 3760 + }, + { + "epoch": 0.25580921320831634, + "grad_norm": 3.6075406074523926, + "learning_rate": 0.0009680238483489606, + "loss": 3.7122, + "step": 3765 + }, + { + "epoch": 0.25614893327897814, + "grad_norm": 0.7793748378753662, + "learning_rate": 0.0009679813833401278, + "loss": 3.7305, + "step": 3770 + }, + { + "epoch": 0.2564886533496399, + "grad_norm": 0.6226781606674194, + "learning_rate": 0.000967938918331295, + "loss": 3.6611, + "step": 3775 + }, + { + "epoch": 0.2568283734203017, + "grad_norm": 0.9450697302818298, + "learning_rate": 0.0009678964533224623, + "loss": 3.8694, + "step": 3780 + }, + { + "epoch": 0.2571680934909634, + "grad_norm": 0.6313878297805786, + "learning_rate": 0.0009678539883136296, + "loss": 3.7633, + "step": 3785 + }, + { + "epoch": 0.2575078135616252, + "grad_norm": 0.5901970267295837, + "learning_rate": 0.0009678115233047968, + "loss": 3.7798, + "step": 3790 + }, + { + "epoch": 0.257847533632287, + "grad_norm": 0.7463569641113281, + "learning_rate": 0.0009677690582959641, + "loss": 3.7769, + "step": 3795 + }, + { + "epoch": 0.25818725370294876, + "grad_norm": 0.6924563050270081, + "learning_rate": 0.0009677265932871315, + "loss": 3.6305, + "step": 3800 + }, + { + "epoch": 0.25852697377361056, + "grad_norm": 0.8145256638526917, + "learning_rate": 0.0009676841282782987, + "loss": 3.6811, + "step": 3805 + }, + { + "epoch": 0.2588666938442723, + "grad_norm": 0.7224252820014954, + "learning_rate": 0.000967641663269466, + "loss": 3.7228, + "step": 3810 + }, + { + "epoch": 0.2592064139149341, + "grad_norm": 0.6247282028198242, + "learning_rate": 0.0009675991982606333, + "loss": 3.7572, + "step": 3815 + }, + { + "epoch": 0.2595461339855959, + "grad_norm": 0.6958258152008057, + "learning_rate": 0.0009675567332518005, + "loss": 3.5988, + "step": 3820 + }, + { + "epoch": 0.25988585405625764, + "grad_norm": 0.6718140244483948, + "learning_rate": 0.0009675142682429678, + "loss": 3.7994, + "step": 3825 + }, + { + "epoch": 0.26022557412691943, + "grad_norm": 0.8125585317611694, + "learning_rate": 0.000967471803234135, + "loss": 3.4968, + "step": 3830 + }, + { + "epoch": 0.2605652941975812, + "grad_norm": 0.6540567874908447, + "learning_rate": 0.0009674293382253024, + "loss": 3.5676, + "step": 3835 + }, + { + "epoch": 0.260905014268243, + "grad_norm": 0.5872313976287842, + "learning_rate": 0.0009673868732164697, + "loss": 3.6956, + "step": 3840 + }, + { + "epoch": 0.2612447343389047, + "grad_norm": 0.6185137629508972, + "learning_rate": 0.0009673444082076369, + "loss": 3.8481, + "step": 3845 + }, + { + "epoch": 0.2615844544095665, + "grad_norm": 0.7442803978919983, + "learning_rate": 0.0009673019431988042, + "loss": 3.8018, + "step": 3850 + }, + { + "epoch": 0.2619241744802283, + "grad_norm": 0.702149510383606, + "learning_rate": 0.0009672594781899715, + "loss": 3.5721, + "step": 3855 + }, + { + "epoch": 0.26226389455089005, + "grad_norm": 0.7930136919021606, + "learning_rate": 0.0009672170131811387, + "loss": 3.8332, + "step": 3860 + }, + { + "epoch": 0.26260361462155185, + "grad_norm": 0.6575914025306702, + "learning_rate": 0.000967174548172306, + "loss": 3.6584, + "step": 3865 + }, + { + "epoch": 0.2629433346922136, + "grad_norm": 0.6223536729812622, + "learning_rate": 0.0009671320831634734, + "loss": 3.4476, + "step": 3870 + }, + { + "epoch": 0.2632830547628754, + "grad_norm": 0.6894667148590088, + "learning_rate": 0.0009670896181546406, + "loss": 3.7942, + "step": 3875 + }, + { + "epoch": 0.2636227748335372, + "grad_norm": 0.5753107070922852, + "learning_rate": 0.0009670471531458078, + "loss": 3.8189, + "step": 3880 + }, + { + "epoch": 0.26396249490419893, + "grad_norm": 0.7406034469604492, + "learning_rate": 0.0009670046881369752, + "loss": 3.6759, + "step": 3885 + }, + { + "epoch": 0.2643022149748607, + "grad_norm": 0.5585180521011353, + "learning_rate": 0.0009669622231281424, + "loss": 3.7618, + "step": 3890 + }, + { + "epoch": 0.26464193504552247, + "grad_norm": 0.7381715774536133, + "learning_rate": 0.0009669197581193096, + "loss": 3.9587, + "step": 3895 + }, + { + "epoch": 0.26498165511618427, + "grad_norm": 0.7363392114639282, + "learning_rate": 0.000966877293110477, + "loss": 3.4923, + "step": 3900 + }, + { + "epoch": 0.26532137518684606, + "grad_norm": 0.9010974764823914, + "learning_rate": 0.0009668348281016443, + "loss": 4.0328, + "step": 3905 + }, + { + "epoch": 0.2656610952575078, + "grad_norm": 0.7546082735061646, + "learning_rate": 0.0009667923630928115, + "loss": 3.6062, + "step": 3910 + }, + { + "epoch": 0.2660008153281696, + "grad_norm": 0.6980230808258057, + "learning_rate": 0.0009667498980839789, + "loss": 3.9225, + "step": 3915 + }, + { + "epoch": 0.26634053539883135, + "grad_norm": 0.7920522689819336, + "learning_rate": 0.0009667074330751461, + "loss": 3.9315, + "step": 3920 + }, + { + "epoch": 0.26668025546949314, + "grad_norm": 0.7503193020820618, + "learning_rate": 0.0009666649680663134, + "loss": 3.7232, + "step": 3925 + }, + { + "epoch": 0.2670199755401549, + "grad_norm": 0.5713961124420166, + "learning_rate": 0.0009666225030574806, + "loss": 3.4191, + "step": 3930 + }, + { + "epoch": 0.2673596956108167, + "grad_norm": 0.6793795228004456, + "learning_rate": 0.0009665800380486479, + "loss": 3.8886, + "step": 3935 + }, + { + "epoch": 0.2676994156814785, + "grad_norm": 0.5702452659606934, + "learning_rate": 0.0009665375730398153, + "loss": 3.8435, + "step": 3940 + }, + { + "epoch": 0.2680391357521402, + "grad_norm": 0.7704566121101379, + "learning_rate": 0.0009664951080309825, + "loss": 3.5918, + "step": 3945 + }, + { + "epoch": 0.268378855822802, + "grad_norm": 0.6938804984092712, + "learning_rate": 0.0009664526430221498, + "loss": 4.0113, + "step": 3950 + }, + { + "epoch": 0.26871857589346376, + "grad_norm": 0.74507737159729, + "learning_rate": 0.0009664101780133171, + "loss": 3.5163, + "step": 3955 + }, + { + "epoch": 0.26905829596412556, + "grad_norm": 0.5978330969810486, + "learning_rate": 0.0009663677130044843, + "loss": 3.8, + "step": 3960 + }, + { + "epoch": 0.26939801603478736, + "grad_norm": 0.9157208204269409, + "learning_rate": 0.0009663252479956515, + "loss": 3.9161, + "step": 3965 + }, + { + "epoch": 0.2697377361054491, + "grad_norm": 0.7541576623916626, + "learning_rate": 0.0009662827829868189, + "loss": 3.5579, + "step": 3970 + }, + { + "epoch": 0.2700774561761109, + "grad_norm": 0.6681669354438782, + "learning_rate": 0.0009662403179779862, + "loss": 3.9983, + "step": 3975 + }, + { + "epoch": 0.27041717624677264, + "grad_norm": 0.6837373971939087, + "learning_rate": 0.0009661978529691534, + "loss": 3.7978, + "step": 3980 + }, + { + "epoch": 0.27075689631743444, + "grad_norm": 0.7061611413955688, + "learning_rate": 0.0009661553879603208, + "loss": 3.9587, + "step": 3985 + }, + { + "epoch": 0.27109661638809623, + "grad_norm": 0.6160033345222473, + "learning_rate": 0.000966112922951488, + "loss": 3.7788, + "step": 3990 + }, + { + "epoch": 0.271436336458758, + "grad_norm": 0.7030746340751648, + "learning_rate": 0.0009660704579426552, + "loss": 3.6304, + "step": 3995 + }, + { + "epoch": 0.2717760565294198, + "grad_norm": 0.7516574263572693, + "learning_rate": 0.0009660279929338226, + "loss": 3.9709, + "step": 4000 + }, + { + "epoch": 0.2721157766000815, + "grad_norm": 0.5817821621894836, + "learning_rate": 0.0009659855279249898, + "loss": 3.6713, + "step": 4005 + }, + { + "epoch": 0.2724554966707433, + "grad_norm": 0.5100502967834473, + "learning_rate": 0.0009659430629161571, + "loss": 3.6554, + "step": 4010 + }, + { + "epoch": 0.27279521674140506, + "grad_norm": 0.7973607182502747, + "learning_rate": 0.0009659005979073245, + "loss": 3.9906, + "step": 4015 + }, + { + "epoch": 0.27313493681206685, + "grad_norm": 0.6986095905303955, + "learning_rate": 0.0009658581328984917, + "loss": 3.5992, + "step": 4020 + }, + { + "epoch": 0.27347465688272865, + "grad_norm": 0.7145535945892334, + "learning_rate": 0.0009658156678896589, + "loss": 3.7215, + "step": 4025 + }, + { + "epoch": 0.2738143769533904, + "grad_norm": 0.644110381603241, + "learning_rate": 0.0009657732028808262, + "loss": 3.4877, + "step": 4030 + }, + { + "epoch": 0.2741540970240522, + "grad_norm": 0.6192665696144104, + "learning_rate": 0.0009657307378719935, + "loss": 3.7874, + "step": 4035 + }, + { + "epoch": 0.27449381709471393, + "grad_norm": 0.5692789554595947, + "learning_rate": 0.0009656882728631607, + "loss": 3.6267, + "step": 4040 + }, + { + "epoch": 0.27483353716537573, + "grad_norm": 0.6869579553604126, + "learning_rate": 0.0009656458078543281, + "loss": 4.0107, + "step": 4045 + }, + { + "epoch": 0.2751732572360375, + "grad_norm": 0.7522791624069214, + "learning_rate": 0.0009656033428454954, + "loss": 3.8558, + "step": 4050 + }, + { + "epoch": 0.27551297730669927, + "grad_norm": 0.5839440822601318, + "learning_rate": 0.0009655608778366626, + "loss": 3.7982, + "step": 4055 + }, + { + "epoch": 0.27585269737736107, + "grad_norm": 0.6837066411972046, + "learning_rate": 0.0009655184128278299, + "loss": 3.7102, + "step": 4060 + }, + { + "epoch": 0.2761924174480228, + "grad_norm": 0.7138178944587708, + "learning_rate": 0.0009654759478189972, + "loss": 3.622, + "step": 4065 + }, + { + "epoch": 0.2765321375186846, + "grad_norm": 0.6989341378211975, + "learning_rate": 0.0009654334828101644, + "loss": 3.5083, + "step": 4070 + }, + { + "epoch": 0.2768718575893464, + "grad_norm": 0.6958532929420471, + "learning_rate": 0.0009653910178013317, + "loss": 3.8488, + "step": 4075 + }, + { + "epoch": 0.27721157766000815, + "grad_norm": 0.5760381817817688, + "learning_rate": 0.000965348552792499, + "loss": 3.731, + "step": 4080 + }, + { + "epoch": 0.27755129773066994, + "grad_norm": 0.6913160085678101, + "learning_rate": 0.0009653060877836663, + "loss": 3.4657, + "step": 4085 + }, + { + "epoch": 0.2778910178013317, + "grad_norm": 0.6547573208808899, + "learning_rate": 0.0009652636227748336, + "loss": 3.6995, + "step": 4090 + }, + { + "epoch": 0.2782307378719935, + "grad_norm": 0.7820612192153931, + "learning_rate": 0.0009652211577660008, + "loss": 3.7452, + "step": 4095 + }, + { + "epoch": 0.2785704579426552, + "grad_norm": 0.6629460453987122, + "learning_rate": 0.0009651786927571681, + "loss": 3.7866, + "step": 4100 + }, + { + "epoch": 0.278910178013317, + "grad_norm": 0.6040059924125671, + "learning_rate": 0.0009651362277483354, + "loss": 3.5589, + "step": 4105 + }, + { + "epoch": 0.2792498980839788, + "grad_norm": 1.0010850429534912, + "learning_rate": 0.0009650937627395026, + "loss": 3.802, + "step": 4110 + }, + { + "epoch": 0.27958961815464056, + "grad_norm": 0.6360946893692017, + "learning_rate": 0.00096505129773067, + "loss": 3.7197, + "step": 4115 + }, + { + "epoch": 0.27992933822530236, + "grad_norm": 2.3139994144439697, + "learning_rate": 0.0009650088327218373, + "loss": 3.6895, + "step": 4120 + }, + { + "epoch": 0.2802690582959641, + "grad_norm": 0.8118367195129395, + "learning_rate": 0.0009649663677130045, + "loss": 3.5454, + "step": 4125 + }, + { + "epoch": 0.2806087783666259, + "grad_norm": 0.702793538570404, + "learning_rate": 0.0009649239027041717, + "loss": 3.6917, + "step": 4130 + }, + { + "epoch": 0.2809484984372877, + "grad_norm": 0.5384190678596497, + "learning_rate": 0.0009648814376953391, + "loss": 3.6009, + "step": 4135 + }, + { + "epoch": 0.28128821850794944, + "grad_norm": 0.6091647148132324, + "learning_rate": 0.0009648389726865063, + "loss": 3.6381, + "step": 4140 + }, + { + "epoch": 0.28162793857861124, + "grad_norm": 0.6185944080352783, + "learning_rate": 0.0009647965076776735, + "loss": 3.7377, + "step": 4145 + }, + { + "epoch": 0.281967658649273, + "grad_norm": 0.8071286678314209, + "learning_rate": 0.000964754042668841, + "loss": 3.9584, + "step": 4150 + }, + { + "epoch": 0.2823073787199348, + "grad_norm": 0.6260259747505188, + "learning_rate": 0.0009647115776600082, + "loss": 3.8657, + "step": 4155 + }, + { + "epoch": 0.2826470987905966, + "grad_norm": 0.7029274702072144, + "learning_rate": 0.0009646691126511754, + "loss": 3.8413, + "step": 4160 + }, + { + "epoch": 0.2829868188612583, + "grad_norm": 0.6549016833305359, + "learning_rate": 0.0009646266476423428, + "loss": 3.5741, + "step": 4165 + }, + { + "epoch": 0.2833265389319201, + "grad_norm": 0.6251670718193054, + "learning_rate": 0.00096458418263351, + "loss": 3.6524, + "step": 4170 + }, + { + "epoch": 0.28366625900258186, + "grad_norm": 0.6982675194740295, + "learning_rate": 0.0009645417176246772, + "loss": 3.7666, + "step": 4175 + }, + { + "epoch": 0.28400597907324365, + "grad_norm": 0.8168516159057617, + "learning_rate": 0.0009644992526158445, + "loss": 3.4167, + "step": 4180 + }, + { + "epoch": 0.2843456991439054, + "grad_norm": 0.6430438756942749, + "learning_rate": 0.0009644567876070119, + "loss": 3.8558, + "step": 4185 + }, + { + "epoch": 0.2846854192145672, + "grad_norm": 0.8260819911956787, + "learning_rate": 0.0009644143225981791, + "loss": 3.4549, + "step": 4190 + }, + { + "epoch": 0.285025139285229, + "grad_norm": 0.6831293702125549, + "learning_rate": 0.0009643718575893464, + "loss": 3.7823, + "step": 4195 + }, + { + "epoch": 0.28536485935589073, + "grad_norm": 0.6138635873794556, + "learning_rate": 0.0009643293925805137, + "loss": 3.7056, + "step": 4200 + }, + { + "epoch": 0.28570457942655253, + "grad_norm": 0.5763000249862671, + "learning_rate": 0.0009642869275716809, + "loss": 3.754, + "step": 4205 + }, + { + "epoch": 0.28604429949721427, + "grad_norm": 0.8776401877403259, + "learning_rate": 0.0009642444625628482, + "loss": 3.8041, + "step": 4210 + }, + { + "epoch": 0.28638401956787607, + "grad_norm": 1.109387993812561, + "learning_rate": 0.0009642019975540154, + "loss": 3.7585, + "step": 4215 + }, + { + "epoch": 0.28672373963853787, + "grad_norm": 0.6992509365081787, + "learning_rate": 0.0009641595325451828, + "loss": 3.9185, + "step": 4220 + }, + { + "epoch": 0.2870634597091996, + "grad_norm": 0.7073003053665161, + "learning_rate": 0.0009641170675363501, + "loss": 3.9487, + "step": 4225 + }, + { + "epoch": 0.2874031797798614, + "grad_norm": 0.6866264939308167, + "learning_rate": 0.0009640746025275173, + "loss": 3.8364, + "step": 4230 + }, + { + "epoch": 0.28774289985052315, + "grad_norm": 0.5466263890266418, + "learning_rate": 0.0009640321375186846, + "loss": 3.5596, + "step": 4235 + }, + { + "epoch": 0.28808261992118495, + "grad_norm": 0.6644207239151001, + "learning_rate": 0.0009639896725098519, + "loss": 3.8898, + "step": 4240 + }, + { + "epoch": 0.28842233999184674, + "grad_norm": 0.6112736463546753, + "learning_rate": 0.0009639472075010191, + "loss": 3.8798, + "step": 4245 + }, + { + "epoch": 0.2887620600625085, + "grad_norm": 0.7702374458312988, + "learning_rate": 0.0009639047424921864, + "loss": 3.6329, + "step": 4250 + }, + { + "epoch": 0.2891017801331703, + "grad_norm": 0.5818788409233093, + "learning_rate": 0.0009638622774833538, + "loss": 3.5738, + "step": 4255 + }, + { + "epoch": 0.289441500203832, + "grad_norm": 0.7824899554252625, + "learning_rate": 0.000963819812474521, + "loss": 3.9873, + "step": 4260 + }, + { + "epoch": 0.2897812202744938, + "grad_norm": 1.1394249200820923, + "learning_rate": 0.0009637773474656884, + "loss": 3.6349, + "step": 4265 + }, + { + "epoch": 0.29012094034515556, + "grad_norm": 0.6994426846504211, + "learning_rate": 0.0009637348824568556, + "loss": 3.7287, + "step": 4270 + }, + { + "epoch": 0.29046066041581736, + "grad_norm": 0.6484062671661377, + "learning_rate": 0.0009636924174480228, + "loss": 3.7825, + "step": 4275 + }, + { + "epoch": 0.29080038048647916, + "grad_norm": 1.0661653280258179, + "learning_rate": 0.0009636499524391901, + "loss": 3.6908, + "step": 4280 + }, + { + "epoch": 0.2911401005571409, + "grad_norm": 0.5649815201759338, + "learning_rate": 0.0009636074874303575, + "loss": 3.8531, + "step": 4285 + }, + { + "epoch": 0.2914798206278027, + "grad_norm": 0.7183372974395752, + "learning_rate": 0.0009635650224215247, + "loss": 3.5815, + "step": 4290 + }, + { + "epoch": 0.29181954069846444, + "grad_norm": 0.6840630173683167, + "learning_rate": 0.000963522557412692, + "loss": 3.955, + "step": 4295 + }, + { + "epoch": 0.29215926076912624, + "grad_norm": 0.6997886300086975, + "learning_rate": 0.0009634800924038593, + "loss": 3.6086, + "step": 4300 + }, + { + "epoch": 0.29249898083978804, + "grad_norm": 0.7394000291824341, + "learning_rate": 0.0009634376273950265, + "loss": 3.4845, + "step": 4305 + }, + { + "epoch": 0.2928387009104498, + "grad_norm": 0.6880769729614258, + "learning_rate": 0.0009633951623861938, + "loss": 3.8555, + "step": 4310 + }, + { + "epoch": 0.2931784209811116, + "grad_norm": 0.9616801142692566, + "learning_rate": 0.000963352697377361, + "loss": 3.6079, + "step": 4315 + }, + { + "epoch": 0.2935181410517733, + "grad_norm": 0.7349090576171875, + "learning_rate": 0.0009633102323685284, + "loss": 3.6728, + "step": 4320 + }, + { + "epoch": 0.2938578611224351, + "grad_norm": 0.6582932472229004, + "learning_rate": 0.0009632677673596957, + "loss": 3.5777, + "step": 4325 + }, + { + "epoch": 0.2941975811930969, + "grad_norm": 0.6528509855270386, + "learning_rate": 0.0009632253023508629, + "loss": 3.7702, + "step": 4330 + }, + { + "epoch": 0.29453730126375866, + "grad_norm": 0.6171696782112122, + "learning_rate": 0.0009631828373420302, + "loss": 3.8234, + "step": 4335 + }, + { + "epoch": 0.29487702133442045, + "grad_norm": 0.9008753299713135, + "learning_rate": 0.0009631403723331975, + "loss": 3.6927, + "step": 4340 + }, + { + "epoch": 0.2952167414050822, + "grad_norm": 0.6457626819610596, + "learning_rate": 0.0009630979073243647, + "loss": 3.2905, + "step": 4345 + }, + { + "epoch": 0.295556461475744, + "grad_norm": 0.7185888886451721, + "learning_rate": 0.000963055442315532, + "loss": 3.6635, + "step": 4350 + }, + { + "epoch": 0.29589618154640573, + "grad_norm": 0.7425805330276489, + "learning_rate": 0.0009630129773066994, + "loss": 3.7653, + "step": 4355 + }, + { + "epoch": 0.29623590161706753, + "grad_norm": 0.6147916913032532, + "learning_rate": 0.0009629705122978666, + "loss": 3.5351, + "step": 4360 + }, + { + "epoch": 0.29657562168772933, + "grad_norm": 0.6081828474998474, + "learning_rate": 0.0009629280472890338, + "loss": 3.8761, + "step": 4365 + }, + { + "epoch": 0.29691534175839107, + "grad_norm": 0.7161690592765808, + "learning_rate": 0.0009628855822802012, + "loss": 3.8387, + "step": 4370 + }, + { + "epoch": 0.29725506182905287, + "grad_norm": 0.7499025464057922, + "learning_rate": 0.0009628431172713684, + "loss": 3.5769, + "step": 4375 + }, + { + "epoch": 0.2975947818997146, + "grad_norm": 0.7925812602043152, + "learning_rate": 0.0009628006522625356, + "loss": 3.7181, + "step": 4380 + }, + { + "epoch": 0.2979345019703764, + "grad_norm": 0.6369455456733704, + "learning_rate": 0.000962758187253703, + "loss": 3.7408, + "step": 4385 + }, + { + "epoch": 0.2982742220410382, + "grad_norm": 0.7347651720046997, + "learning_rate": 0.0009627157222448703, + "loss": 3.9034, + "step": 4390 + }, + { + "epoch": 0.29861394211169995, + "grad_norm": 0.6870957016944885, + "learning_rate": 0.0009626732572360375, + "loss": 3.9102, + "step": 4395 + }, + { + "epoch": 0.29895366218236175, + "grad_norm": 0.717404305934906, + "learning_rate": 0.0009626307922272049, + "loss": 3.6934, + "step": 4400 + }, + { + "epoch": 0.2992933822530235, + "grad_norm": 0.7716221809387207, + "learning_rate": 0.0009625883272183721, + "loss": 3.6996, + "step": 4405 + }, + { + "epoch": 0.2996331023236853, + "grad_norm": 0.8365433216094971, + "learning_rate": 0.0009625458622095393, + "loss": 3.9056, + "step": 4410 + }, + { + "epoch": 0.2999728223943471, + "grad_norm": 0.7925680875778198, + "learning_rate": 0.0009625033972007066, + "loss": 3.9172, + "step": 4415 + }, + { + "epoch": 0.3003125424650088, + "grad_norm": 0.8062108159065247, + "learning_rate": 0.0009624609321918739, + "loss": 3.7353, + "step": 4420 + }, + { + "epoch": 0.3006522625356706, + "grad_norm": 0.5405961871147156, + "learning_rate": 0.0009624184671830412, + "loss": 3.8695, + "step": 4425 + }, + { + "epoch": 0.30099198260633236, + "grad_norm": 0.6823050379753113, + "learning_rate": 0.0009623760021742085, + "loss": 3.9127, + "step": 4430 + }, + { + "epoch": 0.30133170267699416, + "grad_norm": 0.9308080077171326, + "learning_rate": 0.0009623335371653758, + "loss": 3.9683, + "step": 4435 + }, + { + "epoch": 0.3016714227476559, + "grad_norm": 0.7651321291923523, + "learning_rate": 0.000962291072156543, + "loss": 3.6044, + "step": 4440 + }, + { + "epoch": 0.3020111428183177, + "grad_norm": 0.6551792025566101, + "learning_rate": 0.0009622486071477103, + "loss": 3.7825, + "step": 4445 + }, + { + "epoch": 0.3023508628889795, + "grad_norm": 0.6831545233726501, + "learning_rate": 0.0009622061421388776, + "loss": 3.6857, + "step": 4450 + }, + { + "epoch": 0.30269058295964124, + "grad_norm": 0.6022977828979492, + "learning_rate": 0.0009621636771300448, + "loss": 3.5084, + "step": 4455 + }, + { + "epoch": 0.30303030303030304, + "grad_norm": 0.7112560868263245, + "learning_rate": 0.0009621212121212122, + "loss": 3.5855, + "step": 4460 + }, + { + "epoch": 0.3033700231009648, + "grad_norm": 0.6755446791648865, + "learning_rate": 0.0009620787471123794, + "loss": 3.7325, + "step": 4465 + }, + { + "epoch": 0.3037097431716266, + "grad_norm": 1.4436894655227661, + "learning_rate": 0.0009620362821035467, + "loss": 3.5141, + "step": 4470 + }, + { + "epoch": 0.3040494632422884, + "grad_norm": 0.6700942516326904, + "learning_rate": 0.000961993817094714, + "loss": 3.7589, + "step": 4475 + }, + { + "epoch": 0.3043891833129501, + "grad_norm": 0.6899858713150024, + "learning_rate": 0.0009619513520858812, + "loss": 3.9016, + "step": 4480 + }, + { + "epoch": 0.3047289033836119, + "grad_norm": 0.6155400276184082, + "learning_rate": 0.0009619088870770485, + "loss": 3.592, + "step": 4485 + }, + { + "epoch": 0.30506862345427366, + "grad_norm": 0.578019917011261, + "learning_rate": 0.0009618664220682158, + "loss": 4.0385, + "step": 4490 + }, + { + "epoch": 0.30540834352493546, + "grad_norm": 1.3838962316513062, + "learning_rate": 0.0009618239570593831, + "loss": 3.7364, + "step": 4495 + }, + { + "epoch": 0.30574806359559725, + "grad_norm": 0.6865847706794739, + "learning_rate": 0.0009617814920505504, + "loss": 3.7875, + "step": 4500 + }, + { + "epoch": 0.306087783666259, + "grad_norm": 0.698574960231781, + "learning_rate": 0.0009617390270417177, + "loss": 3.5852, + "step": 4505 + }, + { + "epoch": 0.3064275037369208, + "grad_norm": 0.6817865371704102, + "learning_rate": 0.0009616965620328849, + "loss": 4.033, + "step": 4510 + }, + { + "epoch": 0.30676722380758253, + "grad_norm": 0.7040066719055176, + "learning_rate": 0.0009616540970240521, + "loss": 3.8144, + "step": 4515 + }, + { + "epoch": 0.30710694387824433, + "grad_norm": 0.6947503685951233, + "learning_rate": 0.0009616116320152195, + "loss": 3.7897, + "step": 4520 + }, + { + "epoch": 0.3074466639489061, + "grad_norm": 0.7446925640106201, + "learning_rate": 0.0009615691670063867, + "loss": 3.5158, + "step": 4525 + }, + { + "epoch": 0.30778638401956787, + "grad_norm": 0.9600127339363098, + "learning_rate": 0.000961526701997554, + "loss": 3.8045, + "step": 4530 + }, + { + "epoch": 0.30812610409022967, + "grad_norm": 0.7573233246803284, + "learning_rate": 0.0009614842369887214, + "loss": 3.5216, + "step": 4535 + }, + { + "epoch": 0.3084658241608914, + "grad_norm": 0.7382401823997498, + "learning_rate": 0.0009614417719798886, + "loss": 3.6634, + "step": 4540 + }, + { + "epoch": 0.3088055442315532, + "grad_norm": 0.9753466844558716, + "learning_rate": 0.0009613993069710558, + "loss": 3.4437, + "step": 4545 + }, + { + "epoch": 0.30914526430221495, + "grad_norm": 0.6215182542800903, + "learning_rate": 0.0009613568419622232, + "loss": 3.5727, + "step": 4550 + }, + { + "epoch": 0.30948498437287675, + "grad_norm": 0.7607511878013611, + "learning_rate": 0.0009613143769533904, + "loss": 3.8892, + "step": 4555 + }, + { + "epoch": 0.30982470444353855, + "grad_norm": 0.5921122431755066, + "learning_rate": 0.0009612719119445576, + "loss": 3.7165, + "step": 4560 + }, + { + "epoch": 0.3101644245142003, + "grad_norm": 0.8299614191055298, + "learning_rate": 0.000961229446935725, + "loss": 3.8243, + "step": 4565 + }, + { + "epoch": 0.3105041445848621, + "grad_norm": 1.279056429862976, + "learning_rate": 0.0009611869819268923, + "loss": 3.8217, + "step": 4570 + }, + { + "epoch": 0.3108438646555238, + "grad_norm": 0.7296881675720215, + "learning_rate": 0.0009611445169180595, + "loss": 3.5245, + "step": 4575 + }, + { + "epoch": 0.3111835847261856, + "grad_norm": 0.9374946355819702, + "learning_rate": 0.0009611020519092268, + "loss": 4.0029, + "step": 4580 + }, + { + "epoch": 0.3115233047968474, + "grad_norm": 0.6455745100975037, + "learning_rate": 0.0009610595869003941, + "loss": 3.7578, + "step": 4585 + }, + { + "epoch": 0.31186302486750916, + "grad_norm": 0.7967641353607178, + "learning_rate": 0.0009610171218915613, + "loss": 3.7079, + "step": 4590 + }, + { + "epoch": 0.31220274493817096, + "grad_norm": 0.6406934857368469, + "learning_rate": 0.0009609746568827286, + "loss": 3.7325, + "step": 4595 + }, + { + "epoch": 0.3125424650088327, + "grad_norm": 0.6590662002563477, + "learning_rate": 0.000960932191873896, + "loss": 3.8455, + "step": 4600 + }, + { + "epoch": 0.3128821850794945, + "grad_norm": 0.584558367729187, + "learning_rate": 0.0009608897268650633, + "loss": 3.7943, + "step": 4605 + }, + { + "epoch": 0.3132219051501563, + "grad_norm": 0.7745699882507324, + "learning_rate": 0.0009608472618562305, + "loss": 3.6923, + "step": 4610 + }, + { + "epoch": 0.31356162522081804, + "grad_norm": 0.5839930772781372, + "learning_rate": 0.0009608047968473977, + "loss": 3.7063, + "step": 4615 + }, + { + "epoch": 0.31390134529147984, + "grad_norm": 0.6579026579856873, + "learning_rate": 0.0009607623318385651, + "loss": 3.9349, + "step": 4620 + }, + { + "epoch": 0.3142410653621416, + "grad_norm": 0.7334092855453491, + "learning_rate": 0.0009607198668297323, + "loss": 3.7986, + "step": 4625 + }, + { + "epoch": 0.3145807854328034, + "grad_norm": 0.5687442421913147, + "learning_rate": 0.0009606774018208995, + "loss": 3.5316, + "step": 4630 + }, + { + "epoch": 0.3149205055034651, + "grad_norm": 0.8066163063049316, + "learning_rate": 0.000960634936812067, + "loss": 3.6936, + "step": 4635 + }, + { + "epoch": 0.3152602255741269, + "grad_norm": 0.6861526370048523, + "learning_rate": 0.0009605924718032342, + "loss": 3.8663, + "step": 4640 + }, + { + "epoch": 0.3155999456447887, + "grad_norm": 0.6165646910667419, + "learning_rate": 0.0009605500067944014, + "loss": 3.7368, + "step": 4645 + }, + { + "epoch": 0.31593966571545046, + "grad_norm": 0.6109422445297241, + "learning_rate": 0.0009605075417855688, + "loss": 3.6905, + "step": 4650 + }, + { + "epoch": 0.31627938578611225, + "grad_norm": 0.8834192156791687, + "learning_rate": 0.000960465076776736, + "loss": 3.604, + "step": 4655 + }, + { + "epoch": 0.316619105856774, + "grad_norm": 0.5666911602020264, + "learning_rate": 0.0009604226117679032, + "loss": 3.9428, + "step": 4660 + }, + { + "epoch": 0.3169588259274358, + "grad_norm": 0.922512412071228, + "learning_rate": 0.0009603801467590705, + "loss": 3.7212, + "step": 4665 + }, + { + "epoch": 0.3172985459980976, + "grad_norm": 0.6613696813583374, + "learning_rate": 0.0009603376817502379, + "loss": 3.9531, + "step": 4670 + }, + { + "epoch": 0.31763826606875933, + "grad_norm": 0.6953523755073547, + "learning_rate": 0.0009602952167414051, + "loss": 3.8304, + "step": 4675 + }, + { + "epoch": 0.31797798613942113, + "grad_norm": 0.6979256868362427, + "learning_rate": 0.0009602527517325724, + "loss": 3.6581, + "step": 4680 + }, + { + "epoch": 0.3183177062100829, + "grad_norm": 0.5268025994300842, + "learning_rate": 0.0009602102867237397, + "loss": 3.8971, + "step": 4685 + }, + { + "epoch": 0.31865742628074467, + "grad_norm": 0.6296973824501038, + "learning_rate": 0.0009601678217149069, + "loss": 3.8694, + "step": 4690 + }, + { + "epoch": 0.31899714635140647, + "grad_norm": 0.6151579022407532, + "learning_rate": 0.0009601253567060742, + "loss": 3.8435, + "step": 4695 + }, + { + "epoch": 0.3193368664220682, + "grad_norm": 0.6179822087287903, + "learning_rate": 0.0009600828916972415, + "loss": 3.7191, + "step": 4700 + }, + { + "epoch": 0.31967658649273, + "grad_norm": 0.5826281309127808, + "learning_rate": 0.0009600404266884088, + "loss": 3.5952, + "step": 4705 + }, + { + "epoch": 0.32001630656339175, + "grad_norm": 0.8003848791122437, + "learning_rate": 0.0009599979616795761, + "loss": 3.5842, + "step": 4710 + }, + { + "epoch": 0.32035602663405355, + "grad_norm": 0.6263446807861328, + "learning_rate": 0.0009599554966707433, + "loss": 3.7243, + "step": 4715 + }, + { + "epoch": 0.3206957467047153, + "grad_norm": 0.6043445467948914, + "learning_rate": 0.0009599130316619106, + "loss": 3.912, + "step": 4720 + }, + { + "epoch": 0.3210354667753771, + "grad_norm": 0.6658505797386169, + "learning_rate": 0.0009598705666530779, + "loss": 3.8563, + "step": 4725 + }, + { + "epoch": 0.3213751868460389, + "grad_norm": 0.7094211578369141, + "learning_rate": 0.0009598281016442451, + "loss": 3.6251, + "step": 4730 + }, + { + "epoch": 0.3217149069167006, + "grad_norm": 0.6801108717918396, + "learning_rate": 0.0009597856366354124, + "loss": 3.4148, + "step": 4735 + }, + { + "epoch": 0.3220546269873624, + "grad_norm": 0.9091446399688721, + "learning_rate": 0.0009597431716265798, + "loss": 3.6064, + "step": 4740 + }, + { + "epoch": 0.32239434705802417, + "grad_norm": 0.609541654586792, + "learning_rate": 0.000959700706617747, + "loss": 3.5285, + "step": 4745 + }, + { + "epoch": 0.32273406712868596, + "grad_norm": 0.6929119825363159, + "learning_rate": 0.0009596582416089143, + "loss": 3.677, + "step": 4750 + }, + { + "epoch": 0.32307378719934776, + "grad_norm": 0.7696729898452759, + "learning_rate": 0.0009596157766000816, + "loss": 3.907, + "step": 4755 + }, + { + "epoch": 0.3234135072700095, + "grad_norm": 0.734379768371582, + "learning_rate": 0.0009595733115912488, + "loss": 3.5694, + "step": 4760 + }, + { + "epoch": 0.3237532273406713, + "grad_norm": 0.670070469379425, + "learning_rate": 0.000959530846582416, + "loss": 3.7851, + "step": 4765 + }, + { + "epoch": 0.32409294741133304, + "grad_norm": 0.7582732439041138, + "learning_rate": 0.0009594883815735834, + "loss": 3.7556, + "step": 4770 + }, + { + "epoch": 0.32443266748199484, + "grad_norm": 0.5920730829238892, + "learning_rate": 0.0009594459165647507, + "loss": 3.5348, + "step": 4775 + }, + { + "epoch": 0.32477238755265664, + "grad_norm": 0.6227866411209106, + "learning_rate": 0.0009594034515559179, + "loss": 3.8472, + "step": 4780 + }, + { + "epoch": 0.3251121076233184, + "grad_norm": 0.6639100909233093, + "learning_rate": 0.0009593609865470853, + "loss": 3.732, + "step": 4785 + }, + { + "epoch": 0.3254518276939802, + "grad_norm": 0.6550115942955017, + "learning_rate": 0.0009593185215382525, + "loss": 3.5549, + "step": 4790 + }, + { + "epoch": 0.3257915477646419, + "grad_norm": 0.6169217824935913, + "learning_rate": 0.0009592760565294197, + "loss": 3.7479, + "step": 4795 + }, + { + "epoch": 0.3261312678353037, + "grad_norm": 0.8241134285926819, + "learning_rate": 0.000959233591520587, + "loss": 3.4747, + "step": 4800 + }, + { + "epoch": 0.32647098790596546, + "grad_norm": 0.8893460631370544, + "learning_rate": 0.0009591911265117543, + "loss": 3.6325, + "step": 4805 + }, + { + "epoch": 0.32681070797662726, + "grad_norm": 0.6666406989097595, + "learning_rate": 0.0009591486615029216, + "loss": 3.9512, + "step": 4810 + }, + { + "epoch": 0.32715042804728905, + "grad_norm": 0.8277883529663086, + "learning_rate": 0.000959106196494089, + "loss": 3.556, + "step": 4815 + }, + { + "epoch": 0.3274901481179508, + "grad_norm": 0.8062562942504883, + "learning_rate": 0.0009590637314852562, + "loss": 3.737, + "step": 4820 + }, + { + "epoch": 0.3278298681886126, + "grad_norm": 0.5659600496292114, + "learning_rate": 0.0009590212664764234, + "loss": 3.8212, + "step": 4825 + }, + { + "epoch": 0.32816958825927434, + "grad_norm": 0.7723137736320496, + "learning_rate": 0.0009589788014675907, + "loss": 3.7699, + "step": 4830 + }, + { + "epoch": 0.32850930832993613, + "grad_norm": 0.8332376480102539, + "learning_rate": 0.000958936336458758, + "loss": 3.733, + "step": 4835 + }, + { + "epoch": 0.32884902840059793, + "grad_norm": 0.6778247952461243, + "learning_rate": 0.0009588938714499252, + "loss": 3.9619, + "step": 4840 + }, + { + "epoch": 0.3291887484712597, + "grad_norm": 0.7050536274909973, + "learning_rate": 0.0009588514064410926, + "loss": 3.8264, + "step": 4845 + }, + { + "epoch": 0.32952846854192147, + "grad_norm": 0.9085682034492493, + "learning_rate": 0.0009588089414322599, + "loss": 3.7086, + "step": 4850 + }, + { + "epoch": 0.3298681886125832, + "grad_norm": 0.8574063181877136, + "learning_rate": 0.0009587664764234271, + "loss": 3.8815, + "step": 4855 + }, + { + "epoch": 0.330207908683245, + "grad_norm": 0.7238953709602356, + "learning_rate": 0.0009587240114145944, + "loss": 3.4002, + "step": 4860 + }, + { + "epoch": 0.3305476287539068, + "grad_norm": 0.7760982513427734, + "learning_rate": 0.0009586815464057616, + "loss": 3.8384, + "step": 4865 + }, + { + "epoch": 0.33088734882456855, + "grad_norm": 0.8978222608566284, + "learning_rate": 0.0009586390813969289, + "loss": 4.007, + "step": 4870 + }, + { + "epoch": 0.33122706889523035, + "grad_norm": 0.6592036485671997, + "learning_rate": 0.0009585966163880963, + "loss": 3.4802, + "step": 4875 + }, + { + "epoch": 0.3315667889658921, + "grad_norm": 0.7717464566230774, + "learning_rate": 0.0009585541513792635, + "loss": 3.4183, + "step": 4880 + }, + { + "epoch": 0.3319065090365539, + "grad_norm": 0.5275126695632935, + "learning_rate": 0.0009585116863704308, + "loss": 3.8552, + "step": 4885 + }, + { + "epoch": 0.33224622910721563, + "grad_norm": 0.7978081703186035, + "learning_rate": 0.0009584692213615981, + "loss": 3.8261, + "step": 4890 + }, + { + "epoch": 0.3325859491778774, + "grad_norm": 0.5648220777511597, + "learning_rate": 0.0009584267563527653, + "loss": 4.015, + "step": 4895 + }, + { + "epoch": 0.3329256692485392, + "grad_norm": 0.7004973888397217, + "learning_rate": 0.0009583842913439325, + "loss": 3.7186, + "step": 4900 + }, + { + "epoch": 0.33326538931920097, + "grad_norm": 0.7414132952690125, + "learning_rate": 0.0009583418263350999, + "loss": 3.626, + "step": 4905 + }, + { + "epoch": 0.33360510938986276, + "grad_norm": 0.6210420727729797, + "learning_rate": 0.0009582993613262672, + "loss": 3.8391, + "step": 4910 + }, + { + "epoch": 0.3339448294605245, + "grad_norm": 0.7563833594322205, + "learning_rate": 0.0009582568963174344, + "loss": 3.824, + "step": 4915 + }, + { + "epoch": 0.3342845495311863, + "grad_norm": 0.7307597398757935, + "learning_rate": 0.0009582144313086018, + "loss": 3.5793, + "step": 4920 + }, + { + "epoch": 0.3346242696018481, + "grad_norm": 0.6148789525032043, + "learning_rate": 0.000958171966299769, + "loss": 3.8081, + "step": 4925 + }, + { + "epoch": 0.33496398967250984, + "grad_norm": 0.6317515969276428, + "learning_rate": 0.0009581295012909362, + "loss": 4.1875, + "step": 4930 + }, + { + "epoch": 0.33530370974317164, + "grad_norm": 0.7363927960395813, + "learning_rate": 0.0009580870362821036, + "loss": 3.7434, + "step": 4935 + }, + { + "epoch": 0.3356434298138334, + "grad_norm": 0.5728069543838501, + "learning_rate": 0.0009580445712732708, + "loss": 3.6144, + "step": 4940 + }, + { + "epoch": 0.3359831498844952, + "grad_norm": 0.516986072063446, + "learning_rate": 0.0009580021062644382, + "loss": 3.8537, + "step": 4945 + }, + { + "epoch": 0.336322869955157, + "grad_norm": 0.7671200037002563, + "learning_rate": 0.0009579596412556055, + "loss": 3.8065, + "step": 4950 + }, + { + "epoch": 0.3366625900258187, + "grad_norm": 0.5155797600746155, + "learning_rate": 0.0009579171762467727, + "loss": 3.7821, + "step": 4955 + }, + { + "epoch": 0.3370023100964805, + "grad_norm": 0.5971325635910034, + "learning_rate": 0.00095787471123794, + "loss": 3.6163, + "step": 4960 + }, + { + "epoch": 0.33734203016714226, + "grad_norm": 0.5434179306030273, + "learning_rate": 0.0009578322462291072, + "loss": 3.79, + "step": 4965 + }, + { + "epoch": 0.33768175023780406, + "grad_norm": 0.9440898299217224, + "learning_rate": 0.0009577897812202745, + "loss": 3.505, + "step": 4970 + }, + { + "epoch": 0.3380214703084658, + "grad_norm": 0.7863972187042236, + "learning_rate": 0.0009577473162114418, + "loss": 3.6921, + "step": 4975 + }, + { + "epoch": 0.3383611903791276, + "grad_norm": 0.6850526928901672, + "learning_rate": 0.0009577048512026091, + "loss": 3.7286, + "step": 4980 + }, + { + "epoch": 0.3387009104497894, + "grad_norm": 0.6525793671607971, + "learning_rate": 0.0009576623861937764, + "loss": 3.7199, + "step": 4985 + }, + { + "epoch": 0.33904063052045114, + "grad_norm": 1.2761176824569702, + "learning_rate": 0.0009576199211849437, + "loss": 3.7743, + "step": 4990 + }, + { + "epoch": 0.33938035059111293, + "grad_norm": 0.6671357154846191, + "learning_rate": 0.0009575774561761109, + "loss": 3.8975, + "step": 4995 + }, + { + "epoch": 0.3397200706617747, + "grad_norm": 0.7961123585700989, + "learning_rate": 0.0009575349911672781, + "loss": 3.7009, + "step": 5000 + }, + { + "epoch": 0.3400597907324365, + "grad_norm": 1.0096056461334229, + "learning_rate": 0.0009574925261584455, + "loss": 3.9044, + "step": 5005 + }, + { + "epoch": 0.34039951080309827, + "grad_norm": 0.7832284569740295, + "learning_rate": 0.0009574500611496127, + "loss": 3.6719, + "step": 5010 + }, + { + "epoch": 0.34073923087376, + "grad_norm": 0.738598644733429, + "learning_rate": 0.00095740759614078, + "loss": 3.5275, + "step": 5015 + }, + { + "epoch": 0.3410789509444218, + "grad_norm": 0.7370604276657104, + "learning_rate": 0.0009573651311319474, + "loss": 3.5247, + "step": 5020 + }, + { + "epoch": 0.34141867101508355, + "grad_norm": 0.8083518743515015, + "learning_rate": 0.0009573226661231146, + "loss": 3.4899, + "step": 5025 + }, + { + "epoch": 0.34175839108574535, + "grad_norm": 0.7787097692489624, + "learning_rate": 0.0009572802011142818, + "loss": 3.6516, + "step": 5030 + }, + { + "epoch": 0.34209811115640715, + "grad_norm": 0.762360155582428, + "learning_rate": 0.0009572377361054492, + "loss": 3.5661, + "step": 5035 + }, + { + "epoch": 0.3424378312270689, + "grad_norm": 0.8215653300285339, + "learning_rate": 0.0009571952710966164, + "loss": 3.5784, + "step": 5040 + }, + { + "epoch": 0.3427775512977307, + "grad_norm": 0.7317585349082947, + "learning_rate": 0.0009571528060877836, + "loss": 3.7503, + "step": 5045 + }, + { + "epoch": 0.34311727136839243, + "grad_norm": 0.8211789727210999, + "learning_rate": 0.0009571103410789511, + "loss": 3.7904, + "step": 5050 + }, + { + "epoch": 0.3434569914390542, + "grad_norm": 0.49874982237815857, + "learning_rate": 0.0009570678760701183, + "loss": 3.9742, + "step": 5055 + }, + { + "epoch": 0.34379671150971597, + "grad_norm": 0.8592227697372437, + "learning_rate": 0.0009570254110612855, + "loss": 3.5196, + "step": 5060 + }, + { + "epoch": 0.34413643158037777, + "grad_norm": 0.746721625328064, + "learning_rate": 0.0009569829460524528, + "loss": 3.4566, + "step": 5065 + }, + { + "epoch": 0.34447615165103956, + "grad_norm": 0.6125968098640442, + "learning_rate": 0.0009569404810436201, + "loss": 3.6418, + "step": 5070 + }, + { + "epoch": 0.3448158717217013, + "grad_norm": 0.7079349756240845, + "learning_rate": 0.0009568980160347873, + "loss": 3.5957, + "step": 5075 + }, + { + "epoch": 0.3451555917923631, + "grad_norm": 0.77970290184021, + "learning_rate": 0.0009568555510259546, + "loss": 3.6652, + "step": 5080 + }, + { + "epoch": 0.34549531186302485, + "grad_norm": 1.1659437417984009, + "learning_rate": 0.000956813086017122, + "loss": 3.6127, + "step": 5085 + }, + { + "epoch": 0.34583503193368664, + "grad_norm": 0.8484855890274048, + "learning_rate": 0.0009567706210082892, + "loss": 3.5648, + "step": 5090 + }, + { + "epoch": 0.34617475200434844, + "grad_norm": 1.0683693885803223, + "learning_rate": 0.0009567281559994565, + "loss": 3.8022, + "step": 5095 + }, + { + "epoch": 0.3465144720750102, + "grad_norm": 0.6394113898277283, + "learning_rate": 0.0009566856909906237, + "loss": 3.4618, + "step": 5100 + }, + { + "epoch": 0.346854192145672, + "grad_norm": 0.6517549753189087, + "learning_rate": 0.000956643225981791, + "loss": 3.7026, + "step": 5105 + }, + { + "epoch": 0.3471939122163337, + "grad_norm": 0.6326296329498291, + "learning_rate": 0.0009566007609729583, + "loss": 3.7764, + "step": 5110 + }, + { + "epoch": 0.3475336322869955, + "grad_norm": 0.7380351424217224, + "learning_rate": 0.0009565582959641255, + "loss": 3.7525, + "step": 5115 + }, + { + "epoch": 0.3478733523576573, + "grad_norm": 0.6437528133392334, + "learning_rate": 0.0009565158309552929, + "loss": 3.6299, + "step": 5120 + }, + { + "epoch": 0.34821307242831906, + "grad_norm": 0.5986687541007996, + "learning_rate": 0.0009564733659464602, + "loss": 3.8119, + "step": 5125 + }, + { + "epoch": 0.34855279249898086, + "grad_norm": 0.5814089775085449, + "learning_rate": 0.0009564309009376274, + "loss": 3.5218, + "step": 5130 + }, + { + "epoch": 0.3488925125696426, + "grad_norm": 0.677362322807312, + "learning_rate": 0.0009563884359287947, + "loss": 3.5475, + "step": 5135 + }, + { + "epoch": 0.3492322326403044, + "grad_norm": 0.6157347559928894, + "learning_rate": 0.000956345970919962, + "loss": 3.6541, + "step": 5140 + }, + { + "epoch": 0.34957195271096614, + "grad_norm": 0.7887200117111206, + "learning_rate": 0.0009563035059111292, + "loss": 3.4216, + "step": 5145 + }, + { + "epoch": 0.34991167278162794, + "grad_norm": 0.7251083254814148, + "learning_rate": 0.0009562610409022964, + "loss": 3.8344, + "step": 5150 + }, + { + "epoch": 0.35025139285228973, + "grad_norm": 0.9756469130516052, + "learning_rate": 0.0009562185758934639, + "loss": 3.5152, + "step": 5155 + }, + { + "epoch": 0.3505911129229515, + "grad_norm": 0.6757463216781616, + "learning_rate": 0.0009561761108846311, + "loss": 3.9051, + "step": 5160 + }, + { + "epoch": 0.3509308329936133, + "grad_norm": 0.5955451130867004, + "learning_rate": 0.0009561336458757983, + "loss": 3.7723, + "step": 5165 + }, + { + "epoch": 0.351270553064275, + "grad_norm": 0.59389328956604, + "learning_rate": 0.0009560911808669657, + "loss": 3.6294, + "step": 5170 + }, + { + "epoch": 0.3516102731349368, + "grad_norm": 0.6841652989387512, + "learning_rate": 0.0009560487158581329, + "loss": 3.485, + "step": 5175 + }, + { + "epoch": 0.3519499932055986, + "grad_norm": 0.8480034470558167, + "learning_rate": 0.0009560062508493001, + "loss": 3.8257, + "step": 5180 + }, + { + "epoch": 0.35228971327626035, + "grad_norm": 0.8432101011276245, + "learning_rate": 0.0009559637858404675, + "loss": 3.6208, + "step": 5185 + }, + { + "epoch": 0.35262943334692215, + "grad_norm": 0.9456750750541687, + "learning_rate": 0.0009559213208316348, + "loss": 3.809, + "step": 5190 + }, + { + "epoch": 0.3529691534175839, + "grad_norm": 0.6838685274124146, + "learning_rate": 0.000955878855822802, + "loss": 3.6456, + "step": 5195 + }, + { + "epoch": 0.3533088734882457, + "grad_norm": 0.716590940952301, + "learning_rate": 0.0009558363908139693, + "loss": 3.7528, + "step": 5200 + }, + { + "epoch": 0.3536485935589075, + "grad_norm": 0.7408686876296997, + "learning_rate": 0.0009557939258051366, + "loss": 3.5792, + "step": 5205 + }, + { + "epoch": 0.35398831362956923, + "grad_norm": 0.8332806825637817, + "learning_rate": 0.0009557514607963038, + "loss": 3.797, + "step": 5210 + }, + { + "epoch": 0.354328033700231, + "grad_norm": 0.6005328893661499, + "learning_rate": 0.0009557089957874711, + "loss": 3.8753, + "step": 5215 + }, + { + "epoch": 0.35466775377089277, + "grad_norm": 0.8470065593719482, + "learning_rate": 0.0009556665307786384, + "loss": 3.8303, + "step": 5220 + }, + { + "epoch": 0.35500747384155457, + "grad_norm": 0.615916907787323, + "learning_rate": 0.0009556240657698057, + "loss": 3.6057, + "step": 5225 + }, + { + "epoch": 0.3553471939122163, + "grad_norm": 0.5892014503479004, + "learning_rate": 0.000955581600760973, + "loss": 3.6605, + "step": 5230 + }, + { + "epoch": 0.3556869139828781, + "grad_norm": 0.6659533977508545, + "learning_rate": 0.0009555391357521403, + "loss": 3.5359, + "step": 5235 + }, + { + "epoch": 0.3560266340535399, + "grad_norm": 0.5987432599067688, + "learning_rate": 0.0009554966707433075, + "loss": 3.7816, + "step": 5240 + }, + { + "epoch": 0.35636635412420165, + "grad_norm": 0.6748421788215637, + "learning_rate": 0.0009554542057344748, + "loss": 3.5962, + "step": 5245 + }, + { + "epoch": 0.35670607419486344, + "grad_norm": 0.6741753816604614, + "learning_rate": 0.000955411740725642, + "loss": 3.9455, + "step": 5250 + }, + { + "epoch": 0.3570457942655252, + "grad_norm": 0.6069574952125549, + "learning_rate": 0.0009553692757168093, + "loss": 3.5491, + "step": 5255 + }, + { + "epoch": 0.357385514336187, + "grad_norm": 0.6320086717605591, + "learning_rate": 0.0009553268107079767, + "loss": 3.6434, + "step": 5260 + }, + { + "epoch": 0.3577252344068488, + "grad_norm": 0.6192067861557007, + "learning_rate": 0.0009552843456991439, + "loss": 3.4161, + "step": 5265 + }, + { + "epoch": 0.3580649544775105, + "grad_norm": 0.6193461418151855, + "learning_rate": 0.0009552418806903112, + "loss": 3.5882, + "step": 5270 + }, + { + "epoch": 0.3584046745481723, + "grad_norm": 0.6376737952232361, + "learning_rate": 0.0009551994156814785, + "loss": 3.8481, + "step": 5275 + }, + { + "epoch": 0.35874439461883406, + "grad_norm": 0.5878989100456238, + "learning_rate": 0.0009551569506726457, + "loss": 3.6245, + "step": 5280 + }, + { + "epoch": 0.35908411468949586, + "grad_norm": 0.6927847266197205, + "learning_rate": 0.0009551144856638131, + "loss": 3.7532, + "step": 5285 + }, + { + "epoch": 0.35942383476015766, + "grad_norm": 0.8402261137962341, + "learning_rate": 0.0009550720206549803, + "loss": 3.7789, + "step": 5290 + }, + { + "epoch": 0.3597635548308194, + "grad_norm": 0.7300997972488403, + "learning_rate": 0.0009550295556461476, + "loss": 3.9195, + "step": 5295 + }, + { + "epoch": 0.3601032749014812, + "grad_norm": 0.6109755039215088, + "learning_rate": 0.000954987090637315, + "loss": 3.8455, + "step": 5300 + }, + { + "epoch": 0.36044299497214294, + "grad_norm": 0.5649875998497009, + "learning_rate": 0.0009549446256284822, + "loss": 3.9588, + "step": 5305 + }, + { + "epoch": 0.36078271504280474, + "grad_norm": 0.7442036867141724, + "learning_rate": 0.0009549021606196494, + "loss": 3.8603, + "step": 5310 + }, + { + "epoch": 0.3611224351134665, + "grad_norm": 0.5968815088272095, + "learning_rate": 0.0009548596956108167, + "loss": 3.9566, + "step": 5315 + }, + { + "epoch": 0.3614621551841283, + "grad_norm": 0.6159237623214722, + "learning_rate": 0.000954817230601984, + "loss": 3.6103, + "step": 5320 + }, + { + "epoch": 0.3618018752547901, + "grad_norm": 1.258449673652649, + "learning_rate": 0.0009547747655931512, + "loss": 3.9607, + "step": 5325 + }, + { + "epoch": 0.3621415953254518, + "grad_norm": 0.5789459347724915, + "learning_rate": 0.0009547323005843186, + "loss": 3.7212, + "step": 5330 + }, + { + "epoch": 0.3624813153961136, + "grad_norm": 0.9337835907936096, + "learning_rate": 0.0009546898355754859, + "loss": 4.0227, + "step": 5335 + }, + { + "epoch": 0.36282103546677535, + "grad_norm": 0.8357377648353577, + "learning_rate": 0.0009546473705666531, + "loss": 3.6377, + "step": 5340 + }, + { + "epoch": 0.36316075553743715, + "grad_norm": 0.7306584715843201, + "learning_rate": 0.0009546049055578204, + "loss": 3.6089, + "step": 5345 + }, + { + "epoch": 0.36350047560809895, + "grad_norm": 12.268057823181152, + "learning_rate": 0.0009545624405489876, + "loss": 3.7759, + "step": 5350 + }, + { + "epoch": 0.3638401956787607, + "grad_norm": 0.5472047924995422, + "learning_rate": 0.0009545199755401549, + "loss": 3.3878, + "step": 5355 + }, + { + "epoch": 0.3641799157494225, + "grad_norm": 0.7283962965011597, + "learning_rate": 0.0009544775105313222, + "loss": 3.5856, + "step": 5360 + }, + { + "epoch": 0.36451963582008423, + "grad_norm": 0.9926034212112427, + "learning_rate": 0.0009544350455224895, + "loss": 3.8012, + "step": 5365 + }, + { + "epoch": 0.36485935589074603, + "grad_norm": 0.7117539644241333, + "learning_rate": 0.0009543925805136568, + "loss": 3.441, + "step": 5370 + }, + { + "epoch": 0.3651990759614078, + "grad_norm": 0.6438385248184204, + "learning_rate": 0.0009543501155048241, + "loss": 3.6433, + "step": 5375 + }, + { + "epoch": 0.36553879603206957, + "grad_norm": 0.7695574760437012, + "learning_rate": 0.0009543076504959913, + "loss": 3.5963, + "step": 5380 + }, + { + "epoch": 0.36587851610273137, + "grad_norm": 0.5929762125015259, + "learning_rate": 0.0009542651854871585, + "loss": 3.8106, + "step": 5385 + }, + { + "epoch": 0.3662182361733931, + "grad_norm": 0.6758908629417419, + "learning_rate": 0.0009542227204783259, + "loss": 3.6554, + "step": 5390 + }, + { + "epoch": 0.3665579562440549, + "grad_norm": 0.621306300163269, + "learning_rate": 0.0009541802554694931, + "loss": 3.5419, + "step": 5395 + }, + { + "epoch": 0.36689767631471665, + "grad_norm": 0.6142150163650513, + "learning_rate": 0.0009541377904606604, + "loss": 3.6943, + "step": 5400 + }, + { + "epoch": 0.36723739638537845, + "grad_norm": 0.8195553421974182, + "learning_rate": 0.0009540953254518278, + "loss": 3.4791, + "step": 5405 + }, + { + "epoch": 0.36757711645604024, + "grad_norm": 0.7049173712730408, + "learning_rate": 0.000954052860442995, + "loss": 3.5835, + "step": 5410 + }, + { + "epoch": 0.367916836526702, + "grad_norm": 0.6227113008499146, + "learning_rate": 0.0009540103954341622, + "loss": 3.7407, + "step": 5415 + }, + { + "epoch": 0.3682565565973638, + "grad_norm": 0.5784620642662048, + "learning_rate": 0.0009539679304253296, + "loss": 3.6682, + "step": 5420 + }, + { + "epoch": 0.3685962766680255, + "grad_norm": 0.8935307860374451, + "learning_rate": 0.0009539254654164968, + "loss": 3.4359, + "step": 5425 + }, + { + "epoch": 0.3689359967386873, + "grad_norm": 0.5990920662879944, + "learning_rate": 0.000953883000407664, + "loss": 3.7232, + "step": 5430 + }, + { + "epoch": 0.3692757168093491, + "grad_norm": 0.8034440875053406, + "learning_rate": 0.0009538405353988315, + "loss": 3.6245, + "step": 5435 + }, + { + "epoch": 0.36961543688001086, + "grad_norm": 0.765920877456665, + "learning_rate": 0.0009537980703899987, + "loss": 3.4965, + "step": 5440 + }, + { + "epoch": 0.36995515695067266, + "grad_norm": 0.6728287935256958, + "learning_rate": 0.0009537556053811659, + "loss": 3.809, + "step": 5445 + }, + { + "epoch": 0.3702948770213344, + "grad_norm": 0.7797513604164124, + "learning_rate": 0.0009537131403723332, + "loss": 3.5527, + "step": 5450 + }, + { + "epoch": 0.3706345970919962, + "grad_norm": 0.7652395963668823, + "learning_rate": 0.0009536706753635005, + "loss": 3.8327, + "step": 5455 + }, + { + "epoch": 0.370974317162658, + "grad_norm": 0.6470797657966614, + "learning_rate": 0.0009536282103546677, + "loss": 3.6113, + "step": 5460 + }, + { + "epoch": 0.37131403723331974, + "grad_norm": 0.8430959582328796, + "learning_rate": 0.0009535857453458351, + "loss": 3.5643, + "step": 5465 + }, + { + "epoch": 0.37165375730398154, + "grad_norm": 0.8380835652351379, + "learning_rate": 0.0009535432803370024, + "loss": 3.7893, + "step": 5470 + }, + { + "epoch": 0.3719934773746433, + "grad_norm": 0.6937934160232544, + "learning_rate": 0.0009535008153281696, + "loss": 3.8825, + "step": 5475 + }, + { + "epoch": 0.3723331974453051, + "grad_norm": 0.6520200371742249, + "learning_rate": 0.0009534583503193369, + "loss": 3.8422, + "step": 5480 + }, + { + "epoch": 0.3726729175159668, + "grad_norm": 0.7678138613700867, + "learning_rate": 0.0009534158853105042, + "loss": 4.0005, + "step": 5485 + }, + { + "epoch": 0.3730126375866286, + "grad_norm": 0.7923406362533569, + "learning_rate": 0.0009533734203016714, + "loss": 3.6193, + "step": 5490 + }, + { + "epoch": 0.3733523576572904, + "grad_norm": 0.674468457698822, + "learning_rate": 0.0009533309552928387, + "loss": 3.6924, + "step": 5495 + }, + { + "epoch": 0.37369207772795215, + "grad_norm": 0.6791908740997314, + "learning_rate": 0.000953288490284006, + "loss": 3.8278, + "step": 5500 + }, + { + "epoch": 0.37403179779861395, + "grad_norm": 0.7335380911827087, + "learning_rate": 0.0009532460252751733, + "loss": 3.8782, + "step": 5505 + }, + { + "epoch": 0.3743715178692757, + "grad_norm": 0.6129330396652222, + "learning_rate": 0.0009532035602663406, + "loss": 3.5498, + "step": 5510 + }, + { + "epoch": 0.3747112379399375, + "grad_norm": 0.8125602602958679, + "learning_rate": 0.0009531610952575078, + "loss": 3.8319, + "step": 5515 + }, + { + "epoch": 0.3750509580105993, + "grad_norm": 0.7539558410644531, + "learning_rate": 0.0009531186302486751, + "loss": 3.4676, + "step": 5520 + }, + { + "epoch": 0.37539067808126103, + "grad_norm": 0.6741296052932739, + "learning_rate": 0.0009530761652398424, + "loss": 3.5679, + "step": 5525 + }, + { + "epoch": 0.37573039815192283, + "grad_norm": 0.762295663356781, + "learning_rate": 0.0009530337002310096, + "loss": 3.6386, + "step": 5530 + }, + { + "epoch": 0.37607011822258457, + "grad_norm": 0.5959886312484741, + "learning_rate": 0.000952991235222177, + "loss": 4.0023, + "step": 5535 + }, + { + "epoch": 0.37640983829324637, + "grad_norm": 0.6451983451843262, + "learning_rate": 0.0009529487702133443, + "loss": 3.7566, + "step": 5540 + }, + { + "epoch": 0.37674955836390817, + "grad_norm": 0.6125621795654297, + "learning_rate": 0.0009529063052045115, + "loss": 4.0503, + "step": 5545 + }, + { + "epoch": 0.3770892784345699, + "grad_norm": 0.5408923029899597, + "learning_rate": 0.0009528638401956787, + "loss": 3.882, + "step": 5550 + }, + { + "epoch": 0.3774289985052317, + "grad_norm": 0.7112954258918762, + "learning_rate": 0.0009528213751868461, + "loss": 3.9079, + "step": 5555 + }, + { + "epoch": 0.37776871857589345, + "grad_norm": 0.9264182448387146, + "learning_rate": 0.0009527789101780133, + "loss": 3.8527, + "step": 5560 + }, + { + "epoch": 0.37810843864655524, + "grad_norm": 0.8550432920455933, + "learning_rate": 0.0009527364451691805, + "loss": 3.5968, + "step": 5565 + }, + { + "epoch": 0.378448158717217, + "grad_norm": 0.7274702787399292, + "learning_rate": 0.000952693980160348, + "loss": 3.7561, + "step": 5570 + }, + { + "epoch": 0.3787878787878788, + "grad_norm": 0.6404582262039185, + "learning_rate": 0.0009526515151515152, + "loss": 3.5176, + "step": 5575 + }, + { + "epoch": 0.3791275988585406, + "grad_norm": 0.7663081288337708, + "learning_rate": 0.0009526090501426824, + "loss": 3.5577, + "step": 5580 + }, + { + "epoch": 0.3794673189292023, + "grad_norm": 0.6774986386299133, + "learning_rate": 0.0009525665851338498, + "loss": 3.6109, + "step": 5585 + }, + { + "epoch": 0.3798070389998641, + "grad_norm": 0.6545806527137756, + "learning_rate": 0.000952524120125017, + "loss": 3.506, + "step": 5590 + }, + { + "epoch": 0.38014675907052586, + "grad_norm": 0.6958641409873962, + "learning_rate": 0.0009524816551161842, + "loss": 3.5987, + "step": 5595 + }, + { + "epoch": 0.38048647914118766, + "grad_norm": 0.5957985520362854, + "learning_rate": 0.0009524391901073515, + "loss": 3.901, + "step": 5600 + }, + { + "epoch": 0.38082619921184946, + "grad_norm": 0.67305988073349, + "learning_rate": 0.0009523967250985189, + "loss": 3.7973, + "step": 5605 + }, + { + "epoch": 0.3811659192825112, + "grad_norm": 0.7090798020362854, + "learning_rate": 0.0009523542600896861, + "loss": 3.5519, + "step": 5610 + }, + { + "epoch": 0.381505639353173, + "grad_norm": 0.6447020769119263, + "learning_rate": 0.0009523117950808534, + "loss": 3.6189, + "step": 5615 + }, + { + "epoch": 0.38184535942383474, + "grad_norm": 0.6408648490905762, + "learning_rate": 0.0009522693300720207, + "loss": 3.7404, + "step": 5620 + }, + { + "epoch": 0.38218507949449654, + "grad_norm": 0.7665295004844666, + "learning_rate": 0.000952226865063188, + "loss": 3.6038, + "step": 5625 + }, + { + "epoch": 0.38252479956515834, + "grad_norm": 0.7357596755027771, + "learning_rate": 0.0009521844000543552, + "loss": 3.8106, + "step": 5630 + }, + { + "epoch": 0.3828645196358201, + "grad_norm": 0.7217463850975037, + "learning_rate": 0.0009521419350455224, + "loss": 3.5601, + "step": 5635 + }, + { + "epoch": 0.3832042397064819, + "grad_norm": 0.6828171014785767, + "learning_rate": 0.0009520994700366899, + "loss": 3.8198, + "step": 5640 + }, + { + "epoch": 0.3835439597771436, + "grad_norm": 0.7195796966552734, + "learning_rate": 0.0009520570050278571, + "loss": 3.7945, + "step": 5645 + }, + { + "epoch": 0.3838836798478054, + "grad_norm": 0.8151988983154297, + "learning_rate": 0.0009520145400190243, + "loss": 3.6095, + "step": 5650 + }, + { + "epoch": 0.38422339991846716, + "grad_norm": 0.676535964012146, + "learning_rate": 0.0009519720750101917, + "loss": 3.8636, + "step": 5655 + }, + { + "epoch": 0.38456311998912895, + "grad_norm": 0.59398353099823, + "learning_rate": 0.0009519296100013589, + "loss": 3.6462, + "step": 5660 + }, + { + "epoch": 0.38490284005979075, + "grad_norm": 0.7401514053344727, + "learning_rate": 0.0009518871449925261, + "loss": 3.6619, + "step": 5665 + }, + { + "epoch": 0.3852425601304525, + "grad_norm": 0.7680660486221313, + "learning_rate": 0.0009518446799836935, + "loss": 3.5895, + "step": 5670 + }, + { + "epoch": 0.3855822802011143, + "grad_norm": 0.6434188485145569, + "learning_rate": 0.0009518022149748608, + "loss": 3.6502, + "step": 5675 + }, + { + "epoch": 0.38592200027177603, + "grad_norm": 0.735846996307373, + "learning_rate": 0.000951759749966028, + "loss": 3.5815, + "step": 5680 + }, + { + "epoch": 0.38626172034243783, + "grad_norm": 0.623301088809967, + "learning_rate": 0.0009517172849571954, + "loss": 3.5788, + "step": 5685 + }, + { + "epoch": 0.38660144041309963, + "grad_norm": 0.7407659292221069, + "learning_rate": 0.0009516748199483626, + "loss": 3.8412, + "step": 5690 + }, + { + "epoch": 0.38694116048376137, + "grad_norm": 0.6278685331344604, + "learning_rate": 0.0009516323549395298, + "loss": 3.5552, + "step": 5695 + }, + { + "epoch": 0.38728088055442317, + "grad_norm": 0.720552921295166, + "learning_rate": 0.0009515898899306971, + "loss": 3.7116, + "step": 5700 + }, + { + "epoch": 0.3876206006250849, + "grad_norm": 0.7158149480819702, + "learning_rate": 0.0009515474249218644, + "loss": 3.6758, + "step": 5705 + }, + { + "epoch": 0.3879603206957467, + "grad_norm": 0.6728513240814209, + "learning_rate": 0.0009515049599130317, + "loss": 3.7136, + "step": 5710 + }, + { + "epoch": 0.3883000407664085, + "grad_norm": 0.6863958239555359, + "learning_rate": 0.000951462494904199, + "loss": 3.8059, + "step": 5715 + }, + { + "epoch": 0.38863976083707025, + "grad_norm": 0.6744656562805176, + "learning_rate": 0.0009514200298953663, + "loss": 3.7461, + "step": 5720 + }, + { + "epoch": 0.38897948090773204, + "grad_norm": 0.6520145535469055, + "learning_rate": 0.0009513775648865335, + "loss": 3.7768, + "step": 5725 + }, + { + "epoch": 0.3893192009783938, + "grad_norm": 0.6971132755279541, + "learning_rate": 0.0009513350998777008, + "loss": 3.7232, + "step": 5730 + }, + { + "epoch": 0.3896589210490556, + "grad_norm": 0.8385533690452576, + "learning_rate": 0.000951292634868868, + "loss": 3.6967, + "step": 5735 + }, + { + "epoch": 0.3899986411197173, + "grad_norm": 0.6962178945541382, + "learning_rate": 0.0009512501698600353, + "loss": 3.9476, + "step": 5740 + }, + { + "epoch": 0.3903383611903791, + "grad_norm": 0.612159788608551, + "learning_rate": 0.0009512077048512027, + "loss": 3.726, + "step": 5745 + }, + { + "epoch": 0.3906780812610409, + "grad_norm": 0.6508966088294983, + "learning_rate": 0.0009511652398423699, + "loss": 3.7884, + "step": 5750 + }, + { + "epoch": 0.39101780133170266, + "grad_norm": 0.6837339997291565, + "learning_rate": 0.0009511227748335372, + "loss": 3.8072, + "step": 5755 + }, + { + "epoch": 0.39135752140236446, + "grad_norm": 0.7513009905815125, + "learning_rate": 0.0009510803098247045, + "loss": 3.8404, + "step": 5760 + }, + { + "epoch": 0.3916972414730262, + "grad_norm": 0.6866264343261719, + "learning_rate": 0.0009510378448158717, + "loss": 3.5156, + "step": 5765 + }, + { + "epoch": 0.392036961543688, + "grad_norm": 0.7577829957008362, + "learning_rate": 0.000950995379807039, + "loss": 3.9027, + "step": 5770 + }, + { + "epoch": 0.3923766816143498, + "grad_norm": 0.8748649954795837, + "learning_rate": 0.0009509529147982063, + "loss": 3.8692, + "step": 5775 + }, + { + "epoch": 0.39271640168501154, + "grad_norm": 0.7105889916419983, + "learning_rate": 0.0009509104497893736, + "loss": 3.531, + "step": 5780 + }, + { + "epoch": 0.39305612175567334, + "grad_norm": 0.7299677133560181, + "learning_rate": 0.0009508679847805408, + "loss": 3.8598, + "step": 5785 + }, + { + "epoch": 0.3933958418263351, + "grad_norm": 0.7281935811042786, + "learning_rate": 0.0009508255197717082, + "loss": 3.8852, + "step": 5790 + }, + { + "epoch": 0.3937355618969969, + "grad_norm": 0.6584429740905762, + "learning_rate": 0.0009507830547628754, + "loss": 3.7188, + "step": 5795 + }, + { + "epoch": 0.3940752819676587, + "grad_norm": 0.9629406332969666, + "learning_rate": 0.0009507405897540426, + "loss": 3.8586, + "step": 5800 + }, + { + "epoch": 0.3944150020383204, + "grad_norm": 0.7765294909477234, + "learning_rate": 0.00095069812474521, + "loss": 3.842, + "step": 5805 + }, + { + "epoch": 0.3947547221089822, + "grad_norm": 0.7289807200431824, + "learning_rate": 0.0009506556597363772, + "loss": 3.7421, + "step": 5810 + }, + { + "epoch": 0.39509444217964396, + "grad_norm": 0.8604475855827332, + "learning_rate": 0.0009506131947275445, + "loss": 3.434, + "step": 5815 + }, + { + "epoch": 0.39543416225030575, + "grad_norm": 0.676457405090332, + "learning_rate": 0.0009505707297187119, + "loss": 3.8626, + "step": 5820 + }, + { + "epoch": 0.3957738823209675, + "grad_norm": 2.1074578762054443, + "learning_rate": 0.0009505282647098791, + "loss": 3.8085, + "step": 5825 + }, + { + "epoch": 0.3961136023916293, + "grad_norm": 0.8767901062965393, + "learning_rate": 0.0009504857997010463, + "loss": 3.5175, + "step": 5830 + }, + { + "epoch": 0.3964533224622911, + "grad_norm": 0.6856021285057068, + "learning_rate": 0.0009504433346922136, + "loss": 3.8532, + "step": 5835 + }, + { + "epoch": 0.39679304253295283, + "grad_norm": 0.6957550644874573, + "learning_rate": 0.0009504008696833809, + "loss": 3.6103, + "step": 5840 + }, + { + "epoch": 0.39713276260361463, + "grad_norm": 1.1055721044540405, + "learning_rate": 0.0009503584046745481, + "loss": 3.6346, + "step": 5845 + }, + { + "epoch": 0.3974724826742764, + "grad_norm": 0.6748464703559875, + "learning_rate": 0.0009503159396657155, + "loss": 3.5869, + "step": 5850 + }, + { + "epoch": 0.39781220274493817, + "grad_norm": 0.754275381565094, + "learning_rate": 0.0009502734746568828, + "loss": 3.7318, + "step": 5855 + }, + { + "epoch": 0.39815192281559997, + "grad_norm": 0.5751346945762634, + "learning_rate": 0.00095023100964805, + "loss": 3.6952, + "step": 5860 + }, + { + "epoch": 0.3984916428862617, + "grad_norm": 0.571988582611084, + "learning_rate": 0.0009501885446392173, + "loss": 3.8725, + "step": 5865 + }, + { + "epoch": 0.3988313629569235, + "grad_norm": 0.7949030995368958, + "learning_rate": 0.0009501460796303846, + "loss": 3.6394, + "step": 5870 + }, + { + "epoch": 0.39917108302758525, + "grad_norm": 0.5735267996788025, + "learning_rate": 0.0009501036146215518, + "loss": 3.4689, + "step": 5875 + }, + { + "epoch": 0.39951080309824705, + "grad_norm": 0.8305995464324951, + "learning_rate": 0.0009500611496127191, + "loss": 3.7559, + "step": 5880 + }, + { + "epoch": 0.39985052316890884, + "grad_norm": 0.6194843053817749, + "learning_rate": 0.0009500186846038864, + "loss": 3.902, + "step": 5885 + }, + { + "epoch": 0.4001902432395706, + "grad_norm": 0.6379867792129517, + "learning_rate": 0.0009499762195950537, + "loss": 3.7771, + "step": 5890 + }, + { + "epoch": 0.4005299633102324, + "grad_norm": 0.6605093479156494, + "learning_rate": 0.000949933754586221, + "loss": 3.5102, + "step": 5895 + }, + { + "epoch": 0.4008696833808941, + "grad_norm": 0.9110813736915588, + "learning_rate": 0.0009498912895773882, + "loss": 3.662, + "step": 5900 + }, + { + "epoch": 0.4012094034515559, + "grad_norm": 0.6358461976051331, + "learning_rate": 0.0009498488245685555, + "loss": 3.8034, + "step": 5905 + }, + { + "epoch": 0.40154912352221767, + "grad_norm": 0.7812647819519043, + "learning_rate": 0.0009498063595597228, + "loss": 3.7213, + "step": 5910 + }, + { + "epoch": 0.40188884359287946, + "grad_norm": 0.7158457636833191, + "learning_rate": 0.00094976389455089, + "loss": 3.7999, + "step": 5915 + }, + { + "epoch": 0.40222856366354126, + "grad_norm": 1.4320002794265747, + "learning_rate": 0.0009497214295420574, + "loss": 3.8032, + "step": 5920 + }, + { + "epoch": 0.402568283734203, + "grad_norm": 0.6569749712944031, + "learning_rate": 0.0009496789645332247, + "loss": 3.6588, + "step": 5925 + }, + { + "epoch": 0.4029080038048648, + "grad_norm": 0.7718832492828369, + "learning_rate": 0.0009496364995243919, + "loss": 3.6084, + "step": 5930 + }, + { + "epoch": 0.40324772387552654, + "grad_norm": 0.6366055011749268, + "learning_rate": 0.0009495940345155591, + "loss": 3.8482, + "step": 5935 + }, + { + "epoch": 0.40358744394618834, + "grad_norm": 0.6743378639221191, + "learning_rate": 0.0009495515695067265, + "loss": 3.6285, + "step": 5940 + }, + { + "epoch": 0.40392716401685014, + "grad_norm": 0.7103203535079956, + "learning_rate": 0.0009495091044978937, + "loss": 3.728, + "step": 5945 + }, + { + "epoch": 0.4042668840875119, + "grad_norm": 0.6597638130187988, + "learning_rate": 0.0009494666394890609, + "loss": 3.9413, + "step": 5950 + }, + { + "epoch": 0.4046066041581737, + "grad_norm": 0.6881067752838135, + "learning_rate": 0.0009494241744802284, + "loss": 3.8261, + "step": 5955 + }, + { + "epoch": 0.4049463242288354, + "grad_norm": 0.696105420589447, + "learning_rate": 0.0009493817094713956, + "loss": 3.7296, + "step": 5960 + }, + { + "epoch": 0.4052860442994972, + "grad_norm": 0.6859625577926636, + "learning_rate": 0.0009493392444625629, + "loss": 3.8281, + "step": 5965 + }, + { + "epoch": 0.405625764370159, + "grad_norm": 0.7748442888259888, + "learning_rate": 0.0009492967794537302, + "loss": 4.0015, + "step": 5970 + }, + { + "epoch": 0.40596548444082076, + "grad_norm": 0.6900736093521118, + "learning_rate": 0.0009492543144448974, + "loss": 3.4926, + "step": 5975 + }, + { + "epoch": 0.40630520451148255, + "grad_norm": 0.7808459401130676, + "learning_rate": 0.0009492118494360647, + "loss": 3.6567, + "step": 5980 + }, + { + "epoch": 0.4066449245821443, + "grad_norm": 1.1225916147232056, + "learning_rate": 0.000949169384427232, + "loss": 3.5888, + "step": 5985 + }, + { + "epoch": 0.4069846446528061, + "grad_norm": 0.8027675151824951, + "learning_rate": 0.0009491269194183993, + "loss": 4.0031, + "step": 5990 + }, + { + "epoch": 0.40732436472346784, + "grad_norm": 0.6575860977172852, + "learning_rate": 0.0009490844544095666, + "loss": 3.8601, + "step": 5995 + }, + { + "epoch": 0.40766408479412963, + "grad_norm": 0.6641314029693604, + "learning_rate": 0.0009490419894007338, + "loss": 3.6208, + "step": 6000 + }, + { + "epoch": 0.40800380486479143, + "grad_norm": 0.7412399053573608, + "learning_rate": 0.0009489995243919011, + "loss": 3.7681, + "step": 6005 + }, + { + "epoch": 0.4083435249354532, + "grad_norm": 0.8145126104354858, + "learning_rate": 0.0009489570593830684, + "loss": 3.7507, + "step": 6010 + }, + { + "epoch": 0.40868324500611497, + "grad_norm": 0.6852948665618896, + "learning_rate": 0.0009489145943742356, + "loss": 3.6825, + "step": 6015 + }, + { + "epoch": 0.4090229650767767, + "grad_norm": 0.7576074600219727, + "learning_rate": 0.000948872129365403, + "loss": 3.7356, + "step": 6020 + }, + { + "epoch": 0.4093626851474385, + "grad_norm": 0.7892106771469116, + "learning_rate": 0.0009488296643565703, + "loss": 3.5851, + "step": 6025 + }, + { + "epoch": 0.4097024052181003, + "grad_norm": 0.6661109924316406, + "learning_rate": 0.0009487871993477375, + "loss": 3.5746, + "step": 6030 + }, + { + "epoch": 0.41004212528876205, + "grad_norm": 0.7307931184768677, + "learning_rate": 0.0009487447343389047, + "loss": 3.6414, + "step": 6035 + }, + { + "epoch": 0.41038184535942385, + "grad_norm": 0.9471137523651123, + "learning_rate": 0.0009487022693300721, + "loss": 3.7123, + "step": 6040 + }, + { + "epoch": 0.4107215654300856, + "grad_norm": 0.7316438555717468, + "learning_rate": 0.0009486598043212393, + "loss": 3.7229, + "step": 6045 + }, + { + "epoch": 0.4110612855007474, + "grad_norm": 0.7826873660087585, + "learning_rate": 0.0009486173393124065, + "loss": 3.7605, + "step": 6050 + }, + { + "epoch": 0.4114010055714092, + "grad_norm": 0.6444186568260193, + "learning_rate": 0.000948574874303574, + "loss": 3.5857, + "step": 6055 + }, + { + "epoch": 0.4117407256420709, + "grad_norm": 0.7916932702064514, + "learning_rate": 0.0009485324092947412, + "loss": 3.5616, + "step": 6060 + }, + { + "epoch": 0.4120804457127327, + "grad_norm": 0.9611313343048096, + "learning_rate": 0.0009484899442859084, + "loss": 3.832, + "step": 6065 + }, + { + "epoch": 0.41242016578339447, + "grad_norm": 0.8715277314186096, + "learning_rate": 0.0009484474792770758, + "loss": 3.9989, + "step": 6070 + }, + { + "epoch": 0.41275988585405626, + "grad_norm": 0.7321308255195618, + "learning_rate": 0.000948405014268243, + "loss": 3.2912, + "step": 6075 + }, + { + "epoch": 0.413099605924718, + "grad_norm": 0.9948475360870361, + "learning_rate": 0.0009483625492594102, + "loss": 3.9157, + "step": 6080 + }, + { + "epoch": 0.4134393259953798, + "grad_norm": 0.9011125564575195, + "learning_rate": 0.0009483200842505775, + "loss": 3.6813, + "step": 6085 + }, + { + "epoch": 0.4137790460660416, + "grad_norm": 0.6323263645172119, + "learning_rate": 0.0009482776192417449, + "loss": 3.9186, + "step": 6090 + }, + { + "epoch": 0.41411876613670334, + "grad_norm": 0.9080272912979126, + "learning_rate": 0.0009482351542329121, + "loss": 3.6284, + "step": 6095 + }, + { + "epoch": 0.41445848620736514, + "grad_norm": 0.6488545536994934, + "learning_rate": 0.0009481926892240794, + "loss": 3.5882, + "step": 6100 + }, + { + "epoch": 0.4147982062780269, + "grad_norm": 0.6264088153839111, + "learning_rate": 0.0009481502242152467, + "loss": 3.4939, + "step": 6105 + }, + { + "epoch": 0.4151379263486887, + "grad_norm": 0.7170696258544922, + "learning_rate": 0.0009481077592064139, + "loss": 3.6584, + "step": 6110 + }, + { + "epoch": 0.4154776464193505, + "grad_norm": 1.0384855270385742, + "learning_rate": 0.0009480652941975812, + "loss": 3.631, + "step": 6115 + }, + { + "epoch": 0.4158173664900122, + "grad_norm": 0.7056418657302856, + "learning_rate": 0.0009480228291887485, + "loss": 3.9131, + "step": 6120 + }, + { + "epoch": 0.416157086560674, + "grad_norm": 0.638219952583313, + "learning_rate": 0.0009479803641799158, + "loss": 3.7226, + "step": 6125 + }, + { + "epoch": 0.41649680663133576, + "grad_norm": 0.6111626625061035, + "learning_rate": 0.0009479378991710831, + "loss": 3.5262, + "step": 6130 + }, + { + "epoch": 0.41683652670199756, + "grad_norm": 0.886923611164093, + "learning_rate": 0.0009478954341622503, + "loss": 3.5885, + "step": 6135 + }, + { + "epoch": 0.41717624677265935, + "grad_norm": 0.6608405113220215, + "learning_rate": 0.0009478529691534176, + "loss": 3.7589, + "step": 6140 + }, + { + "epoch": 0.4175159668433211, + "grad_norm": 0.7782901525497437, + "learning_rate": 0.0009478105041445849, + "loss": 3.8576, + "step": 6145 + }, + { + "epoch": 0.4178556869139829, + "grad_norm": 1.1632819175720215, + "learning_rate": 0.0009477680391357521, + "loss": 3.8477, + "step": 6150 + }, + { + "epoch": 0.41819540698464464, + "grad_norm": 0.7602625489234924, + "learning_rate": 0.0009477255741269194, + "loss": 3.639, + "step": 6155 + }, + { + "epoch": 0.41853512705530643, + "grad_norm": 0.7379399538040161, + "learning_rate": 0.0009476831091180868, + "loss": 3.7218, + "step": 6160 + }, + { + "epoch": 0.4188748471259682, + "grad_norm": 0.7413083910942078, + "learning_rate": 0.000947640644109254, + "loss": 3.9456, + "step": 6165 + }, + { + "epoch": 0.41921456719663, + "grad_norm": 0.755257785320282, + "learning_rate": 0.0009475981791004213, + "loss": 3.7673, + "step": 6170 + }, + { + "epoch": 0.41955428726729177, + "grad_norm": 0.8536903262138367, + "learning_rate": 0.0009475557140915886, + "loss": 3.8075, + "step": 6175 + }, + { + "epoch": 0.4198940073379535, + "grad_norm": 0.6336054801940918, + "learning_rate": 0.0009475132490827558, + "loss": 3.6452, + "step": 6180 + }, + { + "epoch": 0.4202337274086153, + "grad_norm": 0.5919987559318542, + "learning_rate": 0.000947470784073923, + "loss": 3.693, + "step": 6185 + }, + { + "epoch": 0.42057344747927705, + "grad_norm": 0.7092812657356262, + "learning_rate": 0.0009474283190650904, + "loss": 3.5212, + "step": 6190 + }, + { + "epoch": 0.42091316754993885, + "grad_norm": 0.5136128664016724, + "learning_rate": 0.0009473858540562577, + "loss": 3.742, + "step": 6195 + }, + { + "epoch": 0.42125288762060065, + "grad_norm": 0.710419774055481, + "learning_rate": 0.0009473433890474249, + "loss": 3.6801, + "step": 6200 + }, + { + "epoch": 0.4215926076912624, + "grad_norm": 0.74549400806427, + "learning_rate": 0.0009473009240385923, + "loss": 3.7057, + "step": 6205 + }, + { + "epoch": 0.4219323277619242, + "grad_norm": 0.7171955108642578, + "learning_rate": 0.0009472584590297595, + "loss": 3.6453, + "step": 6210 + }, + { + "epoch": 0.42227204783258593, + "grad_norm": 0.776326060295105, + "learning_rate": 0.0009472159940209267, + "loss": 3.5003, + "step": 6215 + }, + { + "epoch": 0.4226117679032477, + "grad_norm": 0.7289601564407349, + "learning_rate": 0.000947173529012094, + "loss": 3.6845, + "step": 6220 + }, + { + "epoch": 0.4229514879739095, + "grad_norm": 0.694192111492157, + "learning_rate": 0.0009471310640032613, + "loss": 3.5497, + "step": 6225 + }, + { + "epoch": 0.42329120804457127, + "grad_norm": 0.7362591028213501, + "learning_rate": 0.0009470885989944286, + "loss": 3.7047, + "step": 6230 + }, + { + "epoch": 0.42363092811523306, + "grad_norm": 0.645523190498352, + "learning_rate": 0.000947046133985596, + "loss": 3.7183, + "step": 6235 + }, + { + "epoch": 0.4239706481858948, + "grad_norm": 0.7591707110404968, + "learning_rate": 0.0009470036689767632, + "loss": 3.6192, + "step": 6240 + }, + { + "epoch": 0.4243103682565566, + "grad_norm": 2.7486765384674072, + "learning_rate": 0.0009469612039679304, + "loss": 3.6772, + "step": 6245 + }, + { + "epoch": 0.42465008832721834, + "grad_norm": 0.7472816109657288, + "learning_rate": 0.0009469187389590977, + "loss": 3.621, + "step": 6250 + }, + { + "epoch": 0.42498980839788014, + "grad_norm": 0.6765187978744507, + "learning_rate": 0.000946876273950265, + "loss": 3.7256, + "step": 6255 + }, + { + "epoch": 0.42532952846854194, + "grad_norm": 1.0844764709472656, + "learning_rate": 0.0009468338089414322, + "loss": 3.3585, + "step": 6260 + }, + { + "epoch": 0.4256692485392037, + "grad_norm": 0.5765259861946106, + "learning_rate": 0.0009467913439325996, + "loss": 3.548, + "step": 6265 + }, + { + "epoch": 0.4260089686098655, + "grad_norm": 1.046905279159546, + "learning_rate": 0.0009467488789237669, + "loss": 3.8785, + "step": 6270 + }, + { + "epoch": 0.4263486886805272, + "grad_norm": 0.661917507648468, + "learning_rate": 0.0009467064139149341, + "loss": 3.5404, + "step": 6275 + }, + { + "epoch": 0.426688408751189, + "grad_norm": 1.0002559423446655, + "learning_rate": 0.0009466639489061014, + "loss": 3.6421, + "step": 6280 + }, + { + "epoch": 0.4270281288218508, + "grad_norm": 0.9296825528144836, + "learning_rate": 0.0009466214838972686, + "loss": 3.8708, + "step": 6285 + }, + { + "epoch": 0.42736784889251256, + "grad_norm": 1.198769450187683, + "learning_rate": 0.0009465790188884359, + "loss": 3.618, + "step": 6290 + }, + { + "epoch": 0.42770756896317436, + "grad_norm": 0.6735273003578186, + "learning_rate": 0.0009465365538796032, + "loss": 3.8058, + "step": 6295 + }, + { + "epoch": 0.4280472890338361, + "grad_norm": 0.6104165315628052, + "learning_rate": 0.0009464940888707705, + "loss": 3.7062, + "step": 6300 + }, + { + "epoch": 0.4283870091044979, + "grad_norm": 0.6903350949287415, + "learning_rate": 0.0009464516238619379, + "loss": 3.3921, + "step": 6305 + }, + { + "epoch": 0.4287267291751597, + "grad_norm": 0.6779121160507202, + "learning_rate": 0.0009464091588531051, + "loss": 4.0328, + "step": 6310 + }, + { + "epoch": 0.42906644924582144, + "grad_norm": 0.6903079748153687, + "learning_rate": 0.0009463666938442723, + "loss": 3.7839, + "step": 6315 + }, + { + "epoch": 0.42940616931648323, + "grad_norm": 0.7634524703025818, + "learning_rate": 0.0009463242288354397, + "loss": 3.7203, + "step": 6320 + }, + { + "epoch": 0.429745889387145, + "grad_norm": 0.8356140851974487, + "learning_rate": 0.0009462817638266069, + "loss": 3.5624, + "step": 6325 + }, + { + "epoch": 0.4300856094578068, + "grad_norm": 1.1918925046920776, + "learning_rate": 0.0009462392988177741, + "loss": 3.597, + "step": 6330 + }, + { + "epoch": 0.4304253295284685, + "grad_norm": 0.7054935097694397, + "learning_rate": 0.0009461968338089415, + "loss": 3.7081, + "step": 6335 + }, + { + "epoch": 0.4307650495991303, + "grad_norm": 0.7931617498397827, + "learning_rate": 0.0009461543688001088, + "loss": 3.7453, + "step": 6340 + }, + { + "epoch": 0.4311047696697921, + "grad_norm": 0.7819299101829529, + "learning_rate": 0.000946111903791276, + "loss": 3.5882, + "step": 6345 + }, + { + "epoch": 0.43144448974045385, + "grad_norm": 0.7842747569084167, + "learning_rate": 0.0009460694387824433, + "loss": 3.6094, + "step": 6350 + }, + { + "epoch": 0.43178420981111565, + "grad_norm": 0.7885016798973083, + "learning_rate": 0.0009460269737736106, + "loss": 3.6641, + "step": 6355 + }, + { + "epoch": 0.4321239298817774, + "grad_norm": 0.7479512095451355, + "learning_rate": 0.0009459845087647778, + "loss": 3.4524, + "step": 6360 + }, + { + "epoch": 0.4324636499524392, + "grad_norm": 0.6265522837638855, + "learning_rate": 0.0009459420437559451, + "loss": 3.7072, + "step": 6365 + }, + { + "epoch": 0.432803370023101, + "grad_norm": 0.9189678430557251, + "learning_rate": 0.0009458995787471125, + "loss": 3.4026, + "step": 6370 + }, + { + "epoch": 0.43314309009376273, + "grad_norm": 0.8275874853134155, + "learning_rate": 0.0009458571137382797, + "loss": 3.6262, + "step": 6375 + }, + { + "epoch": 0.4334828101644245, + "grad_norm": 0.7689844965934753, + "learning_rate": 0.000945814648729447, + "loss": 3.7545, + "step": 6380 + }, + { + "epoch": 0.43382253023508627, + "grad_norm": 0.9868364930152893, + "learning_rate": 0.0009457721837206142, + "loss": 3.7024, + "step": 6385 + }, + { + "epoch": 0.43416225030574807, + "grad_norm": 0.8647677302360535, + "learning_rate": 0.0009457297187117815, + "loss": 3.6294, + "step": 6390 + }, + { + "epoch": 0.43450197037640986, + "grad_norm": 0.8021628856658936, + "learning_rate": 0.0009456872537029488, + "loss": 3.8862, + "step": 6395 + }, + { + "epoch": 0.4348416904470716, + "grad_norm": 0.8168643116950989, + "learning_rate": 0.000945644788694116, + "loss": 3.6425, + "step": 6400 + }, + { + "epoch": 0.4351814105177334, + "grad_norm": 0.653694212436676, + "learning_rate": 0.0009456023236852834, + "loss": 3.7277, + "step": 6405 + }, + { + "epoch": 0.43552113058839514, + "grad_norm": 0.7840037941932678, + "learning_rate": 0.0009455598586764507, + "loss": 3.6298, + "step": 6410 + }, + { + "epoch": 0.43586085065905694, + "grad_norm": 0.8717862367630005, + "learning_rate": 0.0009455173936676179, + "loss": 3.6109, + "step": 6415 + }, + { + "epoch": 0.4362005707297187, + "grad_norm": 0.587716281414032, + "learning_rate": 0.0009454749286587851, + "loss": 3.569, + "step": 6420 + }, + { + "epoch": 0.4365402908003805, + "grad_norm": 0.7357481122016907, + "learning_rate": 0.0009454324636499525, + "loss": 3.5375, + "step": 6425 + }, + { + "epoch": 0.4368800108710423, + "grad_norm": 0.8100959062576294, + "learning_rate": 0.0009453899986411197, + "loss": 3.8886, + "step": 6430 + }, + { + "epoch": 0.437219730941704, + "grad_norm": 1.7049295902252197, + "learning_rate": 0.0009453475336322869, + "loss": 3.757, + "step": 6435 + }, + { + "epoch": 0.4375594510123658, + "grad_norm": 0.7924312949180603, + "learning_rate": 0.0009453050686234544, + "loss": 3.5864, + "step": 6440 + }, + { + "epoch": 0.43789917108302756, + "grad_norm": 1.9161287546157837, + "learning_rate": 0.0009452626036146216, + "loss": 3.6795, + "step": 6445 + }, + { + "epoch": 0.43823889115368936, + "grad_norm": 0.7814688682556152, + "learning_rate": 0.0009452201386057888, + "loss": 3.6705, + "step": 6450 + }, + { + "epoch": 0.43857861122435116, + "grad_norm": 0.7019619941711426, + "learning_rate": 0.0009451776735969562, + "loss": 3.9125, + "step": 6455 + }, + { + "epoch": 0.4389183312950129, + "grad_norm": 0.7521041035652161, + "learning_rate": 0.0009451352085881234, + "loss": 3.5766, + "step": 6460 + }, + { + "epoch": 0.4392580513656747, + "grad_norm": 0.9290032982826233, + "learning_rate": 0.0009450927435792906, + "loss": 3.6903, + "step": 6465 + }, + { + "epoch": 0.43959777143633644, + "grad_norm": 0.8511236906051636, + "learning_rate": 0.000945050278570458, + "loss": 3.4806, + "step": 6470 + }, + { + "epoch": 0.43993749150699824, + "grad_norm": 0.6680617928504944, + "learning_rate": 0.0009450078135616253, + "loss": 3.86, + "step": 6475 + }, + { + "epoch": 0.44027721157766003, + "grad_norm": 0.6665207147598267, + "learning_rate": 0.0009449653485527925, + "loss": 3.8806, + "step": 6480 + }, + { + "epoch": 0.4406169316483218, + "grad_norm": 0.7572169899940491, + "learning_rate": 0.0009449228835439598, + "loss": 3.4571, + "step": 6485 + }, + { + "epoch": 0.44095665171898357, + "grad_norm": 0.7516233921051025, + "learning_rate": 0.0009448804185351271, + "loss": 3.9447, + "step": 6490 + }, + { + "epoch": 0.4412963717896453, + "grad_norm": 0.9311526417732239, + "learning_rate": 0.0009448379535262943, + "loss": 3.6585, + "step": 6495 + }, + { + "epoch": 0.4416360918603071, + "grad_norm": 0.9275510311126709, + "learning_rate": 0.0009447954885174616, + "loss": 3.9472, + "step": 6500 + }, + { + "epoch": 0.4419758119309689, + "grad_norm": 0.8019735813140869, + "learning_rate": 0.0009447530235086289, + "loss": 3.5139, + "step": 6505 + }, + { + "epoch": 0.44231553200163065, + "grad_norm": 1.6833178997039795, + "learning_rate": 0.0009447105584997962, + "loss": 3.8665, + "step": 6510 + }, + { + "epoch": 0.44265525207229245, + "grad_norm": 0.7209401726722717, + "learning_rate": 0.0009446680934909635, + "loss": 3.5843, + "step": 6515 + }, + { + "epoch": 0.4429949721429542, + "grad_norm": 0.8485277891159058, + "learning_rate": 0.0009446256284821307, + "loss": 3.442, + "step": 6520 + }, + { + "epoch": 0.443334692213616, + "grad_norm": 0.8769036531448364, + "learning_rate": 0.000944583163473298, + "loss": 3.8422, + "step": 6525 + }, + { + "epoch": 0.44367441228427773, + "grad_norm": 0.7882018685340881, + "learning_rate": 0.0009445406984644653, + "loss": 3.656, + "step": 6530 + }, + { + "epoch": 0.44401413235493953, + "grad_norm": 0.7847372889518738, + "learning_rate": 0.0009444982334556325, + "loss": 3.5124, + "step": 6535 + }, + { + "epoch": 0.4443538524256013, + "grad_norm": 0.7315607070922852, + "learning_rate": 0.0009444557684467998, + "loss": 3.5742, + "step": 6540 + }, + { + "epoch": 0.44469357249626307, + "grad_norm": 0.6868340969085693, + "learning_rate": 0.0009444133034379672, + "loss": 3.8345, + "step": 6545 + }, + { + "epoch": 0.44503329256692487, + "grad_norm": 0.8444815278053284, + "learning_rate": 0.0009443708384291344, + "loss": 3.4333, + "step": 6550 + }, + { + "epoch": 0.4453730126375866, + "grad_norm": 0.7547391653060913, + "learning_rate": 0.0009443283734203017, + "loss": 3.8328, + "step": 6555 + }, + { + "epoch": 0.4457127327082484, + "grad_norm": 0.655636727809906, + "learning_rate": 0.000944285908411469, + "loss": 3.942, + "step": 6560 + }, + { + "epoch": 0.4460524527789102, + "grad_norm": 0.6583059430122375, + "learning_rate": 0.0009442434434026362, + "loss": 3.3732, + "step": 6565 + }, + { + "epoch": 0.44639217284957194, + "grad_norm": 0.6351702809333801, + "learning_rate": 0.0009442009783938034, + "loss": 3.5253, + "step": 6570 + }, + { + "epoch": 0.44673189292023374, + "grad_norm": 0.9433460831642151, + "learning_rate": 0.0009441585133849709, + "loss": 3.7905, + "step": 6575 + }, + { + "epoch": 0.4470716129908955, + "grad_norm": 0.7520065307617188, + "learning_rate": 0.0009441160483761381, + "loss": 3.5835, + "step": 6580 + }, + { + "epoch": 0.4474113330615573, + "grad_norm": 1.069012999534607, + "learning_rate": 0.0009440735833673053, + "loss": 3.5903, + "step": 6585 + }, + { + "epoch": 0.4477510531322191, + "grad_norm": 0.6990821957588196, + "learning_rate": 0.0009440311183584727, + "loss": 3.6411, + "step": 6590 + }, + { + "epoch": 0.4480907732028808, + "grad_norm": 0.6778050065040588, + "learning_rate": 0.0009439886533496399, + "loss": 3.6482, + "step": 6595 + }, + { + "epoch": 0.4484304932735426, + "grad_norm": 0.6824806332588196, + "learning_rate": 0.0009439461883408071, + "loss": 3.8699, + "step": 6600 + }, + { + "epoch": 0.44877021334420436, + "grad_norm": 0.6217116117477417, + "learning_rate": 0.0009439037233319745, + "loss": 3.5637, + "step": 6605 + }, + { + "epoch": 0.44910993341486616, + "grad_norm": 0.783353328704834, + "learning_rate": 0.0009438612583231418, + "loss": 3.5869, + "step": 6610 + }, + { + "epoch": 0.4494496534855279, + "grad_norm": 0.6910473108291626, + "learning_rate": 0.000943818793314309, + "loss": 3.8893, + "step": 6615 + }, + { + "epoch": 0.4497893735561897, + "grad_norm": 0.6780043840408325, + "learning_rate": 0.0009437763283054763, + "loss": 3.6721, + "step": 6620 + }, + { + "epoch": 0.4501290936268515, + "grad_norm": 0.7951787710189819, + "learning_rate": 0.0009437338632966436, + "loss": 3.7806, + "step": 6625 + }, + { + "epoch": 0.45046881369751324, + "grad_norm": 0.7661988735198975, + "learning_rate": 0.0009436913982878108, + "loss": 3.8586, + "step": 6630 + }, + { + "epoch": 0.45080853376817503, + "grad_norm": 0.9542354345321655, + "learning_rate": 0.0009436489332789781, + "loss": 3.6989, + "step": 6635 + }, + { + "epoch": 0.4511482538388368, + "grad_norm": 0.8352494239807129, + "learning_rate": 0.0009436064682701454, + "loss": 3.7621, + "step": 6640 + }, + { + "epoch": 0.4514879739094986, + "grad_norm": 0.7874074578285217, + "learning_rate": 0.0009435640032613128, + "loss": 3.4904, + "step": 6645 + }, + { + "epoch": 0.45182769398016037, + "grad_norm": 0.7971878051757812, + "learning_rate": 0.00094352153825248, + "loss": 3.8819, + "step": 6650 + }, + { + "epoch": 0.4521674140508221, + "grad_norm": 0.8122754693031311, + "learning_rate": 0.0009434790732436473, + "loss": 3.6617, + "step": 6655 + }, + { + "epoch": 0.4525071341214839, + "grad_norm": 0.9281842112541199, + "learning_rate": 0.0009434366082348146, + "loss": 3.4884, + "step": 6660 + }, + { + "epoch": 0.45284685419214565, + "grad_norm": 0.6347576975822449, + "learning_rate": 0.0009433941432259818, + "loss": 3.5103, + "step": 6665 + }, + { + "epoch": 0.45318657426280745, + "grad_norm": 0.7310206294059753, + "learning_rate": 0.000943351678217149, + "loss": 3.7996, + "step": 6670 + }, + { + "epoch": 0.45352629433346925, + "grad_norm": 0.7240384221076965, + "learning_rate": 0.0009433092132083164, + "loss": 3.9415, + "step": 6675 + }, + { + "epoch": 0.453866014404131, + "grad_norm": 1.0311484336853027, + "learning_rate": 0.0009432667481994837, + "loss": 3.7467, + "step": 6680 + }, + { + "epoch": 0.4542057344747928, + "grad_norm": 0.8163008689880371, + "learning_rate": 0.0009432242831906509, + "loss": 3.5392, + "step": 6685 + }, + { + "epoch": 0.45454545454545453, + "grad_norm": 0.7271761894226074, + "learning_rate": 0.0009431818181818183, + "loss": 3.5076, + "step": 6690 + }, + { + "epoch": 0.45488517461611633, + "grad_norm": 1.3560280799865723, + "learning_rate": 0.0009431393531729855, + "loss": 3.7702, + "step": 6695 + }, + { + "epoch": 0.45522489468677807, + "grad_norm": 0.8395495414733887, + "learning_rate": 0.0009430968881641527, + "loss": 3.405, + "step": 6700 + }, + { + "epoch": 0.45556461475743987, + "grad_norm": 0.9616559743881226, + "learning_rate": 0.0009430544231553201, + "loss": 3.593, + "step": 6705 + }, + { + "epoch": 0.45590433482810166, + "grad_norm": 0.791356086730957, + "learning_rate": 0.0009430119581464873, + "loss": 3.7463, + "step": 6710 + }, + { + "epoch": 0.4562440548987634, + "grad_norm": 0.7021889686584473, + "learning_rate": 0.0009429694931376546, + "loss": 3.6889, + "step": 6715 + }, + { + "epoch": 0.4565837749694252, + "grad_norm": 0.7087218761444092, + "learning_rate": 0.000942927028128822, + "loss": 3.6237, + "step": 6720 + }, + { + "epoch": 0.45692349504008695, + "grad_norm": 1.0868803262710571, + "learning_rate": 0.0009428845631199892, + "loss": 3.56, + "step": 6725 + }, + { + "epoch": 0.45726321511074874, + "grad_norm": 0.7047871947288513, + "learning_rate": 0.0009428420981111564, + "loss": 3.7595, + "step": 6730 + }, + { + "epoch": 0.45760293518141054, + "grad_norm": 1.0414479970932007, + "learning_rate": 0.0009427996331023237, + "loss": 3.7527, + "step": 6735 + }, + { + "epoch": 0.4579426552520723, + "grad_norm": 0.801582932472229, + "learning_rate": 0.000942757168093491, + "loss": 3.7045, + "step": 6740 + }, + { + "epoch": 0.4582823753227341, + "grad_norm": 0.8411511778831482, + "learning_rate": 0.0009427147030846582, + "loss": 3.697, + "step": 6745 + }, + { + "epoch": 0.4586220953933958, + "grad_norm": 0.6610255837440491, + "learning_rate": 0.0009426722380758256, + "loss": 3.7951, + "step": 6750 + }, + { + "epoch": 0.4589618154640576, + "grad_norm": 0.6865373849868774, + "learning_rate": 0.0009426297730669929, + "loss": 3.7947, + "step": 6755 + }, + { + "epoch": 0.4593015355347194, + "grad_norm": 0.7824287414550781, + "learning_rate": 0.0009425873080581601, + "loss": 3.417, + "step": 6760 + }, + { + "epoch": 0.45964125560538116, + "grad_norm": 0.8662406802177429, + "learning_rate": 0.0009425448430493274, + "loss": 3.7397, + "step": 6765 + }, + { + "epoch": 0.45998097567604296, + "grad_norm": 0.772973358631134, + "learning_rate": 0.0009425023780404946, + "loss": 3.8287, + "step": 6770 + }, + { + "epoch": 0.4603206957467047, + "grad_norm": 0.681059718132019, + "learning_rate": 0.0009424599130316619, + "loss": 3.8886, + "step": 6775 + }, + { + "epoch": 0.4606604158173665, + "grad_norm": 0.8197239637374878, + "learning_rate": 0.0009424174480228292, + "loss": 3.8622, + "step": 6780 + }, + { + "epoch": 0.46100013588802824, + "grad_norm": 0.6595356464385986, + "learning_rate": 0.0009423749830139965, + "loss": 3.7606, + "step": 6785 + }, + { + "epoch": 0.46133985595869004, + "grad_norm": 0.9089701771736145, + "learning_rate": 0.0009423325180051638, + "loss": 3.5524, + "step": 6790 + }, + { + "epoch": 0.46167957602935183, + "grad_norm": 0.8775517344474792, + "learning_rate": 0.0009422900529963311, + "loss": 3.5668, + "step": 6795 + }, + { + "epoch": 0.4620192961000136, + "grad_norm": 0.7642073035240173, + "learning_rate": 0.0009422475879874983, + "loss": 3.7652, + "step": 6800 + }, + { + "epoch": 0.4623590161706754, + "grad_norm": 0.8666463494300842, + "learning_rate": 0.0009422051229786655, + "loss": 3.5679, + "step": 6805 + }, + { + "epoch": 0.4626987362413371, + "grad_norm": 0.8861477375030518, + "learning_rate": 0.0009421626579698329, + "loss": 3.5925, + "step": 6810 + }, + { + "epoch": 0.4630384563119989, + "grad_norm": 0.6292643547058105, + "learning_rate": 0.0009421201929610001, + "loss": 3.427, + "step": 6815 + }, + { + "epoch": 0.4633781763826607, + "grad_norm": 0.7941412925720215, + "learning_rate": 0.0009420777279521674, + "loss": 4.0807, + "step": 6820 + }, + { + "epoch": 0.46371789645332245, + "grad_norm": 0.8130123019218445, + "learning_rate": 0.0009420352629433348, + "loss": 3.8395, + "step": 6825 + }, + { + "epoch": 0.46405761652398425, + "grad_norm": 0.7854185104370117, + "learning_rate": 0.000941992797934502, + "loss": 3.6002, + "step": 6830 + }, + { + "epoch": 0.464397336594646, + "grad_norm": 0.7315665483474731, + "learning_rate": 0.0009419503329256692, + "loss": 3.5257, + "step": 6835 + }, + { + "epoch": 0.4647370566653078, + "grad_norm": 0.6648357510566711, + "learning_rate": 0.0009419078679168366, + "loss": 3.829, + "step": 6840 + }, + { + "epoch": 0.4650767767359696, + "grad_norm": 0.684614360332489, + "learning_rate": 0.0009418654029080038, + "loss": 3.6271, + "step": 6845 + }, + { + "epoch": 0.46541649680663133, + "grad_norm": 0.7598673701286316, + "learning_rate": 0.000941822937899171, + "loss": 3.8068, + "step": 6850 + }, + { + "epoch": 0.4657562168772931, + "grad_norm": 0.8031312227249146, + "learning_rate": 0.0009417804728903385, + "loss": 3.6739, + "step": 6855 + }, + { + "epoch": 0.46609593694795487, + "grad_norm": 0.7127792239189148, + "learning_rate": 0.0009417380078815057, + "loss": 3.7733, + "step": 6860 + }, + { + "epoch": 0.46643565701861667, + "grad_norm": 0.6540017127990723, + "learning_rate": 0.0009416955428726729, + "loss": 3.5903, + "step": 6865 + }, + { + "epoch": 0.4667753770892784, + "grad_norm": 0.8222087621688843, + "learning_rate": 0.0009416530778638402, + "loss": 3.5829, + "step": 6870 + }, + { + "epoch": 0.4671150971599402, + "grad_norm": 0.786706805229187, + "learning_rate": 0.0009416106128550075, + "loss": 3.3733, + "step": 6875 + }, + { + "epoch": 0.467454817230602, + "grad_norm": 0.7892522215843201, + "learning_rate": 0.0009415681478461747, + "loss": 3.4567, + "step": 6880 + }, + { + "epoch": 0.46779453730126375, + "grad_norm": 0.7317526340484619, + "learning_rate": 0.000941525682837342, + "loss": 3.6438, + "step": 6885 + }, + { + "epoch": 0.46813425737192554, + "grad_norm": 0.6503486037254333, + "learning_rate": 0.0009414832178285094, + "loss": 3.7529, + "step": 6890 + }, + { + "epoch": 0.4684739774425873, + "grad_norm": 0.7175481915473938, + "learning_rate": 0.0009414407528196766, + "loss": 3.7208, + "step": 6895 + }, + { + "epoch": 0.4688136975132491, + "grad_norm": 0.8006057143211365, + "learning_rate": 0.0009413982878108439, + "loss": 3.8677, + "step": 6900 + }, + { + "epoch": 0.4691534175839109, + "grad_norm": 0.7676239013671875, + "learning_rate": 0.0009413558228020112, + "loss": 3.8635, + "step": 6905 + }, + { + "epoch": 0.4694931376545726, + "grad_norm": 0.8925326466560364, + "learning_rate": 0.0009413133577931784, + "loss": 3.7591, + "step": 6910 + }, + { + "epoch": 0.4698328577252344, + "grad_norm": 0.7252369523048401, + "learning_rate": 0.0009412708927843457, + "loss": 3.7022, + "step": 6915 + }, + { + "epoch": 0.47017257779589616, + "grad_norm": 0.8102614879608154, + "learning_rate": 0.0009412284277755129, + "loss": 3.5695, + "step": 6920 + }, + { + "epoch": 0.47051229786655796, + "grad_norm": 0.6476667523384094, + "learning_rate": 0.0009411859627666803, + "loss": 3.6827, + "step": 6925 + }, + { + "epoch": 0.47085201793721976, + "grad_norm": 1.0310362577438354, + "learning_rate": 0.0009411434977578476, + "loss": 3.7261, + "step": 6930 + }, + { + "epoch": 0.4711917380078815, + "grad_norm": 0.7481087446212769, + "learning_rate": 0.0009411010327490148, + "loss": 3.62, + "step": 6935 + }, + { + "epoch": 0.4715314580785433, + "grad_norm": 0.6489163637161255, + "learning_rate": 0.0009410585677401821, + "loss": 3.685, + "step": 6940 + }, + { + "epoch": 0.47187117814920504, + "grad_norm": 0.7797324657440186, + "learning_rate": 0.0009410161027313494, + "loss": 3.9003, + "step": 6945 + }, + { + "epoch": 0.47221089821986684, + "grad_norm": 0.825636625289917, + "learning_rate": 0.0009409736377225166, + "loss": 3.601, + "step": 6950 + }, + { + "epoch": 0.4725506182905286, + "grad_norm": 1.3893637657165527, + "learning_rate": 0.0009409311727136838, + "loss": 3.9514, + "step": 6955 + }, + { + "epoch": 0.4728903383611904, + "grad_norm": 1.127408504486084, + "learning_rate": 0.0009408887077048513, + "loss": 3.8569, + "step": 6960 + }, + { + "epoch": 0.4732300584318522, + "grad_norm": 1.2677662372589111, + "learning_rate": 0.0009408462426960185, + "loss": 3.7982, + "step": 6965 + }, + { + "epoch": 0.4735697785025139, + "grad_norm": 0.7119567394256592, + "learning_rate": 0.0009408037776871857, + "loss": 3.5079, + "step": 6970 + }, + { + "epoch": 0.4739094985731757, + "grad_norm": 0.7485243082046509, + "learning_rate": 0.0009407613126783531, + "loss": 3.7139, + "step": 6975 + }, + { + "epoch": 0.47424921864383746, + "grad_norm": 1.8985931873321533, + "learning_rate": 0.0009407188476695203, + "loss": 3.9208, + "step": 6980 + }, + { + "epoch": 0.47458893871449925, + "grad_norm": 0.6684504747390747, + "learning_rate": 0.0009406763826606876, + "loss": 3.6739, + "step": 6985 + }, + { + "epoch": 0.47492865878516105, + "grad_norm": 0.8101637959480286, + "learning_rate": 0.0009406339176518549, + "loss": 3.6928, + "step": 6990 + }, + { + "epoch": 0.4752683788558228, + "grad_norm": 1.3333851099014282, + "learning_rate": 0.0009405914526430222, + "loss": 3.3483, + "step": 6995 + }, + { + "epoch": 0.4756080989264846, + "grad_norm": 0.8930231928825378, + "learning_rate": 0.0009405489876341895, + "loss": 3.3347, + "step": 7000 + }, + { + "epoch": 0.47594781899714633, + "grad_norm": 0.9755032062530518, + "learning_rate": 0.0009405065226253568, + "loss": 4.0481, + "step": 7005 + }, + { + "epoch": 0.47628753906780813, + "grad_norm": 0.8229382038116455, + "learning_rate": 0.000940464057616524, + "loss": 3.8443, + "step": 7010 + }, + { + "epoch": 0.4766272591384699, + "grad_norm": 0.8849020600318909, + "learning_rate": 0.0009404215926076913, + "loss": 3.7291, + "step": 7015 + }, + { + "epoch": 0.47696697920913167, + "grad_norm": 0.6264156699180603, + "learning_rate": 0.0009403791275988585, + "loss": 3.3264, + "step": 7020 + }, + { + "epoch": 0.47730669927979347, + "grad_norm": 0.765312910079956, + "learning_rate": 0.0009403366625900258, + "loss": 3.7668, + "step": 7025 + }, + { + "epoch": 0.4776464193504552, + "grad_norm": 0.8168578743934631, + "learning_rate": 0.0009402941975811932, + "loss": 3.7237, + "step": 7030 + }, + { + "epoch": 0.477986139421117, + "grad_norm": 0.7400926351547241, + "learning_rate": 0.0009402517325723604, + "loss": 3.5315, + "step": 7035 + }, + { + "epoch": 0.47832585949177875, + "grad_norm": 0.657805323600769, + "learning_rate": 0.0009402092675635277, + "loss": 3.5353, + "step": 7040 + }, + { + "epoch": 0.47866557956244055, + "grad_norm": 1.1266076564788818, + "learning_rate": 0.000940166802554695, + "loss": 3.7509, + "step": 7045 + }, + { + "epoch": 0.47900529963310234, + "grad_norm": 0.6874140501022339, + "learning_rate": 0.0009401243375458622, + "loss": 3.6881, + "step": 7050 + }, + { + "epoch": 0.4793450197037641, + "grad_norm": 0.7892255187034607, + "learning_rate": 0.0009400818725370294, + "loss": 3.8323, + "step": 7055 + }, + { + "epoch": 0.4796847397744259, + "grad_norm": 0.6673566699028015, + "learning_rate": 0.0009400394075281968, + "loss": 3.7398, + "step": 7060 + }, + { + "epoch": 0.4800244598450876, + "grad_norm": 0.7166375517845154, + "learning_rate": 0.0009399969425193641, + "loss": 3.6002, + "step": 7065 + }, + { + "epoch": 0.4803641799157494, + "grad_norm": 1.9668477773666382, + "learning_rate": 0.0009399544775105313, + "loss": 3.4477, + "step": 7070 + }, + { + "epoch": 0.4807038999864112, + "grad_norm": 1.105467677116394, + "learning_rate": 0.0009399120125016987, + "loss": 3.7769, + "step": 7075 + }, + { + "epoch": 0.48104362005707296, + "grad_norm": 0.8100324273109436, + "learning_rate": 0.0009398695474928659, + "loss": 3.8853, + "step": 7080 + }, + { + "epoch": 0.48138334012773476, + "grad_norm": 0.855181097984314, + "learning_rate": 0.0009398270824840331, + "loss": 3.9055, + "step": 7085 + }, + { + "epoch": 0.4817230601983965, + "grad_norm": 0.8410840630531311, + "learning_rate": 0.0009397846174752005, + "loss": 3.6061, + "step": 7090 + }, + { + "epoch": 0.4820627802690583, + "grad_norm": 0.6882803440093994, + "learning_rate": 0.0009397421524663677, + "loss": 3.7707, + "step": 7095 + }, + { + "epoch": 0.4824025003397201, + "grad_norm": 0.5635659098625183, + "learning_rate": 0.000939699687457535, + "loss": 3.6181, + "step": 7100 + }, + { + "epoch": 0.48274222041038184, + "grad_norm": 0.960499107837677, + "learning_rate": 0.0009396572224487024, + "loss": 3.7527, + "step": 7105 + }, + { + "epoch": 0.48308194048104364, + "grad_norm": 1.3995314836502075, + "learning_rate": 0.0009396147574398696, + "loss": 3.8983, + "step": 7110 + }, + { + "epoch": 0.4834216605517054, + "grad_norm": 0.8181443214416504, + "learning_rate": 0.0009395722924310368, + "loss": 3.6004, + "step": 7115 + }, + { + "epoch": 0.4837613806223672, + "grad_norm": 0.6821444630622864, + "learning_rate": 0.0009395298274222041, + "loss": 3.7091, + "step": 7120 + }, + { + "epoch": 0.4841011006930289, + "grad_norm": 1.4712451696395874, + "learning_rate": 0.0009394873624133714, + "loss": 3.6254, + "step": 7125 + }, + { + "epoch": 0.4844408207636907, + "grad_norm": 0.9078455567359924, + "learning_rate": 0.0009394448974045386, + "loss": 3.6762, + "step": 7130 + }, + { + "epoch": 0.4847805408343525, + "grad_norm": 0.9781457185745239, + "learning_rate": 0.000939402432395706, + "loss": 3.5707, + "step": 7135 + }, + { + "epoch": 0.48512026090501426, + "grad_norm": 0.8346604108810425, + "learning_rate": 0.0009393599673868733, + "loss": 3.8072, + "step": 7140 + }, + { + "epoch": 0.48545998097567605, + "grad_norm": 0.9316626787185669, + "learning_rate": 0.0009393175023780405, + "loss": 3.7848, + "step": 7145 + }, + { + "epoch": 0.4857997010463378, + "grad_norm": 0.74868243932724, + "learning_rate": 0.0009392750373692078, + "loss": 3.7603, + "step": 7150 + }, + { + "epoch": 0.4861394211169996, + "grad_norm": 0.7628958821296692, + "learning_rate": 0.000939232572360375, + "loss": 3.8041, + "step": 7155 + }, + { + "epoch": 0.4864791411876614, + "grad_norm": 0.7317331433296204, + "learning_rate": 0.0009391901073515423, + "loss": 3.7645, + "step": 7160 + }, + { + "epoch": 0.48681886125832313, + "grad_norm": 0.8249421119689941, + "learning_rate": 0.0009391476423427097, + "loss": 3.7042, + "step": 7165 + }, + { + "epoch": 0.48715858132898493, + "grad_norm": 0.9656422138214111, + "learning_rate": 0.0009391051773338769, + "loss": 3.361, + "step": 7170 + }, + { + "epoch": 0.48749830139964667, + "grad_norm": 0.849381148815155, + "learning_rate": 0.0009390627123250442, + "loss": 3.7454, + "step": 7175 + }, + { + "epoch": 0.48783802147030847, + "grad_norm": 0.7788578867912292, + "learning_rate": 0.0009390202473162115, + "loss": 3.7689, + "step": 7180 + }, + { + "epoch": 0.48817774154097027, + "grad_norm": 0.7999100685119629, + "learning_rate": 0.0009389777823073787, + "loss": 3.5582, + "step": 7185 + }, + { + "epoch": 0.488517461611632, + "grad_norm": 0.6108559966087341, + "learning_rate": 0.000938935317298546, + "loss": 3.7503, + "step": 7190 + }, + { + "epoch": 0.4888571816822938, + "grad_norm": 0.8938859701156616, + "learning_rate": 0.0009388928522897133, + "loss": 3.8461, + "step": 7195 + }, + { + "epoch": 0.48919690175295555, + "grad_norm": 0.8030597567558289, + "learning_rate": 0.0009388503872808806, + "loss": 3.6002, + "step": 7200 + }, + { + "epoch": 0.48953662182361735, + "grad_norm": 0.7490113973617554, + "learning_rate": 0.0009388079222720478, + "loss": 3.6717, + "step": 7205 + }, + { + "epoch": 0.4898763418942791, + "grad_norm": 0.8324754238128662, + "learning_rate": 0.0009387654572632152, + "loss": 3.9046, + "step": 7210 + }, + { + "epoch": 0.4902160619649409, + "grad_norm": 0.6378701329231262, + "learning_rate": 0.0009387229922543824, + "loss": 3.7435, + "step": 7215 + }, + { + "epoch": 0.4905557820356027, + "grad_norm": 0.6696974635124207, + "learning_rate": 0.0009386805272455496, + "loss": 3.6896, + "step": 7220 + }, + { + "epoch": 0.4908955021062644, + "grad_norm": 0.8677157163619995, + "learning_rate": 0.000938638062236717, + "loss": 3.8479, + "step": 7225 + }, + { + "epoch": 0.4912352221769262, + "grad_norm": 0.7982289791107178, + "learning_rate": 0.0009385955972278842, + "loss": 3.4634, + "step": 7230 + }, + { + "epoch": 0.49157494224758796, + "grad_norm": 1.0052379369735718, + "learning_rate": 0.0009385531322190515, + "loss": 3.7482, + "step": 7235 + }, + { + "epoch": 0.49191466231824976, + "grad_norm": 0.8542678952217102, + "learning_rate": 0.0009385106672102189, + "loss": 3.7788, + "step": 7240 + }, + { + "epoch": 0.49225438238891156, + "grad_norm": 0.9654294848442078, + "learning_rate": 0.0009384682022013861, + "loss": 3.8142, + "step": 7245 + }, + { + "epoch": 0.4925941024595733, + "grad_norm": 0.8354349732398987, + "learning_rate": 0.0009384257371925533, + "loss": 3.8907, + "step": 7250 + }, + { + "epoch": 0.4929338225302351, + "grad_norm": 0.7875881195068359, + "learning_rate": 0.0009383832721837206, + "loss": 3.3389, + "step": 7255 + }, + { + "epoch": 0.49327354260089684, + "grad_norm": 0.8593769669532776, + "learning_rate": 0.0009383408071748879, + "loss": 3.8246, + "step": 7260 + }, + { + "epoch": 0.49361326267155864, + "grad_norm": 0.8465110063552856, + "learning_rate": 0.0009382983421660551, + "loss": 3.7153, + "step": 7265 + }, + { + "epoch": 0.49395298274222044, + "grad_norm": 0.7789784073829651, + "learning_rate": 0.0009382558771572225, + "loss": 3.6484, + "step": 7270 + }, + { + "epoch": 0.4942927028128822, + "grad_norm": 0.8830150365829468, + "learning_rate": 0.0009382134121483898, + "loss": 3.6133, + "step": 7275 + }, + { + "epoch": 0.494632422883544, + "grad_norm": 0.7771656513214111, + "learning_rate": 0.000938170947139557, + "loss": 3.7278, + "step": 7280 + }, + { + "epoch": 0.4949721429542057, + "grad_norm": 0.6703338027000427, + "learning_rate": 0.0009381284821307243, + "loss": 3.9752, + "step": 7285 + }, + { + "epoch": 0.4953118630248675, + "grad_norm": 0.7846285700798035, + "learning_rate": 0.0009380860171218916, + "loss": 3.4729, + "step": 7290 + }, + { + "epoch": 0.49565158309552926, + "grad_norm": 0.948172390460968, + "learning_rate": 0.0009380435521130588, + "loss": 3.7144, + "step": 7295 + }, + { + "epoch": 0.49599130316619106, + "grad_norm": 1.0101864337921143, + "learning_rate": 0.0009380010871042261, + "loss": 3.7646, + "step": 7300 + }, + { + "epoch": 0.49633102323685285, + "grad_norm": 0.7544894814491272, + "learning_rate": 0.0009379586220953934, + "loss": 3.6367, + "step": 7305 + }, + { + "epoch": 0.4966707433075146, + "grad_norm": 0.7571282982826233, + "learning_rate": 0.0009379161570865607, + "loss": 3.6955, + "step": 7310 + }, + { + "epoch": 0.4970104633781764, + "grad_norm": 3.134646415710449, + "learning_rate": 0.000937873692077728, + "loss": 3.5329, + "step": 7315 + }, + { + "epoch": 0.49735018344883813, + "grad_norm": 0.8307409882545471, + "learning_rate": 0.0009378312270688952, + "loss": 3.8784, + "step": 7320 + }, + { + "epoch": 0.49768990351949993, + "grad_norm": 0.8937415480613708, + "learning_rate": 0.0009377887620600626, + "loss": 3.6707, + "step": 7325 + }, + { + "epoch": 0.49802962359016173, + "grad_norm": 0.5931734442710876, + "learning_rate": 0.0009377462970512298, + "loss": 3.599, + "step": 7330 + }, + { + "epoch": 0.49836934366082347, + "grad_norm": 0.6827585101127625, + "learning_rate": 0.000937703832042397, + "loss": 3.8399, + "step": 7335 + }, + { + "epoch": 0.49870906373148527, + "grad_norm": 0.8192744851112366, + "learning_rate": 0.0009376613670335645, + "loss": 3.9087, + "step": 7340 + }, + { + "epoch": 0.499048783802147, + "grad_norm": 1.973146677017212, + "learning_rate": 0.0009376189020247317, + "loss": 3.6978, + "step": 7345 + }, + { + "epoch": 0.4993885038728088, + "grad_norm": 0.6945332884788513, + "learning_rate": 0.0009375764370158989, + "loss": 3.391, + "step": 7350 + }, + { + "epoch": 0.4997282239434706, + "grad_norm": 0.7670307755470276, + "learning_rate": 0.0009375339720070662, + "loss": 3.7131, + "step": 7355 + }, + { + "epoch": 0.5000679440141323, + "grad_norm": 0.8199055790901184, + "learning_rate": 0.0009374915069982335, + "loss": 3.7655, + "step": 7360 + }, + { + "epoch": 0.5004076640847941, + "grad_norm": 0.9656381607055664, + "learning_rate": 0.0009374490419894007, + "loss": 3.7649, + "step": 7365 + }, + { + "epoch": 0.5007473841554559, + "grad_norm": 0.9740866422653198, + "learning_rate": 0.000937406576980568, + "loss": 3.9114, + "step": 7370 + }, + { + "epoch": 0.5010871042261177, + "grad_norm": 0.948676347732544, + "learning_rate": 0.0009373641119717354, + "loss": 3.7265, + "step": 7375 + }, + { + "epoch": 0.5014268242967794, + "grad_norm": 0.7805066108703613, + "learning_rate": 0.0009373216469629026, + "loss": 3.9951, + "step": 7380 + }, + { + "epoch": 0.5017665443674413, + "grad_norm": 0.7733291387557983, + "learning_rate": 0.0009372791819540699, + "loss": 4.0057, + "step": 7385 + }, + { + "epoch": 0.502106264438103, + "grad_norm": 0.8172867894172668, + "learning_rate": 0.0009372367169452372, + "loss": 3.4659, + "step": 7390 + }, + { + "epoch": 0.5024459845087648, + "grad_norm": 1.0584605932235718, + "learning_rate": 0.0009371942519364044, + "loss": 3.6907, + "step": 7395 + }, + { + "epoch": 0.5027857045794265, + "grad_norm": 0.8395977020263672, + "learning_rate": 0.0009371517869275717, + "loss": 3.8193, + "step": 7400 + }, + { + "epoch": 0.5031254246500884, + "grad_norm": 0.9776257872581482, + "learning_rate": 0.0009371093219187389, + "loss": 3.6174, + "step": 7405 + }, + { + "epoch": 0.5034651447207501, + "grad_norm": 3.4432694911956787, + "learning_rate": 0.0009370668569099063, + "loss": 3.6921, + "step": 7410 + }, + { + "epoch": 0.5038048647914118, + "grad_norm": 0.7370182871818542, + "learning_rate": 0.0009370243919010736, + "loss": 3.6347, + "step": 7415 + }, + { + "epoch": 0.5041445848620737, + "grad_norm": 0.8768668174743652, + "learning_rate": 0.0009369819268922408, + "loss": 3.8462, + "step": 7420 + }, + { + "epoch": 0.5044843049327354, + "grad_norm": 0.9089639186859131, + "learning_rate": 0.0009369394618834081, + "loss": 3.5258, + "step": 7425 + }, + { + "epoch": 0.5048240250033972, + "grad_norm": 0.7319662570953369, + "learning_rate": 0.0009368969968745754, + "loss": 3.6741, + "step": 7430 + }, + { + "epoch": 0.5051637450740589, + "grad_norm": 0.5772501230239868, + "learning_rate": 0.0009368545318657426, + "loss": 3.7717, + "step": 7435 + }, + { + "epoch": 0.5055034651447208, + "grad_norm": 0.7919349670410156, + "learning_rate": 0.0009368120668569098, + "loss": 3.6611, + "step": 7440 + }, + { + "epoch": 0.5058431852153825, + "grad_norm": 0.6962769627571106, + "learning_rate": 0.0009367696018480773, + "loss": 3.6246, + "step": 7445 + }, + { + "epoch": 0.5061829052860443, + "grad_norm": 0.6427191495895386, + "learning_rate": 0.0009367271368392445, + "loss": 3.7291, + "step": 7450 + }, + { + "epoch": 0.5065226253567061, + "grad_norm": 0.9956004023551941, + "learning_rate": 0.0009366846718304117, + "loss": 3.6286, + "step": 7455 + }, + { + "epoch": 0.5068623454273679, + "grad_norm": 0.9007059335708618, + "learning_rate": 0.0009366422068215791, + "loss": 3.8041, + "step": 7460 + }, + { + "epoch": 0.5072020654980296, + "grad_norm": 0.9285355806350708, + "learning_rate": 0.0009365997418127463, + "loss": 3.7582, + "step": 7465 + }, + { + "epoch": 0.5075417855686915, + "grad_norm": 0.877824068069458, + "learning_rate": 0.0009365572768039135, + "loss": 3.3837, + "step": 7470 + }, + { + "epoch": 0.5078815056393532, + "grad_norm": 0.666251540184021, + "learning_rate": 0.0009365148117950809, + "loss": 4.0161, + "step": 7475 + }, + { + "epoch": 0.5082212257100149, + "grad_norm": 0.811247706413269, + "learning_rate": 0.0009364723467862482, + "loss": 3.4287, + "step": 7480 + }, + { + "epoch": 0.5085609457806767, + "grad_norm": 0.7683402299880981, + "learning_rate": 0.0009364298817774154, + "loss": 4.0752, + "step": 7485 + }, + { + "epoch": 0.5089006658513385, + "grad_norm": 0.7946499586105347, + "learning_rate": 0.0009363874167685828, + "loss": 3.5025, + "step": 7490 + }, + { + "epoch": 0.5092403859220003, + "grad_norm": 0.8931328058242798, + "learning_rate": 0.00093634495175975, + "loss": 3.775, + "step": 7495 + }, + { + "epoch": 0.509580105992662, + "grad_norm": 0.7139407992362976, + "learning_rate": 0.0009363024867509172, + "loss": 3.76, + "step": 7500 + }, + { + "epoch": 0.5099198260633239, + "grad_norm": 6.0580525398254395, + "learning_rate": 0.0009362600217420845, + "loss": 3.7834, + "step": 7505 + }, + { + "epoch": 0.5102595461339856, + "grad_norm": 0.7540703415870667, + "learning_rate": 0.0009362175567332518, + "loss": 3.8597, + "step": 7510 + }, + { + "epoch": 0.5105992662046474, + "grad_norm": 1.1744097471237183, + "learning_rate": 0.0009361750917244191, + "loss": 3.775, + "step": 7515 + }, + { + "epoch": 0.5109389862753091, + "grad_norm": 0.7168186902999878, + "learning_rate": 0.0009361326267155864, + "loss": 3.7443, + "step": 7520 + }, + { + "epoch": 0.511278706345971, + "grad_norm": 1.5193027257919312, + "learning_rate": 0.0009360901617067537, + "loss": 3.7862, + "step": 7525 + }, + { + "epoch": 0.5116184264166327, + "grad_norm": 1.029643177986145, + "learning_rate": 0.0009360476966979209, + "loss": 3.8337, + "step": 7530 + }, + { + "epoch": 0.5119581464872944, + "grad_norm": 1.1100542545318604, + "learning_rate": 0.0009360052316890882, + "loss": 3.7029, + "step": 7535 + }, + { + "epoch": 0.5122978665579563, + "grad_norm": 1.0867546796798706, + "learning_rate": 0.0009359627666802554, + "loss": 3.855, + "step": 7540 + }, + { + "epoch": 0.512637586628618, + "grad_norm": 0.8331971168518066, + "learning_rate": 0.0009359203016714227, + "loss": 3.4543, + "step": 7545 + }, + { + "epoch": 0.5129773066992798, + "grad_norm": 0.7595094442367554, + "learning_rate": 0.0009358778366625901, + "loss": 3.6165, + "step": 7550 + }, + { + "epoch": 0.5133170267699416, + "grad_norm": 2.122424364089966, + "learning_rate": 0.0009358353716537573, + "loss": 3.8286, + "step": 7555 + }, + { + "epoch": 0.5136567468406034, + "grad_norm": 0.9272922873497009, + "learning_rate": 0.0009357929066449246, + "loss": 3.8261, + "step": 7560 + }, + { + "epoch": 0.5139964669112651, + "grad_norm": 0.9774302244186401, + "learning_rate": 0.0009357504416360919, + "loss": 3.7822, + "step": 7565 + }, + { + "epoch": 0.5143361869819268, + "grad_norm": 0.6903049349784851, + "learning_rate": 0.0009357079766272591, + "loss": 3.7987, + "step": 7570 + }, + { + "epoch": 0.5146759070525887, + "grad_norm": 0.8013500571250916, + "learning_rate": 0.0009356655116184264, + "loss": 3.575, + "step": 7575 + }, + { + "epoch": 0.5150156271232504, + "grad_norm": 1.009568452835083, + "learning_rate": 0.0009356230466095937, + "loss": 3.7885, + "step": 7580 + }, + { + "epoch": 0.5153553471939122, + "grad_norm": 0.7884366512298584, + "learning_rate": 0.000935580581600761, + "loss": 3.8031, + "step": 7585 + }, + { + "epoch": 0.515695067264574, + "grad_norm": 0.8781176209449768, + "learning_rate": 0.0009355381165919283, + "loss": 3.573, + "step": 7590 + }, + { + "epoch": 0.5160347873352358, + "grad_norm": 0.9837929010391235, + "learning_rate": 0.0009354956515830956, + "loss": 3.6799, + "step": 7595 + }, + { + "epoch": 0.5163745074058975, + "grad_norm": 0.6819114685058594, + "learning_rate": 0.0009354531865742628, + "loss": 3.7495, + "step": 7600 + }, + { + "epoch": 0.5167142274765593, + "grad_norm": 0.6864736080169678, + "learning_rate": 0.00093541072156543, + "loss": 3.6115, + "step": 7605 + }, + { + "epoch": 0.5170539475472211, + "grad_norm": 1.046215295791626, + "learning_rate": 0.0009353682565565974, + "loss": 3.6609, + "step": 7610 + }, + { + "epoch": 0.5173936676178829, + "grad_norm": 1.337174415588379, + "learning_rate": 0.0009353257915477646, + "loss": 3.3938, + "step": 7615 + }, + { + "epoch": 0.5177333876885446, + "grad_norm": 0.8165753483772278, + "learning_rate": 0.0009352833265389319, + "loss": 3.8146, + "step": 7620 + }, + { + "epoch": 0.5180731077592065, + "grad_norm": 0.6793105602264404, + "learning_rate": 0.0009352408615300993, + "loss": 3.8535, + "step": 7625 + }, + { + "epoch": 0.5184128278298682, + "grad_norm": 0.9664869904518127, + "learning_rate": 0.0009351983965212665, + "loss": 3.4694, + "step": 7630 + }, + { + "epoch": 0.5187525479005299, + "grad_norm": 1.0473023653030396, + "learning_rate": 0.0009351559315124337, + "loss": 3.8503, + "step": 7635 + }, + { + "epoch": 0.5190922679711918, + "grad_norm": 0.6371603012084961, + "learning_rate": 0.000935113466503601, + "loss": 3.812, + "step": 7640 + }, + { + "epoch": 0.5194319880418535, + "grad_norm": 0.8721631169319153, + "learning_rate": 0.0009350710014947683, + "loss": 3.6361, + "step": 7645 + }, + { + "epoch": 0.5197717081125153, + "grad_norm": 0.7898417711257935, + "learning_rate": 0.0009350285364859355, + "loss": 3.7272, + "step": 7650 + }, + { + "epoch": 0.520111428183177, + "grad_norm": 0.8976787328720093, + "learning_rate": 0.000934986071477103, + "loss": 3.5403, + "step": 7655 + }, + { + "epoch": 0.5204511482538389, + "grad_norm": 1.6082552671432495, + "learning_rate": 0.0009349436064682702, + "loss": 3.6175, + "step": 7660 + }, + { + "epoch": 0.5207908683245006, + "grad_norm": 0.830260694026947, + "learning_rate": 0.0009349011414594374, + "loss": 3.4931, + "step": 7665 + }, + { + "epoch": 0.5211305883951624, + "grad_norm": 0.9476523399353027, + "learning_rate": 0.0009348586764506047, + "loss": 3.6765, + "step": 7670 + }, + { + "epoch": 0.5214703084658242, + "grad_norm": 0.7993727326393127, + "learning_rate": 0.000934816211441772, + "loss": 3.5528, + "step": 7675 + }, + { + "epoch": 0.521810028536486, + "grad_norm": 0.6129350662231445, + "learning_rate": 0.0009347737464329393, + "loss": 3.8175, + "step": 7680 + }, + { + "epoch": 0.5221497486071477, + "grad_norm": 1.0025638341903687, + "learning_rate": 0.0009347312814241065, + "loss": 3.7148, + "step": 7685 + }, + { + "epoch": 0.5224894686778094, + "grad_norm": 0.7446555495262146, + "learning_rate": 0.0009346888164152739, + "loss": 3.8676, + "step": 7690 + }, + { + "epoch": 0.5228291887484713, + "grad_norm": 3.560454845428467, + "learning_rate": 0.0009346463514064412, + "loss": 3.4559, + "step": 7695 + }, + { + "epoch": 0.523168908819133, + "grad_norm": 0.6714736819267273, + "learning_rate": 0.0009346038863976084, + "loss": 3.794, + "step": 7700 + }, + { + "epoch": 0.5235086288897948, + "grad_norm": 0.85779869556427, + "learning_rate": 0.0009345614213887756, + "loss": 3.895, + "step": 7705 + }, + { + "epoch": 0.5238483489604566, + "grad_norm": 0.8376686573028564, + "learning_rate": 0.000934518956379943, + "loss": 3.6197, + "step": 7710 + }, + { + "epoch": 0.5241880690311184, + "grad_norm": 1.0659286975860596, + "learning_rate": 0.0009344764913711102, + "loss": 3.7913, + "step": 7715 + }, + { + "epoch": 0.5245277891017801, + "grad_norm": 6.548509120941162, + "learning_rate": 0.0009344340263622774, + "loss": 3.6373, + "step": 7720 + }, + { + "epoch": 0.524867509172442, + "grad_norm": 5.886991024017334, + "learning_rate": 0.0009343915613534449, + "loss": 3.8122, + "step": 7725 + }, + { + "epoch": 0.5252072292431037, + "grad_norm": 0.7771844863891602, + "learning_rate": 0.0009343490963446121, + "loss": 3.6466, + "step": 7730 + }, + { + "epoch": 0.5255469493137654, + "grad_norm": 0.7829844355583191, + "learning_rate": 0.0009343066313357793, + "loss": 3.5896, + "step": 7735 + }, + { + "epoch": 0.5258866693844272, + "grad_norm": 0.7606650590896606, + "learning_rate": 0.0009342641663269467, + "loss": 3.7448, + "step": 7740 + }, + { + "epoch": 0.526226389455089, + "grad_norm": 0.6309944987297058, + "learning_rate": 0.0009342217013181139, + "loss": 3.6158, + "step": 7745 + }, + { + "epoch": 0.5265661095257508, + "grad_norm": 0.7991791367530823, + "learning_rate": 0.0009341792363092811, + "loss": 3.9482, + "step": 7750 + }, + { + "epoch": 0.5269058295964125, + "grad_norm": 1.166977882385254, + "learning_rate": 0.0009341367713004485, + "loss": 3.8754, + "step": 7755 + }, + { + "epoch": 0.5272455496670744, + "grad_norm": 0.951145589351654, + "learning_rate": 0.0009340943062916158, + "loss": 3.1355, + "step": 7760 + }, + { + "epoch": 0.5275852697377361, + "grad_norm": 1.6882327795028687, + "learning_rate": 0.000934051841282783, + "loss": 3.8167, + "step": 7765 + }, + { + "epoch": 0.5279249898083979, + "grad_norm": 1.015091896057129, + "learning_rate": 0.0009340093762739503, + "loss": 3.6939, + "step": 7770 + }, + { + "epoch": 0.5282647098790596, + "grad_norm": 0.8167543411254883, + "learning_rate": 0.0009339669112651176, + "loss": 3.9178, + "step": 7775 + }, + { + "epoch": 0.5286044299497215, + "grad_norm": 0.8711304068565369, + "learning_rate": 0.0009339244462562848, + "loss": 3.3895, + "step": 7780 + }, + { + "epoch": 0.5289441500203832, + "grad_norm": 1.0086500644683838, + "learning_rate": 0.0009338819812474521, + "loss": 3.8921, + "step": 7785 + }, + { + "epoch": 0.5292838700910449, + "grad_norm": 0.8425483703613281, + "learning_rate": 0.0009338395162386195, + "loss": 3.8591, + "step": 7790 + }, + { + "epoch": 0.5296235901617068, + "grad_norm": 1.2312175035476685, + "learning_rate": 0.0009337970512297867, + "loss": 3.912, + "step": 7795 + }, + { + "epoch": 0.5299633102323685, + "grad_norm": 1.2033361196517944, + "learning_rate": 0.000933754586220954, + "loss": 3.5202, + "step": 7800 + }, + { + "epoch": 0.5303030303030303, + "grad_norm": 0.7065937519073486, + "learning_rate": 0.0009337121212121212, + "loss": 3.6259, + "step": 7805 + }, + { + "epoch": 0.5306427503736921, + "grad_norm": 1.1824893951416016, + "learning_rate": 0.0009336696562032885, + "loss": 3.6446, + "step": 7810 + }, + { + "epoch": 0.5309824704443539, + "grad_norm": 1.3385603427886963, + "learning_rate": 0.0009336271911944558, + "loss": 3.6275, + "step": 7815 + }, + { + "epoch": 0.5313221905150156, + "grad_norm": 0.6545684337615967, + "learning_rate": 0.000933584726185623, + "loss": 3.6846, + "step": 7820 + }, + { + "epoch": 0.5316619105856774, + "grad_norm": 0.7343723773956299, + "learning_rate": 0.0009335422611767904, + "loss": 3.7886, + "step": 7825 + }, + { + "epoch": 0.5320016306563392, + "grad_norm": 1.327511191368103, + "learning_rate": 0.0009334997961679577, + "loss": 3.5956, + "step": 7830 + }, + { + "epoch": 0.532341350727001, + "grad_norm": 0.8653922080993652, + "learning_rate": 0.0009334573311591249, + "loss": 3.8504, + "step": 7835 + }, + { + "epoch": 0.5326810707976627, + "grad_norm": 0.9837885499000549, + "learning_rate": 0.0009334148661502921, + "loss": 3.9316, + "step": 7840 + }, + { + "epoch": 0.5330207908683245, + "grad_norm": 0.8776041269302368, + "learning_rate": 0.0009333724011414595, + "loss": 3.6845, + "step": 7845 + }, + { + "epoch": 0.5333605109389863, + "grad_norm": 0.7139930129051208, + "learning_rate": 0.0009333299361326267, + "loss": 3.7757, + "step": 7850 + }, + { + "epoch": 0.533700231009648, + "grad_norm": 0.5540434718132019, + "learning_rate": 0.0009332874711237939, + "loss": 3.8681, + "step": 7855 + }, + { + "epoch": 0.5340399510803098, + "grad_norm": 0.7290138602256775, + "learning_rate": 0.0009332450061149614, + "loss": 3.7144, + "step": 7860 + }, + { + "epoch": 0.5343796711509716, + "grad_norm": 1.5798171758651733, + "learning_rate": 0.0009332025411061286, + "loss": 3.6538, + "step": 7865 + }, + { + "epoch": 0.5347193912216334, + "grad_norm": 0.6140821576118469, + "learning_rate": 0.0009331600760972958, + "loss": 3.8139, + "step": 7870 + }, + { + "epoch": 0.5350591112922951, + "grad_norm": 0.6630499958992004, + "learning_rate": 0.0009331176110884632, + "loss": 3.7556, + "step": 7875 + }, + { + "epoch": 0.535398831362957, + "grad_norm": 0.8805527091026306, + "learning_rate": 0.0009330751460796304, + "loss": 3.8418, + "step": 7880 + }, + { + "epoch": 0.5357385514336187, + "grad_norm": 0.7567583918571472, + "learning_rate": 0.0009330326810707976, + "loss": 3.6942, + "step": 7885 + }, + { + "epoch": 0.5360782715042804, + "grad_norm": 0.8429593443870544, + "learning_rate": 0.000932990216061965, + "loss": 3.9365, + "step": 7890 + }, + { + "epoch": 0.5364179915749423, + "grad_norm": 0.7898145914077759, + "learning_rate": 0.0009329477510531323, + "loss": 3.4877, + "step": 7895 + }, + { + "epoch": 0.536757711645604, + "grad_norm": 0.6723822355270386, + "learning_rate": 0.0009329052860442995, + "loss": 3.4952, + "step": 7900 + }, + { + "epoch": 0.5370974317162658, + "grad_norm": 0.8021454215049744, + "learning_rate": 0.0009328628210354668, + "loss": 3.8208, + "step": 7905 + }, + { + "epoch": 0.5374371517869275, + "grad_norm": 1.1370452642440796, + "learning_rate": 0.0009328203560266341, + "loss": 3.798, + "step": 7910 + }, + { + "epoch": 0.5377768718575894, + "grad_norm": 0.6277890205383301, + "learning_rate": 0.0009327778910178013, + "loss": 3.8627, + "step": 7915 + }, + { + "epoch": 0.5381165919282511, + "grad_norm": 1.3399752378463745, + "learning_rate": 0.0009327354260089686, + "loss": 3.5943, + "step": 7920 + }, + { + "epoch": 0.5384563119989129, + "grad_norm": 0.8896576166152954, + "learning_rate": 0.0009326929610001359, + "loss": 3.8576, + "step": 7925 + }, + { + "epoch": 0.5387960320695747, + "grad_norm": 0.6801732182502747, + "learning_rate": 0.0009326504959913032, + "loss": 3.8824, + "step": 7930 + }, + { + "epoch": 0.5391357521402365, + "grad_norm": 0.6704099178314209, + "learning_rate": 0.0009326080309824705, + "loss": 3.8018, + "step": 7935 + }, + { + "epoch": 0.5394754722108982, + "grad_norm": 0.9045189619064331, + "learning_rate": 0.0009325655659736377, + "loss": 3.7072, + "step": 7940 + }, + { + "epoch": 0.5398151922815599, + "grad_norm": 1.1468932628631592, + "learning_rate": 0.000932523100964805, + "loss": 3.6685, + "step": 7945 + }, + { + "epoch": 0.5401549123522218, + "grad_norm": 0.8054346442222595, + "learning_rate": 0.0009324806359559723, + "loss": 3.8168, + "step": 7950 + }, + { + "epoch": 0.5404946324228835, + "grad_norm": 0.6940960884094238, + "learning_rate": 0.0009324381709471395, + "loss": 3.5268, + "step": 7955 + }, + { + "epoch": 0.5408343524935453, + "grad_norm": 0.7670547962188721, + "learning_rate": 0.0009323957059383068, + "loss": 3.7249, + "step": 7960 + }, + { + "epoch": 0.5411740725642071, + "grad_norm": 0.9753726720809937, + "learning_rate": 0.0009323532409294742, + "loss": 3.6703, + "step": 7965 + }, + { + "epoch": 0.5415137926348689, + "grad_norm": 0.9470634460449219, + "learning_rate": 0.0009323107759206414, + "loss": 3.6562, + "step": 7970 + }, + { + "epoch": 0.5418535127055306, + "grad_norm": 0.7623605728149414, + "learning_rate": 0.0009322683109118087, + "loss": 3.8138, + "step": 7975 + }, + { + "epoch": 0.5421932327761925, + "grad_norm": 0.7771326899528503, + "learning_rate": 0.000932225845902976, + "loss": 3.7095, + "step": 7980 + }, + { + "epoch": 0.5425329528468542, + "grad_norm": 0.7347263693809509, + "learning_rate": 0.0009321833808941432, + "loss": 3.6734, + "step": 7985 + }, + { + "epoch": 0.542872672917516, + "grad_norm": 0.909206211566925, + "learning_rate": 0.0009321409158853104, + "loss": 3.7041, + "step": 7990 + }, + { + "epoch": 0.5432123929881777, + "grad_norm": 1.1828787326812744, + "learning_rate": 0.0009320984508764778, + "loss": 3.7721, + "step": 7995 + }, + { + "epoch": 0.5435521130588395, + "grad_norm": 0.8316271901130676, + "learning_rate": 0.0009320559858676451, + "loss": 3.8443, + "step": 8000 + }, + { + "epoch": 0.5438918331295013, + "grad_norm": 0.7310352921485901, + "learning_rate": 0.0009320135208588123, + "loss": 3.6048, + "step": 8005 + }, + { + "epoch": 0.544231553200163, + "grad_norm": 0.7386995553970337, + "learning_rate": 0.0009319710558499797, + "loss": 3.779, + "step": 8010 + }, + { + "epoch": 0.5445712732708249, + "grad_norm": 0.6649513244628906, + "learning_rate": 0.0009319285908411469, + "loss": 3.6275, + "step": 8015 + }, + { + "epoch": 0.5449109933414866, + "grad_norm": 0.7939846515655518, + "learning_rate": 0.0009318861258323142, + "loss": 3.7219, + "step": 8020 + }, + { + "epoch": 0.5452507134121484, + "grad_norm": 0.9218465089797974, + "learning_rate": 0.0009318436608234815, + "loss": 3.9361, + "step": 8025 + }, + { + "epoch": 0.5455904334828101, + "grad_norm": 1.0145055055618286, + "learning_rate": 0.0009318011958146487, + "loss": 3.7554, + "step": 8030 + }, + { + "epoch": 0.545930153553472, + "grad_norm": 0.793509304523468, + "learning_rate": 0.0009317587308058161, + "loss": 3.7625, + "step": 8035 + }, + { + "epoch": 0.5462698736241337, + "grad_norm": 0.6704797148704529, + "learning_rate": 0.0009317162657969833, + "loss": 3.4795, + "step": 8040 + }, + { + "epoch": 0.5466095936947954, + "grad_norm": 0.7801114320755005, + "learning_rate": 0.0009316738007881506, + "loss": 3.6944, + "step": 8045 + }, + { + "epoch": 0.5469493137654573, + "grad_norm": 0.9635103940963745, + "learning_rate": 0.0009316313357793179, + "loss": 3.5057, + "step": 8050 + }, + { + "epoch": 0.547289033836119, + "grad_norm": 1.3907709121704102, + "learning_rate": 0.0009315888707704851, + "loss": 3.6027, + "step": 8055 + }, + { + "epoch": 0.5476287539067808, + "grad_norm": 0.9257709383964539, + "learning_rate": 0.0009315464057616524, + "loss": 3.6457, + "step": 8060 + }, + { + "epoch": 0.5479684739774426, + "grad_norm": 0.8470078706741333, + "learning_rate": 0.0009315039407528197, + "loss": 3.6378, + "step": 8065 + }, + { + "epoch": 0.5483081940481044, + "grad_norm": 0.9487181305885315, + "learning_rate": 0.000931461475743987, + "loss": 3.7572, + "step": 8070 + }, + { + "epoch": 0.5486479141187661, + "grad_norm": 0.6114799380302429, + "learning_rate": 0.0009314190107351543, + "loss": 3.7352, + "step": 8075 + }, + { + "epoch": 0.5489876341894279, + "grad_norm": 0.6494504809379578, + "learning_rate": 0.0009313765457263216, + "loss": 3.7128, + "step": 8080 + }, + { + "epoch": 0.5493273542600897, + "grad_norm": 0.7182232737541199, + "learning_rate": 0.0009313340807174888, + "loss": 3.6579, + "step": 8085 + }, + { + "epoch": 0.5496670743307515, + "grad_norm": 0.8048247694969177, + "learning_rate": 0.000931291615708656, + "loss": 3.6483, + "step": 8090 + }, + { + "epoch": 0.5500067944014132, + "grad_norm": 1.2954751253128052, + "learning_rate": 0.0009312491506998234, + "loss": 3.6654, + "step": 8095 + }, + { + "epoch": 0.550346514472075, + "grad_norm": 0.9840929508209229, + "learning_rate": 0.0009312066856909906, + "loss": 3.606, + "step": 8100 + }, + { + "epoch": 0.5506862345427368, + "grad_norm": 2.9611222743988037, + "learning_rate": 0.0009311642206821579, + "loss": 3.5107, + "step": 8105 + }, + { + "epoch": 0.5510259546133985, + "grad_norm": 0.7638512253761292, + "learning_rate": 0.0009311217556733253, + "loss": 3.5989, + "step": 8110 + }, + { + "epoch": 0.5513656746840603, + "grad_norm": 1.5318514108657837, + "learning_rate": 0.0009310792906644925, + "loss": 3.6419, + "step": 8115 + }, + { + "epoch": 0.5517053947547221, + "grad_norm": 0.9865680932998657, + "learning_rate": 0.0009310368256556597, + "loss": 3.6727, + "step": 8120 + }, + { + "epoch": 0.5520451148253839, + "grad_norm": 0.762217104434967, + "learning_rate": 0.0009309943606468271, + "loss": 3.8597, + "step": 8125 + }, + { + "epoch": 0.5523848348960456, + "grad_norm": 1.1266885995864868, + "learning_rate": 0.0009309518956379943, + "loss": 3.8076, + "step": 8130 + }, + { + "epoch": 0.5527245549667075, + "grad_norm": 0.9770743250846863, + "learning_rate": 0.0009309094306291615, + "loss": 3.6184, + "step": 8135 + }, + { + "epoch": 0.5530642750373692, + "grad_norm": 0.7908533215522766, + "learning_rate": 0.000930866965620329, + "loss": 3.6969, + "step": 8140 + }, + { + "epoch": 0.553403995108031, + "grad_norm": 0.6796207427978516, + "learning_rate": 0.0009308245006114962, + "loss": 3.7962, + "step": 8145 + }, + { + "epoch": 0.5537437151786928, + "grad_norm": 1.280287265777588, + "learning_rate": 0.0009307820356026634, + "loss": 3.4654, + "step": 8150 + }, + { + "epoch": 0.5540834352493546, + "grad_norm": 0.866417646408081, + "learning_rate": 0.0009307395705938307, + "loss": 3.678, + "step": 8155 + }, + { + "epoch": 0.5544231553200163, + "grad_norm": 0.6913934946060181, + "learning_rate": 0.000930697105584998, + "loss": 3.5478, + "step": 8160 + }, + { + "epoch": 0.554762875390678, + "grad_norm": 0.7851265668869019, + "learning_rate": 0.0009306546405761652, + "loss": 3.9245, + "step": 8165 + }, + { + "epoch": 0.5551025954613399, + "grad_norm": 0.6437773704528809, + "learning_rate": 0.0009306121755673325, + "loss": 4.0281, + "step": 8170 + }, + { + "epoch": 0.5554423155320016, + "grad_norm": 0.7360479831695557, + "learning_rate": 0.0009305697105584999, + "loss": 3.816, + "step": 8175 + }, + { + "epoch": 0.5557820356026634, + "grad_norm": 0.6019051671028137, + "learning_rate": 0.0009305272455496671, + "loss": 3.6643, + "step": 8180 + }, + { + "epoch": 0.5561217556733252, + "grad_norm": 1.143753170967102, + "learning_rate": 0.0009304847805408344, + "loss": 3.5277, + "step": 8185 + }, + { + "epoch": 0.556461475743987, + "grad_norm": 0.7752367258071899, + "learning_rate": 0.0009304423155320016, + "loss": 3.7001, + "step": 8190 + }, + { + "epoch": 0.5568011958146487, + "grad_norm": 0.874198317527771, + "learning_rate": 0.0009303998505231689, + "loss": 3.7288, + "step": 8195 + }, + { + "epoch": 0.5571409158853105, + "grad_norm": 0.9660792946815491, + "learning_rate": 0.0009303573855143362, + "loss": 3.7146, + "step": 8200 + }, + { + "epoch": 0.5574806359559723, + "grad_norm": 0.8832139372825623, + "learning_rate": 0.0009303149205055034, + "loss": 3.6722, + "step": 8205 + }, + { + "epoch": 0.557820356026634, + "grad_norm": 0.7793344259262085, + "learning_rate": 0.0009302724554966708, + "loss": 3.7029, + "step": 8210 + }, + { + "epoch": 0.5581600760972958, + "grad_norm": 1.028097152709961, + "learning_rate": 0.0009302299904878381, + "loss": 3.6757, + "step": 8215 + }, + { + "epoch": 0.5584997961679576, + "grad_norm": 0.6657744646072388, + "learning_rate": 0.0009301875254790053, + "loss": 3.8865, + "step": 8220 + }, + { + "epoch": 0.5588395162386194, + "grad_norm": 0.9394684433937073, + "learning_rate": 0.0009301450604701725, + "loss": 3.5101, + "step": 8225 + }, + { + "epoch": 0.5591792363092811, + "grad_norm": 0.629466712474823, + "learning_rate": 0.0009301025954613399, + "loss": 3.8255, + "step": 8230 + }, + { + "epoch": 0.559518956379943, + "grad_norm": 0.8225659728050232, + "learning_rate": 0.0009300601304525071, + "loss": 3.6394, + "step": 8235 + }, + { + "epoch": 0.5598586764506047, + "grad_norm": 0.6260023713111877, + "learning_rate": 0.0009300176654436743, + "loss": 3.9481, + "step": 8240 + }, + { + "epoch": 0.5601983965212665, + "grad_norm": 0.7845155000686646, + "learning_rate": 0.0009299752004348418, + "loss": 3.4939, + "step": 8245 + }, + { + "epoch": 0.5605381165919282, + "grad_norm": 0.8477449417114258, + "learning_rate": 0.000929932735426009, + "loss": 3.6506, + "step": 8250 + }, + { + "epoch": 0.5608778366625901, + "grad_norm": 1.160852313041687, + "learning_rate": 0.0009298902704171762, + "loss": 3.5319, + "step": 8255 + }, + { + "epoch": 0.5612175567332518, + "grad_norm": 0.6866785287857056, + "learning_rate": 0.0009298478054083436, + "loss": 3.6514, + "step": 8260 + }, + { + "epoch": 0.5615572768039135, + "grad_norm": 0.5965500473976135, + "learning_rate": 0.0009298053403995108, + "loss": 3.7202, + "step": 8265 + }, + { + "epoch": 0.5618969968745754, + "grad_norm": 0.7244465947151184, + "learning_rate": 0.000929762875390678, + "loss": 3.6218, + "step": 8270 + }, + { + "epoch": 0.5622367169452371, + "grad_norm": 0.8582072854042053, + "learning_rate": 0.0009297204103818455, + "loss": 3.4413, + "step": 8275 + }, + { + "epoch": 0.5625764370158989, + "grad_norm": 0.9259071350097656, + "learning_rate": 0.0009296779453730127, + "loss": 3.7276, + "step": 8280 + }, + { + "epoch": 0.5629161570865606, + "grad_norm": 0.779722273349762, + "learning_rate": 0.0009296354803641799, + "loss": 3.6876, + "step": 8285 + }, + { + "epoch": 0.5632558771572225, + "grad_norm": 0.7717862129211426, + "learning_rate": 0.0009295930153553472, + "loss": 3.8275, + "step": 8290 + }, + { + "epoch": 0.5635955972278842, + "grad_norm": 0.7949544787406921, + "learning_rate": 0.0009295505503465145, + "loss": 3.6666, + "step": 8295 + }, + { + "epoch": 0.563935317298546, + "grad_norm": 0.6197950839996338, + "learning_rate": 0.0009295080853376817, + "loss": 3.8307, + "step": 8300 + }, + { + "epoch": 0.5642750373692078, + "grad_norm": 0.6807712912559509, + "learning_rate": 0.000929465620328849, + "loss": 3.7553, + "step": 8305 + }, + { + "epoch": 0.5646147574398696, + "grad_norm": 0.6942443251609802, + "learning_rate": 0.0009294231553200164, + "loss": 3.7168, + "step": 8310 + }, + { + "epoch": 0.5649544775105313, + "grad_norm": 0.8188214302062988, + "learning_rate": 0.0009293806903111836, + "loss": 3.4529, + "step": 8315 + }, + { + "epoch": 0.5652941975811931, + "grad_norm": 0.8197862505912781, + "learning_rate": 0.0009293382253023509, + "loss": 3.4901, + "step": 8320 + }, + { + "epoch": 0.5656339176518549, + "grad_norm": 0.9546991586685181, + "learning_rate": 0.0009292957602935182, + "loss": 3.9682, + "step": 8325 + }, + { + "epoch": 0.5659736377225166, + "grad_norm": 0.887115478515625, + "learning_rate": 0.0009292532952846854, + "loss": 3.7381, + "step": 8330 + }, + { + "epoch": 0.5663133577931784, + "grad_norm": 0.7302711606025696, + "learning_rate": 0.0009292108302758527, + "loss": 3.8742, + "step": 8335 + }, + { + "epoch": 0.5666530778638402, + "grad_norm": 1.0479846000671387, + "learning_rate": 0.0009291683652670199, + "loss": 3.7514, + "step": 8340 + }, + { + "epoch": 0.566992797934502, + "grad_norm": 1.013793706893921, + "learning_rate": 0.0009291259002581873, + "loss": 3.6086, + "step": 8345 + }, + { + "epoch": 0.5673325180051637, + "grad_norm": 0.926574170589447, + "learning_rate": 0.0009290834352493546, + "loss": 3.9455, + "step": 8350 + }, + { + "epoch": 0.5676722380758256, + "grad_norm": 0.6763489842414856, + "learning_rate": 0.0009290409702405218, + "loss": 3.8298, + "step": 8355 + }, + { + "epoch": 0.5680119581464873, + "grad_norm": 0.6750110983848572, + "learning_rate": 0.0009289985052316892, + "loss": 3.5561, + "step": 8360 + }, + { + "epoch": 0.568351678217149, + "grad_norm": 0.6831505298614502, + "learning_rate": 0.0009289560402228564, + "loss": 3.6613, + "step": 8365 + }, + { + "epoch": 0.5686913982878108, + "grad_norm": 0.7711371779441833, + "learning_rate": 0.0009289135752140236, + "loss": 3.6781, + "step": 8370 + }, + { + "epoch": 0.5690311183584726, + "grad_norm": 0.8220307230949402, + "learning_rate": 0.000928871110205191, + "loss": 3.8715, + "step": 8375 + }, + { + "epoch": 0.5693708384291344, + "grad_norm": 0.6971573233604431, + "learning_rate": 0.0009288286451963583, + "loss": 3.8585, + "step": 8380 + }, + { + "epoch": 0.5697105584997961, + "grad_norm": 0.7647146582603455, + "learning_rate": 0.0009287861801875255, + "loss": 3.7368, + "step": 8385 + }, + { + "epoch": 0.570050278570458, + "grad_norm": 0.7522078156471252, + "learning_rate": 0.0009287437151786928, + "loss": 3.6513, + "step": 8390 + }, + { + "epoch": 0.5703899986411197, + "grad_norm": 0.673718273639679, + "learning_rate": 0.0009287012501698601, + "loss": 3.9671, + "step": 8395 + }, + { + "epoch": 0.5707297187117815, + "grad_norm": 0.8873665928840637, + "learning_rate": 0.0009286587851610273, + "loss": 3.5811, + "step": 8400 + }, + { + "epoch": 0.5710694387824433, + "grad_norm": 0.717910885810852, + "learning_rate": 0.0009286163201521946, + "loss": 3.7473, + "step": 8405 + }, + { + "epoch": 0.5714091588531051, + "grad_norm": 0.9158114790916443, + "learning_rate": 0.0009285738551433619, + "loss": 3.759, + "step": 8410 + }, + { + "epoch": 0.5717488789237668, + "grad_norm": 0.7402276396751404, + "learning_rate": 0.0009285313901345292, + "loss": 3.3486, + "step": 8415 + }, + { + "epoch": 0.5720885989944285, + "grad_norm": 0.7738250494003296, + "learning_rate": 0.0009284889251256965, + "loss": 3.7806, + "step": 8420 + }, + { + "epoch": 0.5724283190650904, + "grad_norm": 0.5874295234680176, + "learning_rate": 0.0009284464601168638, + "loss": 3.735, + "step": 8425 + }, + { + "epoch": 0.5727680391357521, + "grad_norm": 0.6737178564071655, + "learning_rate": 0.000928403995108031, + "loss": 3.6863, + "step": 8430 + }, + { + "epoch": 0.5731077592064139, + "grad_norm": 0.9294766187667847, + "learning_rate": 0.0009283615300991983, + "loss": 3.3676, + "step": 8435 + }, + { + "epoch": 0.5734474792770757, + "grad_norm": 0.8388890624046326, + "learning_rate": 0.0009283190650903655, + "loss": 3.6748, + "step": 8440 + }, + { + "epoch": 0.5737871993477375, + "grad_norm": 0.9405798316001892, + "learning_rate": 0.0009282766000815328, + "loss": 3.5638, + "step": 8445 + }, + { + "epoch": 0.5741269194183992, + "grad_norm": 0.8348854184150696, + "learning_rate": 0.0009282341350727002, + "loss": 3.7603, + "step": 8450 + }, + { + "epoch": 0.574466639489061, + "grad_norm": 0.7804384231567383, + "learning_rate": 0.0009281916700638674, + "loss": 3.4435, + "step": 8455 + }, + { + "epoch": 0.5748063595597228, + "grad_norm": 0.7420020699501038, + "learning_rate": 0.0009281492050550347, + "loss": 3.9071, + "step": 8460 + }, + { + "epoch": 0.5751460796303846, + "grad_norm": 0.6835420727729797, + "learning_rate": 0.000928106740046202, + "loss": 3.7762, + "step": 8465 + }, + { + "epoch": 0.5754857997010463, + "grad_norm": 0.658608615398407, + "learning_rate": 0.0009280642750373692, + "loss": 3.5829, + "step": 8470 + }, + { + "epoch": 0.5758255197717081, + "grad_norm": 0.8319673538208008, + "learning_rate": 0.0009280218100285364, + "loss": 3.5933, + "step": 8475 + }, + { + "epoch": 0.5761652398423699, + "grad_norm": 0.9539414048194885, + "learning_rate": 0.0009279793450197038, + "loss": 3.5525, + "step": 8480 + }, + { + "epoch": 0.5765049599130316, + "grad_norm": 0.70899498462677, + "learning_rate": 0.0009279368800108711, + "loss": 3.7797, + "step": 8485 + }, + { + "epoch": 0.5768446799836935, + "grad_norm": 0.76890629529953, + "learning_rate": 0.0009278944150020383, + "loss": 3.465, + "step": 8490 + }, + { + "epoch": 0.5771844000543552, + "grad_norm": 0.8877576589584351, + "learning_rate": 0.0009278519499932057, + "loss": 3.8866, + "step": 8495 + }, + { + "epoch": 0.577524120125017, + "grad_norm": 0.660688579082489, + "learning_rate": 0.0009278094849843729, + "loss": 3.7658, + "step": 8500 + }, + { + "epoch": 0.5778638401956787, + "grad_norm": 0.896030068397522, + "learning_rate": 0.0009277670199755401, + "loss": 3.6203, + "step": 8505 + }, + { + "epoch": 0.5782035602663406, + "grad_norm": 0.7487848401069641, + "learning_rate": 0.0009277245549667075, + "loss": 3.6431, + "step": 8510 + }, + { + "epoch": 0.5785432803370023, + "grad_norm": 0.7852987051010132, + "learning_rate": 0.0009276820899578747, + "loss": 3.651, + "step": 8515 + }, + { + "epoch": 0.578883000407664, + "grad_norm": 0.759682297706604, + "learning_rate": 0.000927639624949042, + "loss": 3.5244, + "step": 8520 + }, + { + "epoch": 0.5792227204783259, + "grad_norm": 0.5944443345069885, + "learning_rate": 0.0009275971599402094, + "loss": 3.6422, + "step": 8525 + }, + { + "epoch": 0.5795624405489876, + "grad_norm": 0.545721173286438, + "learning_rate": 0.0009275546949313766, + "loss": 3.6353, + "step": 8530 + }, + { + "epoch": 0.5799021606196494, + "grad_norm": 0.8526937961578369, + "learning_rate": 0.0009275122299225438, + "loss": 3.6498, + "step": 8535 + }, + { + "epoch": 0.5802418806903111, + "grad_norm": 1.258561372756958, + "learning_rate": 0.0009274697649137111, + "loss": 3.5986, + "step": 8540 + }, + { + "epoch": 0.580581600760973, + "grad_norm": 0.875015914440155, + "learning_rate": 0.0009274272999048784, + "loss": 3.8405, + "step": 8545 + }, + { + "epoch": 0.5809213208316347, + "grad_norm": 0.6685923337936401, + "learning_rate": 0.0009273848348960456, + "loss": 3.4763, + "step": 8550 + }, + { + "epoch": 0.5812610409022965, + "grad_norm": 0.6964318752288818, + "learning_rate": 0.000927342369887213, + "loss": 3.791, + "step": 8555 + }, + { + "epoch": 0.5816007609729583, + "grad_norm": 0.826518177986145, + "learning_rate": 0.0009272999048783803, + "loss": 3.7179, + "step": 8560 + }, + { + "epoch": 0.5819404810436201, + "grad_norm": 0.7018935680389404, + "learning_rate": 0.0009272574398695475, + "loss": 3.626, + "step": 8565 + }, + { + "epoch": 0.5822802011142818, + "grad_norm": 0.8180730938911438, + "learning_rate": 0.0009272149748607148, + "loss": 3.7448, + "step": 8570 + }, + { + "epoch": 0.5826199211849437, + "grad_norm": 0.8528150320053101, + "learning_rate": 0.000927172509851882, + "loss": 3.7044, + "step": 8575 + }, + { + "epoch": 0.5829596412556054, + "grad_norm": 0.7416120767593384, + "learning_rate": 0.0009271300448430493, + "loss": 3.5249, + "step": 8580 + }, + { + "epoch": 0.5832993613262671, + "grad_norm": 0.6952391862869263, + "learning_rate": 0.0009270875798342166, + "loss": 3.7401, + "step": 8585 + }, + { + "epoch": 0.5836390813969289, + "grad_norm": 0.8442528247833252, + "learning_rate": 0.0009270451148253839, + "loss": 3.6977, + "step": 8590 + }, + { + "epoch": 0.5839788014675907, + "grad_norm": 0.6074960827827454, + "learning_rate": 0.0009270026498165512, + "loss": 3.9663, + "step": 8595 + }, + { + "epoch": 0.5843185215382525, + "grad_norm": 0.7610037326812744, + "learning_rate": 0.0009269601848077185, + "loss": 3.6829, + "step": 8600 + }, + { + "epoch": 0.5846582416089142, + "grad_norm": 0.7824552059173584, + "learning_rate": 0.0009269177197988857, + "loss": 3.9931, + "step": 8605 + }, + { + "epoch": 0.5849979616795761, + "grad_norm": 0.5774102210998535, + "learning_rate": 0.000926875254790053, + "loss": 3.6076, + "step": 8610 + }, + { + "epoch": 0.5853376817502378, + "grad_norm": 0.6610915064811707, + "learning_rate": 0.0009268327897812203, + "loss": 3.7273, + "step": 8615 + }, + { + "epoch": 0.5856774018208996, + "grad_norm": 0.9578468799591064, + "learning_rate": 0.0009267903247723875, + "loss": 3.6174, + "step": 8620 + }, + { + "epoch": 0.5860171218915613, + "grad_norm": 0.7981096506118774, + "learning_rate": 0.0009267478597635548, + "loss": 3.6837, + "step": 8625 + }, + { + "epoch": 0.5863568419622232, + "grad_norm": 0.8085305094718933, + "learning_rate": 0.0009267053947547222, + "loss": 3.8751, + "step": 8630 + }, + { + "epoch": 0.5866965620328849, + "grad_norm": 0.6167998909950256, + "learning_rate": 0.0009266629297458894, + "loss": 3.8212, + "step": 8635 + }, + { + "epoch": 0.5870362821035466, + "grad_norm": 0.8082771301269531, + "learning_rate": 0.0009266204647370566, + "loss": 3.3319, + "step": 8640 + }, + { + "epoch": 0.5873760021742085, + "grad_norm": 0.6618968844413757, + "learning_rate": 0.000926577999728224, + "loss": 4.0224, + "step": 8645 + }, + { + "epoch": 0.5877157222448702, + "grad_norm": 0.6083776950836182, + "learning_rate": 0.0009265355347193912, + "loss": 4.0241, + "step": 8650 + }, + { + "epoch": 0.588055442315532, + "grad_norm": 0.6757774949073792, + "learning_rate": 0.0009264930697105584, + "loss": 3.5121, + "step": 8655 + }, + { + "epoch": 0.5883951623861938, + "grad_norm": 1.010972023010254, + "learning_rate": 0.0009264506047017259, + "loss": 3.5277, + "step": 8660 + }, + { + "epoch": 0.5887348824568556, + "grad_norm": 0.6512971520423889, + "learning_rate": 0.0009264081396928931, + "loss": 3.6784, + "step": 8665 + }, + { + "epoch": 0.5890746025275173, + "grad_norm": 0.5896017551422119, + "learning_rate": 0.0009263656746840603, + "loss": 3.8132, + "step": 8670 + }, + { + "epoch": 0.589414322598179, + "grad_norm": 0.6976354718208313, + "learning_rate": 0.0009263232096752276, + "loss": 3.7363, + "step": 8675 + }, + { + "epoch": 0.5897540426688409, + "grad_norm": 0.5934624671936035, + "learning_rate": 0.0009262807446663949, + "loss": 3.7175, + "step": 8680 + }, + { + "epoch": 0.5900937627395026, + "grad_norm": 0.7559554576873779, + "learning_rate": 0.0009262382796575621, + "loss": 3.5124, + "step": 8685 + }, + { + "epoch": 0.5904334828101644, + "grad_norm": 0.7427394986152649, + "learning_rate": 0.0009261958146487294, + "loss": 3.3798, + "step": 8690 + }, + { + "epoch": 0.5907732028808262, + "grad_norm": 0.6653010249137878, + "learning_rate": 0.0009261533496398968, + "loss": 3.654, + "step": 8695 + }, + { + "epoch": 0.591112922951488, + "grad_norm": 0.7774918675422668, + "learning_rate": 0.0009261108846310641, + "loss": 3.7007, + "step": 8700 + }, + { + "epoch": 0.5914526430221497, + "grad_norm": 0.7363237738609314, + "learning_rate": 0.0009260684196222313, + "loss": 3.8055, + "step": 8705 + }, + { + "epoch": 0.5917923630928115, + "grad_norm": 0.7885435819625854, + "learning_rate": 0.0009260259546133986, + "loss": 3.6456, + "step": 8710 + }, + { + "epoch": 0.5921320831634733, + "grad_norm": 0.8036079406738281, + "learning_rate": 0.0009259834896045659, + "loss": 3.53, + "step": 8715 + }, + { + "epoch": 0.5924718032341351, + "grad_norm": 0.9873092770576477, + "learning_rate": 0.0009259410245957331, + "loss": 3.719, + "step": 8720 + }, + { + "epoch": 0.5928115233047968, + "grad_norm": 0.7451286911964417, + "learning_rate": 0.0009258985595869003, + "loss": 3.7688, + "step": 8725 + }, + { + "epoch": 0.5931512433754587, + "grad_norm": 0.8148131370544434, + "learning_rate": 0.0009258560945780678, + "loss": 3.8162, + "step": 8730 + }, + { + "epoch": 0.5934909634461204, + "grad_norm": 0.9994235038757324, + "learning_rate": 0.000925813629569235, + "loss": 3.7565, + "step": 8735 + }, + { + "epoch": 0.5938306835167821, + "grad_norm": 0.7480568885803223, + "learning_rate": 0.0009257711645604022, + "loss": 3.6426, + "step": 8740 + }, + { + "epoch": 0.594170403587444, + "grad_norm": 0.7387261986732483, + "learning_rate": 0.0009257286995515696, + "loss": 3.7159, + "step": 8745 + }, + { + "epoch": 0.5945101236581057, + "grad_norm": 0.7426326274871826, + "learning_rate": 0.0009256862345427368, + "loss": 3.7695, + "step": 8750 + }, + { + "epoch": 0.5948498437287675, + "grad_norm": 0.6524364948272705, + "learning_rate": 0.000925643769533904, + "loss": 3.5758, + "step": 8755 + }, + { + "epoch": 0.5951895637994292, + "grad_norm": 1.2406293153762817, + "learning_rate": 0.0009256013045250714, + "loss": 3.7243, + "step": 8760 + }, + { + "epoch": 0.5955292838700911, + "grad_norm": 0.9576144218444824, + "learning_rate": 0.0009255588395162387, + "loss": 3.6345, + "step": 8765 + }, + { + "epoch": 0.5958690039407528, + "grad_norm": 0.6558308601379395, + "learning_rate": 0.0009255163745074059, + "loss": 3.7384, + "step": 8770 + }, + { + "epoch": 0.5962087240114146, + "grad_norm": 0.9128443598747253, + "learning_rate": 0.0009254739094985732, + "loss": 3.6394, + "step": 8775 + }, + { + "epoch": 0.5965484440820764, + "grad_norm": 0.6811715364456177, + "learning_rate": 0.0009254314444897405, + "loss": 3.6862, + "step": 8780 + }, + { + "epoch": 0.5968881641527382, + "grad_norm": 0.6951560974121094, + "learning_rate": 0.0009253889794809077, + "loss": 3.6192, + "step": 8785 + }, + { + "epoch": 0.5972278842233999, + "grad_norm": 0.726325273513794, + "learning_rate": 0.000925346514472075, + "loss": 3.769, + "step": 8790 + }, + { + "epoch": 0.5975676042940616, + "grad_norm": 0.6010936498641968, + "learning_rate": 0.0009253040494632423, + "loss": 3.7858, + "step": 8795 + }, + { + "epoch": 0.5979073243647235, + "grad_norm": 0.6359979510307312, + "learning_rate": 0.0009252615844544096, + "loss": 3.6635, + "step": 8800 + }, + { + "epoch": 0.5982470444353852, + "grad_norm": 0.7707550525665283, + "learning_rate": 0.0009252191194455769, + "loss": 3.6817, + "step": 8805 + }, + { + "epoch": 0.598586764506047, + "grad_norm": 0.7625561952590942, + "learning_rate": 0.0009251766544367442, + "loss": 3.7249, + "step": 8810 + }, + { + "epoch": 0.5989264845767088, + "grad_norm": 0.9263338446617126, + "learning_rate": 0.0009251341894279114, + "loss": 3.8066, + "step": 8815 + }, + { + "epoch": 0.5992662046473706, + "grad_norm": 0.8002027273178101, + "learning_rate": 0.0009250917244190787, + "loss": 3.4509, + "step": 8820 + }, + { + "epoch": 0.5996059247180323, + "grad_norm": 0.967241108417511, + "learning_rate": 0.0009250492594102459, + "loss": 3.63, + "step": 8825 + }, + { + "epoch": 0.5999456447886942, + "grad_norm": 0.7122082114219666, + "learning_rate": 0.0009250067944014132, + "loss": 3.5557, + "step": 8830 + }, + { + "epoch": 0.6002853648593559, + "grad_norm": 0.6507645845413208, + "learning_rate": 0.0009249643293925806, + "loss": 3.6562, + "step": 8835 + }, + { + "epoch": 0.6006250849300176, + "grad_norm": 2.508281707763672, + "learning_rate": 0.0009249218643837478, + "loss": 3.9142, + "step": 8840 + }, + { + "epoch": 0.6009648050006794, + "grad_norm": 0.606696605682373, + "learning_rate": 0.0009248793993749151, + "loss": 3.6453, + "step": 8845 + }, + { + "epoch": 0.6013045250713412, + "grad_norm": 0.7018161416053772, + "learning_rate": 0.0009248369343660824, + "loss": 3.8992, + "step": 8850 + }, + { + "epoch": 0.601644245142003, + "grad_norm": 0.7630243301391602, + "learning_rate": 0.0009247944693572496, + "loss": 3.7281, + "step": 8855 + }, + { + "epoch": 0.6019839652126647, + "grad_norm": 0.6921541690826416, + "learning_rate": 0.0009247520043484168, + "loss": 3.6395, + "step": 8860 + }, + { + "epoch": 0.6023236852833266, + "grad_norm": 0.6790416836738586, + "learning_rate": 0.0009247095393395843, + "loss": 3.601, + "step": 8865 + }, + { + "epoch": 0.6026634053539883, + "grad_norm": 0.7225310206413269, + "learning_rate": 0.0009246670743307515, + "loss": 3.6465, + "step": 8870 + }, + { + "epoch": 0.6030031254246501, + "grad_norm": 0.8191479444503784, + "learning_rate": 0.0009246246093219187, + "loss": 3.6612, + "step": 8875 + }, + { + "epoch": 0.6033428454953118, + "grad_norm": 0.8516916632652283, + "learning_rate": 0.0009245821443130861, + "loss": 3.4985, + "step": 8880 + }, + { + "epoch": 0.6036825655659737, + "grad_norm": 0.763337254524231, + "learning_rate": 0.0009245396793042533, + "loss": 3.8809, + "step": 8885 + }, + { + "epoch": 0.6040222856366354, + "grad_norm": 0.6327252388000488, + "learning_rate": 0.0009244972142954205, + "loss": 3.6463, + "step": 8890 + }, + { + "epoch": 0.6043620057072971, + "grad_norm": 0.7968856692314148, + "learning_rate": 0.0009244547492865879, + "loss": 3.6589, + "step": 8895 + }, + { + "epoch": 0.604701725777959, + "grad_norm": 0.7092478275299072, + "learning_rate": 0.0009244122842777552, + "loss": 3.8094, + "step": 8900 + }, + { + "epoch": 0.6050414458486207, + "grad_norm": 0.9345669150352478, + "learning_rate": 0.0009243698192689224, + "loss": 3.5164, + "step": 8905 + }, + { + "epoch": 0.6053811659192825, + "grad_norm": 0.6484115719795227, + "learning_rate": 0.0009243273542600898, + "loss": 3.5994, + "step": 8910 + }, + { + "epoch": 0.6057208859899443, + "grad_norm": 1.0086615085601807, + "learning_rate": 0.000924284889251257, + "loss": 3.5933, + "step": 8915 + }, + { + "epoch": 0.6060606060606061, + "grad_norm": 0.5958741903305054, + "learning_rate": 0.0009242424242424242, + "loss": 3.7198, + "step": 8920 + }, + { + "epoch": 0.6064003261312678, + "grad_norm": 0.696983814239502, + "learning_rate": 0.0009241999592335915, + "loss": 3.5835, + "step": 8925 + }, + { + "epoch": 0.6067400462019296, + "grad_norm": 0.8986988663673401, + "learning_rate": 0.0009241574942247588, + "loss": 3.821, + "step": 8930 + }, + { + "epoch": 0.6070797662725914, + "grad_norm": 1.4587080478668213, + "learning_rate": 0.0009241150292159261, + "loss": 3.6971, + "step": 8935 + }, + { + "epoch": 0.6074194863432532, + "grad_norm": 0.6396378874778748, + "learning_rate": 0.0009240725642070934, + "loss": 3.6322, + "step": 8940 + }, + { + "epoch": 0.6077592064139149, + "grad_norm": 0.9018657803535461, + "learning_rate": 0.0009240300991982607, + "loss": 3.7441, + "step": 8945 + }, + { + "epoch": 0.6080989264845768, + "grad_norm": 0.871687114238739, + "learning_rate": 0.0009239876341894279, + "loss": 3.8863, + "step": 8950 + }, + { + "epoch": 0.6084386465552385, + "grad_norm": 1.0030956268310547, + "learning_rate": 0.0009239451691805952, + "loss": 3.6327, + "step": 8955 + }, + { + "epoch": 0.6087783666259002, + "grad_norm": 0.700885534286499, + "learning_rate": 0.0009239027041717624, + "loss": 3.7696, + "step": 8960 + }, + { + "epoch": 0.609118086696562, + "grad_norm": 0.677375853061676, + "learning_rate": 0.0009238602391629297, + "loss": 3.4678, + "step": 8965 + }, + { + "epoch": 0.6094578067672238, + "grad_norm": 0.6456116437911987, + "learning_rate": 0.0009238177741540971, + "loss": 3.7062, + "step": 8970 + }, + { + "epoch": 0.6097975268378856, + "grad_norm": 0.7838855385780334, + "learning_rate": 0.0009237753091452643, + "loss": 3.609, + "step": 8975 + }, + { + "epoch": 0.6101372469085473, + "grad_norm": 1.1101454496383667, + "learning_rate": 0.0009237328441364316, + "loss": 3.5471, + "step": 8980 + }, + { + "epoch": 0.6104769669792092, + "grad_norm": 0.7677356004714966, + "learning_rate": 0.0009236903791275989, + "loss": 3.4993, + "step": 8985 + }, + { + "epoch": 0.6108166870498709, + "grad_norm": 0.6281837224960327, + "learning_rate": 0.0009236479141187661, + "loss": 3.6054, + "step": 8990 + }, + { + "epoch": 0.6111564071205327, + "grad_norm": 1.1245094537734985, + "learning_rate": 0.0009236054491099334, + "loss": 3.9832, + "step": 8995 + }, + { + "epoch": 0.6114961271911945, + "grad_norm": 0.674760103225708, + "learning_rate": 0.0009235629841011007, + "loss": 3.6782, + "step": 9000 + }, + { + "epoch": 0.6118358472618562, + "grad_norm": 1.299085259437561, + "learning_rate": 0.000923520519092268, + "loss": 3.8441, + "step": 9005 + }, + { + "epoch": 0.612175567332518, + "grad_norm": 0.7997239232063293, + "learning_rate": 0.0009234780540834353, + "loss": 3.6219, + "step": 9010 + }, + { + "epoch": 0.6125152874031797, + "grad_norm": 0.7059243321418762, + "learning_rate": 0.0009234355890746026, + "loss": 3.6765, + "step": 9015 + }, + { + "epoch": 0.6128550074738416, + "grad_norm": 0.7399749159812927, + "learning_rate": 0.0009233931240657698, + "loss": 3.7717, + "step": 9020 + }, + { + "epoch": 0.6131947275445033, + "grad_norm": 0.9513211250305176, + "learning_rate": 0.000923350659056937, + "loss": 3.5431, + "step": 9025 + }, + { + "epoch": 0.6135344476151651, + "grad_norm": 0.8005008101463318, + "learning_rate": 0.0009233081940481044, + "loss": 3.611, + "step": 9030 + }, + { + "epoch": 0.6138741676858269, + "grad_norm": 0.8038389086723328, + "learning_rate": 0.0009232657290392716, + "loss": 3.5122, + "step": 9035 + }, + { + "epoch": 0.6142138877564887, + "grad_norm": 1.1867358684539795, + "learning_rate": 0.000923223264030439, + "loss": 3.8666, + "step": 9040 + }, + { + "epoch": 0.6145536078271504, + "grad_norm": 0.8052146434783936, + "learning_rate": 0.0009231807990216063, + "loss": 3.829, + "step": 9045 + }, + { + "epoch": 0.6148933278978121, + "grad_norm": 0.7490555644035339, + "learning_rate": 0.0009231383340127735, + "loss": 3.7249, + "step": 9050 + }, + { + "epoch": 0.615233047968474, + "grad_norm": 0.8196632266044617, + "learning_rate": 0.0009230958690039408, + "loss": 3.8551, + "step": 9055 + }, + { + "epoch": 0.6155727680391357, + "grad_norm": 0.7509273290634155, + "learning_rate": 0.000923053403995108, + "loss": 3.7717, + "step": 9060 + }, + { + "epoch": 0.6159124881097975, + "grad_norm": 0.6380864381790161, + "learning_rate": 0.0009230109389862753, + "loss": 3.8173, + "step": 9065 + }, + { + "epoch": 0.6162522081804593, + "grad_norm": 0.6834045648574829, + "learning_rate": 0.0009229684739774426, + "loss": 3.5823, + "step": 9070 + }, + { + "epoch": 0.6165919282511211, + "grad_norm": 0.8201780915260315, + "learning_rate": 0.0009229260089686099, + "loss": 3.4853, + "step": 9075 + }, + { + "epoch": 0.6169316483217828, + "grad_norm": 0.6504944562911987, + "learning_rate": 0.0009228835439597772, + "loss": 3.7464, + "step": 9080 + }, + { + "epoch": 0.6172713683924447, + "grad_norm": 0.6998506784439087, + "learning_rate": 0.0009228410789509445, + "loss": 3.7638, + "step": 9085 + }, + { + "epoch": 0.6176110884631064, + "grad_norm": 0.6809467673301697, + "learning_rate": 0.0009227986139421117, + "loss": 3.4786, + "step": 9090 + }, + { + "epoch": 0.6179508085337682, + "grad_norm": 0.5954441428184509, + "learning_rate": 0.000922756148933279, + "loss": 3.637, + "step": 9095 + }, + { + "epoch": 0.6182905286044299, + "grad_norm": 0.8184671401977539, + "learning_rate": 0.0009227136839244463, + "loss": 3.7801, + "step": 9100 + }, + { + "epoch": 0.6186302486750918, + "grad_norm": 0.7983341217041016, + "learning_rate": 0.0009226712189156135, + "loss": 4.0072, + "step": 9105 + }, + { + "epoch": 0.6189699687457535, + "grad_norm": 0.8784645795822144, + "learning_rate": 0.0009226287539067809, + "loss": 3.256, + "step": 9110 + }, + { + "epoch": 0.6193096888164152, + "grad_norm": 0.7210829854011536, + "learning_rate": 0.0009225862888979482, + "loss": 3.8032, + "step": 9115 + }, + { + "epoch": 0.6196494088870771, + "grad_norm": 0.6914405226707458, + "learning_rate": 0.0009225438238891154, + "loss": 3.7015, + "step": 9120 + }, + { + "epoch": 0.6199891289577388, + "grad_norm": 0.7073154449462891, + "learning_rate": 0.0009225013588802826, + "loss": 3.7968, + "step": 9125 + }, + { + "epoch": 0.6203288490284006, + "grad_norm": 0.8103522658348083, + "learning_rate": 0.00092245889387145, + "loss": 3.8083, + "step": 9130 + }, + { + "epoch": 0.6206685690990623, + "grad_norm": 0.6959781646728516, + "learning_rate": 0.0009224164288626172, + "loss": 3.8596, + "step": 9135 + }, + { + "epoch": 0.6210082891697242, + "grad_norm": 0.6005171537399292, + "learning_rate": 0.0009223739638537844, + "loss": 3.5595, + "step": 9140 + }, + { + "epoch": 0.6213480092403859, + "grad_norm": 0.7910379767417908, + "learning_rate": 0.0009223314988449519, + "loss": 3.8725, + "step": 9145 + }, + { + "epoch": 0.6216877293110477, + "grad_norm": 0.6003121733665466, + "learning_rate": 0.0009222890338361191, + "loss": 3.5523, + "step": 9150 + }, + { + "epoch": 0.6220274493817095, + "grad_norm": 0.6023795008659363, + "learning_rate": 0.0009222465688272863, + "loss": 3.6321, + "step": 9155 + }, + { + "epoch": 0.6223671694523712, + "grad_norm": 0.6355746984481812, + "learning_rate": 0.0009222041038184537, + "loss": 3.7205, + "step": 9160 + }, + { + "epoch": 0.622706889523033, + "grad_norm": 0.7739025354385376, + "learning_rate": 0.0009221616388096209, + "loss": 3.61, + "step": 9165 + }, + { + "epoch": 0.6230466095936948, + "grad_norm": 0.8389310836791992, + "learning_rate": 0.0009221191738007881, + "loss": 3.6685, + "step": 9170 + }, + { + "epoch": 0.6233863296643566, + "grad_norm": 0.8171815872192383, + "learning_rate": 0.0009220767087919554, + "loss": 3.6865, + "step": 9175 + }, + { + "epoch": 0.6237260497350183, + "grad_norm": 1.1299707889556885, + "learning_rate": 0.0009220342437831228, + "loss": 3.373, + "step": 9180 + }, + { + "epoch": 0.6240657698056801, + "grad_norm": 0.6047030687332153, + "learning_rate": 0.00092199177877429, + "loss": 3.9159, + "step": 9185 + }, + { + "epoch": 0.6244054898763419, + "grad_norm": 0.719158411026001, + "learning_rate": 0.0009219493137654573, + "loss": 3.8218, + "step": 9190 + }, + { + "epoch": 0.6247452099470037, + "grad_norm": 0.7280524969100952, + "learning_rate": 0.0009219068487566246, + "loss": 3.8482, + "step": 9195 + }, + { + "epoch": 0.6250849300176654, + "grad_norm": 0.8552801609039307, + "learning_rate": 0.0009218643837477918, + "loss": 3.538, + "step": 9200 + }, + { + "epoch": 0.6254246500883273, + "grad_norm": 0.7316009998321533, + "learning_rate": 0.0009218219187389591, + "loss": 3.7324, + "step": 9205 + }, + { + "epoch": 0.625764370158989, + "grad_norm": 1.36768639087677, + "learning_rate": 0.0009217794537301263, + "loss": 3.9895, + "step": 9210 + }, + { + "epoch": 0.6261040902296507, + "grad_norm": 0.848290741443634, + "learning_rate": 0.0009217369887212937, + "loss": 3.7114, + "step": 9215 + }, + { + "epoch": 0.6264438103003126, + "grad_norm": 0.7369062900543213, + "learning_rate": 0.000921694523712461, + "loss": 3.8254, + "step": 9220 + }, + { + "epoch": 0.6267835303709743, + "grad_norm": 0.7685941457748413, + "learning_rate": 0.0009216520587036282, + "loss": 4.1522, + "step": 9225 + }, + { + "epoch": 0.6271232504416361, + "grad_norm": 0.9110896587371826, + "learning_rate": 0.0009216095936947955, + "loss": 3.6232, + "step": 9230 + }, + { + "epoch": 0.6274629705122978, + "grad_norm": 0.6928056478500366, + "learning_rate": 0.0009215671286859628, + "loss": 3.6709, + "step": 9235 + }, + { + "epoch": 0.6278026905829597, + "grad_norm": 0.7655983567237854, + "learning_rate": 0.00092152466367713, + "loss": 3.7193, + "step": 9240 + }, + { + "epoch": 0.6281424106536214, + "grad_norm": 0.7509453892707825, + "learning_rate": 0.0009214821986682973, + "loss": 3.7139, + "step": 9245 + }, + { + "epoch": 0.6284821307242832, + "grad_norm": 0.9849753379821777, + "learning_rate": 0.0009214397336594647, + "loss": 3.5676, + "step": 9250 + }, + { + "epoch": 0.628821850794945, + "grad_norm": 0.9442876577377319, + "learning_rate": 0.0009213972686506319, + "loss": 3.4036, + "step": 9255 + }, + { + "epoch": 0.6291615708656068, + "grad_norm": 0.6690783500671387, + "learning_rate": 0.0009213548036417991, + "loss": 3.683, + "step": 9260 + }, + { + "epoch": 0.6295012909362685, + "grad_norm": 0.6942455768585205, + "learning_rate": 0.0009213123386329665, + "loss": 3.5282, + "step": 9265 + }, + { + "epoch": 0.6298410110069302, + "grad_norm": 0.9975630044937134, + "learning_rate": 0.0009212698736241337, + "loss": 3.5586, + "step": 9270 + }, + { + "epoch": 0.6301807310775921, + "grad_norm": 0.811707615852356, + "learning_rate": 0.0009212274086153009, + "loss": 3.6936, + "step": 9275 + }, + { + "epoch": 0.6305204511482538, + "grad_norm": 0.7118427157402039, + "learning_rate": 0.0009211849436064683, + "loss": 3.832, + "step": 9280 + }, + { + "epoch": 0.6308601712189156, + "grad_norm": 0.7813193202018738, + "learning_rate": 0.0009211424785976356, + "loss": 3.8221, + "step": 9285 + }, + { + "epoch": 0.6311998912895774, + "grad_norm": 0.8137817978858948, + "learning_rate": 0.0009211000135888028, + "loss": 3.308, + "step": 9290 + }, + { + "epoch": 0.6315396113602392, + "grad_norm": 0.741693377494812, + "learning_rate": 0.0009210575485799702, + "loss": 3.6189, + "step": 9295 + }, + { + "epoch": 0.6318793314309009, + "grad_norm": 0.9716556668281555, + "learning_rate": 0.0009210150835711374, + "loss": 3.8412, + "step": 9300 + }, + { + "epoch": 0.6322190515015628, + "grad_norm": 0.7751227617263794, + "learning_rate": 0.0009209726185623046, + "loss": 3.6609, + "step": 9305 + }, + { + "epoch": 0.6325587715722245, + "grad_norm": 0.6161702871322632, + "learning_rate": 0.000920930153553472, + "loss": 3.5341, + "step": 9310 + }, + { + "epoch": 0.6328984916428863, + "grad_norm": 1.1263669729232788, + "learning_rate": 0.0009208876885446392, + "loss": 3.5529, + "step": 9315 + }, + { + "epoch": 0.633238211713548, + "grad_norm": 0.7302712798118591, + "learning_rate": 0.0009208452235358065, + "loss": 3.5646, + "step": 9320 + }, + { + "epoch": 0.6335779317842098, + "grad_norm": 1.0156751871109009, + "learning_rate": 0.0009208027585269738, + "loss": 3.7378, + "step": 9325 + }, + { + "epoch": 0.6339176518548716, + "grad_norm": 0.7550612688064575, + "learning_rate": 0.0009207602935181411, + "loss": 3.8317, + "step": 9330 + }, + { + "epoch": 0.6342573719255333, + "grad_norm": 0.9245972037315369, + "learning_rate": 0.0009207178285093083, + "loss": 3.9464, + "step": 9335 + }, + { + "epoch": 0.6345970919961952, + "grad_norm": 0.7483823895454407, + "learning_rate": 0.0009206753635004756, + "loss": 3.8021, + "step": 9340 + }, + { + "epoch": 0.6349368120668569, + "grad_norm": 0.969968855381012, + "learning_rate": 0.0009206328984916429, + "loss": 3.6749, + "step": 9345 + }, + { + "epoch": 0.6352765321375187, + "grad_norm": 0.6859450340270996, + "learning_rate": 0.0009205904334828101, + "loss": 3.7359, + "step": 9350 + }, + { + "epoch": 0.6356162522081804, + "grad_norm": 0.5592966675758362, + "learning_rate": 0.0009205479684739775, + "loss": 3.6339, + "step": 9355 + }, + { + "epoch": 0.6359559722788423, + "grad_norm": 0.736538290977478, + "learning_rate": 0.0009205055034651447, + "loss": 3.6134, + "step": 9360 + }, + { + "epoch": 0.636295692349504, + "grad_norm": 1.0042273998260498, + "learning_rate": 0.000920463038456312, + "loss": 3.8557, + "step": 9365 + }, + { + "epoch": 0.6366354124201657, + "grad_norm": 0.577450156211853, + "learning_rate": 0.0009204205734474793, + "loss": 3.9904, + "step": 9370 + }, + { + "epoch": 0.6369751324908276, + "grad_norm": 0.8989750742912292, + "learning_rate": 0.0009203781084386465, + "loss": 3.6782, + "step": 9375 + }, + { + "epoch": 0.6373148525614893, + "grad_norm": 3.73222017288208, + "learning_rate": 0.0009203356434298139, + "loss": 3.6033, + "step": 9380 + }, + { + "epoch": 0.6376545726321511, + "grad_norm": 0.6811506748199463, + "learning_rate": 0.0009202931784209811, + "loss": 3.7267, + "step": 9385 + }, + { + "epoch": 0.6379942927028129, + "grad_norm": 0.7121829986572266, + "learning_rate": 0.0009202507134121484, + "loss": 3.6227, + "step": 9390 + }, + { + "epoch": 0.6383340127734747, + "grad_norm": 0.9353055357933044, + "learning_rate": 0.0009202082484033158, + "loss": 3.7228, + "step": 9395 + }, + { + "epoch": 0.6386737328441364, + "grad_norm": 0.759248673915863, + "learning_rate": 0.000920165783394483, + "loss": 3.6836, + "step": 9400 + }, + { + "epoch": 0.6390134529147982, + "grad_norm": 0.9816084504127502, + "learning_rate": 0.0009201233183856502, + "loss": 3.6267, + "step": 9405 + }, + { + "epoch": 0.63935317298546, + "grad_norm": 0.7951840162277222, + "learning_rate": 0.0009200808533768175, + "loss": 3.6687, + "step": 9410 + }, + { + "epoch": 0.6396928930561218, + "grad_norm": 0.8741581439971924, + "learning_rate": 0.0009200383883679848, + "loss": 3.5428, + "step": 9415 + }, + { + "epoch": 0.6400326131267835, + "grad_norm": 0.8632642030715942, + "learning_rate": 0.000919995923359152, + "loss": 3.6862, + "step": 9420 + }, + { + "epoch": 0.6403723331974454, + "grad_norm": 0.7152655124664307, + "learning_rate": 0.0009199534583503194, + "loss": 3.731, + "step": 9425 + }, + { + "epoch": 0.6407120532681071, + "grad_norm": 0.8840948343276978, + "learning_rate": 0.0009199109933414867, + "loss": 3.6414, + "step": 9430 + }, + { + "epoch": 0.6410517733387688, + "grad_norm": 1.1780478954315186, + "learning_rate": 0.0009198685283326539, + "loss": 3.5598, + "step": 9435 + }, + { + "epoch": 0.6413914934094306, + "grad_norm": 0.7209941744804382, + "learning_rate": 0.0009198260633238212, + "loss": 3.6987, + "step": 9440 + }, + { + "epoch": 0.6417312134800924, + "grad_norm": 0.813639223575592, + "learning_rate": 0.0009197835983149885, + "loss": 3.6755, + "step": 9445 + }, + { + "epoch": 0.6420709335507542, + "grad_norm": 0.7683884501457214, + "learning_rate": 0.0009197411333061557, + "loss": 3.6572, + "step": 9450 + }, + { + "epoch": 0.6424106536214159, + "grad_norm": 0.8849777579307556, + "learning_rate": 0.0009196986682973231, + "loss": 3.5845, + "step": 9455 + }, + { + "epoch": 0.6427503736920778, + "grad_norm": 0.8575959205627441, + "learning_rate": 0.0009196562032884903, + "loss": 3.7525, + "step": 9460 + }, + { + "epoch": 0.6430900937627395, + "grad_norm": 0.7876947522163391, + "learning_rate": 0.0009196137382796576, + "loss": 3.5424, + "step": 9465 + }, + { + "epoch": 0.6434298138334013, + "grad_norm": 0.8727036714553833, + "learning_rate": 0.0009195712732708249, + "loss": 3.6765, + "step": 9470 + }, + { + "epoch": 0.6437695339040631, + "grad_norm": 1.061108112335205, + "learning_rate": 0.0009195288082619921, + "loss": 3.8255, + "step": 9475 + }, + { + "epoch": 0.6441092539747248, + "grad_norm": 0.6705709099769592, + "learning_rate": 0.0009194863432531594, + "loss": 3.5106, + "step": 9480 + }, + { + "epoch": 0.6444489740453866, + "grad_norm": 0.7401744723320007, + "learning_rate": 0.0009194438782443267, + "loss": 4.1239, + "step": 9485 + }, + { + "epoch": 0.6447886941160483, + "grad_norm": 0.8474775552749634, + "learning_rate": 0.000919401413235494, + "loss": 3.6468, + "step": 9490 + }, + { + "epoch": 0.6451284141867102, + "grad_norm": 0.7192711234092712, + "learning_rate": 0.0009193589482266613, + "loss": 3.6743, + "step": 9495 + }, + { + "epoch": 0.6454681342573719, + "grad_norm": 0.6971704363822937, + "learning_rate": 0.0009193164832178286, + "loss": 3.5711, + "step": 9500 + }, + { + "epoch": 0.6458078543280337, + "grad_norm": 0.5742700099945068, + "learning_rate": 0.0009192740182089958, + "loss": 3.4929, + "step": 9505 + }, + { + "epoch": 0.6461475743986955, + "grad_norm": 0.8801677823066711, + "learning_rate": 0.000919231553200163, + "loss": 3.5568, + "step": 9510 + }, + { + "epoch": 0.6464872944693573, + "grad_norm": 0.7414723038673401, + "learning_rate": 0.0009191890881913304, + "loss": 3.788, + "step": 9515 + }, + { + "epoch": 0.646827014540019, + "grad_norm": 0.9319229125976562, + "learning_rate": 0.0009191466231824976, + "loss": 3.7875, + "step": 9520 + }, + { + "epoch": 0.6471667346106807, + "grad_norm": 0.8426511287689209, + "learning_rate": 0.0009191041581736649, + "loss": 3.7638, + "step": 9525 + }, + { + "epoch": 0.6475064546813426, + "grad_norm": 0.6494197249412537, + "learning_rate": 0.0009190616931648323, + "loss": 3.8266, + "step": 9530 + }, + { + "epoch": 0.6478461747520043, + "grad_norm": 0.7874410152435303, + "learning_rate": 0.0009190192281559995, + "loss": 3.6111, + "step": 9535 + }, + { + "epoch": 0.6481858948226661, + "grad_norm": 0.9318093061447144, + "learning_rate": 0.0009189767631471667, + "loss": 3.8188, + "step": 9540 + }, + { + "epoch": 0.6485256148933279, + "grad_norm": 1.0846035480499268, + "learning_rate": 0.0009189342981383341, + "loss": 3.7991, + "step": 9545 + }, + { + "epoch": 0.6488653349639897, + "grad_norm": 0.718104362487793, + "learning_rate": 0.0009188918331295013, + "loss": 3.8044, + "step": 9550 + }, + { + "epoch": 0.6492050550346514, + "grad_norm": 1.0362039804458618, + "learning_rate": 0.0009188493681206685, + "loss": 3.6264, + "step": 9555 + }, + { + "epoch": 0.6495447751053133, + "grad_norm": 0.7583540081977844, + "learning_rate": 0.000918806903111836, + "loss": 3.7273, + "step": 9560 + }, + { + "epoch": 0.649884495175975, + "grad_norm": 0.8064551949501038, + "learning_rate": 0.0009187644381030032, + "loss": 3.7833, + "step": 9565 + }, + { + "epoch": 0.6502242152466368, + "grad_norm": 0.7205809354782104, + "learning_rate": 0.0009187219730941704, + "loss": 3.6088, + "step": 9570 + }, + { + "epoch": 0.6505639353172985, + "grad_norm": 0.9827800393104553, + "learning_rate": 0.0009186795080853377, + "loss": 3.4921, + "step": 9575 + }, + { + "epoch": 0.6509036553879604, + "grad_norm": 0.726604163646698, + "learning_rate": 0.000918637043076505, + "loss": 3.6566, + "step": 9580 + }, + { + "epoch": 0.6512433754586221, + "grad_norm": 0.9542114734649658, + "learning_rate": 0.0009185945780676722, + "loss": 3.552, + "step": 9585 + }, + { + "epoch": 0.6515830955292838, + "grad_norm": 0.5727741122245789, + "learning_rate": 0.0009185521130588395, + "loss": 3.7092, + "step": 9590 + }, + { + "epoch": 0.6519228155999457, + "grad_norm": 0.868331789970398, + "learning_rate": 0.0009185096480500069, + "loss": 3.7067, + "step": 9595 + }, + { + "epoch": 0.6522625356706074, + "grad_norm": 0.9321011900901794, + "learning_rate": 0.0009184671830411741, + "loss": 3.5555, + "step": 9600 + }, + { + "epoch": 0.6526022557412692, + "grad_norm": 0.6808778047561646, + "learning_rate": 0.0009184247180323414, + "loss": 3.7372, + "step": 9605 + }, + { + "epoch": 0.6529419758119309, + "grad_norm": 1.1563987731933594, + "learning_rate": 0.0009183822530235086, + "loss": 3.7393, + "step": 9610 + }, + { + "epoch": 0.6532816958825928, + "grad_norm": 0.7172516584396362, + "learning_rate": 0.0009183397880146759, + "loss": 3.6702, + "step": 9615 + }, + { + "epoch": 0.6536214159532545, + "grad_norm": 0.8999543190002441, + "learning_rate": 0.0009182973230058432, + "loss": 3.6403, + "step": 9620 + }, + { + "epoch": 0.6539611360239163, + "grad_norm": 0.8441185355186462, + "learning_rate": 0.0009182548579970104, + "loss": 3.8175, + "step": 9625 + }, + { + "epoch": 0.6543008560945781, + "grad_norm": 3.018766164779663, + "learning_rate": 0.0009182123929881778, + "loss": 3.7066, + "step": 9630 + }, + { + "epoch": 0.6546405761652399, + "grad_norm": 0.7492625117301941, + "learning_rate": 0.0009181699279793451, + "loss": 3.6426, + "step": 9635 + }, + { + "epoch": 0.6549802962359016, + "grad_norm": 1.1565454006195068, + "learning_rate": 0.0009181274629705123, + "loss": 3.6956, + "step": 9640 + }, + { + "epoch": 0.6553200163065634, + "grad_norm": 0.693215548992157, + "learning_rate": 0.0009180849979616795, + "loss": 3.563, + "step": 9645 + }, + { + "epoch": 0.6556597363772252, + "grad_norm": 0.891947329044342, + "learning_rate": 0.0009180425329528469, + "loss": 3.2896, + "step": 9650 + }, + { + "epoch": 0.6559994564478869, + "grad_norm": 0.8147547841072083, + "learning_rate": 0.0009180000679440141, + "loss": 3.6974, + "step": 9655 + }, + { + "epoch": 0.6563391765185487, + "grad_norm": 0.7547804713249207, + "learning_rate": 0.0009179576029351813, + "loss": 4.0684, + "step": 9660 + }, + { + "epoch": 0.6566788965892105, + "grad_norm": 1.2950103282928467, + "learning_rate": 0.0009179151379263488, + "loss": 3.5927, + "step": 9665 + }, + { + "epoch": 0.6570186166598723, + "grad_norm": 0.9161685705184937, + "learning_rate": 0.000917872672917516, + "loss": 3.2841, + "step": 9670 + }, + { + "epoch": 0.657358336730534, + "grad_norm": 1.213436245918274, + "learning_rate": 0.0009178302079086832, + "loss": 3.4569, + "step": 9675 + }, + { + "epoch": 0.6576980568011959, + "grad_norm": 0.6643913388252258, + "learning_rate": 0.0009177877428998506, + "loss": 3.9584, + "step": 9680 + }, + { + "epoch": 0.6580377768718576, + "grad_norm": 1.6521121263504028, + "learning_rate": 0.0009177452778910178, + "loss": 3.4733, + "step": 9685 + }, + { + "epoch": 0.6583774969425193, + "grad_norm": 1.1565953493118286, + "learning_rate": 0.000917702812882185, + "loss": 3.4032, + "step": 9690 + }, + { + "epoch": 0.6587172170131811, + "grad_norm": 0.8639934659004211, + "learning_rate": 0.0009176603478733524, + "loss": 3.7065, + "step": 9695 + }, + { + "epoch": 0.6590569370838429, + "grad_norm": 1.243613839149475, + "learning_rate": 0.0009176178828645197, + "loss": 3.6787, + "step": 9700 + }, + { + "epoch": 0.6593966571545047, + "grad_norm": 2.7069647312164307, + "learning_rate": 0.0009175754178556869, + "loss": 3.658, + "step": 9705 + }, + { + "epoch": 0.6597363772251664, + "grad_norm": 0.6198696494102478, + "learning_rate": 0.0009175329528468542, + "loss": 3.6963, + "step": 9710 + }, + { + "epoch": 0.6600760972958283, + "grad_norm": 1.543363332748413, + "learning_rate": 0.0009174904878380215, + "loss": 3.6742, + "step": 9715 + }, + { + "epoch": 0.66041581736649, + "grad_norm": 0.6742938756942749, + "learning_rate": 0.0009174480228291888, + "loss": 4.1696, + "step": 9720 + }, + { + "epoch": 0.6607555374371518, + "grad_norm": 0.805629312992096, + "learning_rate": 0.000917405557820356, + "loss": 3.8163, + "step": 9725 + }, + { + "epoch": 0.6610952575078136, + "grad_norm": 0.829976499080658, + "learning_rate": 0.0009173630928115233, + "loss": 3.4341, + "step": 9730 + }, + { + "epoch": 0.6614349775784754, + "grad_norm": 0.6362728476524353, + "learning_rate": 0.0009173206278026907, + "loss": 3.5633, + "step": 9735 + }, + { + "epoch": 0.6617746976491371, + "grad_norm": 0.964539110660553, + "learning_rate": 0.0009172781627938579, + "loss": 3.6039, + "step": 9740 + }, + { + "epoch": 0.6621144177197988, + "grad_norm": 0.7047832012176514, + "learning_rate": 0.0009172356977850252, + "loss": 3.6552, + "step": 9745 + }, + { + "epoch": 0.6624541377904607, + "grad_norm": 0.9688942432403564, + "learning_rate": 0.0009171932327761925, + "loss": 3.7936, + "step": 9750 + }, + { + "epoch": 0.6627938578611224, + "grad_norm": 0.9333823323249817, + "learning_rate": 0.0009171507677673597, + "loss": 3.8512, + "step": 9755 + }, + { + "epoch": 0.6631335779317842, + "grad_norm": 6.650848865509033, + "learning_rate": 0.0009171083027585269, + "loss": 3.5858, + "step": 9760 + }, + { + "epoch": 0.663473298002446, + "grad_norm": 1.0969469547271729, + "learning_rate": 0.0009170658377496943, + "loss": 3.3736, + "step": 9765 + }, + { + "epoch": 0.6638130180731078, + "grad_norm": 0.7564310431480408, + "learning_rate": 0.0009170233727408616, + "loss": 3.8052, + "step": 9770 + }, + { + "epoch": 0.6641527381437695, + "grad_norm": 4.969074249267578, + "learning_rate": 0.0009169809077320288, + "loss": 3.4517, + "step": 9775 + }, + { + "epoch": 0.6644924582144313, + "grad_norm": 0.6969420909881592, + "learning_rate": 0.0009169384427231962, + "loss": 3.9266, + "step": 9780 + }, + { + "epoch": 0.6648321782850931, + "grad_norm": 0.6856746673583984, + "learning_rate": 0.0009168959777143634, + "loss": 3.882, + "step": 9785 + }, + { + "epoch": 0.6651718983557549, + "grad_norm": 1.7262550592422485, + "learning_rate": 0.0009168535127055306, + "loss": 3.6014, + "step": 9790 + }, + { + "epoch": 0.6655116184264166, + "grad_norm": 0.9507957100868225, + "learning_rate": 0.000916811047696698, + "loss": 3.4152, + "step": 9795 + }, + { + "epoch": 0.6658513384970784, + "grad_norm": 0.7823410630226135, + "learning_rate": 0.0009167685826878652, + "loss": 3.5076, + "step": 9800 + }, + { + "epoch": 0.6661910585677402, + "grad_norm": 0.7469034790992737, + "learning_rate": 0.0009167261176790325, + "loss": 3.6165, + "step": 9805 + }, + { + "epoch": 0.6665307786384019, + "grad_norm": 0.634499192237854, + "learning_rate": 0.0009166836526701998, + "loss": 3.7085, + "step": 9810 + }, + { + "epoch": 0.6668704987090638, + "grad_norm": 0.7854308485984802, + "learning_rate": 0.0009166411876613671, + "loss": 3.8342, + "step": 9815 + }, + { + "epoch": 0.6672102187797255, + "grad_norm": 3.4686992168426514, + "learning_rate": 0.0009165987226525343, + "loss": 3.9239, + "step": 9820 + }, + { + "epoch": 0.6675499388503873, + "grad_norm": 0.7814885377883911, + "learning_rate": 0.0009165562576437016, + "loss": 3.8134, + "step": 9825 + }, + { + "epoch": 0.667889658921049, + "grad_norm": 0.7047469019889832, + "learning_rate": 0.0009165137926348689, + "loss": 3.6561, + "step": 9830 + }, + { + "epoch": 0.6682293789917109, + "grad_norm": 0.8466688394546509, + "learning_rate": 0.0009164713276260361, + "loss": 3.6466, + "step": 9835 + }, + { + "epoch": 0.6685690990623726, + "grad_norm": 0.7483973503112793, + "learning_rate": 0.0009164288626172035, + "loss": 3.7207, + "step": 9840 + }, + { + "epoch": 0.6689088191330343, + "grad_norm": 8.970659255981445, + "learning_rate": 0.0009163863976083708, + "loss": 3.7573, + "step": 9845 + }, + { + "epoch": 0.6692485392036962, + "grad_norm": 0.7494022846221924, + "learning_rate": 0.000916343932599538, + "loss": 3.5658, + "step": 9850 + }, + { + "epoch": 0.6695882592743579, + "grad_norm": 0.968916654586792, + "learning_rate": 0.0009163014675907053, + "loss": 3.9541, + "step": 9855 + }, + { + "epoch": 0.6699279793450197, + "grad_norm": 0.6635299921035767, + "learning_rate": 0.0009162590025818725, + "loss": 3.801, + "step": 9860 + }, + { + "epoch": 0.6702676994156814, + "grad_norm": 1.0608104467391968, + "learning_rate": 0.0009162165375730398, + "loss": 3.8306, + "step": 9865 + }, + { + "epoch": 0.6706074194863433, + "grad_norm": 0.799565851688385, + "learning_rate": 0.0009161740725642071, + "loss": 3.5243, + "step": 9870 + }, + { + "epoch": 0.670947139557005, + "grad_norm": 1.1245368719100952, + "learning_rate": 0.0009161316075553744, + "loss": 4.0492, + "step": 9875 + }, + { + "epoch": 0.6712868596276668, + "grad_norm": 0.9857480525970459, + "learning_rate": 0.0009160891425465417, + "loss": 3.784, + "step": 9880 + }, + { + "epoch": 0.6716265796983286, + "grad_norm": 0.7099499106407166, + "learning_rate": 0.000916046677537709, + "loss": 3.8156, + "step": 9885 + }, + { + "epoch": 0.6719662997689904, + "grad_norm": 3.200536012649536, + "learning_rate": 0.0009160042125288762, + "loss": 3.6041, + "step": 9890 + }, + { + "epoch": 0.6723060198396521, + "grad_norm": 1.211044430732727, + "learning_rate": 0.0009159617475200434, + "loss": 3.5227, + "step": 9895 + }, + { + "epoch": 0.672645739910314, + "grad_norm": 0.7891578078269958, + "learning_rate": 0.0009159192825112108, + "loss": 3.7963, + "step": 9900 + }, + { + "epoch": 0.6729854599809757, + "grad_norm": 1.4241113662719727, + "learning_rate": 0.000915876817502378, + "loss": 3.4813, + "step": 9905 + }, + { + "epoch": 0.6733251800516374, + "grad_norm": 0.7094100713729858, + "learning_rate": 0.0009158343524935453, + "loss": 3.7674, + "step": 9910 + }, + { + "epoch": 0.6736649001222992, + "grad_norm": 0.7429136633872986, + "learning_rate": 0.0009157918874847127, + "loss": 3.7815, + "step": 9915 + }, + { + "epoch": 0.674004620192961, + "grad_norm": 0.8643985390663147, + "learning_rate": 0.0009157494224758799, + "loss": 3.71, + "step": 9920 + }, + { + "epoch": 0.6743443402636228, + "grad_norm": 0.8561012148857117, + "learning_rate": 0.0009157069574670471, + "loss": 3.5881, + "step": 9925 + }, + { + "epoch": 0.6746840603342845, + "grad_norm": 0.5853846669197083, + "learning_rate": 0.0009156644924582145, + "loss": 3.6732, + "step": 9930 + }, + { + "epoch": 0.6750237804049464, + "grad_norm": 0.9918604493141174, + "learning_rate": 0.0009156220274493817, + "loss": 3.7006, + "step": 9935 + }, + { + "epoch": 0.6753635004756081, + "grad_norm": 0.9428429007530212, + "learning_rate": 0.0009155795624405489, + "loss": 3.7501, + "step": 9940 + }, + { + "epoch": 0.6757032205462699, + "grad_norm": 0.7074365615844727, + "learning_rate": 0.0009155370974317164, + "loss": 3.6843, + "step": 9945 + }, + { + "epoch": 0.6760429406169316, + "grad_norm": 0.6907420754432678, + "learning_rate": 0.0009154946324228836, + "loss": 3.8587, + "step": 9950 + }, + { + "epoch": 0.6763826606875935, + "grad_norm": 0.8623411059379578, + "learning_rate": 0.0009154521674140508, + "loss": 3.5507, + "step": 9955 + }, + { + "epoch": 0.6767223807582552, + "grad_norm": 0.7827767729759216, + "learning_rate": 0.0009154097024052181, + "loss": 3.6811, + "step": 9960 + }, + { + "epoch": 0.6770621008289169, + "grad_norm": 0.7575980424880981, + "learning_rate": 0.0009153672373963854, + "loss": 3.7261, + "step": 9965 + }, + { + "epoch": 0.6774018208995788, + "grad_norm": 1.54831862449646, + "learning_rate": 0.0009153247723875526, + "loss": 3.586, + "step": 9970 + }, + { + "epoch": 0.6777415409702405, + "grad_norm": 1.6377944946289062, + "learning_rate": 0.00091528230737872, + "loss": 3.6452, + "step": 9975 + }, + { + "epoch": 0.6780812610409023, + "grad_norm": 0.7046222686767578, + "learning_rate": 0.0009152398423698873, + "loss": 3.6725, + "step": 9980 + }, + { + "epoch": 0.6784209811115641, + "grad_norm": 1.8921164274215698, + "learning_rate": 0.0009151973773610545, + "loss": 3.6119, + "step": 9985 + }, + { + "epoch": 0.6787607011822259, + "grad_norm": 0.6316943168640137, + "learning_rate": 0.0009151549123522218, + "loss": 3.6952, + "step": 9990 + }, + { + "epoch": 0.6791004212528876, + "grad_norm": 1.143775224685669, + "learning_rate": 0.000915112447343389, + "loss": 3.4317, + "step": 9995 + }, + { + "epoch": 0.6794401413235494, + "grad_norm": 0.7595008015632629, + "learning_rate": 0.0009150699823345563, + "loss": 3.5677, + "step": 10000 + }, + { + "epoch": 0.6797798613942112, + "grad_norm": 0.9786950349807739, + "learning_rate": 0.0009150275173257236, + "loss": 3.886, + "step": 10005 + }, + { + "epoch": 0.680119581464873, + "grad_norm": 0.9359726309776306, + "learning_rate": 0.0009149850523168909, + "loss": 3.7313, + "step": 10010 + }, + { + "epoch": 0.6804593015355347, + "grad_norm": 0.8002122044563293, + "learning_rate": 0.0009149425873080582, + "loss": 3.6052, + "step": 10015 + }, + { + "epoch": 0.6807990216061965, + "grad_norm": 1.1193161010742188, + "learning_rate": 0.0009149001222992255, + "loss": 3.5973, + "step": 10020 + }, + { + "epoch": 0.6811387416768583, + "grad_norm": 0.7547570466995239, + "learning_rate": 0.0009148576572903927, + "loss": 3.7883, + "step": 10025 + }, + { + "epoch": 0.68147846174752, + "grad_norm": 0.8184522986412048, + "learning_rate": 0.00091481519228156, + "loss": 3.5993, + "step": 10030 + }, + { + "epoch": 0.6818181818181818, + "grad_norm": 0.7300489544868469, + "learning_rate": 0.0009147727272727273, + "loss": 3.9503, + "step": 10035 + }, + { + "epoch": 0.6821579018888436, + "grad_norm": 0.6881535053253174, + "learning_rate": 0.0009147302622638945, + "loss": 3.6524, + "step": 10040 + }, + { + "epoch": 0.6824976219595054, + "grad_norm": 0.850186288356781, + "learning_rate": 0.0009146877972550618, + "loss": 3.5654, + "step": 10045 + }, + { + "epoch": 0.6828373420301671, + "grad_norm": 1.0091590881347656, + "learning_rate": 0.0009146453322462292, + "loss": 3.4787, + "step": 10050 + }, + { + "epoch": 0.683177062100829, + "grad_norm": 0.7153178453445435, + "learning_rate": 0.0009146028672373964, + "loss": 3.7557, + "step": 10055 + }, + { + "epoch": 0.6835167821714907, + "grad_norm": 0.7237783670425415, + "learning_rate": 0.0009145604022285637, + "loss": 3.863, + "step": 10060 + }, + { + "epoch": 0.6838565022421524, + "grad_norm": 0.7217298746109009, + "learning_rate": 0.000914517937219731, + "loss": 3.7342, + "step": 10065 + }, + { + "epoch": 0.6841962223128143, + "grad_norm": 0.6769093871116638, + "learning_rate": 0.0009144754722108982, + "loss": 3.9304, + "step": 10070 + }, + { + "epoch": 0.684535942383476, + "grad_norm": 1.0798230171203613, + "learning_rate": 0.0009144330072020655, + "loss": 3.3839, + "step": 10075 + }, + { + "epoch": 0.6848756624541378, + "grad_norm": 0.9155729413032532, + "learning_rate": 0.0009143905421932329, + "loss": 3.6162, + "step": 10080 + }, + { + "epoch": 0.6852153825247995, + "grad_norm": 0.8103668093681335, + "learning_rate": 0.0009143480771844001, + "loss": 3.7364, + "step": 10085 + }, + { + "epoch": 0.6855551025954614, + "grad_norm": 0.8245783448219299, + "learning_rate": 0.0009143056121755674, + "loss": 3.6082, + "step": 10090 + }, + { + "epoch": 0.6858948226661231, + "grad_norm": 0.9438191056251526, + "learning_rate": 0.0009142631471667346, + "loss": 3.6008, + "step": 10095 + }, + { + "epoch": 0.6862345427367849, + "grad_norm": 1.074062705039978, + "learning_rate": 0.0009142206821579019, + "loss": 3.7567, + "step": 10100 + }, + { + "epoch": 0.6865742628074467, + "grad_norm": 0.8823865056037903, + "learning_rate": 0.0009141782171490692, + "loss": 3.6116, + "step": 10105 + }, + { + "epoch": 0.6869139828781085, + "grad_norm": 0.7746458649635315, + "learning_rate": 0.0009141357521402364, + "loss": 3.6769, + "step": 10110 + }, + { + "epoch": 0.6872537029487702, + "grad_norm": 0.7455337643623352, + "learning_rate": 0.0009140932871314038, + "loss": 3.821, + "step": 10115 + }, + { + "epoch": 0.6875934230194319, + "grad_norm": 6.852608680725098, + "learning_rate": 0.0009140508221225711, + "loss": 3.4532, + "step": 10120 + }, + { + "epoch": 0.6879331430900938, + "grad_norm": 0.8582409620285034, + "learning_rate": 0.0009140083571137383, + "loss": 3.8375, + "step": 10125 + }, + { + "epoch": 0.6882728631607555, + "grad_norm": 0.8052330613136292, + "learning_rate": 0.0009139658921049056, + "loss": 3.5496, + "step": 10130 + }, + { + "epoch": 0.6886125832314173, + "grad_norm": 0.5799731612205505, + "learning_rate": 0.0009139234270960729, + "loss": 3.7185, + "step": 10135 + }, + { + "epoch": 0.6889523033020791, + "grad_norm": 0.8182925581932068, + "learning_rate": 0.0009138809620872401, + "loss": 3.5338, + "step": 10140 + }, + { + "epoch": 0.6892920233727409, + "grad_norm": 0.7465730905532837, + "learning_rate": 0.0009138384970784073, + "loss": 3.8362, + "step": 10145 + }, + { + "epoch": 0.6896317434434026, + "grad_norm": 0.8224843144416809, + "learning_rate": 0.0009137960320695748, + "loss": 3.7746, + "step": 10150 + }, + { + "epoch": 0.6899714635140645, + "grad_norm": 1.3773318529129028, + "learning_rate": 0.000913753567060742, + "loss": 3.6719, + "step": 10155 + }, + { + "epoch": 0.6903111835847262, + "grad_norm": 0.9082929491996765, + "learning_rate": 0.0009137111020519092, + "loss": 3.6289, + "step": 10160 + }, + { + "epoch": 0.690650903655388, + "grad_norm": 0.9704210758209229, + "learning_rate": 0.0009136686370430766, + "loss": 3.5949, + "step": 10165 + }, + { + "epoch": 0.6909906237260497, + "grad_norm": 0.8680222034454346, + "learning_rate": 0.0009136261720342438, + "loss": 3.6271, + "step": 10170 + }, + { + "epoch": 0.6913303437967115, + "grad_norm": 0.7385654449462891, + "learning_rate": 0.000913583707025411, + "loss": 3.5062, + "step": 10175 + }, + { + "epoch": 0.6916700638673733, + "grad_norm": 0.9444783926010132, + "learning_rate": 0.0009135412420165784, + "loss": 3.9319, + "step": 10180 + }, + { + "epoch": 0.692009783938035, + "grad_norm": 0.7116833329200745, + "learning_rate": 0.0009134987770077457, + "loss": 3.3171, + "step": 10185 + }, + { + "epoch": 0.6923495040086969, + "grad_norm": 0.748099148273468, + "learning_rate": 0.0009134563119989129, + "loss": 3.5612, + "step": 10190 + }, + { + "epoch": 0.6926892240793586, + "grad_norm": 1.1757534742355347, + "learning_rate": 0.0009134138469900802, + "loss": 3.7673, + "step": 10195 + }, + { + "epoch": 0.6930289441500204, + "grad_norm": 0.8531996607780457, + "learning_rate": 0.0009133713819812475, + "loss": 3.691, + "step": 10200 + }, + { + "epoch": 0.6933686642206821, + "grad_norm": 1.682693600654602, + "learning_rate": 0.0009133289169724147, + "loss": 3.4147, + "step": 10205 + }, + { + "epoch": 0.693708384291344, + "grad_norm": 1.2797712087631226, + "learning_rate": 0.000913286451963582, + "loss": 3.6056, + "step": 10210 + }, + { + "epoch": 0.6940481043620057, + "grad_norm": 0.814225971698761, + "learning_rate": 0.0009132439869547493, + "loss": 3.7924, + "step": 10215 + }, + { + "epoch": 0.6943878244326674, + "grad_norm": 4.302161693572998, + "learning_rate": 0.0009132015219459166, + "loss": 3.7744, + "step": 10220 + }, + { + "epoch": 0.6947275445033293, + "grad_norm": 1.2015944719314575, + "learning_rate": 0.0009131590569370839, + "loss": 3.7404, + "step": 10225 + }, + { + "epoch": 0.695067264573991, + "grad_norm": 0.7665482759475708, + "learning_rate": 0.0009131165919282512, + "loss": 3.7591, + "step": 10230 + }, + { + "epoch": 0.6954069846446528, + "grad_norm": 0.8855556845664978, + "learning_rate": 0.0009130741269194184, + "loss": 3.712, + "step": 10235 + }, + { + "epoch": 0.6957467047153146, + "grad_norm": 1.6043473482131958, + "learning_rate": 0.0009130316619105857, + "loss": 3.6952, + "step": 10240 + }, + { + "epoch": 0.6960864247859764, + "grad_norm": 0.7334523797035217, + "learning_rate": 0.0009129891969017529, + "loss": 3.8594, + "step": 10245 + }, + { + "epoch": 0.6964261448566381, + "grad_norm": 0.8549174666404724, + "learning_rate": 0.0009129467318929202, + "loss": 3.6672, + "step": 10250 + }, + { + "epoch": 0.6967658649272999, + "grad_norm": 1.082798719406128, + "learning_rate": 0.0009129042668840876, + "loss": 3.5842, + "step": 10255 + }, + { + "epoch": 0.6971055849979617, + "grad_norm": 1.079723596572876, + "learning_rate": 0.0009128618018752548, + "loss": 3.7358, + "step": 10260 + }, + { + "epoch": 0.6974453050686235, + "grad_norm": 0.9810878038406372, + "learning_rate": 0.0009128193368664221, + "loss": 3.6857, + "step": 10265 + }, + { + "epoch": 0.6977850251392852, + "grad_norm": 0.8088982105255127, + "learning_rate": 0.0009127768718575894, + "loss": 3.8377, + "step": 10270 + }, + { + "epoch": 0.698124745209947, + "grad_norm": 0.7503846883773804, + "learning_rate": 0.0009127344068487566, + "loss": 3.5365, + "step": 10275 + }, + { + "epoch": 0.6984644652806088, + "grad_norm": 0.9040347933769226, + "learning_rate": 0.0009126919418399238, + "loss": 3.6803, + "step": 10280 + }, + { + "epoch": 0.6988041853512705, + "grad_norm": 0.7197749018669128, + "learning_rate": 0.0009126494768310912, + "loss": 3.6487, + "step": 10285 + }, + { + "epoch": 0.6991439054219323, + "grad_norm": 0.8298256397247314, + "learning_rate": 0.0009126070118222585, + "loss": 3.5874, + "step": 10290 + }, + { + "epoch": 0.6994836254925941, + "grad_norm": 1.4128953218460083, + "learning_rate": 0.0009125645468134257, + "loss": 3.3925, + "step": 10295 + }, + { + "epoch": 0.6998233455632559, + "grad_norm": 0.8453262448310852, + "learning_rate": 0.0009125220818045931, + "loss": 3.6807, + "step": 10300 + }, + { + "epoch": 0.7001630656339176, + "grad_norm": 1.091887354850769, + "learning_rate": 0.0009124796167957603, + "loss": 3.6035, + "step": 10305 + }, + { + "epoch": 0.7005027857045795, + "grad_norm": 1.373945713043213, + "learning_rate": 0.0009124371517869275, + "loss": 3.5907, + "step": 10310 + }, + { + "epoch": 0.7008425057752412, + "grad_norm": 1.0452311038970947, + "learning_rate": 0.0009123946867780949, + "loss": 3.7165, + "step": 10315 + }, + { + "epoch": 0.701182225845903, + "grad_norm": 0.75932776927948, + "learning_rate": 0.0009123522217692621, + "loss": 3.5479, + "step": 10320 + }, + { + "epoch": 0.7015219459165648, + "grad_norm": 0.6936649680137634, + "learning_rate": 0.0009123097567604294, + "loss": 3.5975, + "step": 10325 + }, + { + "epoch": 0.7018616659872265, + "grad_norm": 0.9894376993179321, + "learning_rate": 0.0009122672917515968, + "loss": 3.6726, + "step": 10330 + }, + { + "epoch": 0.7022013860578883, + "grad_norm": 0.7210821509361267, + "learning_rate": 0.000912224826742764, + "loss": 3.6174, + "step": 10335 + }, + { + "epoch": 0.70254110612855, + "grad_norm": 0.6754486560821533, + "learning_rate": 0.0009121823617339312, + "loss": 3.5336, + "step": 10340 + }, + { + "epoch": 0.7028808261992119, + "grad_norm": 0.7861688137054443, + "learning_rate": 0.0009121398967250985, + "loss": 3.7957, + "step": 10345 + }, + { + "epoch": 0.7032205462698736, + "grad_norm": 0.6608583927154541, + "learning_rate": 0.0009120974317162658, + "loss": 3.653, + "step": 10350 + }, + { + "epoch": 0.7035602663405354, + "grad_norm": 0.7717050909996033, + "learning_rate": 0.000912054966707433, + "loss": 3.7493, + "step": 10355 + }, + { + "epoch": 0.7038999864111972, + "grad_norm": 1.1313576698303223, + "learning_rate": 0.0009120125016986004, + "loss": 3.9364, + "step": 10360 + }, + { + "epoch": 0.704239706481859, + "grad_norm": 0.8983592987060547, + "learning_rate": 0.0009119700366897677, + "loss": 3.4091, + "step": 10365 + }, + { + "epoch": 0.7045794265525207, + "grad_norm": 0.9312572479248047, + "learning_rate": 0.0009119275716809349, + "loss": 3.7145, + "step": 10370 + }, + { + "epoch": 0.7049191466231824, + "grad_norm": 3.520451307296753, + "learning_rate": 0.0009118851066721022, + "loss": 3.6161, + "step": 10375 + }, + { + "epoch": 0.7052588666938443, + "grad_norm": 0.654140055179596, + "learning_rate": 0.0009118426416632694, + "loss": 3.8116, + "step": 10380 + }, + { + "epoch": 0.705598586764506, + "grad_norm": 0.8536202907562256, + "learning_rate": 0.0009118001766544367, + "loss": 3.7052, + "step": 10385 + }, + { + "epoch": 0.7059383068351678, + "grad_norm": 0.7853626608848572, + "learning_rate": 0.000911757711645604, + "loss": 3.7022, + "step": 10390 + }, + { + "epoch": 0.7062780269058296, + "grad_norm": 0.7837145924568176, + "learning_rate": 0.0009117152466367713, + "loss": 3.7579, + "step": 10395 + }, + { + "epoch": 0.7066177469764914, + "grad_norm": 0.6652942299842834, + "learning_rate": 0.0009116727816279387, + "loss": 3.6918, + "step": 10400 + }, + { + "epoch": 0.7069574670471531, + "grad_norm": 0.782294750213623, + "learning_rate": 0.0009116303166191059, + "loss": 3.5485, + "step": 10405 + }, + { + "epoch": 0.707297187117815, + "grad_norm": 0.6917309761047363, + "learning_rate": 0.0009115878516102731, + "loss": 3.7118, + "step": 10410 + }, + { + "epoch": 0.7076369071884767, + "grad_norm": 0.7393103837966919, + "learning_rate": 0.0009115453866014405, + "loss": 3.7631, + "step": 10415 + }, + { + "epoch": 0.7079766272591385, + "grad_norm": 1.7117981910705566, + "learning_rate": 0.0009115029215926077, + "loss": 3.8354, + "step": 10420 + }, + { + "epoch": 0.7083163473298002, + "grad_norm": 1.0539406538009644, + "learning_rate": 0.0009114604565837749, + "loss": 3.8607, + "step": 10425 + }, + { + "epoch": 0.708656067400462, + "grad_norm": 0.8765161633491516, + "learning_rate": 0.0009114179915749424, + "loss": 3.913, + "step": 10430 + }, + { + "epoch": 0.7089957874711238, + "grad_norm": 0.679240345954895, + "learning_rate": 0.0009113755265661096, + "loss": 3.8422, + "step": 10435 + }, + { + "epoch": 0.7093355075417855, + "grad_norm": 0.7008655667304993, + "learning_rate": 0.0009113330615572768, + "loss": 3.8569, + "step": 10440 + }, + { + "epoch": 0.7096752276124474, + "grad_norm": 0.6589189767837524, + "learning_rate": 0.0009112905965484441, + "loss": 3.8421, + "step": 10445 + }, + { + "epoch": 0.7100149476831091, + "grad_norm": 0.7388678193092346, + "learning_rate": 0.0009112481315396114, + "loss": 3.7708, + "step": 10450 + }, + { + "epoch": 0.7103546677537709, + "grad_norm": 0.7643890380859375, + "learning_rate": 0.0009112056665307786, + "loss": 3.8267, + "step": 10455 + }, + { + "epoch": 0.7106943878244326, + "grad_norm": 3.0079329013824463, + "learning_rate": 0.0009111632015219459, + "loss": 3.6751, + "step": 10460 + }, + { + "epoch": 0.7110341078950945, + "grad_norm": 1.0114420652389526, + "learning_rate": 0.0009111207365131133, + "loss": 3.6703, + "step": 10465 + }, + { + "epoch": 0.7113738279657562, + "grad_norm": 0.6157741546630859, + "learning_rate": 0.0009110782715042805, + "loss": 3.8127, + "step": 10470 + }, + { + "epoch": 0.711713548036418, + "grad_norm": 1.0686148405075073, + "learning_rate": 0.0009110358064954478, + "loss": 3.6019, + "step": 10475 + }, + { + "epoch": 0.7120532681070798, + "grad_norm": 0.7928467988967896, + "learning_rate": 0.000910993341486615, + "loss": 3.68, + "step": 10480 + }, + { + "epoch": 0.7123929881777415, + "grad_norm": 0.9108426570892334, + "learning_rate": 0.0009109508764777823, + "loss": 3.4585, + "step": 10485 + }, + { + "epoch": 0.7127327082484033, + "grad_norm": 0.7484140992164612, + "learning_rate": 0.0009109084114689496, + "loss": 3.7196, + "step": 10490 + }, + { + "epoch": 0.7130724283190651, + "grad_norm": 1.058139443397522, + "learning_rate": 0.0009108659464601168, + "loss": 3.6579, + "step": 10495 + }, + { + "epoch": 0.7134121483897269, + "grad_norm": 0.7690451741218567, + "learning_rate": 0.0009108234814512842, + "loss": 3.7067, + "step": 10500 + }, + { + "epoch": 0.7137518684603886, + "grad_norm": 0.8497024178504944, + "learning_rate": 0.0009107810164424515, + "loss": 3.7172, + "step": 10505 + }, + { + "epoch": 0.7140915885310504, + "grad_norm": 0.6333871483802795, + "learning_rate": 0.0009107385514336187, + "loss": 3.9435, + "step": 10510 + }, + { + "epoch": 0.7144313086017122, + "grad_norm": 0.7330854535102844, + "learning_rate": 0.000910696086424786, + "loss": 3.5719, + "step": 10515 + }, + { + "epoch": 0.714771028672374, + "grad_norm": 1.0492603778839111, + "learning_rate": 0.0009106536214159533, + "loss": 3.6217, + "step": 10520 + }, + { + "epoch": 0.7151107487430357, + "grad_norm": 1.119070291519165, + "learning_rate": 0.0009106111564071205, + "loss": 3.6961, + "step": 10525 + }, + { + "epoch": 0.7154504688136976, + "grad_norm": 0.6877725124359131, + "learning_rate": 0.0009105686913982877, + "loss": 3.5506, + "step": 10530 + }, + { + "epoch": 0.7157901888843593, + "grad_norm": 0.6053353548049927, + "learning_rate": 0.0009105262263894552, + "loss": 3.7111, + "step": 10535 + }, + { + "epoch": 0.716129908955021, + "grad_norm": 0.8503847122192383, + "learning_rate": 0.0009104837613806224, + "loss": 3.7108, + "step": 10540 + }, + { + "epoch": 0.7164696290256828, + "grad_norm": 0.8892068862915039, + "learning_rate": 0.0009104412963717896, + "loss": 3.5369, + "step": 10545 + }, + { + "epoch": 0.7168093490963446, + "grad_norm": 0.8155838251113892, + "learning_rate": 0.000910398831362957, + "loss": 3.7966, + "step": 10550 + }, + { + "epoch": 0.7171490691670064, + "grad_norm": 1.7614003419876099, + "learning_rate": 0.0009103563663541242, + "loss": 3.7736, + "step": 10555 + }, + { + "epoch": 0.7174887892376681, + "grad_norm": 0.640840470790863, + "learning_rate": 0.0009103139013452914, + "loss": 3.7568, + "step": 10560 + }, + { + "epoch": 0.71782850930833, + "grad_norm": 0.6072386503219604, + "learning_rate": 0.0009102714363364589, + "loss": 3.8138, + "step": 10565 + }, + { + "epoch": 0.7181682293789917, + "grad_norm": 0.8307803273200989, + "learning_rate": 0.0009102289713276261, + "loss": 3.901, + "step": 10570 + }, + { + "epoch": 0.7185079494496535, + "grad_norm": 1.1475845575332642, + "learning_rate": 0.0009101865063187933, + "loss": 3.6658, + "step": 10575 + }, + { + "epoch": 0.7188476695203153, + "grad_norm": 0.9262077808380127, + "learning_rate": 0.0009101440413099607, + "loss": 3.7299, + "step": 10580 + }, + { + "epoch": 0.719187389590977, + "grad_norm": 1.0430647134780884, + "learning_rate": 0.0009101015763011279, + "loss": 3.4612, + "step": 10585 + }, + { + "epoch": 0.7195271096616388, + "grad_norm": 0.7741380929946899, + "learning_rate": 0.0009100591112922951, + "loss": 3.8286, + "step": 10590 + }, + { + "epoch": 0.7198668297323005, + "grad_norm": 0.9774120450019836, + "learning_rate": 0.0009100166462834624, + "loss": 3.5167, + "step": 10595 + }, + { + "epoch": 0.7202065498029624, + "grad_norm": 0.7584857940673828, + "learning_rate": 0.0009099741812746298, + "loss": 3.5701, + "step": 10600 + }, + { + "epoch": 0.7205462698736241, + "grad_norm": 0.8413480520248413, + "learning_rate": 0.000909931716265797, + "loss": 3.8187, + "step": 10605 + }, + { + "epoch": 0.7208859899442859, + "grad_norm": 0.665310800075531, + "learning_rate": 0.0009098892512569643, + "loss": 3.7608, + "step": 10610 + }, + { + "epoch": 0.7212257100149477, + "grad_norm": 0.6017893552780151, + "learning_rate": 0.0009098467862481316, + "loss": 3.6939, + "step": 10615 + }, + { + "epoch": 0.7215654300856095, + "grad_norm": 0.7797799110412598, + "learning_rate": 0.0009098043212392988, + "loss": 3.6637, + "step": 10620 + }, + { + "epoch": 0.7219051501562712, + "grad_norm": 1.0015398263931274, + "learning_rate": 0.0009097618562304661, + "loss": 3.878, + "step": 10625 + }, + { + "epoch": 0.722244870226933, + "grad_norm": 0.9935557246208191, + "learning_rate": 0.0009097193912216333, + "loss": 3.7569, + "step": 10630 + }, + { + "epoch": 0.7225845902975948, + "grad_norm": 0.6865654587745667, + "learning_rate": 0.0009096769262128007, + "loss": 3.6062, + "step": 10635 + }, + { + "epoch": 0.7229243103682566, + "grad_norm": 3.791905641555786, + "learning_rate": 0.000909634461203968, + "loss": 3.6453, + "step": 10640 + }, + { + "epoch": 0.7232640304389183, + "grad_norm": 1.3526341915130615, + "learning_rate": 0.0009095919961951352, + "loss": 3.7675, + "step": 10645 + }, + { + "epoch": 0.7236037505095801, + "grad_norm": 1.0163742303848267, + "learning_rate": 0.0009095495311863025, + "loss": 3.5734, + "step": 10650 + }, + { + "epoch": 0.7239434705802419, + "grad_norm": 0.8221908807754517, + "learning_rate": 0.0009095070661774698, + "loss": 3.6041, + "step": 10655 + }, + { + "epoch": 0.7242831906509036, + "grad_norm": 0.7980278730392456, + "learning_rate": 0.000909464601168637, + "loss": 3.2872, + "step": 10660 + }, + { + "epoch": 0.7246229107215655, + "grad_norm": 0.786431610584259, + "learning_rate": 0.0009094221361598043, + "loss": 3.7339, + "step": 10665 + }, + { + "epoch": 0.7249626307922272, + "grad_norm": 0.7921804785728455, + "learning_rate": 0.0009093796711509717, + "loss": 3.6977, + "step": 10670 + }, + { + "epoch": 0.725302350862889, + "grad_norm": 0.8534802198410034, + "learning_rate": 0.0009093372061421389, + "loss": 3.778, + "step": 10675 + }, + { + "epoch": 0.7256420709335507, + "grad_norm": 0.73997563123703, + "learning_rate": 0.0009092947411333061, + "loss": 3.751, + "step": 10680 + }, + { + "epoch": 0.7259817910042126, + "grad_norm": 0.9463241100311279, + "learning_rate": 0.0009092522761244735, + "loss": 3.704, + "step": 10685 + }, + { + "epoch": 0.7263215110748743, + "grad_norm": 0.7088745832443237, + "learning_rate": 0.0009092098111156407, + "loss": 3.7575, + "step": 10690 + }, + { + "epoch": 0.726661231145536, + "grad_norm": 1.01829993724823, + "learning_rate": 0.0009091673461068079, + "loss": 3.7426, + "step": 10695 + }, + { + "epoch": 0.7270009512161979, + "grad_norm": 0.8043103218078613, + "learning_rate": 0.0009091248810979753, + "loss": 3.8787, + "step": 10700 + }, + { + "epoch": 0.7273406712868596, + "grad_norm": 0.8780677914619446, + "learning_rate": 0.0009090824160891426, + "loss": 3.7631, + "step": 10705 + }, + { + "epoch": 0.7276803913575214, + "grad_norm": 0.8562443256378174, + "learning_rate": 0.0009090399510803098, + "loss": 3.9021, + "step": 10710 + }, + { + "epoch": 0.7280201114281831, + "grad_norm": 0.6382040977478027, + "learning_rate": 0.0009089974860714772, + "loss": 3.8719, + "step": 10715 + }, + { + "epoch": 0.728359831498845, + "grad_norm": 1.00100839138031, + "learning_rate": 0.0009089550210626444, + "loss": 3.4644, + "step": 10720 + }, + { + "epoch": 0.7286995515695067, + "grad_norm": 1.4583836793899536, + "learning_rate": 0.0009089125560538116, + "loss": 3.7087, + "step": 10725 + }, + { + "epoch": 0.7290392716401685, + "grad_norm": 0.7512675523757935, + "learning_rate": 0.000908870091044979, + "loss": 3.5266, + "step": 10730 + }, + { + "epoch": 0.7293789917108303, + "grad_norm": 1.338925838470459, + "learning_rate": 0.0009088276260361462, + "loss": 3.8108, + "step": 10735 + }, + { + "epoch": 0.7297187117814921, + "grad_norm": 1.8903957605361938, + "learning_rate": 0.0009087851610273136, + "loss": 3.8596, + "step": 10740 + }, + { + "epoch": 0.7300584318521538, + "grad_norm": 0.746347188949585, + "learning_rate": 0.0009087426960184808, + "loss": 3.6057, + "step": 10745 + }, + { + "epoch": 0.7303981519228157, + "grad_norm": 0.9693640470504761, + "learning_rate": 0.0009087002310096481, + "loss": 3.695, + "step": 10750 + }, + { + "epoch": 0.7307378719934774, + "grad_norm": 0.7466190457344055, + "learning_rate": 0.0009086577660008154, + "loss": 3.7194, + "step": 10755 + }, + { + "epoch": 0.7310775920641391, + "grad_norm": 0.7932271957397461, + "learning_rate": 0.0009086153009919826, + "loss": 3.8626, + "step": 10760 + }, + { + "epoch": 0.7314173121348009, + "grad_norm": 1.12835693359375, + "learning_rate": 0.0009085728359831499, + "loss": 3.7529, + "step": 10765 + }, + { + "epoch": 0.7317570322054627, + "grad_norm": 0.8230791091918945, + "learning_rate": 0.0009085303709743172, + "loss": 3.5801, + "step": 10770 + }, + { + "epoch": 0.7320967522761245, + "grad_norm": 1.0379669666290283, + "learning_rate": 0.0009084879059654845, + "loss": 3.8839, + "step": 10775 + }, + { + "epoch": 0.7324364723467862, + "grad_norm": 0.6634736061096191, + "learning_rate": 0.0009084454409566517, + "loss": 3.6795, + "step": 10780 + }, + { + "epoch": 0.7327761924174481, + "grad_norm": 0.8374850153923035, + "learning_rate": 0.0009084029759478191, + "loss": 3.7551, + "step": 10785 + }, + { + "epoch": 0.7331159124881098, + "grad_norm": 0.8252646327018738, + "learning_rate": 0.0009083605109389863, + "loss": 3.5824, + "step": 10790 + }, + { + "epoch": 0.7334556325587716, + "grad_norm": 0.8450417518615723, + "learning_rate": 0.0009083180459301535, + "loss": 3.6611, + "step": 10795 + }, + { + "epoch": 0.7337953526294333, + "grad_norm": 0.8970220685005188, + "learning_rate": 0.0009082755809213209, + "loss": 3.6258, + "step": 10800 + }, + { + "epoch": 0.7341350727000951, + "grad_norm": 0.8585121631622314, + "learning_rate": 0.0009082331159124881, + "loss": 3.8088, + "step": 10805 + }, + { + "epoch": 0.7344747927707569, + "grad_norm": 0.7040704488754272, + "learning_rate": 0.0009081906509036554, + "loss": 3.8184, + "step": 10810 + }, + { + "epoch": 0.7348145128414186, + "grad_norm": 0.8545712828636169, + "learning_rate": 0.0009081481858948228, + "loss": 3.7435, + "step": 10815 + }, + { + "epoch": 0.7351542329120805, + "grad_norm": 0.6817483901977539, + "learning_rate": 0.00090810572088599, + "loss": 3.7205, + "step": 10820 + }, + { + "epoch": 0.7354939529827422, + "grad_norm": 0.802337646484375, + "learning_rate": 0.0009080632558771572, + "loss": 3.7991, + "step": 10825 + }, + { + "epoch": 0.735833673053404, + "grad_norm": 0.724261999130249, + "learning_rate": 0.0009080207908683245, + "loss": 3.5809, + "step": 10830 + }, + { + "epoch": 0.7361733931240658, + "grad_norm": 1.217659592628479, + "learning_rate": 0.0009079783258594918, + "loss": 3.7884, + "step": 10835 + }, + { + "epoch": 0.7365131131947276, + "grad_norm": 0.9047442078590393, + "learning_rate": 0.000907935860850659, + "loss": 3.5781, + "step": 10840 + }, + { + "epoch": 0.7368528332653893, + "grad_norm": 0.8488001823425293, + "learning_rate": 0.0009078933958418264, + "loss": 3.5537, + "step": 10845 + }, + { + "epoch": 0.737192553336051, + "grad_norm": 0.8094634413719177, + "learning_rate": 0.0009078509308329937, + "loss": 3.7588, + "step": 10850 + }, + { + "epoch": 0.7375322734067129, + "grad_norm": 1.5975488424301147, + "learning_rate": 0.0009078084658241609, + "loss": 3.6672, + "step": 10855 + }, + { + "epoch": 0.7378719934773746, + "grad_norm": 0.7606924772262573, + "learning_rate": 0.0009077660008153282, + "loss": 3.7268, + "step": 10860 + }, + { + "epoch": 0.7382117135480364, + "grad_norm": 0.639672040939331, + "learning_rate": 0.0009077235358064955, + "loss": 3.716, + "step": 10865 + }, + { + "epoch": 0.7385514336186982, + "grad_norm": 0.7177412509918213, + "learning_rate": 0.0009076810707976627, + "loss": 3.6089, + "step": 10870 + }, + { + "epoch": 0.73889115368936, + "grad_norm": 0.8626235127449036, + "learning_rate": 0.00090763860578883, + "loss": 3.7268, + "step": 10875 + }, + { + "epoch": 0.7392308737600217, + "grad_norm": 0.84871506690979, + "learning_rate": 0.0009075961407799973, + "loss": 3.725, + "step": 10880 + }, + { + "epoch": 0.7395705938306835, + "grad_norm": 0.6467830538749695, + "learning_rate": 0.0009075536757711646, + "loss": 3.6621, + "step": 10885 + }, + { + "epoch": 0.7399103139013453, + "grad_norm": 0.8289784789085388, + "learning_rate": 0.0009075112107623319, + "loss": 3.6479, + "step": 10890 + }, + { + "epoch": 0.7402500339720071, + "grad_norm": 0.7648937106132507, + "learning_rate": 0.0009074687457534991, + "loss": 3.6385, + "step": 10895 + }, + { + "epoch": 0.7405897540426688, + "grad_norm": 0.7217904925346375, + "learning_rate": 0.0009074262807446664, + "loss": 3.5578, + "step": 10900 + }, + { + "epoch": 0.7409294741133307, + "grad_norm": 0.7535264492034912, + "learning_rate": 0.0009073838157358337, + "loss": 3.7525, + "step": 10905 + }, + { + "epoch": 0.7412691941839924, + "grad_norm": 0.7339619398117065, + "learning_rate": 0.0009073413507270009, + "loss": 3.6714, + "step": 10910 + }, + { + "epoch": 0.7416089142546541, + "grad_norm": 0.7947880625724792, + "learning_rate": 0.0009072988857181683, + "loss": 3.965, + "step": 10915 + }, + { + "epoch": 0.741948634325316, + "grad_norm": 1.636953592300415, + "learning_rate": 0.0009072564207093356, + "loss": 3.617, + "step": 10920 + }, + { + "epoch": 0.7422883543959777, + "grad_norm": 1.1998404264450073, + "learning_rate": 0.0009072139557005028, + "loss": 3.7413, + "step": 10925 + }, + { + "epoch": 0.7426280744666395, + "grad_norm": 0.9806882739067078, + "learning_rate": 0.00090717149069167, + "loss": 3.6735, + "step": 10930 + }, + { + "epoch": 0.7429677945373012, + "grad_norm": 2.759023666381836, + "learning_rate": 0.0009071290256828374, + "loss": 3.5842, + "step": 10935 + }, + { + "epoch": 0.7433075146079631, + "grad_norm": 1.1711844205856323, + "learning_rate": 0.0009070865606740046, + "loss": 3.6398, + "step": 10940 + }, + { + "epoch": 0.7436472346786248, + "grad_norm": 0.7088755965232849, + "learning_rate": 0.0009070440956651718, + "loss": 3.6239, + "step": 10945 + }, + { + "epoch": 0.7439869547492866, + "grad_norm": 0.663390040397644, + "learning_rate": 0.0009070016306563393, + "loss": 4.053, + "step": 10950 + }, + { + "epoch": 0.7443266748199484, + "grad_norm": 0.7811562418937683, + "learning_rate": 0.0009069591656475065, + "loss": 3.4957, + "step": 10955 + }, + { + "epoch": 0.7446663948906102, + "grad_norm": 0.6137964129447937, + "learning_rate": 0.0009069167006386737, + "loss": 3.588, + "step": 10960 + }, + { + "epoch": 0.7450061149612719, + "grad_norm": 0.7414093613624573, + "learning_rate": 0.0009068742356298411, + "loss": 3.8597, + "step": 10965 + }, + { + "epoch": 0.7453458350319336, + "grad_norm": 0.860612690448761, + "learning_rate": 0.0009068317706210083, + "loss": 3.5892, + "step": 10970 + }, + { + "epoch": 0.7456855551025955, + "grad_norm": 1.3363772630691528, + "learning_rate": 0.0009067893056121755, + "loss": 3.7388, + "step": 10975 + }, + { + "epoch": 0.7460252751732572, + "grad_norm": 1.1836018562316895, + "learning_rate": 0.0009067468406033428, + "loss": 3.6186, + "step": 10980 + }, + { + "epoch": 0.746364995243919, + "grad_norm": 0.7892703413963318, + "learning_rate": 0.0009067043755945102, + "loss": 3.6968, + "step": 10985 + }, + { + "epoch": 0.7467047153145808, + "grad_norm": 0.7137872576713562, + "learning_rate": 0.0009066619105856774, + "loss": 3.5484, + "step": 10990 + }, + { + "epoch": 0.7470444353852426, + "grad_norm": 0.7693912982940674, + "learning_rate": 0.0009066194455768447, + "loss": 3.4623, + "step": 10995 + }, + { + "epoch": 0.7473841554559043, + "grad_norm": 0.6835779547691345, + "learning_rate": 0.000906576980568012, + "loss": 3.472, + "step": 11000 + }, + { + "epoch": 0.7477238755265662, + "grad_norm": 1.3486907482147217, + "learning_rate": 0.0009065345155591792, + "loss": 3.5823, + "step": 11005 + }, + { + "epoch": 0.7480635955972279, + "grad_norm": 1.3197927474975586, + "learning_rate": 0.0009064920505503465, + "loss": 3.6972, + "step": 11010 + }, + { + "epoch": 0.7484033156678896, + "grad_norm": 0.7653129696846008, + "learning_rate": 0.0009064495855415137, + "loss": 3.3573, + "step": 11015 + }, + { + "epoch": 0.7487430357385514, + "grad_norm": 0.832490086555481, + "learning_rate": 0.0009064071205326811, + "loss": 3.7542, + "step": 11020 + }, + { + "epoch": 0.7490827558092132, + "grad_norm": 0.7332544326782227, + "learning_rate": 0.0009063646555238484, + "loss": 3.6438, + "step": 11025 + }, + { + "epoch": 0.749422475879875, + "grad_norm": 1.0927499532699585, + "learning_rate": 0.0009063221905150156, + "loss": 3.6641, + "step": 11030 + }, + { + "epoch": 0.7497621959505367, + "grad_norm": 0.9052128791809082, + "learning_rate": 0.0009062797255061829, + "loss": 3.7397, + "step": 11035 + }, + { + "epoch": 0.7501019160211986, + "grad_norm": 0.7294290065765381, + "learning_rate": 0.0009062372604973502, + "loss": 3.8031, + "step": 11040 + }, + { + "epoch": 0.7504416360918603, + "grad_norm": 1.0512290000915527, + "learning_rate": 0.0009061947954885174, + "loss": 3.6971, + "step": 11045 + }, + { + "epoch": 0.7507813561625221, + "grad_norm": 3.3992373943328857, + "learning_rate": 0.0009061523304796847, + "loss": 3.7365, + "step": 11050 + }, + { + "epoch": 0.7511210762331838, + "grad_norm": 0.9170488119125366, + "learning_rate": 0.0009061098654708521, + "loss": 3.4703, + "step": 11055 + }, + { + "epoch": 0.7514607963038457, + "grad_norm": 0.775380551815033, + "learning_rate": 0.0009060674004620193, + "loss": 3.7283, + "step": 11060 + }, + { + "epoch": 0.7518005163745074, + "grad_norm": 1.198280930519104, + "learning_rate": 0.0009060249354531865, + "loss": 3.8769, + "step": 11065 + }, + { + "epoch": 0.7521402364451691, + "grad_norm": 2.0525169372558594, + "learning_rate": 0.0009059824704443539, + "loss": 3.5898, + "step": 11070 + }, + { + "epoch": 0.752479956515831, + "grad_norm": 0.7144297361373901, + "learning_rate": 0.0009059400054355211, + "loss": 3.7656, + "step": 11075 + }, + { + "epoch": 0.7528196765864927, + "grad_norm": 0.6905569434165955, + "learning_rate": 0.0009058975404266884, + "loss": 3.6303, + "step": 11080 + }, + { + "epoch": 0.7531593966571545, + "grad_norm": 0.9297271966934204, + "learning_rate": 0.0009058550754178557, + "loss": 3.5039, + "step": 11085 + }, + { + "epoch": 0.7534991167278163, + "grad_norm": 1.179353952407837, + "learning_rate": 0.000905812610409023, + "loss": 3.656, + "step": 11090 + }, + { + "epoch": 0.7538388367984781, + "grad_norm": 0.7236602306365967, + "learning_rate": 0.0009057701454001903, + "loss": 3.824, + "step": 11095 + }, + { + "epoch": 0.7541785568691398, + "grad_norm": 0.7393776774406433, + "learning_rate": 0.0009057276803913576, + "loss": 3.5526, + "step": 11100 + }, + { + "epoch": 0.7545182769398016, + "grad_norm": 0.6739901304244995, + "learning_rate": 0.0009056852153825248, + "loss": 3.7601, + "step": 11105 + }, + { + "epoch": 0.7548579970104634, + "grad_norm": 1.1672066450119019, + "learning_rate": 0.0009056427503736921, + "loss": 3.5944, + "step": 11110 + }, + { + "epoch": 0.7551977170811252, + "grad_norm": 0.7919332385063171, + "learning_rate": 0.0009056002853648593, + "loss": 3.6986, + "step": 11115 + }, + { + "epoch": 0.7555374371517869, + "grad_norm": 0.8015375137329102, + "learning_rate": 0.0009055578203560266, + "loss": 3.6726, + "step": 11120 + }, + { + "epoch": 0.7558771572224487, + "grad_norm": 0.6916100978851318, + "learning_rate": 0.000905515355347194, + "loss": 3.8511, + "step": 11125 + }, + { + "epoch": 0.7562168772931105, + "grad_norm": 0.7141310572624207, + "learning_rate": 0.0009054728903383612, + "loss": 3.7437, + "step": 11130 + }, + { + "epoch": 0.7565565973637722, + "grad_norm": 0.7443316578865051, + "learning_rate": 0.0009054304253295285, + "loss": 3.8995, + "step": 11135 + }, + { + "epoch": 0.756896317434434, + "grad_norm": 0.6934722661972046, + "learning_rate": 0.0009053879603206958, + "loss": 3.8128, + "step": 11140 + }, + { + "epoch": 0.7572360375050958, + "grad_norm": 0.9614686369895935, + "learning_rate": 0.000905345495311863, + "loss": 3.5188, + "step": 11145 + }, + { + "epoch": 0.7575757575757576, + "grad_norm": 1.4568780660629272, + "learning_rate": 0.0009053030303030303, + "loss": 3.9663, + "step": 11150 + }, + { + "epoch": 0.7579154776464193, + "grad_norm": 1.0191596746444702, + "learning_rate": 0.0009052605652941977, + "loss": 3.7326, + "step": 11155 + }, + { + "epoch": 0.7582551977170812, + "grad_norm": 0.8418881297111511, + "learning_rate": 0.0009052181002853649, + "loss": 3.6783, + "step": 11160 + }, + { + "epoch": 0.7585949177877429, + "grad_norm": 0.9203734993934631, + "learning_rate": 0.0009051756352765322, + "loss": 3.4804, + "step": 11165 + }, + { + "epoch": 0.7589346378584046, + "grad_norm": 0.8070054650306702, + "learning_rate": 0.0009051331702676995, + "loss": 3.5441, + "step": 11170 + }, + { + "epoch": 0.7592743579290665, + "grad_norm": 0.8074763417243958, + "learning_rate": 0.0009050907052588667, + "loss": 3.6845, + "step": 11175 + }, + { + "epoch": 0.7596140779997282, + "grad_norm": 0.9647040367126465, + "learning_rate": 0.0009050482402500339, + "loss": 3.5161, + "step": 11180 + }, + { + "epoch": 0.75995379807039, + "grad_norm": 0.7623186111450195, + "learning_rate": 0.0009050057752412013, + "loss": 3.5158, + "step": 11185 + }, + { + "epoch": 0.7602935181410517, + "grad_norm": 0.6776531338691711, + "learning_rate": 0.0009049633102323686, + "loss": 3.7268, + "step": 11190 + }, + { + "epoch": 0.7606332382117136, + "grad_norm": 0.6575266122817993, + "learning_rate": 0.0009049208452235358, + "loss": 3.7898, + "step": 11195 + }, + { + "epoch": 0.7609729582823753, + "grad_norm": 2.6512820720672607, + "learning_rate": 0.0009048783802147032, + "loss": 3.5087, + "step": 11200 + }, + { + "epoch": 0.7613126783530371, + "grad_norm": 0.8543487787246704, + "learning_rate": 0.0009048359152058704, + "loss": 3.6902, + "step": 11205 + }, + { + "epoch": 0.7616523984236989, + "grad_norm": 3.5970544815063477, + "learning_rate": 0.0009047934501970376, + "loss": 3.661, + "step": 11210 + }, + { + "epoch": 0.7619921184943607, + "grad_norm": 0.8790910243988037, + "learning_rate": 0.000904750985188205, + "loss": 3.5509, + "step": 11215 + }, + { + "epoch": 0.7623318385650224, + "grad_norm": 0.6596616506576538, + "learning_rate": 0.0009047085201793722, + "loss": 3.7149, + "step": 11220 + }, + { + "epoch": 0.7626715586356841, + "grad_norm": 0.8629494905471802, + "learning_rate": 0.0009046660551705395, + "loss": 3.5913, + "step": 11225 + }, + { + "epoch": 0.763011278706346, + "grad_norm": 0.679855465888977, + "learning_rate": 0.0009046235901617068, + "loss": 3.8356, + "step": 11230 + }, + { + "epoch": 0.7633509987770077, + "grad_norm": 0.6809895038604736, + "learning_rate": 0.0009045811251528741, + "loss": 3.7419, + "step": 11235 + }, + { + "epoch": 0.7636907188476695, + "grad_norm": 0.7851095795631409, + "learning_rate": 0.0009045386601440413, + "loss": 3.7303, + "step": 11240 + }, + { + "epoch": 0.7640304389183313, + "grad_norm": 0.8468301892280579, + "learning_rate": 0.0009044961951352086, + "loss": 3.4092, + "step": 11245 + }, + { + "epoch": 0.7643701589889931, + "grad_norm": 0.7759127616882324, + "learning_rate": 0.0009044537301263759, + "loss": 3.5728, + "step": 11250 + }, + { + "epoch": 0.7647098790596548, + "grad_norm": 0.7105703353881836, + "learning_rate": 0.0009044112651175431, + "loss": 3.8708, + "step": 11255 + }, + { + "epoch": 0.7650495991303167, + "grad_norm": 0.67023104429245, + "learning_rate": 0.0009043688001087105, + "loss": 3.775, + "step": 11260 + }, + { + "epoch": 0.7653893192009784, + "grad_norm": 0.7250822186470032, + "learning_rate": 0.0009043263350998778, + "loss": 3.6947, + "step": 11265 + }, + { + "epoch": 0.7657290392716402, + "grad_norm": 0.8358578085899353, + "learning_rate": 0.000904283870091045, + "loss": 3.7507, + "step": 11270 + }, + { + "epoch": 0.7660687593423019, + "grad_norm": 0.7541988492012024, + "learning_rate": 0.0009042414050822123, + "loss": 3.7711, + "step": 11275 + }, + { + "epoch": 0.7664084794129638, + "grad_norm": 6.97370719909668, + "learning_rate": 0.0009041989400733795, + "loss": 3.6055, + "step": 11280 + }, + { + "epoch": 0.7667481994836255, + "grad_norm": 1.0621757507324219, + "learning_rate": 0.0009041564750645468, + "loss": 3.5172, + "step": 11285 + }, + { + "epoch": 0.7670879195542872, + "grad_norm": 0.7520888447761536, + "learning_rate": 0.0009041140100557141, + "loss": 3.6724, + "step": 11290 + }, + { + "epoch": 0.7674276396249491, + "grad_norm": 0.7989642024040222, + "learning_rate": 0.0009040715450468814, + "loss": 3.5582, + "step": 11295 + }, + { + "epoch": 0.7677673596956108, + "grad_norm": 0.5823965668678284, + "learning_rate": 0.0009040290800380487, + "loss": 3.6233, + "step": 11300 + }, + { + "epoch": 0.7681070797662726, + "grad_norm": 0.8049631714820862, + "learning_rate": 0.000903986615029216, + "loss": 3.5856, + "step": 11305 + }, + { + "epoch": 0.7684467998369343, + "grad_norm": 0.8780303597450256, + "learning_rate": 0.0009039441500203832, + "loss": 3.6896, + "step": 11310 + }, + { + "epoch": 0.7687865199075962, + "grad_norm": 0.994926393032074, + "learning_rate": 0.0009039016850115504, + "loss": 3.5965, + "step": 11315 + }, + { + "epoch": 0.7691262399782579, + "grad_norm": 0.9362306594848633, + "learning_rate": 0.0009038592200027178, + "loss": 3.8483, + "step": 11320 + }, + { + "epoch": 0.7694659600489197, + "grad_norm": 0.8986344933509827, + "learning_rate": 0.000903816754993885, + "loss": 3.6154, + "step": 11325 + }, + { + "epoch": 0.7698056801195815, + "grad_norm": 0.9780957698822021, + "learning_rate": 0.0009037742899850523, + "loss": 3.5373, + "step": 11330 + }, + { + "epoch": 0.7701454001902432, + "grad_norm": 0.9010120630264282, + "learning_rate": 0.0009037318249762197, + "loss": 3.5014, + "step": 11335 + }, + { + "epoch": 0.770485120260905, + "grad_norm": 0.8093063235282898, + "learning_rate": 0.0009036893599673869, + "loss": 3.4193, + "step": 11340 + }, + { + "epoch": 0.7708248403315668, + "grad_norm": 0.8574478626251221, + "learning_rate": 0.0009036468949585541, + "loss": 3.8386, + "step": 11345 + }, + { + "epoch": 0.7711645604022286, + "grad_norm": 0.7611193060874939, + "learning_rate": 0.0009036044299497215, + "loss": 3.6519, + "step": 11350 + }, + { + "epoch": 0.7715042804728903, + "grad_norm": 0.8525367379188538, + "learning_rate": 0.0009035619649408887, + "loss": 3.5253, + "step": 11355 + }, + { + "epoch": 0.7718440005435521, + "grad_norm": 0.8548959493637085, + "learning_rate": 0.0009035194999320559, + "loss": 3.9712, + "step": 11360 + }, + { + "epoch": 0.7721837206142139, + "grad_norm": 0.6429238319396973, + "learning_rate": 0.0009034770349232234, + "loss": 3.7011, + "step": 11365 + }, + { + "epoch": 0.7725234406848757, + "grad_norm": 0.8945921063423157, + "learning_rate": 0.0009034345699143906, + "loss": 3.8966, + "step": 11370 + }, + { + "epoch": 0.7728631607555374, + "grad_norm": 1.0628300905227661, + "learning_rate": 0.0009033921049055578, + "loss": 3.6197, + "step": 11375 + }, + { + "epoch": 0.7732028808261993, + "grad_norm": 1.274933934211731, + "learning_rate": 0.0009033496398967251, + "loss": 3.4399, + "step": 11380 + }, + { + "epoch": 0.773542600896861, + "grad_norm": 1.182814598083496, + "learning_rate": 0.0009033071748878924, + "loss": 3.7439, + "step": 11385 + }, + { + "epoch": 0.7738823209675227, + "grad_norm": 1.0913844108581543, + "learning_rate": 0.0009032647098790596, + "loss": 3.8426, + "step": 11390 + }, + { + "epoch": 0.7742220410381845, + "grad_norm": 1.3744111061096191, + "learning_rate": 0.0009032222448702269, + "loss": 3.7056, + "step": 11395 + }, + { + "epoch": 0.7745617611088463, + "grad_norm": 0.7880125045776367, + "learning_rate": 0.0009031797798613943, + "loss": 3.6926, + "step": 11400 + }, + { + "epoch": 0.7749014811795081, + "grad_norm": 0.7857582569122314, + "learning_rate": 0.0009031373148525615, + "loss": 3.855, + "step": 11405 + }, + { + "epoch": 0.7752412012501698, + "grad_norm": 0.7638903260231018, + "learning_rate": 0.0009030948498437288, + "loss": 3.9538, + "step": 11410 + }, + { + "epoch": 0.7755809213208317, + "grad_norm": 1.0896897315979004, + "learning_rate": 0.000903052384834896, + "loss": 3.7957, + "step": 11415 + }, + { + "epoch": 0.7759206413914934, + "grad_norm": 1.5842639207839966, + "learning_rate": 0.0009030099198260634, + "loss": 3.8637, + "step": 11420 + }, + { + "epoch": 0.7762603614621552, + "grad_norm": 1.6812632083892822, + "learning_rate": 0.0009029674548172306, + "loss": 3.3718, + "step": 11425 + }, + { + "epoch": 0.776600081532817, + "grad_norm": 2.0605804920196533, + "learning_rate": 0.0009029249898083978, + "loss": 3.4116, + "step": 11430 + }, + { + "epoch": 0.7769398016034788, + "grad_norm": 0.7474672794342041, + "learning_rate": 0.0009028825247995653, + "loss": 3.5948, + "step": 11435 + }, + { + "epoch": 0.7772795216741405, + "grad_norm": 0.9415253400802612, + "learning_rate": 0.0009028400597907325, + "loss": 3.738, + "step": 11440 + }, + { + "epoch": 0.7776192417448022, + "grad_norm": 0.946291983127594, + "learning_rate": 0.0009027975947818997, + "loss": 3.5003, + "step": 11445 + }, + { + "epoch": 0.7779589618154641, + "grad_norm": 1.1923933029174805, + "learning_rate": 0.0009027551297730671, + "loss": 3.7299, + "step": 11450 + }, + { + "epoch": 0.7782986818861258, + "grad_norm": 1.4449928998947144, + "learning_rate": 0.0009027126647642343, + "loss": 3.8693, + "step": 11455 + }, + { + "epoch": 0.7786384019567876, + "grad_norm": 0.747748076915741, + "learning_rate": 0.0009026701997554015, + "loss": 3.4474, + "step": 11460 + }, + { + "epoch": 0.7789781220274494, + "grad_norm": 0.8678235411643982, + "learning_rate": 0.0009026277347465688, + "loss": 3.9055, + "step": 11465 + }, + { + "epoch": 0.7793178420981112, + "grad_norm": 0.8195172548294067, + "learning_rate": 0.0009025852697377362, + "loss": 3.6576, + "step": 11470 + }, + { + "epoch": 0.7796575621687729, + "grad_norm": 0.8102189302444458, + "learning_rate": 0.0009025428047289034, + "loss": 3.7615, + "step": 11475 + }, + { + "epoch": 0.7799972822394347, + "grad_norm": 0.786630392074585, + "learning_rate": 0.0009025003397200707, + "loss": 3.7133, + "step": 11480 + }, + { + "epoch": 0.7803370023100965, + "grad_norm": 0.7049959301948547, + "learning_rate": 0.000902457874711238, + "loss": 3.594, + "step": 11485 + }, + { + "epoch": 0.7806767223807582, + "grad_norm": 1.7219754457473755, + "learning_rate": 0.0009024154097024052, + "loss": 3.6476, + "step": 11490 + }, + { + "epoch": 0.78101644245142, + "grad_norm": 16.35957908630371, + "learning_rate": 0.0009023729446935725, + "loss": 3.7201, + "step": 11495 + }, + { + "epoch": 0.7813561625220818, + "grad_norm": 0.9783884882926941, + "learning_rate": 0.0009023304796847398, + "loss": 3.2817, + "step": 11500 + }, + { + "epoch": 0.7816958825927436, + "grad_norm": 0.9637783169746399, + "learning_rate": 0.0009022880146759071, + "loss": 3.6162, + "step": 11505 + }, + { + "epoch": 0.7820356026634053, + "grad_norm": 0.6852649450302124, + "learning_rate": 0.0009022455496670744, + "loss": 3.9328, + "step": 11510 + }, + { + "epoch": 0.7823753227340672, + "grad_norm": 6.640917778015137, + "learning_rate": 0.0009022030846582416, + "loss": 3.7044, + "step": 11515 + }, + { + "epoch": 0.7827150428047289, + "grad_norm": 0.7610793113708496, + "learning_rate": 0.0009021606196494089, + "loss": 3.7768, + "step": 11520 + }, + { + "epoch": 0.7830547628753907, + "grad_norm": 0.6956524848937988, + "learning_rate": 0.0009021181546405762, + "loss": 3.8821, + "step": 11525 + }, + { + "epoch": 0.7833944829460524, + "grad_norm": 0.7520929574966431, + "learning_rate": 0.0009020756896317434, + "loss": 3.6772, + "step": 11530 + }, + { + "epoch": 0.7837342030167143, + "grad_norm": 0.959305465221405, + "learning_rate": 0.0009020332246229107, + "loss": 3.5052, + "step": 11535 + }, + { + "epoch": 0.784073923087376, + "grad_norm": 0.6951141953468323, + "learning_rate": 0.0009019907596140781, + "loss": 3.8615, + "step": 11540 + }, + { + "epoch": 0.7844136431580377, + "grad_norm": 0.8288753628730774, + "learning_rate": 0.0009019482946052453, + "loss": 3.6493, + "step": 11545 + }, + { + "epoch": 0.7847533632286996, + "grad_norm": 0.6916031241416931, + "learning_rate": 0.0009019058295964126, + "loss": 3.756, + "step": 11550 + }, + { + "epoch": 0.7850930832993613, + "grad_norm": 5.888177394866943, + "learning_rate": 0.0009018633645875799, + "loss": 3.6843, + "step": 11555 + }, + { + "epoch": 0.7854328033700231, + "grad_norm": 0.7222215533256531, + "learning_rate": 0.0009018208995787471, + "loss": 3.8328, + "step": 11560 + }, + { + "epoch": 0.7857725234406848, + "grad_norm": 0.7628923654556274, + "learning_rate": 0.0009017784345699143, + "loss": 3.6886, + "step": 11565 + }, + { + "epoch": 0.7861122435113467, + "grad_norm": 1.0630370378494263, + "learning_rate": 0.0009017359695610817, + "loss": 3.447, + "step": 11570 + }, + { + "epoch": 0.7864519635820084, + "grad_norm": 0.8000380396842957, + "learning_rate": 0.000901693504552249, + "loss": 3.65, + "step": 11575 + }, + { + "epoch": 0.7867916836526702, + "grad_norm": 0.8064364194869995, + "learning_rate": 0.0009016510395434162, + "loss": 4.0026, + "step": 11580 + }, + { + "epoch": 0.787131403723332, + "grad_norm": 87.47179412841797, + "learning_rate": 0.0009016085745345836, + "loss": 3.5893, + "step": 11585 + }, + { + "epoch": 0.7874711237939938, + "grad_norm": 1.0113338232040405, + "learning_rate": 0.0009015661095257508, + "loss": 3.6089, + "step": 11590 + }, + { + "epoch": 0.7878108438646555, + "grad_norm": 0.7106664180755615, + "learning_rate": 0.000901523644516918, + "loss": 3.6157, + "step": 11595 + }, + { + "epoch": 0.7881505639353173, + "grad_norm": 0.6645505428314209, + "learning_rate": 0.0009014811795080854, + "loss": 3.5552, + "step": 11600 + }, + { + "epoch": 0.7884902840059791, + "grad_norm": 3.2448296546936035, + "learning_rate": 0.0009014387144992526, + "loss": 3.8646, + "step": 11605 + }, + { + "epoch": 0.7888300040766408, + "grad_norm": 0.7806961536407471, + "learning_rate": 0.0009013962494904199, + "loss": 3.8959, + "step": 11610 + }, + { + "epoch": 0.7891697241473026, + "grad_norm": 0.8047600984573364, + "learning_rate": 0.0009013537844815872, + "loss": 3.4695, + "step": 11615 + }, + { + "epoch": 0.7895094442179644, + "grad_norm": 1.112596869468689, + "learning_rate": 0.0009013113194727545, + "loss": 3.6292, + "step": 11620 + }, + { + "epoch": 0.7898491642886262, + "grad_norm": 0.7855541110038757, + "learning_rate": 0.0009012688544639217, + "loss": 3.543, + "step": 11625 + }, + { + "epoch": 0.7901888843592879, + "grad_norm": 0.7885885238647461, + "learning_rate": 0.000901226389455089, + "loss": 3.6944, + "step": 11630 + }, + { + "epoch": 0.7905286044299498, + "grad_norm": 0.8197888731956482, + "learning_rate": 0.0009011839244462563, + "loss": 3.677, + "step": 11635 + }, + { + "epoch": 0.7908683245006115, + "grad_norm": 0.8078897595405579, + "learning_rate": 0.0009011414594374235, + "loss": 3.7927, + "step": 11640 + }, + { + "epoch": 0.7912080445712733, + "grad_norm": 1.9330220222473145, + "learning_rate": 0.0009010989944285909, + "loss": 3.4802, + "step": 11645 + }, + { + "epoch": 0.791547764641935, + "grad_norm": 0.8217200636863708, + "learning_rate": 0.0009010565294197582, + "loss": 3.7468, + "step": 11650 + }, + { + "epoch": 0.7918874847125968, + "grad_norm": 1.0662391185760498, + "learning_rate": 0.0009010140644109254, + "loss": 3.7532, + "step": 11655 + }, + { + "epoch": 0.7922272047832586, + "grad_norm": 1.107478141784668, + "learning_rate": 0.0009009715994020927, + "loss": 3.7487, + "step": 11660 + }, + { + "epoch": 0.7925669248539203, + "grad_norm": 0.8694434762001038, + "learning_rate": 0.0009009291343932599, + "loss": 3.6482, + "step": 11665 + }, + { + "epoch": 0.7929066449245822, + "grad_norm": 1.516220211982727, + "learning_rate": 0.0009008866693844272, + "loss": 3.6547, + "step": 11670 + }, + { + "epoch": 0.7932463649952439, + "grad_norm": 0.8359872102737427, + "learning_rate": 0.0009008442043755946, + "loss": 3.6229, + "step": 11675 + }, + { + "epoch": 0.7935860850659057, + "grad_norm": 0.7333314418792725, + "learning_rate": 0.0009008017393667618, + "loss": 3.7603, + "step": 11680 + }, + { + "epoch": 0.7939258051365675, + "grad_norm": 0.7693328857421875, + "learning_rate": 0.0009007592743579291, + "loss": 3.9751, + "step": 11685 + }, + { + "epoch": 0.7942655252072293, + "grad_norm": 0.7747409343719482, + "learning_rate": 0.0009007168093490964, + "loss": 3.4737, + "step": 11690 + }, + { + "epoch": 0.794605245277891, + "grad_norm": 1.2370717525482178, + "learning_rate": 0.0009006743443402636, + "loss": 3.7312, + "step": 11695 + }, + { + "epoch": 0.7949449653485527, + "grad_norm": 1.145755410194397, + "learning_rate": 0.0009006318793314308, + "loss": 3.7021, + "step": 11700 + }, + { + "epoch": 0.7952846854192146, + "grad_norm": 1.216779351234436, + "learning_rate": 0.0009005894143225982, + "loss": 3.4789, + "step": 11705 + }, + { + "epoch": 0.7956244054898763, + "grad_norm": 0.7128127217292786, + "learning_rate": 0.0009005469493137655, + "loss": 3.4535, + "step": 11710 + }, + { + "epoch": 0.7959641255605381, + "grad_norm": 0.6719933152198792, + "learning_rate": 0.0009005044843049327, + "loss": 3.7562, + "step": 11715 + }, + { + "epoch": 0.7963038456311999, + "grad_norm": 1.7956007719039917, + "learning_rate": 0.0009004620192961001, + "loss": 3.7448, + "step": 11720 + }, + { + "epoch": 0.7966435657018617, + "grad_norm": 0.9318267107009888, + "learning_rate": 0.0009004195542872673, + "loss": 3.6851, + "step": 11725 + }, + { + "epoch": 0.7969832857725234, + "grad_norm": 1.8655145168304443, + "learning_rate": 0.0009003770892784345, + "loss": 3.7769, + "step": 11730 + }, + { + "epoch": 0.7973230058431852, + "grad_norm": 0.8474512696266174, + "learning_rate": 0.0009003346242696019, + "loss": 3.6509, + "step": 11735 + }, + { + "epoch": 0.797662725913847, + "grad_norm": 1.0019559860229492, + "learning_rate": 0.0009002921592607691, + "loss": 3.5633, + "step": 11740 + }, + { + "epoch": 0.7980024459845088, + "grad_norm": 0.877302348613739, + "learning_rate": 0.0009002496942519364, + "loss": 3.6611, + "step": 11745 + }, + { + "epoch": 0.7983421660551705, + "grad_norm": 1.024214267730713, + "learning_rate": 0.0009002072292431038, + "loss": 3.8209, + "step": 11750 + }, + { + "epoch": 0.7986818861258324, + "grad_norm": 2.550053119659424, + "learning_rate": 0.000900164764234271, + "loss": 3.3611, + "step": 11755 + }, + { + "epoch": 0.7990216061964941, + "grad_norm": 1.1882482767105103, + "learning_rate": 0.0009001222992254383, + "loss": 3.6797, + "step": 11760 + }, + { + "epoch": 0.7993613262671558, + "grad_norm": 0.9128539562225342, + "learning_rate": 0.0009000798342166055, + "loss": 3.7904, + "step": 11765 + }, + { + "epoch": 0.7997010463378177, + "grad_norm": 0.6798457503318787, + "learning_rate": 0.0009000373692077728, + "loss": 3.7553, + "step": 11770 + }, + { + "epoch": 0.8000407664084794, + "grad_norm": 0.9786900281906128, + "learning_rate": 0.0008999949041989401, + "loss": 3.5874, + "step": 11775 + }, + { + "epoch": 0.8003804864791412, + "grad_norm": 0.6784381866455078, + "learning_rate": 0.0008999524391901074, + "loss": 3.8471, + "step": 11780 + }, + { + "epoch": 0.8007202065498029, + "grad_norm": 3.1511030197143555, + "learning_rate": 0.0008999099741812747, + "loss": 3.872, + "step": 11785 + }, + { + "epoch": 0.8010599266204648, + "grad_norm": 0.6527305245399475, + "learning_rate": 0.000899867509172442, + "loss": 3.3656, + "step": 11790 + }, + { + "epoch": 0.8013996466911265, + "grad_norm": 0.8578232526779175, + "learning_rate": 0.0008998250441636092, + "loss": 3.7483, + "step": 11795 + }, + { + "epoch": 0.8017393667617883, + "grad_norm": 0.8288311958312988, + "learning_rate": 0.0008997825791547764, + "loss": 3.8449, + "step": 11800 + }, + { + "epoch": 0.8020790868324501, + "grad_norm": 1.7888271808624268, + "learning_rate": 0.0008997401141459438, + "loss": 3.9227, + "step": 11805 + }, + { + "epoch": 0.8024188069031118, + "grad_norm": 1.4842759370803833, + "learning_rate": 0.000899697649137111, + "loss": 3.4431, + "step": 11810 + }, + { + "epoch": 0.8027585269737736, + "grad_norm": 1.0334523916244507, + "learning_rate": 0.0008996551841282783, + "loss": 3.5816, + "step": 11815 + }, + { + "epoch": 0.8030982470444353, + "grad_norm": 0.8577939867973328, + "learning_rate": 0.0008996127191194457, + "loss": 3.7209, + "step": 11820 + }, + { + "epoch": 0.8034379671150972, + "grad_norm": 0.868172824382782, + "learning_rate": 0.0008995702541106129, + "loss": 3.7975, + "step": 11825 + }, + { + "epoch": 0.8037776871857589, + "grad_norm": 0.9573361873626709, + "learning_rate": 0.0008995277891017801, + "loss": 3.7591, + "step": 11830 + }, + { + "epoch": 0.8041174072564207, + "grad_norm": 0.7482950091362, + "learning_rate": 0.0008994853240929475, + "loss": 3.7724, + "step": 11835 + }, + { + "epoch": 0.8044571273270825, + "grad_norm": 5.594187259674072, + "learning_rate": 0.0008994428590841147, + "loss": 3.7639, + "step": 11840 + }, + { + "epoch": 0.8047968473977443, + "grad_norm": 1.4929182529449463, + "learning_rate": 0.0008994003940752819, + "loss": 3.5914, + "step": 11845 + }, + { + "epoch": 0.805136567468406, + "grad_norm": 1.2356071472167969, + "learning_rate": 0.0008993579290664494, + "loss": 3.7893, + "step": 11850 + }, + { + "epoch": 0.8054762875390679, + "grad_norm": 1.3104801177978516, + "learning_rate": 0.0008993154640576166, + "loss": 3.6528, + "step": 11855 + }, + { + "epoch": 0.8058160076097296, + "grad_norm": 1.1337311267852783, + "learning_rate": 0.0008992729990487838, + "loss": 3.5129, + "step": 11860 + }, + { + "epoch": 0.8061557276803913, + "grad_norm": 0.8900738954544067, + "learning_rate": 0.0008992305340399511, + "loss": 3.6081, + "step": 11865 + }, + { + "epoch": 0.8064954477510531, + "grad_norm": 0.7715288996696472, + "learning_rate": 0.0008991880690311184, + "loss": 3.7922, + "step": 11870 + }, + { + "epoch": 0.8068351678217149, + "grad_norm": 0.7087821960449219, + "learning_rate": 0.0008991456040222856, + "loss": 3.6447, + "step": 11875 + }, + { + "epoch": 0.8071748878923767, + "grad_norm": 0.856303870677948, + "learning_rate": 0.0008991031390134529, + "loss": 3.2649, + "step": 11880 + }, + { + "epoch": 0.8075146079630384, + "grad_norm": 0.7621307969093323, + "learning_rate": 0.0008990606740046203, + "loss": 3.7453, + "step": 11885 + }, + { + "epoch": 0.8078543280337003, + "grad_norm": 0.7358371615409851, + "learning_rate": 0.0008990182089957875, + "loss": 3.3933, + "step": 11890 + }, + { + "epoch": 0.808194048104362, + "grad_norm": 1.152427077293396, + "learning_rate": 0.0008989757439869548, + "loss": 3.6491, + "step": 11895 + }, + { + "epoch": 0.8085337681750238, + "grad_norm": 1.6236579418182373, + "learning_rate": 0.000898933278978122, + "loss": 3.505, + "step": 11900 + }, + { + "epoch": 0.8088734882456855, + "grad_norm": 0.7944048047065735, + "learning_rate": 0.0008988908139692893, + "loss": 3.5157, + "step": 11905 + }, + { + "epoch": 0.8092132083163474, + "grad_norm": 2.7386722564697266, + "learning_rate": 0.0008988483489604566, + "loss": 3.849, + "step": 11910 + }, + { + "epoch": 0.8095529283870091, + "grad_norm": 0.6543929576873779, + "learning_rate": 0.0008988058839516238, + "loss": 3.7691, + "step": 11915 + }, + { + "epoch": 0.8098926484576708, + "grad_norm": 0.7426961064338684, + "learning_rate": 0.0008987634189427912, + "loss": 3.4518, + "step": 11920 + }, + { + "epoch": 0.8102323685283327, + "grad_norm": 0.680490255355835, + "learning_rate": 0.0008987209539339585, + "loss": 3.7327, + "step": 11925 + }, + { + "epoch": 0.8105720885989944, + "grad_norm": 0.9860913753509521, + "learning_rate": 0.0008986784889251257, + "loss": 3.7608, + "step": 11930 + }, + { + "epoch": 0.8109118086696562, + "grad_norm": 0.8084064722061157, + "learning_rate": 0.000898636023916293, + "loss": 3.7959, + "step": 11935 + }, + { + "epoch": 0.811251528740318, + "grad_norm": 1.1407679319381714, + "learning_rate": 0.0008985935589074603, + "loss": 3.5039, + "step": 11940 + }, + { + "epoch": 0.8115912488109798, + "grad_norm": 0.7349080443382263, + "learning_rate": 0.0008985510938986275, + "loss": 3.4194, + "step": 11945 + }, + { + "epoch": 0.8119309688816415, + "grad_norm": 0.8958530426025391, + "learning_rate": 0.0008985086288897947, + "loss": 3.6837, + "step": 11950 + }, + { + "epoch": 0.8122706889523033, + "grad_norm": 2.5459187030792236, + "learning_rate": 0.0008984661638809622, + "loss": 3.609, + "step": 11955 + }, + { + "epoch": 0.8126104090229651, + "grad_norm": 0.810468852519989, + "learning_rate": 0.0008984236988721294, + "loss": 3.6765, + "step": 11960 + }, + { + "epoch": 0.8129501290936268, + "grad_norm": 0.7730026245117188, + "learning_rate": 0.0008983812338632966, + "loss": 3.577, + "step": 11965 + }, + { + "epoch": 0.8132898491642886, + "grad_norm": 1.1334856748580933, + "learning_rate": 0.000898338768854464, + "loss": 3.6766, + "step": 11970 + }, + { + "epoch": 0.8136295692349504, + "grad_norm": 1.0055420398712158, + "learning_rate": 0.0008982963038456312, + "loss": 3.8258, + "step": 11975 + }, + { + "epoch": 0.8139692893056122, + "grad_norm": 1.1260411739349365, + "learning_rate": 0.0008982538388367984, + "loss": 3.5624, + "step": 11980 + }, + { + "epoch": 0.8143090093762739, + "grad_norm": 0.8112538456916809, + "learning_rate": 0.0008982113738279658, + "loss": 3.577, + "step": 11985 + }, + { + "epoch": 0.8146487294469357, + "grad_norm": 0.8501037359237671, + "learning_rate": 0.0008981689088191331, + "loss": 3.8233, + "step": 11990 + }, + { + "epoch": 0.8149884495175975, + "grad_norm": 1.2345054149627686, + "learning_rate": 0.0008981264438103003, + "loss": 3.4861, + "step": 11995 + }, + { + "epoch": 0.8153281695882593, + "grad_norm": 0.8775907158851624, + "learning_rate": 0.0008980839788014677, + "loss": 3.5838, + "step": 12000 + }, + { + "epoch": 0.815667889658921, + "grad_norm": 1.1107614040374756, + "learning_rate": 0.0008980415137926349, + "loss": 3.7822, + "step": 12005 + }, + { + "epoch": 0.8160076097295829, + "grad_norm": 1.473344087600708, + "learning_rate": 0.0008979990487838021, + "loss": 3.6956, + "step": 12010 + }, + { + "epoch": 0.8163473298002446, + "grad_norm": 0.7700946927070618, + "learning_rate": 0.0008979565837749694, + "loss": 3.8535, + "step": 12015 + }, + { + "epoch": 0.8166870498709063, + "grad_norm": 0.8405547142028809, + "learning_rate": 0.0008979141187661367, + "loss": 3.7368, + "step": 12020 + }, + { + "epoch": 0.8170267699415682, + "grad_norm": 0.7274438738822937, + "learning_rate": 0.000897871653757304, + "loss": 3.2918, + "step": 12025 + }, + { + "epoch": 0.8173664900122299, + "grad_norm": 0.9372562170028687, + "learning_rate": 0.0008978291887484713, + "loss": 3.6491, + "step": 12030 + }, + { + "epoch": 0.8177062100828917, + "grad_norm": 1.0108988285064697, + "learning_rate": 0.0008977867237396386, + "loss": 3.5966, + "step": 12035 + }, + { + "epoch": 0.8180459301535534, + "grad_norm": 0.8893322348594666, + "learning_rate": 0.0008977442587308058, + "loss": 3.5553, + "step": 12040 + }, + { + "epoch": 0.8183856502242153, + "grad_norm": 1.1882805824279785, + "learning_rate": 0.0008977017937219731, + "loss": 3.9158, + "step": 12045 + }, + { + "epoch": 0.818725370294877, + "grad_norm": 0.6214559674263, + "learning_rate": 0.0008976593287131403, + "loss": 3.7214, + "step": 12050 + }, + { + "epoch": 0.8190650903655388, + "grad_norm": 0.9441993236541748, + "learning_rate": 0.0008976168637043076, + "loss": 3.6625, + "step": 12055 + }, + { + "epoch": 0.8194048104362006, + "grad_norm": 0.7592012882232666, + "learning_rate": 0.000897574398695475, + "loss": 3.5658, + "step": 12060 + }, + { + "epoch": 0.8197445305068624, + "grad_norm": 0.8846549391746521, + "learning_rate": 0.0008975319336866422, + "loss": 3.8133, + "step": 12065 + }, + { + "epoch": 0.8200842505775241, + "grad_norm": 1.0380690097808838, + "learning_rate": 0.0008974894686778095, + "loss": 3.6635, + "step": 12070 + }, + { + "epoch": 0.8204239706481858, + "grad_norm": 0.7360149621963501, + "learning_rate": 0.0008974470036689768, + "loss": 3.7635, + "step": 12075 + }, + { + "epoch": 0.8207636907188477, + "grad_norm": 1.2694586515426636, + "learning_rate": 0.000897404538660144, + "loss": 3.4643, + "step": 12080 + }, + { + "epoch": 0.8211034107895094, + "grad_norm": 0.7986928224563599, + "learning_rate": 0.0008973620736513113, + "loss": 3.6993, + "step": 12085 + }, + { + "epoch": 0.8214431308601712, + "grad_norm": 0.942507266998291, + "learning_rate": 0.0008973196086424786, + "loss": 3.857, + "step": 12090 + }, + { + "epoch": 0.821782850930833, + "grad_norm": 0.8908464312553406, + "learning_rate": 0.0008972771436336459, + "loss": 3.8681, + "step": 12095 + }, + { + "epoch": 0.8221225710014948, + "grad_norm": 0.8867223262786865, + "learning_rate": 0.0008972346786248133, + "loss": 3.7046, + "step": 12100 + }, + { + "epoch": 0.8224622910721565, + "grad_norm": 0.8820211291313171, + "learning_rate": 0.0008971922136159805, + "loss": 3.5966, + "step": 12105 + }, + { + "epoch": 0.8228020111428184, + "grad_norm": 1.1007713079452515, + "learning_rate": 0.0008971497486071477, + "loss": 3.6767, + "step": 12110 + }, + { + "epoch": 0.8231417312134801, + "grad_norm": 0.8534795045852661, + "learning_rate": 0.000897107283598315, + "loss": 3.7248, + "step": 12115 + }, + { + "epoch": 0.8234814512841419, + "grad_norm": 0.9195171594619751, + "learning_rate": 0.0008970648185894823, + "loss": 3.6925, + "step": 12120 + }, + { + "epoch": 0.8238211713548036, + "grad_norm": 0.982778787612915, + "learning_rate": 0.0008970223535806495, + "loss": 3.6736, + "step": 12125 + }, + { + "epoch": 0.8241608914254654, + "grad_norm": 0.7273781299591064, + "learning_rate": 0.0008969798885718169, + "loss": 3.9274, + "step": 12130 + }, + { + "epoch": 0.8245006114961272, + "grad_norm": 0.9642855525016785, + "learning_rate": 0.0008969374235629842, + "loss": 3.5285, + "step": 12135 + }, + { + "epoch": 0.8248403315667889, + "grad_norm": 1.291280746459961, + "learning_rate": 0.0008968949585541514, + "loss": 3.7009, + "step": 12140 + }, + { + "epoch": 0.8251800516374508, + "grad_norm": 0.7058663964271545, + "learning_rate": 0.0008968524935453187, + "loss": 3.7475, + "step": 12145 + }, + { + "epoch": 0.8255197717081125, + "grad_norm": 1.9695004224777222, + "learning_rate": 0.000896810028536486, + "loss": 3.6172, + "step": 12150 + }, + { + "epoch": 0.8258594917787743, + "grad_norm": 0.8333840370178223, + "learning_rate": 0.0008967675635276532, + "loss": 3.7805, + "step": 12155 + }, + { + "epoch": 0.826199211849436, + "grad_norm": 0.8567728996276855, + "learning_rate": 0.0008967250985188205, + "loss": 3.6374, + "step": 12160 + }, + { + "epoch": 0.8265389319200979, + "grad_norm": 1.1291769742965698, + "learning_rate": 0.0008966826335099878, + "loss": 3.8096, + "step": 12165 + }, + { + "epoch": 0.8268786519907596, + "grad_norm": 0.9778679609298706, + "learning_rate": 0.0008966401685011551, + "loss": 3.6984, + "step": 12170 + }, + { + "epoch": 0.8272183720614213, + "grad_norm": 1.1576662063598633, + "learning_rate": 0.0008965977034923224, + "loss": 3.5487, + "step": 12175 + }, + { + "epoch": 0.8275580921320832, + "grad_norm": 0.6185218095779419, + "learning_rate": 0.0008965552384834896, + "loss": 3.718, + "step": 12180 + }, + { + "epoch": 0.8278978122027449, + "grad_norm": 0.6700850129127502, + "learning_rate": 0.0008965127734746569, + "loss": 3.6459, + "step": 12185 + }, + { + "epoch": 0.8282375322734067, + "grad_norm": 0.8876090049743652, + "learning_rate": 0.0008964703084658242, + "loss": 3.8491, + "step": 12190 + }, + { + "epoch": 0.8285772523440685, + "grad_norm": 0.6527369618415833, + "learning_rate": 0.0008964278434569914, + "loss": 3.8357, + "step": 12195 + }, + { + "epoch": 0.8289169724147303, + "grad_norm": 0.7413597106933594, + "learning_rate": 0.0008963853784481587, + "loss": 3.9019, + "step": 12200 + }, + { + "epoch": 0.829256692485392, + "grad_norm": 0.8062065243721008, + "learning_rate": 0.0008963429134393261, + "loss": 3.924, + "step": 12205 + }, + { + "epoch": 0.8295964125560538, + "grad_norm": 0.9075581431388855, + "learning_rate": 0.0008963004484304933, + "loss": 3.6124, + "step": 12210 + }, + { + "epoch": 0.8299361326267156, + "grad_norm": 0.9779383540153503, + "learning_rate": 0.0008962579834216605, + "loss": 3.6829, + "step": 12215 + }, + { + "epoch": 0.8302758526973774, + "grad_norm": 0.755110502243042, + "learning_rate": 0.0008962155184128279, + "loss": 3.5705, + "step": 12220 + }, + { + "epoch": 0.8306155727680391, + "grad_norm": 0.8037834167480469, + "learning_rate": 0.0008961730534039951, + "loss": 3.785, + "step": 12225 + }, + { + "epoch": 0.830955292838701, + "grad_norm": 0.9082496762275696, + "learning_rate": 0.0008961305883951623, + "loss": 3.5452, + "step": 12230 + }, + { + "epoch": 0.8312950129093627, + "grad_norm": 1.0774654150009155, + "learning_rate": 0.0008960881233863298, + "loss": 3.1426, + "step": 12235 + }, + { + "epoch": 0.8316347329800244, + "grad_norm": 0.8939933180809021, + "learning_rate": 0.000896045658377497, + "loss": 3.5016, + "step": 12240 + }, + { + "epoch": 0.8319744530506862, + "grad_norm": 0.7898679375648499, + "learning_rate": 0.0008960031933686642, + "loss": 3.3459, + "step": 12245 + }, + { + "epoch": 0.832314173121348, + "grad_norm": 0.6527073383331299, + "learning_rate": 0.0008959607283598315, + "loss": 3.6323, + "step": 12250 + }, + { + "epoch": 0.8326538931920098, + "grad_norm": 0.8865939378738403, + "learning_rate": 0.0008959182633509988, + "loss": 3.478, + "step": 12255 + }, + { + "epoch": 0.8329936132626715, + "grad_norm": 1.1525355577468872, + "learning_rate": 0.000895875798342166, + "loss": 3.5154, + "step": 12260 + }, + { + "epoch": 0.8333333333333334, + "grad_norm": 1.543592095375061, + "learning_rate": 0.0008958333333333334, + "loss": 3.7784, + "step": 12265 + }, + { + "epoch": 0.8336730534039951, + "grad_norm": 1.6606383323669434, + "learning_rate": 0.0008957908683245007, + "loss": 3.2701, + "step": 12270 + }, + { + "epoch": 0.8340127734746569, + "grad_norm": 1.9703108072280884, + "learning_rate": 0.0008957484033156679, + "loss": 3.8146, + "step": 12275 + }, + { + "epoch": 0.8343524935453187, + "grad_norm": 0.8228875994682312, + "learning_rate": 0.0008957059383068352, + "loss": 3.6945, + "step": 12280 + }, + { + "epoch": 0.8346922136159804, + "grad_norm": 0.792127251625061, + "learning_rate": 0.0008956634732980025, + "loss": 3.5868, + "step": 12285 + }, + { + "epoch": 0.8350319336866422, + "grad_norm": 0.9458941221237183, + "learning_rate": 0.0008956210082891697, + "loss": 3.6609, + "step": 12290 + }, + { + "epoch": 0.8353716537573039, + "grad_norm": 2.0094172954559326, + "learning_rate": 0.000895578543280337, + "loss": 3.5978, + "step": 12295 + }, + { + "epoch": 0.8357113738279658, + "grad_norm": 0.9320061802864075, + "learning_rate": 0.0008955360782715043, + "loss": 3.7183, + "step": 12300 + }, + { + "epoch": 0.8360510938986275, + "grad_norm": 2.128854990005493, + "learning_rate": 0.0008954936132626716, + "loss": 3.4003, + "step": 12305 + }, + { + "epoch": 0.8363908139692893, + "grad_norm": 0.9589837789535522, + "learning_rate": 0.0008954511482538389, + "loss": 3.8215, + "step": 12310 + }, + { + "epoch": 0.8367305340399511, + "grad_norm": 1.0171260833740234, + "learning_rate": 0.0008954086832450061, + "loss": 3.5688, + "step": 12315 + }, + { + "epoch": 0.8370702541106129, + "grad_norm": 0.746832013130188, + "learning_rate": 0.0008953662182361734, + "loss": 3.7359, + "step": 12320 + }, + { + "epoch": 0.8374099741812746, + "grad_norm": 0.8556201457977295, + "learning_rate": 0.0008953237532273407, + "loss": 3.8403, + "step": 12325 + }, + { + "epoch": 0.8377496942519363, + "grad_norm": 0.7217704653739929, + "learning_rate": 0.0008952812882185079, + "loss": 3.2962, + "step": 12330 + }, + { + "epoch": 0.8380894143225982, + "grad_norm": 1.1214584112167358, + "learning_rate": 0.0008952388232096753, + "loss": 3.8938, + "step": 12335 + }, + { + "epoch": 0.83842913439326, + "grad_norm": 0.7692179679870605, + "learning_rate": 0.0008951963582008426, + "loss": 3.7128, + "step": 12340 + }, + { + "epoch": 0.8387688544639217, + "grad_norm": 0.995206356048584, + "learning_rate": 0.0008951538931920098, + "loss": 3.679, + "step": 12345 + }, + { + "epoch": 0.8391085745345835, + "grad_norm": 1.7244571447372437, + "learning_rate": 0.000895111428183177, + "loss": 3.6521, + "step": 12350 + }, + { + "epoch": 0.8394482946052453, + "grad_norm": 0.7407826781272888, + "learning_rate": 0.0008950689631743444, + "loss": 3.6666, + "step": 12355 + }, + { + "epoch": 0.839788014675907, + "grad_norm": 0.8281946778297424, + "learning_rate": 0.0008950264981655116, + "loss": 3.6433, + "step": 12360 + }, + { + "epoch": 0.8401277347465689, + "grad_norm": 0.610300600528717, + "learning_rate": 0.0008949840331566788, + "loss": 3.9378, + "step": 12365 + }, + { + "epoch": 0.8404674548172306, + "grad_norm": 1.0494595766067505, + "learning_rate": 0.0008949415681478463, + "loss": 3.7256, + "step": 12370 + }, + { + "epoch": 0.8408071748878924, + "grad_norm": 1.2667574882507324, + "learning_rate": 0.0008948991031390135, + "loss": 3.6659, + "step": 12375 + }, + { + "epoch": 0.8411468949585541, + "grad_norm": 1.021226167678833, + "learning_rate": 0.0008948566381301807, + "loss": 3.652, + "step": 12380 + }, + { + "epoch": 0.841486615029216, + "grad_norm": 1.876592993736267, + "learning_rate": 0.0008948141731213481, + "loss": 3.822, + "step": 12385 + }, + { + "epoch": 0.8418263350998777, + "grad_norm": 0.748383104801178, + "learning_rate": 0.0008947717081125153, + "loss": 3.5108, + "step": 12390 + }, + { + "epoch": 0.8421660551705394, + "grad_norm": 0.8421717286109924, + "learning_rate": 0.0008947292431036825, + "loss": 3.5724, + "step": 12395 + }, + { + "epoch": 0.8425057752412013, + "grad_norm": 0.9351264834403992, + "learning_rate": 0.0008946867780948498, + "loss": 3.7664, + "step": 12400 + }, + { + "epoch": 0.842845495311863, + "grad_norm": 0.7131645083427429, + "learning_rate": 0.0008946443130860172, + "loss": 3.666, + "step": 12405 + }, + { + "epoch": 0.8431852153825248, + "grad_norm": 0.8910898566246033, + "learning_rate": 0.0008946018480771844, + "loss": 3.8812, + "step": 12410 + }, + { + "epoch": 0.8435249354531865, + "grad_norm": 0.7854522466659546, + "learning_rate": 0.0008945593830683517, + "loss": 3.7911, + "step": 12415 + }, + { + "epoch": 0.8438646555238484, + "grad_norm": 0.7281268239021301, + "learning_rate": 0.000894516918059519, + "loss": 3.7536, + "step": 12420 + }, + { + "epoch": 0.8442043755945101, + "grad_norm": 0.871080756187439, + "learning_rate": 0.0008944744530506862, + "loss": 3.8731, + "step": 12425 + }, + { + "epoch": 0.8445440956651719, + "grad_norm": 0.6535822153091431, + "learning_rate": 0.0008944319880418535, + "loss": 3.6304, + "step": 12430 + }, + { + "epoch": 0.8448838157358337, + "grad_norm": 0.8274858593940735, + "learning_rate": 0.0008943895230330207, + "loss": 3.8228, + "step": 12435 + }, + { + "epoch": 0.8452235358064955, + "grad_norm": 1.337404489517212, + "learning_rate": 0.0008943470580241882, + "loss": 3.6757, + "step": 12440 + }, + { + "epoch": 0.8455632558771572, + "grad_norm": 0.7477543354034424, + "learning_rate": 0.0008943045930153554, + "loss": 3.5688, + "step": 12445 + }, + { + "epoch": 0.845902975947819, + "grad_norm": 2.138296365737915, + "learning_rate": 0.0008942621280065226, + "loss": 3.6284, + "step": 12450 + }, + { + "epoch": 0.8462426960184808, + "grad_norm": 1.3792190551757812, + "learning_rate": 0.00089421966299769, + "loss": 3.9597, + "step": 12455 + }, + { + "epoch": 0.8465824160891425, + "grad_norm": 0.9175506830215454, + "learning_rate": 0.0008941771979888572, + "loss": 3.6483, + "step": 12460 + }, + { + "epoch": 0.8469221361598043, + "grad_norm": 0.6348444819450378, + "learning_rate": 0.0008941347329800244, + "loss": 3.7185, + "step": 12465 + }, + { + "epoch": 0.8472618562304661, + "grad_norm": 0.8670823574066162, + "learning_rate": 0.0008940922679711918, + "loss": 3.4806, + "step": 12470 + }, + { + "epoch": 0.8476015763011279, + "grad_norm": 1.2782633304595947, + "learning_rate": 0.0008940498029623591, + "loss": 3.2384, + "step": 12475 + }, + { + "epoch": 0.8479412963717896, + "grad_norm": 1.0383013486862183, + "learning_rate": 0.0008940073379535263, + "loss": 3.5404, + "step": 12480 + }, + { + "epoch": 0.8482810164424515, + "grad_norm": 0.9223533868789673, + "learning_rate": 0.0008939648729446937, + "loss": 3.6777, + "step": 12485 + }, + { + "epoch": 0.8486207365131132, + "grad_norm": 0.7466956973075867, + "learning_rate": 0.0008939224079358609, + "loss": 3.7102, + "step": 12490 + }, + { + "epoch": 0.848960456583775, + "grad_norm": 0.7598857283592224, + "learning_rate": 0.0008938799429270281, + "loss": 3.7285, + "step": 12495 + }, + { + "epoch": 0.8493001766544367, + "grad_norm": 1.495668649673462, + "learning_rate": 0.0008938374779181954, + "loss": 3.7081, + "step": 12500 + }, + { + "epoch": 0.8496398967250985, + "grad_norm": 0.8449199795722961, + "learning_rate": 0.0008937950129093627, + "loss": 3.6521, + "step": 12505 + }, + { + "epoch": 0.8499796167957603, + "grad_norm": 0.8988255858421326, + "learning_rate": 0.00089375254790053, + "loss": 3.4967, + "step": 12510 + }, + { + "epoch": 0.850319336866422, + "grad_norm": 0.6855485439300537, + "learning_rate": 0.0008937100828916973, + "loss": 3.5231, + "step": 12515 + }, + { + "epoch": 0.8506590569370839, + "grad_norm": 0.7748841047286987, + "learning_rate": 0.0008936676178828646, + "loss": 3.7082, + "step": 12520 + }, + { + "epoch": 0.8509987770077456, + "grad_norm": 0.9974066615104675, + "learning_rate": 0.0008936251528740318, + "loss": 3.7248, + "step": 12525 + }, + { + "epoch": 0.8513384970784074, + "grad_norm": 0.833838164806366, + "learning_rate": 0.0008935826878651991, + "loss": 3.8281, + "step": 12530 + }, + { + "epoch": 0.8516782171490692, + "grad_norm": 1.0375392436981201, + "learning_rate": 0.0008935402228563663, + "loss": 3.951, + "step": 12535 + }, + { + "epoch": 0.852017937219731, + "grad_norm": 0.8951463103294373, + "learning_rate": 0.0008934977578475336, + "loss": 3.7844, + "step": 12540 + }, + { + "epoch": 0.8523576572903927, + "grad_norm": 0.7702258825302124, + "learning_rate": 0.000893455292838701, + "loss": 3.7887, + "step": 12545 + }, + { + "epoch": 0.8526973773610544, + "grad_norm": 1.7558034658432007, + "learning_rate": 0.0008934128278298682, + "loss": 3.7398, + "step": 12550 + }, + { + "epoch": 0.8530370974317163, + "grad_norm": 0.9577825665473938, + "learning_rate": 0.0008933703628210355, + "loss": 3.739, + "step": 12555 + }, + { + "epoch": 0.853376817502378, + "grad_norm": 1.3343416452407837, + "learning_rate": 0.0008933278978122028, + "loss": 3.4998, + "step": 12560 + }, + { + "epoch": 0.8537165375730398, + "grad_norm": 0.7011745572090149, + "learning_rate": 0.00089328543280337, + "loss": 3.8674, + "step": 12565 + }, + { + "epoch": 0.8540562576437016, + "grad_norm": 0.9189016222953796, + "learning_rate": 0.0008932429677945373, + "loss": 3.5648, + "step": 12570 + }, + { + "epoch": 0.8543959777143634, + "grad_norm": 0.8981935381889343, + "learning_rate": 0.0008932005027857046, + "loss": 3.7673, + "step": 12575 + }, + { + "epoch": 0.8547356977850251, + "grad_norm": 1.0040943622589111, + "learning_rate": 0.0008931580377768719, + "loss": 3.616, + "step": 12580 + }, + { + "epoch": 0.8550754178556869, + "grad_norm": 0.7973901629447937, + "learning_rate": 0.0008931155727680392, + "loss": 3.4455, + "step": 12585 + }, + { + "epoch": 0.8554151379263487, + "grad_norm": 0.8325662612915039, + "learning_rate": 0.0008930731077592065, + "loss": 3.7143, + "step": 12590 + }, + { + "epoch": 0.8557548579970105, + "grad_norm": 0.7751691341400146, + "learning_rate": 0.0008930306427503737, + "loss": 3.7809, + "step": 12595 + }, + { + "epoch": 0.8560945780676722, + "grad_norm": 7.4706196784973145, + "learning_rate": 0.0008929881777415409, + "loss": 3.7586, + "step": 12600 + }, + { + "epoch": 0.856434298138334, + "grad_norm": 0.9639095664024353, + "learning_rate": 0.0008929457127327083, + "loss": 3.4086, + "step": 12605 + }, + { + "epoch": 0.8567740182089958, + "grad_norm": 0.6753028035163879, + "learning_rate": 0.0008929032477238755, + "loss": 3.6616, + "step": 12610 + }, + { + "epoch": 0.8571137382796575, + "grad_norm": 0.7249093055725098, + "learning_rate": 0.0008928607827150428, + "loss": 3.5685, + "step": 12615 + }, + { + "epoch": 0.8574534583503194, + "grad_norm": 1.7473300695419312, + "learning_rate": 0.0008928183177062102, + "loss": 3.9909, + "step": 12620 + }, + { + "epoch": 0.8577931784209811, + "grad_norm": 0.9212319850921631, + "learning_rate": 0.0008927758526973774, + "loss": 3.9006, + "step": 12625 + }, + { + "epoch": 0.8581328984916429, + "grad_norm": 1.0572686195373535, + "learning_rate": 0.0008927333876885446, + "loss": 3.8801, + "step": 12630 + }, + { + "epoch": 0.8584726185623046, + "grad_norm": 1.8485252857208252, + "learning_rate": 0.000892690922679712, + "loss": 3.4499, + "step": 12635 + }, + { + "epoch": 0.8588123386329665, + "grad_norm": 0.7671639323234558, + "learning_rate": 0.0008926484576708792, + "loss": 3.8462, + "step": 12640 + }, + { + "epoch": 0.8591520587036282, + "grad_norm": 1.237829327583313, + "learning_rate": 0.0008926059926620464, + "loss": 3.767, + "step": 12645 + }, + { + "epoch": 0.85949177877429, + "grad_norm": 0.8066554069519043, + "learning_rate": 0.0008925635276532138, + "loss": 3.71, + "step": 12650 + }, + { + "epoch": 0.8598314988449518, + "grad_norm": 0.7689037919044495, + "learning_rate": 0.0008925210626443811, + "loss": 3.5923, + "step": 12655 + }, + { + "epoch": 0.8601712189156135, + "grad_norm": 0.874286949634552, + "learning_rate": 0.0008924785976355483, + "loss": 3.5279, + "step": 12660 + }, + { + "epoch": 0.8605109389862753, + "grad_norm": 0.8563452363014221, + "learning_rate": 0.0008924361326267156, + "loss": 3.6292, + "step": 12665 + }, + { + "epoch": 0.860850659056937, + "grad_norm": 1.5231919288635254, + "learning_rate": 0.0008923936676178829, + "loss": 3.6664, + "step": 12670 + }, + { + "epoch": 0.8611903791275989, + "grad_norm": 0.6971249580383301, + "learning_rate": 0.0008923512026090501, + "loss": 3.7749, + "step": 12675 + }, + { + "epoch": 0.8615300991982606, + "grad_norm": 1.3179380893707275, + "learning_rate": 0.0008923087376002174, + "loss": 3.6636, + "step": 12680 + }, + { + "epoch": 0.8618698192689224, + "grad_norm": 0.7259864211082458, + "learning_rate": 0.0008922662725913848, + "loss": 3.7117, + "step": 12685 + }, + { + "epoch": 0.8622095393395842, + "grad_norm": 1.1027296781539917, + "learning_rate": 0.000892223807582552, + "loss": 3.4609, + "step": 12690 + }, + { + "epoch": 0.862549259410246, + "grad_norm": 0.7830087542533875, + "learning_rate": 0.0008921813425737193, + "loss": 3.7428, + "step": 12695 + }, + { + "epoch": 0.8628889794809077, + "grad_norm": 0.9170404076576233, + "learning_rate": 0.0008921388775648865, + "loss": 3.7073, + "step": 12700 + }, + { + "epoch": 0.8632286995515696, + "grad_norm": 0.7713662385940552, + "learning_rate": 0.0008920964125560538, + "loss": 3.3853, + "step": 12705 + }, + { + "epoch": 0.8635684196222313, + "grad_norm": 0.8968492150306702, + "learning_rate": 0.0008920539475472211, + "loss": 3.4537, + "step": 12710 + }, + { + "epoch": 0.863908139692893, + "grad_norm": 1.0002405643463135, + "learning_rate": 0.0008920114825383883, + "loss": 3.7209, + "step": 12715 + }, + { + "epoch": 0.8642478597635548, + "grad_norm": 1.170670747756958, + "learning_rate": 0.0008919690175295557, + "loss": 3.7614, + "step": 12720 + }, + { + "epoch": 0.8645875798342166, + "grad_norm": 0.9491438865661621, + "learning_rate": 0.000891926552520723, + "loss": 3.8023, + "step": 12725 + }, + { + "epoch": 0.8649272999048784, + "grad_norm": 0.7927387952804565, + "learning_rate": 0.0008918840875118902, + "loss": 3.6425, + "step": 12730 + }, + { + "epoch": 0.8652670199755401, + "grad_norm": 0.9267032742500305, + "learning_rate": 0.0008918416225030574, + "loss": 3.9812, + "step": 12735 + }, + { + "epoch": 0.865606740046202, + "grad_norm": 0.9640617966651917, + "learning_rate": 0.0008917991574942248, + "loss": 3.742, + "step": 12740 + }, + { + "epoch": 0.8659464601168637, + "grad_norm": 0.8019275069236755, + "learning_rate": 0.000891756692485392, + "loss": 3.7103, + "step": 12745 + }, + { + "epoch": 0.8662861801875255, + "grad_norm": 0.6883156895637512, + "learning_rate": 0.0008917142274765592, + "loss": 3.8125, + "step": 12750 + }, + { + "epoch": 0.8666259002581872, + "grad_norm": 0.7106084227561951, + "learning_rate": 0.0008916717624677267, + "loss": 3.5724, + "step": 12755 + }, + { + "epoch": 0.866965620328849, + "grad_norm": 0.7619748115539551, + "learning_rate": 0.0008916292974588939, + "loss": 3.6146, + "step": 12760 + }, + { + "epoch": 0.8673053403995108, + "grad_norm": 0.7439718246459961, + "learning_rate": 0.0008915868324500611, + "loss": 3.7144, + "step": 12765 + }, + { + "epoch": 0.8676450604701725, + "grad_norm": 0.7859118580818176, + "learning_rate": 0.0008915443674412285, + "loss": 3.7746, + "step": 12770 + }, + { + "epoch": 0.8679847805408344, + "grad_norm": 1.412725567817688, + "learning_rate": 0.0008915019024323957, + "loss": 3.8148, + "step": 12775 + }, + { + "epoch": 0.8683245006114961, + "grad_norm": 0.7228496670722961, + "learning_rate": 0.000891459437423563, + "loss": 3.6268, + "step": 12780 + }, + { + "epoch": 0.8686642206821579, + "grad_norm": 0.8363052606582642, + "learning_rate": 0.0008914169724147302, + "loss": 3.8245, + "step": 12785 + }, + { + "epoch": 0.8690039407528197, + "grad_norm": 1.4017343521118164, + "learning_rate": 0.0008913745074058976, + "loss": 3.592, + "step": 12790 + }, + { + "epoch": 0.8693436608234815, + "grad_norm": 0.6179370880126953, + "learning_rate": 0.0008913320423970649, + "loss": 3.724, + "step": 12795 + }, + { + "epoch": 0.8696833808941432, + "grad_norm": 0.8103336691856384, + "learning_rate": 0.0008912895773882321, + "loss": 3.8599, + "step": 12800 + }, + { + "epoch": 0.870023100964805, + "grad_norm": 0.7341117858886719, + "learning_rate": 0.0008912471123793994, + "loss": 3.8265, + "step": 12805 + }, + { + "epoch": 0.8703628210354668, + "grad_norm": 0.8728798627853394, + "learning_rate": 0.0008912046473705667, + "loss": 3.9017, + "step": 12810 + }, + { + "epoch": 0.8707025411061285, + "grad_norm": 1.0656594038009644, + "learning_rate": 0.0008911621823617339, + "loss": 3.6751, + "step": 12815 + }, + { + "epoch": 0.8710422611767903, + "grad_norm": 0.7849941253662109, + "learning_rate": 0.0008911197173529012, + "loss": 3.7614, + "step": 12820 + }, + { + "epoch": 0.8713819812474521, + "grad_norm": 1.0461782217025757, + "learning_rate": 0.0008910772523440686, + "loss": 3.8796, + "step": 12825 + }, + { + "epoch": 0.8717217013181139, + "grad_norm": 0.7844672799110413, + "learning_rate": 0.0008910347873352358, + "loss": 3.6107, + "step": 12830 + }, + { + "epoch": 0.8720614213887756, + "grad_norm": 0.733753502368927, + "learning_rate": 0.000890992322326403, + "loss": 3.4929, + "step": 12835 + }, + { + "epoch": 0.8724011414594374, + "grad_norm": 1.1124409437179565, + "learning_rate": 0.0008909498573175704, + "loss": 3.6283, + "step": 12840 + }, + { + "epoch": 0.8727408615300992, + "grad_norm": 0.8028335571289062, + "learning_rate": 0.0008909073923087376, + "loss": 3.5783, + "step": 12845 + }, + { + "epoch": 0.873080581600761, + "grad_norm": 0.8884498476982117, + "learning_rate": 0.0008908649272999048, + "loss": 3.9649, + "step": 12850 + }, + { + "epoch": 0.8734203016714227, + "grad_norm": 0.9446478486061096, + "learning_rate": 0.0008908224622910723, + "loss": 3.7274, + "step": 12855 + }, + { + "epoch": 0.8737600217420846, + "grad_norm": 0.7945959568023682, + "learning_rate": 0.0008907799972822395, + "loss": 3.6012, + "step": 12860 + }, + { + "epoch": 0.8740997418127463, + "grad_norm": 0.7741202712059021, + "learning_rate": 0.0008907375322734067, + "loss": 3.5847, + "step": 12865 + }, + { + "epoch": 0.874439461883408, + "grad_norm": 1.3905640840530396, + "learning_rate": 0.0008906950672645741, + "loss": 3.4079, + "step": 12870 + }, + { + "epoch": 0.8747791819540699, + "grad_norm": 0.8681594133377075, + "learning_rate": 0.0008906526022557413, + "loss": 3.7226, + "step": 12875 + }, + { + "epoch": 0.8751189020247316, + "grad_norm": 0.8703209757804871, + "learning_rate": 0.0008906101372469085, + "loss": 3.4136, + "step": 12880 + }, + { + "epoch": 0.8754586220953934, + "grad_norm": 2.2236297130584717, + "learning_rate": 0.0008905676722380758, + "loss": 3.6976, + "step": 12885 + }, + { + "epoch": 0.8757983421660551, + "grad_norm": 0.7414122819900513, + "learning_rate": 0.0008905252072292432, + "loss": 3.5893, + "step": 12890 + }, + { + "epoch": 0.876138062236717, + "grad_norm": 0.8235514163970947, + "learning_rate": 0.0008904827422204104, + "loss": 3.7593, + "step": 12895 + }, + { + "epoch": 0.8764777823073787, + "grad_norm": 1.1876682043075562, + "learning_rate": 0.0008904402772115777, + "loss": 3.8826, + "step": 12900 + }, + { + "epoch": 0.8768175023780405, + "grad_norm": 0.9803397059440613, + "learning_rate": 0.000890397812202745, + "loss": 3.6454, + "step": 12905 + }, + { + "epoch": 0.8771572224487023, + "grad_norm": 1.0319056510925293, + "learning_rate": 0.0008903553471939122, + "loss": 3.658, + "step": 12910 + }, + { + "epoch": 0.877496942519364, + "grad_norm": 1.392046570777893, + "learning_rate": 0.0008903128821850795, + "loss": 3.3554, + "step": 12915 + }, + { + "epoch": 0.8778366625900258, + "grad_norm": 1.0085598230361938, + "learning_rate": 0.0008902704171762468, + "loss": 3.6224, + "step": 12920 + }, + { + "epoch": 0.8781763826606876, + "grad_norm": 1.091357946395874, + "learning_rate": 0.0008902279521674141, + "loss": 3.4549, + "step": 12925 + }, + { + "epoch": 0.8785161027313494, + "grad_norm": 0.9164418578147888, + "learning_rate": 0.0008901854871585814, + "loss": 3.6929, + "step": 12930 + }, + { + "epoch": 0.8788558228020111, + "grad_norm": 0.8552413582801819, + "learning_rate": 0.0008901430221497486, + "loss": 3.662, + "step": 12935 + }, + { + "epoch": 0.8791955428726729, + "grad_norm": 1.0654689073562622, + "learning_rate": 0.0008901005571409159, + "loss": 3.6454, + "step": 12940 + }, + { + "epoch": 0.8795352629433347, + "grad_norm": 0.6858550906181335, + "learning_rate": 0.0008900580921320832, + "loss": 3.6554, + "step": 12945 + }, + { + "epoch": 0.8798749830139965, + "grad_norm": 0.8437381386756897, + "learning_rate": 0.0008900156271232504, + "loss": 3.4375, + "step": 12950 + }, + { + "epoch": 0.8802147030846582, + "grad_norm": 0.8676014542579651, + "learning_rate": 0.0008899731621144177, + "loss": 3.734, + "step": 12955 + }, + { + "epoch": 0.8805544231553201, + "grad_norm": 0.9134470224380493, + "learning_rate": 0.0008899306971055851, + "loss": 3.7288, + "step": 12960 + }, + { + "epoch": 0.8808941432259818, + "grad_norm": 0.8111168146133423, + "learning_rate": 0.0008898882320967523, + "loss": 3.7357, + "step": 12965 + }, + { + "epoch": 0.8812338632966435, + "grad_norm": 3.776362657546997, + "learning_rate": 0.0008898457670879196, + "loss": 3.7278, + "step": 12970 + }, + { + "epoch": 0.8815735833673053, + "grad_norm": 6.689911365509033, + "learning_rate": 0.0008898033020790869, + "loss": 3.6838, + "step": 12975 + }, + { + "epoch": 0.8819133034379671, + "grad_norm": 0.7403917908668518, + "learning_rate": 0.0008897608370702541, + "loss": 3.7995, + "step": 12980 + }, + { + "epoch": 0.8822530235086289, + "grad_norm": 0.8998690247535706, + "learning_rate": 0.0008897183720614213, + "loss": 3.5744, + "step": 12985 + }, + { + "epoch": 0.8825927435792906, + "grad_norm": 1.1085846424102783, + "learning_rate": 0.0008896759070525887, + "loss": 3.9121, + "step": 12990 + }, + { + "epoch": 0.8829324636499525, + "grad_norm": 0.7872055172920227, + "learning_rate": 0.000889633442043756, + "loss": 3.689, + "step": 12995 + }, + { + "epoch": 0.8832721837206142, + "grad_norm": 0.7448084354400635, + "learning_rate": 0.0008895909770349232, + "loss": 3.7417, + "step": 13000 + }, + { + "epoch": 0.883611903791276, + "grad_norm": 1.121294379234314, + "learning_rate": 0.0008895485120260906, + "loss": 3.7674, + "step": 13005 + }, + { + "epoch": 0.8839516238619378, + "grad_norm": 0.6590057611465454, + "learning_rate": 0.0008895060470172578, + "loss": 3.4937, + "step": 13010 + }, + { + "epoch": 0.8842913439325996, + "grad_norm": 0.7215134501457214, + "learning_rate": 0.000889463582008425, + "loss": 3.6161, + "step": 13015 + }, + { + "epoch": 0.8846310640032613, + "grad_norm": 0.6728622317314148, + "learning_rate": 0.0008894211169995924, + "loss": 3.5708, + "step": 13020 + }, + { + "epoch": 0.884970784073923, + "grad_norm": 0.889244556427002, + "learning_rate": 0.0008893786519907596, + "loss": 3.5425, + "step": 13025 + }, + { + "epoch": 0.8853105041445849, + "grad_norm": 0.9045710563659668, + "learning_rate": 0.0008893361869819269, + "loss": 3.6133, + "step": 13030 + }, + { + "epoch": 0.8856502242152466, + "grad_norm": 0.86453777551651, + "learning_rate": 0.0008892937219730942, + "loss": 3.7012, + "step": 13035 + }, + { + "epoch": 0.8859899442859084, + "grad_norm": 0.9362204670906067, + "learning_rate": 0.0008892512569642615, + "loss": 3.5917, + "step": 13040 + }, + { + "epoch": 0.8863296643565702, + "grad_norm": 0.7073429226875305, + "learning_rate": 0.0008892087919554287, + "loss": 3.6386, + "step": 13045 + }, + { + "epoch": 0.886669384427232, + "grad_norm": 0.7611358165740967, + "learning_rate": 0.000889166326946596, + "loss": 3.769, + "step": 13050 + }, + { + "epoch": 0.8870091044978937, + "grad_norm": 0.8660561442375183, + "learning_rate": 0.0008891238619377633, + "loss": 3.6306, + "step": 13055 + }, + { + "epoch": 0.8873488245685555, + "grad_norm": 1.0548138618469238, + "learning_rate": 0.0008890813969289305, + "loss": 3.5828, + "step": 13060 + }, + { + "epoch": 0.8876885446392173, + "grad_norm": 0.7869116067886353, + "learning_rate": 0.0008890389319200979, + "loss": 3.7106, + "step": 13065 + }, + { + "epoch": 0.8880282647098791, + "grad_norm": 0.929925799369812, + "learning_rate": 0.0008889964669112652, + "loss": 3.5634, + "step": 13070 + }, + { + "epoch": 0.8883679847805408, + "grad_norm": 0.8506751656532288, + "learning_rate": 0.0008889540019024324, + "loss": 3.7833, + "step": 13075 + }, + { + "epoch": 0.8887077048512027, + "grad_norm": 0.6812782287597656, + "learning_rate": 0.0008889115368935997, + "loss": 3.6973, + "step": 13080 + }, + { + "epoch": 0.8890474249218644, + "grad_norm": 0.8563469052314758, + "learning_rate": 0.0008888690718847669, + "loss": 3.6322, + "step": 13085 + }, + { + "epoch": 0.8893871449925261, + "grad_norm": 5.279268741607666, + "learning_rate": 0.0008888266068759342, + "loss": 3.7121, + "step": 13090 + }, + { + "epoch": 0.889726865063188, + "grad_norm": 0.8281797170639038, + "learning_rate": 0.0008887841418671015, + "loss": 3.6366, + "step": 13095 + }, + { + "epoch": 0.8900665851338497, + "grad_norm": 0.7403798699378967, + "learning_rate": 0.0008887416768582688, + "loss": 3.6392, + "step": 13100 + }, + { + "epoch": 0.8904063052045115, + "grad_norm": 0.790982723236084, + "learning_rate": 0.0008886992118494361, + "loss": 3.629, + "step": 13105 + }, + { + "epoch": 0.8907460252751732, + "grad_norm": 1.0384727716445923, + "learning_rate": 0.0008886567468406034, + "loss": 3.8037, + "step": 13110 + }, + { + "epoch": 0.8910857453458351, + "grad_norm": 0.7752336263656616, + "learning_rate": 0.0008886142818317706, + "loss": 3.6234, + "step": 13115 + }, + { + "epoch": 0.8914254654164968, + "grad_norm": 0.7246715426445007, + "learning_rate": 0.000888571816822938, + "loss": 3.8231, + "step": 13120 + }, + { + "epoch": 0.8917651854871586, + "grad_norm": 1.1059414148330688, + "learning_rate": 0.0008885293518141052, + "loss": 3.4965, + "step": 13125 + }, + { + "epoch": 0.8921049055578204, + "grad_norm": 0.9054039716720581, + "learning_rate": 0.0008884868868052724, + "loss": 3.454, + "step": 13130 + }, + { + "epoch": 0.8924446256284821, + "grad_norm": 1.0211822986602783, + "learning_rate": 0.0008884444217964399, + "loss": 3.5745, + "step": 13135 + }, + { + "epoch": 0.8927843456991439, + "grad_norm": 0.6604037284851074, + "learning_rate": 0.0008884019567876071, + "loss": 3.7203, + "step": 13140 + }, + { + "epoch": 0.8931240657698056, + "grad_norm": 1.4752808809280396, + "learning_rate": 0.0008883594917787743, + "loss": 3.823, + "step": 13145 + }, + { + "epoch": 0.8934637858404675, + "grad_norm": 0.8650283813476562, + "learning_rate": 0.0008883170267699416, + "loss": 3.643, + "step": 13150 + }, + { + "epoch": 0.8938035059111292, + "grad_norm": 0.7194262146949768, + "learning_rate": 0.0008882745617611089, + "loss": 3.8213, + "step": 13155 + }, + { + "epoch": 0.894143225981791, + "grad_norm": 0.849518358707428, + "learning_rate": 0.0008882320967522761, + "loss": 3.4095, + "step": 13160 + }, + { + "epoch": 0.8944829460524528, + "grad_norm": 1.9330400228500366, + "learning_rate": 0.0008881896317434434, + "loss": 3.4546, + "step": 13165 + }, + { + "epoch": 0.8948226661231146, + "grad_norm": 1.4759125709533691, + "learning_rate": 0.0008881471667346108, + "loss": 3.4915, + "step": 13170 + }, + { + "epoch": 0.8951623861937763, + "grad_norm": 0.5968790650367737, + "learning_rate": 0.000888104701725778, + "loss": 3.7624, + "step": 13175 + }, + { + "epoch": 0.8955021062644382, + "grad_norm": 0.9586802124977112, + "learning_rate": 0.0008880622367169453, + "loss": 3.6566, + "step": 13180 + }, + { + "epoch": 0.8958418263350999, + "grad_norm": 0.8077282309532166, + "learning_rate": 0.0008880197717081125, + "loss": 3.8535, + "step": 13185 + }, + { + "epoch": 0.8961815464057616, + "grad_norm": 2.1277451515197754, + "learning_rate": 0.0008879773066992798, + "loss": 3.6182, + "step": 13190 + }, + { + "epoch": 0.8965212664764234, + "grad_norm": 0.901930570602417, + "learning_rate": 0.0008879348416904471, + "loss": 3.758, + "step": 13195 + }, + { + "epoch": 0.8968609865470852, + "grad_norm": 0.7633906602859497, + "learning_rate": 0.0008878923766816143, + "loss": 3.6279, + "step": 13200 + }, + { + "epoch": 0.897200706617747, + "grad_norm": 0.8146492838859558, + "learning_rate": 0.0008878499116727817, + "loss": 3.5136, + "step": 13205 + }, + { + "epoch": 0.8975404266884087, + "grad_norm": 0.702854573726654, + "learning_rate": 0.000887807446663949, + "loss": 3.5898, + "step": 13210 + }, + { + "epoch": 0.8978801467590706, + "grad_norm": 0.9555582404136658, + "learning_rate": 0.0008877649816551162, + "loss": 3.6541, + "step": 13215 + }, + { + "epoch": 0.8982198668297323, + "grad_norm": 0.8376206159591675, + "learning_rate": 0.0008877225166462834, + "loss": 3.7081, + "step": 13220 + }, + { + "epoch": 0.8985595869003941, + "grad_norm": 0.8409804105758667, + "learning_rate": 0.0008876800516374508, + "loss": 3.8191, + "step": 13225 + }, + { + "epoch": 0.8988993069710558, + "grad_norm": 0.9086755514144897, + "learning_rate": 0.000887637586628618, + "loss": 3.8514, + "step": 13230 + }, + { + "epoch": 0.8992390270417177, + "grad_norm": 1.650781512260437, + "learning_rate": 0.0008875951216197852, + "loss": 3.5137, + "step": 13235 + }, + { + "epoch": 0.8995787471123794, + "grad_norm": 0.742462158203125, + "learning_rate": 0.0008875526566109527, + "loss": 3.8492, + "step": 13240 + }, + { + "epoch": 0.8999184671830411, + "grad_norm": 0.767926037311554, + "learning_rate": 0.0008875101916021199, + "loss": 3.4037, + "step": 13245 + }, + { + "epoch": 0.900258187253703, + "grad_norm": 0.8409079313278198, + "learning_rate": 0.0008874677265932871, + "loss": 3.6168, + "step": 13250 + }, + { + "epoch": 0.9005979073243647, + "grad_norm": 0.8704046010971069, + "learning_rate": 0.0008874252615844545, + "loss": 3.6999, + "step": 13255 + }, + { + "epoch": 0.9009376273950265, + "grad_norm": 0.7593732476234436, + "learning_rate": 0.0008873827965756217, + "loss": 3.7242, + "step": 13260 + }, + { + "epoch": 0.9012773474656883, + "grad_norm": 0.7422963380813599, + "learning_rate": 0.0008873403315667889, + "loss": 3.6065, + "step": 13265 + }, + { + "epoch": 0.9016170675363501, + "grad_norm": 1.0189372301101685, + "learning_rate": 0.0008872978665579562, + "loss": 3.6786, + "step": 13270 + }, + { + "epoch": 0.9019567876070118, + "grad_norm": 0.9683796167373657, + "learning_rate": 0.0008872554015491236, + "loss": 3.8129, + "step": 13275 + }, + { + "epoch": 0.9022965076776736, + "grad_norm": 1.0309350490570068, + "learning_rate": 0.0008872129365402908, + "loss": 3.884, + "step": 13280 + }, + { + "epoch": 0.9026362277483354, + "grad_norm": 0.7175368070602417, + "learning_rate": 0.0008871704715314581, + "loss": 3.6692, + "step": 13285 + }, + { + "epoch": 0.9029759478189971, + "grad_norm": 1.0159885883331299, + "learning_rate": 0.0008871280065226254, + "loss": 3.7348, + "step": 13290 + }, + { + "epoch": 0.9033156678896589, + "grad_norm": 1.0885182619094849, + "learning_rate": 0.0008870855415137926, + "loss": 3.8768, + "step": 13295 + }, + { + "epoch": 0.9036553879603207, + "grad_norm": 0.7318781614303589, + "learning_rate": 0.0008870430765049599, + "loss": 3.772, + "step": 13300 + }, + { + "epoch": 0.9039951080309825, + "grad_norm": 3.8705921173095703, + "learning_rate": 0.0008870006114961272, + "loss": 3.4644, + "step": 13305 + }, + { + "epoch": 0.9043348281016442, + "grad_norm": 0.6893762946128845, + "learning_rate": 0.0008869581464872945, + "loss": 3.8524, + "step": 13310 + }, + { + "epoch": 0.904674548172306, + "grad_norm": 0.6884678602218628, + "learning_rate": 0.0008869156814784618, + "loss": 3.9995, + "step": 13315 + }, + { + "epoch": 0.9050142682429678, + "grad_norm": 1.6105153560638428, + "learning_rate": 0.000886873216469629, + "loss": 3.4808, + "step": 13320 + }, + { + "epoch": 0.9053539883136296, + "grad_norm": 0.7443768978118896, + "learning_rate": 0.0008868307514607963, + "loss": 3.5492, + "step": 13325 + }, + { + "epoch": 0.9056937083842913, + "grad_norm": 0.8776019811630249, + "learning_rate": 0.0008867882864519636, + "loss": 3.5964, + "step": 13330 + }, + { + "epoch": 0.9060334284549532, + "grad_norm": 0.9091726541519165, + "learning_rate": 0.0008867458214431308, + "loss": 3.79, + "step": 13335 + }, + { + "epoch": 0.9063731485256149, + "grad_norm": 1.0782090425491333, + "learning_rate": 0.0008867033564342981, + "loss": 3.6298, + "step": 13340 + }, + { + "epoch": 0.9067128685962766, + "grad_norm": 0.7235108017921448, + "learning_rate": 0.0008866608914254655, + "loss": 3.763, + "step": 13345 + }, + { + "epoch": 0.9070525886669385, + "grad_norm": 1.042415738105774, + "learning_rate": 0.0008866184264166327, + "loss": 3.6742, + "step": 13350 + }, + { + "epoch": 0.9073923087376002, + "grad_norm": 0.8211411237716675, + "learning_rate": 0.0008865759614078, + "loss": 3.5995, + "step": 13355 + }, + { + "epoch": 0.907732028808262, + "grad_norm": 0.9127249717712402, + "learning_rate": 0.0008865334963989673, + "loss": 3.5836, + "step": 13360 + }, + { + "epoch": 0.9080717488789237, + "grad_norm": 1.2099254131317139, + "learning_rate": 0.0008864910313901345, + "loss": 3.6074, + "step": 13365 + }, + { + "epoch": 0.9084114689495856, + "grad_norm": 1.1270442008972168, + "learning_rate": 0.0008864485663813017, + "loss": 3.6886, + "step": 13370 + }, + { + "epoch": 0.9087511890202473, + "grad_norm": 0.7445120811462402, + "learning_rate": 0.0008864061013724691, + "loss": 3.2686, + "step": 13375 + }, + { + "epoch": 0.9090909090909091, + "grad_norm": 0.7615050673484802, + "learning_rate": 0.0008863636363636364, + "loss": 3.8553, + "step": 13380 + }, + { + "epoch": 0.9094306291615709, + "grad_norm": 0.871473491191864, + "learning_rate": 0.0008863211713548036, + "loss": 3.9049, + "step": 13385 + }, + { + "epoch": 0.9097703492322327, + "grad_norm": 0.9612770080566406, + "learning_rate": 0.000886278706345971, + "loss": 3.7613, + "step": 13390 + }, + { + "epoch": 0.9101100693028944, + "grad_norm": 6.434609889984131, + "learning_rate": 0.0008862362413371382, + "loss": 3.5481, + "step": 13395 + }, + { + "epoch": 0.9104497893735561, + "grad_norm": 1.103271245956421, + "learning_rate": 0.0008861937763283054, + "loss": 3.418, + "step": 13400 + }, + { + "epoch": 0.910789509444218, + "grad_norm": 1.239216685295105, + "learning_rate": 0.0008861513113194728, + "loss": 4.1605, + "step": 13405 + }, + { + "epoch": 0.9111292295148797, + "grad_norm": 0.8537166118621826, + "learning_rate": 0.00088610884631064, + "loss": 3.4811, + "step": 13410 + }, + { + "epoch": 0.9114689495855415, + "grad_norm": 0.6005226969718933, + "learning_rate": 0.0008860663813018073, + "loss": 3.6195, + "step": 13415 + }, + { + "epoch": 0.9118086696562033, + "grad_norm": 0.7821482419967651, + "learning_rate": 0.0008860239162929747, + "loss": 3.8903, + "step": 13420 + }, + { + "epoch": 0.9121483897268651, + "grad_norm": 0.9275714755058289, + "learning_rate": 0.0008859814512841419, + "loss": 3.6626, + "step": 13425 + }, + { + "epoch": 0.9124881097975268, + "grad_norm": 1.5872000455856323, + "learning_rate": 0.0008859389862753091, + "loss": 3.4835, + "step": 13430 + }, + { + "epoch": 0.9128278298681887, + "grad_norm": 0.9722049832344055, + "learning_rate": 0.0008858965212664764, + "loss": 3.6821, + "step": 13435 + }, + { + "epoch": 0.9131675499388504, + "grad_norm": 0.9756625890731812, + "learning_rate": 0.0008858540562576437, + "loss": 3.4526, + "step": 13440 + }, + { + "epoch": 0.9135072700095122, + "grad_norm": 0.8384751081466675, + "learning_rate": 0.0008858115912488109, + "loss": 3.5765, + "step": 13445 + }, + { + "epoch": 0.9138469900801739, + "grad_norm": 0.7821000814437866, + "learning_rate": 0.0008857691262399783, + "loss": 3.5271, + "step": 13450 + }, + { + "epoch": 0.9141867101508357, + "grad_norm": 0.7574872970581055, + "learning_rate": 0.0008857266612311456, + "loss": 3.5454, + "step": 13455 + }, + { + "epoch": 0.9145264302214975, + "grad_norm": 0.8772850036621094, + "learning_rate": 0.0008856841962223129, + "loss": 3.9148, + "step": 13460 + }, + { + "epoch": 0.9148661502921592, + "grad_norm": 0.8406215310096741, + "learning_rate": 0.0008856417312134801, + "loss": 3.737, + "step": 13465 + }, + { + "epoch": 0.9152058703628211, + "grad_norm": 0.855576753616333, + "learning_rate": 0.0008855992662046473, + "loss": 3.8133, + "step": 13470 + }, + { + "epoch": 0.9155455904334828, + "grad_norm": 6.184667587280273, + "learning_rate": 0.0008855568011958147, + "loss": 3.6333, + "step": 13475 + }, + { + "epoch": 0.9158853105041446, + "grad_norm": 0.835614800453186, + "learning_rate": 0.000885514336186982, + "loss": 3.502, + "step": 13480 + }, + { + "epoch": 0.9162250305748063, + "grad_norm": 1.0329148769378662, + "learning_rate": 0.0008854718711781492, + "loss": 3.6392, + "step": 13485 + }, + { + "epoch": 0.9165647506454682, + "grad_norm": 1.8290917873382568, + "learning_rate": 0.0008854294061693166, + "loss": 3.7724, + "step": 13490 + }, + { + "epoch": 0.9169044707161299, + "grad_norm": 0.9795731902122498, + "learning_rate": 0.0008853869411604838, + "loss": 3.7244, + "step": 13495 + }, + { + "epoch": 0.9172441907867916, + "grad_norm": 0.7607027888298035, + "learning_rate": 0.000885344476151651, + "loss": 3.5909, + "step": 13500 + }, + { + "epoch": 0.9175839108574535, + "grad_norm": 1.6879788637161255, + "learning_rate": 0.0008853020111428184, + "loss": 3.7135, + "step": 13505 + }, + { + "epoch": 0.9179236309281152, + "grad_norm": 1.1392910480499268, + "learning_rate": 0.0008852595461339856, + "loss": 3.9939, + "step": 13510 + }, + { + "epoch": 0.918263350998777, + "grad_norm": 0.8106732964515686, + "learning_rate": 0.0008852170811251529, + "loss": 3.2409, + "step": 13515 + }, + { + "epoch": 0.9186030710694388, + "grad_norm": 0.9688677787780762, + "learning_rate": 0.0008851746161163203, + "loss": 3.7299, + "step": 13520 + }, + { + "epoch": 0.9189427911401006, + "grad_norm": 1.7550389766693115, + "learning_rate": 0.0008851321511074875, + "loss": 3.6679, + "step": 13525 + }, + { + "epoch": 0.9192825112107623, + "grad_norm": 0.894064724445343, + "learning_rate": 0.0008850896860986547, + "loss": 3.386, + "step": 13530 + }, + { + "epoch": 0.9196222312814241, + "grad_norm": 0.8156725764274597, + "learning_rate": 0.000885047221089822, + "loss": 3.8207, + "step": 13535 + }, + { + "epoch": 0.9199619513520859, + "grad_norm": 0.7993063926696777, + "learning_rate": 0.0008850047560809893, + "loss": 3.8328, + "step": 13540 + }, + { + "epoch": 0.9203016714227477, + "grad_norm": 0.9036015868186951, + "learning_rate": 0.0008849622910721565, + "loss": 3.551, + "step": 13545 + }, + { + "epoch": 0.9206413914934094, + "grad_norm": 0.7851384282112122, + "learning_rate": 0.0008849198260633239, + "loss": 3.8335, + "step": 13550 + }, + { + "epoch": 0.9209811115640713, + "grad_norm": 0.7969225645065308, + "learning_rate": 0.0008848773610544912, + "loss": 3.6863, + "step": 13555 + }, + { + "epoch": 0.921320831634733, + "grad_norm": 0.9233984351158142, + "learning_rate": 0.0008848348960456584, + "loss": 3.633, + "step": 13560 + }, + { + "epoch": 0.9216605517053947, + "grad_norm": 1.0682228803634644, + "learning_rate": 0.0008847924310368257, + "loss": 3.7084, + "step": 13565 + }, + { + "epoch": 0.9220002717760565, + "grad_norm": 0.7216653227806091, + "learning_rate": 0.000884749966027993, + "loss": 3.6827, + "step": 13570 + }, + { + "epoch": 0.9223399918467183, + "grad_norm": 0.728904664516449, + "learning_rate": 0.0008847075010191602, + "loss": 3.6725, + "step": 13575 + }, + { + "epoch": 0.9226797119173801, + "grad_norm": 1.3324331045150757, + "learning_rate": 0.0008846650360103275, + "loss": 3.5884, + "step": 13580 + }, + { + "epoch": 0.9230194319880418, + "grad_norm": 0.8203989267349243, + "learning_rate": 0.0008846225710014948, + "loss": 3.4003, + "step": 13585 + }, + { + "epoch": 0.9233591520587037, + "grad_norm": 0.8179434537887573, + "learning_rate": 0.0008845801059926621, + "loss": 3.7058, + "step": 13590 + }, + { + "epoch": 0.9236988721293654, + "grad_norm": 1.7413372993469238, + "learning_rate": 0.0008845376409838294, + "loss": 3.6786, + "step": 13595 + }, + { + "epoch": 0.9240385922000272, + "grad_norm": 0.7709000110626221, + "learning_rate": 0.0008844951759749966, + "loss": 3.5964, + "step": 13600 + }, + { + "epoch": 0.924378312270689, + "grad_norm": 0.8884493708610535, + "learning_rate": 0.0008844527109661639, + "loss": 3.8318, + "step": 13605 + }, + { + "epoch": 0.9247180323413507, + "grad_norm": 0.7240601778030396, + "learning_rate": 0.0008844102459573312, + "loss": 3.8144, + "step": 13610 + }, + { + "epoch": 0.9250577524120125, + "grad_norm": 0.8557507991790771, + "learning_rate": 0.0008843677809484984, + "loss": 3.5952, + "step": 13615 + }, + { + "epoch": 0.9253974724826742, + "grad_norm": 0.8917243480682373, + "learning_rate": 0.0008843253159396657, + "loss": 3.6763, + "step": 13620 + }, + { + "epoch": 0.9257371925533361, + "grad_norm": 0.7802959680557251, + "learning_rate": 0.0008842828509308331, + "loss": 3.7661, + "step": 13625 + }, + { + "epoch": 0.9260769126239978, + "grad_norm": 0.6309027075767517, + "learning_rate": 0.0008842403859220003, + "loss": 3.5654, + "step": 13630 + }, + { + "epoch": 0.9264166326946596, + "grad_norm": 1.106672763824463, + "learning_rate": 0.0008841979209131675, + "loss": 3.7878, + "step": 13635 + }, + { + "epoch": 0.9267563527653214, + "grad_norm": 0.8464555144309998, + "learning_rate": 0.0008841554559043349, + "loss": 3.6785, + "step": 13640 + }, + { + "epoch": 0.9270960728359832, + "grad_norm": 1.3956855535507202, + "learning_rate": 0.0008841129908955021, + "loss": 3.4541, + "step": 13645 + }, + { + "epoch": 0.9274357929066449, + "grad_norm": 0.9163585305213928, + "learning_rate": 0.0008840705258866693, + "loss": 3.8821, + "step": 13650 + }, + { + "epoch": 0.9277755129773066, + "grad_norm": 0.9337664842605591, + "learning_rate": 0.0008840280608778368, + "loss": 3.7544, + "step": 13655 + }, + { + "epoch": 0.9281152330479685, + "grad_norm": 0.6309140920639038, + "learning_rate": 0.000883985595869004, + "loss": 3.8096, + "step": 13660 + }, + { + "epoch": 0.9284549531186302, + "grad_norm": 0.8382517695426941, + "learning_rate": 0.0008839431308601712, + "loss": 3.6736, + "step": 13665 + }, + { + "epoch": 0.928794673189292, + "grad_norm": 0.7094393968582153, + "learning_rate": 0.0008839006658513385, + "loss": 3.8144, + "step": 13670 + }, + { + "epoch": 0.9291343932599538, + "grad_norm": 1.1005032062530518, + "learning_rate": 0.0008838582008425058, + "loss": 3.7603, + "step": 13675 + }, + { + "epoch": 0.9294741133306156, + "grad_norm": 1.0306353569030762, + "learning_rate": 0.000883815735833673, + "loss": 3.7025, + "step": 13680 + }, + { + "epoch": 0.9298138334012773, + "grad_norm": 0.8249132633209229, + "learning_rate": 0.0008837732708248403, + "loss": 3.4602, + "step": 13685 + }, + { + "epoch": 0.9301535534719392, + "grad_norm": 0.6795862913131714, + "learning_rate": 0.0008837308058160077, + "loss": 3.5854, + "step": 13690 + }, + { + "epoch": 0.9304932735426009, + "grad_norm": 0.8747157454490662, + "learning_rate": 0.0008836883408071749, + "loss": 3.5667, + "step": 13695 + }, + { + "epoch": 0.9308329936132627, + "grad_norm": 0.7587159872055054, + "learning_rate": 0.0008836458757983422, + "loss": 3.5191, + "step": 13700 + }, + { + "epoch": 0.9311727136839244, + "grad_norm": 0.8577826023101807, + "learning_rate": 0.0008836034107895095, + "loss": 3.6191, + "step": 13705 + }, + { + "epoch": 0.9315124337545863, + "grad_norm": 0.9343301057815552, + "learning_rate": 0.0008835609457806767, + "loss": 3.6625, + "step": 13710 + }, + { + "epoch": 0.931852153825248, + "grad_norm": 0.8171669840812683, + "learning_rate": 0.000883518480771844, + "loss": 3.766, + "step": 13715 + }, + { + "epoch": 0.9321918738959097, + "grad_norm": 0.8000214695930481, + "learning_rate": 0.0008834760157630112, + "loss": 3.6236, + "step": 13720 + }, + { + "epoch": 0.9325315939665716, + "grad_norm": 0.9358586668968201, + "learning_rate": 0.0008834335507541786, + "loss": 3.4313, + "step": 13725 + }, + { + "epoch": 0.9328713140372333, + "grad_norm": 0.7570367455482483, + "learning_rate": 0.0008833910857453459, + "loss": 3.6079, + "step": 13730 + }, + { + "epoch": 0.9332110341078951, + "grad_norm": 1.0521817207336426, + "learning_rate": 0.0008833486207365131, + "loss": 3.8812, + "step": 13735 + }, + { + "epoch": 0.9335507541785568, + "grad_norm": 0.6636382341384888, + "learning_rate": 0.0008833061557276804, + "loss": 3.6594, + "step": 13740 + }, + { + "epoch": 0.9338904742492187, + "grad_norm": 0.8521907925605774, + "learning_rate": 0.0008832636907188477, + "loss": 3.7076, + "step": 13745 + }, + { + "epoch": 0.9342301943198804, + "grad_norm": 0.919730544090271, + "learning_rate": 0.0008832212257100149, + "loss": 3.237, + "step": 13750 + }, + { + "epoch": 0.9345699143905422, + "grad_norm": 0.7903971076011658, + "learning_rate": 0.0008831787607011821, + "loss": 3.8059, + "step": 13755 + }, + { + "epoch": 0.934909634461204, + "grad_norm": 0.8670375347137451, + "learning_rate": 0.0008831362956923496, + "loss": 3.591, + "step": 13760 + }, + { + "epoch": 0.9352493545318658, + "grad_norm": 1.3178188800811768, + "learning_rate": 0.0008830938306835168, + "loss": 3.6953, + "step": 13765 + }, + { + "epoch": 0.9355890746025275, + "grad_norm": 0.8890280723571777, + "learning_rate": 0.000883051365674684, + "loss": 3.7027, + "step": 13770 + }, + { + "epoch": 0.9359287946731893, + "grad_norm": 0.8231936097145081, + "learning_rate": 0.0008830089006658514, + "loss": 3.4748, + "step": 13775 + }, + { + "epoch": 0.9362685147438511, + "grad_norm": 0.9336920976638794, + "learning_rate": 0.0008829664356570186, + "loss": 3.5724, + "step": 13780 + }, + { + "epoch": 0.9366082348145128, + "grad_norm": 0.6202356219291687, + "learning_rate": 0.0008829239706481858, + "loss": 3.7062, + "step": 13785 + }, + { + "epoch": 0.9369479548851746, + "grad_norm": 0.7148352265357971, + "learning_rate": 0.0008828815056393532, + "loss": 3.6556, + "step": 13790 + }, + { + "epoch": 0.9372876749558364, + "grad_norm": 0.9888379573822021, + "learning_rate": 0.0008828390406305205, + "loss": 3.7264, + "step": 13795 + }, + { + "epoch": 0.9376273950264982, + "grad_norm": 0.8397827744483948, + "learning_rate": 0.0008827965756216878, + "loss": 3.4699, + "step": 13800 + }, + { + "epoch": 0.9379671150971599, + "grad_norm": 0.6670814752578735, + "learning_rate": 0.0008827541106128551, + "loss": 3.6653, + "step": 13805 + }, + { + "epoch": 0.9383068351678218, + "grad_norm": 0.8079091906547546, + "learning_rate": 0.0008827116456040223, + "loss": 3.3031, + "step": 13810 + }, + { + "epoch": 0.9386465552384835, + "grad_norm": 0.8372642397880554, + "learning_rate": 0.0008826691805951896, + "loss": 3.6295, + "step": 13815 + }, + { + "epoch": 0.9389862753091452, + "grad_norm": 0.6966977119445801, + "learning_rate": 0.0008826267155863568, + "loss": 3.7738, + "step": 13820 + }, + { + "epoch": 0.939325995379807, + "grad_norm": 0.9013363122940063, + "learning_rate": 0.0008825842505775241, + "loss": 3.2969, + "step": 13825 + }, + { + "epoch": 0.9396657154504688, + "grad_norm": 1.0251212120056152, + "learning_rate": 0.0008825417855686915, + "loss": 3.6845, + "step": 13830 + }, + { + "epoch": 0.9400054355211306, + "grad_norm": 0.7265050411224365, + "learning_rate": 0.0008824993205598587, + "loss": 3.5742, + "step": 13835 + }, + { + "epoch": 0.9403451555917923, + "grad_norm": 0.6548272371292114, + "learning_rate": 0.000882456855551026, + "loss": 3.6591, + "step": 13840 + }, + { + "epoch": 0.9406848756624542, + "grad_norm": 0.99984210729599, + "learning_rate": 0.0008824143905421933, + "loss": 3.5607, + "step": 13845 + }, + { + "epoch": 0.9410245957331159, + "grad_norm": 0.6612033247947693, + "learning_rate": 0.0008823719255333605, + "loss": 3.812, + "step": 13850 + }, + { + "epoch": 0.9413643158037777, + "grad_norm": 0.662527322769165, + "learning_rate": 0.0008823294605245277, + "loss": 3.8006, + "step": 13855 + }, + { + "epoch": 0.9417040358744395, + "grad_norm": 0.7916511297225952, + "learning_rate": 0.0008822869955156951, + "loss": 3.6858, + "step": 13860 + }, + { + "epoch": 0.9420437559451013, + "grad_norm": 1.1828975677490234, + "learning_rate": 0.0008822445305068624, + "loss": 3.4239, + "step": 13865 + }, + { + "epoch": 0.942383476015763, + "grad_norm": 2.4112625122070312, + "learning_rate": 0.0008822020654980296, + "loss": 3.6426, + "step": 13870 + }, + { + "epoch": 0.9427231960864247, + "grad_norm": 0.8874445557594299, + "learning_rate": 0.000882159600489197, + "loss": 3.6877, + "step": 13875 + }, + { + "epoch": 0.9430629161570866, + "grad_norm": 0.91973876953125, + "learning_rate": 0.0008821171354803642, + "loss": 3.6937, + "step": 13880 + }, + { + "epoch": 0.9434026362277483, + "grad_norm": 0.8752864599227905, + "learning_rate": 0.0008820746704715314, + "loss": 3.4457, + "step": 13885 + }, + { + "epoch": 0.9437423562984101, + "grad_norm": 0.7345873117446899, + "learning_rate": 0.0008820322054626988, + "loss": 3.8366, + "step": 13890 + }, + { + "epoch": 0.9440820763690719, + "grad_norm": 1.2251546382904053, + "learning_rate": 0.000881989740453866, + "loss": 3.8125, + "step": 13895 + }, + { + "epoch": 0.9444217964397337, + "grad_norm": 1.0018945932388306, + "learning_rate": 0.0008819472754450333, + "loss": 3.59, + "step": 13900 + }, + { + "epoch": 0.9447615165103954, + "grad_norm": 0.8775460720062256, + "learning_rate": 0.0008819048104362007, + "loss": 3.649, + "step": 13905 + }, + { + "epoch": 0.9451012365810572, + "grad_norm": 0.8337073922157288, + "learning_rate": 0.0008818623454273679, + "loss": 3.8179, + "step": 13910 + }, + { + "epoch": 0.945440956651719, + "grad_norm": 0.8015432953834534, + "learning_rate": 0.0008818198804185351, + "loss": 3.6612, + "step": 13915 + }, + { + "epoch": 0.9457806767223808, + "grad_norm": 0.8323782682418823, + "learning_rate": 0.0008817774154097024, + "loss": 3.5186, + "step": 13920 + }, + { + "epoch": 0.9461203967930425, + "grad_norm": 0.6812918186187744, + "learning_rate": 0.0008817349504008697, + "loss": 3.8222, + "step": 13925 + }, + { + "epoch": 0.9464601168637043, + "grad_norm": 0.8665046095848083, + "learning_rate": 0.0008816924853920369, + "loss": 3.5126, + "step": 13930 + }, + { + "epoch": 0.9467998369343661, + "grad_norm": 0.8575319051742554, + "learning_rate": 0.0008816500203832043, + "loss": 3.5755, + "step": 13935 + }, + { + "epoch": 0.9471395570050278, + "grad_norm": 1.6602516174316406, + "learning_rate": 0.0008816075553743716, + "loss": 3.8821, + "step": 13940 + }, + { + "epoch": 0.9474792770756897, + "grad_norm": 1.1572099924087524, + "learning_rate": 0.0008815650903655388, + "loss": 3.6644, + "step": 13945 + }, + { + "epoch": 0.9478189971463514, + "grad_norm": 0.8157739043235779, + "learning_rate": 0.0008815226253567061, + "loss": 3.726, + "step": 13950 + }, + { + "epoch": 0.9481587172170132, + "grad_norm": 0.7523817420005798, + "learning_rate": 0.0008814801603478733, + "loss": 3.3739, + "step": 13955 + }, + { + "epoch": 0.9484984372876749, + "grad_norm": 0.7100839018821716, + "learning_rate": 0.0008814376953390406, + "loss": 3.5778, + "step": 13960 + }, + { + "epoch": 0.9488381573583368, + "grad_norm": 0.7609274983406067, + "learning_rate": 0.000881395230330208, + "loss": 3.4127, + "step": 13965 + }, + { + "epoch": 0.9491778774289985, + "grad_norm": 1.1577061414718628, + "learning_rate": 0.0008813527653213752, + "loss": 3.8804, + "step": 13970 + }, + { + "epoch": 0.9495175974996602, + "grad_norm": 0.8294179439544678, + "learning_rate": 0.0008813103003125425, + "loss": 3.6693, + "step": 13975 + }, + { + "epoch": 0.9498573175703221, + "grad_norm": 0.8605481386184692, + "learning_rate": 0.0008812678353037098, + "loss": 3.7811, + "step": 13980 + }, + { + "epoch": 0.9501970376409838, + "grad_norm": 0.9040095210075378, + "learning_rate": 0.000881225370294877, + "loss": 3.5954, + "step": 13985 + }, + { + "epoch": 0.9505367577116456, + "grad_norm": 0.6817090511322021, + "learning_rate": 0.0008811829052860443, + "loss": 3.6321, + "step": 13990 + }, + { + "epoch": 0.9508764777823073, + "grad_norm": 0.8138971328735352, + "learning_rate": 0.0008811404402772116, + "loss": 3.7651, + "step": 13995 + }, + { + "epoch": 0.9512161978529692, + "grad_norm": 0.7704663276672363, + "learning_rate": 0.0008810979752683789, + "loss": 3.5433, + "step": 14000 + }, + { + "epoch": 0.9515559179236309, + "grad_norm": 0.6035419702529907, + "learning_rate": 0.0008810555102595462, + "loss": 3.3425, + "step": 14005 + }, + { + "epoch": 0.9518956379942927, + "grad_norm": 0.7641417980194092, + "learning_rate": 0.0008810130452507135, + "loss": 3.4852, + "step": 14010 + }, + { + "epoch": 0.9522353580649545, + "grad_norm": 0.8048719763755798, + "learning_rate": 0.0008809705802418807, + "loss": 3.7314, + "step": 14015 + }, + { + "epoch": 0.9525750781356163, + "grad_norm": 0.7399217486381531, + "learning_rate": 0.0008809281152330479, + "loss": 3.6636, + "step": 14020 + }, + { + "epoch": 0.952914798206278, + "grad_norm": 0.816810667514801, + "learning_rate": 0.0008808856502242153, + "loss": 3.7608, + "step": 14025 + }, + { + "epoch": 0.9532545182769399, + "grad_norm": 0.6726149320602417, + "learning_rate": 0.0008808431852153825, + "loss": 3.6351, + "step": 14030 + }, + { + "epoch": 0.9535942383476016, + "grad_norm": 0.7571033239364624, + "learning_rate": 0.0008808007202065498, + "loss": 3.6505, + "step": 14035 + }, + { + "epoch": 0.9539339584182633, + "grad_norm": 0.6782366633415222, + "learning_rate": 0.0008807582551977172, + "loss": 3.8236, + "step": 14040 + }, + { + "epoch": 0.9542736784889251, + "grad_norm": 1.115885853767395, + "learning_rate": 0.0008807157901888844, + "loss": 3.7044, + "step": 14045 + }, + { + "epoch": 0.9546133985595869, + "grad_norm": 0.8412324786186218, + "learning_rate": 0.0008806733251800516, + "loss": 3.9145, + "step": 14050 + }, + { + "epoch": 0.9549531186302487, + "grad_norm": 0.77687007188797, + "learning_rate": 0.000880630860171219, + "loss": 3.6443, + "step": 14055 + }, + { + "epoch": 0.9552928387009104, + "grad_norm": 0.7641855478286743, + "learning_rate": 0.0008805883951623862, + "loss": 3.53, + "step": 14060 + }, + { + "epoch": 0.9556325587715723, + "grad_norm": 0.7573676109313965, + "learning_rate": 0.0008805459301535534, + "loss": 3.6638, + "step": 14065 + }, + { + "epoch": 0.955972278842234, + "grad_norm": 0.781079888343811, + "learning_rate": 0.0008805034651447208, + "loss": 3.5952, + "step": 14070 + }, + { + "epoch": 0.9563119989128958, + "grad_norm": 0.7804728746414185, + "learning_rate": 0.0008804610001358881, + "loss": 3.7907, + "step": 14075 + }, + { + "epoch": 0.9566517189835575, + "grad_norm": 1.077364444732666, + "learning_rate": 0.0008804185351270553, + "loss": 3.6663, + "step": 14080 + }, + { + "epoch": 0.9569914390542194, + "grad_norm": 0.7271178364753723, + "learning_rate": 0.0008803760701182226, + "loss": 3.5515, + "step": 14085 + }, + { + "epoch": 0.9573311591248811, + "grad_norm": 0.8874168992042542, + "learning_rate": 0.0008803336051093899, + "loss": 3.647, + "step": 14090 + }, + { + "epoch": 0.9576708791955428, + "grad_norm": 0.7073732614517212, + "learning_rate": 0.0008802911401005571, + "loss": 3.7275, + "step": 14095 + }, + { + "epoch": 0.9580105992662047, + "grad_norm": 1.1617673635482788, + "learning_rate": 0.0008802486750917244, + "loss": 3.5342, + "step": 14100 + }, + { + "epoch": 0.9583503193368664, + "grad_norm": 0.9121617078781128, + "learning_rate": 0.0008802062100828918, + "loss": 3.4911, + "step": 14105 + }, + { + "epoch": 0.9586900394075282, + "grad_norm": 0.7611116170883179, + "learning_rate": 0.000880163745074059, + "loss": 3.4952, + "step": 14110 + }, + { + "epoch": 0.95902975947819, + "grad_norm": 0.8260694146156311, + "learning_rate": 0.0008801212800652263, + "loss": 3.6234, + "step": 14115 + }, + { + "epoch": 0.9593694795488518, + "grad_norm": 0.8095779418945312, + "learning_rate": 0.0008800788150563935, + "loss": 4.0281, + "step": 14120 + }, + { + "epoch": 0.9597091996195135, + "grad_norm": 0.894602358341217, + "learning_rate": 0.0008800363500475608, + "loss": 3.8609, + "step": 14125 + }, + { + "epoch": 0.9600489196901753, + "grad_norm": 1.0150920152664185, + "learning_rate": 0.0008799938850387281, + "loss": 3.6769, + "step": 14130 + }, + { + "epoch": 0.9603886397608371, + "grad_norm": 2.2357912063598633, + "learning_rate": 0.0008799514200298953, + "loss": 3.681, + "step": 14135 + }, + { + "epoch": 0.9607283598314988, + "grad_norm": 0.8003286123275757, + "learning_rate": 0.0008799089550210628, + "loss": 3.8388, + "step": 14140 + }, + { + "epoch": 0.9610680799021606, + "grad_norm": 0.6359890699386597, + "learning_rate": 0.00087986649001223, + "loss": 3.5662, + "step": 14145 + }, + { + "epoch": 0.9614077999728224, + "grad_norm": 0.6818934679031372, + "learning_rate": 0.0008798240250033972, + "loss": 3.8263, + "step": 14150 + }, + { + "epoch": 0.9617475200434842, + "grad_norm": 0.9057003855705261, + "learning_rate": 0.0008797815599945646, + "loss": 3.66, + "step": 14155 + }, + { + "epoch": 0.9620872401141459, + "grad_norm": 0.7551100254058838, + "learning_rate": 0.0008797390949857318, + "loss": 3.7164, + "step": 14160 + }, + { + "epoch": 0.9624269601848077, + "grad_norm": 0.7440225481987, + "learning_rate": 0.000879696629976899, + "loss": 3.7412, + "step": 14165 + }, + { + "epoch": 0.9627666802554695, + "grad_norm": 0.7717633843421936, + "learning_rate": 0.0008796541649680663, + "loss": 3.6196, + "step": 14170 + }, + { + "epoch": 0.9631064003261313, + "grad_norm": 0.7985604405403137, + "learning_rate": 0.0008796116999592337, + "loss": 3.6836, + "step": 14175 + }, + { + "epoch": 0.963446120396793, + "grad_norm": 0.7992721199989319, + "learning_rate": 0.0008795692349504009, + "loss": 3.6446, + "step": 14180 + }, + { + "epoch": 0.9637858404674549, + "grad_norm": 0.8389267921447754, + "learning_rate": 0.0008795267699415682, + "loss": 3.8309, + "step": 14185 + }, + { + "epoch": 0.9641255605381166, + "grad_norm": 0.8056051135063171, + "learning_rate": 0.0008794843049327355, + "loss": 3.5781, + "step": 14190 + }, + { + "epoch": 0.9644652806087783, + "grad_norm": 0.7733875513076782, + "learning_rate": 0.0008794418399239027, + "loss": 3.6502, + "step": 14195 + }, + { + "epoch": 0.9648050006794402, + "grad_norm": 2.093970775604248, + "learning_rate": 0.00087939937491507, + "loss": 3.5789, + "step": 14200 + }, + { + "epoch": 0.9651447207501019, + "grad_norm": 0.9151191711425781, + "learning_rate": 0.0008793569099062372, + "loss": 3.5379, + "step": 14205 + }, + { + "epoch": 0.9654844408207637, + "grad_norm": 0.765770673751831, + "learning_rate": 0.0008793144448974046, + "loss": 3.6453, + "step": 14210 + }, + { + "epoch": 0.9658241608914254, + "grad_norm": 1.0559406280517578, + "learning_rate": 0.0008792719798885719, + "loss": 3.6769, + "step": 14215 + }, + { + "epoch": 0.9661638809620873, + "grad_norm": 0.9609495997428894, + "learning_rate": 0.0008792295148797391, + "loss": 3.8035, + "step": 14220 + }, + { + "epoch": 0.966503601032749, + "grad_norm": 0.8505371809005737, + "learning_rate": 0.0008791870498709064, + "loss": 3.6685, + "step": 14225 + }, + { + "epoch": 0.9668433211034108, + "grad_norm": 0.8088500499725342, + "learning_rate": 0.0008791445848620737, + "loss": 3.6269, + "step": 14230 + }, + { + "epoch": 0.9671830411740726, + "grad_norm": 0.6623719334602356, + "learning_rate": 0.0008791021198532409, + "loss": 3.7227, + "step": 14235 + }, + { + "epoch": 0.9675227612447344, + "grad_norm": 0.9269905686378479, + "learning_rate": 0.0008790596548444082, + "loss": 3.4804, + "step": 14240 + }, + { + "epoch": 0.9678624813153961, + "grad_norm": 0.7461432218551636, + "learning_rate": 0.0008790171898355756, + "loss": 3.5073, + "step": 14245 + }, + { + "epoch": 0.9682022013860578, + "grad_norm": 0.686406672000885, + "learning_rate": 0.0008789747248267428, + "loss": 3.7627, + "step": 14250 + }, + { + "epoch": 0.9685419214567197, + "grad_norm": 0.7554128170013428, + "learning_rate": 0.00087893225981791, + "loss": 3.7856, + "step": 14255 + }, + { + "epoch": 0.9688816415273814, + "grad_norm": 0.7116771340370178, + "learning_rate": 0.0008788897948090774, + "loss": 3.6954, + "step": 14260 + }, + { + "epoch": 0.9692213615980432, + "grad_norm": 0.6576034426689148, + "learning_rate": 0.0008788473298002446, + "loss": 3.5698, + "step": 14265 + }, + { + "epoch": 0.969561081668705, + "grad_norm": 0.9415531754493713, + "learning_rate": 0.0008788048647914118, + "loss": 3.5838, + "step": 14270 + }, + { + "epoch": 0.9699008017393668, + "grad_norm": 0.8466349840164185, + "learning_rate": 0.0008787623997825792, + "loss": 3.768, + "step": 14275 + }, + { + "epoch": 0.9702405218100285, + "grad_norm": 0.9635273218154907, + "learning_rate": 0.0008787199347737465, + "loss": 3.5905, + "step": 14280 + }, + { + "epoch": 0.9705802418806904, + "grad_norm": 0.8305516242980957, + "learning_rate": 0.0008786774697649137, + "loss": 3.643, + "step": 14285 + }, + { + "epoch": 0.9709199619513521, + "grad_norm": 0.640354573726654, + "learning_rate": 0.0008786350047560811, + "loss": 3.6625, + "step": 14290 + }, + { + "epoch": 0.9712596820220138, + "grad_norm": 0.7893102169036865, + "learning_rate": 0.0008785925397472483, + "loss": 3.6282, + "step": 14295 + }, + { + "epoch": 0.9715994020926756, + "grad_norm": 0.8902161121368408, + "learning_rate": 0.0008785500747384155, + "loss": 3.6412, + "step": 14300 + }, + { + "epoch": 0.9719391221633374, + "grad_norm": 0.6888489723205566, + "learning_rate": 0.0008785076097295828, + "loss": 3.823, + "step": 14305 + }, + { + "epoch": 0.9722788422339992, + "grad_norm": 0.8645451068878174, + "learning_rate": 0.0008784651447207501, + "loss": 3.666, + "step": 14310 + }, + { + "epoch": 0.9726185623046609, + "grad_norm": 0.8414638042449951, + "learning_rate": 0.0008784226797119174, + "loss": 3.4548, + "step": 14315 + }, + { + "epoch": 0.9729582823753228, + "grad_norm": 0.7070507407188416, + "learning_rate": 0.0008783802147030847, + "loss": 3.6038, + "step": 14320 + }, + { + "epoch": 0.9732980024459845, + "grad_norm": 0.9197102785110474, + "learning_rate": 0.000878337749694252, + "loss": 3.1827, + "step": 14325 + }, + { + "epoch": 0.9736377225166463, + "grad_norm": 0.9254668951034546, + "learning_rate": 0.0008782952846854192, + "loss": 3.7677, + "step": 14330 + }, + { + "epoch": 0.973977442587308, + "grad_norm": 0.6447213292121887, + "learning_rate": 0.0008782528196765865, + "loss": 3.7592, + "step": 14335 + }, + { + "epoch": 0.9743171626579699, + "grad_norm": 0.5707345008850098, + "learning_rate": 0.0008782103546677538, + "loss": 3.6348, + "step": 14340 + }, + { + "epoch": 0.9746568827286316, + "grad_norm": 0.8413156867027283, + "learning_rate": 0.000878167889658921, + "loss": 3.8071, + "step": 14345 + }, + { + "epoch": 0.9749966027992933, + "grad_norm": 0.8275386095046997, + "learning_rate": 0.0008781254246500884, + "loss": 3.6867, + "step": 14350 + }, + { + "epoch": 0.9753363228699552, + "grad_norm": 0.7195150256156921, + "learning_rate": 0.0008780829596412556, + "loss": 3.3275, + "step": 14355 + }, + { + "epoch": 0.9756760429406169, + "grad_norm": 0.6811602115631104, + "learning_rate": 0.0008780404946324229, + "loss": 3.7898, + "step": 14360 + }, + { + "epoch": 0.9760157630112787, + "grad_norm": 0.6916821599006653, + "learning_rate": 0.0008779980296235902, + "loss": 3.6729, + "step": 14365 + }, + { + "epoch": 0.9763554830819405, + "grad_norm": 0.7175919413566589, + "learning_rate": 0.0008779555646147574, + "loss": 3.7737, + "step": 14370 + }, + { + "epoch": 0.9766952031526023, + "grad_norm": 0.8398088812828064, + "learning_rate": 0.0008779130996059247, + "loss": 3.4601, + "step": 14375 + }, + { + "epoch": 0.977034923223264, + "grad_norm": 0.7387117743492126, + "learning_rate": 0.000877870634597092, + "loss": 3.7371, + "step": 14380 + }, + { + "epoch": 0.9773746432939258, + "grad_norm": 0.6802313923835754, + "learning_rate": 0.0008778281695882593, + "loss": 3.368, + "step": 14385 + }, + { + "epoch": 0.9777143633645876, + "grad_norm": 0.9588213562965393, + "learning_rate": 0.0008777857045794266, + "loss": 3.4924, + "step": 14390 + }, + { + "epoch": 0.9780540834352494, + "grad_norm": 1.0174281597137451, + "learning_rate": 0.0008777432395705939, + "loss": 3.7799, + "step": 14395 + }, + { + "epoch": 0.9783938035059111, + "grad_norm": 0.8835201859474182, + "learning_rate": 0.0008777007745617611, + "loss": 3.6366, + "step": 14400 + }, + { + "epoch": 0.978733523576573, + "grad_norm": 0.7212127447128296, + "learning_rate": 0.0008776583095529283, + "loss": 3.4659, + "step": 14405 + }, + { + "epoch": 0.9790732436472347, + "grad_norm": 0.7751473188400269, + "learning_rate": 0.0008776158445440957, + "loss": 3.5266, + "step": 14410 + }, + { + "epoch": 0.9794129637178964, + "grad_norm": 0.7671706676483154, + "learning_rate": 0.0008775733795352629, + "loss": 3.7281, + "step": 14415 + }, + { + "epoch": 0.9797526837885582, + "grad_norm": 1.0055835247039795, + "learning_rate": 0.0008775309145264302, + "loss": 3.3887, + "step": 14420 + }, + { + "epoch": 0.98009240385922, + "grad_norm": 0.7222244143486023, + "learning_rate": 0.0008774884495175976, + "loss": 3.4246, + "step": 14425 + }, + { + "epoch": 0.9804321239298818, + "grad_norm": 0.8830586075782776, + "learning_rate": 0.0008774459845087648, + "loss": 3.73, + "step": 14430 + }, + { + "epoch": 0.9807718440005435, + "grad_norm": 1.598350167274475, + "learning_rate": 0.000877403519499932, + "loss": 3.2986, + "step": 14435 + }, + { + "epoch": 0.9811115640712054, + "grad_norm": 0.7917866110801697, + "learning_rate": 0.0008773610544910994, + "loss": 3.6402, + "step": 14440 + }, + { + "epoch": 0.9814512841418671, + "grad_norm": 0.7206060886383057, + "learning_rate": 0.0008773185894822666, + "loss": 3.8055, + "step": 14445 + }, + { + "epoch": 0.9817910042125289, + "grad_norm": 1.0332262516021729, + "learning_rate": 0.0008772761244734338, + "loss": 3.366, + "step": 14450 + }, + { + "epoch": 0.9821307242831907, + "grad_norm": 0.5614504218101501, + "learning_rate": 0.0008772336594646012, + "loss": 3.7558, + "step": 14455 + }, + { + "epoch": 0.9824704443538524, + "grad_norm": 0.8473602533340454, + "learning_rate": 0.0008771911944557685, + "loss": 3.5177, + "step": 14460 + }, + { + "epoch": 0.9828101644245142, + "grad_norm": 0.64482581615448, + "learning_rate": 0.0008771487294469357, + "loss": 3.6836, + "step": 14465 + }, + { + "epoch": 0.9831498844951759, + "grad_norm": 1.353542685508728, + "learning_rate": 0.000877106264438103, + "loss": 3.6789, + "step": 14470 + }, + { + "epoch": 0.9834896045658378, + "grad_norm": 1.7296388149261475, + "learning_rate": 0.0008770637994292703, + "loss": 3.427, + "step": 14475 + }, + { + "epoch": 0.9838293246364995, + "grad_norm": 0.8996850848197937, + "learning_rate": 0.0008770213344204376, + "loss": 3.5071, + "step": 14480 + }, + { + "epoch": 0.9841690447071613, + "grad_norm": 0.6983279585838318, + "learning_rate": 0.0008769788694116048, + "loss": 3.7391, + "step": 14485 + }, + { + "epoch": 0.9845087647778231, + "grad_norm": 0.7334893941879272, + "learning_rate": 0.0008769364044027722, + "loss": 3.5943, + "step": 14490 + }, + { + "epoch": 0.9848484848484849, + "grad_norm": 1.382614016532898, + "learning_rate": 0.0008768939393939395, + "loss": 3.789, + "step": 14495 + }, + { + "epoch": 0.9851882049191466, + "grad_norm": 0.8171646595001221, + "learning_rate": 0.0008768514743851067, + "loss": 3.6549, + "step": 14500 + }, + { + "epoch": 0.9855279249898083, + "grad_norm": 0.8168349266052246, + "learning_rate": 0.0008768090093762739, + "loss": 3.5184, + "step": 14505 + }, + { + "epoch": 0.9858676450604702, + "grad_norm": 0.9689443707466125, + "learning_rate": 0.0008767665443674413, + "loss": 3.5, + "step": 14510 + }, + { + "epoch": 0.9862073651311319, + "grad_norm": 0.7656502723693848, + "learning_rate": 0.0008767240793586085, + "loss": 3.7818, + "step": 14515 + }, + { + "epoch": 0.9865470852017937, + "grad_norm": 0.9008796811103821, + "learning_rate": 0.0008766816143497757, + "loss": 3.526, + "step": 14520 + }, + { + "epoch": 0.9868868052724555, + "grad_norm": 0.7302773594856262, + "learning_rate": 0.0008766391493409432, + "loss": 3.4609, + "step": 14525 + }, + { + "epoch": 0.9872265253431173, + "grad_norm": 0.8998444676399231, + "learning_rate": 0.0008765966843321104, + "loss": 3.7165, + "step": 14530 + }, + { + "epoch": 0.987566245413779, + "grad_norm": 1.2108217477798462, + "learning_rate": 0.0008765542193232776, + "loss": 3.5773, + "step": 14535 + }, + { + "epoch": 0.9879059654844409, + "grad_norm": 1.1694490909576416, + "learning_rate": 0.000876511754314445, + "loss": 3.7066, + "step": 14540 + }, + { + "epoch": 0.9882456855551026, + "grad_norm": 0.688947319984436, + "learning_rate": 0.0008764692893056122, + "loss": 3.5175, + "step": 14545 + }, + { + "epoch": 0.9885854056257644, + "grad_norm": 1.004987120628357, + "learning_rate": 0.0008764268242967794, + "loss": 3.5854, + "step": 14550 + }, + { + "epoch": 0.9889251256964261, + "grad_norm": 1.1951470375061035, + "learning_rate": 0.0008763843592879469, + "loss": 3.6282, + "step": 14555 + }, + { + "epoch": 0.989264845767088, + "grad_norm": 0.7731999754905701, + "learning_rate": 0.0008763418942791141, + "loss": 3.6051, + "step": 14560 + }, + { + "epoch": 0.9896045658377497, + "grad_norm": 0.6846461296081543, + "learning_rate": 0.0008762994292702813, + "loss": 3.754, + "step": 14565 + }, + { + "epoch": 0.9899442859084114, + "grad_norm": 0.5805031061172485, + "learning_rate": 0.0008762569642614486, + "loss": 3.4636, + "step": 14570 + }, + { + "epoch": 0.9902840059790733, + "grad_norm": 0.8409395813941956, + "learning_rate": 0.0008762144992526159, + "loss": 3.7109, + "step": 14575 + }, + { + "epoch": 0.990623726049735, + "grad_norm": 0.8794849514961243, + "learning_rate": 0.0008761720342437831, + "loss": 3.572, + "step": 14580 + }, + { + "epoch": 0.9909634461203968, + "grad_norm": 0.8855902552604675, + "learning_rate": 0.0008761295692349504, + "loss": 3.6108, + "step": 14585 + }, + { + "epoch": 0.9913031661910585, + "grad_norm": 0.6604815721511841, + "learning_rate": 0.0008760871042261178, + "loss": 3.7763, + "step": 14590 + }, + { + "epoch": 0.9916428862617204, + "grad_norm": 0.7075049877166748, + "learning_rate": 0.000876044639217285, + "loss": 3.5551, + "step": 14595 + }, + { + "epoch": 0.9919826063323821, + "grad_norm": 1.19882333278656, + "learning_rate": 0.0008760021742084523, + "loss": 3.7255, + "step": 14600 + }, + { + "epoch": 0.9923223264030439, + "grad_norm": 0.920500636100769, + "learning_rate": 0.0008759597091996195, + "loss": 3.8716, + "step": 14605 + }, + { + "epoch": 0.9926620464737057, + "grad_norm": 0.9802355170249939, + "learning_rate": 0.0008759172441907868, + "loss": 3.5886, + "step": 14610 + }, + { + "epoch": 0.9930017665443674, + "grad_norm": 0.7546324133872986, + "learning_rate": 0.0008758747791819541, + "loss": 3.7325, + "step": 14615 + }, + { + "epoch": 0.9933414866150292, + "grad_norm": 0.7007372379302979, + "learning_rate": 0.0008758323141731213, + "loss": 3.541, + "step": 14620 + }, + { + "epoch": 0.993681206685691, + "grad_norm": 0.7954425811767578, + "learning_rate": 0.0008757898491642887, + "loss": 3.6988, + "step": 14625 + }, + { + "epoch": 0.9940209267563528, + "grad_norm": 0.7672184109687805, + "learning_rate": 0.000875747384155456, + "loss": 3.8338, + "step": 14630 + }, + { + "epoch": 0.9943606468270145, + "grad_norm": 0.9527212381362915, + "learning_rate": 0.0008757049191466232, + "loss": 3.5615, + "step": 14635 + }, + { + "epoch": 0.9947003668976763, + "grad_norm": 0.8215718269348145, + "learning_rate": 0.0008756624541377904, + "loss": 3.8083, + "step": 14640 + }, + { + "epoch": 0.9950400869683381, + "grad_norm": 1.2067676782608032, + "learning_rate": 0.0008756199891289578, + "loss": 3.6209, + "step": 14645 + }, + { + "epoch": 0.9953798070389999, + "grad_norm": 0.7917919158935547, + "learning_rate": 0.000875577524120125, + "loss": 3.5736, + "step": 14650 + }, + { + "epoch": 0.9957195271096616, + "grad_norm": 0.9128100872039795, + "learning_rate": 0.0008755350591112922, + "loss": 3.6215, + "step": 14655 + }, + { + "epoch": 0.9960592471803235, + "grad_norm": 0.7717034220695496, + "learning_rate": 0.0008754925941024597, + "loss": 3.6486, + "step": 14660 + }, + { + "epoch": 0.9963989672509852, + "grad_norm": 0.7634798288345337, + "learning_rate": 0.0008754501290936269, + "loss": 3.8336, + "step": 14665 + }, + { + "epoch": 0.9967386873216469, + "grad_norm": 0.7274680733680725, + "learning_rate": 0.0008754076640847941, + "loss": 3.7638, + "step": 14670 + }, + { + "epoch": 0.9970784073923087, + "grad_norm": 0.7382634282112122, + "learning_rate": 0.0008753651990759615, + "loss": 3.7377, + "step": 14675 + }, + { + "epoch": 0.9974181274629705, + "grad_norm": 0.743107259273529, + "learning_rate": 0.0008753227340671287, + "loss": 3.7184, + "step": 14680 + }, + { + "epoch": 0.9977578475336323, + "grad_norm": 0.8931724429130554, + "learning_rate": 0.0008752802690582959, + "loss": 3.6193, + "step": 14685 + }, + { + "epoch": 0.998097567604294, + "grad_norm": 0.5748050212860107, + "learning_rate": 0.0008752378040494632, + "loss": 3.7684, + "step": 14690 + }, + { + "epoch": 0.9984372876749559, + "grad_norm": 0.8331864476203918, + "learning_rate": 0.0008751953390406306, + "loss": 3.9459, + "step": 14695 + }, + { + "epoch": 0.9987770077456176, + "grad_norm": 0.7781587839126587, + "learning_rate": 0.0008751528740317978, + "loss": 3.8221, + "step": 14700 + }, + { + "epoch": 0.9991167278162794, + "grad_norm": 1.0691982507705688, + "learning_rate": 0.0008751104090229651, + "loss": 3.6112, + "step": 14705 + }, + { + "epoch": 0.9994564478869412, + "grad_norm": 0.9747766852378845, + "learning_rate": 0.0008750679440141324, + "loss": 3.7022, + "step": 14710 + }, + { + "epoch": 0.999796167957603, + "grad_norm": 0.820070207118988, + "learning_rate": 0.0008750254790052996, + "loss": 3.8419, + "step": 14715 + }, + { + "epoch": 1.0, + "eval_bertscore": { + "f1": 0.8381534741476809, + "precision": 0.8350563258703786, + "recall": 0.8423021950917208 + }, + "eval_bleu_4": 0.0204193666824347, + "eval_exact_match": 9.690861517588914e-05, + "eval_loss": 3.501424789428711, + "eval_meteor": 0.08368346610482402, + "eval_rouge": { + "rouge1": 0.11350856223882408, + "rouge2": 0.01748823344153523, + "rougeL": 0.0994305933081177, + "rougeLsum": 0.09939504392279817 + }, + "eval_runtime": 3799.1101, + "eval_samples_per_second": 2.716, + "eval_steps_per_second": 0.34, + "step": 14718 + }, + { + "epoch": 1.0001358880282647, + "grad_norm": 0.8635953664779663, + "learning_rate": 0.0008749830139964669, + "loss": 3.625, + "step": 14720 + }, + { + "epoch": 1.0004756080989265, + "grad_norm": 0.7356550693511963, + "learning_rate": 0.0008749405489876342, + "loss": 3.6107, + "step": 14725 + }, + { + "epoch": 1.0008153281695882, + "grad_norm": 1.0746026039123535, + "learning_rate": 0.0008748980839788015, + "loss": 3.5392, + "step": 14730 + }, + { + "epoch": 1.00115504824025, + "grad_norm": 0.6924282908439636, + "learning_rate": 0.0008748556189699688, + "loss": 3.6395, + "step": 14735 + }, + { + "epoch": 1.0014947683109119, + "grad_norm": 0.9120036959648132, + "learning_rate": 0.000874813153961136, + "loss": 3.625, + "step": 14740 + }, + { + "epoch": 1.0018344883815735, + "grad_norm": 1.085540771484375, + "learning_rate": 0.0008747706889523033, + "loss": 3.6414, + "step": 14745 + }, + { + "epoch": 1.0021742084522354, + "grad_norm": 0.7921756505966187, + "learning_rate": 0.0008747282239434706, + "loss": 3.3817, + "step": 14750 + }, + { + "epoch": 1.0025139285228972, + "grad_norm": 0.9197180867195129, + "learning_rate": 0.0008746857589346378, + "loss": 3.4746, + "step": 14755 + }, + { + "epoch": 1.0028536485935589, + "grad_norm": 1.0967633724212646, + "learning_rate": 0.0008746432939258051, + "loss": 3.7374, + "step": 14760 + }, + { + "epoch": 1.0031933686642207, + "grad_norm": 0.7022275328636169, + "learning_rate": 0.0008746008289169725, + "loss": 3.6564, + "step": 14765 + }, + { + "epoch": 1.0035330887348826, + "grad_norm": 0.7609009146690369, + "learning_rate": 0.0008745583639081397, + "loss": 3.4658, + "step": 14770 + }, + { + "epoch": 1.0038728088055442, + "grad_norm": 0.8393577933311462, + "learning_rate": 0.000874515898899307, + "loss": 3.718, + "step": 14775 + }, + { + "epoch": 1.004212528876206, + "grad_norm": 1.5050324201583862, + "learning_rate": 0.0008744734338904743, + "loss": 3.6466, + "step": 14780 + }, + { + "epoch": 1.0045522489468677, + "grad_norm": 0.9276782274246216, + "learning_rate": 0.0008744309688816415, + "loss": 3.6929, + "step": 14785 + }, + { + "epoch": 1.0048919690175295, + "grad_norm": 0.9778801798820496, + "learning_rate": 0.0008743885038728087, + "loss": 3.7286, + "step": 14790 + }, + { + "epoch": 1.0052316890881914, + "grad_norm": 0.8289713263511658, + "learning_rate": 0.0008743460388639761, + "loss": 3.4104, + "step": 14795 + }, + { + "epoch": 1.005571409158853, + "grad_norm": 0.7857244610786438, + "learning_rate": 0.0008743035738551434, + "loss": 3.5886, + "step": 14800 + }, + { + "epoch": 1.0059111292295149, + "grad_norm": 2.2191922664642334, + "learning_rate": 0.0008742611088463106, + "loss": 3.5865, + "step": 14805 + }, + { + "epoch": 1.0062508493001767, + "grad_norm": 1.045069694519043, + "learning_rate": 0.000874218643837478, + "loss": 3.4953, + "step": 14810 + }, + { + "epoch": 1.0065905693708384, + "grad_norm": 0.7998431921005249, + "learning_rate": 0.0008741761788286452, + "loss": 3.4237, + "step": 14815 + }, + { + "epoch": 1.0069302894415002, + "grad_norm": 1.0983692407608032, + "learning_rate": 0.0008741337138198124, + "loss": 3.9104, + "step": 14820 + }, + { + "epoch": 1.007270009512162, + "grad_norm": 1.054249882698059, + "learning_rate": 0.0008740912488109798, + "loss": 3.6365, + "step": 14825 + }, + { + "epoch": 1.0076097295828237, + "grad_norm": 0.8268421292304993, + "learning_rate": 0.000874048783802147, + "loss": 3.6579, + "step": 14830 + }, + { + "epoch": 1.0079494496534855, + "grad_norm": 0.9890739917755127, + "learning_rate": 0.0008740063187933144, + "loss": 3.6686, + "step": 14835 + }, + { + "epoch": 1.0082891697241474, + "grad_norm": 0.7856653332710266, + "learning_rate": 0.0008739638537844817, + "loss": 3.5731, + "step": 14840 + }, + { + "epoch": 1.008628889794809, + "grad_norm": 0.8003954291343689, + "learning_rate": 0.0008739213887756489, + "loss": 3.9485, + "step": 14845 + }, + { + "epoch": 1.0089686098654709, + "grad_norm": 1.0611882209777832, + "learning_rate": 0.0008738789237668162, + "loss": 3.7569, + "step": 14850 + }, + { + "epoch": 1.0093083299361327, + "grad_norm": 0.7606302499771118, + "learning_rate": 0.0008738364587579834, + "loss": 3.8321, + "step": 14855 + }, + { + "epoch": 1.0096480500067944, + "grad_norm": 0.9010590314865112, + "learning_rate": 0.0008737939937491507, + "loss": 3.6095, + "step": 14860 + }, + { + "epoch": 1.0099877700774562, + "grad_norm": 0.9839968681335449, + "learning_rate": 0.000873751528740318, + "loss": 3.5869, + "step": 14865 + }, + { + "epoch": 1.0103274901481178, + "grad_norm": 0.884035050868988, + "learning_rate": 0.0008737090637314853, + "loss": 3.5727, + "step": 14870 + }, + { + "epoch": 1.0106672102187797, + "grad_norm": 0.9170675873756409, + "learning_rate": 0.0008736665987226526, + "loss": 3.643, + "step": 14875 + }, + { + "epoch": 1.0110069302894416, + "grad_norm": 0.9580128192901611, + "learning_rate": 0.0008736241337138199, + "loss": 3.5574, + "step": 14880 + }, + { + "epoch": 1.0113466503601032, + "grad_norm": 0.7552632093429565, + "learning_rate": 0.0008735816687049871, + "loss": 3.7948, + "step": 14885 + }, + { + "epoch": 1.011686370430765, + "grad_norm": 0.8886842131614685, + "learning_rate": 0.0008735392036961543, + "loss": 3.8444, + "step": 14890 + }, + { + "epoch": 1.012026090501427, + "grad_norm": 0.7021509408950806, + "learning_rate": 0.0008734967386873217, + "loss": 3.5689, + "step": 14895 + }, + { + "epoch": 1.0123658105720885, + "grad_norm": 0.7798046469688416, + "learning_rate": 0.0008734542736784889, + "loss": 3.336, + "step": 14900 + }, + { + "epoch": 1.0127055306427504, + "grad_norm": 0.7485750317573547, + "learning_rate": 0.0008734118086696562, + "loss": 3.8953, + "step": 14905 + }, + { + "epoch": 1.0130452507134122, + "grad_norm": 0.7091067433357239, + "learning_rate": 0.0008733693436608236, + "loss": 3.9294, + "step": 14910 + }, + { + "epoch": 1.0133849707840739, + "grad_norm": 0.7800996899604797, + "learning_rate": 0.0008733268786519908, + "loss": 3.5407, + "step": 14915 + }, + { + "epoch": 1.0137246908547357, + "grad_norm": 0.9742063283920288, + "learning_rate": 0.000873284413643158, + "loss": 3.3899, + "step": 14920 + }, + { + "epoch": 1.0140644109253976, + "grad_norm": 0.7232404947280884, + "learning_rate": 0.0008732419486343254, + "loss": 3.6032, + "step": 14925 + }, + { + "epoch": 1.0144041309960592, + "grad_norm": 0.9668446183204651, + "learning_rate": 0.0008731994836254926, + "loss": 3.4827, + "step": 14930 + }, + { + "epoch": 1.014743851066721, + "grad_norm": 0.9133487939834595, + "learning_rate": 0.0008731570186166598, + "loss": 3.4154, + "step": 14935 + }, + { + "epoch": 1.015083571137383, + "grad_norm": 0.7239922881126404, + "learning_rate": 0.0008731145536078273, + "loss": 3.3123, + "step": 14940 + }, + { + "epoch": 1.0154232912080445, + "grad_norm": 0.7745734453201294, + "learning_rate": 0.0008730720885989945, + "loss": 3.868, + "step": 14945 + }, + { + "epoch": 1.0157630112787064, + "grad_norm": 0.7706164121627808, + "learning_rate": 0.0008730296235901617, + "loss": 3.7817, + "step": 14950 + }, + { + "epoch": 1.016102731349368, + "grad_norm": 0.734239399433136, + "learning_rate": 0.000872987158581329, + "loss": 3.5991, + "step": 14955 + }, + { + "epoch": 1.0164424514200299, + "grad_norm": 0.7745044827461243, + "learning_rate": 0.0008729446935724963, + "loss": 3.6409, + "step": 14960 + }, + { + "epoch": 1.0167821714906917, + "grad_norm": 0.9256260991096497, + "learning_rate": 0.0008729022285636635, + "loss": 3.5508, + "step": 14965 + }, + { + "epoch": 1.0171218915613534, + "grad_norm": 0.8038735389709473, + "learning_rate": 0.0008728597635548308, + "loss": 3.4701, + "step": 14970 + }, + { + "epoch": 1.0174616116320152, + "grad_norm": 0.9272863268852234, + "learning_rate": 0.0008728172985459982, + "loss": 3.5885, + "step": 14975 + }, + { + "epoch": 1.017801331702677, + "grad_norm": 0.7427102327346802, + "learning_rate": 0.0008727748335371654, + "loss": 3.6044, + "step": 14980 + }, + { + "epoch": 1.0181410517733387, + "grad_norm": 0.8176618218421936, + "learning_rate": 0.0008727323685283327, + "loss": 3.8535, + "step": 14985 + }, + { + "epoch": 1.0184807718440005, + "grad_norm": 0.8393145799636841, + "learning_rate": 0.0008726899035195, + "loss": 3.6163, + "step": 14990 + }, + { + "epoch": 1.0188204919146624, + "grad_norm": 1.0643292665481567, + "learning_rate": 0.0008726474385106672, + "loss": 3.6557, + "step": 14995 + }, + { + "epoch": 1.019160211985324, + "grad_norm": 0.7359565496444702, + "learning_rate": 0.0008726049735018345, + "loss": 3.6195, + "step": 15000 + }, + { + "epoch": 1.0194999320559859, + "grad_norm": 0.7306636571884155, + "learning_rate": 0.0008725625084930017, + "loss": 3.4327, + "step": 15005 + }, + { + "epoch": 1.0198396521266477, + "grad_norm": 0.7349743247032166, + "learning_rate": 0.0008725200434841691, + "loss": 3.6665, + "step": 15010 + }, + { + "epoch": 1.0201793721973094, + "grad_norm": 0.6671044230461121, + "learning_rate": 0.0008724775784753364, + "loss": 3.647, + "step": 15015 + }, + { + "epoch": 1.0205190922679712, + "grad_norm": 1.2748668193817139, + "learning_rate": 0.0008724351134665036, + "loss": 3.3336, + "step": 15020 + }, + { + "epoch": 1.020858812338633, + "grad_norm": 0.9750664234161377, + "learning_rate": 0.0008723926484576709, + "loss": 3.6654, + "step": 15025 + }, + { + "epoch": 1.0211985324092947, + "grad_norm": 0.9526344537734985, + "learning_rate": 0.0008723501834488382, + "loss": 3.3167, + "step": 15030 + }, + { + "epoch": 1.0215382524799566, + "grad_norm": 0.7460962533950806, + "learning_rate": 0.0008723077184400054, + "loss": 3.9022, + "step": 15035 + }, + { + "epoch": 1.0218779725506182, + "grad_norm": 0.7792105674743652, + "learning_rate": 0.0008722652534311726, + "loss": 3.5521, + "step": 15040 + }, + { + "epoch": 1.02221769262128, + "grad_norm": 0.7971070408821106, + "learning_rate": 0.0008722227884223401, + "loss": 3.6695, + "step": 15045 + }, + { + "epoch": 1.022557412691942, + "grad_norm": 0.9763092398643494, + "learning_rate": 0.0008721803234135073, + "loss": 3.622, + "step": 15050 + }, + { + "epoch": 1.0228971327626035, + "grad_norm": 0.7915216684341431, + "learning_rate": 0.0008721378584046745, + "loss": 3.5439, + "step": 15055 + }, + { + "epoch": 1.0232368528332654, + "grad_norm": 0.8394269943237305, + "learning_rate": 0.0008720953933958419, + "loss": 3.582, + "step": 15060 + }, + { + "epoch": 1.0235765729039272, + "grad_norm": 0.9936326742172241, + "learning_rate": 0.0008720529283870091, + "loss": 3.5888, + "step": 15065 + }, + { + "epoch": 1.0239162929745889, + "grad_norm": 0.976253092288971, + "learning_rate": 0.0008720104633781763, + "loss": 3.5825, + "step": 15070 + }, + { + "epoch": 1.0242560130452507, + "grad_norm": 1.0398626327514648, + "learning_rate": 0.0008719679983693437, + "loss": 3.5866, + "step": 15075 + }, + { + "epoch": 1.0245957331159126, + "grad_norm": 0.7951353788375854, + "learning_rate": 0.000871925533360511, + "loss": 3.8302, + "step": 15080 + }, + { + "epoch": 1.0249354531865742, + "grad_norm": 0.7903410196304321, + "learning_rate": 0.0008718830683516782, + "loss": 3.4905, + "step": 15085 + }, + { + "epoch": 1.025275173257236, + "grad_norm": 0.8046399354934692, + "learning_rate": 0.0008718406033428455, + "loss": 3.6093, + "step": 15090 + }, + { + "epoch": 1.025614893327898, + "grad_norm": 1.0281310081481934, + "learning_rate": 0.0008717981383340128, + "loss": 3.8742, + "step": 15095 + }, + { + "epoch": 1.0259546133985595, + "grad_norm": 0.6996908783912659, + "learning_rate": 0.00087175567332518, + "loss": 3.6023, + "step": 15100 + }, + { + "epoch": 1.0262943334692214, + "grad_norm": 0.785322904586792, + "learning_rate": 0.0008717132083163473, + "loss": 3.5748, + "step": 15105 + }, + { + "epoch": 1.0266340535398832, + "grad_norm": 0.8924939632415771, + "learning_rate": 0.0008716707433075146, + "loss": 3.6664, + "step": 15110 + }, + { + "epoch": 1.0269737736105449, + "grad_norm": 1.2890547513961792, + "learning_rate": 0.0008716282782986819, + "loss": 3.5811, + "step": 15115 + }, + { + "epoch": 1.0273134936812067, + "grad_norm": 0.6925711035728455, + "learning_rate": 0.0008715858132898492, + "loss": 3.6863, + "step": 15120 + }, + { + "epoch": 1.0276532137518684, + "grad_norm": 0.8621943593025208, + "learning_rate": 0.0008715433482810165, + "loss": 3.5405, + "step": 15125 + }, + { + "epoch": 1.0279929338225302, + "grad_norm": 0.8417950868606567, + "learning_rate": 0.0008715008832721837, + "loss": 3.7407, + "step": 15130 + }, + { + "epoch": 1.028332653893192, + "grad_norm": 0.8515392541885376, + "learning_rate": 0.000871458418263351, + "loss": 3.6746, + "step": 15135 + }, + { + "epoch": 1.0286723739638537, + "grad_norm": 0.653509795665741, + "learning_rate": 0.0008714159532545182, + "loss": 3.699, + "step": 15140 + }, + { + "epoch": 1.0290120940345155, + "grad_norm": 0.724891722202301, + "learning_rate": 0.0008713734882456855, + "loss": 3.5059, + "step": 15145 + }, + { + "epoch": 1.0293518141051774, + "grad_norm": 0.8254064917564392, + "learning_rate": 0.0008713310232368529, + "loss": 3.757, + "step": 15150 + }, + { + "epoch": 1.029691534175839, + "grad_norm": 0.7997815608978271, + "learning_rate": 0.0008712885582280201, + "loss": 3.6241, + "step": 15155 + }, + { + "epoch": 1.0300312542465009, + "grad_norm": 0.6828070878982544, + "learning_rate": 0.0008712460932191874, + "loss": 3.7626, + "step": 15160 + }, + { + "epoch": 1.0303709743171627, + "grad_norm": 0.7282760143280029, + "learning_rate": 0.0008712036282103547, + "loss": 3.5814, + "step": 15165 + }, + { + "epoch": 1.0307106943878244, + "grad_norm": 0.627254843711853, + "learning_rate": 0.0008711611632015219, + "loss": 3.7564, + "step": 15170 + }, + { + "epoch": 1.0310504144584862, + "grad_norm": 0.9231497049331665, + "learning_rate": 0.0008711186981926893, + "loss": 3.4034, + "step": 15175 + }, + { + "epoch": 1.031390134529148, + "grad_norm": 0.9329749345779419, + "learning_rate": 0.0008710762331838566, + "loss": 3.6913, + "step": 15180 + }, + { + "epoch": 1.0317298545998097, + "grad_norm": 0.6896570324897766, + "learning_rate": 0.0008710337681750238, + "loss": 3.6709, + "step": 15185 + }, + { + "epoch": 1.0320695746704716, + "grad_norm": 0.8407685160636902, + "learning_rate": 0.0008709913031661911, + "loss": 3.7709, + "step": 15190 + }, + { + "epoch": 1.0324092947411334, + "grad_norm": 0.798702597618103, + "learning_rate": 0.0008709488381573584, + "loss": 3.6941, + "step": 15195 + }, + { + "epoch": 1.032749014811795, + "grad_norm": 0.8163991570472717, + "learning_rate": 0.0008709063731485256, + "loss": 3.5403, + "step": 15200 + }, + { + "epoch": 1.033088734882457, + "grad_norm": 0.790916383266449, + "learning_rate": 0.0008708639081396929, + "loss": 3.7373, + "step": 15205 + }, + { + "epoch": 1.0334284549531185, + "grad_norm": 0.9498236179351807, + "learning_rate": 0.0008708214431308602, + "loss": 3.6599, + "step": 15210 + }, + { + "epoch": 1.0337681750237804, + "grad_norm": 0.6762052178382874, + "learning_rate": 0.0008707789781220275, + "loss": 3.6779, + "step": 15215 + }, + { + "epoch": 1.0341078950944422, + "grad_norm": 0.7648082971572876, + "learning_rate": 0.0008707365131131948, + "loss": 3.8165, + "step": 15220 + }, + { + "epoch": 1.0344476151651039, + "grad_norm": 0.8258038759231567, + "learning_rate": 0.0008706940481043621, + "loss": 3.2511, + "step": 15225 + }, + { + "epoch": 1.0347873352357657, + "grad_norm": 0.8653243780136108, + "learning_rate": 0.0008706515830955293, + "loss": 3.636, + "step": 15230 + }, + { + "epoch": 1.0351270553064276, + "grad_norm": 0.7162191271781921, + "learning_rate": 0.0008706091180866966, + "loss": 3.854, + "step": 15235 + }, + { + "epoch": 1.0354667753770892, + "grad_norm": 0.857960045337677, + "learning_rate": 0.0008705666530778638, + "loss": 3.5439, + "step": 15240 + }, + { + "epoch": 1.035806495447751, + "grad_norm": 1.0894333124160767, + "learning_rate": 0.0008705241880690311, + "loss": 3.8655, + "step": 15245 + }, + { + "epoch": 1.036146215518413, + "grad_norm": 0.682677149772644, + "learning_rate": 0.0008704817230601985, + "loss": 3.7218, + "step": 15250 + }, + { + "epoch": 1.0364859355890745, + "grad_norm": 0.7342495322227478, + "learning_rate": 0.0008704392580513657, + "loss": 3.6575, + "step": 15255 + }, + { + "epoch": 1.0368256556597364, + "grad_norm": 0.8051048517227173, + "learning_rate": 0.000870396793042533, + "loss": 3.7448, + "step": 15260 + }, + { + "epoch": 1.0371653757303982, + "grad_norm": 0.6888447999954224, + "learning_rate": 0.0008703543280337003, + "loss": 3.6063, + "step": 15265 + }, + { + "epoch": 1.0375050958010599, + "grad_norm": 1.0684417486190796, + "learning_rate": 0.0008703118630248675, + "loss": 3.6057, + "step": 15270 + }, + { + "epoch": 1.0378448158717217, + "grad_norm": 0.9023168087005615, + "learning_rate": 0.0008702693980160347, + "loss": 3.3609, + "step": 15275 + }, + { + "epoch": 1.0381845359423836, + "grad_norm": 1.139523983001709, + "learning_rate": 0.0008702269330072021, + "loss": 3.6831, + "step": 15280 + }, + { + "epoch": 1.0385242560130452, + "grad_norm": 0.860252857208252, + "learning_rate": 0.0008701844679983694, + "loss": 3.7156, + "step": 15285 + }, + { + "epoch": 1.038863976083707, + "grad_norm": 0.6485652923583984, + "learning_rate": 0.0008701420029895366, + "loss": 3.9258, + "step": 15290 + }, + { + "epoch": 1.0392036961543687, + "grad_norm": 0.9340188503265381, + "learning_rate": 0.000870099537980704, + "loss": 3.6028, + "step": 15295 + }, + { + "epoch": 1.0395434162250305, + "grad_norm": 0.7440124750137329, + "learning_rate": 0.0008700570729718712, + "loss": 3.623, + "step": 15300 + }, + { + "epoch": 1.0398831362956924, + "grad_norm": 2.039768934249878, + "learning_rate": 0.0008700146079630384, + "loss": 3.6893, + "step": 15305 + }, + { + "epoch": 1.040222856366354, + "grad_norm": 1.0455355644226074, + "learning_rate": 0.0008699721429542058, + "loss": 3.3212, + "step": 15310 + }, + { + "epoch": 1.0405625764370159, + "grad_norm": 0.9251106381416321, + "learning_rate": 0.000869929677945373, + "loss": 3.7865, + "step": 15315 + }, + { + "epoch": 1.0409022965076777, + "grad_norm": 1.0914902687072754, + "learning_rate": 0.0008698872129365403, + "loss": 3.5292, + "step": 15320 + }, + { + "epoch": 1.0412420165783394, + "grad_norm": 0.9012385606765747, + "learning_rate": 0.0008698447479277077, + "loss": 3.5812, + "step": 15325 + }, + { + "epoch": 1.0415817366490012, + "grad_norm": 0.8007596135139465, + "learning_rate": 0.0008698022829188749, + "loss": 3.6419, + "step": 15330 + }, + { + "epoch": 1.041921456719663, + "grad_norm": 0.9741922616958618, + "learning_rate": 0.0008697598179100421, + "loss": 3.7753, + "step": 15335 + }, + { + "epoch": 1.0422611767903247, + "grad_norm": 0.7994995713233948, + "learning_rate": 0.0008697173529012094, + "loss": 3.9091, + "step": 15340 + }, + { + "epoch": 1.0426008968609866, + "grad_norm": 0.7818124294281006, + "learning_rate": 0.0008696748878923767, + "loss": 3.6, + "step": 15345 + }, + { + "epoch": 1.0429406169316484, + "grad_norm": 1.2978562116622925, + "learning_rate": 0.0008696324228835439, + "loss": 3.4725, + "step": 15350 + }, + { + "epoch": 1.04328033700231, + "grad_norm": 0.8045937418937683, + "learning_rate": 0.0008695899578747113, + "loss": 3.3523, + "step": 15355 + }, + { + "epoch": 1.043620057072972, + "grad_norm": 0.6882285475730896, + "learning_rate": 0.0008695474928658786, + "loss": 3.6876, + "step": 15360 + }, + { + "epoch": 1.0439597771436337, + "grad_norm": 5.161287784576416, + "learning_rate": 0.0008695050278570458, + "loss": 3.3709, + "step": 15365 + }, + { + "epoch": 1.0442994972142954, + "grad_norm": 0.952535092830658, + "learning_rate": 0.0008694625628482131, + "loss": 3.7402, + "step": 15370 + }, + { + "epoch": 1.0446392172849572, + "grad_norm": 0.8316429853439331, + "learning_rate": 0.0008694200978393803, + "loss": 3.6016, + "step": 15375 + }, + { + "epoch": 1.0449789373556189, + "grad_norm": 0.7014027237892151, + "learning_rate": 0.0008693776328305476, + "loss": 3.4609, + "step": 15380 + }, + { + "epoch": 1.0453186574262807, + "grad_norm": 0.9950093030929565, + "learning_rate": 0.0008693351678217149, + "loss": 3.4544, + "step": 15385 + }, + { + "epoch": 1.0456583774969426, + "grad_norm": 0.8154751658439636, + "learning_rate": 0.0008692927028128822, + "loss": 3.4839, + "step": 15390 + }, + { + "epoch": 1.0459980975676042, + "grad_norm": 0.7050445079803467, + "learning_rate": 0.0008692502378040495, + "loss": 3.4533, + "step": 15395 + }, + { + "epoch": 1.046337817638266, + "grad_norm": 1.300398588180542, + "learning_rate": 0.0008692077727952168, + "loss": 3.4721, + "step": 15400 + }, + { + "epoch": 1.046677537708928, + "grad_norm": 0.9842777848243713, + "learning_rate": 0.000869165307786384, + "loss": 3.5049, + "step": 15405 + }, + { + "epoch": 1.0470172577795895, + "grad_norm": 1.1697839498519897, + "learning_rate": 0.0008691228427775513, + "loss": 3.6207, + "step": 15410 + }, + { + "epoch": 1.0473569778502514, + "grad_norm": 0.7548295855522156, + "learning_rate": 0.0008690803777687186, + "loss": 3.9181, + "step": 15415 + }, + { + "epoch": 1.0476966979209132, + "grad_norm": 0.9170308709144592, + "learning_rate": 0.0008690379127598858, + "loss": 3.434, + "step": 15420 + }, + { + "epoch": 1.0480364179915749, + "grad_norm": 0.6952283978462219, + "learning_rate": 0.0008689954477510532, + "loss": 3.5821, + "step": 15425 + }, + { + "epoch": 1.0483761380622367, + "grad_norm": 0.9291009306907654, + "learning_rate": 0.0008689529827422205, + "loss": 3.417, + "step": 15430 + }, + { + "epoch": 1.0487158581328986, + "grad_norm": 0.777187705039978, + "learning_rate": 0.0008689105177333877, + "loss": 3.5713, + "step": 15435 + }, + { + "epoch": 1.0490555782035602, + "grad_norm": 0.6541638374328613, + "learning_rate": 0.0008688680527245549, + "loss": 3.5258, + "step": 15440 + }, + { + "epoch": 1.049395298274222, + "grad_norm": 0.8184323310852051, + "learning_rate": 0.0008688255877157223, + "loss": 3.4299, + "step": 15445 + }, + { + "epoch": 1.049735018344884, + "grad_norm": 0.8719306588172913, + "learning_rate": 0.0008687831227068895, + "loss": 3.761, + "step": 15450 + }, + { + "epoch": 1.0500747384155455, + "grad_norm": 0.8403478264808655, + "learning_rate": 0.0008687406576980567, + "loss": 3.6937, + "step": 15455 + }, + { + "epoch": 1.0504144584862074, + "grad_norm": 1.1352113485336304, + "learning_rate": 0.0008686981926892242, + "loss": 3.3911, + "step": 15460 + }, + { + "epoch": 1.050754178556869, + "grad_norm": 0.8563434481620789, + "learning_rate": 0.0008686557276803914, + "loss": 3.8448, + "step": 15465 + }, + { + "epoch": 1.0510938986275309, + "grad_norm": 0.8117370009422302, + "learning_rate": 0.0008686132626715586, + "loss": 3.8048, + "step": 15470 + }, + { + "epoch": 1.0514336186981927, + "grad_norm": 0.8282987475395203, + "learning_rate": 0.000868570797662726, + "loss": 3.4767, + "step": 15475 + }, + { + "epoch": 1.0517733387688544, + "grad_norm": 0.7735786437988281, + "learning_rate": 0.0008685283326538932, + "loss": 3.7295, + "step": 15480 + }, + { + "epoch": 1.0521130588395162, + "grad_norm": 0.7515607476234436, + "learning_rate": 0.0008684858676450604, + "loss": 3.7812, + "step": 15485 + }, + { + "epoch": 1.052452778910178, + "grad_norm": 0.6671996116638184, + "learning_rate": 0.0008684434026362277, + "loss": 3.7486, + "step": 15490 + }, + { + "epoch": 1.0527924989808397, + "grad_norm": 0.9180407524108887, + "learning_rate": 0.0008684009376273951, + "loss": 3.5754, + "step": 15495 + }, + { + "epoch": 1.0531322190515016, + "grad_norm": 0.7398190498352051, + "learning_rate": 0.0008683584726185623, + "loss": 3.4129, + "step": 15500 + }, + { + "epoch": 1.0534719391221634, + "grad_norm": 0.8888563513755798, + "learning_rate": 0.0008683160076097296, + "loss": 3.7205, + "step": 15505 + }, + { + "epoch": 1.053811659192825, + "grad_norm": 0.6766151785850525, + "learning_rate": 0.0008682735426008969, + "loss": 3.895, + "step": 15510 + }, + { + "epoch": 1.054151379263487, + "grad_norm": 0.8470721244812012, + "learning_rate": 0.0008682310775920642, + "loss": 3.5278, + "step": 15515 + }, + { + "epoch": 1.0544910993341488, + "grad_norm": 0.8622351884841919, + "learning_rate": 0.0008681886125832314, + "loss": 3.7996, + "step": 15520 + }, + { + "epoch": 1.0548308194048104, + "grad_norm": 0.9552620053291321, + "learning_rate": 0.0008681461475743986, + "loss": 3.6605, + "step": 15525 + }, + { + "epoch": 1.0551705394754722, + "grad_norm": 0.6540444493293762, + "learning_rate": 0.0008681036825655661, + "loss": 3.5894, + "step": 15530 + }, + { + "epoch": 1.055510259546134, + "grad_norm": 0.9241417050361633, + "learning_rate": 0.0008680612175567333, + "loss": 3.8359, + "step": 15535 + }, + { + "epoch": 1.0558499796167957, + "grad_norm": 0.8745183944702148, + "learning_rate": 0.0008680187525479005, + "loss": 3.7336, + "step": 15540 + }, + { + "epoch": 1.0561896996874576, + "grad_norm": 0.773281455039978, + "learning_rate": 0.0008679762875390679, + "loss": 3.6772, + "step": 15545 + }, + { + "epoch": 1.0565294197581192, + "grad_norm": 0.8701076507568359, + "learning_rate": 0.0008679338225302351, + "loss": 3.7153, + "step": 15550 + }, + { + "epoch": 1.056869139828781, + "grad_norm": 0.7325215339660645, + "learning_rate": 0.0008678913575214023, + "loss": 3.4492, + "step": 15555 + }, + { + "epoch": 1.057208859899443, + "grad_norm": 0.8060164451599121, + "learning_rate": 0.0008678488925125697, + "loss": 3.3348, + "step": 15560 + }, + { + "epoch": 1.0575485799701045, + "grad_norm": 0.8601083755493164, + "learning_rate": 0.000867806427503737, + "loss": 3.8161, + "step": 15565 + }, + { + "epoch": 1.0578883000407664, + "grad_norm": 1.020885944366455, + "learning_rate": 0.0008677639624949042, + "loss": 3.6374, + "step": 15570 + }, + { + "epoch": 1.0582280201114282, + "grad_norm": 1.8093150854110718, + "learning_rate": 0.0008677214974860716, + "loss": 3.4302, + "step": 15575 + }, + { + "epoch": 1.0585677401820899, + "grad_norm": 0.7937633395195007, + "learning_rate": 0.0008676790324772388, + "loss": 3.638, + "step": 15580 + }, + { + "epoch": 1.0589074602527517, + "grad_norm": 0.7711955904960632, + "learning_rate": 0.000867636567468406, + "loss": 3.7284, + "step": 15585 + }, + { + "epoch": 1.0592471803234136, + "grad_norm": 0.7265394926071167, + "learning_rate": 0.0008675941024595733, + "loss": 3.7851, + "step": 15590 + }, + { + "epoch": 1.0595869003940752, + "grad_norm": 0.8184247612953186, + "learning_rate": 0.0008675516374507406, + "loss": 3.7141, + "step": 15595 + }, + { + "epoch": 1.059926620464737, + "grad_norm": 0.9044502973556519, + "learning_rate": 0.0008675091724419079, + "loss": 3.5797, + "step": 15600 + }, + { + "epoch": 1.060266340535399, + "grad_norm": 0.6830253005027771, + "learning_rate": 0.0008674667074330752, + "loss": 3.5252, + "step": 15605 + }, + { + "epoch": 1.0606060606060606, + "grad_norm": 0.8944404125213623, + "learning_rate": 0.0008674242424242425, + "loss": 3.6994, + "step": 15610 + }, + { + "epoch": 1.0609457806767224, + "grad_norm": 0.7353104948997498, + "learning_rate": 0.0008673817774154097, + "loss": 3.7032, + "step": 15615 + }, + { + "epoch": 1.0612855007473843, + "grad_norm": 0.7380582094192505, + "learning_rate": 0.000867339312406577, + "loss": 3.9983, + "step": 15620 + }, + { + "epoch": 1.061625220818046, + "grad_norm": 0.8265982866287231, + "learning_rate": 0.0008672968473977442, + "loss": 3.597, + "step": 15625 + }, + { + "epoch": 1.0619649408887077, + "grad_norm": 0.7482130527496338, + "learning_rate": 0.0008672543823889115, + "loss": 3.7222, + "step": 15630 + }, + { + "epoch": 1.0623046609593694, + "grad_norm": 1.041133999824524, + "learning_rate": 0.0008672119173800789, + "loss": 3.7337, + "step": 15635 + }, + { + "epoch": 1.0626443810300312, + "grad_norm": 1.0080825090408325, + "learning_rate": 0.0008671694523712461, + "loss": 3.4739, + "step": 15640 + }, + { + "epoch": 1.062984101100693, + "grad_norm": 0.8892780542373657, + "learning_rate": 0.0008671269873624134, + "loss": 3.5258, + "step": 15645 + }, + { + "epoch": 1.0633238211713547, + "grad_norm": 0.9092037677764893, + "learning_rate": 0.0008670845223535807, + "loss": 3.6483, + "step": 15650 + }, + { + "epoch": 1.0636635412420166, + "grad_norm": 0.9016504287719727, + "learning_rate": 0.0008670420573447479, + "loss": 3.635, + "step": 15655 + }, + { + "epoch": 1.0640032613126784, + "grad_norm": 1.3748823404312134, + "learning_rate": 0.0008669995923359152, + "loss": 3.5499, + "step": 15660 + }, + { + "epoch": 1.06434298138334, + "grad_norm": 0.6727843880653381, + "learning_rate": 0.0008669571273270826, + "loss": 3.5387, + "step": 15665 + }, + { + "epoch": 1.064682701454002, + "grad_norm": 0.6666795015335083, + "learning_rate": 0.0008669146623182498, + "loss": 3.4707, + "step": 15670 + }, + { + "epoch": 1.0650224215246638, + "grad_norm": 1.2119091749191284, + "learning_rate": 0.000866872197309417, + "loss": 3.6234, + "step": 15675 + }, + { + "epoch": 1.0653621415953254, + "grad_norm": 0.8031150698661804, + "learning_rate": 0.0008668297323005844, + "loss": 3.8277, + "step": 15680 + }, + { + "epoch": 1.0657018616659872, + "grad_norm": 0.8668376207351685, + "learning_rate": 0.0008667872672917516, + "loss": 3.3959, + "step": 15685 + }, + { + "epoch": 1.066041581736649, + "grad_norm": 0.6989830732345581, + "learning_rate": 0.0008667448022829188, + "loss": 3.8552, + "step": 15690 + }, + { + "epoch": 1.0663813018073107, + "grad_norm": 1.4462339878082275, + "learning_rate": 0.0008667023372740862, + "loss": 3.6881, + "step": 15695 + }, + { + "epoch": 1.0667210218779726, + "grad_norm": 0.9396673440933228, + "learning_rate": 0.0008666598722652535, + "loss": 3.6808, + "step": 15700 + }, + { + "epoch": 1.0670607419486344, + "grad_norm": 1.7965008020401, + "learning_rate": 0.0008666174072564207, + "loss": 3.2489, + "step": 15705 + }, + { + "epoch": 1.067400462019296, + "grad_norm": 0.797441303730011, + "learning_rate": 0.0008665749422475881, + "loss": 3.6139, + "step": 15710 + }, + { + "epoch": 1.067740182089958, + "grad_norm": 0.7148320078849792, + "learning_rate": 0.0008665324772387553, + "loss": 3.5044, + "step": 15715 + }, + { + "epoch": 1.0680799021606195, + "grad_norm": 0.7780255675315857, + "learning_rate": 0.0008664900122299225, + "loss": 3.576, + "step": 15720 + }, + { + "epoch": 1.0684196222312814, + "grad_norm": 0.8445881009101868, + "learning_rate": 0.0008664475472210898, + "loss": 3.6259, + "step": 15725 + }, + { + "epoch": 1.0687593423019432, + "grad_norm": 0.855167031288147, + "learning_rate": 0.0008664050822122571, + "loss": 3.6185, + "step": 15730 + }, + { + "epoch": 1.0690990623726049, + "grad_norm": 0.6781607866287231, + "learning_rate": 0.0008663626172034244, + "loss": 3.8138, + "step": 15735 + }, + { + "epoch": 1.0694387824432667, + "grad_norm": 0.7494103312492371, + "learning_rate": 0.0008663201521945917, + "loss": 3.7026, + "step": 15740 + }, + { + "epoch": 1.0697785025139286, + "grad_norm": 0.9175722002983093, + "learning_rate": 0.000866277687185759, + "loss": 3.4927, + "step": 15745 + }, + { + "epoch": 1.0701182225845902, + "grad_norm": 0.6961216330528259, + "learning_rate": 0.0008662352221769262, + "loss": 3.3283, + "step": 15750 + }, + { + "epoch": 1.070457942655252, + "grad_norm": 0.8230640888214111, + "learning_rate": 0.0008661927571680935, + "loss": 3.8827, + "step": 15755 + }, + { + "epoch": 1.070797662725914, + "grad_norm": 0.9357647895812988, + "learning_rate": 0.0008661502921592608, + "loss": 3.5325, + "step": 15760 + }, + { + "epoch": 1.0711373827965756, + "grad_norm": 0.840833306312561, + "learning_rate": 0.000866107827150428, + "loss": 3.6716, + "step": 15765 + }, + { + "epoch": 1.0714771028672374, + "grad_norm": 0.8124556541442871, + "learning_rate": 0.0008660653621415954, + "loss": 3.4026, + "step": 15770 + }, + { + "epoch": 1.0718168229378993, + "grad_norm": 0.7972529530525208, + "learning_rate": 0.0008660228971327626, + "loss": 3.6875, + "step": 15775 + }, + { + "epoch": 1.072156543008561, + "grad_norm": 1.0712969303131104, + "learning_rate": 0.0008659804321239299, + "loss": 3.602, + "step": 15780 + }, + { + "epoch": 1.0724962630792227, + "grad_norm": 0.8613400459289551, + "learning_rate": 0.0008659379671150972, + "loss": 3.4832, + "step": 15785 + }, + { + "epoch": 1.0728359831498846, + "grad_norm": 0.9217832088470459, + "learning_rate": 0.0008658955021062644, + "loss": 3.6408, + "step": 15790 + }, + { + "epoch": 1.0731757032205462, + "grad_norm": 0.8431523442268372, + "learning_rate": 0.0008658530370974317, + "loss": 3.8588, + "step": 15795 + }, + { + "epoch": 1.073515423291208, + "grad_norm": 0.9289519786834717, + "learning_rate": 0.000865810572088599, + "loss": 3.6217, + "step": 15800 + }, + { + "epoch": 1.07385514336187, + "grad_norm": 0.7121431231498718, + "learning_rate": 0.0008657681070797663, + "loss": 3.655, + "step": 15805 + }, + { + "epoch": 1.0741948634325316, + "grad_norm": 0.7740139365196228, + "learning_rate": 0.0008657256420709336, + "loss": 3.524, + "step": 15810 + }, + { + "epoch": 1.0745345835031934, + "grad_norm": 0.9154728651046753, + "learning_rate": 0.0008656831770621009, + "loss": 3.6548, + "step": 15815 + }, + { + "epoch": 1.074874303573855, + "grad_norm": 0.6473360657691956, + "learning_rate": 0.0008656407120532681, + "loss": 3.6962, + "step": 15820 + }, + { + "epoch": 1.075214023644517, + "grad_norm": 0.8365759253501892, + "learning_rate": 0.0008655982470444353, + "loss": 3.6452, + "step": 15825 + }, + { + "epoch": 1.0755537437151788, + "grad_norm": 0.7894644141197205, + "learning_rate": 0.0008655557820356027, + "loss": 3.656, + "step": 15830 + }, + { + "epoch": 1.0758934637858404, + "grad_norm": 0.7554925084114075, + "learning_rate": 0.0008655133170267699, + "loss": 3.4169, + "step": 15835 + }, + { + "epoch": 1.0762331838565022, + "grad_norm": 0.682033121585846, + "learning_rate": 0.0008654708520179372, + "loss": 3.6037, + "step": 15840 + }, + { + "epoch": 1.076572903927164, + "grad_norm": 0.9009461998939514, + "learning_rate": 0.0008654283870091046, + "loss": 3.1913, + "step": 15845 + }, + { + "epoch": 1.0769126239978257, + "grad_norm": 0.791874349117279, + "learning_rate": 0.0008653859220002718, + "loss": 3.3567, + "step": 15850 + }, + { + "epoch": 1.0772523440684876, + "grad_norm": 0.724998950958252, + "learning_rate": 0.0008653434569914391, + "loss": 3.6087, + "step": 15855 + }, + { + "epoch": 1.0775920641391494, + "grad_norm": 0.8859862685203552, + "learning_rate": 0.0008653009919826064, + "loss": 3.8405, + "step": 15860 + }, + { + "epoch": 1.077931784209811, + "grad_norm": 0.8791608810424805, + "learning_rate": 0.0008652585269737736, + "loss": 3.361, + "step": 15865 + }, + { + "epoch": 1.078271504280473, + "grad_norm": 0.826759934425354, + "learning_rate": 0.0008652160619649409, + "loss": 3.712, + "step": 15870 + }, + { + "epoch": 1.0786112243511348, + "grad_norm": 0.8054351210594177, + "learning_rate": 0.0008651735969561082, + "loss": 3.5739, + "step": 15875 + }, + { + "epoch": 1.0789509444217964, + "grad_norm": 0.9845974445343018, + "learning_rate": 0.0008651311319472755, + "loss": 3.4905, + "step": 15880 + }, + { + "epoch": 1.0792906644924583, + "grad_norm": 0.775550901889801, + "learning_rate": 0.0008650886669384428, + "loss": 3.7116, + "step": 15885 + }, + { + "epoch": 1.0796303845631199, + "grad_norm": 0.6726576685905457, + "learning_rate": 0.00086504620192961, + "loss": 3.5282, + "step": 15890 + }, + { + "epoch": 1.0799701046337817, + "grad_norm": 0.8075435757637024, + "learning_rate": 0.0008650037369207773, + "loss": 3.7082, + "step": 15895 + }, + { + "epoch": 1.0803098247044436, + "grad_norm": 0.9395415186882019, + "learning_rate": 0.0008649612719119446, + "loss": 3.2468, + "step": 15900 + }, + { + "epoch": 1.0806495447751052, + "grad_norm": 0.7269135117530823, + "learning_rate": 0.0008649188069031118, + "loss": 3.8944, + "step": 15905 + }, + { + "epoch": 1.080989264845767, + "grad_norm": 0.8821495175361633, + "learning_rate": 0.0008648763418942792, + "loss": 3.8836, + "step": 15910 + }, + { + "epoch": 1.081328984916429, + "grad_norm": 0.7838165760040283, + "learning_rate": 0.0008648338768854465, + "loss": 3.6889, + "step": 15915 + }, + { + "epoch": 1.0816687049870906, + "grad_norm": 0.8660455346107483, + "learning_rate": 0.0008647914118766137, + "loss": 3.5453, + "step": 15920 + }, + { + "epoch": 1.0820084250577524, + "grad_norm": 0.7564034461975098, + "learning_rate": 0.0008647489468677809, + "loss": 3.6311, + "step": 15925 + }, + { + "epoch": 1.0823481451284143, + "grad_norm": 0.7242792844772339, + "learning_rate": 0.0008647064818589483, + "loss": 3.532, + "step": 15930 + }, + { + "epoch": 1.082687865199076, + "grad_norm": 0.7514887452125549, + "learning_rate": 0.0008646640168501155, + "loss": 3.4765, + "step": 15935 + }, + { + "epoch": 1.0830275852697377, + "grad_norm": 1.032804250717163, + "learning_rate": 0.0008646215518412827, + "loss": 3.6304, + "step": 15940 + }, + { + "epoch": 1.0833673053403996, + "grad_norm": 0.8599602580070496, + "learning_rate": 0.0008645790868324502, + "loss": 3.6612, + "step": 15945 + }, + { + "epoch": 1.0837070254110612, + "grad_norm": 0.7160202264785767, + "learning_rate": 0.0008645366218236174, + "loss": 3.5995, + "step": 15950 + }, + { + "epoch": 1.084046745481723, + "grad_norm": 0.7802814245223999, + "learning_rate": 0.0008644941568147846, + "loss": 3.669, + "step": 15955 + }, + { + "epoch": 1.084386465552385, + "grad_norm": 0.7167781591415405, + "learning_rate": 0.000864451691805952, + "loss": 3.5472, + "step": 15960 + }, + { + "epoch": 1.0847261856230466, + "grad_norm": 0.7563731670379639, + "learning_rate": 0.0008644092267971192, + "loss": 3.6312, + "step": 15965 + }, + { + "epoch": 1.0850659056937084, + "grad_norm": 0.7367556691169739, + "learning_rate": 0.0008643667617882864, + "loss": 3.5568, + "step": 15970 + }, + { + "epoch": 1.0854056257643703, + "grad_norm": 0.7233540415763855, + "learning_rate": 0.0008643242967794537, + "loss": 3.7646, + "step": 15975 + }, + { + "epoch": 1.085745345835032, + "grad_norm": 0.704897940158844, + "learning_rate": 0.0008642818317706211, + "loss": 3.7845, + "step": 15980 + }, + { + "epoch": 1.0860850659056938, + "grad_norm": 0.8350036144256592, + "learning_rate": 0.0008642393667617883, + "loss": 3.6337, + "step": 15985 + }, + { + "epoch": 1.0864247859763554, + "grad_norm": 0.6900033950805664, + "learning_rate": 0.0008641969017529556, + "loss": 3.5762, + "step": 15990 + }, + { + "epoch": 1.0867645060470172, + "grad_norm": 0.7977431416511536, + "learning_rate": 0.0008641544367441229, + "loss": 3.6783, + "step": 15995 + }, + { + "epoch": 1.087104226117679, + "grad_norm": 0.8137974143028259, + "learning_rate": 0.0008641119717352901, + "loss": 3.7409, + "step": 16000 + }, + { + "epoch": 1.0874439461883407, + "grad_norm": 0.8998544812202454, + "learning_rate": 0.0008640695067264574, + "loss": 3.6384, + "step": 16005 + }, + { + "epoch": 1.0877836662590026, + "grad_norm": 0.7980680465698242, + "learning_rate": 0.0008640270417176246, + "loss": 3.8776, + "step": 16010 + }, + { + "epoch": 1.0881233863296644, + "grad_norm": 0.6561006903648376, + "learning_rate": 0.000863984576708792, + "loss": 3.4956, + "step": 16015 + }, + { + "epoch": 1.088463106400326, + "grad_norm": 0.5969799160957336, + "learning_rate": 0.0008639421116999593, + "loss": 3.5969, + "step": 16020 + }, + { + "epoch": 1.088802826470988, + "grad_norm": 0.824756383895874, + "learning_rate": 0.0008638996466911265, + "loss": 3.6241, + "step": 16025 + }, + { + "epoch": 1.0891425465416498, + "grad_norm": 0.8661601543426514, + "learning_rate": 0.0008638571816822938, + "loss": 3.6212, + "step": 16030 + }, + { + "epoch": 1.0894822666123114, + "grad_norm": 0.8394104242324829, + "learning_rate": 0.0008638147166734611, + "loss": 3.5966, + "step": 16035 + }, + { + "epoch": 1.0898219866829733, + "grad_norm": 0.8741182684898376, + "learning_rate": 0.0008637722516646283, + "loss": 3.4223, + "step": 16040 + }, + { + "epoch": 1.090161706753635, + "grad_norm": 0.7803946733474731, + "learning_rate": 0.0008637297866557956, + "loss": 3.4778, + "step": 16045 + }, + { + "epoch": 1.0905014268242967, + "grad_norm": 0.6031648516654968, + "learning_rate": 0.000863687321646963, + "loss": 3.862, + "step": 16050 + }, + { + "epoch": 1.0908411468949586, + "grad_norm": 0.6892852783203125, + "learning_rate": 0.0008636448566381302, + "loss": 3.5967, + "step": 16055 + }, + { + "epoch": 1.0911808669656202, + "grad_norm": 0.8592513799667358, + "learning_rate": 0.0008636023916292974, + "loss": 3.3707, + "step": 16060 + }, + { + "epoch": 1.091520587036282, + "grad_norm": 0.8277498483657837, + "learning_rate": 0.0008635599266204648, + "loss": 3.6386, + "step": 16065 + }, + { + "epoch": 1.091860307106944, + "grad_norm": 2.0371274948120117, + "learning_rate": 0.000863517461611632, + "loss": 3.5039, + "step": 16070 + }, + { + "epoch": 1.0922000271776056, + "grad_norm": 0.7140267491340637, + "learning_rate": 0.0008634749966027992, + "loss": 3.6246, + "step": 16075 + }, + { + "epoch": 1.0925397472482674, + "grad_norm": 0.6711429357528687, + "learning_rate": 0.0008634325315939666, + "loss": 3.5917, + "step": 16080 + }, + { + "epoch": 1.0928794673189293, + "grad_norm": 0.8528093099594116, + "learning_rate": 0.0008633900665851339, + "loss": 3.5662, + "step": 16085 + }, + { + "epoch": 1.093219187389591, + "grad_norm": 1.503667950630188, + "learning_rate": 0.0008633476015763011, + "loss": 3.3591, + "step": 16090 + }, + { + "epoch": 1.0935589074602527, + "grad_norm": 0.8576028943061829, + "learning_rate": 0.0008633051365674685, + "loss": 3.7892, + "step": 16095 + }, + { + "epoch": 1.0938986275309146, + "grad_norm": 1.0386401414871216, + "learning_rate": 0.0008632626715586357, + "loss": 3.7382, + "step": 16100 + }, + { + "epoch": 1.0942383476015762, + "grad_norm": 0.777122437953949, + "learning_rate": 0.0008632202065498029, + "loss": 3.5138, + "step": 16105 + }, + { + "epoch": 1.094578067672238, + "grad_norm": 0.8399432301521301, + "learning_rate": 0.0008631777415409702, + "loss": 3.5598, + "step": 16110 + }, + { + "epoch": 1.0949177877429, + "grad_norm": 0.9027761816978455, + "learning_rate": 0.0008631352765321375, + "loss": 3.6889, + "step": 16115 + }, + { + "epoch": 1.0952575078135616, + "grad_norm": 0.73099684715271, + "learning_rate": 0.0008630928115233048, + "loss": 3.6822, + "step": 16120 + }, + { + "epoch": 1.0955972278842234, + "grad_norm": 0.887734055519104, + "learning_rate": 0.0008630503465144721, + "loss": 3.4904, + "step": 16125 + }, + { + "epoch": 1.0959369479548853, + "grad_norm": 0.7868804335594177, + "learning_rate": 0.0008630078815056394, + "loss": 3.8223, + "step": 16130 + }, + { + "epoch": 1.096276668025547, + "grad_norm": 1.0577092170715332, + "learning_rate": 0.0008629654164968066, + "loss": 3.4639, + "step": 16135 + }, + { + "epoch": 1.0966163880962088, + "grad_norm": 0.6253250241279602, + "learning_rate": 0.0008629229514879739, + "loss": 3.75, + "step": 16140 + }, + { + "epoch": 1.0969561081668706, + "grad_norm": 0.8179384469985962, + "learning_rate": 0.0008628804864791412, + "loss": 3.5636, + "step": 16145 + }, + { + "epoch": 1.0972958282375322, + "grad_norm": 0.7240017056465149, + "learning_rate": 0.0008628380214703084, + "loss": 3.6615, + "step": 16150 + }, + { + "epoch": 1.097635548308194, + "grad_norm": 0.784687340259552, + "learning_rate": 0.0008627955564614758, + "loss": 3.7314, + "step": 16155 + }, + { + "epoch": 1.0979752683788557, + "grad_norm": 0.6818301677703857, + "learning_rate": 0.000862753091452643, + "loss": 3.5592, + "step": 16160 + }, + { + "epoch": 1.0983149884495176, + "grad_norm": 0.9793298244476318, + "learning_rate": 0.0008627106264438103, + "loss": 3.5025, + "step": 16165 + }, + { + "epoch": 1.0986547085201794, + "grad_norm": 1.1576931476593018, + "learning_rate": 0.0008626681614349776, + "loss": 3.6291, + "step": 16170 + }, + { + "epoch": 1.098994428590841, + "grad_norm": 1.016464114189148, + "learning_rate": 0.0008626256964261448, + "loss": 3.724, + "step": 16175 + }, + { + "epoch": 1.099334148661503, + "grad_norm": 0.8881516456604004, + "learning_rate": 0.0008625832314173121, + "loss": 3.6412, + "step": 16180 + }, + { + "epoch": 1.0996738687321648, + "grad_norm": 0.6917318105697632, + "learning_rate": 0.0008625407664084794, + "loss": 3.7253, + "step": 16185 + }, + { + "epoch": 1.1000135888028264, + "grad_norm": 0.8780159950256348, + "learning_rate": 0.0008624983013996467, + "loss": 3.4274, + "step": 16190 + }, + { + "epoch": 1.1003533088734883, + "grad_norm": 0.7432769536972046, + "learning_rate": 0.0008624558363908141, + "loss": 3.6838, + "step": 16195 + }, + { + "epoch": 1.10069302894415, + "grad_norm": 0.802573025226593, + "learning_rate": 0.0008624133713819813, + "loss": 3.6626, + "step": 16200 + }, + { + "epoch": 1.1010327490148117, + "grad_norm": 0.7606637477874756, + "learning_rate": 0.0008623709063731485, + "loss": 3.6076, + "step": 16205 + }, + { + "epoch": 1.1013724690854736, + "grad_norm": 0.7026821374893188, + "learning_rate": 0.0008623284413643159, + "loss": 3.4194, + "step": 16210 + }, + { + "epoch": 1.1017121891561354, + "grad_norm": 0.6719335913658142, + "learning_rate": 0.0008622859763554831, + "loss": 3.4139, + "step": 16215 + }, + { + "epoch": 1.102051909226797, + "grad_norm": 0.8041277527809143, + "learning_rate": 0.0008622435113466503, + "loss": 3.7013, + "step": 16220 + }, + { + "epoch": 1.102391629297459, + "grad_norm": 0.8182024359703064, + "learning_rate": 0.0008622010463378177, + "loss": 3.8294, + "step": 16225 + }, + { + "epoch": 1.1027313493681206, + "grad_norm": 0.7067434191703796, + "learning_rate": 0.000862158581328985, + "loss": 3.6723, + "step": 16230 + }, + { + "epoch": 1.1030710694387824, + "grad_norm": 0.9231346845626831, + "learning_rate": 0.0008621161163201522, + "loss": 3.6599, + "step": 16235 + }, + { + "epoch": 1.1034107895094443, + "grad_norm": 0.6359360218048096, + "learning_rate": 0.0008620736513113195, + "loss": 3.7619, + "step": 16240 + }, + { + "epoch": 1.103750509580106, + "grad_norm": 0.7554166316986084, + "learning_rate": 0.0008620311863024868, + "loss": 3.7817, + "step": 16245 + }, + { + "epoch": 1.1040902296507678, + "grad_norm": 0.7820613980293274, + "learning_rate": 0.000861988721293654, + "loss": 3.7275, + "step": 16250 + }, + { + "epoch": 1.1044299497214296, + "grad_norm": 0.8319656252861023, + "learning_rate": 0.0008619462562848214, + "loss": 3.65, + "step": 16255 + }, + { + "epoch": 1.1047696697920912, + "grad_norm": 21.17509651184082, + "learning_rate": 0.0008619037912759887, + "loss": 3.5016, + "step": 16260 + }, + { + "epoch": 1.105109389862753, + "grad_norm": 0.8816961646080017, + "learning_rate": 0.0008618613262671559, + "loss": 3.4664, + "step": 16265 + }, + { + "epoch": 1.105449109933415, + "grad_norm": 0.9653134942054749, + "learning_rate": 0.0008618188612583232, + "loss": 3.4988, + "step": 16270 + }, + { + "epoch": 1.1057888300040766, + "grad_norm": 0.7862028479576111, + "learning_rate": 0.0008617763962494904, + "loss": 3.6665, + "step": 16275 + }, + { + "epoch": 1.1061285500747384, + "grad_norm": 0.9713080525398254, + "learning_rate": 0.0008617339312406577, + "loss": 3.5714, + "step": 16280 + }, + { + "epoch": 1.1064682701454003, + "grad_norm": 0.8657552599906921, + "learning_rate": 0.000861691466231825, + "loss": 3.5994, + "step": 16285 + }, + { + "epoch": 1.106807990216062, + "grad_norm": 0.8849959969520569, + "learning_rate": 0.0008616490012229923, + "loss": 3.5815, + "step": 16290 + }, + { + "epoch": 1.1071477102867238, + "grad_norm": 0.8014518618583679, + "learning_rate": 0.0008616065362141596, + "loss": 3.2865, + "step": 16295 + }, + { + "epoch": 1.1074874303573856, + "grad_norm": 0.9684229493141174, + "learning_rate": 0.0008615640712053269, + "loss": 3.374, + "step": 16300 + }, + { + "epoch": 1.1078271504280472, + "grad_norm": 1.1310782432556152, + "learning_rate": 0.0008615216061964941, + "loss": 3.5702, + "step": 16305 + }, + { + "epoch": 1.108166870498709, + "grad_norm": 0.7408681511878967, + "learning_rate": 0.0008614791411876613, + "loss": 3.8609, + "step": 16310 + }, + { + "epoch": 1.108506590569371, + "grad_norm": 0.799790620803833, + "learning_rate": 0.0008614366761788287, + "loss": 3.4911, + "step": 16315 + }, + { + "epoch": 1.1088463106400326, + "grad_norm": 0.7795165181159973, + "learning_rate": 0.0008613942111699959, + "loss": 3.7212, + "step": 16320 + }, + { + "epoch": 1.1091860307106944, + "grad_norm": 0.6737672090530396, + "learning_rate": 0.0008613517461611632, + "loss": 3.7207, + "step": 16325 + }, + { + "epoch": 1.109525750781356, + "grad_norm": 0.9147469997406006, + "learning_rate": 0.0008613092811523306, + "loss": 3.572, + "step": 16330 + }, + { + "epoch": 1.109865470852018, + "grad_norm": 0.5638120770454407, + "learning_rate": 0.0008612668161434978, + "loss": 3.7112, + "step": 16335 + }, + { + "epoch": 1.1102051909226798, + "grad_norm": 0.8296770453453064, + "learning_rate": 0.000861224351134665, + "loss": 3.46, + "step": 16340 + }, + { + "epoch": 1.1105449109933414, + "grad_norm": 0.8826504349708557, + "learning_rate": 0.0008611818861258324, + "loss": 3.7987, + "step": 16345 + }, + { + "epoch": 1.1108846310640033, + "grad_norm": 0.7032650113105774, + "learning_rate": 0.0008611394211169996, + "loss": 3.6871, + "step": 16350 + }, + { + "epoch": 1.1112243511346651, + "grad_norm": 1.0232462882995605, + "learning_rate": 0.0008610969561081668, + "loss": 3.6665, + "step": 16355 + }, + { + "epoch": 1.1115640712053267, + "grad_norm": 0.7778091430664062, + "learning_rate": 0.0008610544910993343, + "loss": 3.4323, + "step": 16360 + }, + { + "epoch": 1.1119037912759886, + "grad_norm": 0.7016257643699646, + "learning_rate": 0.0008610120260905015, + "loss": 3.6359, + "step": 16365 + }, + { + "epoch": 1.1122435113466504, + "grad_norm": 0.8150584697723389, + "learning_rate": 0.0008609695610816687, + "loss": 3.5146, + "step": 16370 + }, + { + "epoch": 1.112583231417312, + "grad_norm": 0.7057853937149048, + "learning_rate": 0.000860927096072836, + "loss": 3.3827, + "step": 16375 + }, + { + "epoch": 1.112922951487974, + "grad_norm": 0.9030365943908691, + "learning_rate": 0.0008608846310640033, + "loss": 3.7423, + "step": 16380 + }, + { + "epoch": 1.1132626715586358, + "grad_norm": 0.8879201412200928, + "learning_rate": 0.0008608421660551705, + "loss": 3.6915, + "step": 16385 + }, + { + "epoch": 1.1136023916292974, + "grad_norm": 0.768269956111908, + "learning_rate": 0.0008607997010463378, + "loss": 3.8513, + "step": 16390 + }, + { + "epoch": 1.1139421116999593, + "grad_norm": 0.7520266771316528, + "learning_rate": 0.0008607572360375052, + "loss": 3.8602, + "step": 16395 + }, + { + "epoch": 1.114281831770621, + "grad_norm": 0.7475628852844238, + "learning_rate": 0.0008607147710286724, + "loss": 3.4871, + "step": 16400 + }, + { + "epoch": 1.1146215518412828, + "grad_norm": 1.7916303873062134, + "learning_rate": 0.0008606723060198397, + "loss": 3.5601, + "step": 16405 + }, + { + "epoch": 1.1149612719119446, + "grad_norm": 0.7615216374397278, + "learning_rate": 0.000860629841011007, + "loss": 3.7309, + "step": 16410 + }, + { + "epoch": 1.1153009919826062, + "grad_norm": 0.8642655611038208, + "learning_rate": 0.0008605873760021742, + "loss": 3.6074, + "step": 16415 + }, + { + "epoch": 1.115640712053268, + "grad_norm": 0.833690345287323, + "learning_rate": 0.0008605449109933415, + "loss": 3.7721, + "step": 16420 + }, + { + "epoch": 1.11598043212393, + "grad_norm": 0.6924107670783997, + "learning_rate": 0.0008605024459845087, + "loss": 3.4532, + "step": 16425 + }, + { + "epoch": 1.1163201521945916, + "grad_norm": 0.9325452446937561, + "learning_rate": 0.0008604599809756761, + "loss": 3.7371, + "step": 16430 + }, + { + "epoch": 1.1166598722652534, + "grad_norm": 0.9256713390350342, + "learning_rate": 0.0008604175159668434, + "loss": 3.524, + "step": 16435 + }, + { + "epoch": 1.1169995923359153, + "grad_norm": 0.8062615990638733, + "learning_rate": 0.0008603750509580106, + "loss": 3.6315, + "step": 16440 + }, + { + "epoch": 1.117339312406577, + "grad_norm": 0.9539016485214233, + "learning_rate": 0.0008603325859491779, + "loss": 3.5954, + "step": 16445 + }, + { + "epoch": 1.1176790324772388, + "grad_norm": 0.9835814237594604, + "learning_rate": 0.0008602901209403452, + "loss": 3.9943, + "step": 16450 + }, + { + "epoch": 1.1180187525479006, + "grad_norm": 0.7676118016242981, + "learning_rate": 0.0008602476559315124, + "loss": 3.6656, + "step": 16455 + }, + { + "epoch": 1.1183584726185622, + "grad_norm": 1.1830953359603882, + "learning_rate": 0.0008602051909226796, + "loss": 3.6766, + "step": 16460 + }, + { + "epoch": 1.118698192689224, + "grad_norm": 0.7889232039451599, + "learning_rate": 0.0008601627259138471, + "loss": 3.8535, + "step": 16465 + }, + { + "epoch": 1.119037912759886, + "grad_norm": 0.9148766398429871, + "learning_rate": 0.0008601202609050143, + "loss": 3.3512, + "step": 16470 + }, + { + "epoch": 1.1193776328305476, + "grad_norm": 1.1237456798553467, + "learning_rate": 0.0008600777958961815, + "loss": 3.6986, + "step": 16475 + }, + { + "epoch": 1.1197173529012094, + "grad_norm": 0.7358617186546326, + "learning_rate": 0.0008600353308873489, + "loss": 3.6707, + "step": 16480 + }, + { + "epoch": 1.1200570729718713, + "grad_norm": 1.004721760749817, + "learning_rate": 0.0008599928658785161, + "loss": 3.5836, + "step": 16485 + }, + { + "epoch": 1.120396793042533, + "grad_norm": 0.6926807761192322, + "learning_rate": 0.0008599504008696833, + "loss": 3.5393, + "step": 16490 + }, + { + "epoch": 1.1207365131131948, + "grad_norm": 0.710195779800415, + "learning_rate": 0.0008599079358608507, + "loss": 3.6593, + "step": 16495 + }, + { + "epoch": 1.1210762331838564, + "grad_norm": 0.798686146736145, + "learning_rate": 0.000859865470852018, + "loss": 3.6687, + "step": 16500 + }, + { + "epoch": 1.1214159532545183, + "grad_norm": 0.7609354257583618, + "learning_rate": 0.0008598230058431852, + "loss": 3.636, + "step": 16505 + }, + { + "epoch": 1.1217556733251801, + "grad_norm": 0.7079862356185913, + "learning_rate": 0.0008597805408343525, + "loss": 3.5538, + "step": 16510 + }, + { + "epoch": 1.1220953933958417, + "grad_norm": 0.7339089512825012, + "learning_rate": 0.0008597380758255198, + "loss": 3.6521, + "step": 16515 + }, + { + "epoch": 1.1224351134665036, + "grad_norm": 0.7709986567497253, + "learning_rate": 0.000859695610816687, + "loss": 3.7699, + "step": 16520 + }, + { + "epoch": 1.1227748335371655, + "grad_norm": 0.8063376545906067, + "learning_rate": 0.0008596531458078543, + "loss": 3.5019, + "step": 16525 + }, + { + "epoch": 1.123114553607827, + "grad_norm": 1.043861746788025, + "learning_rate": 0.0008596106807990216, + "loss": 3.5688, + "step": 16530 + }, + { + "epoch": 1.123454273678489, + "grad_norm": 0.8596996068954468, + "learning_rate": 0.000859568215790189, + "loss": 3.5837, + "step": 16535 + }, + { + "epoch": 1.1237939937491508, + "grad_norm": 0.78280109167099, + "learning_rate": 0.0008595257507813562, + "loss": 3.8415, + "step": 16540 + }, + { + "epoch": 1.1241337138198124, + "grad_norm": 0.8170922994613647, + "learning_rate": 0.0008594832857725235, + "loss": 3.6423, + "step": 16545 + }, + { + "epoch": 1.1244734338904743, + "grad_norm": 0.8567583560943604, + "learning_rate": 0.0008594408207636908, + "loss": 3.5889, + "step": 16550 + }, + { + "epoch": 1.1248131539611361, + "grad_norm": 0.9866414070129395, + "learning_rate": 0.000859398355754858, + "loss": 3.6437, + "step": 16555 + }, + { + "epoch": 1.1251528740317978, + "grad_norm": 0.8666515350341797, + "learning_rate": 0.0008593558907460252, + "loss": 3.6796, + "step": 16560 + }, + { + "epoch": 1.1254925941024596, + "grad_norm": 0.9461397528648376, + "learning_rate": 0.0008593134257371926, + "loss": 3.7611, + "step": 16565 + }, + { + "epoch": 1.1258323141731212, + "grad_norm": 0.8283349871635437, + "learning_rate": 0.0008592709607283599, + "loss": 3.56, + "step": 16570 + }, + { + "epoch": 1.126172034243783, + "grad_norm": 0.8518447875976562, + "learning_rate": 0.0008592284957195271, + "loss": 3.7552, + "step": 16575 + }, + { + "epoch": 1.126511754314445, + "grad_norm": 0.8528580665588379, + "learning_rate": 0.0008591860307106945, + "loss": 3.7868, + "step": 16580 + }, + { + "epoch": 1.1268514743851066, + "grad_norm": 0.6408374905586243, + "learning_rate": 0.0008591435657018617, + "loss": 3.7224, + "step": 16585 + }, + { + "epoch": 1.1271911944557684, + "grad_norm": 0.853872537612915, + "learning_rate": 0.0008591011006930289, + "loss": 3.5575, + "step": 16590 + }, + { + "epoch": 1.1275309145264303, + "grad_norm": 0.8306583166122437, + "learning_rate": 0.0008590586356841963, + "loss": 3.9019, + "step": 16595 + }, + { + "epoch": 1.127870634597092, + "grad_norm": 1.2525982856750488, + "learning_rate": 0.0008590161706753635, + "loss": 3.5223, + "step": 16600 + }, + { + "epoch": 1.1282103546677538, + "grad_norm": 0.8633636236190796, + "learning_rate": 0.0008589737056665308, + "loss": 3.4292, + "step": 16605 + }, + { + "epoch": 1.1285500747384156, + "grad_norm": 0.8228412866592407, + "learning_rate": 0.0008589312406576981, + "loss": 3.6021, + "step": 16610 + }, + { + "epoch": 1.1288897948090773, + "grad_norm": 0.8219278454780579, + "learning_rate": 0.0008588887756488654, + "loss": 3.6794, + "step": 16615 + }, + { + "epoch": 1.129229514879739, + "grad_norm": 1.0117695331573486, + "learning_rate": 0.0008588463106400326, + "loss": 3.4624, + "step": 16620 + }, + { + "epoch": 1.129569234950401, + "grad_norm": 1.118656039237976, + "learning_rate": 0.0008588038456311999, + "loss": 3.6851, + "step": 16625 + }, + { + "epoch": 1.1299089550210626, + "grad_norm": 0.8691116571426392, + "learning_rate": 0.0008587613806223672, + "loss": 3.5823, + "step": 16630 + }, + { + "epoch": 1.1302486750917244, + "grad_norm": 0.8138573169708252, + "learning_rate": 0.0008587189156135344, + "loss": 3.7302, + "step": 16635 + }, + { + "epoch": 1.1305883951623863, + "grad_norm": 0.8142614364624023, + "learning_rate": 0.0008586764506047018, + "loss": 3.6892, + "step": 16640 + }, + { + "epoch": 1.130928115233048, + "grad_norm": 0.7584553956985474, + "learning_rate": 0.0008586339855958691, + "loss": 3.531, + "step": 16645 + }, + { + "epoch": 1.1312678353037098, + "grad_norm": 0.9203415513038635, + "learning_rate": 0.0008585915205870363, + "loss": 3.8447, + "step": 16650 + }, + { + "epoch": 1.1316075553743716, + "grad_norm": 0.770386815071106, + "learning_rate": 0.0008585490555782036, + "loss": 3.3881, + "step": 16655 + }, + { + "epoch": 1.1319472754450333, + "grad_norm": 0.7741190195083618, + "learning_rate": 0.0008585065905693708, + "loss": 3.6307, + "step": 16660 + }, + { + "epoch": 1.1322869955156951, + "grad_norm": 0.6954805254936218, + "learning_rate": 0.0008584641255605381, + "loss": 3.593, + "step": 16665 + }, + { + "epoch": 1.1326267155863567, + "grad_norm": 0.8596692681312561, + "learning_rate": 0.0008584216605517054, + "loss": 3.6929, + "step": 16670 + }, + { + "epoch": 1.1329664356570186, + "grad_norm": 0.819408118724823, + "learning_rate": 0.0008583791955428727, + "loss": 3.64, + "step": 16675 + }, + { + "epoch": 1.1333061557276805, + "grad_norm": 1.0247679948806763, + "learning_rate": 0.00085833673053404, + "loss": 3.6088, + "step": 16680 + }, + { + "epoch": 1.133645875798342, + "grad_norm": 0.7671202421188354, + "learning_rate": 0.0008582942655252073, + "loss": 3.6303, + "step": 16685 + }, + { + "epoch": 1.133985595869004, + "grad_norm": 0.8971432447433472, + "learning_rate": 0.0008582518005163745, + "loss": 3.7277, + "step": 16690 + }, + { + "epoch": 1.1343253159396658, + "grad_norm": 0.7380162477493286, + "learning_rate": 0.0008582093355075417, + "loss": 3.5407, + "step": 16695 + }, + { + "epoch": 1.1346650360103274, + "grad_norm": 1.3466129302978516, + "learning_rate": 0.0008581668704987091, + "loss": 3.3349, + "step": 16700 + }, + { + "epoch": 1.1350047560809893, + "grad_norm": 0.7564104199409485, + "learning_rate": 0.0008581244054898763, + "loss": 3.6664, + "step": 16705 + }, + { + "epoch": 1.1353444761516511, + "grad_norm": 0.9620634913444519, + "learning_rate": 0.0008580819404810436, + "loss": 3.7406, + "step": 16710 + }, + { + "epoch": 1.1356841962223128, + "grad_norm": 0.721673846244812, + "learning_rate": 0.000858039475472211, + "loss": 3.5365, + "step": 16715 + }, + { + "epoch": 1.1360239162929746, + "grad_norm": 0.7465148568153381, + "learning_rate": 0.0008579970104633782, + "loss": 3.8301, + "step": 16720 + }, + { + "epoch": 1.1363636363636362, + "grad_norm": 0.8791694641113281, + "learning_rate": 0.0008579545454545454, + "loss": 3.6285, + "step": 16725 + }, + { + "epoch": 1.136703356434298, + "grad_norm": 0.7738268971443176, + "learning_rate": 0.0008579120804457128, + "loss": 3.7493, + "step": 16730 + }, + { + "epoch": 1.13704307650496, + "grad_norm": 0.9750152826309204, + "learning_rate": 0.00085786961543688, + "loss": 3.7962, + "step": 16735 + }, + { + "epoch": 1.1373827965756216, + "grad_norm": 2.011751890182495, + "learning_rate": 0.0008578271504280472, + "loss": 3.5232, + "step": 16740 + }, + { + "epoch": 1.1377225166462834, + "grad_norm": 0.88902747631073, + "learning_rate": 0.0008577846854192147, + "loss": 3.5958, + "step": 16745 + }, + { + "epoch": 1.1380622367169453, + "grad_norm": 0.9654760956764221, + "learning_rate": 0.0008577422204103819, + "loss": 3.6575, + "step": 16750 + }, + { + "epoch": 1.138401956787607, + "grad_norm": 1.0040602684020996, + "learning_rate": 0.0008576997554015491, + "loss": 3.8901, + "step": 16755 + }, + { + "epoch": 1.1387416768582688, + "grad_norm": 0.9915438890457153, + "learning_rate": 0.0008576572903927164, + "loss": 3.5477, + "step": 16760 + }, + { + "epoch": 1.1390813969289306, + "grad_norm": 0.7121073007583618, + "learning_rate": 0.0008576148253838837, + "loss": 3.9085, + "step": 16765 + }, + { + "epoch": 1.1394211169995923, + "grad_norm": 1.034774661064148, + "learning_rate": 0.0008575723603750509, + "loss": 3.6523, + "step": 16770 + }, + { + "epoch": 1.139760837070254, + "grad_norm": 0.7872732877731323, + "learning_rate": 0.0008575298953662182, + "loss": 3.5303, + "step": 16775 + }, + { + "epoch": 1.140100557140916, + "grad_norm": 0.766667366027832, + "learning_rate": 0.0008574874303573856, + "loss": 3.5771, + "step": 16780 + }, + { + "epoch": 1.1404402772115776, + "grad_norm": 0.8747329115867615, + "learning_rate": 0.0008574449653485528, + "loss": 3.6362, + "step": 16785 + }, + { + "epoch": 1.1407799972822394, + "grad_norm": 0.7882775068283081, + "learning_rate": 0.0008574025003397201, + "loss": 3.7866, + "step": 16790 + }, + { + "epoch": 1.1411197173529013, + "grad_norm": 1.009997844696045, + "learning_rate": 0.0008573600353308873, + "loss": 3.4335, + "step": 16795 + }, + { + "epoch": 1.141459437423563, + "grad_norm": 0.8503082990646362, + "learning_rate": 0.0008573175703220546, + "loss": 3.7175, + "step": 16800 + }, + { + "epoch": 1.1417991574942248, + "grad_norm": 0.9235851764678955, + "learning_rate": 0.0008572751053132219, + "loss": 3.6923, + "step": 16805 + }, + { + "epoch": 1.1421388775648866, + "grad_norm": 0.8511438369750977, + "learning_rate": 0.0008572326403043891, + "loss": 3.713, + "step": 16810 + }, + { + "epoch": 1.1424785976355483, + "grad_norm": 1.3942556381225586, + "learning_rate": 0.0008571901752955565, + "loss": 3.6219, + "step": 16815 + }, + { + "epoch": 1.1428183177062101, + "grad_norm": 0.7534943222999573, + "learning_rate": 0.0008571477102867238, + "loss": 3.4717, + "step": 16820 + }, + { + "epoch": 1.143158037776872, + "grad_norm": 0.8964465260505676, + "learning_rate": 0.000857105245277891, + "loss": 3.5183, + "step": 16825 + }, + { + "epoch": 1.1434977578475336, + "grad_norm": 0.7877064943313599, + "learning_rate": 0.0008570627802690583, + "loss": 3.5627, + "step": 16830 + }, + { + "epoch": 1.1438374779181955, + "grad_norm": 1.1212519407272339, + "learning_rate": 0.0008570203152602256, + "loss": 3.7749, + "step": 16835 + }, + { + "epoch": 1.144177197988857, + "grad_norm": 0.7497186064720154, + "learning_rate": 0.0008569778502513928, + "loss": 3.7101, + "step": 16840 + }, + { + "epoch": 1.144516918059519, + "grad_norm": 2.6848554611206055, + "learning_rate": 0.00085693538524256, + "loss": 3.6506, + "step": 16845 + }, + { + "epoch": 1.1448566381301808, + "grad_norm": 0.81125408411026, + "learning_rate": 0.0008568929202337275, + "loss": 3.7755, + "step": 16850 + }, + { + "epoch": 1.1451963582008424, + "grad_norm": 0.9101123213768005, + "learning_rate": 0.0008568504552248947, + "loss": 3.5634, + "step": 16855 + }, + { + "epoch": 1.1455360782715043, + "grad_norm": 0.6549205780029297, + "learning_rate": 0.0008568079902160619, + "loss": 3.7316, + "step": 16860 + }, + { + "epoch": 1.1458757983421661, + "grad_norm": 0.83475261926651, + "learning_rate": 0.0008567655252072293, + "loss": 3.4136, + "step": 16865 + }, + { + "epoch": 1.1462155184128278, + "grad_norm": 0.8266447186470032, + "learning_rate": 0.0008567230601983965, + "loss": 3.5383, + "step": 16870 + }, + { + "epoch": 1.1465552384834896, + "grad_norm": 1.1312578916549683, + "learning_rate": 0.0008566805951895638, + "loss": 3.8844, + "step": 16875 + }, + { + "epoch": 1.1468949585541515, + "grad_norm": 0.8602306842803955, + "learning_rate": 0.0008566381301807312, + "loss": 3.6204, + "step": 16880 + }, + { + "epoch": 1.147234678624813, + "grad_norm": 0.8610104918479919, + "learning_rate": 0.0008565956651718984, + "loss": 3.657, + "step": 16885 + }, + { + "epoch": 1.147574398695475, + "grad_norm": 0.8559669256210327, + "learning_rate": 0.0008565532001630657, + "loss": 3.6957, + "step": 16890 + }, + { + "epoch": 1.1479141187661366, + "grad_norm": 0.8816953301429749, + "learning_rate": 0.000856510735154233, + "loss": 3.6052, + "step": 16895 + }, + { + "epoch": 1.1482538388367984, + "grad_norm": 0.7338087558746338, + "learning_rate": 0.0008564682701454002, + "loss": 3.5228, + "step": 16900 + }, + { + "epoch": 1.1485935589074603, + "grad_norm": 0.646308422088623, + "learning_rate": 0.0008564258051365675, + "loss": 3.6746, + "step": 16905 + }, + { + "epoch": 1.148933278978122, + "grad_norm": 0.8172942996025085, + "learning_rate": 0.0008563833401277347, + "loss": 3.5247, + "step": 16910 + }, + { + "epoch": 1.1492729990487838, + "grad_norm": 1.034718632698059, + "learning_rate": 0.0008563408751189021, + "loss": 3.5556, + "step": 16915 + }, + { + "epoch": 1.1496127191194456, + "grad_norm": 0.8653772473335266, + "learning_rate": 0.0008562984101100694, + "loss": 3.5308, + "step": 16920 + }, + { + "epoch": 1.1499524391901073, + "grad_norm": 1.1585173606872559, + "learning_rate": 0.0008562559451012366, + "loss": 3.5779, + "step": 16925 + }, + { + "epoch": 1.150292159260769, + "grad_norm": 1.0344363451004028, + "learning_rate": 0.0008562134800924039, + "loss": 3.6363, + "step": 16930 + }, + { + "epoch": 1.150631879331431, + "grad_norm": 1.2529666423797607, + "learning_rate": 0.0008561710150835712, + "loss": 3.54, + "step": 16935 + }, + { + "epoch": 1.1509715994020926, + "grad_norm": 0.7306733727455139, + "learning_rate": 0.0008561285500747384, + "loss": 3.7216, + "step": 16940 + }, + { + "epoch": 1.1513113194727544, + "grad_norm": 0.7992914319038391, + "learning_rate": 0.0008560860850659056, + "loss": 3.4651, + "step": 16945 + }, + { + "epoch": 1.1516510395434163, + "grad_norm": 0.7225033640861511, + "learning_rate": 0.0008560436200570731, + "loss": 3.6823, + "step": 16950 + }, + { + "epoch": 1.151990759614078, + "grad_norm": 0.7301670908927917, + "learning_rate": 0.0008560011550482403, + "loss": 3.3418, + "step": 16955 + }, + { + "epoch": 1.1523304796847398, + "grad_norm": 0.9155261516571045, + "learning_rate": 0.0008559586900394075, + "loss": 3.4561, + "step": 16960 + }, + { + "epoch": 1.1526701997554016, + "grad_norm": 0.962055504322052, + "learning_rate": 0.0008559162250305749, + "loss": 3.5986, + "step": 16965 + }, + { + "epoch": 1.1530099198260633, + "grad_norm": 0.8913617730140686, + "learning_rate": 0.0008558737600217421, + "loss": 3.5035, + "step": 16970 + }, + { + "epoch": 1.1533496398967251, + "grad_norm": 0.944158136844635, + "learning_rate": 0.0008558312950129093, + "loss": 3.5333, + "step": 16975 + }, + { + "epoch": 1.153689359967387, + "grad_norm": 1.0185366868972778, + "learning_rate": 0.0008557888300040767, + "loss": 3.6026, + "step": 16980 + }, + { + "epoch": 1.1540290800380486, + "grad_norm": 1.439587950706482, + "learning_rate": 0.000855746364995244, + "loss": 3.8101, + "step": 16985 + }, + { + "epoch": 1.1543688001087105, + "grad_norm": 1.1215914487838745, + "learning_rate": 0.0008557038999864112, + "loss": 3.4887, + "step": 16990 + }, + { + "epoch": 1.1547085201793723, + "grad_norm": 1.2692121267318726, + "learning_rate": 0.0008556614349775786, + "loss": 3.5968, + "step": 16995 + }, + { + "epoch": 1.155048240250034, + "grad_norm": 0.9150411486625671, + "learning_rate": 0.0008556189699687458, + "loss": 3.5848, + "step": 17000 + }, + { + "epoch": 1.1553879603206958, + "grad_norm": 1.4083114862442017, + "learning_rate": 0.000855576504959913, + "loss": 3.653, + "step": 17005 + }, + { + "epoch": 1.1557276803913574, + "grad_norm": 0.9078852534294128, + "learning_rate": 0.0008555340399510803, + "loss": 3.7396, + "step": 17010 + }, + { + "epoch": 1.1560674004620193, + "grad_norm": 0.8285259008407593, + "learning_rate": 0.0008554915749422476, + "loss": 3.7799, + "step": 17015 + }, + { + "epoch": 1.1564071205326811, + "grad_norm": 0.6048609018325806, + "learning_rate": 0.0008554491099334149, + "loss": 3.6569, + "step": 17020 + }, + { + "epoch": 1.1567468406033428, + "grad_norm": 1.0835025310516357, + "learning_rate": 0.0008554066449245822, + "loss": 3.8416, + "step": 17025 + }, + { + "epoch": 1.1570865606740046, + "grad_norm": 1.3159259557724, + "learning_rate": 0.0008553641799157495, + "loss": 3.5665, + "step": 17030 + }, + { + "epoch": 1.1574262807446665, + "grad_norm": 1.063502550125122, + "learning_rate": 0.0008553217149069167, + "loss": 3.5922, + "step": 17035 + }, + { + "epoch": 1.157766000815328, + "grad_norm": 1.0917658805847168, + "learning_rate": 0.000855279249898084, + "loss": 3.3665, + "step": 17040 + }, + { + "epoch": 1.15810572088599, + "grad_norm": 1.282745599746704, + "learning_rate": 0.0008552367848892512, + "loss": 3.8931, + "step": 17045 + }, + { + "epoch": 1.1584454409566518, + "grad_norm": 0.9100906848907471, + "learning_rate": 0.0008551943198804185, + "loss": 3.3497, + "step": 17050 + }, + { + "epoch": 1.1587851610273134, + "grad_norm": 1.7236732244491577, + "learning_rate": 0.0008551518548715859, + "loss": 3.6626, + "step": 17055 + }, + { + "epoch": 1.1591248810979753, + "grad_norm": 0.8754659295082092, + "learning_rate": 0.0008551093898627531, + "loss": 3.6369, + "step": 17060 + }, + { + "epoch": 1.159464601168637, + "grad_norm": 0.8582232594490051, + "learning_rate": 0.0008550669248539204, + "loss": 3.7551, + "step": 17065 + }, + { + "epoch": 1.1598043212392988, + "grad_norm": 0.7621845602989197, + "learning_rate": 0.0008550244598450877, + "loss": 3.5718, + "step": 17070 + }, + { + "epoch": 1.1601440413099606, + "grad_norm": 0.7973881959915161, + "learning_rate": 0.0008549819948362549, + "loss": 3.748, + "step": 17075 + }, + { + "epoch": 1.1604837613806223, + "grad_norm": 2.1430253982543945, + "learning_rate": 0.0008549395298274222, + "loss": 3.3996, + "step": 17080 + }, + { + "epoch": 1.1608234814512841, + "grad_norm": 1.0586857795715332, + "learning_rate": 0.0008548970648185895, + "loss": 3.5756, + "step": 17085 + }, + { + "epoch": 1.161163201521946, + "grad_norm": 0.7825150489807129, + "learning_rate": 0.0008548545998097568, + "loss": 3.5194, + "step": 17090 + }, + { + "epoch": 1.1615029215926076, + "grad_norm": 0.7929461002349854, + "learning_rate": 0.000854812134800924, + "loss": 3.5636, + "step": 17095 + }, + { + "epoch": 1.1618426416632694, + "grad_norm": 0.8021267652511597, + "learning_rate": 0.0008547696697920914, + "loss": 3.5552, + "step": 17100 + }, + { + "epoch": 1.1621823617339313, + "grad_norm": 0.7752456068992615, + "learning_rate": 0.0008547272047832586, + "loss": 3.5965, + "step": 17105 + }, + { + "epoch": 1.162522081804593, + "grad_norm": 0.7759296894073486, + "learning_rate": 0.0008546847397744258, + "loss": 3.4726, + "step": 17110 + }, + { + "epoch": 1.1628618018752548, + "grad_norm": 2.4652488231658936, + "learning_rate": 0.0008546422747655932, + "loss": 3.476, + "step": 17115 + }, + { + "epoch": 1.1632015219459166, + "grad_norm": 0.7785670757293701, + "learning_rate": 0.0008545998097567604, + "loss": 3.589, + "step": 17120 + }, + { + "epoch": 1.1635412420165783, + "grad_norm": 0.7526088356971741, + "learning_rate": 0.0008545573447479277, + "loss": 3.5264, + "step": 17125 + }, + { + "epoch": 1.1638809620872401, + "grad_norm": 0.6999656558036804, + "learning_rate": 0.0008545148797390951, + "loss": 3.8938, + "step": 17130 + }, + { + "epoch": 1.164220682157902, + "grad_norm": 0.8884168267250061, + "learning_rate": 0.0008544724147302623, + "loss": 3.3884, + "step": 17135 + }, + { + "epoch": 1.1645604022285636, + "grad_norm": 0.8296924829483032, + "learning_rate": 0.0008544299497214295, + "loss": 3.8405, + "step": 17140 + }, + { + "epoch": 1.1649001222992255, + "grad_norm": 0.8450603485107422, + "learning_rate": 0.0008543874847125968, + "loss": 3.6565, + "step": 17145 + }, + { + "epoch": 1.1652398423698873, + "grad_norm": 0.9728952646255493, + "learning_rate": 0.0008543450197037641, + "loss": 3.6436, + "step": 17150 + }, + { + "epoch": 1.165579562440549, + "grad_norm": 0.8507139682769775, + "learning_rate": 0.0008543025546949313, + "loss": 3.6495, + "step": 17155 + }, + { + "epoch": 1.1659192825112108, + "grad_norm": 0.728164792060852, + "learning_rate": 0.0008542600896860987, + "loss": 3.8071, + "step": 17160 + }, + { + "epoch": 1.1662590025818727, + "grad_norm": 0.971473753452301, + "learning_rate": 0.000854217624677266, + "loss": 3.6427, + "step": 17165 + }, + { + "epoch": 1.1665987226525343, + "grad_norm": 0.6229178309440613, + "learning_rate": 0.0008541751596684332, + "loss": 3.489, + "step": 17170 + }, + { + "epoch": 1.1669384427231961, + "grad_norm": 0.9602677226066589, + "learning_rate": 0.0008541326946596005, + "loss": 3.8143, + "step": 17175 + }, + { + "epoch": 1.1672781627938578, + "grad_norm": 2.633232593536377, + "learning_rate": 0.0008540902296507678, + "loss": 3.8185, + "step": 17180 + }, + { + "epoch": 1.1676178828645196, + "grad_norm": 0.764832615852356, + "learning_rate": 0.000854047764641935, + "loss": 3.6915, + "step": 17185 + }, + { + "epoch": 1.1679576029351815, + "grad_norm": 0.9543217420578003, + "learning_rate": 0.0008540052996331023, + "loss": 3.4057, + "step": 17190 + }, + { + "epoch": 1.168297323005843, + "grad_norm": 1.017743468284607, + "learning_rate": 0.0008539628346242696, + "loss": 3.527, + "step": 17195 + }, + { + "epoch": 1.168637043076505, + "grad_norm": 1.7346428632736206, + "learning_rate": 0.0008539203696154369, + "loss": 3.8696, + "step": 17200 + }, + { + "epoch": 1.1689767631471668, + "grad_norm": 0.6835223436355591, + "learning_rate": 0.0008538779046066042, + "loss": 3.5876, + "step": 17205 + }, + { + "epoch": 1.1693164832178284, + "grad_norm": 0.7966700196266174, + "learning_rate": 0.0008538354395977714, + "loss": 3.6236, + "step": 17210 + }, + { + "epoch": 1.1696562032884903, + "grad_norm": 0.7694163918495178, + "learning_rate": 0.0008537929745889388, + "loss": 3.5424, + "step": 17215 + }, + { + "epoch": 1.1699959233591521, + "grad_norm": 7.493810176849365, + "learning_rate": 0.000853750509580106, + "loss": 3.703, + "step": 17220 + }, + { + "epoch": 1.1703356434298138, + "grad_norm": 1.1680781841278076, + "learning_rate": 0.0008537080445712732, + "loss": 3.5469, + "step": 17225 + }, + { + "epoch": 1.1706753635004756, + "grad_norm": 1.880760669708252, + "learning_rate": 0.0008536655795624407, + "loss": 3.5878, + "step": 17230 + }, + { + "epoch": 1.1710150835711373, + "grad_norm": 5.972364902496338, + "learning_rate": 0.0008536231145536079, + "loss": 3.4208, + "step": 17235 + }, + { + "epoch": 1.1713548036417991, + "grad_norm": 1.596903681755066, + "learning_rate": 0.0008535806495447751, + "loss": 3.3627, + "step": 17240 + }, + { + "epoch": 1.171694523712461, + "grad_norm": 2.3864011764526367, + "learning_rate": 0.0008535381845359424, + "loss": 3.5494, + "step": 17245 + }, + { + "epoch": 1.1720342437831226, + "grad_norm": 0.8851460218429565, + "learning_rate": 0.0008534957195271097, + "loss": 3.6108, + "step": 17250 + }, + { + "epoch": 1.1723739638537845, + "grad_norm": 1.5603492259979248, + "learning_rate": 0.0008534532545182769, + "loss": 3.8096, + "step": 17255 + }, + { + "epoch": 1.1727136839244463, + "grad_norm": 0.7604426741600037, + "learning_rate": 0.0008534107895094442, + "loss": 3.6608, + "step": 17260 + }, + { + "epoch": 1.173053403995108, + "grad_norm": 0.8749967217445374, + "learning_rate": 0.0008533683245006116, + "loss": 3.6138, + "step": 17265 + }, + { + "epoch": 1.1733931240657698, + "grad_norm": 0.9765446186065674, + "learning_rate": 0.0008533258594917788, + "loss": 3.5996, + "step": 17270 + }, + { + "epoch": 1.1737328441364316, + "grad_norm": 0.9182650446891785, + "learning_rate": 0.0008532833944829461, + "loss": 3.3802, + "step": 17275 + }, + { + "epoch": 1.1740725642070933, + "grad_norm": 1.864100456237793, + "learning_rate": 0.0008532409294741134, + "loss": 3.8625, + "step": 17280 + }, + { + "epoch": 1.1744122842777551, + "grad_norm": 0.9317402243614197, + "learning_rate": 0.0008531984644652806, + "loss": 3.5741, + "step": 17285 + }, + { + "epoch": 1.174752004348417, + "grad_norm": 0.7591428756713867, + "learning_rate": 0.0008531559994564479, + "loss": 3.6428, + "step": 17290 + }, + { + "epoch": 1.1750917244190786, + "grad_norm": 0.8667760491371155, + "learning_rate": 0.0008531135344476151, + "loss": 3.7521, + "step": 17295 + }, + { + "epoch": 1.1754314444897405, + "grad_norm": 0.8764576315879822, + "learning_rate": 0.0008530710694387825, + "loss": 3.7382, + "step": 17300 + }, + { + "epoch": 1.1757711645604023, + "grad_norm": 0.7918635010719299, + "learning_rate": 0.0008530286044299498, + "loss": 3.7379, + "step": 17305 + }, + { + "epoch": 1.176110884631064, + "grad_norm": 0.9020955562591553, + "learning_rate": 0.000852986139421117, + "loss": 3.7546, + "step": 17310 + }, + { + "epoch": 1.1764506047017258, + "grad_norm": 0.9699614644050598, + "learning_rate": 0.0008529436744122843, + "loss": 3.6168, + "step": 17315 + }, + { + "epoch": 1.1767903247723877, + "grad_norm": 0.6874472498893738, + "learning_rate": 0.0008529012094034516, + "loss": 3.8207, + "step": 17320 + }, + { + "epoch": 1.1771300448430493, + "grad_norm": 0.8439446687698364, + "learning_rate": 0.0008528587443946188, + "loss": 3.6292, + "step": 17325 + }, + { + "epoch": 1.1774697649137111, + "grad_norm": 0.7672669291496277, + "learning_rate": 0.000852816279385786, + "loss": 3.8878, + "step": 17330 + }, + { + "epoch": 1.177809484984373, + "grad_norm": 0.7623843550682068, + "learning_rate": 0.0008527738143769535, + "loss": 3.7638, + "step": 17335 + }, + { + "epoch": 1.1781492050550346, + "grad_norm": 3.8027656078338623, + "learning_rate": 0.0008527313493681207, + "loss": 3.6529, + "step": 17340 + }, + { + "epoch": 1.1784889251256965, + "grad_norm": 1.0129262208938599, + "learning_rate": 0.0008526888843592879, + "loss": 3.5983, + "step": 17345 + }, + { + "epoch": 1.178828645196358, + "grad_norm": 0.684468150138855, + "learning_rate": 0.0008526464193504553, + "loss": 3.5052, + "step": 17350 + }, + { + "epoch": 1.17916836526702, + "grad_norm": 0.7793314456939697, + "learning_rate": 0.0008526039543416225, + "loss": 3.4746, + "step": 17355 + }, + { + "epoch": 1.1795080853376818, + "grad_norm": 0.8944817185401917, + "learning_rate": 0.0008525614893327897, + "loss": 3.7936, + "step": 17360 + }, + { + "epoch": 1.1798478054083434, + "grad_norm": 1.2124223709106445, + "learning_rate": 0.0008525190243239571, + "loss": 3.313, + "step": 17365 + }, + { + "epoch": 1.1801875254790053, + "grad_norm": 0.9932569861412048, + "learning_rate": 0.0008524765593151244, + "loss": 3.5552, + "step": 17370 + }, + { + "epoch": 1.1805272455496671, + "grad_norm": 0.84503173828125, + "learning_rate": 0.0008524340943062916, + "loss": 3.5939, + "step": 17375 + }, + { + "epoch": 1.1808669656203288, + "grad_norm": 0.7551472783088684, + "learning_rate": 0.000852391629297459, + "loss": 3.6712, + "step": 17380 + }, + { + "epoch": 1.1812066856909906, + "grad_norm": 0.9820950031280518, + "learning_rate": 0.0008523491642886262, + "loss": 3.3883, + "step": 17385 + }, + { + "epoch": 1.1815464057616525, + "grad_norm": 2.210876941680908, + "learning_rate": 0.0008523066992797934, + "loss": 3.5562, + "step": 17390 + }, + { + "epoch": 1.1818861258323141, + "grad_norm": 0.9046183824539185, + "learning_rate": 0.0008522642342709607, + "loss": 3.4567, + "step": 17395 + }, + { + "epoch": 1.182225845902976, + "grad_norm": 0.7804042100906372, + "learning_rate": 0.000852221769262128, + "loss": 3.4418, + "step": 17400 + }, + { + "epoch": 1.1825655659736376, + "grad_norm": 1.4414958953857422, + "learning_rate": 0.0008521793042532953, + "loss": 3.4449, + "step": 17405 + }, + { + "epoch": 1.1829052860442995, + "grad_norm": 0.8063602447509766, + "learning_rate": 0.0008521368392444626, + "loss": 3.7806, + "step": 17410 + }, + { + "epoch": 1.1832450061149613, + "grad_norm": 0.7786731719970703, + "learning_rate": 0.0008520943742356299, + "loss": 3.5172, + "step": 17415 + }, + { + "epoch": 1.183584726185623, + "grad_norm": 1.3335820436477661, + "learning_rate": 0.0008520519092267971, + "loss": 3.5369, + "step": 17420 + }, + { + "epoch": 1.1839244462562848, + "grad_norm": 0.7360917329788208, + "learning_rate": 0.0008520094442179644, + "loss": 3.5619, + "step": 17425 + }, + { + "epoch": 1.1842641663269466, + "grad_norm": 0.7670918107032776, + "learning_rate": 0.0008519669792091316, + "loss": 3.8127, + "step": 17430 + }, + { + "epoch": 1.1846038863976083, + "grad_norm": 0.6371151208877563, + "learning_rate": 0.0008519245142002989, + "loss": 3.9432, + "step": 17435 + }, + { + "epoch": 1.1849436064682701, + "grad_norm": 0.916063129901886, + "learning_rate": 0.0008518820491914663, + "loss": 3.5995, + "step": 17440 + }, + { + "epoch": 1.185283326538932, + "grad_norm": 0.7930320501327515, + "learning_rate": 0.0008518395841826335, + "loss": 3.3264, + "step": 17445 + }, + { + "epoch": 1.1856230466095936, + "grad_norm": 0.86455237865448, + "learning_rate": 0.0008517971191738008, + "loss": 3.5734, + "step": 17450 + }, + { + "epoch": 1.1859627666802555, + "grad_norm": 0.8365366458892822, + "learning_rate": 0.0008517546541649681, + "loss": 3.743, + "step": 17455 + }, + { + "epoch": 1.1863024867509173, + "grad_norm": 0.6427135467529297, + "learning_rate": 0.0008517121891561353, + "loss": 3.6662, + "step": 17460 + }, + { + "epoch": 1.186642206821579, + "grad_norm": 0.952447772026062, + "learning_rate": 0.0008516697241473026, + "loss": 3.6493, + "step": 17465 + }, + { + "epoch": 1.1869819268922408, + "grad_norm": 0.765146791934967, + "learning_rate": 0.00085162725913847, + "loss": 3.8014, + "step": 17470 + }, + { + "epoch": 1.1873216469629027, + "grad_norm": 0.9991302490234375, + "learning_rate": 0.0008515847941296372, + "loss": 3.2053, + "step": 17475 + }, + { + "epoch": 1.1876613670335643, + "grad_norm": 0.9211218953132629, + "learning_rate": 0.0008515423291208044, + "loss": 3.6564, + "step": 17480 + }, + { + "epoch": 1.1880010871042261, + "grad_norm": 3.253815174102783, + "learning_rate": 0.0008514998641119718, + "loss": 3.7077, + "step": 17485 + }, + { + "epoch": 1.188340807174888, + "grad_norm": 0.8603718876838684, + "learning_rate": 0.000851457399103139, + "loss": 3.3403, + "step": 17490 + }, + { + "epoch": 1.1886805272455496, + "grad_norm": 0.8207667469978333, + "learning_rate": 0.0008514149340943062, + "loss": 3.4288, + "step": 17495 + }, + { + "epoch": 1.1890202473162115, + "grad_norm": 0.7828772068023682, + "learning_rate": 0.0008513724690854736, + "loss": 3.4702, + "step": 17500 + }, + { + "epoch": 1.1893599673868733, + "grad_norm": 0.790790319442749, + "learning_rate": 0.0008513300040766409, + "loss": 3.4339, + "step": 17505 + }, + { + "epoch": 1.189699687457535, + "grad_norm": 0.742437481880188, + "learning_rate": 0.0008512875390678081, + "loss": 3.5319, + "step": 17510 + }, + { + "epoch": 1.1900394075281968, + "grad_norm": 0.7208537459373474, + "learning_rate": 0.0008512450740589755, + "loss": 3.8602, + "step": 17515 + }, + { + "epoch": 1.1903791275988584, + "grad_norm": 0.8345505595207214, + "learning_rate": 0.0008512026090501427, + "loss": 3.5534, + "step": 17520 + }, + { + "epoch": 1.1907188476695203, + "grad_norm": 1.2964720726013184, + "learning_rate": 0.0008511601440413099, + "loss": 3.633, + "step": 17525 + }, + { + "epoch": 1.1910585677401822, + "grad_norm": 0.7335799932479858, + "learning_rate": 0.0008511176790324772, + "loss": 3.652, + "step": 17530 + }, + { + "epoch": 1.1913982878108438, + "grad_norm": 0.8151165843009949, + "learning_rate": 0.0008510752140236445, + "loss": 3.7206, + "step": 17535 + }, + { + "epoch": 1.1917380078815056, + "grad_norm": 0.8891379833221436, + "learning_rate": 0.0008510327490148118, + "loss": 3.5112, + "step": 17540 + }, + { + "epoch": 1.1920777279521675, + "grad_norm": 0.9424435496330261, + "learning_rate": 0.0008509902840059791, + "loss": 3.6306, + "step": 17545 + }, + { + "epoch": 1.1924174480228291, + "grad_norm": 0.8631328344345093, + "learning_rate": 0.0008509478189971464, + "loss": 3.7135, + "step": 17550 + }, + { + "epoch": 1.192757168093491, + "grad_norm": 1.1773486137390137, + "learning_rate": 0.0008509053539883137, + "loss": 3.6368, + "step": 17555 + }, + { + "epoch": 1.1930968881641528, + "grad_norm": 0.8781136870384216, + "learning_rate": 0.0008508628889794809, + "loss": 3.6872, + "step": 17560 + }, + { + "epoch": 1.1934366082348145, + "grad_norm": 0.9004260897636414, + "learning_rate": 0.0008508204239706482, + "loss": 3.3981, + "step": 17565 + }, + { + "epoch": 1.1937763283054763, + "grad_norm": 0.8195340633392334, + "learning_rate": 0.0008507779589618155, + "loss": 3.6932, + "step": 17570 + }, + { + "epoch": 1.194116048376138, + "grad_norm": 1.0165120363235474, + "learning_rate": 0.0008507354939529828, + "loss": 3.3475, + "step": 17575 + }, + { + "epoch": 1.1944557684467998, + "grad_norm": 0.7121320366859436, + "learning_rate": 0.00085069302894415, + "loss": 3.6105, + "step": 17580 + }, + { + "epoch": 1.1947954885174616, + "grad_norm": 1.007370114326477, + "learning_rate": 0.0008506505639353174, + "loss": 3.7749, + "step": 17585 + }, + { + "epoch": 1.1951352085881233, + "grad_norm": 0.9096401333808899, + "learning_rate": 0.0008506080989264846, + "loss": 3.7979, + "step": 17590 + }, + { + "epoch": 1.1954749286587851, + "grad_norm": 0.8496130704879761, + "learning_rate": 0.0008505656339176518, + "loss": 3.7891, + "step": 17595 + }, + { + "epoch": 1.195814648729447, + "grad_norm": 0.6904847621917725, + "learning_rate": 0.0008505231689088192, + "loss": 3.697, + "step": 17600 + }, + { + "epoch": 1.1961543688001086, + "grad_norm": 0.9885188341140747, + "learning_rate": 0.0008504807038999864, + "loss": 3.4673, + "step": 17605 + }, + { + "epoch": 1.1964940888707705, + "grad_norm": 0.7779517769813538, + "learning_rate": 0.0008504382388911537, + "loss": 3.1831, + "step": 17610 + }, + { + "epoch": 1.1968338089414323, + "grad_norm": 1.2768638134002686, + "learning_rate": 0.0008503957738823211, + "loss": 3.6647, + "step": 17615 + }, + { + "epoch": 1.197173529012094, + "grad_norm": 0.8547077775001526, + "learning_rate": 0.0008503533088734883, + "loss": 3.5678, + "step": 17620 + }, + { + "epoch": 1.1975132490827558, + "grad_norm": 0.9448862075805664, + "learning_rate": 0.0008503108438646555, + "loss": 3.3435, + "step": 17625 + }, + { + "epoch": 1.1978529691534177, + "grad_norm": 0.7758262157440186, + "learning_rate": 0.0008502683788558229, + "loss": 3.5426, + "step": 17630 + }, + { + "epoch": 1.1981926892240793, + "grad_norm": 0.6783838868141174, + "learning_rate": 0.0008502259138469901, + "loss": 3.4621, + "step": 17635 + }, + { + "epoch": 1.1985324092947411, + "grad_norm": 0.8388721942901611, + "learning_rate": 0.0008501834488381573, + "loss": 3.7712, + "step": 17640 + }, + { + "epoch": 1.198872129365403, + "grad_norm": 1.356605887413025, + "learning_rate": 0.0008501409838293247, + "loss": 3.4928, + "step": 17645 + }, + { + "epoch": 1.1992118494360646, + "grad_norm": 0.8604318499565125, + "learning_rate": 0.000850098518820492, + "loss": 3.5861, + "step": 17650 + }, + { + "epoch": 1.1995515695067265, + "grad_norm": 0.758072555065155, + "learning_rate": 0.0008500560538116592, + "loss": 3.6657, + "step": 17655 + }, + { + "epoch": 1.1998912895773883, + "grad_norm": 0.816387414932251, + "learning_rate": 0.0008500135888028265, + "loss": 3.6745, + "step": 17660 + }, + { + "epoch": 1.20023100964805, + "grad_norm": 0.8752282857894897, + "learning_rate": 0.0008499711237939938, + "loss": 3.7374, + "step": 17665 + }, + { + "epoch": 1.2005707297187118, + "grad_norm": 0.6966831684112549, + "learning_rate": 0.000849928658785161, + "loss": 3.5319, + "step": 17670 + }, + { + "epoch": 1.2009104497893737, + "grad_norm": 0.8720716834068298, + "learning_rate": 0.0008498861937763283, + "loss": 3.5579, + "step": 17675 + }, + { + "epoch": 1.2012501698600353, + "grad_norm": 0.8371026515960693, + "learning_rate": 0.0008498437287674957, + "loss": 3.379, + "step": 17680 + }, + { + "epoch": 1.2015898899306972, + "grad_norm": 1.0002015829086304, + "learning_rate": 0.0008498012637586629, + "loss": 3.757, + "step": 17685 + }, + { + "epoch": 1.2019296100013588, + "grad_norm": 0.7890312075614929, + "learning_rate": 0.0008497587987498302, + "loss": 3.8396, + "step": 17690 + }, + { + "epoch": 1.2022693300720206, + "grad_norm": 0.7668339014053345, + "learning_rate": 0.0008497163337409974, + "loss": 3.4249, + "step": 17695 + }, + { + "epoch": 1.2026090501426825, + "grad_norm": 0.6268309950828552, + "learning_rate": 0.0008496738687321647, + "loss": 3.3876, + "step": 17700 + }, + { + "epoch": 1.2029487702133441, + "grad_norm": 0.8368004560470581, + "learning_rate": 0.000849631403723332, + "loss": 3.3393, + "step": 17705 + }, + { + "epoch": 1.203288490284006, + "grad_norm": 0.8494678139686584, + "learning_rate": 0.0008495889387144992, + "loss": 3.9974, + "step": 17710 + }, + { + "epoch": 1.2036282103546678, + "grad_norm": 2.8528928756713867, + "learning_rate": 0.0008495464737056666, + "loss": 3.3394, + "step": 17715 + }, + { + "epoch": 1.2039679304253295, + "grad_norm": 0.7073647379875183, + "learning_rate": 0.0008495040086968339, + "loss": 3.7781, + "step": 17720 + }, + { + "epoch": 1.2043076504959913, + "grad_norm": 0.8016384243965149, + "learning_rate": 0.0008494615436880011, + "loss": 3.7749, + "step": 17725 + }, + { + "epoch": 1.2046473705666532, + "grad_norm": 0.8881009817123413, + "learning_rate": 0.0008494190786791683, + "loss": 3.5789, + "step": 17730 + }, + { + "epoch": 1.2049870906373148, + "grad_norm": 1.0380406379699707, + "learning_rate": 0.0008493766136703357, + "loss": 3.5601, + "step": 17735 + }, + { + "epoch": 1.2053268107079766, + "grad_norm": 0.8030280470848083, + "learning_rate": 0.0008493341486615029, + "loss": 3.5167, + "step": 17740 + }, + { + "epoch": 1.2056665307786383, + "grad_norm": 0.6908677816390991, + "learning_rate": 0.0008492916836526701, + "loss": 3.636, + "step": 17745 + }, + { + "epoch": 1.2060062508493001, + "grad_norm": 0.7932618856430054, + "learning_rate": 0.0008492492186438376, + "loss": 3.7781, + "step": 17750 + }, + { + "epoch": 1.206345970919962, + "grad_norm": 1.4347883462905884, + "learning_rate": 0.0008492067536350048, + "loss": 3.4018, + "step": 17755 + }, + { + "epoch": 1.2066856909906236, + "grad_norm": 2.564614772796631, + "learning_rate": 0.000849164288626172, + "loss": 3.68, + "step": 17760 + }, + { + "epoch": 1.2070254110612855, + "grad_norm": 0.7498965263366699, + "learning_rate": 0.0008491218236173394, + "loss": 3.7144, + "step": 17765 + }, + { + "epoch": 1.2073651311319473, + "grad_norm": 1.0869518518447876, + "learning_rate": 0.0008490793586085066, + "loss": 3.4769, + "step": 17770 + }, + { + "epoch": 1.207704851202609, + "grad_norm": 3.0085835456848145, + "learning_rate": 0.0008490368935996738, + "loss": 3.8415, + "step": 17775 + }, + { + "epoch": 1.2080445712732708, + "grad_norm": 0.8655113577842712, + "learning_rate": 0.0008489944285908411, + "loss": 3.6228, + "step": 17780 + }, + { + "epoch": 1.2083842913439327, + "grad_norm": 10.374820709228516, + "learning_rate": 0.0008489519635820085, + "loss": 3.5565, + "step": 17785 + }, + { + "epoch": 1.2087240114145943, + "grad_norm": 1.2534915208816528, + "learning_rate": 0.0008489094985731757, + "loss": 3.9803, + "step": 17790 + }, + { + "epoch": 1.2090637314852561, + "grad_norm": 0.7677974700927734, + "learning_rate": 0.000848867033564343, + "loss": 3.6673, + "step": 17795 + }, + { + "epoch": 1.209403451555918, + "grad_norm": 0.8613688945770264, + "learning_rate": 0.0008488245685555103, + "loss": 3.8141, + "step": 17800 + }, + { + "epoch": 1.2097431716265796, + "grad_norm": 0.9198121428489685, + "learning_rate": 0.0008487821035466775, + "loss": 3.3198, + "step": 17805 + }, + { + "epoch": 1.2100828916972415, + "grad_norm": 2.8198814392089844, + "learning_rate": 0.0008487396385378448, + "loss": 3.5049, + "step": 17810 + }, + { + "epoch": 1.2104226117679033, + "grad_norm": 1.3424034118652344, + "learning_rate": 0.000848697173529012, + "loss": 3.7246, + "step": 17815 + }, + { + "epoch": 1.210762331838565, + "grad_norm": 0.7301501035690308, + "learning_rate": 0.0008486547085201794, + "loss": 3.6335, + "step": 17820 + }, + { + "epoch": 1.2111020519092268, + "grad_norm": 5.553789138793945, + "learning_rate": 0.0008486122435113467, + "loss": 3.6726, + "step": 17825 + }, + { + "epoch": 1.2114417719798887, + "grad_norm": 1.5404613018035889, + "learning_rate": 0.0008485697785025139, + "loss": 3.7583, + "step": 17830 + }, + { + "epoch": 1.2117814920505503, + "grad_norm": 0.8810997009277344, + "learning_rate": 0.0008485273134936812, + "loss": 3.6484, + "step": 17835 + }, + { + "epoch": 1.2121212121212122, + "grad_norm": 0.8727395534515381, + "learning_rate": 0.0008484848484848485, + "loss": 3.8148, + "step": 17840 + }, + { + "epoch": 1.212460932191874, + "grad_norm": 0.8452165126800537, + "learning_rate": 0.0008484423834760157, + "loss": 3.7303, + "step": 17845 + }, + { + "epoch": 1.2128006522625356, + "grad_norm": 0.8771845102310181, + "learning_rate": 0.000848399918467183, + "loss": 3.6504, + "step": 17850 + }, + { + "epoch": 1.2131403723331975, + "grad_norm": 0.9076299071311951, + "learning_rate": 0.0008483574534583504, + "loss": 3.8004, + "step": 17855 + }, + { + "epoch": 1.2134800924038591, + "grad_norm": 0.7213129997253418, + "learning_rate": 0.0008483149884495176, + "loss": 3.565, + "step": 17860 + }, + { + "epoch": 1.213819812474521, + "grad_norm": 1.2307398319244385, + "learning_rate": 0.0008482725234406849, + "loss": 3.8107, + "step": 17865 + }, + { + "epoch": 1.2141595325451828, + "grad_norm": 0.8960558772087097, + "learning_rate": 0.0008482300584318522, + "loss": 3.4757, + "step": 17870 + }, + { + "epoch": 1.2144992526158445, + "grad_norm": 0.7587851285934448, + "learning_rate": 0.0008481875934230194, + "loss": 3.854, + "step": 17875 + }, + { + "epoch": 1.2148389726865063, + "grad_norm": 0.7378130555152893, + "learning_rate": 0.0008481451284141866, + "loss": 3.7177, + "step": 17880 + }, + { + "epoch": 1.2151786927571682, + "grad_norm": 0.9377250671386719, + "learning_rate": 0.000848102663405354, + "loss": 3.4566, + "step": 17885 + }, + { + "epoch": 1.2155184128278298, + "grad_norm": 0.6432008147239685, + "learning_rate": 0.0008480601983965213, + "loss": 3.6867, + "step": 17890 + }, + { + "epoch": 1.2158581328984917, + "grad_norm": 0.7586209177970886, + "learning_rate": 0.0008480177333876886, + "loss": 3.8213, + "step": 17895 + }, + { + "epoch": 1.2161978529691535, + "grad_norm": 0.7203810214996338, + "learning_rate": 0.0008479752683788559, + "loss": 3.7752, + "step": 17900 + }, + { + "epoch": 1.2165375730398151, + "grad_norm": 1.1642966270446777, + "learning_rate": 0.0008479328033700231, + "loss": 3.6636, + "step": 17905 + }, + { + "epoch": 1.216877293110477, + "grad_norm": 0.7937231659889221, + "learning_rate": 0.0008478903383611904, + "loss": 3.5277, + "step": 17910 + }, + { + "epoch": 1.2172170131811386, + "grad_norm": 1.0121312141418457, + "learning_rate": 0.0008478478733523577, + "loss": 3.585, + "step": 17915 + }, + { + "epoch": 1.2175567332518005, + "grad_norm": 0.7786787152290344, + "learning_rate": 0.0008478054083435249, + "loss": 3.7303, + "step": 17920 + }, + { + "epoch": 1.2178964533224623, + "grad_norm": 1.0357321500778198, + "learning_rate": 0.0008477629433346923, + "loss": 3.7174, + "step": 17925 + }, + { + "epoch": 1.218236173393124, + "grad_norm": 0.8810575604438782, + "learning_rate": 0.0008477204783258595, + "loss": 3.4362, + "step": 17930 + }, + { + "epoch": 1.2185758934637858, + "grad_norm": 0.7524788975715637, + "learning_rate": 0.0008476780133170268, + "loss": 3.7864, + "step": 17935 + }, + { + "epoch": 1.2189156135344477, + "grad_norm": 0.8959475159645081, + "learning_rate": 0.0008476355483081941, + "loss": 3.4437, + "step": 17940 + }, + { + "epoch": 1.2192553336051093, + "grad_norm": 0.6720718741416931, + "learning_rate": 0.0008475930832993613, + "loss": 3.9562, + "step": 17945 + }, + { + "epoch": 1.2195950536757711, + "grad_norm": 0.816149890422821, + "learning_rate": 0.0008475506182905286, + "loss": 3.5219, + "step": 17950 + }, + { + "epoch": 1.219934773746433, + "grad_norm": 2.802460193634033, + "learning_rate": 0.000847508153281696, + "loss": 3.5521, + "step": 17955 + }, + { + "epoch": 1.2202744938170946, + "grad_norm": 0.7368342280387878, + "learning_rate": 0.0008474656882728632, + "loss": 3.6967, + "step": 17960 + }, + { + "epoch": 1.2206142138877565, + "grad_norm": 0.935755729675293, + "learning_rate": 0.0008474232232640305, + "loss": 3.5699, + "step": 17965 + }, + { + "epoch": 1.2209539339584183, + "grad_norm": 0.679125189781189, + "learning_rate": 0.0008473807582551978, + "loss": 3.6766, + "step": 17970 + }, + { + "epoch": 1.22129365402908, + "grad_norm": 0.9291485548019409, + "learning_rate": 0.000847338293246365, + "loss": 3.695, + "step": 17975 + }, + { + "epoch": 1.2216333740997418, + "grad_norm": 0.652441680431366, + "learning_rate": 0.0008472958282375322, + "loss": 3.6368, + "step": 17980 + }, + { + "epoch": 1.2219730941704037, + "grad_norm": 1.0129534006118774, + "learning_rate": 0.0008472533632286996, + "loss": 3.4031, + "step": 17985 + }, + { + "epoch": 1.2223128142410653, + "grad_norm": 0.8840212225914001, + "learning_rate": 0.0008472108982198669, + "loss": 3.7429, + "step": 17990 + }, + { + "epoch": 1.2226525343117272, + "grad_norm": 0.9972313642501831, + "learning_rate": 0.0008471684332110341, + "loss": 3.7065, + "step": 17995 + }, + { + "epoch": 1.222992254382389, + "grad_norm": 1.400220274925232, + "learning_rate": 0.0008471259682022015, + "loss": 3.6157, + "step": 18000 + }, + { + "epoch": 1.2233319744530506, + "grad_norm": 0.8838474154472351, + "learning_rate": 0.0008470835031933687, + "loss": 3.5444, + "step": 18005 + }, + { + "epoch": 1.2236716945237125, + "grad_norm": 0.8083096146583557, + "learning_rate": 0.0008470410381845359, + "loss": 3.6164, + "step": 18010 + }, + { + "epoch": 1.2240114145943743, + "grad_norm": 0.6595573425292969, + "learning_rate": 0.0008469985731757033, + "loss": 3.4771, + "step": 18015 + }, + { + "epoch": 1.224351134665036, + "grad_norm": 0.6899487972259521, + "learning_rate": 0.0008469561081668705, + "loss": 3.7901, + "step": 18020 + }, + { + "epoch": 1.2246908547356978, + "grad_norm": 0.8112339377403259, + "learning_rate": 0.0008469136431580378, + "loss": 3.5362, + "step": 18025 + }, + { + "epoch": 1.2250305748063597, + "grad_norm": 0.9344635605812073, + "learning_rate": 0.0008468711781492051, + "loss": 3.4623, + "step": 18030 + }, + { + "epoch": 1.2253702948770213, + "grad_norm": 0.7305163741111755, + "learning_rate": 0.0008468287131403724, + "loss": 3.9629, + "step": 18035 + }, + { + "epoch": 1.2257100149476832, + "grad_norm": 0.9904261827468872, + "learning_rate": 0.0008467862481315396, + "loss": 3.7765, + "step": 18040 + }, + { + "epoch": 1.2260497350183448, + "grad_norm": 0.7735642790794373, + "learning_rate": 0.0008467437831227069, + "loss": 3.7925, + "step": 18045 + }, + { + "epoch": 1.2263894550890067, + "grad_norm": 0.7642565965652466, + "learning_rate": 0.0008467013181138742, + "loss": 3.5578, + "step": 18050 + }, + { + "epoch": 1.2267291751596685, + "grad_norm": 0.8942404389381409, + "learning_rate": 0.0008466588531050414, + "loss": 3.6719, + "step": 18055 + }, + { + "epoch": 1.2270688952303301, + "grad_norm": 1.0126745700836182, + "learning_rate": 0.0008466163880962088, + "loss": 3.3308, + "step": 18060 + }, + { + "epoch": 1.227408615300992, + "grad_norm": 0.765275239944458, + "learning_rate": 0.0008465739230873761, + "loss": 3.5693, + "step": 18065 + }, + { + "epoch": 1.2277483353716538, + "grad_norm": 1.321358323097229, + "learning_rate": 0.0008465314580785433, + "loss": 3.758, + "step": 18070 + }, + { + "epoch": 1.2280880554423155, + "grad_norm": 0.8480993509292603, + "learning_rate": 0.0008464889930697106, + "loss": 3.7312, + "step": 18075 + }, + { + "epoch": 1.2284277755129773, + "grad_norm": 0.8472058773040771, + "learning_rate": 0.0008464465280608778, + "loss": 3.4252, + "step": 18080 + }, + { + "epoch": 1.228767495583639, + "grad_norm": 0.8045173287391663, + "learning_rate": 0.0008464040630520451, + "loss": 3.5508, + "step": 18085 + }, + { + "epoch": 1.2291072156543008, + "grad_norm": 0.9411778450012207, + "learning_rate": 0.0008463615980432124, + "loss": 3.6773, + "step": 18090 + }, + { + "epoch": 1.2294469357249627, + "grad_norm": 0.8735380172729492, + "learning_rate": 0.0008463191330343797, + "loss": 3.6757, + "step": 18095 + }, + { + "epoch": 1.2297866557956243, + "grad_norm": 0.7130782604217529, + "learning_rate": 0.000846276668025547, + "loss": 3.8401, + "step": 18100 + }, + { + "epoch": 1.2301263758662861, + "grad_norm": 0.9153977632522583, + "learning_rate": 0.0008462342030167143, + "loss": 3.7362, + "step": 18105 + }, + { + "epoch": 1.230466095936948, + "grad_norm": 0.7222729921340942, + "learning_rate": 0.0008461917380078815, + "loss": 3.606, + "step": 18110 + }, + { + "epoch": 1.2308058160076096, + "grad_norm": 0.7029339075088501, + "learning_rate": 0.0008461492729990487, + "loss": 3.7076, + "step": 18115 + }, + { + "epoch": 1.2311455360782715, + "grad_norm": 0.8207064270973206, + "learning_rate": 0.0008461068079902161, + "loss": 3.8793, + "step": 18120 + }, + { + "epoch": 1.2314852561489333, + "grad_norm": 0.8649730682373047, + "learning_rate": 0.0008460643429813833, + "loss": 3.4853, + "step": 18125 + }, + { + "epoch": 1.231824976219595, + "grad_norm": 0.9953786730766296, + "learning_rate": 0.0008460218779725506, + "loss": 3.5368, + "step": 18130 + }, + { + "epoch": 1.2321646962902568, + "grad_norm": 0.745816171169281, + "learning_rate": 0.000845979412963718, + "loss": 3.6434, + "step": 18135 + }, + { + "epoch": 1.2325044163609187, + "grad_norm": 0.8291403651237488, + "learning_rate": 0.0008459369479548852, + "loss": 3.4113, + "step": 18140 + }, + { + "epoch": 1.2328441364315803, + "grad_norm": 1.1412771940231323, + "learning_rate": 0.0008458944829460524, + "loss": 3.5915, + "step": 18145 + }, + { + "epoch": 1.2331838565022422, + "grad_norm": 0.8733716011047363, + "learning_rate": 0.0008458520179372198, + "loss": 3.6662, + "step": 18150 + }, + { + "epoch": 1.233523576572904, + "grad_norm": 0.7471228837966919, + "learning_rate": 0.000845809552928387, + "loss": 3.483, + "step": 18155 + }, + { + "epoch": 1.2338632966435656, + "grad_norm": 1.6504098176956177, + "learning_rate": 0.0008457670879195542, + "loss": 3.6134, + "step": 18160 + }, + { + "epoch": 1.2342030167142275, + "grad_norm": 1.4210253953933716, + "learning_rate": 0.0008457246229107217, + "loss": 3.3989, + "step": 18165 + }, + { + "epoch": 1.2345427367848893, + "grad_norm": 0.9557244181632996, + "learning_rate": 0.0008456821579018889, + "loss": 3.5461, + "step": 18170 + }, + { + "epoch": 1.234882456855551, + "grad_norm": 0.881510317325592, + "learning_rate": 0.0008456396928930561, + "loss": 3.6149, + "step": 18175 + }, + { + "epoch": 1.2352221769262128, + "grad_norm": 0.9930477142333984, + "learning_rate": 0.0008455972278842234, + "loss": 3.6329, + "step": 18180 + }, + { + "epoch": 1.2355618969968747, + "grad_norm": 0.9029057621955872, + "learning_rate": 0.0008455547628753907, + "loss": 3.6404, + "step": 18185 + }, + { + "epoch": 1.2359016170675363, + "grad_norm": 0.7720837593078613, + "learning_rate": 0.0008455122978665579, + "loss": 3.511, + "step": 18190 + }, + { + "epoch": 1.2362413371381982, + "grad_norm": 0.9271489381790161, + "learning_rate": 0.0008454698328577252, + "loss": 3.6233, + "step": 18195 + }, + { + "epoch": 1.23658105720886, + "grad_norm": 0.7605803608894348, + "learning_rate": 0.0008454273678488926, + "loss": 3.6811, + "step": 18200 + }, + { + "epoch": 1.2369207772795217, + "grad_norm": 0.8124759197235107, + "learning_rate": 0.0008453849028400598, + "loss": 3.4208, + "step": 18205 + }, + { + "epoch": 1.2372604973501835, + "grad_norm": 0.9307723045349121, + "learning_rate": 0.0008453424378312271, + "loss": 3.59, + "step": 18210 + }, + { + "epoch": 1.2376002174208451, + "grad_norm": 0.7233723402023315, + "learning_rate": 0.0008452999728223943, + "loss": 3.5635, + "step": 18215 + }, + { + "epoch": 1.237939937491507, + "grad_norm": 0.7466638088226318, + "learning_rate": 0.0008452575078135616, + "loss": 3.4461, + "step": 18220 + }, + { + "epoch": 1.2382796575621688, + "grad_norm": 0.8962716460227966, + "learning_rate": 0.0008452150428047289, + "loss": 3.8297, + "step": 18225 + }, + { + "epoch": 1.2386193776328305, + "grad_norm": 1.0340651273727417, + "learning_rate": 0.0008451725777958961, + "loss": 3.6565, + "step": 18230 + }, + { + "epoch": 1.2389590977034923, + "grad_norm": 0.7753156423568726, + "learning_rate": 0.0008451301127870636, + "loss": 3.7198, + "step": 18235 + }, + { + "epoch": 1.2392988177741542, + "grad_norm": 0.6185409426689148, + "learning_rate": 0.0008450876477782308, + "loss": 3.5466, + "step": 18240 + }, + { + "epoch": 1.2396385378448158, + "grad_norm": 0.9427835941314697, + "learning_rate": 0.000845045182769398, + "loss": 3.5347, + "step": 18245 + }, + { + "epoch": 1.2399782579154777, + "grad_norm": 0.8883402347564697, + "learning_rate": 0.0008450027177605654, + "loss": 3.6476, + "step": 18250 + }, + { + "epoch": 1.2403179779861393, + "grad_norm": 0.9068079590797424, + "learning_rate": 0.0008449602527517326, + "loss": 3.634, + "step": 18255 + }, + { + "epoch": 1.2406576980568012, + "grad_norm": 0.8822120428085327, + "learning_rate": 0.0008449177877428998, + "loss": 3.4382, + "step": 18260 + }, + { + "epoch": 1.240997418127463, + "grad_norm": 0.8714054226875305, + "learning_rate": 0.0008448753227340671, + "loss": 3.8631, + "step": 18265 + }, + { + "epoch": 1.2413371381981246, + "grad_norm": 0.8798202276229858, + "learning_rate": 0.0008448328577252345, + "loss": 3.3534, + "step": 18270 + }, + { + "epoch": 1.2416768582687865, + "grad_norm": 0.9595583081245422, + "learning_rate": 0.0008447903927164017, + "loss": 3.506, + "step": 18275 + }, + { + "epoch": 1.2420165783394483, + "grad_norm": 0.7908827662467957, + "learning_rate": 0.000844747927707569, + "loss": 3.7485, + "step": 18280 + }, + { + "epoch": 1.24235629841011, + "grad_norm": 0.8436384201049805, + "learning_rate": 0.0008447054626987363, + "loss": 3.9602, + "step": 18285 + }, + { + "epoch": 1.2426960184807718, + "grad_norm": 0.9083687663078308, + "learning_rate": 0.0008446629976899035, + "loss": 3.6702, + "step": 18290 + }, + { + "epoch": 1.2430357385514337, + "grad_norm": 0.7771982550621033, + "learning_rate": 0.0008446205326810708, + "loss": 3.5023, + "step": 18295 + }, + { + "epoch": 1.2433754586220953, + "grad_norm": 0.8990803360939026, + "learning_rate": 0.0008445780676722381, + "loss": 3.7242, + "step": 18300 + }, + { + "epoch": 1.2437151786927572, + "grad_norm": 0.7780101299285889, + "learning_rate": 0.0008445356026634054, + "loss": 3.7996, + "step": 18305 + }, + { + "epoch": 1.244054898763419, + "grad_norm": 0.9294916987419128, + "learning_rate": 0.0008444931376545727, + "loss": 3.7713, + "step": 18310 + }, + { + "epoch": 1.2443946188340806, + "grad_norm": 0.7924439311027527, + "learning_rate": 0.00084445067264574, + "loss": 3.8392, + "step": 18315 + }, + { + "epoch": 1.2447343389047425, + "grad_norm": 1.612853765487671, + "learning_rate": 0.0008444082076369072, + "loss": 3.6889, + "step": 18320 + }, + { + "epoch": 1.2450740589754044, + "grad_norm": 0.7830581068992615, + "learning_rate": 0.0008443657426280745, + "loss": 3.4884, + "step": 18325 + }, + { + "epoch": 1.245413779046066, + "grad_norm": 0.7962588667869568, + "learning_rate": 0.0008443232776192417, + "loss": 3.4492, + "step": 18330 + }, + { + "epoch": 1.2457534991167278, + "grad_norm": 1.3123995065689087, + "learning_rate": 0.000844280812610409, + "loss": 3.5741, + "step": 18335 + }, + { + "epoch": 1.2460932191873897, + "grad_norm": 0.8696267008781433, + "learning_rate": 0.0008442383476015764, + "loss": 3.5459, + "step": 18340 + }, + { + "epoch": 1.2464329392580513, + "grad_norm": 0.7638306617736816, + "learning_rate": 0.0008441958825927436, + "loss": 3.594, + "step": 18345 + }, + { + "epoch": 1.2467726593287132, + "grad_norm": 0.8447389006614685, + "learning_rate": 0.0008441534175839109, + "loss": 3.4698, + "step": 18350 + }, + { + "epoch": 1.247112379399375, + "grad_norm": 0.716288149356842, + "learning_rate": 0.0008441109525750782, + "loss": 3.8621, + "step": 18355 + }, + { + "epoch": 1.2474520994700367, + "grad_norm": 0.7221596837043762, + "learning_rate": 0.0008440684875662454, + "loss": 3.7811, + "step": 18360 + }, + { + "epoch": 1.2477918195406985, + "grad_norm": 0.6570661664009094, + "learning_rate": 0.0008440260225574126, + "loss": 3.6572, + "step": 18365 + }, + { + "epoch": 1.2481315396113604, + "grad_norm": 0.8833088874816895, + "learning_rate": 0.00084398355754858, + "loss": 3.4318, + "step": 18370 + }, + { + "epoch": 1.248471259682022, + "grad_norm": 2.111253023147583, + "learning_rate": 0.0008439410925397473, + "loss": 3.7862, + "step": 18375 + }, + { + "epoch": 1.2488109797526838, + "grad_norm": 1.6731688976287842, + "learning_rate": 0.0008438986275309145, + "loss": 3.3248, + "step": 18380 + }, + { + "epoch": 1.2491506998233455, + "grad_norm": 0.6982871890068054, + "learning_rate": 0.0008438561625220819, + "loss": 3.8462, + "step": 18385 + }, + { + "epoch": 1.2494904198940073, + "grad_norm": 0.8089686632156372, + "learning_rate": 0.0008438136975132491, + "loss": 3.4609, + "step": 18390 + }, + { + "epoch": 1.2498301399646692, + "grad_norm": 0.7364463210105896, + "learning_rate": 0.0008437712325044163, + "loss": 3.5474, + "step": 18395 + }, + { + "epoch": 1.2501698600353308, + "grad_norm": 0.6618789434432983, + "learning_rate": 0.0008437287674955837, + "loss": 3.6525, + "step": 18400 + }, + { + "epoch": 1.2505095801059927, + "grad_norm": 1.1544857025146484, + "learning_rate": 0.0008436863024867509, + "loss": 3.7402, + "step": 18405 + }, + { + "epoch": 1.2508493001766543, + "grad_norm": 0.6474128365516663, + "learning_rate": 0.0008436438374779182, + "loss": 3.724, + "step": 18410 + }, + { + "epoch": 1.2511890202473162, + "grad_norm": 0.7408311367034912, + "learning_rate": 0.0008436013724690856, + "loss": 3.3269, + "step": 18415 + }, + { + "epoch": 1.251528740317978, + "grad_norm": 0.9004170894622803, + "learning_rate": 0.0008435589074602528, + "loss": 3.4703, + "step": 18420 + }, + { + "epoch": 1.2518684603886396, + "grad_norm": 1.001169204711914, + "learning_rate": 0.00084351644245142, + "loss": 3.6058, + "step": 18425 + }, + { + "epoch": 1.2522081804593015, + "grad_norm": 1.0033454895019531, + "learning_rate": 0.0008434739774425873, + "loss": 3.3701, + "step": 18430 + }, + { + "epoch": 1.2525479005299633, + "grad_norm": 1.0031142234802246, + "learning_rate": 0.0008434315124337546, + "loss": 3.7816, + "step": 18435 + }, + { + "epoch": 1.252887620600625, + "grad_norm": 0.8103834986686707, + "learning_rate": 0.0008433890474249218, + "loss": 3.5997, + "step": 18440 + }, + { + "epoch": 1.2532273406712868, + "grad_norm": 0.870478093624115, + "learning_rate": 0.0008433465824160892, + "loss": 3.515, + "step": 18445 + }, + { + "epoch": 1.2535670607419487, + "grad_norm": 0.6764253973960876, + "learning_rate": 0.0008433041174072565, + "loss": 3.79, + "step": 18450 + }, + { + "epoch": 1.2539067808126103, + "grad_norm": 0.7315080761909485, + "learning_rate": 0.0008432616523984237, + "loss": 3.4807, + "step": 18455 + }, + { + "epoch": 1.2542465008832722, + "grad_norm": 1.4298170804977417, + "learning_rate": 0.000843219187389591, + "loss": 3.8018, + "step": 18460 + }, + { + "epoch": 1.254586220953934, + "grad_norm": 1.0117830038070679, + "learning_rate": 0.0008431767223807582, + "loss": 3.6612, + "step": 18465 + }, + { + "epoch": 1.2549259410245956, + "grad_norm": 1.1819978952407837, + "learning_rate": 0.0008431342573719255, + "loss": 3.3462, + "step": 18470 + }, + { + "epoch": 1.2552656610952575, + "grad_norm": 0.9841753840446472, + "learning_rate": 0.0008430917923630928, + "loss": 3.6861, + "step": 18475 + }, + { + "epoch": 1.2556053811659194, + "grad_norm": 1.1169837713241577, + "learning_rate": 0.0008430493273542601, + "loss": 3.5949, + "step": 18480 + }, + { + "epoch": 1.255945101236581, + "grad_norm": 0.8819544911384583, + "learning_rate": 0.0008430068623454274, + "loss": 3.4074, + "step": 18485 + }, + { + "epoch": 1.2562848213072428, + "grad_norm": 1.1495901346206665, + "learning_rate": 0.0008429643973365947, + "loss": 3.6429, + "step": 18490 + }, + { + "epoch": 1.2566245413779047, + "grad_norm": 1.0049405097961426, + "learning_rate": 0.0008429219323277619, + "loss": 3.6618, + "step": 18495 + }, + { + "epoch": 1.2569642614485663, + "grad_norm": 0.7330548763275146, + "learning_rate": 0.0008428794673189292, + "loss": 3.4593, + "step": 18500 + }, + { + "epoch": 1.2573039815192282, + "grad_norm": 1.1443344354629517, + "learning_rate": 0.0008428370023100965, + "loss": 3.6317, + "step": 18505 + }, + { + "epoch": 1.25764370158989, + "grad_norm": 0.867304801940918, + "learning_rate": 0.0008427945373012637, + "loss": 3.6558, + "step": 18510 + }, + { + "epoch": 1.2579834216605517, + "grad_norm": 0.876792848110199, + "learning_rate": 0.000842752072292431, + "loss": 3.8291, + "step": 18515 + }, + { + "epoch": 1.2583231417312135, + "grad_norm": 1.71896493434906, + "learning_rate": 0.0008427096072835984, + "loss": 3.6977, + "step": 18520 + }, + { + "epoch": 1.2586628618018754, + "grad_norm": 0.7942525148391724, + "learning_rate": 0.0008426671422747656, + "loss": 3.8734, + "step": 18525 + }, + { + "epoch": 1.259002581872537, + "grad_norm": 0.6905406713485718, + "learning_rate": 0.0008426246772659328, + "loss": 3.6691, + "step": 18530 + }, + { + "epoch": 1.2593423019431988, + "grad_norm": 0.7500548958778381, + "learning_rate": 0.0008425822122571002, + "loss": 3.9068, + "step": 18535 + }, + { + "epoch": 1.2596820220138607, + "grad_norm": 0.822364866733551, + "learning_rate": 0.0008425397472482674, + "loss": 3.611, + "step": 18540 + }, + { + "epoch": 1.2600217420845223, + "grad_norm": 0.8194557428359985, + "learning_rate": 0.0008424972822394346, + "loss": 3.7225, + "step": 18545 + }, + { + "epoch": 1.2603614621551842, + "grad_norm": 0.9240332245826721, + "learning_rate": 0.0008424548172306021, + "loss": 3.8984, + "step": 18550 + }, + { + "epoch": 1.260701182225846, + "grad_norm": 0.9217618703842163, + "learning_rate": 0.0008424123522217693, + "loss": 3.4859, + "step": 18555 + }, + { + "epoch": 1.2610409022965077, + "grad_norm": 0.6778313517570496, + "learning_rate": 0.0008423698872129365, + "loss": 3.9416, + "step": 18560 + }, + { + "epoch": 1.2613806223671695, + "grad_norm": 0.8785179257392883, + "learning_rate": 0.0008423274222041038, + "loss": 3.6826, + "step": 18565 + }, + { + "epoch": 1.2617203424378312, + "grad_norm": 0.8147216439247131, + "learning_rate": 0.0008422849571952711, + "loss": 3.7592, + "step": 18570 + }, + { + "epoch": 1.262060062508493, + "grad_norm": 0.8724662065505981, + "learning_rate": 0.0008422424921864384, + "loss": 3.4872, + "step": 18575 + }, + { + "epoch": 1.2623997825791546, + "grad_norm": 0.8504079580307007, + "learning_rate": 0.0008422000271776057, + "loss": 3.4962, + "step": 18580 + }, + { + "epoch": 1.2627395026498165, + "grad_norm": 1.1918799877166748, + "learning_rate": 0.000842157562168773, + "loss": 3.7573, + "step": 18585 + }, + { + "epoch": 1.2630792227204783, + "grad_norm": 0.9112948775291443, + "learning_rate": 0.0008421150971599403, + "loss": 3.7774, + "step": 18590 + }, + { + "epoch": 1.26341894279114, + "grad_norm": 0.83143150806427, + "learning_rate": 0.0008420726321511075, + "loss": 3.5513, + "step": 18595 + }, + { + "epoch": 1.2637586628618018, + "grad_norm": 0.8087087273597717, + "learning_rate": 0.0008420301671422748, + "loss": 3.745, + "step": 18600 + }, + { + "epoch": 1.2640983829324637, + "grad_norm": 0.8781185150146484, + "learning_rate": 0.0008419877021334421, + "loss": 3.507, + "step": 18605 + }, + { + "epoch": 1.2644381030031253, + "grad_norm": 0.6325780749320984, + "learning_rate": 0.0008419452371246093, + "loss": 3.3865, + "step": 18610 + }, + { + "epoch": 1.2647778230737872, + "grad_norm": 0.8836072683334351, + "learning_rate": 0.0008419027721157766, + "loss": 3.7734, + "step": 18615 + }, + { + "epoch": 1.265117543144449, + "grad_norm": 0.8629582524299622, + "learning_rate": 0.000841860307106944, + "loss": 3.7886, + "step": 18620 + }, + { + "epoch": 1.2654572632151107, + "grad_norm": 0.8805978298187256, + "learning_rate": 0.0008418178420981112, + "loss": 3.4427, + "step": 18625 + }, + { + "epoch": 1.2657969832857725, + "grad_norm": 0.9512669444084167, + "learning_rate": 0.0008417753770892784, + "loss": 3.3798, + "step": 18630 + }, + { + "epoch": 1.2661367033564344, + "grad_norm": 0.8685581684112549, + "learning_rate": 0.0008417329120804458, + "loss": 3.8713, + "step": 18635 + }, + { + "epoch": 1.266476423427096, + "grad_norm": 0.7789969444274902, + "learning_rate": 0.000841690447071613, + "loss": 3.3721, + "step": 18640 + }, + { + "epoch": 1.2668161434977578, + "grad_norm": 0.7959485054016113, + "learning_rate": 0.0008416479820627802, + "loss": 3.3812, + "step": 18645 + }, + { + "epoch": 1.2671558635684197, + "grad_norm": 0.9449043869972229, + "learning_rate": 0.0008416055170539477, + "loss": 3.6233, + "step": 18650 + }, + { + "epoch": 1.2674955836390813, + "grad_norm": 0.668481171131134, + "learning_rate": 0.0008415630520451149, + "loss": 3.7671, + "step": 18655 + }, + { + "epoch": 1.2678353037097432, + "grad_norm": 0.839979887008667, + "learning_rate": 0.0008415205870362821, + "loss": 3.7694, + "step": 18660 + }, + { + "epoch": 1.268175023780405, + "grad_norm": 0.9389800429344177, + "learning_rate": 0.0008414781220274494, + "loss": 3.6999, + "step": 18665 + }, + { + "epoch": 1.2685147438510667, + "grad_norm": 1.1306737661361694, + "learning_rate": 0.0008414356570186167, + "loss": 3.5589, + "step": 18670 + }, + { + "epoch": 1.2688544639217285, + "grad_norm": 0.771314799785614, + "learning_rate": 0.0008413931920097839, + "loss": 3.4752, + "step": 18675 + }, + { + "epoch": 1.2691941839923904, + "grad_norm": 0.9299555420875549, + "learning_rate": 0.0008413507270009512, + "loss": 3.5053, + "step": 18680 + }, + { + "epoch": 1.269533904063052, + "grad_norm": 0.8187107443809509, + "learning_rate": 0.0008413082619921186, + "loss": 3.8558, + "step": 18685 + }, + { + "epoch": 1.2698736241337139, + "grad_norm": 1.8932468891143799, + "learning_rate": 0.0008412657969832858, + "loss": 3.8907, + "step": 18690 + }, + { + "epoch": 1.2702133442043757, + "grad_norm": 0.9495120048522949, + "learning_rate": 0.0008412233319744531, + "loss": 3.6481, + "step": 18695 + }, + { + "epoch": 1.2705530642750373, + "grad_norm": 0.8422479033470154, + "learning_rate": 0.0008411808669656204, + "loss": 3.8434, + "step": 18700 + }, + { + "epoch": 1.2708927843456992, + "grad_norm": 0.6998239755630493, + "learning_rate": 0.0008411384019567876, + "loss": 3.5922, + "step": 18705 + }, + { + "epoch": 1.271232504416361, + "grad_norm": 0.7931044697761536, + "learning_rate": 0.0008410959369479549, + "loss": 3.6347, + "step": 18710 + }, + { + "epoch": 1.2715722244870227, + "grad_norm": 0.98618483543396, + "learning_rate": 0.0008410534719391221, + "loss": 3.483, + "step": 18715 + }, + { + "epoch": 1.2719119445576845, + "grad_norm": 1.4689273834228516, + "learning_rate": 0.0008410110069302895, + "loss": 3.4692, + "step": 18720 + }, + { + "epoch": 1.2722516646283464, + "grad_norm": 0.7294166684150696, + "learning_rate": 0.0008409685419214568, + "loss": 3.5618, + "step": 18725 + }, + { + "epoch": 1.272591384699008, + "grad_norm": 0.986870527267456, + "learning_rate": 0.000840926076912624, + "loss": 3.4794, + "step": 18730 + }, + { + "epoch": 1.2729311047696699, + "grad_norm": 1.0757185220718384, + "learning_rate": 0.0008408836119037913, + "loss": 3.5533, + "step": 18735 + }, + { + "epoch": 1.2732708248403315, + "grad_norm": 0.7239949703216553, + "learning_rate": 0.0008408411468949586, + "loss": 3.3803, + "step": 18740 + }, + { + "epoch": 1.2736105449109933, + "grad_norm": 0.9407927393913269, + "learning_rate": 0.0008407986818861258, + "loss": 3.5135, + "step": 18745 + }, + { + "epoch": 1.273950264981655, + "grad_norm": 0.9007251262664795, + "learning_rate": 0.000840756216877293, + "loss": 3.6895, + "step": 18750 + }, + { + "epoch": 1.2742899850523168, + "grad_norm": 1.447066068649292, + "learning_rate": 0.0008407137518684605, + "loss": 3.6948, + "step": 18755 + }, + { + "epoch": 1.2746297051229787, + "grad_norm": 0.7781888246536255, + "learning_rate": 0.0008406712868596277, + "loss": 3.6503, + "step": 18760 + }, + { + "epoch": 1.2749694251936403, + "grad_norm": 1.0265910625457764, + "learning_rate": 0.0008406288218507949, + "loss": 3.4891, + "step": 18765 + }, + { + "epoch": 1.2753091452643022, + "grad_norm": 0.7800175547599792, + "learning_rate": 0.0008405863568419623, + "loss": 3.4718, + "step": 18770 + }, + { + "epoch": 1.275648865334964, + "grad_norm": 0.943182647228241, + "learning_rate": 0.0008405438918331295, + "loss": 3.6423, + "step": 18775 + }, + { + "epoch": 1.2759885854056257, + "grad_norm": 0.8891888856887817, + "learning_rate": 0.0008405014268242967, + "loss": 3.8814, + "step": 18780 + }, + { + "epoch": 1.2763283054762875, + "grad_norm": 0.9731910228729248, + "learning_rate": 0.0008404589618154641, + "loss": 3.7375, + "step": 18785 + }, + { + "epoch": 1.2766680255469494, + "grad_norm": 0.8911810517311096, + "learning_rate": 0.0008404164968066314, + "loss": 3.9072, + "step": 18790 + }, + { + "epoch": 1.277007745617611, + "grad_norm": 0.6762071847915649, + "learning_rate": 0.0008403740317977986, + "loss": 3.671, + "step": 18795 + }, + { + "epoch": 1.2773474656882728, + "grad_norm": 0.9559621810913086, + "learning_rate": 0.000840331566788966, + "loss": 3.5413, + "step": 18800 + }, + { + "epoch": 1.2776871857589347, + "grad_norm": 0.7364009022712708, + "learning_rate": 0.0008402891017801332, + "loss": 3.6753, + "step": 18805 + }, + { + "epoch": 1.2780269058295963, + "grad_norm": 0.7066251039505005, + "learning_rate": 0.0008402466367713004, + "loss": 3.5796, + "step": 18810 + }, + { + "epoch": 1.2783666259002582, + "grad_norm": 0.8029266595840454, + "learning_rate": 0.0008402041717624677, + "loss": 3.6567, + "step": 18815 + }, + { + "epoch": 1.27870634597092, + "grad_norm": 0.8868125677108765, + "learning_rate": 0.000840161706753635, + "loss": 3.7226, + "step": 18820 + }, + { + "epoch": 1.2790460660415817, + "grad_norm": 0.7348118424415588, + "learning_rate": 0.0008401192417448023, + "loss": 3.6647, + "step": 18825 + }, + { + "epoch": 1.2793857861122435, + "grad_norm": 1.0088237524032593, + "learning_rate": 0.0008400767767359696, + "loss": 3.5265, + "step": 18830 + }, + { + "epoch": 1.2797255061829054, + "grad_norm": 1.6056684255599976, + "learning_rate": 0.0008400343117271369, + "loss": 3.8512, + "step": 18835 + }, + { + "epoch": 1.280065226253567, + "grad_norm": 0.8223760724067688, + "learning_rate": 0.0008399918467183041, + "loss": 3.7661, + "step": 18840 + }, + { + "epoch": 1.2804049463242289, + "grad_norm": 0.7191720008850098, + "learning_rate": 0.0008399493817094714, + "loss": 3.5785, + "step": 18845 + }, + { + "epoch": 1.2807446663948907, + "grad_norm": 1.026749610900879, + "learning_rate": 0.0008399069167006386, + "loss": 3.5328, + "step": 18850 + }, + { + "epoch": 1.2810843864655523, + "grad_norm": 0.6405095458030701, + "learning_rate": 0.0008398644516918059, + "loss": 3.6255, + "step": 18855 + }, + { + "epoch": 1.2814241065362142, + "grad_norm": 0.9298734068870544, + "learning_rate": 0.0008398219866829733, + "loss": 3.6699, + "step": 18860 + }, + { + "epoch": 1.281763826606876, + "grad_norm": 0.7743015885353088, + "learning_rate": 0.0008397795216741405, + "loss": 3.5715, + "step": 18865 + }, + { + "epoch": 1.2821035466775377, + "grad_norm": 0.7801750898361206, + "learning_rate": 0.0008397370566653078, + "loss": 3.7601, + "step": 18870 + }, + { + "epoch": 1.2824432667481995, + "grad_norm": 0.7174571752548218, + "learning_rate": 0.0008396945916564751, + "loss": 3.6115, + "step": 18875 + }, + { + "epoch": 1.2827829868188614, + "grad_norm": 0.8859849572181702, + "learning_rate": 0.0008396521266476423, + "loss": 3.4274, + "step": 18880 + }, + { + "epoch": 1.283122706889523, + "grad_norm": 0.823289692401886, + "learning_rate": 0.0008396096616388096, + "loss": 3.7089, + "step": 18885 + }, + { + "epoch": 1.2834624269601849, + "grad_norm": 0.9599798917770386, + "learning_rate": 0.0008395671966299769, + "loss": 3.3319, + "step": 18890 + }, + { + "epoch": 1.2838021470308467, + "grad_norm": 0.7811353802680969, + "learning_rate": 0.0008395247316211442, + "loss": 3.6249, + "step": 18895 + }, + { + "epoch": 1.2841418671015083, + "grad_norm": 0.8638587594032288, + "learning_rate": 0.0008394822666123114, + "loss": 3.575, + "step": 18900 + }, + { + "epoch": 1.2844815871721702, + "grad_norm": 1.471102237701416, + "learning_rate": 0.0008394398016034788, + "loss": 3.583, + "step": 18905 + }, + { + "epoch": 1.2848213072428318, + "grad_norm": 0.6785420179367065, + "learning_rate": 0.000839397336594646, + "loss": 3.8226, + "step": 18910 + }, + { + "epoch": 1.2851610273134937, + "grad_norm": 0.8843326568603516, + "learning_rate": 0.0008393548715858133, + "loss": 3.1594, + "step": 18915 + }, + { + "epoch": 1.2855007473841553, + "grad_norm": 0.7196475863456726, + "learning_rate": 0.0008393124065769806, + "loss": 3.5177, + "step": 18920 + }, + { + "epoch": 1.2858404674548172, + "grad_norm": 0.9442929029464722, + "learning_rate": 0.0008392699415681478, + "loss": 3.5674, + "step": 18925 + }, + { + "epoch": 1.286180187525479, + "grad_norm": 1.0013362169265747, + "learning_rate": 0.0008392274765593152, + "loss": 3.5787, + "step": 18930 + }, + { + "epoch": 1.2865199075961407, + "grad_norm": 0.780559778213501, + "learning_rate": 0.0008391850115504825, + "loss": 3.4753, + "step": 18935 + }, + { + "epoch": 1.2868596276668025, + "grad_norm": 1.0348625183105469, + "learning_rate": 0.0008391425465416497, + "loss": 3.7592, + "step": 18940 + }, + { + "epoch": 1.2871993477374644, + "grad_norm": 0.8198745846748352, + "learning_rate": 0.000839100081532817, + "loss": 3.5791, + "step": 18945 + }, + { + "epoch": 1.287539067808126, + "grad_norm": 0.8706223368644714, + "learning_rate": 0.0008390576165239842, + "loss": 3.8354, + "step": 18950 + }, + { + "epoch": 1.2878787878787878, + "grad_norm": 0.6900718212127686, + "learning_rate": 0.0008390151515151515, + "loss": 3.7682, + "step": 18955 + }, + { + "epoch": 1.2882185079494497, + "grad_norm": 0.99193274974823, + "learning_rate": 0.0008389726865063188, + "loss": 3.4471, + "step": 18960 + }, + { + "epoch": 1.2885582280201113, + "grad_norm": 0.8024405241012573, + "learning_rate": 0.0008389302214974861, + "loss": 3.4529, + "step": 18965 + }, + { + "epoch": 1.2888979480907732, + "grad_norm": 0.8801078796386719, + "learning_rate": 0.0008388877564886534, + "loss": 3.7201, + "step": 18970 + }, + { + "epoch": 1.289237668161435, + "grad_norm": 0.746083676815033, + "learning_rate": 0.0008388452914798207, + "loss": 3.6776, + "step": 18975 + }, + { + "epoch": 1.2895773882320967, + "grad_norm": 1.5363494157791138, + "learning_rate": 0.0008388028264709879, + "loss": 3.6466, + "step": 18980 + }, + { + "epoch": 1.2899171083027585, + "grad_norm": 0.6721611022949219, + "learning_rate": 0.0008387603614621552, + "loss": 3.669, + "step": 18985 + }, + { + "epoch": 1.2902568283734204, + "grad_norm": 0.720432460308075, + "learning_rate": 0.0008387178964533225, + "loss": 3.4773, + "step": 18990 + }, + { + "epoch": 1.290596548444082, + "grad_norm": 1.061953067779541, + "learning_rate": 0.0008386754314444897, + "loss": 3.997, + "step": 18995 + }, + { + "epoch": 1.2909362685147439, + "grad_norm": 0.761386513710022, + "learning_rate": 0.000838632966435657, + "loss": 3.496, + "step": 19000 + }, + { + "epoch": 1.2912759885854057, + "grad_norm": 0.8305133581161499, + "learning_rate": 0.0008385905014268244, + "loss": 3.5356, + "step": 19005 + }, + { + "epoch": 1.2916157086560673, + "grad_norm": 0.7632470726966858, + "learning_rate": 0.0008385480364179916, + "loss": 3.6356, + "step": 19010 + }, + { + "epoch": 1.2919554287267292, + "grad_norm": 0.9063440561294556, + "learning_rate": 0.0008385055714091588, + "loss": 3.6967, + "step": 19015 + }, + { + "epoch": 1.292295148797391, + "grad_norm": 0.9601851105690002, + "learning_rate": 0.0008384631064003262, + "loss": 3.7234, + "step": 19020 + }, + { + "epoch": 1.2926348688680527, + "grad_norm": 1.3528258800506592, + "learning_rate": 0.0008384206413914934, + "loss": 3.6734, + "step": 19025 + }, + { + "epoch": 1.2929745889387145, + "grad_norm": 0.9065192341804504, + "learning_rate": 0.0008383781763826606, + "loss": 3.7763, + "step": 19030 + }, + { + "epoch": 1.2933143090093764, + "grad_norm": 0.9659169912338257, + "learning_rate": 0.0008383357113738281, + "loss": 3.9401, + "step": 19035 + }, + { + "epoch": 1.293654029080038, + "grad_norm": 0.8508115410804749, + "learning_rate": 0.0008382932463649953, + "loss": 3.6517, + "step": 19040 + }, + { + "epoch": 1.2939937491506999, + "grad_norm": 0.7941755056381226, + "learning_rate": 0.0008382507813561625, + "loss": 3.803, + "step": 19045 + }, + { + "epoch": 1.2943334692213617, + "grad_norm": 0.7235051393508911, + "learning_rate": 0.0008382083163473299, + "loss": 3.3518, + "step": 19050 + }, + { + "epoch": 1.2946731892920234, + "grad_norm": 3.7901744842529297, + "learning_rate": 0.0008381658513384971, + "loss": 3.6621, + "step": 19055 + }, + { + "epoch": 1.2950129093626852, + "grad_norm": 0.7474851012229919, + "learning_rate": 0.0008381233863296643, + "loss": 3.7553, + "step": 19060 + }, + { + "epoch": 1.295352629433347, + "grad_norm": 0.7516352534294128, + "learning_rate": 0.0008380809213208316, + "loss": 3.4176, + "step": 19065 + }, + { + "epoch": 1.2956923495040087, + "grad_norm": 0.8967887163162231, + "learning_rate": 0.000838038456311999, + "loss": 3.499, + "step": 19070 + }, + { + "epoch": 1.2960320695746705, + "grad_norm": 1.0398764610290527, + "learning_rate": 0.0008379959913031662, + "loss": 3.5029, + "step": 19075 + }, + { + "epoch": 1.2963717896453322, + "grad_norm": 0.7808350920677185, + "learning_rate": 0.0008379535262943335, + "loss": 3.8588, + "step": 19080 + }, + { + "epoch": 1.296711509715994, + "grad_norm": 0.8476454019546509, + "learning_rate": 0.0008379110612855008, + "loss": 3.8217, + "step": 19085 + }, + { + "epoch": 1.2970512297866557, + "grad_norm": 3.1767690181732178, + "learning_rate": 0.000837868596276668, + "loss": 3.6801, + "step": 19090 + }, + { + "epoch": 1.2973909498573175, + "grad_norm": 0.7458839416503906, + "learning_rate": 0.0008378261312678353, + "loss": 3.4079, + "step": 19095 + }, + { + "epoch": 1.2977306699279794, + "grad_norm": 1.2557902336120605, + "learning_rate": 0.0008377836662590025, + "loss": 3.528, + "step": 19100 + }, + { + "epoch": 1.298070389998641, + "grad_norm": 0.8059072494506836, + "learning_rate": 0.0008377412012501699, + "loss": 3.5666, + "step": 19105 + }, + { + "epoch": 1.2984101100693028, + "grad_norm": 0.9414175152778625, + "learning_rate": 0.0008376987362413372, + "loss": 3.535, + "step": 19110 + }, + { + "epoch": 1.2987498301399647, + "grad_norm": 0.9976986646652222, + "learning_rate": 0.0008376562712325044, + "loss": 3.5132, + "step": 19115 + }, + { + "epoch": 1.2990895502106263, + "grad_norm": 0.7399609684944153, + "learning_rate": 0.0008376138062236717, + "loss": 3.702, + "step": 19120 + }, + { + "epoch": 1.2994292702812882, + "grad_norm": 0.746816098690033, + "learning_rate": 0.000837571341214839, + "loss": 3.6675, + "step": 19125 + }, + { + "epoch": 1.29976899035195, + "grad_norm": 0.7854049205780029, + "learning_rate": 0.0008375288762060062, + "loss": 3.6947, + "step": 19130 + }, + { + "epoch": 1.3001087104226117, + "grad_norm": 0.9822672009468079, + "learning_rate": 0.0008374864111971734, + "loss": 3.5687, + "step": 19135 + }, + { + "epoch": 1.3004484304932735, + "grad_norm": 0.8535969257354736, + "learning_rate": 0.0008374439461883409, + "loss": 3.4407, + "step": 19140 + }, + { + "epoch": 1.3007881505639354, + "grad_norm": 0.9285767674446106, + "learning_rate": 0.0008374014811795081, + "loss": 3.707, + "step": 19145 + }, + { + "epoch": 1.301127870634597, + "grad_norm": 0.7824862003326416, + "learning_rate": 0.0008373590161706753, + "loss": 3.7798, + "step": 19150 + }, + { + "epoch": 1.3014675907052589, + "grad_norm": 0.7672741413116455, + "learning_rate": 0.0008373165511618427, + "loss": 3.5057, + "step": 19155 + }, + { + "epoch": 1.3018073107759207, + "grad_norm": 0.8770267367362976, + "learning_rate": 0.0008372740861530099, + "loss": 3.6849, + "step": 19160 + }, + { + "epoch": 1.3021470308465823, + "grad_norm": 1.0106281042099, + "learning_rate": 0.0008372316211441771, + "loss": 3.5227, + "step": 19165 + }, + { + "epoch": 1.3024867509172442, + "grad_norm": 0.8849656581878662, + "learning_rate": 0.0008371891561353446, + "loss": 3.9317, + "step": 19170 + }, + { + "epoch": 1.302826470987906, + "grad_norm": 0.8273836970329285, + "learning_rate": 0.0008371466911265118, + "loss": 3.5142, + "step": 19175 + }, + { + "epoch": 1.3031661910585677, + "grad_norm": 1.0824638605117798, + "learning_rate": 0.000837104226117679, + "loss": 3.5279, + "step": 19180 + }, + { + "epoch": 1.3035059111292295, + "grad_norm": 0.7543104887008667, + "learning_rate": 0.0008370617611088464, + "loss": 3.5396, + "step": 19185 + }, + { + "epoch": 1.3038456311998914, + "grad_norm": 0.7501362562179565, + "learning_rate": 0.0008370192961000136, + "loss": 3.9378, + "step": 19190 + }, + { + "epoch": 1.304185351270553, + "grad_norm": 0.759842038154602, + "learning_rate": 0.0008369768310911808, + "loss": 3.5307, + "step": 19195 + }, + { + "epoch": 1.3045250713412149, + "grad_norm": 0.8994619846343994, + "learning_rate": 0.0008369343660823481, + "loss": 3.9167, + "step": 19200 + }, + { + "epoch": 1.3048647914118767, + "grad_norm": 1.0880717039108276, + "learning_rate": 0.0008368919010735155, + "loss": 3.6136, + "step": 19205 + }, + { + "epoch": 1.3052045114825384, + "grad_norm": 0.7082797884941101, + "learning_rate": 0.0008368494360646827, + "loss": 3.6299, + "step": 19210 + }, + { + "epoch": 1.3055442315532002, + "grad_norm": 0.7700816988945007, + "learning_rate": 0.00083680697105585, + "loss": 3.8674, + "step": 19215 + }, + { + "epoch": 1.305883951623862, + "grad_norm": 0.7759992480278015, + "learning_rate": 0.0008367645060470173, + "loss": 3.4494, + "step": 19220 + }, + { + "epoch": 1.3062236716945237, + "grad_norm": 0.8918359875679016, + "learning_rate": 0.0008367220410381845, + "loss": 3.7901, + "step": 19225 + }, + { + "epoch": 1.3065633917651855, + "grad_norm": 2.300729513168335, + "learning_rate": 0.0008366795760293518, + "loss": 3.6712, + "step": 19230 + }, + { + "epoch": 1.3069031118358474, + "grad_norm": 1.0091993808746338, + "learning_rate": 0.000836637111020519, + "loss": 3.3453, + "step": 19235 + }, + { + "epoch": 1.307242831906509, + "grad_norm": 0.7243125438690186, + "learning_rate": 0.0008365946460116864, + "loss": 3.5832, + "step": 19240 + }, + { + "epoch": 1.3075825519771709, + "grad_norm": 0.7430331707000732, + "learning_rate": 0.0008365521810028537, + "loss": 3.7651, + "step": 19245 + }, + { + "epoch": 1.3079222720478325, + "grad_norm": 1.9062978029251099, + "learning_rate": 0.0008365097159940209, + "loss": 3.8237, + "step": 19250 + }, + { + "epoch": 1.3082619921184944, + "grad_norm": 0.7068991661071777, + "learning_rate": 0.0008364672509851883, + "loss": 3.8731, + "step": 19255 + }, + { + "epoch": 1.308601712189156, + "grad_norm": 1.2697590589523315, + "learning_rate": 0.0008364247859763555, + "loss": 3.8133, + "step": 19260 + }, + { + "epoch": 1.3089414322598178, + "grad_norm": 0.8661418557167053, + "learning_rate": 0.0008363823209675227, + "loss": 3.1856, + "step": 19265 + }, + { + "epoch": 1.3092811523304797, + "grad_norm": 1.2155336141586304, + "learning_rate": 0.0008363398559586901, + "loss": 3.5142, + "step": 19270 + }, + { + "epoch": 1.3096208724011413, + "grad_norm": 0.5832532048225403, + "learning_rate": 0.0008362973909498574, + "loss": 3.5184, + "step": 19275 + }, + { + "epoch": 1.3099605924718032, + "grad_norm": 0.8530396819114685, + "learning_rate": 0.0008362549259410246, + "loss": 3.2229, + "step": 19280 + }, + { + "epoch": 1.310300312542465, + "grad_norm": 0.8273479342460632, + "learning_rate": 0.000836212460932192, + "loss": 3.656, + "step": 19285 + }, + { + "epoch": 1.3106400326131267, + "grad_norm": 1.3767008781433105, + "learning_rate": 0.0008361699959233592, + "loss": 3.6, + "step": 19290 + }, + { + "epoch": 1.3109797526837885, + "grad_norm": 0.7706455588340759, + "learning_rate": 0.0008361275309145264, + "loss": 3.6458, + "step": 19295 + }, + { + "epoch": 1.3113194727544504, + "grad_norm": 0.7214418649673462, + "learning_rate": 0.0008360850659056937, + "loss": 3.6441, + "step": 19300 + }, + { + "epoch": 1.311659192825112, + "grad_norm": 0.6593573689460754, + "learning_rate": 0.000836042600896861, + "loss": 3.7781, + "step": 19305 + }, + { + "epoch": 1.3119989128957739, + "grad_norm": 0.6926696300506592, + "learning_rate": 0.0008360001358880283, + "loss": 3.4172, + "step": 19310 + }, + { + "epoch": 1.3123386329664357, + "grad_norm": 0.9231415390968323, + "learning_rate": 0.0008359576708791956, + "loss": 3.4628, + "step": 19315 + }, + { + "epoch": 1.3126783530370973, + "grad_norm": 0.664509654045105, + "learning_rate": 0.0008359152058703629, + "loss": 3.527, + "step": 19320 + }, + { + "epoch": 1.3130180731077592, + "grad_norm": 0.7721145749092102, + "learning_rate": 0.0008358727408615301, + "loss": 3.6316, + "step": 19325 + }, + { + "epoch": 1.313357793178421, + "grad_norm": 1.0311793088912964, + "learning_rate": 0.0008358302758526974, + "loss": 3.6091, + "step": 19330 + }, + { + "epoch": 1.3136975132490827, + "grad_norm": 0.93877774477005, + "learning_rate": 0.0008357878108438647, + "loss": 3.4597, + "step": 19335 + }, + { + "epoch": 1.3140372333197445, + "grad_norm": 0.9329226613044739, + "learning_rate": 0.0008357453458350319, + "loss": 3.6775, + "step": 19340 + }, + { + "epoch": 1.3143769533904064, + "grad_norm": 0.9558454751968384, + "learning_rate": 0.0008357028808261993, + "loss": 3.7005, + "step": 19345 + }, + { + "epoch": 1.314716673461068, + "grad_norm": 1.8134329319000244, + "learning_rate": 0.0008356604158173665, + "loss": 3.5576, + "step": 19350 + }, + { + "epoch": 1.3150563935317299, + "grad_norm": 1.1518900394439697, + "learning_rate": 0.0008356179508085338, + "loss": 3.5975, + "step": 19355 + }, + { + "epoch": 1.3153961136023917, + "grad_norm": 1.0108559131622314, + "learning_rate": 0.0008355754857997011, + "loss": 3.5175, + "step": 19360 + }, + { + "epoch": 1.3157358336730534, + "grad_norm": 0.9091687798500061, + "learning_rate": 0.0008355330207908683, + "loss": 3.7377, + "step": 19365 + }, + { + "epoch": 1.3160755537437152, + "grad_norm": 0.7748539447784424, + "learning_rate": 0.0008354905557820356, + "loss": 3.7376, + "step": 19370 + }, + { + "epoch": 1.316415273814377, + "grad_norm": 0.7644347548484802, + "learning_rate": 0.0008354480907732029, + "loss": 3.9485, + "step": 19375 + }, + { + "epoch": 1.3167549938850387, + "grad_norm": 0.8928879499435425, + "learning_rate": 0.0008354056257643702, + "loss": 3.6535, + "step": 19380 + }, + { + "epoch": 1.3170947139557005, + "grad_norm": 0.7633859515190125, + "learning_rate": 0.0008353631607555375, + "loss": 3.7145, + "step": 19385 + }, + { + "epoch": 1.3174344340263624, + "grad_norm": 1.004683256149292, + "learning_rate": 0.0008353206957467048, + "loss": 3.677, + "step": 19390 + }, + { + "epoch": 1.317774154097024, + "grad_norm": 1.224363923072815, + "learning_rate": 0.000835278230737872, + "loss": 3.2911, + "step": 19395 + }, + { + "epoch": 1.3181138741676859, + "grad_norm": 0.7706454992294312, + "learning_rate": 0.0008352357657290392, + "loss": 3.8322, + "step": 19400 + }, + { + "epoch": 1.3184535942383477, + "grad_norm": 1.0382957458496094, + "learning_rate": 0.0008351933007202066, + "loss": 3.6499, + "step": 19405 + }, + { + "epoch": 1.3187933143090094, + "grad_norm": 0.9529634118080139, + "learning_rate": 0.0008351508357113738, + "loss": 3.6193, + "step": 19410 + }, + { + "epoch": 1.3191330343796712, + "grad_norm": 0.9824109673500061, + "learning_rate": 0.0008351083707025411, + "loss": 3.5582, + "step": 19415 + }, + { + "epoch": 1.3194727544503329, + "grad_norm": 0.7511589527130127, + "learning_rate": 0.0008350659056937085, + "loss": 3.7914, + "step": 19420 + }, + { + "epoch": 1.3198124745209947, + "grad_norm": 1.3449985980987549, + "learning_rate": 0.0008350234406848757, + "loss": 3.5377, + "step": 19425 + }, + { + "epoch": 1.3201521945916566, + "grad_norm": 0.7778506278991699, + "learning_rate": 0.0008349809756760429, + "loss": 3.7603, + "step": 19430 + }, + { + "epoch": 1.3204919146623182, + "grad_norm": 0.7010202407836914, + "learning_rate": 0.0008349385106672103, + "loss": 3.6515, + "step": 19435 + }, + { + "epoch": 1.32083163473298, + "grad_norm": 0.9390833973884583, + "learning_rate": 0.0008348960456583775, + "loss": 3.5299, + "step": 19440 + }, + { + "epoch": 1.3211713548036417, + "grad_norm": 0.8722630739212036, + "learning_rate": 0.0008348535806495447, + "loss": 3.6038, + "step": 19445 + }, + { + "epoch": 1.3215110748743035, + "grad_norm": 0.8852182030677795, + "learning_rate": 0.0008348111156407121, + "loss": 3.6416, + "step": 19450 + }, + { + "epoch": 1.3218507949449654, + "grad_norm": 0.8159305453300476, + "learning_rate": 0.0008347686506318794, + "loss": 3.8608, + "step": 19455 + }, + { + "epoch": 1.322190515015627, + "grad_norm": 0.8102395534515381, + "learning_rate": 0.0008347261856230466, + "loss": 3.8264, + "step": 19460 + }, + { + "epoch": 1.3225302350862889, + "grad_norm": 0.7249893546104431, + "learning_rate": 0.0008346837206142139, + "loss": 3.5293, + "step": 19465 + }, + { + "epoch": 1.3228699551569507, + "grad_norm": 1.4663506746292114, + "learning_rate": 0.0008346412556053812, + "loss": 3.8016, + "step": 19470 + }, + { + "epoch": 1.3232096752276123, + "grad_norm": 0.9181190133094788, + "learning_rate": 0.0008345987905965484, + "loss": 3.5921, + "step": 19475 + }, + { + "epoch": 1.3235493952982742, + "grad_norm": 3.3967654705047607, + "learning_rate": 0.0008345563255877157, + "loss": 3.5958, + "step": 19480 + }, + { + "epoch": 1.323889115368936, + "grad_norm": 0.9114540815353394, + "learning_rate": 0.0008345138605788831, + "loss": 3.6338, + "step": 19485 + }, + { + "epoch": 1.3242288354395977, + "grad_norm": 0.7916017770767212, + "learning_rate": 0.0008344713955700503, + "loss": 3.4627, + "step": 19490 + }, + { + "epoch": 1.3245685555102595, + "grad_norm": 0.7891761660575867, + "learning_rate": 0.0008344289305612176, + "loss": 3.6942, + "step": 19495 + }, + { + "epoch": 1.3249082755809214, + "grad_norm": 0.7993078231811523, + "learning_rate": 0.0008343864655523848, + "loss": 3.7123, + "step": 19500 + }, + { + "epoch": 1.325247995651583, + "grad_norm": 1.2489020824432373, + "learning_rate": 0.0008343440005435521, + "loss": 3.4263, + "step": 19505 + }, + { + "epoch": 1.3255877157222449, + "grad_norm": 1.1415542364120483, + "learning_rate": 0.0008343015355347194, + "loss": 3.5019, + "step": 19510 + }, + { + "epoch": 1.3259274357929067, + "grad_norm": 0.6854630708694458, + "learning_rate": 0.0008342590705258866, + "loss": 3.6906, + "step": 19515 + }, + { + "epoch": 1.3262671558635684, + "grad_norm": 0.7478453516960144, + "learning_rate": 0.000834216605517054, + "loss": 3.5565, + "step": 19520 + }, + { + "epoch": 1.3266068759342302, + "grad_norm": 0.7282377481460571, + "learning_rate": 0.0008341741405082213, + "loss": 3.5382, + "step": 19525 + }, + { + "epoch": 1.326946596004892, + "grad_norm": 0.9372228384017944, + "learning_rate": 0.0008341316754993885, + "loss": 3.4152, + "step": 19530 + }, + { + "epoch": 1.3272863160755537, + "grad_norm": 0.8004382252693176, + "learning_rate": 0.0008340892104905557, + "loss": 3.6709, + "step": 19535 + }, + { + "epoch": 1.3276260361462155, + "grad_norm": 0.9729716777801514, + "learning_rate": 0.0008340467454817231, + "loss": 3.6408, + "step": 19540 + }, + { + "epoch": 1.3279657562168774, + "grad_norm": 0.574894905090332, + "learning_rate": 0.0008340042804728903, + "loss": 3.5192, + "step": 19545 + }, + { + "epoch": 1.328305476287539, + "grad_norm": 0.7566072940826416, + "learning_rate": 0.0008339618154640575, + "loss": 3.4148, + "step": 19550 + }, + { + "epoch": 1.3286451963582009, + "grad_norm": 1.0383200645446777, + "learning_rate": 0.000833919350455225, + "loss": 3.7664, + "step": 19555 + }, + { + "epoch": 1.3289849164288627, + "grad_norm": 1.513988733291626, + "learning_rate": 0.0008338768854463922, + "loss": 3.7069, + "step": 19560 + }, + { + "epoch": 1.3293246364995244, + "grad_norm": 1.3090567588806152, + "learning_rate": 0.0008338344204375594, + "loss": 3.4956, + "step": 19565 + }, + { + "epoch": 1.3296643565701862, + "grad_norm": 0.7339676022529602, + "learning_rate": 0.0008337919554287268, + "loss": 3.5868, + "step": 19570 + }, + { + "epoch": 1.330004076640848, + "grad_norm": 0.8436015844345093, + "learning_rate": 0.000833749490419894, + "loss": 3.7834, + "step": 19575 + }, + { + "epoch": 1.3303437967115097, + "grad_norm": 0.7996444702148438, + "learning_rate": 0.0008337070254110612, + "loss": 3.5127, + "step": 19580 + }, + { + "epoch": 1.3306835167821716, + "grad_norm": 0.8504420518875122, + "learning_rate": 0.0008336645604022285, + "loss": 3.6038, + "step": 19585 + }, + { + "epoch": 1.3310232368528332, + "grad_norm": 0.7843599915504456, + "learning_rate": 0.0008336220953933959, + "loss": 3.5667, + "step": 19590 + }, + { + "epoch": 1.331362956923495, + "grad_norm": 0.7322075366973877, + "learning_rate": 0.0008335796303845632, + "loss": 3.6758, + "step": 19595 + }, + { + "epoch": 1.331702676994157, + "grad_norm": 0.8108183145523071, + "learning_rate": 0.0008335371653757304, + "loss": 3.6849, + "step": 19600 + }, + { + "epoch": 1.3320423970648185, + "grad_norm": 0.9602959752082825, + "learning_rate": 0.0008334947003668977, + "loss": 3.6558, + "step": 19605 + }, + { + "epoch": 1.3323821171354804, + "grad_norm": 0.8836127519607544, + "learning_rate": 0.000833452235358065, + "loss": 3.5397, + "step": 19610 + }, + { + "epoch": 1.332721837206142, + "grad_norm": 0.8938495516777039, + "learning_rate": 0.0008334097703492322, + "loss": 3.3683, + "step": 19615 + }, + { + "epoch": 1.3330615572768039, + "grad_norm": 0.8161758184432983, + "learning_rate": 0.0008333673053403995, + "loss": 3.8448, + "step": 19620 + }, + { + "epoch": 1.3334012773474657, + "grad_norm": 0.927852213382721, + "learning_rate": 0.0008333248403315669, + "loss": 3.778, + "step": 19625 + }, + { + "epoch": 1.3337409974181273, + "grad_norm": 1.045912742614746, + "learning_rate": 0.0008332823753227341, + "loss": 4.0483, + "step": 19630 + }, + { + "epoch": 1.3340807174887892, + "grad_norm": 0.9028273820877075, + "learning_rate": 0.0008332399103139013, + "loss": 3.6994, + "step": 19635 + }, + { + "epoch": 1.334420437559451, + "grad_norm": 0.8321359753608704, + "learning_rate": 0.0008331974453050687, + "loss": 3.5622, + "step": 19640 + }, + { + "epoch": 1.3347601576301127, + "grad_norm": 0.7420857548713684, + "learning_rate": 0.0008331549802962359, + "loss": 3.6311, + "step": 19645 + }, + { + "epoch": 1.3350998777007745, + "grad_norm": 0.9606713056564331, + "learning_rate": 0.0008331125152874031, + "loss": 3.3942, + "step": 19650 + }, + { + "epoch": 1.3354395977714364, + "grad_norm": 0.7483904957771301, + "learning_rate": 0.0008330700502785706, + "loss": 3.4491, + "step": 19655 + }, + { + "epoch": 1.335779317842098, + "grad_norm": 0.8392571806907654, + "learning_rate": 0.0008330275852697378, + "loss": 3.7821, + "step": 19660 + }, + { + "epoch": 1.3361190379127599, + "grad_norm": 1.0173481702804565, + "learning_rate": 0.000832985120260905, + "loss": 3.726, + "step": 19665 + }, + { + "epoch": 1.3364587579834217, + "grad_norm": 0.905332088470459, + "learning_rate": 0.0008329426552520724, + "loss": 3.7561, + "step": 19670 + }, + { + "epoch": 1.3367984780540834, + "grad_norm": 1.1367340087890625, + "learning_rate": 0.0008329001902432396, + "loss": 3.7775, + "step": 19675 + }, + { + "epoch": 1.3371381981247452, + "grad_norm": 1.0209132432937622, + "learning_rate": 0.0008328577252344068, + "loss": 3.7185, + "step": 19680 + }, + { + "epoch": 1.337477918195407, + "grad_norm": 0.7754608988761902, + "learning_rate": 0.0008328152602255741, + "loss": 3.8013, + "step": 19685 + }, + { + "epoch": 1.3378176382660687, + "grad_norm": 0.7911690473556519, + "learning_rate": 0.0008327727952167415, + "loss": 3.6321, + "step": 19690 + }, + { + "epoch": 1.3381573583367306, + "grad_norm": 0.6917429566383362, + "learning_rate": 0.0008327303302079087, + "loss": 3.7086, + "step": 19695 + }, + { + "epoch": 1.3384970784073924, + "grad_norm": 1.0019865036010742, + "learning_rate": 0.000832687865199076, + "loss": 3.4491, + "step": 19700 + }, + { + "epoch": 1.338836798478054, + "grad_norm": 0.7981930375099182, + "learning_rate": 0.0008326454001902433, + "loss": 3.703, + "step": 19705 + }, + { + "epoch": 1.3391765185487159, + "grad_norm": 1.3944613933563232, + "learning_rate": 0.0008326029351814105, + "loss": 3.5023, + "step": 19710 + }, + { + "epoch": 1.3395162386193777, + "grad_norm": 0.7388907074928284, + "learning_rate": 0.0008325604701725778, + "loss": 3.4653, + "step": 19715 + }, + { + "epoch": 1.3398559586900394, + "grad_norm": 1.0520060062408447, + "learning_rate": 0.0008325180051637451, + "loss": 3.6993, + "step": 19720 + }, + { + "epoch": 1.3401956787607012, + "grad_norm": 0.7282788753509521, + "learning_rate": 0.0008324755401549124, + "loss": 3.7358, + "step": 19725 + }, + { + "epoch": 1.340535398831363, + "grad_norm": 0.7743442058563232, + "learning_rate": 0.0008324330751460797, + "loss": 3.5194, + "step": 19730 + }, + { + "epoch": 1.3408751189020247, + "grad_norm": 0.771087110042572, + "learning_rate": 0.000832390610137247, + "loss": 3.6726, + "step": 19735 + }, + { + "epoch": 1.3412148389726866, + "grad_norm": 2.1107139587402344, + "learning_rate": 0.0008323481451284142, + "loss": 3.3392, + "step": 19740 + }, + { + "epoch": 1.3415545590433484, + "grad_norm": 0.7359451651573181, + "learning_rate": 0.0008323056801195815, + "loss": 3.7288, + "step": 19745 + }, + { + "epoch": 1.34189427911401, + "grad_norm": 1.036449909210205, + "learning_rate": 0.0008322632151107487, + "loss": 3.4754, + "step": 19750 + }, + { + "epoch": 1.342233999184672, + "grad_norm": 0.8496060371398926, + "learning_rate": 0.000832220750101916, + "loss": 3.7369, + "step": 19755 + }, + { + "epoch": 1.3425737192553335, + "grad_norm": 1.0397948026657104, + "learning_rate": 0.0008321782850930834, + "loss": 3.5993, + "step": 19760 + }, + { + "epoch": 1.3429134393259954, + "grad_norm": 1.02755606174469, + "learning_rate": 0.0008321358200842506, + "loss": 3.3733, + "step": 19765 + }, + { + "epoch": 1.3432531593966572, + "grad_norm": 0.9234209656715393, + "learning_rate": 0.0008320933550754179, + "loss": 3.4907, + "step": 19770 + }, + { + "epoch": 1.3435928794673189, + "grad_norm": 1.6648839712142944, + "learning_rate": 0.0008320508900665852, + "loss": 3.5315, + "step": 19775 + }, + { + "epoch": 1.3439325995379807, + "grad_norm": 0.7281484007835388, + "learning_rate": 0.0008320084250577524, + "loss": 3.6876, + "step": 19780 + }, + { + "epoch": 1.3442723196086424, + "grad_norm": 0.8084084391593933, + "learning_rate": 0.0008319659600489196, + "loss": 3.6864, + "step": 19785 + }, + { + "epoch": 1.3446120396793042, + "grad_norm": 0.6933193206787109, + "learning_rate": 0.000831923495040087, + "loss": 3.7692, + "step": 19790 + }, + { + "epoch": 1.344951759749966, + "grad_norm": 0.7642212510108948, + "learning_rate": 0.0008318810300312543, + "loss": 3.7358, + "step": 19795 + }, + { + "epoch": 1.3452914798206277, + "grad_norm": 1.1583991050720215, + "learning_rate": 0.0008318385650224215, + "loss": 3.5249, + "step": 19800 + }, + { + "epoch": 1.3456311998912895, + "grad_norm": 0.7731074690818787, + "learning_rate": 0.0008317961000135889, + "loss": 3.6161, + "step": 19805 + }, + { + "epoch": 1.3459709199619514, + "grad_norm": 0.7493581175804138, + "learning_rate": 0.0008317536350047561, + "loss": 3.4031, + "step": 19810 + }, + { + "epoch": 1.346310640032613, + "grad_norm": 0.9421755075454712, + "learning_rate": 0.0008317111699959233, + "loss": 3.6214, + "step": 19815 + }, + { + "epoch": 1.3466503601032749, + "grad_norm": 0.8453624844551086, + "learning_rate": 0.0008316687049870907, + "loss": 3.5466, + "step": 19820 + }, + { + "epoch": 1.3469900801739367, + "grad_norm": 0.7539472579956055, + "learning_rate": 0.0008316262399782579, + "loss": 3.5655, + "step": 19825 + }, + { + "epoch": 1.3473298002445984, + "grad_norm": 0.785384476184845, + "learning_rate": 0.0008315837749694252, + "loss": 3.5493, + "step": 19830 + }, + { + "epoch": 1.3476695203152602, + "grad_norm": 0.6150581240653992, + "learning_rate": 0.0008315413099605926, + "loss": 3.7445, + "step": 19835 + }, + { + "epoch": 1.348009240385922, + "grad_norm": 0.7939785122871399, + "learning_rate": 0.0008314988449517598, + "loss": 3.558, + "step": 19840 + }, + { + "epoch": 1.3483489604565837, + "grad_norm": 0.896045446395874, + "learning_rate": 0.000831456379942927, + "loss": 3.5743, + "step": 19845 + }, + { + "epoch": 1.3486886805272456, + "grad_norm": 0.9773976802825928, + "learning_rate": 0.0008314139149340943, + "loss": 3.7642, + "step": 19850 + }, + { + "epoch": 1.3490284005979074, + "grad_norm": 0.725162148475647, + "learning_rate": 0.0008313714499252616, + "loss": 3.693, + "step": 19855 + }, + { + "epoch": 1.349368120668569, + "grad_norm": 0.8170601725578308, + "learning_rate": 0.0008313289849164288, + "loss": 3.6087, + "step": 19860 + }, + { + "epoch": 1.349707840739231, + "grad_norm": 0.965078592300415, + "learning_rate": 0.0008312865199075962, + "loss": 3.7533, + "step": 19865 + }, + { + "epoch": 1.3500475608098927, + "grad_norm": 0.8299670219421387, + "learning_rate": 0.0008312440548987635, + "loss": 3.713, + "step": 19870 + }, + { + "epoch": 1.3503872808805544, + "grad_norm": 0.8961119055747986, + "learning_rate": 0.0008312015898899307, + "loss": 3.8778, + "step": 19875 + }, + { + "epoch": 1.3507270009512162, + "grad_norm": 0.8540622591972351, + "learning_rate": 0.000831159124881098, + "loss": 3.928, + "step": 19880 + }, + { + "epoch": 1.351066721021878, + "grad_norm": 0.9417552351951599, + "learning_rate": 0.0008311166598722652, + "loss": 3.7619, + "step": 19885 + }, + { + "epoch": 1.3514064410925397, + "grad_norm": 0.8376766443252563, + "learning_rate": 0.0008310741948634325, + "loss": 3.4543, + "step": 19890 + }, + { + "epoch": 1.3517461611632016, + "grad_norm": 0.7920647263526917, + "learning_rate": 0.0008310317298545998, + "loss": 3.7001, + "step": 19895 + }, + { + "epoch": 1.3520858812338634, + "grad_norm": 0.8727974891662598, + "learning_rate": 0.0008309892648457671, + "loss": 3.5563, + "step": 19900 + }, + { + "epoch": 1.352425601304525, + "grad_norm": 1.4218459129333496, + "learning_rate": 0.0008309467998369344, + "loss": 3.6041, + "step": 19905 + }, + { + "epoch": 1.352765321375187, + "grad_norm": 0.9833852052688599, + "learning_rate": 0.0008309043348281017, + "loss": 3.6877, + "step": 19910 + }, + { + "epoch": 1.3531050414458488, + "grad_norm": 0.7281168103218079, + "learning_rate": 0.0008308618698192689, + "loss": 3.6883, + "step": 19915 + }, + { + "epoch": 1.3534447615165104, + "grad_norm": 1.0514922142028809, + "learning_rate": 0.0008308194048104362, + "loss": 3.7067, + "step": 19920 + }, + { + "epoch": 1.3537844815871722, + "grad_norm": 0.9565787315368652, + "learning_rate": 0.0008307769398016035, + "loss": 3.5998, + "step": 19925 + }, + { + "epoch": 1.3541242016578339, + "grad_norm": 0.7510514855384827, + "learning_rate": 0.0008307344747927707, + "loss": 3.8489, + "step": 19930 + }, + { + "epoch": 1.3544639217284957, + "grad_norm": 0.7497783899307251, + "learning_rate": 0.0008306920097839382, + "loss": 3.7901, + "step": 19935 + }, + { + "epoch": 1.3548036417991576, + "grad_norm": 0.8029617667198181, + "learning_rate": 0.0008306495447751054, + "loss": 3.5973, + "step": 19940 + }, + { + "epoch": 1.3551433618698192, + "grad_norm": 0.7661157250404358, + "learning_rate": 0.0008306070797662726, + "loss": 3.7219, + "step": 19945 + }, + { + "epoch": 1.355483081940481, + "grad_norm": 0.7473312616348267, + "learning_rate": 0.0008305646147574399, + "loss": 3.6941, + "step": 19950 + }, + { + "epoch": 1.3558228020111427, + "grad_norm": 0.7058807611465454, + "learning_rate": 0.0008305221497486072, + "loss": 3.6592, + "step": 19955 + }, + { + "epoch": 1.3561625220818045, + "grad_norm": 0.7740005850791931, + "learning_rate": 0.0008304796847397744, + "loss": 3.7558, + "step": 19960 + }, + { + "epoch": 1.3565022421524664, + "grad_norm": 0.8686883449554443, + "learning_rate": 0.0008304372197309417, + "loss": 3.8264, + "step": 19965 + }, + { + "epoch": 1.356841962223128, + "grad_norm": 0.9508535861968994, + "learning_rate": 0.0008303947547221091, + "loss": 3.6278, + "step": 19970 + }, + { + "epoch": 1.3571816822937899, + "grad_norm": 0.8972076773643494, + "learning_rate": 0.0008303522897132763, + "loss": 3.5942, + "step": 19975 + }, + { + "epoch": 1.3575214023644517, + "grad_norm": 0.7137060165405273, + "learning_rate": 0.0008303098247044436, + "loss": 3.7111, + "step": 19980 + }, + { + "epoch": 1.3578611224351134, + "grad_norm": 0.7540163993835449, + "learning_rate": 0.0008302673596956108, + "loss": 3.5098, + "step": 19985 + }, + { + "epoch": 1.3582008425057752, + "grad_norm": 0.6191962957382202, + "learning_rate": 0.0008302248946867781, + "loss": 3.567, + "step": 19990 + }, + { + "epoch": 1.358540562576437, + "grad_norm": 1.0202337503433228, + "learning_rate": 0.0008301824296779454, + "loss": 3.4531, + "step": 19995 + }, + { + "epoch": 1.3588802826470987, + "grad_norm": 0.8122200965881348, + "learning_rate": 0.0008301399646691126, + "loss": 3.7727, + "step": 20000 + }, + { + "epoch": 1.3592200027177606, + "grad_norm": 0.7568419575691223, + "learning_rate": 0.00083009749966028, + "loss": 3.8636, + "step": 20005 + }, + { + "epoch": 1.3595597227884224, + "grad_norm": 0.9622646570205688, + "learning_rate": 0.0008300550346514473, + "loss": 3.7696, + "step": 20010 + }, + { + "epoch": 1.359899442859084, + "grad_norm": 0.7535944581031799, + "learning_rate": 0.0008300125696426145, + "loss": 3.8556, + "step": 20015 + }, + { + "epoch": 1.360239162929746, + "grad_norm": 0.7176857590675354, + "learning_rate": 0.0008299701046337818, + "loss": 3.7539, + "step": 20020 + }, + { + "epoch": 1.3605788830004077, + "grad_norm": 1.5292377471923828, + "learning_rate": 0.0008299276396249491, + "loss": 3.5624, + "step": 20025 + }, + { + "epoch": 1.3609186030710694, + "grad_norm": 0.6986082792282104, + "learning_rate": 0.0008298851746161163, + "loss": 3.6364, + "step": 20030 + }, + { + "epoch": 1.3612583231417312, + "grad_norm": 0.7370955348014832, + "learning_rate": 0.0008298427096072835, + "loss": 3.772, + "step": 20035 + }, + { + "epoch": 1.361598043212393, + "grad_norm": 0.8615638017654419, + "learning_rate": 0.000829800244598451, + "loss": 3.7706, + "step": 20040 + }, + { + "epoch": 1.3619377632830547, + "grad_norm": 0.8507272601127625, + "learning_rate": 0.0008297577795896182, + "loss": 3.5582, + "step": 20045 + }, + { + "epoch": 1.3622774833537166, + "grad_norm": 0.9351620078086853, + "learning_rate": 0.0008297153145807854, + "loss": 3.7316, + "step": 20050 + }, + { + "epoch": 1.3626172034243784, + "grad_norm": 0.8998017907142639, + "learning_rate": 0.0008296728495719528, + "loss": 3.6514, + "step": 20055 + }, + { + "epoch": 1.36295692349504, + "grad_norm": 0.6098034977912903, + "learning_rate": 0.00082963038456312, + "loss": 3.6465, + "step": 20060 + }, + { + "epoch": 1.363296643565702, + "grad_norm": 0.8833012580871582, + "learning_rate": 0.0008295879195542872, + "loss": 3.7565, + "step": 20065 + }, + { + "epoch": 1.3636363636363638, + "grad_norm": 1.0289818048477173, + "learning_rate": 0.0008295454545454546, + "loss": 3.6553, + "step": 20070 + }, + { + "epoch": 1.3639760837070254, + "grad_norm": 0.9692005515098572, + "learning_rate": 0.0008295029895366219, + "loss": 3.6894, + "step": 20075 + }, + { + "epoch": 1.3643158037776872, + "grad_norm": 0.7152788639068604, + "learning_rate": 0.0008294605245277891, + "loss": 3.5481, + "step": 20080 + }, + { + "epoch": 1.364655523848349, + "grad_norm": 0.8993133902549744, + "learning_rate": 0.0008294180595189564, + "loss": 3.4693, + "step": 20085 + }, + { + "epoch": 1.3649952439190107, + "grad_norm": 0.8196595907211304, + "learning_rate": 0.0008293755945101237, + "loss": 3.5787, + "step": 20090 + }, + { + "epoch": 1.3653349639896726, + "grad_norm": 0.7238662242889404, + "learning_rate": 0.0008293331295012909, + "loss": 3.5447, + "step": 20095 + }, + { + "epoch": 1.3656746840603342, + "grad_norm": 1.0158884525299072, + "learning_rate": 0.0008292906644924582, + "loss": 3.6347, + "step": 20100 + }, + { + "epoch": 1.366014404130996, + "grad_norm": 0.9779278635978699, + "learning_rate": 0.0008292481994836255, + "loss": 3.7077, + "step": 20105 + }, + { + "epoch": 1.366354124201658, + "grad_norm": 1.117322564125061, + "learning_rate": 0.0008292057344747928, + "loss": 3.4578, + "step": 20110 + }, + { + "epoch": 1.3666938442723195, + "grad_norm": 1.8293728828430176, + "learning_rate": 0.0008291632694659601, + "loss": 3.6426, + "step": 20115 + }, + { + "epoch": 1.3670335643429814, + "grad_norm": 0.8472502827644348, + "learning_rate": 0.0008291208044571274, + "loss": 3.1927, + "step": 20120 + }, + { + "epoch": 1.367373284413643, + "grad_norm": 0.7554154396057129, + "learning_rate": 0.0008290783394482946, + "loss": 3.4672, + "step": 20125 + }, + { + "epoch": 1.3677130044843049, + "grad_norm": 0.790399968624115, + "learning_rate": 0.0008290358744394619, + "loss": 3.5768, + "step": 20130 + }, + { + "epoch": 1.3680527245549667, + "grad_norm": 0.8999508023262024, + "learning_rate": 0.0008289934094306291, + "loss": 3.6848, + "step": 20135 + }, + { + "epoch": 1.3683924446256284, + "grad_norm": 0.9056254625320435, + "learning_rate": 0.0008289509444217964, + "loss": 3.3525, + "step": 20140 + }, + { + "epoch": 1.3687321646962902, + "grad_norm": 3.9266421794891357, + "learning_rate": 0.0008289084794129638, + "loss": 3.524, + "step": 20145 + }, + { + "epoch": 1.369071884766952, + "grad_norm": 0.9506530165672302, + "learning_rate": 0.000828866014404131, + "loss": 3.4021, + "step": 20150 + }, + { + "epoch": 1.3694116048376137, + "grad_norm": 0.9955437183380127, + "learning_rate": 0.0008288235493952983, + "loss": 3.8126, + "step": 20155 + }, + { + "epoch": 1.3697513249082756, + "grad_norm": 0.8767886161804199, + "learning_rate": 0.0008287810843864656, + "loss": 3.6191, + "step": 20160 + }, + { + "epoch": 1.3700910449789374, + "grad_norm": 0.7537773847579956, + "learning_rate": 0.0008287386193776328, + "loss": 3.6686, + "step": 20165 + }, + { + "epoch": 1.370430765049599, + "grad_norm": 0.8429709672927856, + "learning_rate": 0.0008286961543688, + "loss": 3.4358, + "step": 20170 + }, + { + "epoch": 1.370770485120261, + "grad_norm": 0.6619389057159424, + "learning_rate": 0.0008286536893599674, + "loss": 3.4888, + "step": 20175 + }, + { + "epoch": 1.3711102051909227, + "grad_norm": 1.3760391473770142, + "learning_rate": 0.0008286112243511347, + "loss": 3.6753, + "step": 20180 + }, + { + "epoch": 1.3714499252615844, + "grad_norm": 0.7743170261383057, + "learning_rate": 0.0008285687593423019, + "loss": 3.3126, + "step": 20185 + }, + { + "epoch": 1.3717896453322462, + "grad_norm": 0.9974854588508606, + "learning_rate": 0.0008285262943334693, + "loss": 3.7557, + "step": 20190 + }, + { + "epoch": 1.372129365402908, + "grad_norm": 0.983306348323822, + "learning_rate": 0.0008284838293246365, + "loss": 3.4007, + "step": 20195 + }, + { + "epoch": 1.3724690854735697, + "grad_norm": 1.081614375114441, + "learning_rate": 0.0008284413643158037, + "loss": 3.3639, + "step": 20200 + }, + { + "epoch": 1.3728088055442316, + "grad_norm": 0.7151765823364258, + "learning_rate": 0.0008283988993069711, + "loss": 3.4608, + "step": 20205 + }, + { + "epoch": 1.3731485256148934, + "grad_norm": 0.7609778642654419, + "learning_rate": 0.0008283564342981383, + "loss": 3.5031, + "step": 20210 + }, + { + "epoch": 1.373488245685555, + "grad_norm": 0.8248502612113953, + "learning_rate": 0.0008283139692893056, + "loss": 3.5884, + "step": 20215 + }, + { + "epoch": 1.373827965756217, + "grad_norm": 1.5455756187438965, + "learning_rate": 0.000828271504280473, + "loss": 3.392, + "step": 20220 + }, + { + "epoch": 1.3741676858268788, + "grad_norm": 0.7607990503311157, + "learning_rate": 0.0008282290392716402, + "loss": 3.684, + "step": 20225 + }, + { + "epoch": 1.3745074058975404, + "grad_norm": 3.544597864151001, + "learning_rate": 0.0008281865742628074, + "loss": 3.3773, + "step": 20230 + }, + { + "epoch": 1.3748471259682022, + "grad_norm": 1.0479261875152588, + "learning_rate": 0.0008281441092539747, + "loss": 3.574, + "step": 20235 + }, + { + "epoch": 1.375186846038864, + "grad_norm": 0.7941829562187195, + "learning_rate": 0.000828101644245142, + "loss": 3.5016, + "step": 20240 + }, + { + "epoch": 1.3755265661095257, + "grad_norm": 1.0169727802276611, + "learning_rate": 0.0008280591792363092, + "loss": 3.5307, + "step": 20245 + }, + { + "epoch": 1.3758662861801876, + "grad_norm": 0.9216648936271667, + "learning_rate": 0.0008280167142274766, + "loss": 3.5364, + "step": 20250 + }, + { + "epoch": 1.3762060062508494, + "grad_norm": 0.8656046986579895, + "learning_rate": 0.0008279742492186439, + "loss": 3.519, + "step": 20255 + }, + { + "epoch": 1.376545726321511, + "grad_norm": 0.9340004324913025, + "learning_rate": 0.0008279317842098111, + "loss": 3.4805, + "step": 20260 + }, + { + "epoch": 1.376885446392173, + "grad_norm": 1.1377168893814087, + "learning_rate": 0.0008278893192009784, + "loss": 3.6431, + "step": 20265 + }, + { + "epoch": 1.3772251664628345, + "grad_norm": 0.9111413359642029, + "learning_rate": 0.0008278468541921456, + "loss": 3.8543, + "step": 20270 + }, + { + "epoch": 1.3775648865334964, + "grad_norm": 1.0429924726486206, + "learning_rate": 0.000827804389183313, + "loss": 3.6353, + "step": 20275 + }, + { + "epoch": 1.3779046066041583, + "grad_norm": 0.7700743079185486, + "learning_rate": 0.0008277619241744803, + "loss": 3.6987, + "step": 20280 + }, + { + "epoch": 1.3782443266748199, + "grad_norm": 1.1103097200393677, + "learning_rate": 0.0008277194591656475, + "loss": 3.5317, + "step": 20285 + }, + { + "epoch": 1.3785840467454817, + "grad_norm": 1.1317721605300903, + "learning_rate": 0.0008276769941568149, + "loss": 3.8176, + "step": 20290 + }, + { + "epoch": 1.3789237668161434, + "grad_norm": 0.9521926045417786, + "learning_rate": 0.0008276345291479821, + "loss": 3.7743, + "step": 20295 + }, + { + "epoch": 1.3792634868868052, + "grad_norm": 1.0168817043304443, + "learning_rate": 0.0008275920641391493, + "loss": 3.8019, + "step": 20300 + }, + { + "epoch": 1.379603206957467, + "grad_norm": 0.8945667743682861, + "learning_rate": 0.0008275495991303167, + "loss": 3.4125, + "step": 20305 + }, + { + "epoch": 1.3799429270281287, + "grad_norm": 0.7459174394607544, + "learning_rate": 0.0008275071341214839, + "loss": 3.5869, + "step": 20310 + }, + { + "epoch": 1.3802826470987906, + "grad_norm": 0.7691378593444824, + "learning_rate": 0.0008274646691126512, + "loss": 3.7097, + "step": 20315 + }, + { + "epoch": 1.3806223671694524, + "grad_norm": 1.0497788190841675, + "learning_rate": 0.0008274222041038186, + "loss": 3.6395, + "step": 20320 + }, + { + "epoch": 1.380962087240114, + "grad_norm": 1.243142008781433, + "learning_rate": 0.0008273797390949858, + "loss": 3.5531, + "step": 20325 + }, + { + "epoch": 1.381301807310776, + "grad_norm": 0.7451356053352356, + "learning_rate": 0.000827337274086153, + "loss": 3.5585, + "step": 20330 + }, + { + "epoch": 1.3816415273814378, + "grad_norm": 1.9425336122512817, + "learning_rate": 0.0008272948090773203, + "loss": 3.5113, + "step": 20335 + }, + { + "epoch": 1.3819812474520994, + "grad_norm": 0.7421507239341736, + "learning_rate": 0.0008272523440684876, + "loss": 3.7413, + "step": 20340 + }, + { + "epoch": 1.3823209675227612, + "grad_norm": 0.7696300745010376, + "learning_rate": 0.0008272098790596548, + "loss": 3.5647, + "step": 20345 + }, + { + "epoch": 1.382660687593423, + "grad_norm": 0.7083591222763062, + "learning_rate": 0.0008271674140508222, + "loss": 3.5757, + "step": 20350 + }, + { + "epoch": 1.3830004076640847, + "grad_norm": 0.8059934377670288, + "learning_rate": 0.0008271249490419895, + "loss": 3.4186, + "step": 20355 + }, + { + "epoch": 1.3833401277347466, + "grad_norm": 0.8010045289993286, + "learning_rate": 0.0008270824840331567, + "loss": 3.4904, + "step": 20360 + }, + { + "epoch": 1.3836798478054084, + "grad_norm": 0.7706525921821594, + "learning_rate": 0.000827040019024324, + "loss": 3.7753, + "step": 20365 + }, + { + "epoch": 1.38401956787607, + "grad_norm": 1.0286221504211426, + "learning_rate": 0.0008269975540154912, + "loss": 3.679, + "step": 20370 + }, + { + "epoch": 1.384359287946732, + "grad_norm": 0.9037805199623108, + "learning_rate": 0.0008269550890066585, + "loss": 3.6565, + "step": 20375 + }, + { + "epoch": 1.3846990080173938, + "grad_norm": 0.7697586417198181, + "learning_rate": 0.0008269126239978258, + "loss": 3.8985, + "step": 20380 + }, + { + "epoch": 1.3850387280880554, + "grad_norm": 0.7971416115760803, + "learning_rate": 0.0008268701589889931, + "loss": 3.7845, + "step": 20385 + }, + { + "epoch": 1.3853784481587172, + "grad_norm": 0.6265209317207336, + "learning_rate": 0.0008268276939801604, + "loss": 3.501, + "step": 20390 + }, + { + "epoch": 1.385718168229379, + "grad_norm": 0.8635155558586121, + "learning_rate": 0.0008267852289713277, + "loss": 3.6134, + "step": 20395 + }, + { + "epoch": 1.3860578883000407, + "grad_norm": 0.8051210045814514, + "learning_rate": 0.0008267427639624949, + "loss": 3.5318, + "step": 20400 + }, + { + "epoch": 1.3863976083707026, + "grad_norm": 1.0002379417419434, + "learning_rate": 0.0008267002989536622, + "loss": 3.5288, + "step": 20405 + }, + { + "epoch": 1.3867373284413644, + "grad_norm": 0.8467056155204773, + "learning_rate": 0.0008266578339448295, + "loss": 3.5673, + "step": 20410 + }, + { + "epoch": 1.387077048512026, + "grad_norm": 0.9337602853775024, + "learning_rate": 0.0008266153689359967, + "loss": 3.7144, + "step": 20415 + }, + { + "epoch": 1.387416768582688, + "grad_norm": 0.7187165021896362, + "learning_rate": 0.000826572903927164, + "loss": 3.7148, + "step": 20420 + }, + { + "epoch": 1.3877564886533498, + "grad_norm": 0.8400835394859314, + "learning_rate": 0.0008265304389183314, + "loss": 3.5658, + "step": 20425 + }, + { + "epoch": 1.3880962087240114, + "grad_norm": 0.7464248538017273, + "learning_rate": 0.0008264879739094986, + "loss": 3.5879, + "step": 20430 + }, + { + "epoch": 1.3884359287946733, + "grad_norm": 0.819556474685669, + "learning_rate": 0.0008264455089006658, + "loss": 3.774, + "step": 20435 + }, + { + "epoch": 1.3887756488653349, + "grad_norm": 0.8895112872123718, + "learning_rate": 0.0008264030438918332, + "loss": 3.5174, + "step": 20440 + }, + { + "epoch": 1.3891153689359967, + "grad_norm": 0.7157255411148071, + "learning_rate": 0.0008263605788830004, + "loss": 3.76, + "step": 20445 + }, + { + "epoch": 1.3894550890066586, + "grad_norm": 0.9935908913612366, + "learning_rate": 0.0008263181138741676, + "loss": 3.6405, + "step": 20450 + }, + { + "epoch": 1.3897948090773202, + "grad_norm": 0.798677384853363, + "learning_rate": 0.0008262756488653351, + "loss": 3.4065, + "step": 20455 + }, + { + "epoch": 1.390134529147982, + "grad_norm": 0.9333176016807556, + "learning_rate": 0.0008262331838565023, + "loss": 3.8207, + "step": 20460 + }, + { + "epoch": 1.3904742492186437, + "grad_norm": 0.6949829459190369, + "learning_rate": 0.0008261907188476695, + "loss": 3.5882, + "step": 20465 + }, + { + "epoch": 1.3908139692893056, + "grad_norm": 0.8420959711074829, + "learning_rate": 0.0008261482538388369, + "loss": 3.4167, + "step": 20470 + }, + { + "epoch": 1.3911536893599674, + "grad_norm": 0.7815521955490112, + "learning_rate": 0.0008261057888300041, + "loss": 3.6029, + "step": 20475 + }, + { + "epoch": 1.391493409430629, + "grad_norm": 0.9085712432861328, + "learning_rate": 0.0008260633238211713, + "loss": 3.6257, + "step": 20480 + }, + { + "epoch": 1.391833129501291, + "grad_norm": 0.9145505428314209, + "learning_rate": 0.0008260208588123386, + "loss": 3.4722, + "step": 20485 + }, + { + "epoch": 1.3921728495719528, + "grad_norm": 0.8195421695709229, + "learning_rate": 0.000825978393803506, + "loss": 3.6158, + "step": 20490 + }, + { + "epoch": 1.3925125696426144, + "grad_norm": 0.872031033039093, + "learning_rate": 0.0008259359287946732, + "loss": 3.469, + "step": 20495 + }, + { + "epoch": 1.3928522897132762, + "grad_norm": 1.4268487691879272, + "learning_rate": 0.0008258934637858405, + "loss": 3.5495, + "step": 20500 + }, + { + "epoch": 1.393192009783938, + "grad_norm": 0.8297573924064636, + "learning_rate": 0.0008258509987770078, + "loss": 3.5599, + "step": 20505 + }, + { + "epoch": 1.3935317298545997, + "grad_norm": 0.824938952922821, + "learning_rate": 0.000825808533768175, + "loss": 3.5409, + "step": 20510 + }, + { + "epoch": 1.3938714499252616, + "grad_norm": 0.8266805410385132, + "learning_rate": 0.0008257660687593423, + "loss": 3.5616, + "step": 20515 + }, + { + "epoch": 1.3942111699959234, + "grad_norm": 0.9786876440048218, + "learning_rate": 0.0008257236037505095, + "loss": 3.7859, + "step": 20520 + }, + { + "epoch": 1.394550890066585, + "grad_norm": 1.1329405307769775, + "learning_rate": 0.0008256811387416769, + "loss": 3.6537, + "step": 20525 + }, + { + "epoch": 1.394890610137247, + "grad_norm": 1.0957200527191162, + "learning_rate": 0.0008256386737328442, + "loss": 3.3437, + "step": 20530 + }, + { + "epoch": 1.3952303302079088, + "grad_norm": 0.8033496737480164, + "learning_rate": 0.0008255962087240114, + "loss": 3.6547, + "step": 20535 + }, + { + "epoch": 1.3955700502785704, + "grad_norm": 0.7064065933227539, + "learning_rate": 0.0008255537437151787, + "loss": 3.7429, + "step": 20540 + }, + { + "epoch": 1.3959097703492322, + "grad_norm": 0.7505779266357422, + "learning_rate": 0.000825511278706346, + "loss": 3.6715, + "step": 20545 + }, + { + "epoch": 1.396249490419894, + "grad_norm": 0.8774320483207703, + "learning_rate": 0.0008254688136975132, + "loss": 3.333, + "step": 20550 + }, + { + "epoch": 1.3965892104905557, + "grad_norm": 0.6276168823242188, + "learning_rate": 0.0008254263486886804, + "loss": 3.4375, + "step": 20555 + }, + { + "epoch": 1.3969289305612176, + "grad_norm": 0.6749109625816345, + "learning_rate": 0.0008253838836798479, + "loss": 3.5985, + "step": 20560 + }, + { + "epoch": 1.3972686506318794, + "grad_norm": 0.7648650407791138, + "learning_rate": 0.0008253414186710151, + "loss": 3.6375, + "step": 20565 + }, + { + "epoch": 1.397608370702541, + "grad_norm": 0.8186941146850586, + "learning_rate": 0.0008252989536621823, + "loss": 3.5453, + "step": 20570 + }, + { + "epoch": 1.397948090773203, + "grad_norm": 0.7318486571311951, + "learning_rate": 0.0008252564886533497, + "loss": 3.2976, + "step": 20575 + }, + { + "epoch": 1.3982878108438648, + "grad_norm": 0.9368587732315063, + "learning_rate": 0.0008252140236445169, + "loss": 3.6553, + "step": 20580 + }, + { + "epoch": 1.3986275309145264, + "grad_norm": 0.9344570636749268, + "learning_rate": 0.0008251715586356841, + "loss": 3.6233, + "step": 20585 + }, + { + "epoch": 1.3989672509851883, + "grad_norm": 0.8090149164199829, + "learning_rate": 0.0008251290936268515, + "loss": 3.6855, + "step": 20590 + }, + { + "epoch": 1.3993069710558501, + "grad_norm": 1.348703384399414, + "learning_rate": 0.0008250866286180188, + "loss": 3.4964, + "step": 20595 + }, + { + "epoch": 1.3996466911265117, + "grad_norm": 1.0034469366073608, + "learning_rate": 0.000825044163609186, + "loss": 3.5434, + "step": 20600 + }, + { + "epoch": 1.3999864111971736, + "grad_norm": 0.9152267575263977, + "learning_rate": 0.0008250016986003534, + "loss": 3.7884, + "step": 20605 + }, + { + "epoch": 1.4003261312678352, + "grad_norm": 0.8734991550445557, + "learning_rate": 0.0008249592335915206, + "loss": 3.6744, + "step": 20610 + }, + { + "epoch": 1.400665851338497, + "grad_norm": 0.6777758598327637, + "learning_rate": 0.0008249167685826879, + "loss": 3.7508, + "step": 20615 + }, + { + "epoch": 1.401005571409159, + "grad_norm": 0.7659586071968079, + "learning_rate": 0.0008248743035738551, + "loss": 3.5722, + "step": 20620 + }, + { + "epoch": 1.4013452914798206, + "grad_norm": 0.959725558757782, + "learning_rate": 0.0008248318385650224, + "loss": 3.5768, + "step": 20625 + }, + { + "epoch": 1.4016850115504824, + "grad_norm": 0.5695216059684753, + "learning_rate": 0.0008247893735561898, + "loss": 3.6181, + "step": 20630 + }, + { + "epoch": 1.402024731621144, + "grad_norm": 0.8693369030952454, + "learning_rate": 0.000824746908547357, + "loss": 3.4066, + "step": 20635 + }, + { + "epoch": 1.402364451691806, + "grad_norm": 0.7927300333976746, + "learning_rate": 0.0008247044435385243, + "loss": 4.0636, + "step": 20640 + }, + { + "epoch": 1.4027041717624678, + "grad_norm": 0.9798035025596619, + "learning_rate": 0.0008246619785296916, + "loss": 3.9302, + "step": 20645 + }, + { + "epoch": 1.4030438918331294, + "grad_norm": 0.9714420437812805, + "learning_rate": 0.0008246195135208588, + "loss": 3.4737, + "step": 20650 + }, + { + "epoch": 1.4033836119037912, + "grad_norm": 0.7763824462890625, + "learning_rate": 0.000824577048512026, + "loss": 3.4242, + "step": 20655 + }, + { + "epoch": 1.403723331974453, + "grad_norm": 0.8162153363227844, + "learning_rate": 0.0008245345835031934, + "loss": 3.6222, + "step": 20660 + }, + { + "epoch": 1.4040630520451147, + "grad_norm": 0.8270282745361328, + "learning_rate": 0.0008244921184943607, + "loss": 3.757, + "step": 20665 + }, + { + "epoch": 1.4044027721157766, + "grad_norm": 0.8101107478141785, + "learning_rate": 0.0008244496534855279, + "loss": 3.5851, + "step": 20670 + }, + { + "epoch": 1.4047424921864384, + "grad_norm": 0.8065143823623657, + "learning_rate": 0.0008244071884766953, + "loss": 3.5528, + "step": 20675 + }, + { + "epoch": 1.4050822122571, + "grad_norm": 0.7318259477615356, + "learning_rate": 0.0008243647234678625, + "loss": 3.739, + "step": 20680 + }, + { + "epoch": 1.405421932327762, + "grad_norm": 1.1027981042861938, + "learning_rate": 0.0008243222584590297, + "loss": 3.586, + "step": 20685 + }, + { + "epoch": 1.4057616523984238, + "grad_norm": 1.04094398021698, + "learning_rate": 0.0008242797934501971, + "loss": 3.4476, + "step": 20690 + }, + { + "epoch": 1.4061013724690854, + "grad_norm": 0.8588951230049133, + "learning_rate": 0.0008242373284413643, + "loss": 3.6069, + "step": 20695 + }, + { + "epoch": 1.4064410925397473, + "grad_norm": 0.9245058298110962, + "learning_rate": 0.0008241948634325316, + "loss": 3.6423, + "step": 20700 + }, + { + "epoch": 1.406780812610409, + "grad_norm": 1.0811768770217896, + "learning_rate": 0.000824152398423699, + "loss": 3.4884, + "step": 20705 + }, + { + "epoch": 1.4071205326810707, + "grad_norm": 0.8094359636306763, + "learning_rate": 0.0008241099334148662, + "loss": 3.8573, + "step": 20710 + }, + { + "epoch": 1.4074602527517326, + "grad_norm": 0.8182267546653748, + "learning_rate": 0.0008240674684060334, + "loss": 3.3837, + "step": 20715 + }, + { + "epoch": 1.4077999728223944, + "grad_norm": 0.6878779530525208, + "learning_rate": 0.0008240250033972007, + "loss": 3.3033, + "step": 20720 + }, + { + "epoch": 1.408139692893056, + "grad_norm": 0.7361664175987244, + "learning_rate": 0.000823982538388368, + "loss": 3.5176, + "step": 20725 + }, + { + "epoch": 1.408479412963718, + "grad_norm": 1.0483208894729614, + "learning_rate": 0.0008239400733795352, + "loss": 3.4674, + "step": 20730 + }, + { + "epoch": 1.4088191330343798, + "grad_norm": 0.8424781560897827, + "learning_rate": 0.0008238976083707026, + "loss": 3.6651, + "step": 20735 + }, + { + "epoch": 1.4091588531050414, + "grad_norm": 0.9902977347373962, + "learning_rate": 0.0008238551433618699, + "loss": 3.4855, + "step": 20740 + }, + { + "epoch": 1.4094985731757033, + "grad_norm": 0.9176404476165771, + "learning_rate": 0.0008238126783530371, + "loss": 3.5121, + "step": 20745 + }, + { + "epoch": 1.4098382932463651, + "grad_norm": 1.2062361240386963, + "learning_rate": 0.0008237702133442044, + "loss": 3.6757, + "step": 20750 + }, + { + "epoch": 1.4101780133170267, + "grad_norm": 0.8620239496231079, + "learning_rate": 0.0008237277483353717, + "loss": 3.4297, + "step": 20755 + }, + { + "epoch": 1.4105177333876886, + "grad_norm": 0.7602016925811768, + "learning_rate": 0.0008236852833265389, + "loss": 3.5532, + "step": 20760 + }, + { + "epoch": 1.4108574534583505, + "grad_norm": 0.8381115198135376, + "learning_rate": 0.0008236428183177062, + "loss": 3.7047, + "step": 20765 + }, + { + "epoch": 1.411197173529012, + "grad_norm": 0.8198944926261902, + "learning_rate": 0.0008236003533088735, + "loss": 3.6948, + "step": 20770 + }, + { + "epoch": 1.411536893599674, + "grad_norm": 0.9184916019439697, + "learning_rate": 0.0008235578883000408, + "loss": 3.5166, + "step": 20775 + }, + { + "epoch": 1.4118766136703356, + "grad_norm": 0.7563937306404114, + "learning_rate": 0.0008235154232912081, + "loss": 3.8841, + "step": 20780 + }, + { + "epoch": 1.4122163337409974, + "grad_norm": 1.1413289308547974, + "learning_rate": 0.0008234729582823753, + "loss": 3.5892, + "step": 20785 + }, + { + "epoch": 1.4125560538116593, + "grad_norm": 1.0585217475891113, + "learning_rate": 0.0008234304932735426, + "loss": 3.531, + "step": 20790 + }, + { + "epoch": 1.412895773882321, + "grad_norm": 0.7346665263175964, + "learning_rate": 0.0008233880282647099, + "loss": 3.7766, + "step": 20795 + }, + { + "epoch": 1.4132354939529828, + "grad_norm": 0.8998989462852478, + "learning_rate": 0.0008233455632558771, + "loss": 3.5497, + "step": 20800 + }, + { + "epoch": 1.4135752140236444, + "grad_norm": 0.7291684150695801, + "learning_rate": 0.0008233030982470445, + "loss": 3.5563, + "step": 20805 + }, + { + "epoch": 1.4139149340943062, + "grad_norm": 0.8629082441329956, + "learning_rate": 0.0008232606332382118, + "loss": 3.4234, + "step": 20810 + }, + { + "epoch": 1.414254654164968, + "grad_norm": 1.0397497415542603, + "learning_rate": 0.000823218168229379, + "loss": 3.7712, + "step": 20815 + }, + { + "epoch": 1.4145943742356297, + "grad_norm": 1.0735012292861938, + "learning_rate": 0.0008231757032205462, + "loss": 3.6501, + "step": 20820 + }, + { + "epoch": 1.4149340943062916, + "grad_norm": 0.8563271164894104, + "learning_rate": 0.0008231332382117136, + "loss": 3.7, + "step": 20825 + }, + { + "epoch": 1.4152738143769534, + "grad_norm": 0.9112222790718079, + "learning_rate": 0.0008230907732028808, + "loss": 3.6677, + "step": 20830 + }, + { + "epoch": 1.415613534447615, + "grad_norm": 0.7145978808403015, + "learning_rate": 0.000823048308194048, + "loss": 3.6044, + "step": 20835 + }, + { + "epoch": 1.415953254518277, + "grad_norm": 0.739173412322998, + "learning_rate": 0.0008230058431852155, + "loss": 3.79, + "step": 20840 + }, + { + "epoch": 1.4162929745889388, + "grad_norm": 1.0261688232421875, + "learning_rate": 0.0008229633781763827, + "loss": 3.83, + "step": 20845 + }, + { + "epoch": 1.4166326946596004, + "grad_norm": 1.3438572883605957, + "learning_rate": 0.0008229209131675499, + "loss": 3.4498, + "step": 20850 + }, + { + "epoch": 1.4169724147302623, + "grad_norm": 0.7410697937011719, + "learning_rate": 0.0008228784481587173, + "loss": 3.3386, + "step": 20855 + }, + { + "epoch": 1.417312134800924, + "grad_norm": 0.8338222503662109, + "learning_rate": 0.0008228359831498845, + "loss": 3.5874, + "step": 20860 + }, + { + "epoch": 1.4176518548715857, + "grad_norm": 0.9488686919212341, + "learning_rate": 0.0008227935181410517, + "loss": 3.6481, + "step": 20865 + }, + { + "epoch": 1.4179915749422476, + "grad_norm": 0.750819206237793, + "learning_rate": 0.0008227510531322191, + "loss": 3.6444, + "step": 20870 + }, + { + "epoch": 1.4183312950129094, + "grad_norm": 0.8117307424545288, + "learning_rate": 0.0008227085881233864, + "loss": 3.6442, + "step": 20875 + }, + { + "epoch": 1.418671015083571, + "grad_norm": 0.8307729363441467, + "learning_rate": 0.0008226661231145536, + "loss": 3.6969, + "step": 20880 + }, + { + "epoch": 1.419010735154233, + "grad_norm": 0.7594723105430603, + "learning_rate": 0.0008226236581057209, + "loss": 3.6485, + "step": 20885 + }, + { + "epoch": 1.4193504552248948, + "grad_norm": 0.7602181434631348, + "learning_rate": 0.0008225811930968882, + "loss": 3.6384, + "step": 20890 + }, + { + "epoch": 1.4196901752955564, + "grad_norm": 0.9653111100196838, + "learning_rate": 0.0008225387280880554, + "loss": 3.8369, + "step": 20895 + }, + { + "epoch": 1.4200298953662183, + "grad_norm": 0.9489719867706299, + "learning_rate": 0.0008224962630792227, + "loss": 3.7231, + "step": 20900 + }, + { + "epoch": 1.4203696154368801, + "grad_norm": 0.9118810296058655, + "learning_rate": 0.0008224537980703901, + "loss": 3.4382, + "step": 20905 + }, + { + "epoch": 1.4207093355075417, + "grad_norm": 0.893730878829956, + "learning_rate": 0.0008224113330615573, + "loss": 3.788, + "step": 20910 + }, + { + "epoch": 1.4210490555782036, + "grad_norm": 0.7363401055335999, + "learning_rate": 0.0008223688680527246, + "loss": 3.5311, + "step": 20915 + }, + { + "epoch": 1.4213887756488655, + "grad_norm": 0.8287197351455688, + "learning_rate": 0.0008223264030438918, + "loss": 3.7329, + "step": 20920 + }, + { + "epoch": 1.421728495719527, + "grad_norm": 1.3203455209732056, + "learning_rate": 0.0008222839380350591, + "loss": 3.6321, + "step": 20925 + }, + { + "epoch": 1.422068215790189, + "grad_norm": 0.870648980140686, + "learning_rate": 0.0008222414730262264, + "loss": 3.5652, + "step": 20930 + }, + { + "epoch": 1.4224079358608508, + "grad_norm": 0.9471779465675354, + "learning_rate": 0.0008221990080173936, + "loss": 3.5954, + "step": 20935 + }, + { + "epoch": 1.4227476559315124, + "grad_norm": 0.7863608598709106, + "learning_rate": 0.000822156543008561, + "loss": 3.4203, + "step": 20940 + }, + { + "epoch": 1.4230873760021743, + "grad_norm": 0.8741341233253479, + "learning_rate": 0.0008221140779997283, + "loss": 3.4349, + "step": 20945 + }, + { + "epoch": 1.423427096072836, + "grad_norm": 1.0044779777526855, + "learning_rate": 0.0008220716129908955, + "loss": 3.4346, + "step": 20950 + }, + { + "epoch": 1.4237668161434978, + "grad_norm": 0.7439308166503906, + "learning_rate": 0.0008220291479820629, + "loss": 3.8433, + "step": 20955 + }, + { + "epoch": 1.4241065362141596, + "grad_norm": 0.8374903202056885, + "learning_rate": 0.0008219866829732301, + "loss": 3.7173, + "step": 20960 + }, + { + "epoch": 1.4244462562848212, + "grad_norm": 1.6416101455688477, + "learning_rate": 0.0008219442179643973, + "loss": 3.5514, + "step": 20965 + }, + { + "epoch": 1.424785976355483, + "grad_norm": 1.4563789367675781, + "learning_rate": 0.0008219017529555646, + "loss": 3.7034, + "step": 20970 + }, + { + "epoch": 1.4251256964261447, + "grad_norm": 0.7204855680465698, + "learning_rate": 0.000821859287946732, + "loss": 3.4418, + "step": 20975 + }, + { + "epoch": 1.4254654164968066, + "grad_norm": 0.8681305646896362, + "learning_rate": 0.0008218168229378992, + "loss": 3.8385, + "step": 20980 + }, + { + "epoch": 1.4258051365674684, + "grad_norm": 2.1982781887054443, + "learning_rate": 0.0008217743579290665, + "loss": 3.5276, + "step": 20985 + }, + { + "epoch": 1.42614485663813, + "grad_norm": 1.1381447315216064, + "learning_rate": 0.0008217318929202338, + "loss": 3.4344, + "step": 20990 + }, + { + "epoch": 1.426484576708792, + "grad_norm": 0.7454198002815247, + "learning_rate": 0.000821689427911401, + "loss": 3.352, + "step": 20995 + }, + { + "epoch": 1.4268242967794538, + "grad_norm": 0.7620583772659302, + "learning_rate": 0.0008216469629025683, + "loss": 3.5581, + "step": 21000 + }, + { + "epoch": 1.4271640168501154, + "grad_norm": 0.7967021465301514, + "learning_rate": 0.0008216044978937355, + "loss": 3.7029, + "step": 21005 + }, + { + "epoch": 1.4275037369207773, + "grad_norm": 0.8864526152610779, + "learning_rate": 0.0008215620328849029, + "loss": 3.7122, + "step": 21010 + }, + { + "epoch": 1.427843456991439, + "grad_norm": 0.7317259907722473, + "learning_rate": 0.0008215195678760702, + "loss": 3.8906, + "step": 21015 + }, + { + "epoch": 1.4281831770621007, + "grad_norm": 0.8627270460128784, + "learning_rate": 0.0008214771028672374, + "loss": 3.4338, + "step": 21020 + }, + { + "epoch": 1.4285228971327626, + "grad_norm": 1.137544870376587, + "learning_rate": 0.0008214346378584047, + "loss": 3.54, + "step": 21025 + }, + { + "epoch": 1.4288626172034244, + "grad_norm": 0.8415155410766602, + "learning_rate": 0.000821392172849572, + "loss": 3.6983, + "step": 21030 + }, + { + "epoch": 1.429202337274086, + "grad_norm": 0.784503161907196, + "learning_rate": 0.0008213497078407392, + "loss": 3.6645, + "step": 21035 + }, + { + "epoch": 1.429542057344748, + "grad_norm": 0.7717975378036499, + "learning_rate": 0.0008213072428319065, + "loss": 3.8199, + "step": 21040 + }, + { + "epoch": 1.4298817774154098, + "grad_norm": 0.6637728214263916, + "learning_rate": 0.0008212647778230739, + "loss": 3.6402, + "step": 21045 + }, + { + "epoch": 1.4302214974860714, + "grad_norm": 0.7215205430984497, + "learning_rate": 0.0008212223128142411, + "loss": 3.8438, + "step": 21050 + }, + { + "epoch": 1.4305612175567333, + "grad_norm": 0.848948061466217, + "learning_rate": 0.0008211798478054083, + "loss": 3.74, + "step": 21055 + }, + { + "epoch": 1.4309009376273951, + "grad_norm": 0.9796594381332397, + "learning_rate": 0.0008211373827965757, + "loss": 3.6688, + "step": 21060 + }, + { + "epoch": 1.4312406576980568, + "grad_norm": 0.658288836479187, + "learning_rate": 0.0008210949177877429, + "loss": 3.6781, + "step": 21065 + }, + { + "epoch": 1.4315803777687186, + "grad_norm": 0.8367807865142822, + "learning_rate": 0.0008210524527789101, + "loss": 3.478, + "step": 21070 + }, + { + "epoch": 1.4319200978393805, + "grad_norm": 0.9000985026359558, + "learning_rate": 0.0008210099877700775, + "loss": 3.493, + "step": 21075 + }, + { + "epoch": 1.432259817910042, + "grad_norm": 1.0970194339752197, + "learning_rate": 0.0008209675227612448, + "loss": 3.5552, + "step": 21080 + }, + { + "epoch": 1.432599537980704, + "grad_norm": 1.3618448972702026, + "learning_rate": 0.000820925057752412, + "loss": 3.6232, + "step": 21085 + }, + { + "epoch": 1.4329392580513658, + "grad_norm": 0.9046831130981445, + "learning_rate": 0.0008208825927435794, + "loss": 3.5048, + "step": 21090 + }, + { + "epoch": 1.4332789781220274, + "grad_norm": 0.7496330142021179, + "learning_rate": 0.0008208401277347466, + "loss": 3.5051, + "step": 21095 + }, + { + "epoch": 1.4336186981926893, + "grad_norm": 1.044731616973877, + "learning_rate": 0.0008207976627259138, + "loss": 3.3247, + "step": 21100 + }, + { + "epoch": 1.4339584182633511, + "grad_norm": 0.8243861198425293, + "learning_rate": 0.0008207551977170811, + "loss": 3.4564, + "step": 21105 + }, + { + "epoch": 1.4342981383340128, + "grad_norm": 0.6760514378547668, + "learning_rate": 0.0008207127327082484, + "loss": 3.6602, + "step": 21110 + }, + { + "epoch": 1.4346378584046746, + "grad_norm": 0.792583167552948, + "learning_rate": 0.0008206702676994157, + "loss": 3.5704, + "step": 21115 + }, + { + "epoch": 1.4349775784753362, + "grad_norm": 0.9021437168121338, + "learning_rate": 0.000820627802690583, + "loss": 3.7041, + "step": 21120 + }, + { + "epoch": 1.435317298545998, + "grad_norm": 0.7091696262359619, + "learning_rate": 0.0008205853376817503, + "loss": 3.5509, + "step": 21125 + }, + { + "epoch": 1.43565701861666, + "grad_norm": 1.3741788864135742, + "learning_rate": 0.0008205428726729175, + "loss": 3.7019, + "step": 21130 + }, + { + "epoch": 1.4359967386873216, + "grad_norm": 0.8589696884155273, + "learning_rate": 0.0008205004076640848, + "loss": 3.6868, + "step": 21135 + }, + { + "epoch": 1.4363364587579834, + "grad_norm": 0.8531078100204468, + "learning_rate": 0.0008204579426552521, + "loss": 3.6759, + "step": 21140 + }, + { + "epoch": 1.436676178828645, + "grad_norm": 0.9978375434875488, + "learning_rate": 0.0008204154776464193, + "loss": 3.6129, + "step": 21145 + }, + { + "epoch": 1.437015898899307, + "grad_norm": 1.1907693147659302, + "learning_rate": 0.0008203730126375867, + "loss": 3.5617, + "step": 21150 + }, + { + "epoch": 1.4373556189699688, + "grad_norm": 0.7551723122596741, + "learning_rate": 0.000820330547628754, + "loss": 3.5167, + "step": 21155 + }, + { + "epoch": 1.4376953390406304, + "grad_norm": 0.9808022975921631, + "learning_rate": 0.0008202880826199212, + "loss": 3.513, + "step": 21160 + }, + { + "epoch": 1.4380350591112923, + "grad_norm": 1.1801602840423584, + "learning_rate": 0.0008202456176110885, + "loss": 3.6696, + "step": 21165 + }, + { + "epoch": 1.438374779181954, + "grad_norm": 4.158160209655762, + "learning_rate": 0.0008202031526022557, + "loss": 4.0072, + "step": 21170 + }, + { + "epoch": 1.4387144992526157, + "grad_norm": 1.271159052848816, + "learning_rate": 0.000820160687593423, + "loss": 3.6832, + "step": 21175 + }, + { + "epoch": 1.4390542193232776, + "grad_norm": 0.6326258778572083, + "learning_rate": 0.0008201182225845903, + "loss": 3.598, + "step": 21180 + }, + { + "epoch": 1.4393939393939394, + "grad_norm": 0.7248108983039856, + "learning_rate": 0.0008200757575757576, + "loss": 3.7626, + "step": 21185 + }, + { + "epoch": 1.439733659464601, + "grad_norm": 1.2428157329559326, + "learning_rate": 0.0008200332925669249, + "loss": 3.4822, + "step": 21190 + }, + { + "epoch": 1.440073379535263, + "grad_norm": 0.877465546131134, + "learning_rate": 0.0008199908275580922, + "loss": 3.6096, + "step": 21195 + }, + { + "epoch": 1.4404130996059248, + "grad_norm": 0.8538775444030762, + "learning_rate": 0.0008199483625492594, + "loss": 3.5902, + "step": 21200 + }, + { + "epoch": 1.4407528196765864, + "grad_norm": 1.1947253942489624, + "learning_rate": 0.0008199058975404266, + "loss": 3.9568, + "step": 21205 + }, + { + "epoch": 1.4410925397472483, + "grad_norm": 0.7671557068824768, + "learning_rate": 0.000819863432531594, + "loss": 3.5745, + "step": 21210 + }, + { + "epoch": 1.4414322598179101, + "grad_norm": 0.7160440683364868, + "learning_rate": 0.0008198209675227612, + "loss": 3.7933, + "step": 21215 + }, + { + "epoch": 1.4417719798885718, + "grad_norm": 0.8745556473731995, + "learning_rate": 0.0008197785025139285, + "loss": 3.7365, + "step": 21220 + }, + { + "epoch": 1.4421116999592336, + "grad_norm": 0.8915352821350098, + "learning_rate": 0.0008197360375050959, + "loss": 3.7093, + "step": 21225 + }, + { + "epoch": 1.4424514200298955, + "grad_norm": 1.2859364748001099, + "learning_rate": 0.0008196935724962631, + "loss": 3.6747, + "step": 21230 + }, + { + "epoch": 1.442791140100557, + "grad_norm": 1.118139386177063, + "learning_rate": 0.0008196511074874303, + "loss": 3.6023, + "step": 21235 + }, + { + "epoch": 1.443130860171219, + "grad_norm": 0.8791807889938354, + "learning_rate": 0.0008196086424785977, + "loss": 3.5695, + "step": 21240 + }, + { + "epoch": 1.4434705802418808, + "grad_norm": 0.8088740110397339, + "learning_rate": 0.0008195661774697649, + "loss": 3.7669, + "step": 21245 + }, + { + "epoch": 1.4438103003125424, + "grad_norm": 0.8321335911750793, + "learning_rate": 0.0008195237124609321, + "loss": 3.4593, + "step": 21250 + }, + { + "epoch": 1.4441500203832043, + "grad_norm": 0.9366874694824219, + "learning_rate": 0.0008194812474520996, + "loss": 3.7584, + "step": 21255 + }, + { + "epoch": 1.4444897404538661, + "grad_norm": 0.8399820923805237, + "learning_rate": 0.0008194387824432668, + "loss": 3.4822, + "step": 21260 + }, + { + "epoch": 1.4448294605245278, + "grad_norm": 0.8835268616676331, + "learning_rate": 0.000819396317434434, + "loss": 3.4674, + "step": 21265 + }, + { + "epoch": 1.4451691805951896, + "grad_norm": 0.7493513822555542, + "learning_rate": 0.0008193538524256013, + "loss": 3.6053, + "step": 21270 + }, + { + "epoch": 1.4455089006658515, + "grad_norm": 1.2194836139678955, + "learning_rate": 0.0008193113874167686, + "loss": 3.5036, + "step": 21275 + }, + { + "epoch": 1.445848620736513, + "grad_norm": 0.7187374234199524, + "learning_rate": 0.0008192689224079358, + "loss": 3.5007, + "step": 21280 + }, + { + "epoch": 1.446188340807175, + "grad_norm": 0.9049649238586426, + "learning_rate": 0.0008192264573991031, + "loss": 3.7869, + "step": 21285 + }, + { + "epoch": 1.4465280608778366, + "grad_norm": 0.7639972567558289, + "learning_rate": 0.0008191839923902705, + "loss": 3.9195, + "step": 21290 + }, + { + "epoch": 1.4468677809484984, + "grad_norm": 0.8590627908706665, + "learning_rate": 0.0008191415273814378, + "loss": 3.5509, + "step": 21295 + }, + { + "epoch": 1.4472075010191603, + "grad_norm": 0.7979729771614075, + "learning_rate": 0.000819099062372605, + "loss": 3.677, + "step": 21300 + }, + { + "epoch": 1.447547221089822, + "grad_norm": 0.9269406199455261, + "learning_rate": 0.0008190565973637722, + "loss": 3.5252, + "step": 21305 + }, + { + "epoch": 1.4478869411604838, + "grad_norm": 0.9275681972503662, + "learning_rate": 0.0008190141323549396, + "loss": 3.5098, + "step": 21310 + }, + { + "epoch": 1.4482266612311454, + "grad_norm": 0.8810122609138489, + "learning_rate": 0.0008189716673461068, + "loss": 3.5391, + "step": 21315 + }, + { + "epoch": 1.4485663813018073, + "grad_norm": 0.8511087894439697, + "learning_rate": 0.000818929202337274, + "loss": 3.5226, + "step": 21320 + }, + { + "epoch": 1.4489061013724691, + "grad_norm": 0.9180045127868652, + "learning_rate": 0.0008188867373284415, + "loss": 3.7932, + "step": 21325 + }, + { + "epoch": 1.4492458214431307, + "grad_norm": 0.8433115482330322, + "learning_rate": 0.0008188442723196087, + "loss": 3.5941, + "step": 21330 + }, + { + "epoch": 1.4495855415137926, + "grad_norm": 0.939642608165741, + "learning_rate": 0.0008188018073107759, + "loss": 3.8022, + "step": 21335 + }, + { + "epoch": 1.4499252615844545, + "grad_norm": 0.9234473705291748, + "learning_rate": 0.0008187593423019433, + "loss": 3.4986, + "step": 21340 + }, + { + "epoch": 1.450264981655116, + "grad_norm": 0.7766485810279846, + "learning_rate": 0.0008187168772931105, + "loss": 3.7319, + "step": 21345 + }, + { + "epoch": 1.450604701725778, + "grad_norm": 2.1646480560302734, + "learning_rate": 0.0008186744122842777, + "loss": 3.565, + "step": 21350 + }, + { + "epoch": 1.4509444217964398, + "grad_norm": 1.0277379751205444, + "learning_rate": 0.0008186319472754452, + "loss": 3.6447, + "step": 21355 + }, + { + "epoch": 1.4512841418671014, + "grad_norm": 0.9476875066757202, + "learning_rate": 0.0008185894822666124, + "loss": 3.6489, + "step": 21360 + }, + { + "epoch": 1.4516238619377633, + "grad_norm": 0.7743902802467346, + "learning_rate": 0.0008185470172577796, + "loss": 3.5181, + "step": 21365 + }, + { + "epoch": 1.4519635820084251, + "grad_norm": 0.7930931448936462, + "learning_rate": 0.0008185045522489469, + "loss": 3.9411, + "step": 21370 + }, + { + "epoch": 1.4523033020790868, + "grad_norm": 0.8143540620803833, + "learning_rate": 0.0008184620872401142, + "loss": 3.5559, + "step": 21375 + }, + { + "epoch": 1.4526430221497486, + "grad_norm": 1.3362376689910889, + "learning_rate": 0.0008184196222312814, + "loss": 3.3382, + "step": 21380 + }, + { + "epoch": 1.4529827422204105, + "grad_norm": 0.9951709508895874, + "learning_rate": 0.0008183771572224487, + "loss": 3.7575, + "step": 21385 + }, + { + "epoch": 1.453322462291072, + "grad_norm": 0.8831445574760437, + "learning_rate": 0.0008183346922136161, + "loss": 3.7641, + "step": 21390 + }, + { + "epoch": 1.453662182361734, + "grad_norm": 0.9894601106643677, + "learning_rate": 0.0008182922272047833, + "loss": 3.3111, + "step": 21395 + }, + { + "epoch": 1.4540019024323958, + "grad_norm": 0.8318936824798584, + "learning_rate": 0.0008182497621959506, + "loss": 3.4487, + "step": 21400 + }, + { + "epoch": 1.4543416225030574, + "grad_norm": 0.8804715871810913, + "learning_rate": 0.0008182072971871178, + "loss": 3.6597, + "step": 21405 + }, + { + "epoch": 1.4546813425737193, + "grad_norm": 0.8738488554954529, + "learning_rate": 0.0008181648321782851, + "loss": 3.6583, + "step": 21410 + }, + { + "epoch": 1.4550210626443811, + "grad_norm": 0.9786257147789001, + "learning_rate": 0.0008181223671694524, + "loss": 3.5933, + "step": 21415 + }, + { + "epoch": 1.4553607827150428, + "grad_norm": 0.9737648367881775, + "learning_rate": 0.0008180799021606196, + "loss": 3.4951, + "step": 21420 + }, + { + "epoch": 1.4557005027857046, + "grad_norm": 0.8127761483192444, + "learning_rate": 0.000818037437151787, + "loss": 3.4703, + "step": 21425 + }, + { + "epoch": 1.4560402228563665, + "grad_norm": 1.0311543941497803, + "learning_rate": 0.0008179949721429543, + "loss": 3.6113, + "step": 21430 + }, + { + "epoch": 1.456379942927028, + "grad_norm": 1.067214012145996, + "learning_rate": 0.0008179525071341215, + "loss": 3.4785, + "step": 21435 + }, + { + "epoch": 1.45671966299769, + "grad_norm": 0.9844973087310791, + "learning_rate": 0.0008179100421252888, + "loss": 3.7457, + "step": 21440 + }, + { + "epoch": 1.4570593830683518, + "grad_norm": 0.8678098917007446, + "learning_rate": 0.0008178675771164561, + "loss": 3.5818, + "step": 21445 + }, + { + "epoch": 1.4573991031390134, + "grad_norm": 0.7208032011985779, + "learning_rate": 0.0008178251121076233, + "loss": 3.8525, + "step": 21450 + }, + { + "epoch": 1.4577388232096753, + "grad_norm": 2.097007989883423, + "learning_rate": 0.0008177826470987905, + "loss": 3.7151, + "step": 21455 + }, + { + "epoch": 1.458078543280337, + "grad_norm": 0.911551833152771, + "learning_rate": 0.000817740182089958, + "loss": 3.4568, + "step": 21460 + }, + { + "epoch": 1.4584182633509988, + "grad_norm": 0.884174108505249, + "learning_rate": 0.0008176977170811252, + "loss": 3.7931, + "step": 21465 + }, + { + "epoch": 1.4587579834216606, + "grad_norm": 2.025204658508301, + "learning_rate": 0.0008176552520722924, + "loss": 3.4686, + "step": 21470 + }, + { + "epoch": 1.4590977034923223, + "grad_norm": 0.9569721817970276, + "learning_rate": 0.0008176127870634598, + "loss": 3.4445, + "step": 21475 + }, + { + "epoch": 1.4594374235629841, + "grad_norm": 0.9006094336509705, + "learning_rate": 0.000817570322054627, + "loss": 3.7951, + "step": 21480 + }, + { + "epoch": 1.4597771436336457, + "grad_norm": 1.010255217552185, + "learning_rate": 0.0008175278570457942, + "loss": 3.676, + "step": 21485 + }, + { + "epoch": 1.4601168637043076, + "grad_norm": 1.0475374460220337, + "learning_rate": 0.0008174853920369616, + "loss": 3.9279, + "step": 21490 + }, + { + "epoch": 1.4604565837749695, + "grad_norm": 1.5161429643630981, + "learning_rate": 0.0008174429270281289, + "loss": 3.8212, + "step": 21495 + }, + { + "epoch": 1.460796303845631, + "grad_norm": 0.7942125797271729, + "learning_rate": 0.0008174004620192961, + "loss": 3.4719, + "step": 21500 + }, + { + "epoch": 1.461136023916293, + "grad_norm": 1.2833255529403687, + "learning_rate": 0.0008173579970104634, + "loss": 3.5714, + "step": 21505 + }, + { + "epoch": 1.4614757439869548, + "grad_norm": 0.845539927482605, + "learning_rate": 0.0008173155320016307, + "loss": 3.8547, + "step": 21510 + }, + { + "epoch": 1.4618154640576164, + "grad_norm": 0.8955182433128357, + "learning_rate": 0.0008172730669927979, + "loss": 3.7075, + "step": 21515 + }, + { + "epoch": 1.4621551841282783, + "grad_norm": 1.1749134063720703, + "learning_rate": 0.0008172306019839652, + "loss": 3.4106, + "step": 21520 + }, + { + "epoch": 1.4624949041989401, + "grad_norm": 0.8004704117774963, + "learning_rate": 0.0008171881369751325, + "loss": 3.8432, + "step": 21525 + }, + { + "epoch": 1.4628346242696018, + "grad_norm": 1.074240803718567, + "learning_rate": 0.0008171456719662998, + "loss": 3.5803, + "step": 21530 + }, + { + "epoch": 1.4631743443402636, + "grad_norm": 2.3231027126312256, + "learning_rate": 0.0008171032069574671, + "loss": 3.7486, + "step": 21535 + }, + { + "epoch": 1.4635140644109255, + "grad_norm": 2.1603333950042725, + "learning_rate": 0.0008170607419486344, + "loss": 3.6802, + "step": 21540 + }, + { + "epoch": 1.463853784481587, + "grad_norm": 0.8936594128608704, + "learning_rate": 0.0008170182769398016, + "loss": 3.4125, + "step": 21545 + }, + { + "epoch": 1.464193504552249, + "grad_norm": 0.8998619914054871, + "learning_rate": 0.0008169758119309689, + "loss": 3.842, + "step": 21550 + }, + { + "epoch": 1.4645332246229108, + "grad_norm": 0.9896082878112793, + "learning_rate": 0.0008169333469221361, + "loss": 3.5288, + "step": 21555 + }, + { + "epoch": 1.4648729446935724, + "grad_norm": 0.7342512011528015, + "learning_rate": 0.0008168908819133034, + "loss": 3.701, + "step": 21560 + }, + { + "epoch": 1.4652126647642343, + "grad_norm": 0.7940000891685486, + "learning_rate": 0.0008168484169044708, + "loss": 3.8444, + "step": 21565 + }, + { + "epoch": 1.4655523848348961, + "grad_norm": 1.2127254009246826, + "learning_rate": 0.000816805951895638, + "loss": 3.7677, + "step": 21570 + }, + { + "epoch": 1.4658921049055578, + "grad_norm": 0.7867873311042786, + "learning_rate": 0.0008167634868868053, + "loss": 3.7963, + "step": 21575 + }, + { + "epoch": 1.4662318249762196, + "grad_norm": 0.9748193621635437, + "learning_rate": 0.0008167210218779726, + "loss": 3.9508, + "step": 21580 + }, + { + "epoch": 1.4665715450468815, + "grad_norm": 0.743140459060669, + "learning_rate": 0.0008166785568691398, + "loss": 3.6243, + "step": 21585 + }, + { + "epoch": 1.466911265117543, + "grad_norm": 0.7631969451904297, + "learning_rate": 0.000816636091860307, + "loss": 3.5035, + "step": 21590 + }, + { + "epoch": 1.467250985188205, + "grad_norm": 0.8958408832550049, + "learning_rate": 0.0008165936268514744, + "loss": 3.6533, + "step": 21595 + }, + { + "epoch": 1.4675907052588668, + "grad_norm": 0.8481122851371765, + "learning_rate": 0.0008165511618426417, + "loss": 3.7868, + "step": 21600 + }, + { + "epoch": 1.4679304253295284, + "grad_norm": 1.8762288093566895, + "learning_rate": 0.0008165086968338089, + "loss": 3.6979, + "step": 21605 + }, + { + "epoch": 1.4682701454001903, + "grad_norm": 1.0441898107528687, + "learning_rate": 0.0008164662318249763, + "loss": 3.7121, + "step": 21610 + }, + { + "epoch": 1.4686098654708521, + "grad_norm": 0.8871060609817505, + "learning_rate": 0.0008164237668161435, + "loss": 3.4828, + "step": 21615 + }, + { + "epoch": 1.4689495855415138, + "grad_norm": 0.9585723280906677, + "learning_rate": 0.0008163813018073107, + "loss": 3.5887, + "step": 21620 + }, + { + "epoch": 1.4692893056121756, + "grad_norm": 0.8343510627746582, + "learning_rate": 0.0008163388367984781, + "loss": 3.3523, + "step": 21625 + }, + { + "epoch": 1.4696290256828373, + "grad_norm": 0.7199278473854065, + "learning_rate": 0.0008162963717896453, + "loss": 3.3846, + "step": 21630 + }, + { + "epoch": 1.4699687457534991, + "grad_norm": 0.9149731993675232, + "learning_rate": 0.0008162539067808127, + "loss": 3.7782, + "step": 21635 + }, + { + "epoch": 1.470308465824161, + "grad_norm": 0.9607492089271545, + "learning_rate": 0.00081621144177198, + "loss": 3.3856, + "step": 21640 + }, + { + "epoch": 1.4706481858948226, + "grad_norm": 0.7553168535232544, + "learning_rate": 0.0008161689767631472, + "loss": 3.4264, + "step": 21645 + }, + { + "epoch": 1.4709879059654845, + "grad_norm": 0.8708540797233582, + "learning_rate": 0.0008161265117543145, + "loss": 3.6059, + "step": 21650 + }, + { + "epoch": 1.471327626036146, + "grad_norm": 0.8537190556526184, + "learning_rate": 0.0008160840467454817, + "loss": 3.5533, + "step": 21655 + }, + { + "epoch": 1.471667346106808, + "grad_norm": 0.8778212666511536, + "learning_rate": 0.000816041581736649, + "loss": 3.4298, + "step": 21660 + }, + { + "epoch": 1.4720070661774698, + "grad_norm": 0.9492563009262085, + "learning_rate": 0.0008159991167278163, + "loss": 3.7448, + "step": 21665 + }, + { + "epoch": 1.4723467862481314, + "grad_norm": 0.9947162866592407, + "learning_rate": 0.0008159566517189836, + "loss": 3.8784, + "step": 21670 + }, + { + "epoch": 1.4726865063187933, + "grad_norm": 0.7637332677841187, + "learning_rate": 0.0008159141867101509, + "loss": 3.5681, + "step": 21675 + }, + { + "epoch": 1.4730262263894551, + "grad_norm": 0.7588942646980286, + "learning_rate": 0.0008158717217013182, + "loss": 3.5862, + "step": 21680 + }, + { + "epoch": 1.4733659464601168, + "grad_norm": 0.7561425566673279, + "learning_rate": 0.0008158292566924854, + "loss": 3.6088, + "step": 21685 + }, + { + "epoch": 1.4737056665307786, + "grad_norm": 0.8163308501243591, + "learning_rate": 0.0008157867916836526, + "loss": 3.5612, + "step": 21690 + }, + { + "epoch": 1.4740453866014405, + "grad_norm": 0.904755175113678, + "learning_rate": 0.00081574432667482, + "loss": 3.7589, + "step": 21695 + }, + { + "epoch": 1.474385106672102, + "grad_norm": 0.7634891867637634, + "learning_rate": 0.0008157018616659872, + "loss": 3.5614, + "step": 21700 + }, + { + "epoch": 1.474724826742764, + "grad_norm": 0.8672685623168945, + "learning_rate": 0.0008156593966571545, + "loss": 3.5227, + "step": 21705 + }, + { + "epoch": 1.4750645468134258, + "grad_norm": 0.7154973745346069, + "learning_rate": 0.0008156169316483219, + "loss": 3.595, + "step": 21710 + }, + { + "epoch": 1.4754042668840874, + "grad_norm": 0.6683629751205444, + "learning_rate": 0.0008155744666394891, + "loss": 3.7091, + "step": 21715 + }, + { + "epoch": 1.4757439869547493, + "grad_norm": 1.1849454641342163, + "learning_rate": 0.0008155320016306563, + "loss": 3.5867, + "step": 21720 + }, + { + "epoch": 1.4760837070254111, + "grad_norm": 1.078500747680664, + "learning_rate": 0.0008154895366218237, + "loss": 3.6121, + "step": 21725 + }, + { + "epoch": 1.4764234270960728, + "grad_norm": 0.6529405117034912, + "learning_rate": 0.0008154470716129909, + "loss": 3.5905, + "step": 21730 + }, + { + "epoch": 1.4767631471667346, + "grad_norm": 0.7779830098152161, + "learning_rate": 0.0008154046066041581, + "loss": 3.5447, + "step": 21735 + }, + { + "epoch": 1.4771028672373965, + "grad_norm": 0.8573459982872009, + "learning_rate": 0.0008153621415953256, + "loss": 3.4908, + "step": 21740 + }, + { + "epoch": 1.477442587308058, + "grad_norm": 0.7864373326301575, + "learning_rate": 0.0008153196765864928, + "loss": 3.5203, + "step": 21745 + }, + { + "epoch": 1.47778230737872, + "grad_norm": 0.9067627787590027, + "learning_rate": 0.00081527721157766, + "loss": 3.6091, + "step": 21750 + }, + { + "epoch": 1.4781220274493818, + "grad_norm": 0.8581112027168274, + "learning_rate": 0.0008152347465688273, + "loss": 3.6542, + "step": 21755 + }, + { + "epoch": 1.4784617475200434, + "grad_norm": 0.7228453755378723, + "learning_rate": 0.0008151922815599946, + "loss": 3.7159, + "step": 21760 + }, + { + "epoch": 1.4788014675907053, + "grad_norm": 0.8718706965446472, + "learning_rate": 0.0008151498165511618, + "loss": 3.9812, + "step": 21765 + }, + { + "epoch": 1.4791411876613672, + "grad_norm": 0.7447729110717773, + "learning_rate": 0.0008151073515423291, + "loss": 3.5932, + "step": 21770 + }, + { + "epoch": 1.4794809077320288, + "grad_norm": 0.8128122091293335, + "learning_rate": 0.0008150648865334965, + "loss": 3.55, + "step": 21775 + }, + { + "epoch": 1.4798206278026906, + "grad_norm": 0.7510889172554016, + "learning_rate": 0.0008150224215246637, + "loss": 3.4492, + "step": 21780 + }, + { + "epoch": 1.4801603478733525, + "grad_norm": 0.6616800427436829, + "learning_rate": 0.000814979956515831, + "loss": 3.8175, + "step": 21785 + }, + { + "epoch": 1.4805000679440141, + "grad_norm": 4.121511936187744, + "learning_rate": 0.0008149374915069982, + "loss": 3.6536, + "step": 21790 + }, + { + "epoch": 1.480839788014676, + "grad_norm": 0.7207654118537903, + "learning_rate": 0.0008148950264981655, + "loss": 3.7986, + "step": 21795 + }, + { + "epoch": 1.4811795080853376, + "grad_norm": 0.9350873231887817, + "learning_rate": 0.0008148525614893328, + "loss": 3.6416, + "step": 21800 + }, + { + "epoch": 1.4815192281559995, + "grad_norm": 0.6798318028450012, + "learning_rate": 0.0008148100964805, + "loss": 3.8267, + "step": 21805 + }, + { + "epoch": 1.4818589482266613, + "grad_norm": 0.6271336674690247, + "learning_rate": 0.0008147676314716674, + "loss": 3.6608, + "step": 21810 + }, + { + "epoch": 1.482198668297323, + "grad_norm": 1.019269585609436, + "learning_rate": 0.0008147251664628347, + "loss": 3.835, + "step": 21815 + }, + { + "epoch": 1.4825383883679848, + "grad_norm": 0.7201334238052368, + "learning_rate": 0.0008146827014540019, + "loss": 3.4257, + "step": 21820 + }, + { + "epoch": 1.4828781084386464, + "grad_norm": 0.74535071849823, + "learning_rate": 0.0008146402364451692, + "loss": 3.8975, + "step": 21825 + }, + { + "epoch": 1.4832178285093083, + "grad_norm": 1.012795090675354, + "learning_rate": 0.0008145977714363365, + "loss": 3.67, + "step": 21830 + }, + { + "epoch": 1.4835575485799701, + "grad_norm": 1.1203855276107788, + "learning_rate": 0.0008145553064275037, + "loss": 3.7473, + "step": 21835 + }, + { + "epoch": 1.4838972686506318, + "grad_norm": 1.1410645246505737, + "learning_rate": 0.0008145128414186709, + "loss": 3.6254, + "step": 21840 + }, + { + "epoch": 1.4842369887212936, + "grad_norm": 0.8051386475563049, + "learning_rate": 0.0008144703764098384, + "loss": 3.6534, + "step": 21845 + }, + { + "epoch": 1.4845767087919555, + "grad_norm": 0.9999968409538269, + "learning_rate": 0.0008144279114010056, + "loss": 3.5057, + "step": 21850 + }, + { + "epoch": 1.484916428862617, + "grad_norm": 1.6259433031082153, + "learning_rate": 0.0008143854463921728, + "loss": 3.8802, + "step": 21855 + }, + { + "epoch": 1.485256148933279, + "grad_norm": 0.9761654138565063, + "learning_rate": 0.0008143429813833402, + "loss": 3.6888, + "step": 21860 + }, + { + "epoch": 1.4855958690039408, + "grad_norm": 0.9813851714134216, + "learning_rate": 0.0008143005163745074, + "loss": 3.8576, + "step": 21865 + }, + { + "epoch": 1.4859355890746024, + "grad_norm": 1.0830596685409546, + "learning_rate": 0.0008142580513656746, + "loss": 3.4483, + "step": 21870 + }, + { + "epoch": 1.4862753091452643, + "grad_norm": 0.8879253268241882, + "learning_rate": 0.000814215586356842, + "loss": 3.513, + "step": 21875 + }, + { + "epoch": 1.4866150292159261, + "grad_norm": 0.9507722854614258, + "learning_rate": 0.0008141731213480093, + "loss": 3.7061, + "step": 21880 + }, + { + "epoch": 1.4869547492865878, + "grad_norm": 0.7310115694999695, + "learning_rate": 0.0008141306563391765, + "loss": 3.3683, + "step": 21885 + }, + { + "epoch": 1.4872944693572496, + "grad_norm": 0.8615754842758179, + "learning_rate": 0.0008140881913303439, + "loss": 3.3743, + "step": 21890 + }, + { + "epoch": 1.4876341894279115, + "grad_norm": 1.008418321609497, + "learning_rate": 0.0008140457263215111, + "loss": 3.667, + "step": 21895 + }, + { + "epoch": 1.487973909498573, + "grad_norm": 1.0290607213974, + "learning_rate": 0.0008140032613126783, + "loss": 3.4433, + "step": 21900 + }, + { + "epoch": 1.488313629569235, + "grad_norm": 0.9199928045272827, + "learning_rate": 0.0008139607963038456, + "loss": 3.6276, + "step": 21905 + }, + { + "epoch": 1.4886533496398968, + "grad_norm": 0.7884456515312195, + "learning_rate": 0.0008139183312950129, + "loss": 3.2784, + "step": 21910 + }, + { + "epoch": 1.4889930697105584, + "grad_norm": 0.9601994752883911, + "learning_rate": 0.0008138758662861802, + "loss": 3.5129, + "step": 21915 + }, + { + "epoch": 1.4893327897812203, + "grad_norm": 0.9612324237823486, + "learning_rate": 0.0008138334012773475, + "loss": 3.4873, + "step": 21920 + }, + { + "epoch": 1.4896725098518822, + "grad_norm": 1.1331292390823364, + "learning_rate": 0.0008137909362685148, + "loss": 3.3827, + "step": 21925 + }, + { + "epoch": 1.4900122299225438, + "grad_norm": 0.9957653880119324, + "learning_rate": 0.000813748471259682, + "loss": 3.7574, + "step": 21930 + }, + { + "epoch": 1.4903519499932056, + "grad_norm": 0.7885183095932007, + "learning_rate": 0.0008137060062508493, + "loss": 3.7064, + "step": 21935 + }, + { + "epoch": 1.4906916700638675, + "grad_norm": 0.8184173107147217, + "learning_rate": 0.0008136635412420165, + "loss": 3.7318, + "step": 21940 + }, + { + "epoch": 1.4910313901345291, + "grad_norm": 0.8275867104530334, + "learning_rate": 0.0008136210762331838, + "loss": 3.5079, + "step": 21945 + }, + { + "epoch": 1.491371110205191, + "grad_norm": 1.0202633142471313, + "learning_rate": 0.0008135786112243512, + "loss": 3.6899, + "step": 21950 + }, + { + "epoch": 1.4917108302758528, + "grad_norm": 0.795844316482544, + "learning_rate": 0.0008135361462155184, + "loss": 3.6045, + "step": 21955 + }, + { + "epoch": 1.4920505503465145, + "grad_norm": 0.760252833366394, + "learning_rate": 0.0008134936812066857, + "loss": 3.629, + "step": 21960 + }, + { + "epoch": 1.4923902704171763, + "grad_norm": 0.8256102204322815, + "learning_rate": 0.000813451216197853, + "loss": 3.5065, + "step": 21965 + }, + { + "epoch": 1.492729990487838, + "grad_norm": 44.190853118896484, + "learning_rate": 0.0008134087511890202, + "loss": 3.5083, + "step": 21970 + }, + { + "epoch": 1.4930697105584998, + "grad_norm": 0.9282599091529846, + "learning_rate": 0.0008133662861801876, + "loss": 3.738, + "step": 21975 + }, + { + "epoch": 1.4934094306291616, + "grad_norm": 0.8702162504196167, + "learning_rate": 0.0008133238211713549, + "loss": 3.6939, + "step": 21980 + }, + { + "epoch": 1.4937491506998233, + "grad_norm": 1.1665629148483276, + "learning_rate": 0.0008132813561625221, + "loss": 3.3626, + "step": 21985 + }, + { + "epoch": 1.4940888707704851, + "grad_norm": 0.8936403393745422, + "learning_rate": 0.0008132388911536895, + "loss": 3.5156, + "step": 21990 + }, + { + "epoch": 1.4944285908411468, + "grad_norm": 0.8541828393936157, + "learning_rate": 0.0008131964261448567, + "loss": 3.639, + "step": 21995 + }, + { + "epoch": 1.4947683109118086, + "grad_norm": 0.8075734972953796, + "learning_rate": 0.0008131539611360239, + "loss": 3.7343, + "step": 22000 + }, + { + "epoch": 1.4951080309824705, + "grad_norm": 2.090625762939453, + "learning_rate": 0.0008131114961271912, + "loss": 3.7884, + "step": 22005 + }, + { + "epoch": 1.495447751053132, + "grad_norm": 0.9424380660057068, + "learning_rate": 0.0008130690311183585, + "loss": 3.5661, + "step": 22010 + }, + { + "epoch": 1.495787471123794, + "grad_norm": 0.8785465955734253, + "learning_rate": 0.0008130265661095258, + "loss": 3.6538, + "step": 22015 + }, + { + "epoch": 1.4961271911944558, + "grad_norm": 0.7870867848396301, + "learning_rate": 0.0008129841011006931, + "loss": 3.6814, + "step": 22020 + }, + { + "epoch": 1.4964669112651174, + "grad_norm": 0.862997829914093, + "learning_rate": 0.0008129416360918604, + "loss": 3.5377, + "step": 22025 + }, + { + "epoch": 1.4968066313357793, + "grad_norm": 0.8783271312713623, + "learning_rate": 0.0008128991710830276, + "loss": 3.3761, + "step": 22030 + }, + { + "epoch": 1.4971463514064411, + "grad_norm": 0.9661562442779541, + "learning_rate": 0.0008128567060741949, + "loss": 3.85, + "step": 22035 + }, + { + "epoch": 1.4974860714771028, + "grad_norm": 1.013491153717041, + "learning_rate": 0.0008128142410653621, + "loss": 3.5776, + "step": 22040 + }, + { + "epoch": 1.4978257915477646, + "grad_norm": 1.3911138772964478, + "learning_rate": 0.0008127717760565294, + "loss": 3.6554, + "step": 22045 + }, + { + "epoch": 1.4981655116184265, + "grad_norm": 0.8019959926605225, + "learning_rate": 0.0008127293110476968, + "loss": 3.6283, + "step": 22050 + }, + { + "epoch": 1.4985052316890881, + "grad_norm": 0.806312084197998, + "learning_rate": 0.000812686846038864, + "loss": 3.4803, + "step": 22055 + }, + { + "epoch": 1.49884495175975, + "grad_norm": 0.9606966376304626, + "learning_rate": 0.0008126443810300313, + "loss": 3.6998, + "step": 22060 + }, + { + "epoch": 1.4991846718304118, + "grad_norm": 0.7715378403663635, + "learning_rate": 0.0008126019160211986, + "loss": 3.9484, + "step": 22065 + }, + { + "epoch": 1.4995243919010735, + "grad_norm": 2.489417552947998, + "learning_rate": 0.0008125594510123658, + "loss": 3.4609, + "step": 22070 + }, + { + "epoch": 1.4998641119717353, + "grad_norm": 4.652007102966309, + "learning_rate": 0.000812516986003533, + "loss": 3.7212, + "step": 22075 + }, + { + "epoch": 1.5002038320423972, + "grad_norm": 0.8743653297424316, + "learning_rate": 0.0008124745209947004, + "loss": 3.7734, + "step": 22080 + }, + { + "epoch": 1.5005435521130588, + "grad_norm": 0.8106618523597717, + "learning_rate": 0.0008124320559858677, + "loss": 3.4748, + "step": 22085 + }, + { + "epoch": 1.5008832721837206, + "grad_norm": 0.9319062829017639, + "learning_rate": 0.0008123895909770349, + "loss": 3.5262, + "step": 22090 + }, + { + "epoch": 1.5012229922543825, + "grad_norm": 1.1792935132980347, + "learning_rate": 0.0008123471259682023, + "loss": 3.6257, + "step": 22095 + }, + { + "epoch": 1.5015627123250441, + "grad_norm": 0.8822785019874573, + "learning_rate": 0.0008123046609593695, + "loss": 3.4636, + "step": 22100 + }, + { + "epoch": 1.501902432395706, + "grad_norm": 1.0079622268676758, + "learning_rate": 0.0008122621959505367, + "loss": 3.7314, + "step": 22105 + }, + { + "epoch": 1.5022421524663678, + "grad_norm": 0.7648638486862183, + "learning_rate": 0.0008122197309417041, + "loss": 3.7662, + "step": 22110 + }, + { + "epoch": 1.5025818725370295, + "grad_norm": 0.9404330253601074, + "learning_rate": 0.0008121772659328713, + "loss": 3.6336, + "step": 22115 + }, + { + "epoch": 1.5029215926076913, + "grad_norm": 0.7802832126617432, + "learning_rate": 0.0008121348009240386, + "loss": 3.4796, + "step": 22120 + }, + { + "epoch": 1.5032613126783532, + "grad_norm": 0.8400198817253113, + "learning_rate": 0.000812092335915206, + "loss": 3.6483, + "step": 22125 + }, + { + "epoch": 1.5036010327490148, + "grad_norm": 0.7932274341583252, + "learning_rate": 0.0008120498709063732, + "loss": 3.4743, + "step": 22130 + }, + { + "epoch": 1.5039407528196764, + "grad_norm": 1.0990383625030518, + "learning_rate": 0.0008120074058975404, + "loss": 3.6033, + "step": 22135 + }, + { + "epoch": 1.5042804728903385, + "grad_norm": 0.9605428576469421, + "learning_rate": 0.0008119649408887077, + "loss": 3.6626, + "step": 22140 + }, + { + "epoch": 1.5046201929610001, + "grad_norm": 1.0197434425354004, + "learning_rate": 0.000811922475879875, + "loss": 3.7524, + "step": 22145 + }, + { + "epoch": 1.5049599130316618, + "grad_norm": 1.150770902633667, + "learning_rate": 0.0008118800108710422, + "loss": 3.7737, + "step": 22150 + }, + { + "epoch": 1.5052996331023238, + "grad_norm": 1.3884674310684204, + "learning_rate": 0.0008118375458622096, + "loss": 3.595, + "step": 22155 + }, + { + "epoch": 1.5056393531729855, + "grad_norm": 0.9235925078392029, + "learning_rate": 0.0008117950808533769, + "loss": 3.6805, + "step": 22160 + }, + { + "epoch": 1.505979073243647, + "grad_norm": 2.927175760269165, + "learning_rate": 0.0008117526158445441, + "loss": 3.782, + "step": 22165 + }, + { + "epoch": 1.506318793314309, + "grad_norm": 0.9864637851715088, + "learning_rate": 0.0008117101508357114, + "loss": 3.4585, + "step": 22170 + }, + { + "epoch": 1.5066585133849708, + "grad_norm": 1.06930673122406, + "learning_rate": 0.0008116676858268787, + "loss": 3.5276, + "step": 22175 + }, + { + "epoch": 1.5069982334556324, + "grad_norm": 0.8951318264007568, + "learning_rate": 0.0008116252208180459, + "loss": 3.5389, + "step": 22180 + }, + { + "epoch": 1.5073379535262943, + "grad_norm": 0.8548993468284607, + "learning_rate": 0.0008115827558092132, + "loss": 3.7338, + "step": 22185 + }, + { + "epoch": 1.5076776735969561, + "grad_norm": 1.6443783044815063, + "learning_rate": 0.0008115402908003805, + "loss": 3.3473, + "step": 22190 + }, + { + "epoch": 1.5080173936676178, + "grad_norm": 0.7876196503639221, + "learning_rate": 0.0008114978257915478, + "loss": 3.7573, + "step": 22195 + }, + { + "epoch": 1.5083571137382796, + "grad_norm": 0.731320858001709, + "learning_rate": 0.0008114553607827151, + "loss": 3.6303, + "step": 22200 + }, + { + "epoch": 1.5086968338089415, + "grad_norm": 1.1729212999343872, + "learning_rate": 0.0008114128957738823, + "loss": 3.7865, + "step": 22205 + }, + { + "epoch": 1.5090365538796031, + "grad_norm": 1.0513426065444946, + "learning_rate": 0.0008113704307650496, + "loss": 3.5633, + "step": 22210 + }, + { + "epoch": 1.509376273950265, + "grad_norm": 0.7319819927215576, + "learning_rate": 0.0008113279657562169, + "loss": 3.7131, + "step": 22215 + }, + { + "epoch": 1.5097159940209268, + "grad_norm": 0.888192355632782, + "learning_rate": 0.0008112855007473841, + "loss": 3.5013, + "step": 22220 + }, + { + "epoch": 1.5100557140915885, + "grad_norm": 0.9850166440010071, + "learning_rate": 0.0008112430357385515, + "loss": 3.5801, + "step": 22225 + }, + { + "epoch": 1.5103954341622503, + "grad_norm": 1.0577248334884644, + "learning_rate": 0.0008112005707297188, + "loss": 3.5202, + "step": 22230 + }, + { + "epoch": 1.5107351542329122, + "grad_norm": 1.1939570903778076, + "learning_rate": 0.000811158105720886, + "loss": 3.4592, + "step": 22235 + }, + { + "epoch": 1.5110748743035738, + "grad_norm": 0.7630566358566284, + "learning_rate": 0.0008111156407120532, + "loss": 3.6854, + "step": 22240 + }, + { + "epoch": 1.5114145943742356, + "grad_norm": 0.8147478699684143, + "learning_rate": 0.0008110731757032206, + "loss": 3.4123, + "step": 22245 + }, + { + "epoch": 1.5117543144448975, + "grad_norm": 0.8937450647354126, + "learning_rate": 0.0008110307106943878, + "loss": 3.744, + "step": 22250 + }, + { + "epoch": 1.5120940345155591, + "grad_norm": 0.9313421845436096, + "learning_rate": 0.000810988245685555, + "loss": 3.7744, + "step": 22255 + }, + { + "epoch": 1.512433754586221, + "grad_norm": 0.8605682253837585, + "learning_rate": 0.0008109457806767225, + "loss": 3.5162, + "step": 22260 + }, + { + "epoch": 1.5127734746568828, + "grad_norm": 0.7299453020095825, + "learning_rate": 0.0008109033156678897, + "loss": 3.581, + "step": 22265 + }, + { + "epoch": 1.5131131947275445, + "grad_norm": 0.8222042918205261, + "learning_rate": 0.0008108608506590569, + "loss": 3.6304, + "step": 22270 + }, + { + "epoch": 1.5134529147982063, + "grad_norm": 1.1752972602844238, + "learning_rate": 0.0008108183856502243, + "loss": 3.4502, + "step": 22275 + }, + { + "epoch": 1.5137926348688682, + "grad_norm": 1.8046059608459473, + "learning_rate": 0.0008107759206413915, + "loss": 3.4202, + "step": 22280 + }, + { + "epoch": 1.5141323549395298, + "grad_norm": 0.8829478025436401, + "learning_rate": 0.0008107334556325587, + "loss": 3.7306, + "step": 22285 + }, + { + "epoch": 1.5144720750101917, + "grad_norm": 0.8909691572189331, + "learning_rate": 0.000810690990623726, + "loss": 3.6082, + "step": 22290 + }, + { + "epoch": 1.5148117950808535, + "grad_norm": 1.0571426153182983, + "learning_rate": 0.0008106485256148934, + "loss": 3.4456, + "step": 22295 + }, + { + "epoch": 1.5151515151515151, + "grad_norm": 1.4158740043640137, + "learning_rate": 0.0008106060606060606, + "loss": 3.6156, + "step": 22300 + }, + { + "epoch": 1.5154912352221768, + "grad_norm": 0.9626006484031677, + "learning_rate": 0.0008105635955972279, + "loss": 3.8215, + "step": 22305 + }, + { + "epoch": 1.5158309552928388, + "grad_norm": 1.6804776191711426, + "learning_rate": 0.0008105211305883952, + "loss": 3.5964, + "step": 22310 + }, + { + "epoch": 1.5161706753635005, + "grad_norm": 0.9848420023918152, + "learning_rate": 0.0008104786655795624, + "loss": 3.7577, + "step": 22315 + }, + { + "epoch": 1.516510395434162, + "grad_norm": 0.817219614982605, + "learning_rate": 0.0008104362005707297, + "loss": 3.6377, + "step": 22320 + }, + { + "epoch": 1.5168501155048242, + "grad_norm": 0.9835785627365112, + "learning_rate": 0.000810393735561897, + "loss": 3.6006, + "step": 22325 + }, + { + "epoch": 1.5171898355754858, + "grad_norm": 1.1600946187973022, + "learning_rate": 0.0008103512705530644, + "loss": 3.5452, + "step": 22330 + }, + { + "epoch": 1.5175295556461474, + "grad_norm": 0.9806091785430908, + "learning_rate": 0.0008103088055442316, + "loss": 3.6534, + "step": 22335 + }, + { + "epoch": 1.5178692757168093, + "grad_norm": 1.0273557901382446, + "learning_rate": 0.0008102663405353988, + "loss": 3.6165, + "step": 22340 + }, + { + "epoch": 1.5182089957874711, + "grad_norm": 0.796470046043396, + "learning_rate": 0.0008102238755265662, + "loss": 3.5717, + "step": 22345 + }, + { + "epoch": 1.5185487158581328, + "grad_norm": 0.9569476842880249, + "learning_rate": 0.0008101814105177334, + "loss": 3.5608, + "step": 22350 + }, + { + "epoch": 1.5188884359287946, + "grad_norm": 1.0431102514266968, + "learning_rate": 0.0008101389455089006, + "loss": 3.5303, + "step": 22355 + }, + { + "epoch": 1.5192281559994565, + "grad_norm": 0.7515780925750732, + "learning_rate": 0.000810096480500068, + "loss": 3.6888, + "step": 22360 + }, + { + "epoch": 1.5195678760701181, + "grad_norm": 0.9718908071517944, + "learning_rate": 0.0008100540154912353, + "loss": 3.4985, + "step": 22365 + }, + { + "epoch": 1.51990759614078, + "grad_norm": 1.195468544960022, + "learning_rate": 0.0008100115504824025, + "loss": 3.5277, + "step": 22370 + }, + { + "epoch": 1.5202473162114418, + "grad_norm": 0.7781392335891724, + "learning_rate": 0.0008099690854735699, + "loss": 3.7776, + "step": 22375 + }, + { + "epoch": 1.5205870362821035, + "grad_norm": 0.8142921328544617, + "learning_rate": 0.0008099266204647371, + "loss": 3.7104, + "step": 22380 + }, + { + "epoch": 1.5209267563527653, + "grad_norm": 0.7960123419761658, + "learning_rate": 0.0008098841554559043, + "loss": 3.6966, + "step": 22385 + }, + { + "epoch": 1.5212664764234272, + "grad_norm": 1.0521873235702515, + "learning_rate": 0.0008098416904470716, + "loss": 3.6761, + "step": 22390 + }, + { + "epoch": 1.5216061964940888, + "grad_norm": 1.4839521646499634, + "learning_rate": 0.0008097992254382389, + "loss": 3.7996, + "step": 22395 + }, + { + "epoch": 1.5219459165647506, + "grad_norm": 0.9616371393203735, + "learning_rate": 0.0008097567604294062, + "loss": 3.75, + "step": 22400 + }, + { + "epoch": 1.5222856366354125, + "grad_norm": 1.0493491888046265, + "learning_rate": 0.0008097142954205735, + "loss": 3.6799, + "step": 22405 + }, + { + "epoch": 1.5226253567060741, + "grad_norm": 0.8211082816123962, + "learning_rate": 0.0008096718304117408, + "loss": 3.8219, + "step": 22410 + }, + { + "epoch": 1.522965076776736, + "grad_norm": 0.8053097724914551, + "learning_rate": 0.000809629365402908, + "loss": 3.6452, + "step": 22415 + }, + { + "epoch": 1.5233047968473978, + "grad_norm": 0.9654778838157654, + "learning_rate": 0.0008095869003940753, + "loss": 3.6568, + "step": 22420 + }, + { + "epoch": 1.5236445169180595, + "grad_norm": 0.8423956036567688, + "learning_rate": 0.0008095444353852425, + "loss": 3.752, + "step": 22425 + }, + { + "epoch": 1.5239842369887213, + "grad_norm": 13.296154022216797, + "learning_rate": 0.0008095019703764098, + "loss": 3.3864, + "step": 22430 + }, + { + "epoch": 1.5243239570593832, + "grad_norm": 1.3225758075714111, + "learning_rate": 0.0008094595053675772, + "loss": 3.5206, + "step": 22435 + }, + { + "epoch": 1.5246636771300448, + "grad_norm": 0.8590450882911682, + "learning_rate": 0.0008094170403587444, + "loss": 3.7009, + "step": 22440 + }, + { + "epoch": 1.5250033972007067, + "grad_norm": 0.8763025999069214, + "learning_rate": 0.0008093745753499117, + "loss": 3.6687, + "step": 22445 + }, + { + "epoch": 1.5253431172713685, + "grad_norm": 1.0289757251739502, + "learning_rate": 0.000809332110341079, + "loss": 3.4923, + "step": 22450 + }, + { + "epoch": 1.5256828373420301, + "grad_norm": 0.8347643613815308, + "learning_rate": 0.0008092896453322462, + "loss": 3.5813, + "step": 22455 + }, + { + "epoch": 1.526022557412692, + "grad_norm": 0.9816063046455383, + "learning_rate": 0.0008092471803234135, + "loss": 3.7323, + "step": 22460 + }, + { + "epoch": 1.5263622774833538, + "grad_norm": 0.7496846914291382, + "learning_rate": 0.0008092047153145808, + "loss": 3.6675, + "step": 22465 + }, + { + "epoch": 1.5267019975540155, + "grad_norm": 0.8033835291862488, + "learning_rate": 0.0008091622503057481, + "loss": 3.5864, + "step": 22470 + }, + { + "epoch": 1.527041717624677, + "grad_norm": 1.1840670108795166, + "learning_rate": 0.0008091197852969153, + "loss": 3.887, + "step": 22475 + }, + { + "epoch": 1.5273814376953392, + "grad_norm": 0.8926982879638672, + "learning_rate": 0.0008090773202880827, + "loss": 3.6433, + "step": 22480 + }, + { + "epoch": 1.5277211577660008, + "grad_norm": 0.8521348834037781, + "learning_rate": 0.0008090348552792499, + "loss": 3.8104, + "step": 22485 + }, + { + "epoch": 1.5280608778366624, + "grad_norm": 1.0022072792053223, + "learning_rate": 0.0008089923902704171, + "loss": 3.7542, + "step": 22490 + }, + { + "epoch": 1.5284005979073245, + "grad_norm": 0.9392140507698059, + "learning_rate": 0.0008089499252615845, + "loss": 3.6139, + "step": 22495 + }, + { + "epoch": 1.5287403179779862, + "grad_norm": 1.1814770698547363, + "learning_rate": 0.0008089074602527517, + "loss": 3.6327, + "step": 22500 + }, + { + "epoch": 1.5290800380486478, + "grad_norm": 0.7807260155677795, + "learning_rate": 0.000808864995243919, + "loss": 3.3598, + "step": 22505 + }, + { + "epoch": 1.5294197581193096, + "grad_norm": 0.8285924792289734, + "learning_rate": 0.0008088225302350864, + "loss": 3.5522, + "step": 22510 + }, + { + "epoch": 1.5297594781899715, + "grad_norm": 0.7884901762008667, + "learning_rate": 0.0008087800652262536, + "loss": 3.4582, + "step": 22515 + }, + { + "epoch": 1.5300991982606331, + "grad_norm": 1.1802457571029663, + "learning_rate": 0.0008087376002174208, + "loss": 3.2962, + "step": 22520 + }, + { + "epoch": 1.530438918331295, + "grad_norm": 0.7172727584838867, + "learning_rate": 0.0008086951352085881, + "loss": 3.7662, + "step": 22525 + }, + { + "epoch": 1.5307786384019568, + "grad_norm": 0.6931126713752747, + "learning_rate": 0.0008086526701997554, + "loss": 3.5946, + "step": 22530 + }, + { + "epoch": 1.5311183584726185, + "grad_norm": 1.9215635061264038, + "learning_rate": 0.0008086102051909226, + "loss": 3.783, + "step": 22535 + }, + { + "epoch": 1.5314580785432803, + "grad_norm": 0.9454769492149353, + "learning_rate": 0.00080856774018209, + "loss": 3.565, + "step": 22540 + }, + { + "epoch": 1.5317977986139422, + "grad_norm": 1.076009750366211, + "learning_rate": 0.0008085252751732573, + "loss": 3.5372, + "step": 22545 + }, + { + "epoch": 1.5321375186846038, + "grad_norm": 0.9424208998680115, + "learning_rate": 0.0008084828101644245, + "loss": 3.5803, + "step": 22550 + }, + { + "epoch": 1.5324772387552656, + "grad_norm": 0.7671887278556824, + "learning_rate": 0.0008084403451555918, + "loss": 3.7085, + "step": 22555 + }, + { + "epoch": 1.5328169588259275, + "grad_norm": 0.7315728664398193, + "learning_rate": 0.0008083978801467591, + "loss": 3.6087, + "step": 22560 + }, + { + "epoch": 1.5331566788965891, + "grad_norm": 1.044740080833435, + "learning_rate": 0.0008083554151379263, + "loss": 3.5571, + "step": 22565 + }, + { + "epoch": 1.533496398967251, + "grad_norm": 0.9926623702049255, + "learning_rate": 0.0008083129501290937, + "loss": 3.7927, + "step": 22570 + }, + { + "epoch": 1.5338361190379128, + "grad_norm": 0.8281869888305664, + "learning_rate": 0.000808270485120261, + "loss": 3.5189, + "step": 22575 + }, + { + "epoch": 1.5341758391085745, + "grad_norm": 1.3996199369430542, + "learning_rate": 0.0008082280201114282, + "loss": 3.6841, + "step": 22580 + }, + { + "epoch": 1.5345155591792363, + "grad_norm": 1.8536888360977173, + "learning_rate": 0.0008081855551025955, + "loss": 3.7203, + "step": 22585 + }, + { + "epoch": 1.5348552792498982, + "grad_norm": 1.1468416452407837, + "learning_rate": 0.0008081430900937627, + "loss": 3.9712, + "step": 22590 + }, + { + "epoch": 1.5351949993205598, + "grad_norm": 0.9454309344291687, + "learning_rate": 0.00080810062508493, + "loss": 3.9385, + "step": 22595 + }, + { + "epoch": 1.5355347193912217, + "grad_norm": 0.758994460105896, + "learning_rate": 0.0008080581600760973, + "loss": 3.5691, + "step": 22600 + }, + { + "epoch": 1.5358744394618835, + "grad_norm": 0.7544907331466675, + "learning_rate": 0.0008080156950672646, + "loss": 3.632, + "step": 22605 + }, + { + "epoch": 1.5362141595325451, + "grad_norm": 1.500591516494751, + "learning_rate": 0.0008079732300584319, + "loss": 3.5932, + "step": 22610 + }, + { + "epoch": 1.536553879603207, + "grad_norm": 0.7420836091041565, + "learning_rate": 0.0008079307650495992, + "loss": 3.9243, + "step": 22615 + }, + { + "epoch": 1.5368935996738688, + "grad_norm": 0.8717135190963745, + "learning_rate": 0.0008078883000407664, + "loss": 3.6164, + "step": 22620 + }, + { + "epoch": 1.5372333197445305, + "grad_norm": 1.1612741947174072, + "learning_rate": 0.0008078458350319336, + "loss": 3.5868, + "step": 22625 + }, + { + "epoch": 1.5375730398151923, + "grad_norm": 0.9587457776069641, + "learning_rate": 0.000807803370023101, + "loss": 3.6252, + "step": 22630 + }, + { + "epoch": 1.5379127598858542, + "grad_norm": 0.8992137312889099, + "learning_rate": 0.0008077609050142682, + "loss": 3.7543, + "step": 22635 + }, + { + "epoch": 1.5382524799565158, + "grad_norm": 0.785990834236145, + "learning_rate": 0.0008077184400054355, + "loss": 3.5381, + "step": 22640 + }, + { + "epoch": 1.5385922000271774, + "grad_norm": 2.1124203205108643, + "learning_rate": 0.0008076759749966029, + "loss": 3.6217, + "step": 22645 + }, + { + "epoch": 1.5389319200978395, + "grad_norm": 1.0307828187942505, + "learning_rate": 0.0008076335099877701, + "loss": 3.665, + "step": 22650 + }, + { + "epoch": 1.5392716401685012, + "grad_norm": 1.2666990756988525, + "learning_rate": 0.0008075910449789373, + "loss": 3.5478, + "step": 22655 + }, + { + "epoch": 1.5396113602391628, + "grad_norm": 0.8385536670684814, + "learning_rate": 0.0008075485799701047, + "loss": 3.6572, + "step": 22660 + }, + { + "epoch": 1.5399510803098249, + "grad_norm": 0.8621832728385925, + "learning_rate": 0.0008075061149612719, + "loss": 3.5894, + "step": 22665 + }, + { + "epoch": 1.5402908003804865, + "grad_norm": 0.6969078779220581, + "learning_rate": 0.0008074636499524392, + "loss": 3.4891, + "step": 22670 + }, + { + "epoch": 1.5406305204511481, + "grad_norm": 1.7248413562774658, + "learning_rate": 0.0008074211849436066, + "loss": 3.6176, + "step": 22675 + }, + { + "epoch": 1.54097024052181, + "grad_norm": 0.7885979413986206, + "learning_rate": 0.0008073787199347738, + "loss": 3.5696, + "step": 22680 + }, + { + "epoch": 1.5413099605924718, + "grad_norm": 1.5965110063552856, + "learning_rate": 0.0008073362549259411, + "loss": 3.7555, + "step": 22685 + }, + { + "epoch": 1.5416496806631335, + "grad_norm": 0.784146785736084, + "learning_rate": 0.0008072937899171083, + "loss": 3.3155, + "step": 22690 + }, + { + "epoch": 1.5419894007337953, + "grad_norm": 1.2182210683822632, + "learning_rate": 0.0008072513249082756, + "loss": 3.8097, + "step": 22695 + }, + { + "epoch": 1.5423291208044572, + "grad_norm": 0.9736958742141724, + "learning_rate": 0.0008072088598994429, + "loss": 3.558, + "step": 22700 + }, + { + "epoch": 1.5426688408751188, + "grad_norm": 0.5929139852523804, + "learning_rate": 0.0008071663948906101, + "loss": 3.628, + "step": 22705 + }, + { + "epoch": 1.5430085609457806, + "grad_norm": 1.2482004165649414, + "learning_rate": 0.0008071239298817775, + "loss": 3.5474, + "step": 22710 + }, + { + "epoch": 1.5433482810164425, + "grad_norm": 0.8275907635688782, + "learning_rate": 0.0008070814648729448, + "loss": 3.6432, + "step": 22715 + }, + { + "epoch": 1.5436880010871041, + "grad_norm": 0.9764878749847412, + "learning_rate": 0.000807038999864112, + "loss": 3.4316, + "step": 22720 + }, + { + "epoch": 1.544027721157766, + "grad_norm": 1.013533115386963, + "learning_rate": 0.0008069965348552792, + "loss": 3.6586, + "step": 22725 + }, + { + "epoch": 1.5443674412284278, + "grad_norm": 1.2373079061508179, + "learning_rate": 0.0008069540698464466, + "loss": 3.7482, + "step": 22730 + }, + { + "epoch": 1.5447071612990895, + "grad_norm": 0.7629674077033997, + "learning_rate": 0.0008069116048376138, + "loss": 3.5597, + "step": 22735 + }, + { + "epoch": 1.5450468813697513, + "grad_norm": 0.6417026519775391, + "learning_rate": 0.000806869139828781, + "loss": 3.5827, + "step": 22740 + }, + { + "epoch": 1.5453866014404132, + "grad_norm": 1.7202032804489136, + "learning_rate": 0.0008068266748199485, + "loss": 3.6263, + "step": 22745 + }, + { + "epoch": 1.5457263215110748, + "grad_norm": 0.6892850995063782, + "learning_rate": 0.0008067842098111157, + "loss": 3.7504, + "step": 22750 + }, + { + "epoch": 1.5460660415817367, + "grad_norm": 0.8357324004173279, + "learning_rate": 0.0008067417448022829, + "loss": 3.5557, + "step": 22755 + }, + { + "epoch": 1.5464057616523985, + "grad_norm": 0.876130998134613, + "learning_rate": 0.0008066992797934503, + "loss": 3.5252, + "step": 22760 + }, + { + "epoch": 1.5467454817230601, + "grad_norm": 0.8455385565757751, + "learning_rate": 0.0008066568147846175, + "loss": 3.4376, + "step": 22765 + }, + { + "epoch": 1.547085201793722, + "grad_norm": 0.862780749797821, + "learning_rate": 0.0008066143497757847, + "loss": 3.5089, + "step": 22770 + }, + { + "epoch": 1.5474249218643839, + "grad_norm": 0.9155977964401245, + "learning_rate": 0.000806571884766952, + "loss": 3.7824, + "step": 22775 + }, + { + "epoch": 1.5477646419350455, + "grad_norm": 1.0503411293029785, + "learning_rate": 0.0008065294197581194, + "loss": 3.5253, + "step": 22780 + }, + { + "epoch": 1.5481043620057073, + "grad_norm": 1.2907692193984985, + "learning_rate": 0.0008064869547492866, + "loss": 3.7616, + "step": 22785 + }, + { + "epoch": 1.5484440820763692, + "grad_norm": 0.9911143183708191, + "learning_rate": 0.0008064444897404539, + "loss": 3.6041, + "step": 22790 + }, + { + "epoch": 1.5487838021470308, + "grad_norm": 0.9603139758110046, + "learning_rate": 0.0008064020247316212, + "loss": 3.8913, + "step": 22795 + }, + { + "epoch": 1.5491235222176927, + "grad_norm": 0.9356276988983154, + "learning_rate": 0.0008063595597227884, + "loss": 3.3038, + "step": 22800 + }, + { + "epoch": 1.5494632422883545, + "grad_norm": 0.9622668027877808, + "learning_rate": 0.0008063170947139557, + "loss": 3.3986, + "step": 22805 + }, + { + "epoch": 1.5498029623590162, + "grad_norm": 0.8473086953163147, + "learning_rate": 0.000806274629705123, + "loss": 3.4507, + "step": 22810 + }, + { + "epoch": 1.5501426824296778, + "grad_norm": 0.9318678975105286, + "learning_rate": 0.0008062321646962903, + "loss": 3.6777, + "step": 22815 + }, + { + "epoch": 1.5504824025003399, + "grad_norm": 0.7008505463600159, + "learning_rate": 0.0008061896996874576, + "loss": 3.5634, + "step": 22820 + }, + { + "epoch": 1.5508221225710015, + "grad_norm": 0.9625828862190247, + "learning_rate": 0.0008061472346786248, + "loss": 3.5853, + "step": 22825 + }, + { + "epoch": 1.5511618426416631, + "grad_norm": 1.0728516578674316, + "learning_rate": 0.0008061047696697921, + "loss": 3.7753, + "step": 22830 + }, + { + "epoch": 1.5515015627123252, + "grad_norm": 0.9589217901229858, + "learning_rate": 0.0008060623046609594, + "loss": 3.7163, + "step": 22835 + }, + { + "epoch": 1.5518412827829868, + "grad_norm": 1.330429196357727, + "learning_rate": 0.0008060198396521266, + "loss": 3.6416, + "step": 22840 + }, + { + "epoch": 1.5521810028536485, + "grad_norm": 0.9001039266586304, + "learning_rate": 0.0008059773746432939, + "loss": 3.7091, + "step": 22845 + }, + { + "epoch": 1.5525207229243103, + "grad_norm": 0.9324317574501038, + "learning_rate": 0.0008059349096344613, + "loss": 3.6015, + "step": 22850 + }, + { + "epoch": 1.5528604429949722, + "grad_norm": 1.0072044134140015, + "learning_rate": 0.0008058924446256285, + "loss": 3.4328, + "step": 22855 + }, + { + "epoch": 1.5532001630656338, + "grad_norm": 1.4578256607055664, + "learning_rate": 0.0008058499796167958, + "loss": 3.6767, + "step": 22860 + }, + { + "epoch": 1.5535398831362957, + "grad_norm": 0.8969048261642456, + "learning_rate": 0.0008058075146079631, + "loss": 3.7119, + "step": 22865 + }, + { + "epoch": 1.5538796032069575, + "grad_norm": 1.6967047452926636, + "learning_rate": 0.0008057650495991303, + "loss": 3.5337, + "step": 22870 + }, + { + "epoch": 1.5542193232776191, + "grad_norm": 0.7999581098556519, + "learning_rate": 0.0008057225845902975, + "loss": 3.684, + "step": 22875 + }, + { + "epoch": 1.554559043348281, + "grad_norm": 0.7820613384246826, + "learning_rate": 0.0008056801195814649, + "loss": 3.6337, + "step": 22880 + }, + { + "epoch": 1.5548987634189428, + "grad_norm": 0.7836406230926514, + "learning_rate": 0.0008056376545726322, + "loss": 3.5899, + "step": 22885 + }, + { + "epoch": 1.5552384834896045, + "grad_norm": 1.4860467910766602, + "learning_rate": 0.0008055951895637994, + "loss": 3.6698, + "step": 22890 + }, + { + "epoch": 1.5555782035602663, + "grad_norm": 1.0068683624267578, + "learning_rate": 0.0008055527245549668, + "loss": 3.732, + "step": 22895 + }, + { + "epoch": 1.5559179236309282, + "grad_norm": 0.9511059522628784, + "learning_rate": 0.000805510259546134, + "loss": 3.5187, + "step": 22900 + }, + { + "epoch": 1.5562576437015898, + "grad_norm": 0.6855207681655884, + "learning_rate": 0.0008054677945373012, + "loss": 3.6602, + "step": 22905 + }, + { + "epoch": 1.5565973637722517, + "grad_norm": 0.8003137707710266, + "learning_rate": 0.0008054253295284686, + "loss": 3.6585, + "step": 22910 + }, + { + "epoch": 1.5569370838429135, + "grad_norm": 0.8115288615226746, + "learning_rate": 0.0008053828645196358, + "loss": 3.7257, + "step": 22915 + }, + { + "epoch": 1.5572768039135751, + "grad_norm": 1.423439621925354, + "learning_rate": 0.0008053403995108031, + "loss": 3.5754, + "step": 22920 + }, + { + "epoch": 1.557616523984237, + "grad_norm": 1.6858798265457153, + "learning_rate": 0.0008052979345019704, + "loss": 3.7301, + "step": 22925 + }, + { + "epoch": 1.5579562440548989, + "grad_norm": 0.7612317204475403, + "learning_rate": 0.0008052554694931377, + "loss": 3.5535, + "step": 22930 + }, + { + "epoch": 1.5582959641255605, + "grad_norm": 0.8697726130485535, + "learning_rate": 0.0008052130044843049, + "loss": 3.5451, + "step": 22935 + }, + { + "epoch": 1.5586356841962223, + "grad_norm": 0.6498113870620728, + "learning_rate": 0.0008051705394754722, + "loss": 3.7092, + "step": 22940 + }, + { + "epoch": 1.5589754042668842, + "grad_norm": 0.8934295177459717, + "learning_rate": 0.0008051280744666395, + "loss": 3.603, + "step": 22945 + }, + { + "epoch": 1.5593151243375458, + "grad_norm": 0.6928650736808777, + "learning_rate": 0.0008050856094578067, + "loss": 3.5475, + "step": 22950 + }, + { + "epoch": 1.5596548444082077, + "grad_norm": 0.8298101425170898, + "learning_rate": 0.0008050431444489741, + "loss": 3.6046, + "step": 22955 + }, + { + "epoch": 1.5599945644788695, + "grad_norm": 0.9894862174987793, + "learning_rate": 0.0008050006794401414, + "loss": 3.634, + "step": 22960 + }, + { + "epoch": 1.5603342845495312, + "grad_norm": 1.3444044589996338, + "learning_rate": 0.0008049582144313086, + "loss": 3.6892, + "step": 22965 + }, + { + "epoch": 1.560674004620193, + "grad_norm": 1.8900312185287476, + "learning_rate": 0.0008049157494224759, + "loss": 3.7891, + "step": 22970 + }, + { + "epoch": 1.5610137246908549, + "grad_norm": 0.7968646287918091, + "learning_rate": 0.0008048732844136431, + "loss": 3.3982, + "step": 22975 + }, + { + "epoch": 1.5613534447615165, + "grad_norm": 0.8287606835365295, + "learning_rate": 0.0008048308194048104, + "loss": 3.4049, + "step": 22980 + }, + { + "epoch": 1.5616931648321781, + "grad_norm": 1.1667578220367432, + "learning_rate": 0.0008047883543959777, + "loss": 3.5603, + "step": 22985 + }, + { + "epoch": 1.5620328849028402, + "grad_norm": 0.7978909015655518, + "learning_rate": 0.000804745889387145, + "loss": 3.8638, + "step": 22990 + }, + { + "epoch": 1.5623726049735018, + "grad_norm": 0.910234808921814, + "learning_rate": 0.0008047034243783123, + "loss": 3.5163, + "step": 22995 + }, + { + "epoch": 1.5627123250441635, + "grad_norm": 0.955987274646759, + "learning_rate": 0.0008046609593694796, + "loss": 3.5148, + "step": 23000 + }, + { + "epoch": 1.5630520451148255, + "grad_norm": 0.8619040846824646, + "learning_rate": 0.0008046184943606468, + "loss": 3.5625, + "step": 23005 + }, + { + "epoch": 1.5633917651854872, + "grad_norm": 0.8423206806182861, + "learning_rate": 0.0008045760293518142, + "loss": 3.4994, + "step": 23010 + }, + { + "epoch": 1.5637314852561488, + "grad_norm": 0.9177834391593933, + "learning_rate": 0.0008045335643429814, + "loss": 3.9807, + "step": 23015 + }, + { + "epoch": 1.5640712053268107, + "grad_norm": 1.0530548095703125, + "learning_rate": 0.0008044910993341486, + "loss": 3.7496, + "step": 23020 + }, + { + "epoch": 1.5644109253974725, + "grad_norm": 0.995296061038971, + "learning_rate": 0.000804448634325316, + "loss": 3.4806, + "step": 23025 + }, + { + "epoch": 1.5647506454681341, + "grad_norm": 1.233062505722046, + "learning_rate": 0.0008044061693164833, + "loss": 3.4626, + "step": 23030 + }, + { + "epoch": 1.565090365538796, + "grad_norm": 0.8018514513969421, + "learning_rate": 0.0008043637043076505, + "loss": 3.8425, + "step": 23035 + }, + { + "epoch": 1.5654300856094578, + "grad_norm": 0.806217610836029, + "learning_rate": 0.0008043212392988178, + "loss": 3.6307, + "step": 23040 + }, + { + "epoch": 1.5657698056801195, + "grad_norm": 0.8837574124336243, + "learning_rate": 0.0008042787742899851, + "loss": 3.5121, + "step": 23045 + }, + { + "epoch": 1.5661095257507813, + "grad_norm": 0.7693446278572083, + "learning_rate": 0.0008042363092811523, + "loss": 3.6844, + "step": 23050 + }, + { + "epoch": 1.5664492458214432, + "grad_norm": 1.0842504501342773, + "learning_rate": 0.0008041938442723196, + "loss": 3.6045, + "step": 23055 + }, + { + "epoch": 1.5667889658921048, + "grad_norm": 0.666137158870697, + "learning_rate": 0.000804151379263487, + "loss": 3.7396, + "step": 23060 + }, + { + "epoch": 1.5671286859627667, + "grad_norm": 0.9801398515701294, + "learning_rate": 0.0008041089142546542, + "loss": 3.8525, + "step": 23065 + }, + { + "epoch": 1.5674684060334285, + "grad_norm": 0.9786010384559631, + "learning_rate": 0.0008040664492458215, + "loss": 3.8985, + "step": 23070 + }, + { + "epoch": 1.5678081261040901, + "grad_norm": 1.0461175441741943, + "learning_rate": 0.0008040239842369887, + "loss": 3.5778, + "step": 23075 + }, + { + "epoch": 1.568147846174752, + "grad_norm": 0.7922382950782776, + "learning_rate": 0.000803981519228156, + "loss": 3.722, + "step": 23080 + }, + { + "epoch": 1.5684875662454139, + "grad_norm": 0.8120085000991821, + "learning_rate": 0.0008039390542193233, + "loss": 3.6013, + "step": 23085 + }, + { + "epoch": 1.5688272863160755, + "grad_norm": 0.8989419341087341, + "learning_rate": 0.0008038965892104905, + "loss": 3.6531, + "step": 23090 + }, + { + "epoch": 1.5691670063867373, + "grad_norm": 0.9200401306152344, + "learning_rate": 0.0008038541242016579, + "loss": 3.6092, + "step": 23095 + }, + { + "epoch": 1.5695067264573992, + "grad_norm": 0.7439908981323242, + "learning_rate": 0.0008038116591928252, + "loss": 3.6983, + "step": 23100 + }, + { + "epoch": 1.5698464465280608, + "grad_norm": 1.0173051357269287, + "learning_rate": 0.0008037691941839924, + "loss": 3.7334, + "step": 23105 + }, + { + "epoch": 1.5701861665987227, + "grad_norm": 0.6860843896865845, + "learning_rate": 0.0008037267291751596, + "loss": 3.6004, + "step": 23110 + }, + { + "epoch": 1.5705258866693845, + "grad_norm": 1.1148924827575684, + "learning_rate": 0.000803684264166327, + "loss": 3.6645, + "step": 23115 + }, + { + "epoch": 1.5708656067400462, + "grad_norm": 0.7811193466186523, + "learning_rate": 0.0008036417991574942, + "loss": 3.5053, + "step": 23120 + }, + { + "epoch": 1.571205326810708, + "grad_norm": 0.7312164306640625, + "learning_rate": 0.0008035993341486614, + "loss": 3.7217, + "step": 23125 + }, + { + "epoch": 1.5715450468813699, + "grad_norm": 0.7451415657997131, + "learning_rate": 0.0008035568691398289, + "loss": 3.7186, + "step": 23130 + }, + { + "epoch": 1.5718847669520315, + "grad_norm": 0.7153777480125427, + "learning_rate": 0.0008035144041309961, + "loss": 3.8071, + "step": 23135 + }, + { + "epoch": 1.5722244870226934, + "grad_norm": 1.0929551124572754, + "learning_rate": 0.0008034719391221633, + "loss": 3.6357, + "step": 23140 + }, + { + "epoch": 1.5725642070933552, + "grad_norm": 0.7249621748924255, + "learning_rate": 0.0008034294741133307, + "loss": 3.6326, + "step": 23145 + }, + { + "epoch": 1.5729039271640168, + "grad_norm": 2.8112776279449463, + "learning_rate": 0.0008033870091044979, + "loss": 3.7499, + "step": 23150 + }, + { + "epoch": 1.5732436472346785, + "grad_norm": 1.1247221231460571, + "learning_rate": 0.0008033445440956651, + "loss": 3.8276, + "step": 23155 + }, + { + "epoch": 1.5735833673053405, + "grad_norm": 0.8950327634811401, + "learning_rate": 0.0008033020790868326, + "loss": 3.5158, + "step": 23160 + }, + { + "epoch": 1.5739230873760022, + "grad_norm": 0.7450421452522278, + "learning_rate": 0.0008032596140779998, + "loss": 3.6468, + "step": 23165 + }, + { + "epoch": 1.5742628074466638, + "grad_norm": 1.545096516609192, + "learning_rate": 0.000803217149069167, + "loss": 3.7397, + "step": 23170 + }, + { + "epoch": 1.5746025275173259, + "grad_norm": 1.2321829795837402, + "learning_rate": 0.0008031746840603343, + "loss": 3.7426, + "step": 23175 + }, + { + "epoch": 1.5749422475879875, + "grad_norm": 1.1646597385406494, + "learning_rate": 0.0008031322190515016, + "loss": 3.8136, + "step": 23180 + }, + { + "epoch": 1.5752819676586491, + "grad_norm": 0.7760124802589417, + "learning_rate": 0.0008030897540426688, + "loss": 3.6527, + "step": 23185 + }, + { + "epoch": 1.575621687729311, + "grad_norm": 2.333987236022949, + "learning_rate": 0.0008030472890338361, + "loss": 3.7422, + "step": 23190 + }, + { + "epoch": 1.5759614077999728, + "grad_norm": 0.7689435482025146, + "learning_rate": 0.0008030048240250035, + "loss": 3.9986, + "step": 23195 + }, + { + "epoch": 1.5763011278706345, + "grad_norm": 0.9107362031936646, + "learning_rate": 0.0008029623590161707, + "loss": 3.3347, + "step": 23200 + }, + { + "epoch": 1.5766408479412963, + "grad_norm": 0.9274855256080627, + "learning_rate": 0.000802919894007338, + "loss": 3.6721, + "step": 23205 + }, + { + "epoch": 1.5769805680119582, + "grad_norm": 1.2673982381820679, + "learning_rate": 0.0008028774289985052, + "loss": 3.4871, + "step": 23210 + }, + { + "epoch": 1.5773202880826198, + "grad_norm": 1.9633291959762573, + "learning_rate": 0.0008028349639896725, + "loss": 3.4842, + "step": 23215 + }, + { + "epoch": 1.5776600081532817, + "grad_norm": 1.1052364110946655, + "learning_rate": 0.0008027924989808398, + "loss": 3.5478, + "step": 23220 + }, + { + "epoch": 1.5779997282239435, + "grad_norm": 0.7918109893798828, + "learning_rate": 0.000802750033972007, + "loss": 3.6834, + "step": 23225 + }, + { + "epoch": 1.5783394482946052, + "grad_norm": 1.271414041519165, + "learning_rate": 0.0008027075689631744, + "loss": 3.4974, + "step": 23230 + }, + { + "epoch": 1.578679168365267, + "grad_norm": 1.0142359733581543, + "learning_rate": 0.0008026651039543417, + "loss": 3.7193, + "step": 23235 + }, + { + "epoch": 1.5790188884359289, + "grad_norm": 0.9054991006851196, + "learning_rate": 0.0008026226389455089, + "loss": 3.5532, + "step": 23240 + }, + { + "epoch": 1.5793586085065905, + "grad_norm": 0.9334622621536255, + "learning_rate": 0.0008025801739366762, + "loss": 3.4267, + "step": 23245 + }, + { + "epoch": 1.5796983285772523, + "grad_norm": 0.7576130628585815, + "learning_rate": 0.0008025377089278435, + "loss": 3.657, + "step": 23250 + }, + { + "epoch": 1.5800380486479142, + "grad_norm": 0.8967666625976562, + "learning_rate": 0.0008024952439190107, + "loss": 3.7651, + "step": 23255 + }, + { + "epoch": 1.5803777687185758, + "grad_norm": 0.886087954044342, + "learning_rate": 0.0008024527789101779, + "loss": 3.6664, + "step": 23260 + }, + { + "epoch": 1.5807174887892377, + "grad_norm": 1.066821813583374, + "learning_rate": 0.0008024103139013454, + "loss": 3.6648, + "step": 23265 + }, + { + "epoch": 1.5810572088598995, + "grad_norm": 2.746002197265625, + "learning_rate": 0.0008023678488925126, + "loss": 3.7318, + "step": 23270 + }, + { + "epoch": 1.5813969289305612, + "grad_norm": 0.7996529340744019, + "learning_rate": 0.0008023253838836798, + "loss": 3.9165, + "step": 23275 + }, + { + "epoch": 1.581736649001223, + "grad_norm": 0.9263038039207458, + "learning_rate": 0.0008022829188748472, + "loss": 3.8473, + "step": 23280 + }, + { + "epoch": 1.5820763690718849, + "grad_norm": 1.0041640996932983, + "learning_rate": 0.0008022404538660144, + "loss": 3.6553, + "step": 23285 + }, + { + "epoch": 1.5824160891425465, + "grad_norm": 1.1628309488296509, + "learning_rate": 0.0008021979888571816, + "loss": 3.1556, + "step": 23290 + }, + { + "epoch": 1.5827558092132084, + "grad_norm": 1.1976736783981323, + "learning_rate": 0.000802155523848349, + "loss": 3.5365, + "step": 23295 + }, + { + "epoch": 1.5830955292838702, + "grad_norm": 0.9240379333496094, + "learning_rate": 0.0008021130588395163, + "loss": 3.6256, + "step": 23300 + }, + { + "epoch": 1.5834352493545318, + "grad_norm": 1.0488568544387817, + "learning_rate": 0.0008020705938306835, + "loss": 3.6563, + "step": 23305 + }, + { + "epoch": 1.5837749694251937, + "grad_norm": 0.8769745230674744, + "learning_rate": 0.0008020281288218509, + "loss": 3.4754, + "step": 23310 + }, + { + "epoch": 1.5841146894958555, + "grad_norm": 0.9473564028739929, + "learning_rate": 0.0008019856638130181, + "loss": 3.3395, + "step": 23315 + }, + { + "epoch": 1.5844544095665172, + "grad_norm": 0.7586822509765625, + "learning_rate": 0.0008019431988041853, + "loss": 3.3944, + "step": 23320 + }, + { + "epoch": 1.5847941296371788, + "grad_norm": 0.846683144569397, + "learning_rate": 0.0008019007337953526, + "loss": 3.5852, + "step": 23325 + }, + { + "epoch": 1.5851338497078409, + "grad_norm": 0.7387902140617371, + "learning_rate": 0.0008018582687865199, + "loss": 3.7006, + "step": 23330 + }, + { + "epoch": 1.5854735697785025, + "grad_norm": 0.8863653540611267, + "learning_rate": 0.0008018158037776872, + "loss": 3.7039, + "step": 23335 + }, + { + "epoch": 1.5858132898491641, + "grad_norm": 1.0591938495635986, + "learning_rate": 0.0008017733387688545, + "loss": 3.4639, + "step": 23340 + }, + { + "epoch": 1.5861530099198262, + "grad_norm": 0.8726239204406738, + "learning_rate": 0.0008017308737600218, + "loss": 3.5635, + "step": 23345 + }, + { + "epoch": 1.5864927299904878, + "grad_norm": 0.8549723625183105, + "learning_rate": 0.0008016884087511891, + "loss": 3.9902, + "step": 23350 + }, + { + "epoch": 1.5868324500611495, + "grad_norm": 1.5374599695205688, + "learning_rate": 0.0008016459437423563, + "loss": 3.6925, + "step": 23355 + }, + { + "epoch": 1.5871721701318113, + "grad_norm": 1.0196396112442017, + "learning_rate": 0.0008016034787335235, + "loss": 3.4372, + "step": 23360 + }, + { + "epoch": 1.5875118902024732, + "grad_norm": 0.7873160243034363, + "learning_rate": 0.0008015610137246909, + "loss": 3.7331, + "step": 23365 + }, + { + "epoch": 1.5878516102731348, + "grad_norm": 0.7569318413734436, + "learning_rate": 0.0008015185487158582, + "loss": 3.7141, + "step": 23370 + }, + { + "epoch": 1.5881913303437967, + "grad_norm": 0.7585000395774841, + "learning_rate": 0.0008014760837070254, + "loss": 3.5053, + "step": 23375 + }, + { + "epoch": 1.5885310504144585, + "grad_norm": 2.3421552181243896, + "learning_rate": 0.0008014336186981928, + "loss": 3.6515, + "step": 23380 + }, + { + "epoch": 1.5888707704851202, + "grad_norm": 1.0552809238433838, + "learning_rate": 0.00080139115368936, + "loss": 3.7835, + "step": 23385 + }, + { + "epoch": 1.589210490555782, + "grad_norm": 1.1268529891967773, + "learning_rate": 0.0008013486886805272, + "loss": 3.5088, + "step": 23390 + }, + { + "epoch": 1.5895502106264439, + "grad_norm": 0.914889931678772, + "learning_rate": 0.0008013062236716946, + "loss": 3.6977, + "step": 23395 + }, + { + "epoch": 1.5898899306971055, + "grad_norm": 1.0253980159759521, + "learning_rate": 0.0008012637586628618, + "loss": 3.5166, + "step": 23400 + }, + { + "epoch": 1.5902296507677673, + "grad_norm": 0.8665760159492493, + "learning_rate": 0.0008012212936540291, + "loss": 3.8314, + "step": 23405 + }, + { + "epoch": 1.5905693708384292, + "grad_norm": 0.936165988445282, + "learning_rate": 0.0008011788286451965, + "loss": 3.8013, + "step": 23410 + }, + { + "epoch": 1.5909090909090908, + "grad_norm": 0.7086284756660461, + "learning_rate": 0.0008011363636363637, + "loss": 3.6688, + "step": 23415 + }, + { + "epoch": 1.5912488109797527, + "grad_norm": 0.8857320547103882, + "learning_rate": 0.0008010938986275309, + "loss": 3.4335, + "step": 23420 + }, + { + "epoch": 1.5915885310504145, + "grad_norm": 1.0179429054260254, + "learning_rate": 0.0008010514336186982, + "loss": 3.7738, + "step": 23425 + }, + { + "epoch": 1.5919282511210762, + "grad_norm": 1.4826998710632324, + "learning_rate": 0.0008010089686098655, + "loss": 3.5936, + "step": 23430 + }, + { + "epoch": 1.592267971191738, + "grad_norm": 0.9242985248565674, + "learning_rate": 0.0008009665036010327, + "loss": 3.2941, + "step": 23435 + }, + { + "epoch": 1.5926076912623999, + "grad_norm": 1.3576202392578125, + "learning_rate": 0.0008009240385922001, + "loss": 3.2814, + "step": 23440 + }, + { + "epoch": 1.5929474113330615, + "grad_norm": 0.9796791076660156, + "learning_rate": 0.0008008815735833674, + "loss": 3.5618, + "step": 23445 + }, + { + "epoch": 1.5932871314037234, + "grad_norm": 0.8101518750190735, + "learning_rate": 0.0008008391085745346, + "loss": 3.833, + "step": 23450 + }, + { + "epoch": 1.5936268514743852, + "grad_norm": 0.9805428385734558, + "learning_rate": 0.0008007966435657019, + "loss": 3.4768, + "step": 23455 + }, + { + "epoch": 1.5939665715450468, + "grad_norm": 0.9744020104408264, + "learning_rate": 0.0008007541785568691, + "loss": 3.7696, + "step": 23460 + }, + { + "epoch": 1.5943062916157087, + "grad_norm": 0.6675955057144165, + "learning_rate": 0.0008007117135480364, + "loss": 3.4117, + "step": 23465 + }, + { + "epoch": 1.5946460116863705, + "grad_norm": 1.0053808689117432, + "learning_rate": 0.0008006692485392037, + "loss": 3.7145, + "step": 23470 + }, + { + "epoch": 1.5949857317570322, + "grad_norm": 0.9080557227134705, + "learning_rate": 0.000800626783530371, + "loss": 3.2919, + "step": 23475 + }, + { + "epoch": 1.595325451827694, + "grad_norm": 0.8128993511199951, + "learning_rate": 0.0008005843185215383, + "loss": 3.6347, + "step": 23480 + }, + { + "epoch": 1.5956651718983559, + "grad_norm": 1.1896580457687378, + "learning_rate": 0.0008005418535127056, + "loss": 3.5171, + "step": 23485 + }, + { + "epoch": 1.5960048919690175, + "grad_norm": 0.8397215008735657, + "learning_rate": 0.0008004993885038728, + "loss": 3.4268, + "step": 23490 + }, + { + "epoch": 1.5963446120396791, + "grad_norm": 0.8102633357048035, + "learning_rate": 0.00080045692349504, + "loss": 3.5101, + "step": 23495 + }, + { + "epoch": 1.5966843321103412, + "grad_norm": 1.5935547351837158, + "learning_rate": 0.0008004144584862074, + "loss": 3.7055, + "step": 23500 + }, + { + "epoch": 1.5970240521810029, + "grad_norm": 1.0748968124389648, + "learning_rate": 0.0008003719934773746, + "loss": 3.3178, + "step": 23505 + }, + { + "epoch": 1.5973637722516645, + "grad_norm": 51.05748748779297, + "learning_rate": 0.0008003295284685419, + "loss": 3.6339, + "step": 23510 + }, + { + "epoch": 1.5977034923223266, + "grad_norm": 1.2285377979278564, + "learning_rate": 0.0008002870634597093, + "loss": 3.7413, + "step": 23515 + }, + { + "epoch": 1.5980432123929882, + "grad_norm": 0.9696277379989624, + "learning_rate": 0.0008002445984508765, + "loss": 3.6362, + "step": 23520 + }, + { + "epoch": 1.5983829324636498, + "grad_norm": 1.0059013366699219, + "learning_rate": 0.0008002021334420437, + "loss": 3.7718, + "step": 23525 + }, + { + "epoch": 1.5987226525343117, + "grad_norm": 1.3645857572555542, + "learning_rate": 0.0008001596684332111, + "loss": 3.4715, + "step": 23530 + }, + { + "epoch": 1.5990623726049735, + "grad_norm": 0.8606336116790771, + "learning_rate": 0.0008001172034243783, + "loss": 3.3936, + "step": 23535 + }, + { + "epoch": 1.5994020926756352, + "grad_norm": 1.0308597087860107, + "learning_rate": 0.0008000747384155455, + "loss": 3.8029, + "step": 23540 + }, + { + "epoch": 1.599741812746297, + "grad_norm": 0.9579052925109863, + "learning_rate": 0.000800032273406713, + "loss": 3.4656, + "step": 23545 + }, + { + "epoch": 1.6000815328169589, + "grad_norm": 0.8312612771987915, + "learning_rate": 0.0007999898083978802, + "loss": 3.8852, + "step": 23550 + }, + { + "epoch": 1.6004212528876205, + "grad_norm": 0.7538661956787109, + "learning_rate": 0.0007999473433890474, + "loss": 3.641, + "step": 23555 + }, + { + "epoch": 1.6007609729582823, + "grad_norm": 1.1336020231246948, + "learning_rate": 0.0007999048783802147, + "loss": 3.6772, + "step": 23560 + }, + { + "epoch": 1.6011006930289442, + "grad_norm": 1.305827260017395, + "learning_rate": 0.000799862413371382, + "loss": 3.5811, + "step": 23565 + }, + { + "epoch": 1.6014404130996058, + "grad_norm": 0.8289303183555603, + "learning_rate": 0.0007998199483625492, + "loss": 3.7448, + "step": 23570 + }, + { + "epoch": 1.6017801331702677, + "grad_norm": 1.020234227180481, + "learning_rate": 0.0007997774833537165, + "loss": 3.6153, + "step": 23575 + }, + { + "epoch": 1.6021198532409295, + "grad_norm": 1.488905906677246, + "learning_rate": 0.0007997350183448839, + "loss": 3.4622, + "step": 23580 + }, + { + "epoch": 1.6024595733115912, + "grad_norm": 0.9271208643913269, + "learning_rate": 0.0007996925533360511, + "loss": 3.9867, + "step": 23585 + }, + { + "epoch": 1.602799293382253, + "grad_norm": 0.940216064453125, + "learning_rate": 0.0007996500883272184, + "loss": 3.7772, + "step": 23590 + }, + { + "epoch": 1.6031390134529149, + "grad_norm": 0.8570595383644104, + "learning_rate": 0.0007996076233183857, + "loss": 3.57, + "step": 23595 + }, + { + "epoch": 1.6034787335235765, + "grad_norm": 0.9865108132362366, + "learning_rate": 0.0007995651583095529, + "loss": 3.8086, + "step": 23600 + }, + { + "epoch": 1.6038184535942384, + "grad_norm": 0.835692822933197, + "learning_rate": 0.0007995226933007202, + "loss": 3.7469, + "step": 23605 + }, + { + "epoch": 1.6041581736649002, + "grad_norm": 0.7875521183013916, + "learning_rate": 0.0007994802282918874, + "loss": 3.6248, + "step": 23610 + }, + { + "epoch": 1.6044978937355618, + "grad_norm": 1.1715832948684692, + "learning_rate": 0.0007994377632830548, + "loss": 3.2905, + "step": 23615 + }, + { + "epoch": 1.6048376138062237, + "grad_norm": 1.0136688947677612, + "learning_rate": 0.0007993952982742221, + "loss": 3.5795, + "step": 23620 + }, + { + "epoch": 1.6051773338768855, + "grad_norm": 0.7369644045829773, + "learning_rate": 0.0007993528332653893, + "loss": 3.7182, + "step": 23625 + }, + { + "epoch": 1.6055170539475472, + "grad_norm": 0.7304967641830444, + "learning_rate": 0.0007993103682565566, + "loss": 3.6482, + "step": 23630 + }, + { + "epoch": 1.605856774018209, + "grad_norm": 0.883631706237793, + "learning_rate": 0.0007992679032477239, + "loss": 3.4561, + "step": 23635 + }, + { + "epoch": 1.6061964940888709, + "grad_norm": 0.8991934657096863, + "learning_rate": 0.0007992254382388911, + "loss": 3.7719, + "step": 23640 + }, + { + "epoch": 1.6065362141595325, + "grad_norm": 0.6554173231124878, + "learning_rate": 0.0007991829732300583, + "loss": 3.4212, + "step": 23645 + }, + { + "epoch": 1.6068759342301944, + "grad_norm": 1.0237566232681274, + "learning_rate": 0.0007991405082212258, + "loss": 3.6227, + "step": 23650 + }, + { + "epoch": 1.6072156543008562, + "grad_norm": 0.8674492239952087, + "learning_rate": 0.000799098043212393, + "loss": 3.6712, + "step": 23655 + }, + { + "epoch": 1.6075553743715179, + "grad_norm": 0.7446072697639465, + "learning_rate": 0.0007990555782035602, + "loss": 3.7225, + "step": 23660 + }, + { + "epoch": 1.6078950944421795, + "grad_norm": 0.732405960559845, + "learning_rate": 0.0007990131131947276, + "loss": 3.4367, + "step": 23665 + }, + { + "epoch": 1.6082348145128416, + "grad_norm": 0.8097562789916992, + "learning_rate": 0.0007989706481858948, + "loss": 3.4835, + "step": 23670 + }, + { + "epoch": 1.6085745345835032, + "grad_norm": 0.9072396755218506, + "learning_rate": 0.000798928183177062, + "loss": 3.4175, + "step": 23675 + }, + { + "epoch": 1.6089142546541648, + "grad_norm": 1.0019091367721558, + "learning_rate": 0.0007988857181682295, + "loss": 3.5773, + "step": 23680 + }, + { + "epoch": 1.609253974724827, + "grad_norm": 0.8127030730247498, + "learning_rate": 0.0007988432531593967, + "loss": 3.6151, + "step": 23685 + }, + { + "epoch": 1.6095936947954885, + "grad_norm": 0.6904447674751282, + "learning_rate": 0.000798800788150564, + "loss": 3.6993, + "step": 23690 + }, + { + "epoch": 1.6099334148661502, + "grad_norm": 0.8293530344963074, + "learning_rate": 0.0007987583231417313, + "loss": 3.2441, + "step": 23695 + }, + { + "epoch": 1.610273134936812, + "grad_norm": 0.8480821847915649, + "learning_rate": 0.0007987158581328985, + "loss": 3.6264, + "step": 23700 + }, + { + "epoch": 1.6106128550074739, + "grad_norm": 0.5983982682228088, + "learning_rate": 0.0007986733931240658, + "loss": 3.5845, + "step": 23705 + }, + { + "epoch": 1.6109525750781355, + "grad_norm": 0.757638156414032, + "learning_rate": 0.000798630928115233, + "loss": 3.6406, + "step": 23710 + }, + { + "epoch": 1.6112922951487973, + "grad_norm": 0.7428051829338074, + "learning_rate": 0.0007985884631064004, + "loss": 3.6964, + "step": 23715 + }, + { + "epoch": 1.6116320152194592, + "grad_norm": 0.8134360909461975, + "learning_rate": 0.0007985459980975677, + "loss": 3.8953, + "step": 23720 + }, + { + "epoch": 1.6119717352901208, + "grad_norm": 0.7559777498245239, + "learning_rate": 0.0007985035330887349, + "loss": 3.5425, + "step": 23725 + }, + { + "epoch": 1.6123114553607827, + "grad_norm": 0.9046006202697754, + "learning_rate": 0.0007984610680799022, + "loss": 3.8438, + "step": 23730 + }, + { + "epoch": 1.6126511754314445, + "grad_norm": 0.9858770966529846, + "learning_rate": 0.0007984186030710695, + "loss": 3.6159, + "step": 23735 + }, + { + "epoch": 1.6129908955021062, + "grad_norm": 0.7351667881011963, + "learning_rate": 0.0007983761380622367, + "loss": 3.3538, + "step": 23740 + }, + { + "epoch": 1.613330615572768, + "grad_norm": 0.8889757990837097, + "learning_rate": 0.000798333673053404, + "loss": 3.6714, + "step": 23745 + }, + { + "epoch": 1.6136703356434299, + "grad_norm": 0.8211771249771118, + "learning_rate": 0.0007982912080445714, + "loss": 3.7144, + "step": 23750 + }, + { + "epoch": 1.6140100557140915, + "grad_norm": 0.7915773987770081, + "learning_rate": 0.0007982487430357386, + "loss": 3.78, + "step": 23755 + }, + { + "epoch": 1.6143497757847534, + "grad_norm": 0.7084634304046631, + "learning_rate": 0.0007982062780269058, + "loss": 3.6234, + "step": 23760 + }, + { + "epoch": 1.6146894958554152, + "grad_norm": 0.8499560356140137, + "learning_rate": 0.0007981638130180732, + "loss": 3.5703, + "step": 23765 + }, + { + "epoch": 1.6150292159260768, + "grad_norm": 0.8613821268081665, + "learning_rate": 0.0007981213480092404, + "loss": 3.314, + "step": 23770 + }, + { + "epoch": 1.6153689359967387, + "grad_norm": 0.759160578250885, + "learning_rate": 0.0007980788830004076, + "loss": 3.6486, + "step": 23775 + }, + { + "epoch": 1.6157086560674006, + "grad_norm": 0.8055202960968018, + "learning_rate": 0.000798036417991575, + "loss": 3.6462, + "step": 23780 + }, + { + "epoch": 1.6160483761380622, + "grad_norm": 0.7821147441864014, + "learning_rate": 0.0007979939529827423, + "loss": 3.4886, + "step": 23785 + }, + { + "epoch": 1.616388096208724, + "grad_norm": 0.8772496581077576, + "learning_rate": 0.0007979514879739095, + "loss": 3.5972, + "step": 23790 + }, + { + "epoch": 1.6167278162793859, + "grad_norm": 1.0764036178588867, + "learning_rate": 0.0007979090229650769, + "loss": 3.3691, + "step": 23795 + }, + { + "epoch": 1.6170675363500475, + "grad_norm": 0.9633224606513977, + "learning_rate": 0.0007978665579562441, + "loss": 3.6717, + "step": 23800 + }, + { + "epoch": 1.6174072564207094, + "grad_norm": 0.7704792618751526, + "learning_rate": 0.0007978240929474113, + "loss": 3.4692, + "step": 23805 + }, + { + "epoch": 1.6177469764913712, + "grad_norm": 1.0446765422821045, + "learning_rate": 0.0007977816279385786, + "loss": 3.5989, + "step": 23810 + }, + { + "epoch": 1.6180866965620329, + "grad_norm": 0.7427679300308228, + "learning_rate": 0.0007977391629297459, + "loss": 3.8131, + "step": 23815 + }, + { + "epoch": 1.6184264166326947, + "grad_norm": 0.7944449782371521, + "learning_rate": 0.0007976966979209132, + "loss": 3.4794, + "step": 23820 + }, + { + "epoch": 1.6187661367033566, + "grad_norm": 0.8769071102142334, + "learning_rate": 0.0007976542329120805, + "loss": 3.646, + "step": 23825 + }, + { + "epoch": 1.6191058567740182, + "grad_norm": 0.9226222038269043, + "learning_rate": 0.0007976117679032478, + "loss": 3.6245, + "step": 23830 + }, + { + "epoch": 1.6194455768446798, + "grad_norm": 0.7831975817680359, + "learning_rate": 0.000797569302894415, + "loss": 3.5635, + "step": 23835 + }, + { + "epoch": 1.619785296915342, + "grad_norm": 1.2923822402954102, + "learning_rate": 0.0007975268378855823, + "loss": 3.4353, + "step": 23840 + }, + { + "epoch": 1.6201250169860035, + "grad_norm": 1.9435033798217773, + "learning_rate": 0.0007974843728767495, + "loss": 3.3165, + "step": 23845 + }, + { + "epoch": 1.6204647370566652, + "grad_norm": 0.7486109733581543, + "learning_rate": 0.0007974419078679168, + "loss": 3.7958, + "step": 23850 + }, + { + "epoch": 1.6208044571273272, + "grad_norm": 0.905042290687561, + "learning_rate": 0.0007973994428590842, + "loss": 3.7929, + "step": 23855 + }, + { + "epoch": 1.6211441771979889, + "grad_norm": 0.8462935090065002, + "learning_rate": 0.0007973569778502514, + "loss": 3.4852, + "step": 23860 + }, + { + "epoch": 1.6214838972686505, + "grad_norm": 0.7317774295806885, + "learning_rate": 0.0007973145128414187, + "loss": 3.4457, + "step": 23865 + }, + { + "epoch": 1.6218236173393124, + "grad_norm": 0.7766490578651428, + "learning_rate": 0.000797272047832586, + "loss": 3.4813, + "step": 23870 + }, + { + "epoch": 1.6221633374099742, + "grad_norm": 0.9738821387290955, + "learning_rate": 0.0007972295828237532, + "loss": 3.5736, + "step": 23875 + }, + { + "epoch": 1.6225030574806358, + "grad_norm": 0.8721387386322021, + "learning_rate": 0.0007971871178149205, + "loss": 3.7161, + "step": 23880 + }, + { + "epoch": 1.6228427775512977, + "grad_norm": 0.8313774466514587, + "learning_rate": 0.0007971446528060878, + "loss": 3.6662, + "step": 23885 + }, + { + "epoch": 1.6231824976219595, + "grad_norm": 1.0236966609954834, + "learning_rate": 0.0007971021877972551, + "loss": 3.7131, + "step": 23890 + }, + { + "epoch": 1.6235222176926212, + "grad_norm": 1.1392419338226318, + "learning_rate": 0.0007970597227884223, + "loss": 3.3905, + "step": 23895 + }, + { + "epoch": 1.623861937763283, + "grad_norm": 0.7545613646507263, + "learning_rate": 0.0007970172577795897, + "loss": 3.6117, + "step": 23900 + }, + { + "epoch": 1.6242016578339449, + "grad_norm": 0.7725341320037842, + "learning_rate": 0.0007969747927707569, + "loss": 3.6647, + "step": 23905 + }, + { + "epoch": 1.6245413779046065, + "grad_norm": 0.8368087410926819, + "learning_rate": 0.0007969323277619241, + "loss": 3.58, + "step": 23910 + }, + { + "epoch": 1.6248810979752684, + "grad_norm": 1.0467458963394165, + "learning_rate": 0.0007968898627530915, + "loss": 3.6169, + "step": 23915 + }, + { + "epoch": 1.6252208180459302, + "grad_norm": 0.7514470219612122, + "learning_rate": 0.0007968473977442587, + "loss": 3.5628, + "step": 23920 + }, + { + "epoch": 1.6255605381165918, + "grad_norm": 0.7083759307861328, + "learning_rate": 0.000796804932735426, + "loss": 3.662, + "step": 23925 + }, + { + "epoch": 1.6259002581872537, + "grad_norm": 0.7236567735671997, + "learning_rate": 0.0007967624677265934, + "loss": 3.5875, + "step": 23930 + }, + { + "epoch": 1.6262399782579156, + "grad_norm": 0.9053975939750671, + "learning_rate": 0.0007967200027177606, + "loss": 3.4448, + "step": 23935 + }, + { + "epoch": 1.6265796983285772, + "grad_norm": 0.6816624402999878, + "learning_rate": 0.0007966775377089278, + "loss": 3.5639, + "step": 23940 + }, + { + "epoch": 1.626919418399239, + "grad_norm": 1.0644323825836182, + "learning_rate": 0.0007966350727000951, + "loss": 3.6094, + "step": 23945 + }, + { + "epoch": 1.627259138469901, + "grad_norm": 0.911760687828064, + "learning_rate": 0.0007965926076912624, + "loss": 3.3631, + "step": 23950 + }, + { + "epoch": 1.6275988585405625, + "grad_norm": 1.0427849292755127, + "learning_rate": 0.0007965501426824296, + "loss": 3.6936, + "step": 23955 + }, + { + "epoch": 1.6279385786112244, + "grad_norm": 0.842400848865509, + "learning_rate": 0.000796507677673597, + "loss": 3.913, + "step": 23960 + }, + { + "epoch": 1.6282782986818862, + "grad_norm": 1.030480146408081, + "learning_rate": 0.0007964652126647643, + "loss": 3.618, + "step": 23965 + }, + { + "epoch": 1.6286180187525479, + "grad_norm": 0.7657821774482727, + "learning_rate": 0.0007964227476559315, + "loss": 3.7221, + "step": 23970 + }, + { + "epoch": 1.6289577388232097, + "grad_norm": 0.8917818665504456, + "learning_rate": 0.0007963802826470988, + "loss": 3.586, + "step": 23975 + }, + { + "epoch": 1.6292974588938716, + "grad_norm": 0.8667722344398499, + "learning_rate": 0.0007963378176382661, + "loss": 3.5886, + "step": 23980 + }, + { + "epoch": 1.6296371789645332, + "grad_norm": 0.9347371459007263, + "learning_rate": 0.0007962953526294333, + "loss": 3.895, + "step": 23985 + }, + { + "epoch": 1.629976899035195, + "grad_norm": 1.1037589311599731, + "learning_rate": 0.0007962528876206006, + "loss": 3.7443, + "step": 23990 + }, + { + "epoch": 1.630316619105857, + "grad_norm": 0.8355764746665955, + "learning_rate": 0.000796210422611768, + "loss": 3.6586, + "step": 23995 + }, + { + "epoch": 1.6306563391765185, + "grad_norm": 1.8527227640151978, + "learning_rate": 0.0007961679576029352, + "loss": 3.6009, + "step": 24000 + }, + { + "epoch": 1.6309960592471802, + "grad_norm": 1.0354634523391724, + "learning_rate": 0.0007961254925941025, + "loss": 3.4729, + "step": 24005 + }, + { + "epoch": 1.6313357793178422, + "grad_norm": 0.9577877521514893, + "learning_rate": 0.0007960830275852697, + "loss": 3.4487, + "step": 24010 + }, + { + "epoch": 1.6316754993885039, + "grad_norm": 0.8647545576095581, + "learning_rate": 0.000796040562576437, + "loss": 3.4622, + "step": 24015 + }, + { + "epoch": 1.6320152194591655, + "grad_norm": 0.9539531469345093, + "learning_rate": 0.0007959980975676043, + "loss": 3.7307, + "step": 24020 + }, + { + "epoch": 1.6323549395298276, + "grad_norm": 0.9936141967773438, + "learning_rate": 0.0007959556325587715, + "loss": 3.6429, + "step": 24025 + }, + { + "epoch": 1.6326946596004892, + "grad_norm": 0.7562452554702759, + "learning_rate": 0.000795913167549939, + "loss": 3.7228, + "step": 24030 + }, + { + "epoch": 1.6330343796711508, + "grad_norm": 0.6935852766036987, + "learning_rate": 0.0007958707025411062, + "loss": 3.6721, + "step": 24035 + }, + { + "epoch": 1.633374099741813, + "grad_norm": 0.8108484745025635, + "learning_rate": 0.0007958282375322734, + "loss": 3.5455, + "step": 24040 + }, + { + "epoch": 1.6337138198124745, + "grad_norm": 0.8274190425872803, + "learning_rate": 0.0007957857725234408, + "loss": 3.5713, + "step": 24045 + }, + { + "epoch": 1.6340535398831362, + "grad_norm": 1.0683006048202515, + "learning_rate": 0.000795743307514608, + "loss": 3.4565, + "step": 24050 + }, + { + "epoch": 1.634393259953798, + "grad_norm": 0.9223059415817261, + "learning_rate": 0.0007957008425057752, + "loss": 3.2271, + "step": 24055 + }, + { + "epoch": 1.6347329800244599, + "grad_norm": 1.1195131540298462, + "learning_rate": 0.0007956583774969425, + "loss": 3.6638, + "step": 24060 + }, + { + "epoch": 1.6350727000951215, + "grad_norm": 0.7707493901252747, + "learning_rate": 0.0007956159124881099, + "loss": 3.8254, + "step": 24065 + }, + { + "epoch": 1.6354124201657834, + "grad_norm": 1.104349970817566, + "learning_rate": 0.0007955734474792771, + "loss": 3.7439, + "step": 24070 + }, + { + "epoch": 1.6357521402364452, + "grad_norm": 0.6951128244400024, + "learning_rate": 0.0007955309824704444, + "loss": 3.6323, + "step": 24075 + }, + { + "epoch": 1.6360918603071068, + "grad_norm": 0.9689163565635681, + "learning_rate": 0.0007954885174616117, + "loss": 3.8728, + "step": 24080 + }, + { + "epoch": 1.6364315803777687, + "grad_norm": 0.917448103427887, + "learning_rate": 0.0007954460524527789, + "loss": 4.08, + "step": 24085 + }, + { + "epoch": 1.6367713004484306, + "grad_norm": 2.319807767868042, + "learning_rate": 0.0007954035874439462, + "loss": 3.6446, + "step": 24090 + }, + { + "epoch": 1.6371110205190922, + "grad_norm": 0.9266855120658875, + "learning_rate": 0.0007953611224351134, + "loss": 3.5803, + "step": 24095 + }, + { + "epoch": 1.637450740589754, + "grad_norm": 0.9528455138206482, + "learning_rate": 0.0007953186574262808, + "loss": 3.6649, + "step": 24100 + }, + { + "epoch": 1.637790460660416, + "grad_norm": 0.8594325184822083, + "learning_rate": 0.0007952761924174481, + "loss": 3.5385, + "step": 24105 + }, + { + "epoch": 1.6381301807310775, + "grad_norm": 0.935585081577301, + "learning_rate": 0.0007952337274086153, + "loss": 3.4835, + "step": 24110 + }, + { + "epoch": 1.6384699008017394, + "grad_norm": 0.9347378015518188, + "learning_rate": 0.0007951912623997826, + "loss": 3.5169, + "step": 24115 + }, + { + "epoch": 1.6388096208724012, + "grad_norm": 0.9570009708404541, + "learning_rate": 0.0007951487973909499, + "loss": 3.5181, + "step": 24120 + }, + { + "epoch": 1.6391493409430629, + "grad_norm": 0.8882436752319336, + "learning_rate": 0.0007951063323821171, + "loss": 3.5786, + "step": 24125 + }, + { + "epoch": 1.6394890610137247, + "grad_norm": 0.9020746946334839, + "learning_rate": 0.0007950638673732843, + "loss": 3.653, + "step": 24130 + }, + { + "epoch": 1.6398287810843866, + "grad_norm": 0.6291216015815735, + "learning_rate": 0.0007950214023644518, + "loss": 3.5078, + "step": 24135 + }, + { + "epoch": 1.6401685011550482, + "grad_norm": 0.9524011015892029, + "learning_rate": 0.000794978937355619, + "loss": 3.7782, + "step": 24140 + }, + { + "epoch": 1.64050822122571, + "grad_norm": 1.1547197103500366, + "learning_rate": 0.0007949364723467862, + "loss": 3.4975, + "step": 24145 + }, + { + "epoch": 1.640847941296372, + "grad_norm": 0.9637807607650757, + "learning_rate": 0.0007948940073379536, + "loss": 3.7987, + "step": 24150 + }, + { + "epoch": 1.6411876613670335, + "grad_norm": 0.928939700126648, + "learning_rate": 0.0007948515423291208, + "loss": 3.4597, + "step": 24155 + }, + { + "epoch": 1.6415273814376954, + "grad_norm": 0.9121633768081665, + "learning_rate": 0.000794809077320288, + "loss": 3.1828, + "step": 24160 + }, + { + "epoch": 1.6418671015083572, + "grad_norm": 2.2096896171569824, + "learning_rate": 0.0007947666123114554, + "loss": 3.6896, + "step": 24165 + }, + { + "epoch": 1.6422068215790189, + "grad_norm": 0.8442529439926147, + "learning_rate": 0.0007947241473026227, + "loss": 3.2535, + "step": 24170 + }, + { + "epoch": 1.6425465416496805, + "grad_norm": 0.8707417845726013, + "learning_rate": 0.0007946816822937899, + "loss": 3.5612, + "step": 24175 + }, + { + "epoch": 1.6428862617203426, + "grad_norm": 0.8863127827644348, + "learning_rate": 0.0007946392172849573, + "loss": 3.5251, + "step": 24180 + }, + { + "epoch": 1.6432259817910042, + "grad_norm": 1.0329296588897705, + "learning_rate": 0.0007945967522761245, + "loss": 3.7008, + "step": 24185 + }, + { + "epoch": 1.6435657018616658, + "grad_norm": 0.9131726622581482, + "learning_rate": 0.0007945542872672917, + "loss": 3.2332, + "step": 24190 + }, + { + "epoch": 1.643905421932328, + "grad_norm": 0.8410009741783142, + "learning_rate": 0.000794511822258459, + "loss": 3.5078, + "step": 24195 + }, + { + "epoch": 1.6442451420029895, + "grad_norm": 1.0598950386047363, + "learning_rate": 0.0007944693572496263, + "loss": 3.8795, + "step": 24200 + }, + { + "epoch": 1.6445848620736512, + "grad_norm": 0.836259126663208, + "learning_rate": 0.0007944268922407936, + "loss": 3.7727, + "step": 24205 + }, + { + "epoch": 1.6449245821443133, + "grad_norm": 0.9779436588287354, + "learning_rate": 0.0007943844272319609, + "loss": 3.4638, + "step": 24210 + }, + { + "epoch": 1.6452643022149749, + "grad_norm": 0.9968096613883972, + "learning_rate": 0.0007943419622231282, + "loss": 3.575, + "step": 24215 + }, + { + "epoch": 1.6456040222856365, + "grad_norm": 0.8281976580619812, + "learning_rate": 0.0007942994972142954, + "loss": 3.5921, + "step": 24220 + }, + { + "epoch": 1.6459437423562984, + "grad_norm": 1.3908189535140991, + "learning_rate": 0.0007942570322054627, + "loss": 3.5911, + "step": 24225 + }, + { + "epoch": 1.6462834624269602, + "grad_norm": 1.1226708889007568, + "learning_rate": 0.00079421456719663, + "loss": 3.6616, + "step": 24230 + }, + { + "epoch": 1.6466231824976219, + "grad_norm": 0.9824207425117493, + "learning_rate": 0.0007941721021877972, + "loss": 3.501, + "step": 24235 + }, + { + "epoch": 1.6469629025682837, + "grad_norm": 1.5835187435150146, + "learning_rate": 0.0007941296371789646, + "loss": 3.7243, + "step": 24240 + }, + { + "epoch": 1.6473026226389456, + "grad_norm": 0.8599257469177246, + "learning_rate": 0.0007940871721701318, + "loss": 3.4914, + "step": 24245 + }, + { + "epoch": 1.6476423427096072, + "grad_norm": 0.685516893863678, + "learning_rate": 0.0007940447071612991, + "loss": 3.7324, + "step": 24250 + }, + { + "epoch": 1.647982062780269, + "grad_norm": 1.0108672380447388, + "learning_rate": 0.0007940022421524664, + "loss": 3.4902, + "step": 24255 + }, + { + "epoch": 1.648321782850931, + "grad_norm": 0.907278835773468, + "learning_rate": 0.0007939597771436336, + "loss": 3.6695, + "step": 24260 + }, + { + "epoch": 1.6486615029215925, + "grad_norm": 0.8084161281585693, + "learning_rate": 0.0007939173121348009, + "loss": 3.6995, + "step": 24265 + }, + { + "epoch": 1.6490012229922544, + "grad_norm": 1.344904899597168, + "learning_rate": 0.0007938748471259683, + "loss": 3.5237, + "step": 24270 + }, + { + "epoch": 1.6493409430629162, + "grad_norm": 0.7513027191162109, + "learning_rate": 0.0007938323821171355, + "loss": 3.5218, + "step": 24275 + }, + { + "epoch": 1.6496806631335779, + "grad_norm": 0.9992022514343262, + "learning_rate": 0.0007937899171083028, + "loss": 3.8445, + "step": 24280 + }, + { + "epoch": 1.6500203832042397, + "grad_norm": 0.788295328617096, + "learning_rate": 0.0007937474520994701, + "loss": 3.4659, + "step": 24285 + }, + { + "epoch": 1.6503601032749016, + "grad_norm": 1.026893138885498, + "learning_rate": 0.0007937049870906373, + "loss": 3.482, + "step": 24290 + }, + { + "epoch": 1.6506998233455632, + "grad_norm": 0.9260799884796143, + "learning_rate": 0.0007936625220818045, + "loss": 3.8298, + "step": 24295 + }, + { + "epoch": 1.651039543416225, + "grad_norm": 0.8436681032180786, + "learning_rate": 0.0007936200570729719, + "loss": 3.4303, + "step": 24300 + }, + { + "epoch": 1.651379263486887, + "grad_norm": 0.7937162518501282, + "learning_rate": 0.0007935775920641392, + "loss": 3.5674, + "step": 24305 + }, + { + "epoch": 1.6517189835575485, + "grad_norm": 1.0211962461471558, + "learning_rate": 0.0007935351270553064, + "loss": 3.6888, + "step": 24310 + }, + { + "epoch": 1.6520587036282104, + "grad_norm": 0.7790192365646362, + "learning_rate": 0.0007934926620464738, + "loss": 3.6848, + "step": 24315 + }, + { + "epoch": 1.6523984236988722, + "grad_norm": 1.0438449382781982, + "learning_rate": 0.000793450197037641, + "loss": 3.4933, + "step": 24320 + }, + { + "epoch": 1.6527381437695339, + "grad_norm": 1.036513328552246, + "learning_rate": 0.0007934077320288082, + "loss": 3.5504, + "step": 24325 + }, + { + "epoch": 1.6530778638401957, + "grad_norm": 0.8681591749191284, + "learning_rate": 0.0007933652670199756, + "loss": 3.4734, + "step": 24330 + }, + { + "epoch": 1.6534175839108576, + "grad_norm": 1.102919101715088, + "learning_rate": 0.0007933228020111428, + "loss": 3.5633, + "step": 24335 + }, + { + "epoch": 1.6537573039815192, + "grad_norm": 0.9221736788749695, + "learning_rate": 0.0007932803370023101, + "loss": 3.4217, + "step": 24340 + }, + { + "epoch": 1.6540970240521808, + "grad_norm": 0.8063276410102844, + "learning_rate": 0.0007932378719934774, + "loss": 3.5708, + "step": 24345 + }, + { + "epoch": 1.654436744122843, + "grad_norm": 1.0694186687469482, + "learning_rate": 0.0007931954069846447, + "loss": 3.6861, + "step": 24350 + }, + { + "epoch": 1.6547764641935045, + "grad_norm": 1.0067116022109985, + "learning_rate": 0.0007931529419758119, + "loss": 3.5822, + "step": 24355 + }, + { + "epoch": 1.6551161842641662, + "grad_norm": 0.9750124216079712, + "learning_rate": 0.0007931104769669792, + "loss": 3.7232, + "step": 24360 + }, + { + "epoch": 1.6554559043348283, + "grad_norm": 0.8720276355743408, + "learning_rate": 0.0007930680119581465, + "loss": 3.5734, + "step": 24365 + }, + { + "epoch": 1.6557956244054899, + "grad_norm": 0.8657500147819519, + "learning_rate": 0.0007930255469493138, + "loss": 3.6362, + "step": 24370 + }, + { + "epoch": 1.6561353444761515, + "grad_norm": 0.8843256831169128, + "learning_rate": 0.0007929830819404811, + "loss": 3.5553, + "step": 24375 + }, + { + "epoch": 1.6564750645468136, + "grad_norm": 0.7166230082511902, + "learning_rate": 0.0007929406169316484, + "loss": 3.4389, + "step": 24380 + }, + { + "epoch": 1.6568147846174752, + "grad_norm": 0.9719016551971436, + "learning_rate": 0.0007928981519228157, + "loss": 3.836, + "step": 24385 + }, + { + "epoch": 1.6571545046881369, + "grad_norm": 0.9254566431045532, + "learning_rate": 0.0007928556869139829, + "loss": 3.5425, + "step": 24390 + }, + { + "epoch": 1.6574942247587987, + "grad_norm": 0.8964767456054688, + "learning_rate": 0.0007928132219051501, + "loss": 3.7351, + "step": 24395 + }, + { + "epoch": 1.6578339448294606, + "grad_norm": 0.7842639684677124, + "learning_rate": 0.0007927707568963175, + "loss": 3.7234, + "step": 24400 + }, + { + "epoch": 1.6581736649001222, + "grad_norm": 1.0432732105255127, + "learning_rate": 0.0007927282918874847, + "loss": 3.5309, + "step": 24405 + }, + { + "epoch": 1.658513384970784, + "grad_norm": 0.8915448188781738, + "learning_rate": 0.000792685826878652, + "loss": 3.6623, + "step": 24410 + }, + { + "epoch": 1.658853105041446, + "grad_norm": 1.486798882484436, + "learning_rate": 0.0007926433618698194, + "loss": 3.582, + "step": 24415 + }, + { + "epoch": 1.6591928251121075, + "grad_norm": 0.8449773788452148, + "learning_rate": 0.0007926008968609866, + "loss": 3.6036, + "step": 24420 + }, + { + "epoch": 1.6595325451827694, + "grad_norm": 0.9151912331581116, + "learning_rate": 0.0007925584318521538, + "loss": 3.6627, + "step": 24425 + }, + { + "epoch": 1.6598722652534312, + "grad_norm": 0.8255609273910522, + "learning_rate": 0.0007925159668433212, + "loss": 3.7531, + "step": 24430 + }, + { + "epoch": 1.6602119853240929, + "grad_norm": 0.8707639575004578, + "learning_rate": 0.0007924735018344884, + "loss": 3.6422, + "step": 24435 + }, + { + "epoch": 1.6605517053947547, + "grad_norm": 1.099408745765686, + "learning_rate": 0.0007924310368256556, + "loss": 3.5605, + "step": 24440 + }, + { + "epoch": 1.6608914254654166, + "grad_norm": 1.8394436836242676, + "learning_rate": 0.000792388571816823, + "loss": 3.5707, + "step": 24445 + }, + { + "epoch": 1.6612311455360782, + "grad_norm": 0.7939387559890747, + "learning_rate": 0.0007923461068079903, + "loss": 3.7819, + "step": 24450 + }, + { + "epoch": 1.66157086560674, + "grad_norm": 1.0110630989074707, + "learning_rate": 0.0007923036417991575, + "loss": 3.5795, + "step": 24455 + }, + { + "epoch": 1.661910585677402, + "grad_norm": 0.8170263767242432, + "learning_rate": 0.0007922611767903248, + "loss": 3.77, + "step": 24460 + }, + { + "epoch": 1.6622503057480635, + "grad_norm": 1.02459716796875, + "learning_rate": 0.0007922187117814921, + "loss": 3.609, + "step": 24465 + }, + { + "epoch": 1.6625900258187254, + "grad_norm": 1.1937389373779297, + "learning_rate": 0.0007921762467726593, + "loss": 3.8032, + "step": 24470 + }, + { + "epoch": 1.6629297458893872, + "grad_norm": 0.9913656115531921, + "learning_rate": 0.0007921337817638266, + "loss": 3.7429, + "step": 24475 + }, + { + "epoch": 1.6632694659600489, + "grad_norm": 0.7464697957038879, + "learning_rate": 0.000792091316754994, + "loss": 3.8337, + "step": 24480 + }, + { + "epoch": 1.6636091860307107, + "grad_norm": 1.0407745838165283, + "learning_rate": 0.0007920488517461612, + "loss": 3.5077, + "step": 24485 + }, + { + "epoch": 1.6639489061013726, + "grad_norm": 0.9377779364585876, + "learning_rate": 0.0007920063867373285, + "loss": 3.5667, + "step": 24490 + }, + { + "epoch": 1.6642886261720342, + "grad_norm": 1.397944450378418, + "learning_rate": 0.0007919639217284957, + "loss": 3.5412, + "step": 24495 + }, + { + "epoch": 1.664628346242696, + "grad_norm": 0.805711567401886, + "learning_rate": 0.000791921456719663, + "loss": 3.84, + "step": 24500 + }, + { + "epoch": 1.664968066313358, + "grad_norm": 0.9098036885261536, + "learning_rate": 0.0007918789917108303, + "loss": 3.5562, + "step": 24505 + }, + { + "epoch": 1.6653077863840196, + "grad_norm": 0.705431342124939, + "learning_rate": 0.0007918365267019975, + "loss": 3.5241, + "step": 24510 + }, + { + "epoch": 1.6656475064546812, + "grad_norm": 0.7368111610412598, + "learning_rate": 0.0007917940616931649, + "loss": 3.447, + "step": 24515 + }, + { + "epoch": 1.6659872265253433, + "grad_norm": 0.948047935962677, + "learning_rate": 0.0007917515966843322, + "loss": 3.4755, + "step": 24520 + }, + { + "epoch": 1.6663269465960049, + "grad_norm": 0.659034252166748, + "learning_rate": 0.0007917091316754994, + "loss": 3.764, + "step": 24525 + }, + { + "epoch": 1.6666666666666665, + "grad_norm": 0.8233663439750671, + "learning_rate": 0.0007916666666666666, + "loss": 3.7284, + "step": 24530 + }, + { + "epoch": 1.6670063867373286, + "grad_norm": 0.8553111553192139, + "learning_rate": 0.000791624201657834, + "loss": 3.7634, + "step": 24535 + }, + { + "epoch": 1.6673461068079902, + "grad_norm": 0.8617802858352661, + "learning_rate": 0.0007915817366490012, + "loss": 3.6164, + "step": 24540 + }, + { + "epoch": 1.6676858268786519, + "grad_norm": 1.92367684841156, + "learning_rate": 0.0007915392716401684, + "loss": 3.65, + "step": 24545 + }, + { + "epoch": 1.668025546949314, + "grad_norm": 1.0957508087158203, + "learning_rate": 0.0007914968066313359, + "loss": 3.5995, + "step": 24550 + }, + { + "epoch": 1.6683652670199756, + "grad_norm": 0.8375377058982849, + "learning_rate": 0.0007914543416225031, + "loss": 3.84, + "step": 24555 + }, + { + "epoch": 1.6687049870906372, + "grad_norm": 0.7838262915611267, + "learning_rate": 0.0007914118766136703, + "loss": 3.5936, + "step": 24560 + }, + { + "epoch": 1.669044707161299, + "grad_norm": 0.6568529009819031, + "learning_rate": 0.0007913694116048377, + "loss": 3.7518, + "step": 24565 + }, + { + "epoch": 1.669384427231961, + "grad_norm": 0.9234344363212585, + "learning_rate": 0.0007913269465960049, + "loss": 3.8272, + "step": 24570 + }, + { + "epoch": 1.6697241473026225, + "grad_norm": 0.8516789674758911, + "learning_rate": 0.0007912844815871721, + "loss": 3.5669, + "step": 24575 + }, + { + "epoch": 1.6700638673732844, + "grad_norm": 1.0142862796783447, + "learning_rate": 0.0007912420165783394, + "loss": 3.6851, + "step": 24580 + }, + { + "epoch": 1.6704035874439462, + "grad_norm": 0.9487866759300232, + "learning_rate": 0.0007911995515695068, + "loss": 3.6546, + "step": 24585 + }, + { + "epoch": 1.6707433075146079, + "grad_norm": 0.7897478938102722, + "learning_rate": 0.000791157086560674, + "loss": 3.4394, + "step": 24590 + }, + { + "epoch": 1.6710830275852697, + "grad_norm": 1.2754398584365845, + "learning_rate": 0.0007911146215518413, + "loss": 3.6394, + "step": 24595 + }, + { + "epoch": 1.6714227476559316, + "grad_norm": 0.8909595608711243, + "learning_rate": 0.0007910721565430086, + "loss": 3.4293, + "step": 24600 + }, + { + "epoch": 1.6717624677265932, + "grad_norm": 1.1150822639465332, + "learning_rate": 0.0007910296915341758, + "loss": 3.608, + "step": 24605 + }, + { + "epoch": 1.672102187797255, + "grad_norm": 0.9409980773925781, + "learning_rate": 0.0007909872265253431, + "loss": 3.7637, + "step": 24610 + }, + { + "epoch": 1.672441907867917, + "grad_norm": 1.0394517183303833, + "learning_rate": 0.0007909447615165104, + "loss": 3.7946, + "step": 24615 + }, + { + "epoch": 1.6727816279385785, + "grad_norm": 0.9402737617492676, + "learning_rate": 0.0007909022965076777, + "loss": 3.4, + "step": 24620 + }, + { + "epoch": 1.6731213480092404, + "grad_norm": 0.8615217208862305, + "learning_rate": 0.000790859831498845, + "loss": 3.6784, + "step": 24625 + }, + { + "epoch": 1.6734610680799022, + "grad_norm": 0.9433040022850037, + "learning_rate": 0.0007908173664900122, + "loss": 3.7704, + "step": 24630 + }, + { + "epoch": 1.6738007881505639, + "grad_norm": 0.7118695974349976, + "learning_rate": 0.0007907749014811795, + "loss": 3.6867, + "step": 24635 + }, + { + "epoch": 1.6741405082212257, + "grad_norm": 0.7933318614959717, + "learning_rate": 0.0007907324364723468, + "loss": 3.5965, + "step": 24640 + }, + { + "epoch": 1.6744802282918876, + "grad_norm": 0.7382358908653259, + "learning_rate": 0.000790689971463514, + "loss": 3.6258, + "step": 24645 + }, + { + "epoch": 1.6748199483625492, + "grad_norm": 0.7156182527542114, + "learning_rate": 0.0007906475064546813, + "loss": 3.4533, + "step": 24650 + }, + { + "epoch": 1.675159668433211, + "grad_norm": 2.3593029975891113, + "learning_rate": 0.0007906050414458487, + "loss": 3.7235, + "step": 24655 + }, + { + "epoch": 1.675499388503873, + "grad_norm": 0.8176868557929993, + "learning_rate": 0.0007905625764370159, + "loss": 3.5293, + "step": 24660 + }, + { + "epoch": 1.6758391085745346, + "grad_norm": 0.8658609390258789, + "learning_rate": 0.0007905201114281832, + "loss": 3.6861, + "step": 24665 + }, + { + "epoch": 1.6761788286451964, + "grad_norm": 1.0936497449874878, + "learning_rate": 0.0007904776464193505, + "loss": 3.4858, + "step": 24670 + }, + { + "epoch": 1.6765185487158583, + "grad_norm": 0.7908972501754761, + "learning_rate": 0.0007904351814105177, + "loss": 3.7076, + "step": 24675 + }, + { + "epoch": 1.67685826878652, + "grad_norm": 1.1707217693328857, + "learning_rate": 0.0007903927164016849, + "loss": 3.4585, + "step": 24680 + }, + { + "epoch": 1.6771979888571815, + "grad_norm": 0.922609806060791, + "learning_rate": 0.0007903502513928523, + "loss": 3.8582, + "step": 24685 + }, + { + "epoch": 1.6775377089278436, + "grad_norm": 0.8792892098426819, + "learning_rate": 0.0007903077863840196, + "loss": 3.6767, + "step": 24690 + }, + { + "epoch": 1.6778774289985052, + "grad_norm": 0.862585723400116, + "learning_rate": 0.0007902653213751868, + "loss": 3.709, + "step": 24695 + }, + { + "epoch": 1.6782171490691669, + "grad_norm": 0.9991616606712341, + "learning_rate": 0.0007902228563663542, + "loss": 3.6167, + "step": 24700 + }, + { + "epoch": 1.678556869139829, + "grad_norm": 0.8119308352470398, + "learning_rate": 0.0007901803913575214, + "loss": 3.6309, + "step": 24705 + }, + { + "epoch": 1.6788965892104906, + "grad_norm": 0.7031501531600952, + "learning_rate": 0.0007901379263486887, + "loss": 3.6361, + "step": 24710 + }, + { + "epoch": 1.6792363092811522, + "grad_norm": 0.8699216246604919, + "learning_rate": 0.000790095461339856, + "loss": 3.8782, + "step": 24715 + }, + { + "epoch": 1.6795760293518143, + "grad_norm": 0.8439099788665771, + "learning_rate": 0.0007900529963310232, + "loss": 3.6235, + "step": 24720 + }, + { + "epoch": 1.679915749422476, + "grad_norm": 1.0415631532669067, + "learning_rate": 0.0007900105313221906, + "loss": 3.42, + "step": 24725 + }, + { + "epoch": 1.6802554694931375, + "grad_norm": 1.0736912488937378, + "learning_rate": 0.0007899680663133579, + "loss": 3.6379, + "step": 24730 + }, + { + "epoch": 1.6805951895637994, + "grad_norm": 1.1264219284057617, + "learning_rate": 0.0007899256013045251, + "loss": 3.5814, + "step": 24735 + }, + { + "epoch": 1.6809349096344612, + "grad_norm": 0.7785015106201172, + "learning_rate": 0.0007898831362956924, + "loss": 3.5889, + "step": 24740 + }, + { + "epoch": 1.6812746297051229, + "grad_norm": 0.828604519367218, + "learning_rate": 0.0007898406712868596, + "loss": 3.8711, + "step": 24745 + }, + { + "epoch": 1.6816143497757847, + "grad_norm": 0.9275208115577698, + "learning_rate": 0.0007897982062780269, + "loss": 3.6378, + "step": 24750 + }, + { + "epoch": 1.6819540698464466, + "grad_norm": 0.9512251019477844, + "learning_rate": 0.0007897557412691942, + "loss": 3.5914, + "step": 24755 + }, + { + "epoch": 1.6822937899171082, + "grad_norm": 0.8005484938621521, + "learning_rate": 0.0007897132762603615, + "loss": 3.4447, + "step": 24760 + }, + { + "epoch": 1.68263350998777, + "grad_norm": 0.7597029209136963, + "learning_rate": 0.0007896708112515288, + "loss": 3.1614, + "step": 24765 + }, + { + "epoch": 1.682973230058432, + "grad_norm": 0.8426936268806458, + "learning_rate": 0.0007896283462426961, + "loss": 3.5529, + "step": 24770 + }, + { + "epoch": 1.6833129501290935, + "grad_norm": 0.813234269618988, + "learning_rate": 0.0007895858812338633, + "loss": 3.6662, + "step": 24775 + }, + { + "epoch": 1.6836526701997554, + "grad_norm": 0.9698591232299805, + "learning_rate": 0.0007895434162250305, + "loss": 3.5314, + "step": 24780 + }, + { + "epoch": 1.6839923902704172, + "grad_norm": 1.0913695096969604, + "learning_rate": 0.0007895009512161979, + "loss": 3.5848, + "step": 24785 + }, + { + "epoch": 1.6843321103410789, + "grad_norm": 0.6830967664718628, + "learning_rate": 0.0007894584862073651, + "loss": 3.4369, + "step": 24790 + }, + { + "epoch": 1.6846718304117407, + "grad_norm": 0.8661881685256958, + "learning_rate": 0.0007894160211985324, + "loss": 3.608, + "step": 24795 + }, + { + "epoch": 1.6850115504824026, + "grad_norm": 0.7323229312896729, + "learning_rate": 0.0007893735561896998, + "loss": 3.4296, + "step": 24800 + }, + { + "epoch": 1.6853512705530642, + "grad_norm": 1.3891593217849731, + "learning_rate": 0.000789331091180867, + "loss": 3.3683, + "step": 24805 + }, + { + "epoch": 1.685690990623726, + "grad_norm": 1.674864411354065, + "learning_rate": 0.0007892886261720342, + "loss": 3.199, + "step": 24810 + }, + { + "epoch": 1.686030710694388, + "grad_norm": 0.7719573378562927, + "learning_rate": 0.0007892461611632016, + "loss": 3.6464, + "step": 24815 + }, + { + "epoch": 1.6863704307650496, + "grad_norm": 1.207993507385254, + "learning_rate": 0.0007892036961543688, + "loss": 3.5438, + "step": 24820 + }, + { + "epoch": 1.6867101508357114, + "grad_norm": 1.1320436000823975, + "learning_rate": 0.000789161231145536, + "loss": 3.6096, + "step": 24825 + }, + { + "epoch": 1.6870498709063733, + "grad_norm": 0.8345465064048767, + "learning_rate": 0.0007891187661367035, + "loss": 3.8119, + "step": 24830 + }, + { + "epoch": 1.687389590977035, + "grad_norm": 0.7946677803993225, + "learning_rate": 0.0007890763011278707, + "loss": 3.5182, + "step": 24835 + }, + { + "epoch": 1.6877293110476967, + "grad_norm": 0.7344543933868408, + "learning_rate": 0.0007890338361190379, + "loss": 3.6934, + "step": 24840 + }, + { + "epoch": 1.6880690311183586, + "grad_norm": 1.0828746557235718, + "learning_rate": 0.0007889913711102052, + "loss": 3.6823, + "step": 24845 + }, + { + "epoch": 1.6884087511890202, + "grad_norm": 0.9360087513923645, + "learning_rate": 0.0007889489061013725, + "loss": 3.4799, + "step": 24850 + }, + { + "epoch": 1.6887484712596819, + "grad_norm": 0.7124038338661194, + "learning_rate": 0.0007889064410925397, + "loss": 3.4875, + "step": 24855 + }, + { + "epoch": 1.689088191330344, + "grad_norm": 0.7972016334533691, + "learning_rate": 0.0007888639760837071, + "loss": 3.7074, + "step": 24860 + }, + { + "epoch": 1.6894279114010056, + "grad_norm": 0.8951337933540344, + "learning_rate": 0.0007888215110748744, + "loss": 3.4117, + "step": 24865 + }, + { + "epoch": 1.6897676314716672, + "grad_norm": 0.8420272469520569, + "learning_rate": 0.0007887790460660416, + "loss": 3.4128, + "step": 24870 + }, + { + "epoch": 1.6901073515423293, + "grad_norm": 0.6531159281730652, + "learning_rate": 0.0007887365810572089, + "loss": 3.5992, + "step": 24875 + }, + { + "epoch": 1.690447071612991, + "grad_norm": 0.7660381197929382, + "learning_rate": 0.0007886941160483761, + "loss": 3.4279, + "step": 24880 + }, + { + "epoch": 1.6907867916836525, + "grad_norm": 0.7737146615982056, + "learning_rate": 0.0007886516510395434, + "loss": 3.3479, + "step": 24885 + }, + { + "epoch": 1.6911265117543146, + "grad_norm": 0.8405500054359436, + "learning_rate": 0.0007886091860307107, + "loss": 3.5609, + "step": 24890 + }, + { + "epoch": 1.6914662318249762, + "grad_norm": 0.9663074016571045, + "learning_rate": 0.000788566721021878, + "loss": 3.6688, + "step": 24895 + }, + { + "epoch": 1.6918059518956379, + "grad_norm": 0.8301217555999756, + "learning_rate": 0.0007885242560130453, + "loss": 3.431, + "step": 24900 + }, + { + "epoch": 1.6921456719662997, + "grad_norm": 0.8733827471733093, + "learning_rate": 0.0007884817910042126, + "loss": 3.4668, + "step": 24905 + }, + { + "epoch": 1.6924853920369616, + "grad_norm": 0.7199254035949707, + "learning_rate": 0.0007884393259953798, + "loss": 3.7519, + "step": 24910 + }, + { + "epoch": 1.6928251121076232, + "grad_norm": 0.8966487646102905, + "learning_rate": 0.000788396860986547, + "loss": 3.5421, + "step": 24915 + }, + { + "epoch": 1.693164832178285, + "grad_norm": 0.7796667218208313, + "learning_rate": 0.0007883543959777144, + "loss": 3.6116, + "step": 24920 + }, + { + "epoch": 1.693504552248947, + "grad_norm": 0.8656803965568542, + "learning_rate": 0.0007883119309688816, + "loss": 3.9229, + "step": 24925 + }, + { + "epoch": 1.6938442723196085, + "grad_norm": 0.9291998147964478, + "learning_rate": 0.0007882694659600489, + "loss": 3.2287, + "step": 24930 + }, + { + "epoch": 1.6941839923902704, + "grad_norm": 0.7663198113441467, + "learning_rate": 0.0007882270009512163, + "loss": 3.8697, + "step": 24935 + }, + { + "epoch": 1.6945237124609323, + "grad_norm": 0.9700276255607605, + "learning_rate": 0.0007881845359423835, + "loss": 3.5059, + "step": 24940 + }, + { + "epoch": 1.6948634325315939, + "grad_norm": 0.8098610639572144, + "learning_rate": 0.0007881420709335507, + "loss": 3.6999, + "step": 24945 + }, + { + "epoch": 1.6952031526022557, + "grad_norm": 1.127705693244934, + "learning_rate": 0.0007880996059247181, + "loss": 3.7003, + "step": 24950 + }, + { + "epoch": 1.6955428726729176, + "grad_norm": 0.8115562796592712, + "learning_rate": 0.0007880571409158853, + "loss": 3.6591, + "step": 24955 + }, + { + "epoch": 1.6958825927435792, + "grad_norm": 0.8117164969444275, + "learning_rate": 0.0007880146759070525, + "loss": 3.2934, + "step": 24960 + }, + { + "epoch": 1.696222312814241, + "grad_norm": 0.7299073338508606, + "learning_rate": 0.00078797221089822, + "loss": 3.4468, + "step": 24965 + }, + { + "epoch": 1.696562032884903, + "grad_norm": 0.7534223198890686, + "learning_rate": 0.0007879297458893872, + "loss": 3.7497, + "step": 24970 + }, + { + "epoch": 1.6969017529555646, + "grad_norm": 0.7504227757453918, + "learning_rate": 0.0007878872808805544, + "loss": 3.6254, + "step": 24975 + }, + { + "epoch": 1.6972414730262264, + "grad_norm": 0.9052001237869263, + "learning_rate": 0.0007878448158717217, + "loss": 3.4899, + "step": 24980 + }, + { + "epoch": 1.6975811930968883, + "grad_norm": 0.9598838090896606, + "learning_rate": 0.000787802350862889, + "loss": 3.3261, + "step": 24985 + }, + { + "epoch": 1.69792091316755, + "grad_norm": 0.8577004671096802, + "learning_rate": 0.0007877598858540562, + "loss": 3.5583, + "step": 24990 + }, + { + "epoch": 1.6982606332382117, + "grad_norm": 1.034424066543579, + "learning_rate": 0.0007877174208452235, + "loss": 3.7597, + "step": 24995 + }, + { + "epoch": 1.6986003533088736, + "grad_norm": 0.7616882920265198, + "learning_rate": 0.0007876749558363909, + "loss": 3.7198, + "step": 25000 + }, + { + "epoch": 1.6989400733795352, + "grad_norm": 0.7755090594291687, + "learning_rate": 0.0007876324908275581, + "loss": 3.5673, + "step": 25005 + }, + { + "epoch": 1.699279793450197, + "grad_norm": 0.9648683071136475, + "learning_rate": 0.0007875900258187254, + "loss": 3.7855, + "step": 25010 + }, + { + "epoch": 1.699619513520859, + "grad_norm": 0.9244015216827393, + "learning_rate": 0.0007875475608098927, + "loss": 3.436, + "step": 25015 + }, + { + "epoch": 1.6999592335915206, + "grad_norm": 0.868840217590332, + "learning_rate": 0.0007875050958010599, + "loss": 3.7161, + "step": 25020 + }, + { + "epoch": 1.7002989536621822, + "grad_norm": 0.9129986763000488, + "learning_rate": 0.0007874626307922272, + "loss": 3.6825, + "step": 25025 + }, + { + "epoch": 1.7006386737328443, + "grad_norm": 1.2084019184112549, + "learning_rate": 0.0007874201657833944, + "loss": 3.628, + "step": 25030 + }, + { + "epoch": 1.700978393803506, + "grad_norm": 0.8145015239715576, + "learning_rate": 0.0007873777007745618, + "loss": 3.256, + "step": 25035 + }, + { + "epoch": 1.7013181138741675, + "grad_norm": 0.8580953478813171, + "learning_rate": 0.0007873352357657291, + "loss": 3.724, + "step": 25040 + }, + { + "epoch": 1.7016578339448296, + "grad_norm": 0.9365824460983276, + "learning_rate": 0.0007872927707568963, + "loss": 3.6328, + "step": 25045 + }, + { + "epoch": 1.7019975540154912, + "grad_norm": 1.0616331100463867, + "learning_rate": 0.0007872503057480637, + "loss": 3.5926, + "step": 25050 + }, + { + "epoch": 1.7023372740861529, + "grad_norm": 1.2011743783950806, + "learning_rate": 0.0007872078407392309, + "loss": 3.5311, + "step": 25055 + }, + { + "epoch": 1.702676994156815, + "grad_norm": 0.9170088171958923, + "learning_rate": 0.0007871653757303981, + "loss": 3.5167, + "step": 25060 + }, + { + "epoch": 1.7030167142274766, + "grad_norm": 0.7168837785720825, + "learning_rate": 0.0007871229107215655, + "loss": 3.3605, + "step": 25065 + }, + { + "epoch": 1.7033564342981382, + "grad_norm": 0.7276991009712219, + "learning_rate": 0.0007870804457127328, + "loss": 3.6938, + "step": 25070 + }, + { + "epoch": 1.7036961543688, + "grad_norm": 0.8171664476394653, + "learning_rate": 0.0007870379807039, + "loss": 3.3423, + "step": 25075 + }, + { + "epoch": 1.704035874439462, + "grad_norm": 0.7256236672401428, + "learning_rate": 0.0007869955156950673, + "loss": 3.498, + "step": 25080 + }, + { + "epoch": 1.7043755945101235, + "grad_norm": 0.8292992115020752, + "learning_rate": 0.0007869530506862346, + "loss": 3.7187, + "step": 25085 + }, + { + "epoch": 1.7047153145807854, + "grad_norm": 0.7140609622001648, + "learning_rate": 0.0007869105856774018, + "loss": 3.6975, + "step": 25090 + }, + { + "epoch": 1.7050550346514473, + "grad_norm": 0.8792794942855835, + "learning_rate": 0.0007868681206685691, + "loss": 3.7577, + "step": 25095 + }, + { + "epoch": 1.7053947547221089, + "grad_norm": 0.7570245862007141, + "learning_rate": 0.0007868256556597364, + "loss": 3.6283, + "step": 25100 + }, + { + "epoch": 1.7057344747927707, + "grad_norm": 0.9570472836494446, + "learning_rate": 0.0007867831906509037, + "loss": 3.6202, + "step": 25105 + }, + { + "epoch": 1.7060741948634326, + "grad_norm": 0.6486772894859314, + "learning_rate": 0.000786740725642071, + "loss": 3.6275, + "step": 25110 + }, + { + "epoch": 1.7064139149340942, + "grad_norm": 0.8136416077613831, + "learning_rate": 0.0007866982606332383, + "loss": 3.4998, + "step": 25115 + }, + { + "epoch": 1.706753635004756, + "grad_norm": 0.9321552515029907, + "learning_rate": 0.0007866557956244055, + "loss": 3.4543, + "step": 25120 + }, + { + "epoch": 1.707093355075418, + "grad_norm": 1.0802842378616333, + "learning_rate": 0.0007866133306155728, + "loss": 3.5594, + "step": 25125 + }, + { + "epoch": 1.7074330751460796, + "grad_norm": 0.7569277882575989, + "learning_rate": 0.00078657086560674, + "loss": 3.8712, + "step": 25130 + }, + { + "epoch": 1.7077727952167414, + "grad_norm": 0.8030095100402832, + "learning_rate": 0.0007865284005979073, + "loss": 3.4874, + "step": 25135 + }, + { + "epoch": 1.7081125152874033, + "grad_norm": 0.8289168477058411, + "learning_rate": 0.0007864859355890747, + "loss": 3.7909, + "step": 25140 + }, + { + "epoch": 1.708452235358065, + "grad_norm": 1.2423521280288696, + "learning_rate": 0.0007864434705802419, + "loss": 3.5996, + "step": 25145 + }, + { + "epoch": 1.7087919554287267, + "grad_norm": 0.8918646574020386, + "learning_rate": 0.0007864010055714092, + "loss": 3.9593, + "step": 25150 + }, + { + "epoch": 1.7091316754993886, + "grad_norm": 1.0012123584747314, + "learning_rate": 0.0007863585405625765, + "loss": 3.7475, + "step": 25155 + }, + { + "epoch": 1.7094713955700502, + "grad_norm": 1.2975878715515137, + "learning_rate": 0.0007863160755537437, + "loss": 3.749, + "step": 25160 + }, + { + "epoch": 1.709811115640712, + "grad_norm": 0.8570579290390015, + "learning_rate": 0.0007862736105449109, + "loss": 3.568, + "step": 25165 + }, + { + "epoch": 1.710150835711374, + "grad_norm": 0.8899945616722107, + "learning_rate": 0.0007862311455360783, + "loss": 3.718, + "step": 25170 + }, + { + "epoch": 1.7104905557820356, + "grad_norm": 0.9734316468238831, + "learning_rate": 0.0007861886805272456, + "loss": 3.7305, + "step": 25175 + }, + { + "epoch": 1.7108302758526974, + "grad_norm": 0.8933466672897339, + "learning_rate": 0.0007861462155184128, + "loss": 3.7253, + "step": 25180 + }, + { + "epoch": 1.7111699959233593, + "grad_norm": 0.8621127009391785, + "learning_rate": 0.0007861037505095802, + "loss": 3.8209, + "step": 25185 + }, + { + "epoch": 1.711509715994021, + "grad_norm": 1.1854227781295776, + "learning_rate": 0.0007860612855007474, + "loss": 3.6119, + "step": 25190 + }, + { + "epoch": 1.7118494360646825, + "grad_norm": 1.0156502723693848, + "learning_rate": 0.0007860188204919146, + "loss": 3.7828, + "step": 25195 + }, + { + "epoch": 1.7121891561353446, + "grad_norm": 0.9271770715713501, + "learning_rate": 0.000785976355483082, + "loss": 3.5079, + "step": 25200 + }, + { + "epoch": 1.7125288762060062, + "grad_norm": 0.9001640677452087, + "learning_rate": 0.0007859338904742492, + "loss": 3.4672, + "step": 25205 + }, + { + "epoch": 1.7128685962766679, + "grad_norm": 0.882731020450592, + "learning_rate": 0.0007858914254654165, + "loss": 3.6034, + "step": 25210 + }, + { + "epoch": 1.71320831634733, + "grad_norm": 0.8454843163490295, + "learning_rate": 0.0007858489604565839, + "loss": 3.5299, + "step": 25215 + }, + { + "epoch": 1.7135480364179916, + "grad_norm": 0.7633732557296753, + "learning_rate": 0.0007858064954477511, + "loss": 3.6641, + "step": 25220 + }, + { + "epoch": 1.7138877564886532, + "grad_norm": 0.9066680073738098, + "learning_rate": 0.0007857640304389183, + "loss": 3.6847, + "step": 25225 + }, + { + "epoch": 1.7142274765593153, + "grad_norm": 0.8993523120880127, + "learning_rate": 0.0007857215654300856, + "loss": 3.3965, + "step": 25230 + }, + { + "epoch": 1.714567196629977, + "grad_norm": 0.8086879253387451, + "learning_rate": 0.0007856791004212529, + "loss": 3.5346, + "step": 25235 + }, + { + "epoch": 1.7149069167006386, + "grad_norm": 1.2318345308303833, + "learning_rate": 0.0007856366354124201, + "loss": 4.0851, + "step": 25240 + }, + { + "epoch": 1.7152466367713004, + "grad_norm": 0.6680339574813843, + "learning_rate": 0.0007855941704035875, + "loss": 3.5625, + "step": 25245 + }, + { + "epoch": 1.7155863568419623, + "grad_norm": 0.7271459698677063, + "learning_rate": 0.0007855517053947548, + "loss": 3.6285, + "step": 25250 + }, + { + "epoch": 1.7159260769126239, + "grad_norm": 0.9027031064033508, + "learning_rate": 0.000785509240385922, + "loss": 3.5296, + "step": 25255 + }, + { + "epoch": 1.7162657969832857, + "grad_norm": 0.8412736058235168, + "learning_rate": 0.0007854667753770893, + "loss": 3.44, + "step": 25260 + }, + { + "epoch": 1.7166055170539476, + "grad_norm": 0.8910761475563049, + "learning_rate": 0.0007854243103682565, + "loss": 3.7127, + "step": 25265 + }, + { + "epoch": 1.7169452371246092, + "grad_norm": 0.716374397277832, + "learning_rate": 0.0007853818453594238, + "loss": 3.6474, + "step": 25270 + }, + { + "epoch": 1.717284957195271, + "grad_norm": 0.7394686937332153, + "learning_rate": 0.0007853393803505911, + "loss": 3.5211, + "step": 25275 + }, + { + "epoch": 1.717624677265933, + "grad_norm": 1.0245121717453003, + "learning_rate": 0.0007852969153417584, + "loss": 3.6877, + "step": 25280 + }, + { + "epoch": 1.7179643973365946, + "grad_norm": 0.711330235004425, + "learning_rate": 0.0007852544503329257, + "loss": 3.7462, + "step": 25285 + }, + { + "epoch": 1.7183041174072564, + "grad_norm": 0.8598629236221313, + "learning_rate": 0.000785211985324093, + "loss": 3.761, + "step": 25290 + }, + { + "epoch": 1.7186438374779183, + "grad_norm": 0.9318746328353882, + "learning_rate": 0.0007851695203152602, + "loss": 3.3737, + "step": 25295 + }, + { + "epoch": 1.71898355754858, + "grad_norm": 0.6782599687576294, + "learning_rate": 0.0007851270553064275, + "loss": 3.6125, + "step": 25300 + }, + { + "epoch": 1.7193232776192418, + "grad_norm": 1.48186194896698, + "learning_rate": 0.0007850845902975948, + "loss": 3.4413, + "step": 25305 + }, + { + "epoch": 1.7196629976899036, + "grad_norm": 0.9846222400665283, + "learning_rate": 0.000785042125288762, + "loss": 3.7362, + "step": 25310 + }, + { + "epoch": 1.7200027177605652, + "grad_norm": 0.9732329249382019, + "learning_rate": 0.0007849996602799293, + "loss": 3.6995, + "step": 25315 + }, + { + "epoch": 1.720342437831227, + "grad_norm": 0.7859788537025452, + "learning_rate": 0.0007849571952710967, + "loss": 3.6746, + "step": 25320 + }, + { + "epoch": 1.720682157901889, + "grad_norm": 0.8149735331535339, + "learning_rate": 0.0007849147302622639, + "loss": 3.4913, + "step": 25325 + }, + { + "epoch": 1.7210218779725506, + "grad_norm": 0.7564394474029541, + "learning_rate": 0.0007848722652534311, + "loss": 3.8136, + "step": 25330 + }, + { + "epoch": 1.7213615980432124, + "grad_norm": 1.0305190086364746, + "learning_rate": 0.0007848298002445985, + "loss": 3.4983, + "step": 25335 + }, + { + "epoch": 1.7217013181138743, + "grad_norm": 0.7466962933540344, + "learning_rate": 0.0007847873352357657, + "loss": 3.6435, + "step": 25340 + }, + { + "epoch": 1.722041038184536, + "grad_norm": 0.837145209312439, + "learning_rate": 0.0007847448702269329, + "loss": 3.4404, + "step": 25345 + }, + { + "epoch": 1.7223807582551978, + "grad_norm": 0.772117018699646, + "learning_rate": 0.0007847024052181004, + "loss": 3.3389, + "step": 25350 + }, + { + "epoch": 1.7227204783258596, + "grad_norm": 0.9325253367424011, + "learning_rate": 0.0007846599402092676, + "loss": 3.4587, + "step": 25355 + }, + { + "epoch": 1.7230601983965212, + "grad_norm": 0.8070598840713501, + "learning_rate": 0.0007846174752004348, + "loss": 3.4012, + "step": 25360 + }, + { + "epoch": 1.7233999184671829, + "grad_norm": 0.8725023865699768, + "learning_rate": 0.0007845750101916021, + "loss": 3.5858, + "step": 25365 + }, + { + "epoch": 1.723739638537845, + "grad_norm": 2.5675835609436035, + "learning_rate": 0.0007845325451827694, + "loss": 3.8135, + "step": 25370 + }, + { + "epoch": 1.7240793586085066, + "grad_norm": 0.956696093082428, + "learning_rate": 0.0007844900801739366, + "loss": 3.5513, + "step": 25375 + }, + { + "epoch": 1.7244190786791682, + "grad_norm": 0.9640530347824097, + "learning_rate": 0.000784447615165104, + "loss": 3.5469, + "step": 25380 + }, + { + "epoch": 1.7247587987498303, + "grad_norm": 1.0938912630081177, + "learning_rate": 0.0007844051501562713, + "loss": 3.7052, + "step": 25385 + }, + { + "epoch": 1.725098518820492, + "grad_norm": 0.7502402067184448, + "learning_rate": 0.0007843626851474386, + "loss": 3.7198, + "step": 25390 + }, + { + "epoch": 1.7254382388911536, + "grad_norm": 1.082424283027649, + "learning_rate": 0.0007843202201386058, + "loss": 3.6527, + "step": 25395 + }, + { + "epoch": 1.7257779589618156, + "grad_norm": 0.7903933525085449, + "learning_rate": 0.0007842777551297731, + "loss": 3.5476, + "step": 25400 + }, + { + "epoch": 1.7261176790324773, + "grad_norm": 0.9569440484046936, + "learning_rate": 0.0007842352901209404, + "loss": 3.5726, + "step": 25405 + }, + { + "epoch": 1.726457399103139, + "grad_norm": 0.7713733315467834, + "learning_rate": 0.0007841928251121076, + "loss": 3.6635, + "step": 25410 + }, + { + "epoch": 1.7267971191738007, + "grad_norm": 0.8318759799003601, + "learning_rate": 0.000784150360103275, + "loss": 3.6996, + "step": 25415 + }, + { + "epoch": 1.7271368392444626, + "grad_norm": 0.9367022514343262, + "learning_rate": 0.0007841078950944423, + "loss": 3.8604, + "step": 25420 + }, + { + "epoch": 1.7274765593151242, + "grad_norm": 0.8763725161552429, + "learning_rate": 0.0007840654300856095, + "loss": 3.6309, + "step": 25425 + }, + { + "epoch": 1.727816279385786, + "grad_norm": 0.9216275811195374, + "learning_rate": 0.0007840229650767767, + "loss": 3.499, + "step": 25430 + }, + { + "epoch": 1.728155999456448, + "grad_norm": 1.0109037160873413, + "learning_rate": 0.0007839805000679441, + "loss": 3.6353, + "step": 25435 + }, + { + "epoch": 1.7284957195271096, + "grad_norm": 0.69557124376297, + "learning_rate": 0.0007839380350591113, + "loss": 3.7098, + "step": 25440 + }, + { + "epoch": 1.7288354395977714, + "grad_norm": 0.8115939497947693, + "learning_rate": 0.0007838955700502785, + "loss": 3.5327, + "step": 25445 + }, + { + "epoch": 1.7291751596684333, + "grad_norm": 1.075798749923706, + "learning_rate": 0.000783853105041446, + "loss": 3.6241, + "step": 25450 + }, + { + "epoch": 1.729514879739095, + "grad_norm": 0.8638855814933777, + "learning_rate": 0.0007838106400326132, + "loss": 3.716, + "step": 25455 + }, + { + "epoch": 1.7298545998097568, + "grad_norm": 0.7389013171195984, + "learning_rate": 0.0007837681750237804, + "loss": 3.9045, + "step": 25460 + }, + { + "epoch": 1.7301943198804186, + "grad_norm": 0.9173335433006287, + "learning_rate": 0.0007837257100149478, + "loss": 3.6819, + "step": 25465 + }, + { + "epoch": 1.7305340399510802, + "grad_norm": 1.1047354936599731, + "learning_rate": 0.000783683245006115, + "loss": 3.4661, + "step": 25470 + }, + { + "epoch": 1.730873760021742, + "grad_norm": 0.8324397802352905, + "learning_rate": 0.0007836407799972822, + "loss": 3.5738, + "step": 25475 + }, + { + "epoch": 1.731213480092404, + "grad_norm": 1.0180611610412598, + "learning_rate": 0.0007835983149884495, + "loss": 3.4663, + "step": 25480 + }, + { + "epoch": 1.7315532001630656, + "grad_norm": 0.9008068442344666, + "learning_rate": 0.0007835558499796169, + "loss": 3.5779, + "step": 25485 + }, + { + "epoch": 1.7318929202337274, + "grad_norm": 1.0287599563598633, + "learning_rate": 0.0007835133849707841, + "loss": 3.6283, + "step": 25490 + }, + { + "epoch": 1.7322326403043893, + "grad_norm": 0.9672912955284119, + "learning_rate": 0.0007834709199619514, + "loss": 3.412, + "step": 25495 + }, + { + "epoch": 1.732572360375051, + "grad_norm": 1.0840785503387451, + "learning_rate": 0.0007834284549531187, + "loss": 3.6638, + "step": 25500 + }, + { + "epoch": 1.7329120804457128, + "grad_norm": 1.008786916732788, + "learning_rate": 0.0007833859899442859, + "loss": 3.7185, + "step": 25505 + }, + { + "epoch": 1.7332518005163746, + "grad_norm": 0.811100423336029, + "learning_rate": 0.0007833435249354532, + "loss": 3.8903, + "step": 25510 + }, + { + "epoch": 1.7335915205870362, + "grad_norm": 0.905464231967926, + "learning_rate": 0.0007833010599266204, + "loss": 3.4721, + "step": 25515 + }, + { + "epoch": 1.733931240657698, + "grad_norm": 0.7067092061042786, + "learning_rate": 0.0007832585949177878, + "loss": 3.639, + "step": 25520 + }, + { + "epoch": 1.73427096072836, + "grad_norm": 0.9283978939056396, + "learning_rate": 0.0007832161299089551, + "loss": 3.2637, + "step": 25525 + }, + { + "epoch": 1.7346106807990216, + "grad_norm": 0.8091094493865967, + "learning_rate": 0.0007831736649001223, + "loss": 3.3166, + "step": 25530 + }, + { + "epoch": 1.7349504008696832, + "grad_norm": 0.9639005661010742, + "learning_rate": 0.0007831311998912896, + "loss": 3.382, + "step": 25535 + }, + { + "epoch": 1.7352901209403453, + "grad_norm": 0.9067379236221313, + "learning_rate": 0.0007830887348824569, + "loss": 3.521, + "step": 25540 + }, + { + "epoch": 1.735629841011007, + "grad_norm": 1.1528699398040771, + "learning_rate": 0.0007830462698736241, + "loss": 3.5495, + "step": 25545 + }, + { + "epoch": 1.7359695610816686, + "grad_norm": 1.3540722131729126, + "learning_rate": 0.0007830038048647913, + "loss": 3.6872, + "step": 25550 + }, + { + "epoch": 1.7363092811523306, + "grad_norm": 0.7633983492851257, + "learning_rate": 0.0007829613398559588, + "loss": 3.6953, + "step": 25555 + }, + { + "epoch": 1.7366490012229923, + "grad_norm": 0.7218027114868164, + "learning_rate": 0.000782918874847126, + "loss": 3.6425, + "step": 25560 + }, + { + "epoch": 1.736988721293654, + "grad_norm": 0.8587518930435181, + "learning_rate": 0.0007828764098382932, + "loss": 3.5772, + "step": 25565 + }, + { + "epoch": 1.737328441364316, + "grad_norm": 1.1589277982711792, + "learning_rate": 0.0007828339448294606, + "loss": 3.3, + "step": 25570 + }, + { + "epoch": 1.7376681614349776, + "grad_norm": 0.8263728022575378, + "learning_rate": 0.0007827914798206278, + "loss": 3.4872, + "step": 25575 + }, + { + "epoch": 1.7380078815056392, + "grad_norm": 0.9674803018569946, + "learning_rate": 0.000782749014811795, + "loss": 3.6417, + "step": 25580 + }, + { + "epoch": 1.738347601576301, + "grad_norm": 1.4880969524383545, + "learning_rate": 0.0007827065498029624, + "loss": 3.5134, + "step": 25585 + }, + { + "epoch": 1.738687321646963, + "grad_norm": 1.052046537399292, + "learning_rate": 0.0007826640847941297, + "loss": 3.7721, + "step": 25590 + }, + { + "epoch": 1.7390270417176246, + "grad_norm": 0.8075477480888367, + "learning_rate": 0.0007826216197852969, + "loss": 3.7144, + "step": 25595 + }, + { + "epoch": 1.7393667617882864, + "grad_norm": 1.185079574584961, + "learning_rate": 0.0007825791547764643, + "loss": 3.9093, + "step": 25600 + }, + { + "epoch": 1.7397064818589483, + "grad_norm": 0.9164443612098694, + "learning_rate": 0.0007825366897676315, + "loss": 3.623, + "step": 25605 + }, + { + "epoch": 1.74004620192961, + "grad_norm": 0.9363973140716553, + "learning_rate": 0.0007824942247587987, + "loss": 3.597, + "step": 25610 + }, + { + "epoch": 1.7403859220002718, + "grad_norm": 0.8684784770011902, + "learning_rate": 0.000782451759749966, + "loss": 3.8104, + "step": 25615 + }, + { + "epoch": 1.7407256420709336, + "grad_norm": 0.8852021098136902, + "learning_rate": 0.0007824092947411333, + "loss": 3.4868, + "step": 25620 + }, + { + "epoch": 1.7410653621415952, + "grad_norm": 1.8354719877243042, + "learning_rate": 0.0007823668297323006, + "loss": 3.6706, + "step": 25625 + }, + { + "epoch": 1.741405082212257, + "grad_norm": 0.8952626585960388, + "learning_rate": 0.0007823243647234679, + "loss": 3.7335, + "step": 25630 + }, + { + "epoch": 1.741744802282919, + "grad_norm": 1.5697295665740967, + "learning_rate": 0.0007822818997146352, + "loss": 3.5452, + "step": 25635 + }, + { + "epoch": 1.7420845223535806, + "grad_norm": 0.7047613859176636, + "learning_rate": 0.0007822394347058024, + "loss": 3.4494, + "step": 25640 + }, + { + "epoch": 1.7424242424242424, + "grad_norm": 0.9039536714553833, + "learning_rate": 0.0007821969696969697, + "loss": 3.4186, + "step": 25645 + }, + { + "epoch": 1.7427639624949043, + "grad_norm": 0.9211353659629822, + "learning_rate": 0.000782154504688137, + "loss": 3.3789, + "step": 25650 + }, + { + "epoch": 1.743103682565566, + "grad_norm": 0.7146081924438477, + "learning_rate": 0.0007821120396793042, + "loss": 3.643, + "step": 25655 + }, + { + "epoch": 1.7434434026362278, + "grad_norm": 0.8163903951644897, + "learning_rate": 0.0007820695746704716, + "loss": 3.4956, + "step": 25660 + }, + { + "epoch": 1.7437831227068896, + "grad_norm": 1.0551904439926147, + "learning_rate": 0.0007820271096616388, + "loss": 3.5657, + "step": 25665 + }, + { + "epoch": 1.7441228427775513, + "grad_norm": 0.8210049271583557, + "learning_rate": 0.0007819846446528061, + "loss": 3.9947, + "step": 25670 + }, + { + "epoch": 1.744462562848213, + "grad_norm": 0.815860390663147, + "learning_rate": 0.0007819421796439734, + "loss": 3.4684, + "step": 25675 + }, + { + "epoch": 1.744802282918875, + "grad_norm": 0.8983619809150696, + "learning_rate": 0.0007818997146351406, + "loss": 3.7298, + "step": 25680 + }, + { + "epoch": 1.7451420029895366, + "grad_norm": 0.8563128113746643, + "learning_rate": 0.0007818572496263079, + "loss": 3.3687, + "step": 25685 + }, + { + "epoch": 1.7454817230601984, + "grad_norm": 0.9592525959014893, + "learning_rate": 0.0007818147846174752, + "loss": 3.7007, + "step": 25690 + }, + { + "epoch": 1.7458214431308603, + "grad_norm": 0.7552351951599121, + "learning_rate": 0.0007817723196086425, + "loss": 3.6563, + "step": 25695 + }, + { + "epoch": 1.746161163201522, + "grad_norm": 0.8797276616096497, + "learning_rate": 0.0007817298545998098, + "loss": 3.5357, + "step": 25700 + }, + { + "epoch": 1.7465008832721836, + "grad_norm": 0.8985642194747925, + "learning_rate": 0.0007816873895909771, + "loss": 3.7454, + "step": 25705 + }, + { + "epoch": 1.7468406033428456, + "grad_norm": 0.9358516931533813, + "learning_rate": 0.0007816449245821443, + "loss": 3.4075, + "step": 25710 + }, + { + "epoch": 1.7471803234135073, + "grad_norm": 0.9598605632781982, + "learning_rate": 0.0007816024595733115, + "loss": 3.5979, + "step": 25715 + }, + { + "epoch": 1.747520043484169, + "grad_norm": 3.204887866973877, + "learning_rate": 0.0007815599945644789, + "loss": 3.5259, + "step": 25720 + }, + { + "epoch": 1.747859763554831, + "grad_norm": 0.8831291198730469, + "learning_rate": 0.0007815175295556461, + "loss": 3.5218, + "step": 25725 + }, + { + "epoch": 1.7481994836254926, + "grad_norm": 0.7682862281799316, + "learning_rate": 0.0007814750645468135, + "loss": 3.6461, + "step": 25730 + }, + { + "epoch": 1.7485392036961542, + "grad_norm": 0.6871035695075989, + "learning_rate": 0.0007814325995379808, + "loss": 3.2896, + "step": 25735 + }, + { + "epoch": 1.7488789237668163, + "grad_norm": 0.7829475402832031, + "learning_rate": 0.000781390134529148, + "loss": 3.7015, + "step": 25740 + }, + { + "epoch": 1.749218643837478, + "grad_norm": 0.6390740275382996, + "learning_rate": 0.0007813476695203153, + "loss": 3.285, + "step": 25745 + }, + { + "epoch": 1.7495583639081396, + "grad_norm": 0.9304513931274414, + "learning_rate": 0.0007813052045114826, + "loss": 3.7594, + "step": 25750 + }, + { + "epoch": 1.7498980839788014, + "grad_norm": 1.132741093635559, + "learning_rate": 0.0007812627395026498, + "loss": 3.5237, + "step": 25755 + }, + { + "epoch": 1.7502378040494633, + "grad_norm": 1.568610668182373, + "learning_rate": 0.0007812202744938171, + "loss": 3.6244, + "step": 25760 + }, + { + "epoch": 1.750577524120125, + "grad_norm": 0.9372501373291016, + "learning_rate": 0.0007811778094849844, + "loss": 3.5168, + "step": 25765 + }, + { + "epoch": 1.7509172441907868, + "grad_norm": 0.7375500798225403, + "learning_rate": 0.0007811353444761517, + "loss": 3.3374, + "step": 25770 + }, + { + "epoch": 1.7512569642614486, + "grad_norm": 0.9484035968780518, + "learning_rate": 0.000781092879467319, + "loss": 3.7108, + "step": 25775 + }, + { + "epoch": 1.7515966843321102, + "grad_norm": 0.8634046316146851, + "learning_rate": 0.0007810504144584862, + "loss": 3.6578, + "step": 25780 + }, + { + "epoch": 1.751936404402772, + "grad_norm": 0.8717957735061646, + "learning_rate": 0.0007810079494496535, + "loss": 3.4543, + "step": 25785 + }, + { + "epoch": 1.752276124473434, + "grad_norm": 0.8090501427650452, + "learning_rate": 0.0007809654844408208, + "loss": 3.781, + "step": 25790 + }, + { + "epoch": 1.7526158445440956, + "grad_norm": 0.8110304474830627, + "learning_rate": 0.000780923019431988, + "loss": 3.6415, + "step": 25795 + }, + { + "epoch": 1.7529555646147574, + "grad_norm": 0.7835606932640076, + "learning_rate": 0.0007808805544231554, + "loss": 3.6532, + "step": 25800 + }, + { + "epoch": 1.7532952846854193, + "grad_norm": 1.2483454942703247, + "learning_rate": 0.0007808380894143227, + "loss": 3.6573, + "step": 25805 + }, + { + "epoch": 1.753635004756081, + "grad_norm": 0.7581899166107178, + "learning_rate": 0.0007807956244054899, + "loss": 3.7744, + "step": 25810 + }, + { + "epoch": 1.7539747248267428, + "grad_norm": 1.0076439380645752, + "learning_rate": 0.0007807531593966571, + "loss": 3.8244, + "step": 25815 + }, + { + "epoch": 1.7543144448974046, + "grad_norm": 0.9983475804328918, + "learning_rate": 0.0007807106943878245, + "loss": 3.8544, + "step": 25820 + }, + { + "epoch": 1.7546541649680663, + "grad_norm": 1.0102927684783936, + "learning_rate": 0.0007806682293789917, + "loss": 3.5101, + "step": 25825 + }, + { + "epoch": 1.754993885038728, + "grad_norm": 1.167184591293335, + "learning_rate": 0.0007806257643701589, + "loss": 3.6922, + "step": 25830 + }, + { + "epoch": 1.75533360510939, + "grad_norm": 0.8384759426116943, + "learning_rate": 0.0007805832993613264, + "loss": 3.6572, + "step": 25835 + }, + { + "epoch": 1.7556733251800516, + "grad_norm": 0.7975273728370667, + "learning_rate": 0.0007805408343524936, + "loss": 3.6317, + "step": 25840 + }, + { + "epoch": 1.7560130452507134, + "grad_norm": 1.1237077713012695, + "learning_rate": 0.0007804983693436608, + "loss": 3.6961, + "step": 25845 + }, + { + "epoch": 1.7563527653213753, + "grad_norm": 0.851542055606842, + "learning_rate": 0.0007804559043348282, + "loss": 3.527, + "step": 25850 + }, + { + "epoch": 1.756692485392037, + "grad_norm": 1.3945471048355103, + "learning_rate": 0.0007804134393259954, + "loss": 3.6489, + "step": 25855 + }, + { + "epoch": 1.7570322054626988, + "grad_norm": 1.315050482749939, + "learning_rate": 0.0007803709743171626, + "loss": 3.6273, + "step": 25860 + }, + { + "epoch": 1.7573719255333606, + "grad_norm": 1.0731977224349976, + "learning_rate": 0.0007803285093083299, + "loss": 3.6821, + "step": 25865 + }, + { + "epoch": 1.7577116456040223, + "grad_norm": 0.7589316368103027, + "learning_rate": 0.0007802860442994973, + "loss": 3.4292, + "step": 25870 + }, + { + "epoch": 1.758051365674684, + "grad_norm": 1.2083330154418945, + "learning_rate": 0.0007802435792906645, + "loss": 3.3793, + "step": 25875 + }, + { + "epoch": 1.758391085745346, + "grad_norm": 0.8640690445899963, + "learning_rate": 0.0007802011142818318, + "loss": 3.4375, + "step": 25880 + }, + { + "epoch": 1.7587308058160076, + "grad_norm": 0.7859452366828918, + "learning_rate": 0.0007801586492729991, + "loss": 3.6506, + "step": 25885 + }, + { + "epoch": 1.7590705258866692, + "grad_norm": 0.8886356949806213, + "learning_rate": 0.0007801161842641663, + "loss": 3.444, + "step": 25890 + }, + { + "epoch": 1.7594102459573313, + "grad_norm": 1.5000760555267334, + "learning_rate": 0.0007800737192553336, + "loss": 3.6421, + "step": 25895 + }, + { + "epoch": 1.759749966027993, + "grad_norm": 3.777399778366089, + "learning_rate": 0.0007800312542465008, + "loss": 3.2754, + "step": 25900 + }, + { + "epoch": 1.7600896860986546, + "grad_norm": 1.2245557308197021, + "learning_rate": 0.0007799887892376682, + "loss": 3.5433, + "step": 25905 + }, + { + "epoch": 1.7604294061693166, + "grad_norm": 0.9506595134735107, + "learning_rate": 0.0007799463242288355, + "loss": 3.6563, + "step": 25910 + }, + { + "epoch": 1.7607691262399783, + "grad_norm": 1.03312349319458, + "learning_rate": 0.0007799038592200027, + "loss": 3.7397, + "step": 25915 + }, + { + "epoch": 1.76110884631064, + "grad_norm": 0.807860255241394, + "learning_rate": 0.00077986139421117, + "loss": 3.485, + "step": 25920 + }, + { + "epoch": 1.7614485663813018, + "grad_norm": 0.9521588087081909, + "learning_rate": 0.0007798189292023373, + "loss": 3.5696, + "step": 25925 + }, + { + "epoch": 1.7617882864519636, + "grad_norm": 0.829430341720581, + "learning_rate": 0.0007797764641935045, + "loss": 3.659, + "step": 25930 + }, + { + "epoch": 1.7621280065226252, + "grad_norm": 0.8915740847587585, + "learning_rate": 0.0007797339991846718, + "loss": 3.6361, + "step": 25935 + }, + { + "epoch": 1.762467726593287, + "grad_norm": 0.9075105786323547, + "learning_rate": 0.0007796915341758392, + "loss": 3.4581, + "step": 25940 + }, + { + "epoch": 1.762807446663949, + "grad_norm": 0.7310864925384521, + "learning_rate": 0.0007796490691670064, + "loss": 3.9386, + "step": 25945 + }, + { + "epoch": 1.7631471667346106, + "grad_norm": 1.213394284248352, + "learning_rate": 0.0007796066041581736, + "loss": 3.5004, + "step": 25950 + }, + { + "epoch": 1.7634868868052724, + "grad_norm": 0.835250735282898, + "learning_rate": 0.000779564139149341, + "loss": 3.8188, + "step": 25955 + }, + { + "epoch": 1.7638266068759343, + "grad_norm": 0.8052058219909668, + "learning_rate": 0.0007795216741405082, + "loss": 3.7281, + "step": 25960 + }, + { + "epoch": 1.764166326946596, + "grad_norm": 0.7974103689193726, + "learning_rate": 0.0007794792091316754, + "loss": 3.566, + "step": 25965 + }, + { + "epoch": 1.7645060470172578, + "grad_norm": 0.9005211591720581, + "learning_rate": 0.0007794367441228429, + "loss": 3.8846, + "step": 25970 + }, + { + "epoch": 1.7648457670879196, + "grad_norm": 0.7942432165145874, + "learning_rate": 0.0007793942791140101, + "loss": 3.5281, + "step": 25975 + }, + { + "epoch": 1.7651854871585813, + "grad_norm": 0.7167492508888245, + "learning_rate": 0.0007793518141051773, + "loss": 3.4482, + "step": 25980 + }, + { + "epoch": 1.765525207229243, + "grad_norm": 1.034829020500183, + "learning_rate": 0.0007793093490963447, + "loss": 3.4977, + "step": 25985 + }, + { + "epoch": 1.765864927299905, + "grad_norm": 0.968509316444397, + "learning_rate": 0.0007792668840875119, + "loss": 3.524, + "step": 25990 + }, + { + "epoch": 1.7662046473705666, + "grad_norm": 0.8943504095077515, + "learning_rate": 0.0007792244190786791, + "loss": 3.9223, + "step": 25995 + }, + { + "epoch": 1.7665443674412284, + "grad_norm": 0.7722302079200745, + "learning_rate": 0.0007791819540698464, + "loss": 3.7378, + "step": 26000 + }, + { + "epoch": 1.7668840875118903, + "grad_norm": 0.7472617030143738, + "learning_rate": 0.0007791394890610138, + "loss": 3.5882, + "step": 26005 + }, + { + "epoch": 1.767223807582552, + "grad_norm": 0.9298312664031982, + "learning_rate": 0.000779097024052181, + "loss": 3.8509, + "step": 26010 + }, + { + "epoch": 1.7675635276532138, + "grad_norm": 0.7147135734558105, + "learning_rate": 0.0007790545590433483, + "loss": 3.5533, + "step": 26015 + }, + { + "epoch": 1.7679032477238756, + "grad_norm": 0.902371346950531, + "learning_rate": 0.0007790120940345156, + "loss": 3.4996, + "step": 26020 + }, + { + "epoch": 1.7682429677945373, + "grad_norm": 0.775912880897522, + "learning_rate": 0.0007789696290256828, + "loss": 3.2739, + "step": 26025 + }, + { + "epoch": 1.7685826878651991, + "grad_norm": 0.7736325860023499, + "learning_rate": 0.0007789271640168501, + "loss": 3.4853, + "step": 26030 + }, + { + "epoch": 1.768922407935861, + "grad_norm": 0.739092230796814, + "learning_rate": 0.0007788846990080174, + "loss": 3.5761, + "step": 26035 + }, + { + "epoch": 1.7692621280065226, + "grad_norm": 0.813132643699646, + "learning_rate": 0.0007788422339991847, + "loss": 3.6797, + "step": 26040 + }, + { + "epoch": 1.7696018480771842, + "grad_norm": 0.8204823136329651, + "learning_rate": 0.000778799768990352, + "loss": 3.5608, + "step": 26045 + }, + { + "epoch": 1.7699415681478463, + "grad_norm": 1.0364375114440918, + "learning_rate": 0.0007787573039815192, + "loss": 3.6712, + "step": 26050 + }, + { + "epoch": 1.770281288218508, + "grad_norm": 0.7703670263290405, + "learning_rate": 0.0007787148389726865, + "loss": 3.3849, + "step": 26055 + }, + { + "epoch": 1.7706210082891696, + "grad_norm": 0.8427909016609192, + "learning_rate": 0.0007786723739638538, + "loss": 3.6009, + "step": 26060 + }, + { + "epoch": 1.7709607283598316, + "grad_norm": 0.8401391506195068, + "learning_rate": 0.000778629908955021, + "loss": 3.5513, + "step": 26065 + }, + { + "epoch": 1.7713004484304933, + "grad_norm": 0.8893247842788696, + "learning_rate": 0.0007785874439461884, + "loss": 3.3767, + "step": 26070 + }, + { + "epoch": 1.771640168501155, + "grad_norm": 1.0901212692260742, + "learning_rate": 0.0007785449789373557, + "loss": 3.4769, + "step": 26075 + }, + { + "epoch": 1.771979888571817, + "grad_norm": 0.9976149797439575, + "learning_rate": 0.0007785025139285229, + "loss": 3.7512, + "step": 26080 + }, + { + "epoch": 1.7723196086424786, + "grad_norm": 0.7497579455375671, + "learning_rate": 0.0007784600489196903, + "loss": 3.4341, + "step": 26085 + }, + { + "epoch": 1.7726593287131402, + "grad_norm": 0.7696861624717712, + "learning_rate": 0.0007784175839108575, + "loss": 3.6361, + "step": 26090 + }, + { + "epoch": 1.772999048783802, + "grad_norm": 1.311822533607483, + "learning_rate": 0.0007783751189020247, + "loss": 3.3312, + "step": 26095 + }, + { + "epoch": 1.773338768854464, + "grad_norm": 0.8057903051376343, + "learning_rate": 0.000778332653893192, + "loss": 3.5161, + "step": 26100 + }, + { + "epoch": 1.7736784889251256, + "grad_norm": 0.5596839189529419, + "learning_rate": 0.0007782901888843593, + "loss": 3.643, + "step": 26105 + }, + { + "epoch": 1.7740182089957874, + "grad_norm": 1.091204047203064, + "learning_rate": 0.0007782477238755266, + "loss": 3.203, + "step": 26110 + }, + { + "epoch": 1.7743579290664493, + "grad_norm": 3.4007673263549805, + "learning_rate": 0.0007782052588666939, + "loss": 3.5878, + "step": 26115 + }, + { + "epoch": 1.774697649137111, + "grad_norm": 2.3639938831329346, + "learning_rate": 0.0007781627938578612, + "loss": 3.7566, + "step": 26120 + }, + { + "epoch": 1.7750373692077728, + "grad_norm": 0.8306202292442322, + "learning_rate": 0.0007781203288490284, + "loss": 3.6895, + "step": 26125 + }, + { + "epoch": 1.7753770892784346, + "grad_norm": 1.246132731437683, + "learning_rate": 0.0007780778638401957, + "loss": 3.5499, + "step": 26130 + }, + { + "epoch": 1.7757168093490963, + "grad_norm": 0.8069747090339661, + "learning_rate": 0.000778035398831363, + "loss": 3.5961, + "step": 26135 + }, + { + "epoch": 1.7760565294197581, + "grad_norm": 0.761080801486969, + "learning_rate": 0.0007779929338225302, + "loss": 3.6282, + "step": 26140 + }, + { + "epoch": 1.77639624949042, + "grad_norm": 0.9705154299736023, + "learning_rate": 0.0007779504688136976, + "loss": 3.5841, + "step": 26145 + }, + { + "epoch": 1.7767359695610816, + "grad_norm": 0.9187883138656616, + "learning_rate": 0.0007779080038048648, + "loss": 3.4097, + "step": 26150 + }, + { + "epoch": 1.7770756896317434, + "grad_norm": 0.8528329730033875, + "learning_rate": 0.0007778655387960321, + "loss": 3.3493, + "step": 26155 + }, + { + "epoch": 1.7774154097024053, + "grad_norm": 0.740311861038208, + "learning_rate": 0.0007778230737871994, + "loss": 3.5759, + "step": 26160 + }, + { + "epoch": 1.777755129773067, + "grad_norm": 0.9098858833312988, + "learning_rate": 0.0007777806087783666, + "loss": 3.6686, + "step": 26165 + }, + { + "epoch": 1.7780948498437288, + "grad_norm": 0.989059329032898, + "learning_rate": 0.0007777381437695339, + "loss": 3.6815, + "step": 26170 + }, + { + "epoch": 1.7784345699143906, + "grad_norm": 0.9316731691360474, + "learning_rate": 0.0007776956787607012, + "loss": 3.3554, + "step": 26175 + }, + { + "epoch": 1.7787742899850523, + "grad_norm": 0.7356194257736206, + "learning_rate": 0.0007776532137518685, + "loss": 3.883, + "step": 26180 + }, + { + "epoch": 1.7791140100557141, + "grad_norm": 0.7641822695732117, + "learning_rate": 0.0007776107487430358, + "loss": 3.7066, + "step": 26185 + }, + { + "epoch": 1.779453730126376, + "grad_norm": 0.9342180490493774, + "learning_rate": 0.0007775682837342031, + "loss": 3.4949, + "step": 26190 + }, + { + "epoch": 1.7797934501970376, + "grad_norm": 0.8679904341697693, + "learning_rate": 0.0007775258187253703, + "loss": 3.3175, + "step": 26195 + }, + { + "epoch": 1.7801331702676995, + "grad_norm": 1.0978342294692993, + "learning_rate": 0.0007774833537165375, + "loss": 3.4248, + "step": 26200 + }, + { + "epoch": 1.7804728903383613, + "grad_norm": 0.791932225227356, + "learning_rate": 0.0007774408887077049, + "loss": 3.483, + "step": 26205 + }, + { + "epoch": 1.780812610409023, + "grad_norm": 1.1206365823745728, + "learning_rate": 0.0007773984236988721, + "loss": 3.6317, + "step": 26210 + }, + { + "epoch": 1.7811523304796846, + "grad_norm": 0.9562692642211914, + "learning_rate": 0.0007773559586900394, + "loss": 3.4922, + "step": 26215 + }, + { + "epoch": 1.7814920505503467, + "grad_norm": 1.0970159769058228, + "learning_rate": 0.0007773134936812068, + "loss": 3.5252, + "step": 26220 + }, + { + "epoch": 1.7818317706210083, + "grad_norm": 0.6391739249229431, + "learning_rate": 0.000777271028672374, + "loss": 3.6473, + "step": 26225 + }, + { + "epoch": 1.78217149069167, + "grad_norm": 0.7666692137718201, + "learning_rate": 0.0007772285636635412, + "loss": 3.7515, + "step": 26230 + }, + { + "epoch": 1.782511210762332, + "grad_norm": 0.6616684198379517, + "learning_rate": 0.0007771860986547086, + "loss": 3.6492, + "step": 26235 + }, + { + "epoch": 1.7828509308329936, + "grad_norm": 1.3172047138214111, + "learning_rate": 0.0007771436336458758, + "loss": 3.5426, + "step": 26240 + }, + { + "epoch": 1.7831906509036552, + "grad_norm": 0.8805871605873108, + "learning_rate": 0.000777101168637043, + "loss": 3.7075, + "step": 26245 + }, + { + "epoch": 1.7835303709743173, + "grad_norm": 0.8232247233390808, + "learning_rate": 0.0007770587036282105, + "loss": 3.7654, + "step": 26250 + }, + { + "epoch": 1.783870091044979, + "grad_norm": 0.6485000252723694, + "learning_rate": 0.0007770162386193777, + "loss": 3.4879, + "step": 26255 + }, + { + "epoch": 1.7842098111156406, + "grad_norm": 0.9163903594017029, + "learning_rate": 0.0007769737736105449, + "loss": 3.5246, + "step": 26260 + }, + { + "epoch": 1.7845495311863024, + "grad_norm": 0.7492279410362244, + "learning_rate": 0.0007769313086017122, + "loss": 3.5252, + "step": 26265 + }, + { + "epoch": 1.7848892512569643, + "grad_norm": 0.7342423796653748, + "learning_rate": 0.0007768888435928795, + "loss": 3.7, + "step": 26270 + }, + { + "epoch": 1.785228971327626, + "grad_norm": 1.0257717370986938, + "learning_rate": 0.0007768463785840467, + "loss": 3.6694, + "step": 26275 + }, + { + "epoch": 1.7855686913982878, + "grad_norm": 0.6728167533874512, + "learning_rate": 0.000776803913575214, + "loss": 3.7653, + "step": 26280 + }, + { + "epoch": 1.7859084114689496, + "grad_norm": 1.01872718334198, + "learning_rate": 0.0007767614485663814, + "loss": 3.6343, + "step": 26285 + }, + { + "epoch": 1.7862481315396113, + "grad_norm": 0.7362704277038574, + "learning_rate": 0.0007767189835575486, + "loss": 3.6759, + "step": 26290 + }, + { + "epoch": 1.7865878516102731, + "grad_norm": 1.366279125213623, + "learning_rate": 0.0007766765185487159, + "loss": 3.644, + "step": 26295 + }, + { + "epoch": 1.786927571680935, + "grad_norm": 0.820401668548584, + "learning_rate": 0.0007766340535398831, + "loss": 3.404, + "step": 26300 + }, + { + "epoch": 1.7872672917515966, + "grad_norm": 1.2093595266342163, + "learning_rate": 0.0007765915885310504, + "loss": 3.5584, + "step": 26305 + }, + { + "epoch": 1.7876070118222585, + "grad_norm": 0.9943270683288574, + "learning_rate": 0.0007765491235222177, + "loss": 3.6304, + "step": 26310 + }, + { + "epoch": 1.7879467318929203, + "grad_norm": 0.8318180441856384, + "learning_rate": 0.0007765066585133849, + "loss": 3.6815, + "step": 26315 + }, + { + "epoch": 1.788286451963582, + "grad_norm": 0.8164183497428894, + "learning_rate": 0.0007764641935045523, + "loss": 3.4947, + "step": 26320 + }, + { + "epoch": 1.7886261720342438, + "grad_norm": 0.8470618724822998, + "learning_rate": 0.0007764217284957196, + "loss": 3.4427, + "step": 26325 + }, + { + "epoch": 1.7889658921049056, + "grad_norm": 0.968076229095459, + "learning_rate": 0.0007763792634868868, + "loss": 3.6362, + "step": 26330 + }, + { + "epoch": 1.7893056121755673, + "grad_norm": 0.8567678928375244, + "learning_rate": 0.000776336798478054, + "loss": 3.5284, + "step": 26335 + }, + { + "epoch": 1.7896453322462291, + "grad_norm": 0.9097279906272888, + "learning_rate": 0.0007762943334692214, + "loss": 3.6648, + "step": 26340 + }, + { + "epoch": 1.789985052316891, + "grad_norm": 0.740727424621582, + "learning_rate": 0.0007762518684603886, + "loss": 3.632, + "step": 26345 + }, + { + "epoch": 1.7903247723875526, + "grad_norm": 0.830374002456665, + "learning_rate": 0.0007762094034515558, + "loss": 3.9525, + "step": 26350 + }, + { + "epoch": 1.7906644924582145, + "grad_norm": 0.7706518173217773, + "learning_rate": 0.0007761669384427233, + "loss": 3.8259, + "step": 26355 + }, + { + "epoch": 1.7910042125288763, + "grad_norm": 0.7739277482032776, + "learning_rate": 0.0007761244734338905, + "loss": 3.6415, + "step": 26360 + }, + { + "epoch": 1.791343932599538, + "grad_norm": 0.8040427565574646, + "learning_rate": 0.0007760820084250577, + "loss": 3.7262, + "step": 26365 + }, + { + "epoch": 1.7916836526701998, + "grad_norm": 1.1802363395690918, + "learning_rate": 0.0007760395434162251, + "loss": 3.7065, + "step": 26370 + }, + { + "epoch": 1.7920233727408617, + "grad_norm": 0.746063768863678, + "learning_rate": 0.0007759970784073923, + "loss": 3.6451, + "step": 26375 + }, + { + "epoch": 1.7923630928115233, + "grad_norm": 1.1472550630569458, + "learning_rate": 0.0007759546133985595, + "loss": 3.7605, + "step": 26380 + }, + { + "epoch": 1.792702812882185, + "grad_norm": 0.8310592174530029, + "learning_rate": 0.0007759121483897269, + "loss": 3.6762, + "step": 26385 + }, + { + "epoch": 1.793042532952847, + "grad_norm": 1.0211249589920044, + "learning_rate": 0.0007758696833808942, + "loss": 3.6768, + "step": 26390 + }, + { + "epoch": 1.7933822530235086, + "grad_norm": 0.6641800999641418, + "learning_rate": 0.0007758272183720614, + "loss": 3.7767, + "step": 26395 + }, + { + "epoch": 1.7937219730941703, + "grad_norm": 0.8496041297912598, + "learning_rate": 0.0007757847533632287, + "loss": 3.5218, + "step": 26400 + }, + { + "epoch": 1.7940616931648323, + "grad_norm": 0.7252931594848633, + "learning_rate": 0.000775742288354396, + "loss": 3.5727, + "step": 26405 + }, + { + "epoch": 1.794401413235494, + "grad_norm": 0.861052930355072, + "learning_rate": 0.0007756998233455633, + "loss": 3.5623, + "step": 26410 + }, + { + "epoch": 1.7947411333061556, + "grad_norm": 0.6104414463043213, + "learning_rate": 0.0007756573583367305, + "loss": 3.5555, + "step": 26415 + }, + { + "epoch": 1.7950808533768177, + "grad_norm": 0.8131604790687561, + "learning_rate": 0.0007756148933278978, + "loss": 3.4779, + "step": 26420 + }, + { + "epoch": 1.7954205734474793, + "grad_norm": 0.9060520529747009, + "learning_rate": 0.0007755724283190652, + "loss": 3.401, + "step": 26425 + }, + { + "epoch": 1.795760293518141, + "grad_norm": 0.878742516040802, + "learning_rate": 0.0007755299633102324, + "loss": 3.5545, + "step": 26430 + }, + { + "epoch": 1.7961000135888028, + "grad_norm": 1.2023383378982544, + "learning_rate": 0.0007754874983013997, + "loss": 3.415, + "step": 26435 + }, + { + "epoch": 1.7964397336594646, + "grad_norm": 1.2565702199935913, + "learning_rate": 0.000775445033292567, + "loss": 3.7117, + "step": 26440 + }, + { + "epoch": 1.7967794537301263, + "grad_norm": 0.7067636847496033, + "learning_rate": 0.0007754025682837342, + "loss": 3.3881, + "step": 26445 + }, + { + "epoch": 1.7971191738007881, + "grad_norm": 1.422179937362671, + "learning_rate": 0.0007753601032749014, + "loss": 3.2169, + "step": 26450 + }, + { + "epoch": 1.79745889387145, + "grad_norm": 1.0218615531921387, + "learning_rate": 0.0007753176382660688, + "loss": 3.5967, + "step": 26455 + }, + { + "epoch": 1.7977986139421116, + "grad_norm": 0.8417220711708069, + "learning_rate": 0.0007752751732572361, + "loss": 3.7578, + "step": 26460 + }, + { + "epoch": 1.7981383340127735, + "grad_norm": 0.6974185705184937, + "learning_rate": 0.0007752327082484033, + "loss": 3.5118, + "step": 26465 + }, + { + "epoch": 1.7984780540834353, + "grad_norm": 0.9856672286987305, + "learning_rate": 0.0007751902432395707, + "loss": 3.6269, + "step": 26470 + }, + { + "epoch": 1.798817774154097, + "grad_norm": 0.9373189210891724, + "learning_rate": 0.0007751477782307379, + "loss": 3.6082, + "step": 26475 + }, + { + "epoch": 1.7991574942247588, + "grad_norm": 0.8289579153060913, + "learning_rate": 0.0007751053132219051, + "loss": 3.7031, + "step": 26480 + }, + { + "epoch": 1.7994972142954206, + "grad_norm": 0.9893999099731445, + "learning_rate": 0.0007750628482130725, + "loss": 3.5097, + "step": 26485 + }, + { + "epoch": 1.7998369343660823, + "grad_norm": 0.9149765968322754, + "learning_rate": 0.0007750203832042397, + "loss": 3.5405, + "step": 26490 + }, + { + "epoch": 1.8001766544367441, + "grad_norm": 0.706072986125946, + "learning_rate": 0.000774977918195407, + "loss": 3.6303, + "step": 26495 + }, + { + "epoch": 1.800516374507406, + "grad_norm": 0.7087363004684448, + "learning_rate": 0.0007749354531865743, + "loss": 3.5855, + "step": 26500 + }, + { + "epoch": 1.8008560945780676, + "grad_norm": 0.9747774004936218, + "learning_rate": 0.0007748929881777416, + "loss": 3.4446, + "step": 26505 + }, + { + "epoch": 1.8011958146487295, + "grad_norm": 0.7663552165031433, + "learning_rate": 0.0007748505231689088, + "loss": 3.9202, + "step": 26510 + }, + { + "epoch": 1.8015355347193913, + "grad_norm": 0.83504718542099, + "learning_rate": 0.0007748080581600761, + "loss": 3.6159, + "step": 26515 + }, + { + "epoch": 1.801875254790053, + "grad_norm": 0.6961431503295898, + "learning_rate": 0.0007747655931512434, + "loss": 3.2866, + "step": 26520 + }, + { + "epoch": 1.8022149748607148, + "grad_norm": 0.8577984571456909, + "learning_rate": 0.0007747231281424106, + "loss": 3.7548, + "step": 26525 + }, + { + "epoch": 1.8025546949313767, + "grad_norm": 0.7688853144645691, + "learning_rate": 0.000774680663133578, + "loss": 3.4588, + "step": 26530 + }, + { + "epoch": 1.8028944150020383, + "grad_norm": 0.9180909991264343, + "learning_rate": 0.0007746381981247453, + "loss": 3.785, + "step": 26535 + }, + { + "epoch": 1.8032341350727001, + "grad_norm": 0.8309755325317383, + "learning_rate": 0.0007745957331159125, + "loss": 3.7327, + "step": 26540 + }, + { + "epoch": 1.803573855143362, + "grad_norm": 0.7985311150550842, + "learning_rate": 0.0007745532681070798, + "loss": 3.635, + "step": 26545 + }, + { + "epoch": 1.8039135752140236, + "grad_norm": 1.130068063735962, + "learning_rate": 0.000774510803098247, + "loss": 3.522, + "step": 26550 + }, + { + "epoch": 1.8042532952846853, + "grad_norm": 0.9112940430641174, + "learning_rate": 0.0007744683380894143, + "loss": 3.8311, + "step": 26555 + }, + { + "epoch": 1.8045930153553473, + "grad_norm": 1.0229567289352417, + "learning_rate": 0.0007744258730805817, + "loss": 3.7484, + "step": 26560 + }, + { + "epoch": 1.804932735426009, + "grad_norm": 0.8279697895050049, + "learning_rate": 0.0007743834080717489, + "loss": 3.5041, + "step": 26565 + }, + { + "epoch": 1.8052724554966706, + "grad_norm": 0.7603811025619507, + "learning_rate": 0.0007743409430629162, + "loss": 3.6527, + "step": 26570 + }, + { + "epoch": 1.8056121755673327, + "grad_norm": 0.9912105202674866, + "learning_rate": 0.0007742984780540835, + "loss": 3.4704, + "step": 26575 + }, + { + "epoch": 1.8059518956379943, + "grad_norm": 0.8260603547096252, + "learning_rate": 0.0007742560130452507, + "loss": 3.5106, + "step": 26580 + }, + { + "epoch": 1.806291615708656, + "grad_norm": 0.7465791702270508, + "learning_rate": 0.0007742135480364179, + "loss": 3.5074, + "step": 26585 + }, + { + "epoch": 1.806631335779318, + "grad_norm": 1.007486343383789, + "learning_rate": 0.0007741710830275853, + "loss": 3.6652, + "step": 26590 + }, + { + "epoch": 1.8069710558499796, + "grad_norm": 0.8247768878936768, + "learning_rate": 0.0007741286180187526, + "loss": 3.6207, + "step": 26595 + }, + { + "epoch": 1.8073107759206413, + "grad_norm": 0.9325391054153442, + "learning_rate": 0.0007740861530099198, + "loss": 3.7965, + "step": 26600 + }, + { + "epoch": 1.8076504959913031, + "grad_norm": 1.2678993940353394, + "learning_rate": 0.0007740436880010872, + "loss": 3.5615, + "step": 26605 + }, + { + "epoch": 1.807990216061965, + "grad_norm": 0.7268818020820618, + "learning_rate": 0.0007740012229922544, + "loss": 3.5981, + "step": 26610 + }, + { + "epoch": 1.8083299361326266, + "grad_norm": 0.7097742557525635, + "learning_rate": 0.0007739587579834216, + "loss": 3.4925, + "step": 26615 + }, + { + "epoch": 1.8086696562032885, + "grad_norm": 0.772057294845581, + "learning_rate": 0.000773916292974589, + "loss": 3.6636, + "step": 26620 + }, + { + "epoch": 1.8090093762739503, + "grad_norm": 0.803548276424408, + "learning_rate": 0.0007738738279657562, + "loss": 3.4405, + "step": 26625 + }, + { + "epoch": 1.809349096344612, + "grad_norm": 0.6985668540000916, + "learning_rate": 0.0007738313629569235, + "loss": 3.6018, + "step": 26630 + }, + { + "epoch": 1.8096888164152738, + "grad_norm": 0.8016760349273682, + "learning_rate": 0.0007737888979480909, + "loss": 3.9352, + "step": 26635 + }, + { + "epoch": 1.8100285364859356, + "grad_norm": 0.8703891038894653, + "learning_rate": 0.0007737464329392581, + "loss": 3.5596, + "step": 26640 + }, + { + "epoch": 1.8103682565565973, + "grad_norm": 0.834832489490509, + "learning_rate": 0.0007737039679304253, + "loss": 3.6899, + "step": 26645 + }, + { + "epoch": 1.8107079766272591, + "grad_norm": 0.7041547894477844, + "learning_rate": 0.0007736615029215926, + "loss": 3.6926, + "step": 26650 + }, + { + "epoch": 1.811047696697921, + "grad_norm": 0.7675163745880127, + "learning_rate": 0.0007736190379127599, + "loss": 3.6088, + "step": 26655 + }, + { + "epoch": 1.8113874167685826, + "grad_norm": 2.7944114208221436, + "learning_rate": 0.0007735765729039271, + "loss": 3.5574, + "step": 26660 + }, + { + "epoch": 1.8117271368392445, + "grad_norm": 0.6680983901023865, + "learning_rate": 0.0007735341078950945, + "loss": 3.4886, + "step": 26665 + }, + { + "epoch": 1.8120668569099063, + "grad_norm": 0.8312240839004517, + "learning_rate": 0.0007734916428862618, + "loss": 3.5767, + "step": 26670 + }, + { + "epoch": 1.812406576980568, + "grad_norm": 0.9335630536079407, + "learning_rate": 0.000773449177877429, + "loss": 3.5908, + "step": 26675 + }, + { + "epoch": 1.8127462970512298, + "grad_norm": 0.8477770090103149, + "learning_rate": 0.0007734067128685963, + "loss": 3.4267, + "step": 26680 + }, + { + "epoch": 1.8130860171218917, + "grad_norm": 0.8280338644981384, + "learning_rate": 0.0007733642478597635, + "loss": 3.4877, + "step": 26685 + }, + { + "epoch": 1.8134257371925533, + "grad_norm": 0.7250208854675293, + "learning_rate": 0.0007733217828509308, + "loss": 3.5619, + "step": 26690 + }, + { + "epoch": 1.8137654572632151, + "grad_norm": 0.8724890947341919, + "learning_rate": 0.0007732793178420981, + "loss": 3.994, + "step": 26695 + }, + { + "epoch": 1.814105177333877, + "grad_norm": 0.7015180587768555, + "learning_rate": 0.0007732368528332654, + "loss": 3.6746, + "step": 26700 + }, + { + "epoch": 1.8144448974045386, + "grad_norm": 0.7386777400970459, + "learning_rate": 0.0007731943878244327, + "loss": 3.3649, + "step": 26705 + }, + { + "epoch": 1.8147846174752005, + "grad_norm": 0.7724874019622803, + "learning_rate": 0.0007731519228156, + "loss": 3.7754, + "step": 26710 + }, + { + "epoch": 1.8151243375458623, + "grad_norm": 1.1906942129135132, + "learning_rate": 0.0007731094578067672, + "loss": 3.5221, + "step": 26715 + }, + { + "epoch": 1.815464057616524, + "grad_norm": 1.0118272304534912, + "learning_rate": 0.0007730669927979345, + "loss": 3.5368, + "step": 26720 + }, + { + "epoch": 1.8158037776871856, + "grad_norm": 0.876660168170929, + "learning_rate": 0.0007730245277891018, + "loss": 3.6559, + "step": 26725 + }, + { + "epoch": 1.8161434977578477, + "grad_norm": 0.9719272255897522, + "learning_rate": 0.000772982062780269, + "loss": 3.7132, + "step": 26730 + }, + { + "epoch": 1.8164832178285093, + "grad_norm": 0.8586097955703735, + "learning_rate": 0.0007729395977714363, + "loss": 3.7356, + "step": 26735 + }, + { + "epoch": 1.816822937899171, + "grad_norm": 0.8096010088920593, + "learning_rate": 0.0007728971327626037, + "loss": 3.2139, + "step": 26740 + }, + { + "epoch": 1.817162657969833, + "grad_norm": 0.6867233514785767, + "learning_rate": 0.0007728546677537709, + "loss": 3.7706, + "step": 26745 + }, + { + "epoch": 1.8175023780404946, + "grad_norm": 1.5170011520385742, + "learning_rate": 0.0007728122027449382, + "loss": 3.876, + "step": 26750 + }, + { + "epoch": 1.8178420981111563, + "grad_norm": 0.6503229141235352, + "learning_rate": 0.0007727697377361055, + "loss": 3.6268, + "step": 26755 + }, + { + "epoch": 1.8181818181818183, + "grad_norm": 1.088379144668579, + "learning_rate": 0.0007727272727272727, + "loss": 3.5683, + "step": 26760 + }, + { + "epoch": 1.81852153825248, + "grad_norm": 0.9574974775314331, + "learning_rate": 0.00077268480771844, + "loss": 3.4579, + "step": 26765 + }, + { + "epoch": 1.8188612583231416, + "grad_norm": 0.9170920252799988, + "learning_rate": 0.0007726423427096074, + "loss": 3.6672, + "step": 26770 + }, + { + "epoch": 1.8192009783938035, + "grad_norm": 0.9261990189552307, + "learning_rate": 0.0007725998777007746, + "loss": 3.7014, + "step": 26775 + }, + { + "epoch": 1.8195406984644653, + "grad_norm": 1.0048258304595947, + "learning_rate": 0.0007725574126919419, + "loss": 3.3778, + "step": 26780 + }, + { + "epoch": 1.819880418535127, + "grad_norm": 0.7642540335655212, + "learning_rate": 0.0007725149476831091, + "loss": 3.5766, + "step": 26785 + }, + { + "epoch": 1.8202201386057888, + "grad_norm": 1.089959740638733, + "learning_rate": 0.0007724724826742764, + "loss": 3.5016, + "step": 26790 + }, + { + "epoch": 1.8205598586764506, + "grad_norm": 1.5794767141342163, + "learning_rate": 0.0007724300176654437, + "loss": 3.4705, + "step": 26795 + }, + { + "epoch": 1.8208995787471123, + "grad_norm": 0.9247173070907593, + "learning_rate": 0.0007723875526566109, + "loss": 3.6616, + "step": 26800 + }, + { + "epoch": 1.8212392988177741, + "grad_norm": 0.9356890320777893, + "learning_rate": 0.0007723450876477783, + "loss": 3.7823, + "step": 26805 + }, + { + "epoch": 1.821579018888436, + "grad_norm": 1.1007574796676636, + "learning_rate": 0.0007723026226389456, + "loss": 3.6459, + "step": 26810 + }, + { + "epoch": 1.8219187389590976, + "grad_norm": 0.8401316404342651, + "learning_rate": 0.0007722601576301128, + "loss": 3.5881, + "step": 26815 + }, + { + "epoch": 1.8222584590297595, + "grad_norm": 0.7598063349723816, + "learning_rate": 0.0007722176926212801, + "loss": 3.4313, + "step": 26820 + }, + { + "epoch": 1.8225981791004213, + "grad_norm": 0.8231205344200134, + "learning_rate": 0.0007721752276124474, + "loss": 3.4215, + "step": 26825 + }, + { + "epoch": 1.822937899171083, + "grad_norm": 0.842429518699646, + "learning_rate": 0.0007721327626036146, + "loss": 3.5588, + "step": 26830 + }, + { + "epoch": 1.8232776192417448, + "grad_norm": 0.843315839767456, + "learning_rate": 0.0007720902975947818, + "loss": 3.5675, + "step": 26835 + }, + { + "epoch": 1.8236173393124067, + "grad_norm": 0.8193457126617432, + "learning_rate": 0.0007720478325859493, + "loss": 3.6766, + "step": 26840 + }, + { + "epoch": 1.8239570593830683, + "grad_norm": 0.8713749051094055, + "learning_rate": 0.0007720053675771165, + "loss": 3.6384, + "step": 26845 + }, + { + "epoch": 1.8242967794537301, + "grad_norm": 0.8145918846130371, + "learning_rate": 0.0007719629025682837, + "loss": 3.7354, + "step": 26850 + }, + { + "epoch": 1.824636499524392, + "grad_norm": 0.7795513272285461, + "learning_rate": 0.0007719204375594511, + "loss": 3.6932, + "step": 26855 + }, + { + "epoch": 1.8249762195950536, + "grad_norm": 0.874422550201416, + "learning_rate": 0.0007718779725506183, + "loss": 3.4648, + "step": 26860 + }, + { + "epoch": 1.8253159396657155, + "grad_norm": 0.7691762447357178, + "learning_rate": 0.0007718355075417855, + "loss": 3.7515, + "step": 26865 + }, + { + "epoch": 1.8256556597363773, + "grad_norm": 0.9575505256652832, + "learning_rate": 0.0007717930425329529, + "loss": 3.7314, + "step": 26870 + }, + { + "epoch": 1.825995379807039, + "grad_norm": 1.2624891996383667, + "learning_rate": 0.0007717505775241202, + "loss": 3.9587, + "step": 26875 + }, + { + "epoch": 1.8263350998777008, + "grad_norm": 0.9109483361244202, + "learning_rate": 0.0007717081125152874, + "loss": 3.638, + "step": 26880 + }, + { + "epoch": 1.8266748199483627, + "grad_norm": 0.7420251965522766, + "learning_rate": 0.0007716656475064548, + "loss": 3.559, + "step": 26885 + }, + { + "epoch": 1.8270145400190243, + "grad_norm": 1.038568139076233, + "learning_rate": 0.000771623182497622, + "loss": 3.575, + "step": 26890 + }, + { + "epoch": 1.827354260089686, + "grad_norm": 0.8054929971694946, + "learning_rate": 0.0007715807174887892, + "loss": 3.5465, + "step": 26895 + }, + { + "epoch": 1.827693980160348, + "grad_norm": 1.322400450706482, + "learning_rate": 0.0007715382524799565, + "loss": 3.4497, + "step": 26900 + }, + { + "epoch": 1.8280337002310096, + "grad_norm": 0.9718524813652039, + "learning_rate": 0.0007714957874711238, + "loss": 3.5916, + "step": 26905 + }, + { + "epoch": 1.8283734203016713, + "grad_norm": 0.8023597598075867, + "learning_rate": 0.0007714533224622911, + "loss": 3.5082, + "step": 26910 + }, + { + "epoch": 1.8287131403723333, + "grad_norm": 0.9028297662734985, + "learning_rate": 0.0007714108574534584, + "loss": 3.7235, + "step": 26915 + }, + { + "epoch": 1.829052860442995, + "grad_norm": 0.8934625387191772, + "learning_rate": 0.0007713683924446257, + "loss": 3.4892, + "step": 26920 + }, + { + "epoch": 1.8293925805136566, + "grad_norm": 1.4386545419692993, + "learning_rate": 0.0007713259274357929, + "loss": 3.4851, + "step": 26925 + }, + { + "epoch": 1.8297323005843187, + "grad_norm": 0.8062783479690552, + "learning_rate": 0.0007712834624269602, + "loss": 3.6672, + "step": 26930 + }, + { + "epoch": 1.8300720206549803, + "grad_norm": 0.6392088532447815, + "learning_rate": 0.0007712409974181274, + "loss": 3.6979, + "step": 26935 + }, + { + "epoch": 1.830411740725642, + "grad_norm": 0.7586618661880493, + "learning_rate": 0.0007711985324092947, + "loss": 3.5235, + "step": 26940 + }, + { + "epoch": 1.8307514607963038, + "grad_norm": 0.7700725197792053, + "learning_rate": 0.0007711560674004621, + "loss": 3.5141, + "step": 26945 + }, + { + "epoch": 1.8310911808669657, + "grad_norm": 0.9625568985939026, + "learning_rate": 0.0007711136023916293, + "loss": 3.3685, + "step": 26950 + }, + { + "epoch": 1.8314309009376273, + "grad_norm": 1.4263578653335571, + "learning_rate": 0.0007710711373827966, + "loss": 3.5274, + "step": 26955 + }, + { + "epoch": 1.8317706210082891, + "grad_norm": 0.6960394382476807, + "learning_rate": 0.0007710286723739639, + "loss": 3.6193, + "step": 26960 + }, + { + "epoch": 1.832110341078951, + "grad_norm": 0.7709155082702637, + "learning_rate": 0.0007709862073651311, + "loss": 3.4351, + "step": 26965 + }, + { + "epoch": 1.8324500611496126, + "grad_norm": 0.8902915716171265, + "learning_rate": 0.0007709437423562983, + "loss": 3.6356, + "step": 26970 + }, + { + "epoch": 1.8327897812202745, + "grad_norm": 0.9700208306312561, + "learning_rate": 0.0007709012773474657, + "loss": 3.8918, + "step": 26975 + }, + { + "epoch": 1.8331295012909363, + "grad_norm": 0.73200523853302, + "learning_rate": 0.000770858812338633, + "loss": 3.6007, + "step": 26980 + }, + { + "epoch": 1.833469221361598, + "grad_norm": 1.791385293006897, + "learning_rate": 0.0007708163473298002, + "loss": 3.5307, + "step": 26985 + }, + { + "epoch": 1.8338089414322598, + "grad_norm": 0.8827402591705322, + "learning_rate": 0.0007707738823209676, + "loss": 3.5124, + "step": 26990 + }, + { + "epoch": 1.8341486615029217, + "grad_norm": 0.7849318385124207, + "learning_rate": 0.0007707314173121348, + "loss": 3.2828, + "step": 26995 + }, + { + "epoch": 1.8344883815735833, + "grad_norm": 1.2764041423797607, + "learning_rate": 0.000770688952303302, + "loss": 3.7065, + "step": 27000 + }, + { + "epoch": 1.8348281016442451, + "grad_norm": 0.892051100730896, + "learning_rate": 0.0007706464872944694, + "loss": 3.3017, + "step": 27005 + }, + { + "epoch": 1.835167821714907, + "grad_norm": 1.0181549787521362, + "learning_rate": 0.0007706040222856366, + "loss": 3.524, + "step": 27010 + }, + { + "epoch": 1.8355075417855686, + "grad_norm": 0.7881715297698975, + "learning_rate": 0.0007705615572768039, + "loss": 3.6031, + "step": 27015 + }, + { + "epoch": 1.8358472618562305, + "grad_norm": 1.4318002462387085, + "learning_rate": 0.0007705190922679713, + "loss": 3.678, + "step": 27020 + }, + { + "epoch": 1.8361869819268923, + "grad_norm": 3.092374086380005, + "learning_rate": 0.0007704766272591385, + "loss": 3.6537, + "step": 27025 + }, + { + "epoch": 1.836526701997554, + "grad_norm": 1.129860281944275, + "learning_rate": 0.0007704341622503057, + "loss": 3.402, + "step": 27030 + }, + { + "epoch": 1.8368664220682158, + "grad_norm": 0.622101902961731, + "learning_rate": 0.000770391697241473, + "loss": 3.6286, + "step": 27035 + }, + { + "epoch": 1.8372061421388777, + "grad_norm": 0.7821914553642273, + "learning_rate": 0.0007703492322326403, + "loss": 3.2474, + "step": 27040 + }, + { + "epoch": 1.8375458622095393, + "grad_norm": 0.996071457862854, + "learning_rate": 0.0007703067672238075, + "loss": 3.5382, + "step": 27045 + }, + { + "epoch": 1.8378855822802012, + "grad_norm": 0.7692259550094604, + "learning_rate": 0.0007702643022149749, + "loss": 3.6814, + "step": 27050 + }, + { + "epoch": 1.838225302350863, + "grad_norm": 0.9140481948852539, + "learning_rate": 0.0007702218372061422, + "loss": 3.566, + "step": 27055 + }, + { + "epoch": 1.8385650224215246, + "grad_norm": 3.619688034057617, + "learning_rate": 0.0007701793721973094, + "loss": 3.4249, + "step": 27060 + }, + { + "epoch": 1.8389047424921863, + "grad_norm": 0.9598721861839294, + "learning_rate": 0.0007701369071884767, + "loss": 3.6482, + "step": 27065 + }, + { + "epoch": 1.8392444625628483, + "grad_norm": 0.8211422562599182, + "learning_rate": 0.000770094442179644, + "loss": 3.6153, + "step": 27070 + }, + { + "epoch": 1.83958418263351, + "grad_norm": 1.1440249681472778, + "learning_rate": 0.0007700519771708112, + "loss": 3.6286, + "step": 27075 + }, + { + "epoch": 1.8399239027041716, + "grad_norm": 0.8672448396682739, + "learning_rate": 0.0007700095121619785, + "loss": 3.6143, + "step": 27080 + }, + { + "epoch": 1.8402636227748337, + "grad_norm": 1.1416536569595337, + "learning_rate": 0.0007699670471531458, + "loss": 3.5142, + "step": 27085 + }, + { + "epoch": 1.8406033428454953, + "grad_norm": 1.0361298322677612, + "learning_rate": 0.0007699245821443132, + "loss": 3.6066, + "step": 27090 + }, + { + "epoch": 1.840943062916157, + "grad_norm": 0.8177396655082703, + "learning_rate": 0.0007698821171354804, + "loss": 3.4067, + "step": 27095 + }, + { + "epoch": 1.841282782986819, + "grad_norm": 1.0287342071533203, + "learning_rate": 0.0007698396521266476, + "loss": 3.5735, + "step": 27100 + }, + { + "epoch": 1.8416225030574807, + "grad_norm": 0.839390754699707, + "learning_rate": 0.000769797187117815, + "loss": 3.5407, + "step": 27105 + }, + { + "epoch": 1.8419622231281423, + "grad_norm": 0.7465947270393372, + "learning_rate": 0.0007697547221089822, + "loss": 3.7772, + "step": 27110 + }, + { + "epoch": 1.8423019431988041, + "grad_norm": 0.911277711391449, + "learning_rate": 0.0007697122571001494, + "loss": 3.4055, + "step": 27115 + }, + { + "epoch": 1.842641663269466, + "grad_norm": 0.940079391002655, + "learning_rate": 0.0007696697920913169, + "loss": 3.7562, + "step": 27120 + }, + { + "epoch": 1.8429813833401276, + "grad_norm": 1.1584961414337158, + "learning_rate": 0.0007696273270824841, + "loss": 3.6399, + "step": 27125 + }, + { + "epoch": 1.8433211034107895, + "grad_norm": 0.8440088629722595, + "learning_rate": 0.0007695848620736513, + "loss": 3.7136, + "step": 27130 + }, + { + "epoch": 1.8436608234814513, + "grad_norm": 0.746200442314148, + "learning_rate": 0.0007695423970648186, + "loss": 3.6443, + "step": 27135 + }, + { + "epoch": 1.844000543552113, + "grad_norm": 0.8213841915130615, + "learning_rate": 0.0007694999320559859, + "loss": 3.6862, + "step": 27140 + }, + { + "epoch": 1.8443402636227748, + "grad_norm": 0.6806025505065918, + "learning_rate": 0.0007694574670471531, + "loss": 3.3967, + "step": 27145 + }, + { + "epoch": 1.8446799836934367, + "grad_norm": 1.1042819023132324, + "learning_rate": 0.0007694150020383205, + "loss": 3.7189, + "step": 27150 + }, + { + "epoch": 1.8450197037640983, + "grad_norm": 1.0057493448257446, + "learning_rate": 0.0007693725370294878, + "loss": 3.7434, + "step": 27155 + }, + { + "epoch": 1.8453594238347601, + "grad_norm": 0.8743189573287964, + "learning_rate": 0.000769330072020655, + "loss": 3.5987, + "step": 27160 + }, + { + "epoch": 1.845699143905422, + "grad_norm": 1.1106281280517578, + "learning_rate": 0.0007692876070118223, + "loss": 3.2995, + "step": 27165 + }, + { + "epoch": 1.8460388639760836, + "grad_norm": 0.868264377117157, + "learning_rate": 0.0007692451420029896, + "loss": 3.5978, + "step": 27170 + }, + { + "epoch": 1.8463785840467455, + "grad_norm": 0.8783383369445801, + "learning_rate": 0.0007692026769941568, + "loss": 3.8498, + "step": 27175 + }, + { + "epoch": 1.8467183041174073, + "grad_norm": 1.074454426765442, + "learning_rate": 0.0007691602119853241, + "loss": 3.8683, + "step": 27180 + }, + { + "epoch": 1.847058024188069, + "grad_norm": 1.0404958724975586, + "learning_rate": 0.0007691177469764914, + "loss": 3.6995, + "step": 27185 + }, + { + "epoch": 1.8473977442587308, + "grad_norm": 0.853009819984436, + "learning_rate": 0.0007690752819676587, + "loss": 3.8654, + "step": 27190 + }, + { + "epoch": 1.8477374643293927, + "grad_norm": 0.7700687646865845, + "learning_rate": 0.000769032816958826, + "loss": 3.8463, + "step": 27195 + }, + { + "epoch": 1.8480771844000543, + "grad_norm": 0.6830011010169983, + "learning_rate": 0.0007689903519499932, + "loss": 3.5871, + "step": 27200 + }, + { + "epoch": 1.8484169044707162, + "grad_norm": 0.6956087350845337, + "learning_rate": 0.0007689478869411605, + "loss": 3.8807, + "step": 27205 + }, + { + "epoch": 1.848756624541378, + "grad_norm": 0.7649178504943848, + "learning_rate": 0.0007689054219323278, + "loss": 3.5997, + "step": 27210 + }, + { + "epoch": 1.8490963446120396, + "grad_norm": 0.64471834897995, + "learning_rate": 0.000768862956923495, + "loss": 3.3776, + "step": 27215 + }, + { + "epoch": 1.8494360646827015, + "grad_norm": 0.9000673890113831, + "learning_rate": 0.0007688204919146624, + "loss": 3.8378, + "step": 27220 + }, + { + "epoch": 1.8497757847533634, + "grad_norm": 1.0492181777954102, + "learning_rate": 0.0007687780269058297, + "loss": 3.4325, + "step": 27225 + }, + { + "epoch": 1.850115504824025, + "grad_norm": 1.0854588747024536, + "learning_rate": 0.0007687355618969969, + "loss": 3.4353, + "step": 27230 + }, + { + "epoch": 1.8504552248946866, + "grad_norm": 0.7448928356170654, + "learning_rate": 0.0007686930968881641, + "loss": 3.6821, + "step": 27235 + }, + { + "epoch": 1.8507949449653487, + "grad_norm": 0.9552714824676514, + "learning_rate": 0.0007686506318793315, + "loss": 3.6605, + "step": 27240 + }, + { + "epoch": 1.8511346650360103, + "grad_norm": 0.8704216480255127, + "learning_rate": 0.0007686081668704987, + "loss": 3.6212, + "step": 27245 + }, + { + "epoch": 1.851474385106672, + "grad_norm": 1.2198938131332397, + "learning_rate": 0.0007685657018616659, + "loss": 3.4653, + "step": 27250 + }, + { + "epoch": 1.851814105177334, + "grad_norm": 1.2459924221038818, + "learning_rate": 0.0007685232368528334, + "loss": 3.4805, + "step": 27255 + }, + { + "epoch": 1.8521538252479957, + "grad_norm": 0.7156299948692322, + "learning_rate": 0.0007684807718440006, + "loss": 3.6415, + "step": 27260 + }, + { + "epoch": 1.8524935453186573, + "grad_norm": 1.0380462408065796, + "learning_rate": 0.0007684383068351678, + "loss": 3.7782, + "step": 27265 + }, + { + "epoch": 1.8528332653893194, + "grad_norm": 1.0427687168121338, + "learning_rate": 0.0007683958418263352, + "loss": 3.677, + "step": 27270 + }, + { + "epoch": 1.853172985459981, + "grad_norm": 0.9309741854667664, + "learning_rate": 0.0007683533768175024, + "loss": 3.8922, + "step": 27275 + }, + { + "epoch": 1.8535127055306426, + "grad_norm": 0.8546794056892395, + "learning_rate": 0.0007683109118086696, + "loss": 3.6222, + "step": 27280 + }, + { + "epoch": 1.8538524256013045, + "grad_norm": 1.0085939168930054, + "learning_rate": 0.0007682684467998369, + "loss": 3.6291, + "step": 27285 + }, + { + "epoch": 1.8541921456719663, + "grad_norm": 0.6203987002372742, + "learning_rate": 0.0007682259817910043, + "loss": 3.7207, + "step": 27290 + }, + { + "epoch": 1.854531865742628, + "grad_norm": 1.034547209739685, + "learning_rate": 0.0007681835167821715, + "loss": 3.4743, + "step": 27295 + }, + { + "epoch": 1.8548715858132898, + "grad_norm": 1.2599289417266846, + "learning_rate": 0.0007681410517733388, + "loss": 3.6557, + "step": 27300 + }, + { + "epoch": 1.8552113058839517, + "grad_norm": 0.7946342825889587, + "learning_rate": 0.0007680985867645061, + "loss": 3.5512, + "step": 27305 + }, + { + "epoch": 1.8555510259546133, + "grad_norm": 1.008371353149414, + "learning_rate": 0.0007680561217556733, + "loss": 3.5165, + "step": 27310 + }, + { + "epoch": 1.8558907460252752, + "grad_norm": 1.0323691368103027, + "learning_rate": 0.0007680136567468406, + "loss": 3.8126, + "step": 27315 + }, + { + "epoch": 1.856230466095937, + "grad_norm": 0.8112845420837402, + "learning_rate": 0.0007679711917380078, + "loss": 3.4403, + "step": 27320 + }, + { + "epoch": 1.8565701861665986, + "grad_norm": 0.8098148703575134, + "learning_rate": 0.0007679287267291752, + "loss": 3.4563, + "step": 27325 + }, + { + "epoch": 1.8569099062372605, + "grad_norm": 0.9876975417137146, + "learning_rate": 0.0007678862617203425, + "loss": 3.4914, + "step": 27330 + }, + { + "epoch": 1.8572496263079223, + "grad_norm": 1.1242824792861938, + "learning_rate": 0.0007678437967115097, + "loss": 3.5018, + "step": 27335 + }, + { + "epoch": 1.857589346378584, + "grad_norm": 0.7763880491256714, + "learning_rate": 0.000767801331702677, + "loss": 3.481, + "step": 27340 + }, + { + "epoch": 1.8579290664492458, + "grad_norm": 0.887474775314331, + "learning_rate": 0.0007677588666938443, + "loss": 3.4525, + "step": 27345 + }, + { + "epoch": 1.8582687865199077, + "grad_norm": 1.0389174222946167, + "learning_rate": 0.0007677164016850115, + "loss": 3.4386, + "step": 27350 + }, + { + "epoch": 1.8586085065905693, + "grad_norm": 0.8689044117927551, + "learning_rate": 0.0007676739366761788, + "loss": 3.6042, + "step": 27355 + }, + { + "epoch": 1.8589482266612312, + "grad_norm": 0.7528037428855896, + "learning_rate": 0.0007676314716673462, + "loss": 3.5482, + "step": 27360 + }, + { + "epoch": 1.859287946731893, + "grad_norm": 0.8691056370735168, + "learning_rate": 0.0007675890066585134, + "loss": 3.5505, + "step": 27365 + }, + { + "epoch": 1.8596276668025546, + "grad_norm": 0.7504808306694031, + "learning_rate": 0.0007675465416496806, + "loss": 3.4806, + "step": 27370 + }, + { + "epoch": 1.8599673868732165, + "grad_norm": 1.110742449760437, + "learning_rate": 0.000767504076640848, + "loss": 3.5491, + "step": 27375 + }, + { + "epoch": 1.8603071069438784, + "grad_norm": 0.7816097736358643, + "learning_rate": 0.0007674616116320152, + "loss": 3.4334, + "step": 27380 + }, + { + "epoch": 1.86064682701454, + "grad_norm": 0.9115440845489502, + "learning_rate": 0.0007674191466231824, + "loss": 3.5536, + "step": 27385 + }, + { + "epoch": 1.8609865470852018, + "grad_norm": 0.9597583413124084, + "learning_rate": 0.0007673766816143498, + "loss": 3.2818, + "step": 27390 + }, + { + "epoch": 1.8613262671558637, + "grad_norm": 0.7826088666915894, + "learning_rate": 0.0007673342166055171, + "loss": 3.7394, + "step": 27395 + }, + { + "epoch": 1.8616659872265253, + "grad_norm": 0.9445761442184448, + "learning_rate": 0.0007672917515966843, + "loss": 3.8531, + "step": 27400 + }, + { + "epoch": 1.862005707297187, + "grad_norm": 1.0347726345062256, + "learning_rate": 0.0007672492865878517, + "loss": 3.3727, + "step": 27405 + }, + { + "epoch": 1.862345427367849, + "grad_norm": 0.9945400357246399, + "learning_rate": 0.0007672068215790189, + "loss": 3.8067, + "step": 27410 + }, + { + "epoch": 1.8626851474385107, + "grad_norm": 0.8428159356117249, + "learning_rate": 0.0007671643565701861, + "loss": 3.8321, + "step": 27415 + }, + { + "epoch": 1.8630248675091723, + "grad_norm": 0.8342534303665161, + "learning_rate": 0.0007671218915613534, + "loss": 3.7449, + "step": 27420 + }, + { + "epoch": 1.8633645875798344, + "grad_norm": 0.7847055196762085, + "learning_rate": 0.0007670794265525207, + "loss": 3.6294, + "step": 27425 + }, + { + "epoch": 1.863704307650496, + "grad_norm": 0.9415319561958313, + "learning_rate": 0.0007670369615436881, + "loss": 3.6916, + "step": 27430 + }, + { + "epoch": 1.8640440277211576, + "grad_norm": 0.6695097088813782, + "learning_rate": 0.0007669944965348553, + "loss": 3.6119, + "step": 27435 + }, + { + "epoch": 1.8643837477918197, + "grad_norm": 0.9680308699607849, + "learning_rate": 0.0007669520315260226, + "loss": 3.8666, + "step": 27440 + }, + { + "epoch": 1.8647234678624813, + "grad_norm": 0.8927963376045227, + "learning_rate": 0.0007669095665171899, + "loss": 3.5948, + "step": 27445 + }, + { + "epoch": 1.865063187933143, + "grad_norm": 0.9845446944236755, + "learning_rate": 0.0007668671015083571, + "loss": 3.3865, + "step": 27450 + }, + { + "epoch": 1.8654029080038048, + "grad_norm": 0.9280561804771423, + "learning_rate": 0.0007668246364995244, + "loss": 3.4073, + "step": 27455 + }, + { + "epoch": 1.8657426280744667, + "grad_norm": 0.6795822978019714, + "learning_rate": 0.0007667821714906917, + "loss": 3.5863, + "step": 27460 + }, + { + "epoch": 1.8660823481451283, + "grad_norm": 1.4276690483093262, + "learning_rate": 0.000766739706481859, + "loss": 3.3912, + "step": 27465 + }, + { + "epoch": 1.8664220682157902, + "grad_norm": 0.8691829442977905, + "learning_rate": 0.0007666972414730262, + "loss": 3.5612, + "step": 27470 + }, + { + "epoch": 1.866761788286452, + "grad_norm": 0.927778422832489, + "learning_rate": 0.0007666547764641936, + "loss": 3.3653, + "step": 27475 + }, + { + "epoch": 1.8671015083571136, + "grad_norm": 2.222081422805786, + "learning_rate": 0.0007666123114553608, + "loss": 3.6839, + "step": 27480 + }, + { + "epoch": 1.8674412284277755, + "grad_norm": 1.2115776538848877, + "learning_rate": 0.000766569846446528, + "loss": 3.507, + "step": 27485 + }, + { + "epoch": 1.8677809484984373, + "grad_norm": 0.9850777387619019, + "learning_rate": 0.0007665273814376954, + "loss": 3.482, + "step": 27490 + }, + { + "epoch": 1.868120668569099, + "grad_norm": 3.00408935546875, + "learning_rate": 0.0007664849164288626, + "loss": 3.8213, + "step": 27495 + }, + { + "epoch": 1.8684603886397608, + "grad_norm": 1.016135573387146, + "learning_rate": 0.0007664424514200299, + "loss": 3.4416, + "step": 27500 + }, + { + "epoch": 1.8688001087104227, + "grad_norm": 0.9369787573814392, + "learning_rate": 0.0007663999864111973, + "loss": 3.7815, + "step": 27505 + }, + { + "epoch": 1.8691398287810843, + "grad_norm": 1.1467502117156982, + "learning_rate": 0.0007663575214023645, + "loss": 3.3465, + "step": 27510 + }, + { + "epoch": 1.8694795488517462, + "grad_norm": 0.9237188100814819, + "learning_rate": 0.0007663150563935317, + "loss": 3.6249, + "step": 27515 + }, + { + "epoch": 1.869819268922408, + "grad_norm": 0.9433103799819946, + "learning_rate": 0.000766272591384699, + "loss": 3.4141, + "step": 27520 + }, + { + "epoch": 1.8701589889930696, + "grad_norm": 0.7327950596809387, + "learning_rate": 0.0007662301263758663, + "loss": 3.503, + "step": 27525 + }, + { + "epoch": 1.8704987090637315, + "grad_norm": 0.7121861577033997, + "learning_rate": 0.0007661876613670335, + "loss": 3.7001, + "step": 27530 + }, + { + "epoch": 1.8708384291343934, + "grad_norm": 8.888404846191406, + "learning_rate": 0.0007661451963582009, + "loss": 3.2281, + "step": 27535 + }, + { + "epoch": 1.871178149205055, + "grad_norm": 0.8614112138748169, + "learning_rate": 0.0007661027313493682, + "loss": 3.8553, + "step": 27540 + }, + { + "epoch": 1.8715178692757168, + "grad_norm": 0.701536238193512, + "learning_rate": 0.0007660602663405354, + "loss": 3.5742, + "step": 27545 + }, + { + "epoch": 1.8718575893463787, + "grad_norm": 0.8840484619140625, + "learning_rate": 0.0007660178013317027, + "loss": 3.8435, + "step": 27550 + }, + { + "epoch": 1.8721973094170403, + "grad_norm": 0.8737392425537109, + "learning_rate": 0.00076597533632287, + "loss": 3.5616, + "step": 27555 + }, + { + "epoch": 1.8725370294877022, + "grad_norm": 0.9750568866729736, + "learning_rate": 0.0007659328713140372, + "loss": 3.1655, + "step": 27560 + }, + { + "epoch": 1.872876749558364, + "grad_norm": 1.8826932907104492, + "learning_rate": 0.0007658904063052045, + "loss": 3.6648, + "step": 27565 + }, + { + "epoch": 1.8732164696290257, + "grad_norm": 0.9589025974273682, + "learning_rate": 0.0007658479412963718, + "loss": 3.4366, + "step": 27570 + }, + { + "epoch": 1.8735561896996873, + "grad_norm": 0.9324145317077637, + "learning_rate": 0.0007658054762875391, + "loss": 3.7053, + "step": 27575 + }, + { + "epoch": 1.8738959097703494, + "grad_norm": 1.2258570194244385, + "learning_rate": 0.0007657630112787064, + "loss": 3.7311, + "step": 27580 + }, + { + "epoch": 1.874235629841011, + "grad_norm": 0.8182900547981262, + "learning_rate": 0.0007657205462698736, + "loss": 3.5832, + "step": 27585 + }, + { + "epoch": 1.8745753499116726, + "grad_norm": 0.7030863761901855, + "learning_rate": 0.0007656780812610409, + "loss": 3.431, + "step": 27590 + }, + { + "epoch": 1.8749150699823347, + "grad_norm": 0.8426514863967896, + "learning_rate": 0.0007656356162522082, + "loss": 3.4716, + "step": 27595 + }, + { + "epoch": 1.8752547900529963, + "grad_norm": 2.751703977584839, + "learning_rate": 0.0007655931512433754, + "loss": 3.4714, + "step": 27600 + }, + { + "epoch": 1.875594510123658, + "grad_norm": 2.7049918174743652, + "learning_rate": 0.0007655506862345428, + "loss": 3.753, + "step": 27605 + }, + { + "epoch": 1.87593423019432, + "grad_norm": 0.8007530570030212, + "learning_rate": 0.0007655082212257101, + "loss": 3.4382, + "step": 27610 + }, + { + "epoch": 1.8762739502649817, + "grad_norm": 0.7956873774528503, + "learning_rate": 0.0007654657562168773, + "loss": 3.7092, + "step": 27615 + }, + { + "epoch": 1.8766136703356433, + "grad_norm": 0.8245481848716736, + "learning_rate": 0.0007654232912080445, + "loss": 3.5895, + "step": 27620 + }, + { + "epoch": 1.8769533904063052, + "grad_norm": 1.086917519569397, + "learning_rate": 0.0007653808261992119, + "loss": 3.7614, + "step": 27625 + }, + { + "epoch": 1.877293110476967, + "grad_norm": 0.7628483176231384, + "learning_rate": 0.0007653383611903791, + "loss": 3.5879, + "step": 27630 + }, + { + "epoch": 1.8776328305476286, + "grad_norm": 0.8953080177307129, + "learning_rate": 0.0007652958961815463, + "loss": 3.253, + "step": 27635 + }, + { + "epoch": 1.8779725506182905, + "grad_norm": 0.6786538362503052, + "learning_rate": 0.0007652534311727138, + "loss": 3.6148, + "step": 27640 + }, + { + "epoch": 1.8783122706889523, + "grad_norm": 1.1295231580734253, + "learning_rate": 0.000765210966163881, + "loss": 3.3177, + "step": 27645 + }, + { + "epoch": 1.878651990759614, + "grad_norm": 1.1180208921432495, + "learning_rate": 0.0007651685011550482, + "loss": 3.4648, + "step": 27650 + }, + { + "epoch": 1.8789917108302758, + "grad_norm": 1.1307436227798462, + "learning_rate": 0.0007651260361462156, + "loss": 3.6143, + "step": 27655 + }, + { + "epoch": 1.8793314309009377, + "grad_norm": 1.5474539995193481, + "learning_rate": 0.0007650835711373828, + "loss": 3.6104, + "step": 27660 + }, + { + "epoch": 1.8796711509715993, + "grad_norm": 1.0800511837005615, + "learning_rate": 0.00076504110612855, + "loss": 3.7443, + "step": 27665 + }, + { + "epoch": 1.8800108710422612, + "grad_norm": 0.7681803107261658, + "learning_rate": 0.0007649986411197175, + "loss": 3.5591, + "step": 27670 + }, + { + "epoch": 1.880350591112923, + "grad_norm": 1.0217119455337524, + "learning_rate": 0.0007649561761108847, + "loss": 3.461, + "step": 27675 + }, + { + "epoch": 1.8806903111835847, + "grad_norm": 0.8130105137825012, + "learning_rate": 0.0007649137111020519, + "loss": 3.7187, + "step": 27680 + }, + { + "epoch": 1.8810300312542465, + "grad_norm": 0.7654531002044678, + "learning_rate": 0.0007648712460932192, + "loss": 3.58, + "step": 27685 + }, + { + "epoch": 1.8813697513249084, + "grad_norm": 0.702591061592102, + "learning_rate": 0.0007648287810843865, + "loss": 3.568, + "step": 27690 + }, + { + "epoch": 1.88170947139557, + "grad_norm": 1.335799217224121, + "learning_rate": 0.0007647863160755537, + "loss": 3.8125, + "step": 27695 + }, + { + "epoch": 1.8820491914662318, + "grad_norm": 0.9578614234924316, + "learning_rate": 0.000764743851066721, + "loss": 3.4757, + "step": 27700 + }, + { + "epoch": 1.8823889115368937, + "grad_norm": 0.9437412023544312, + "learning_rate": 0.0007647013860578884, + "loss": 3.5117, + "step": 27705 + }, + { + "epoch": 1.8827286316075553, + "grad_norm": 0.9306308031082153, + "learning_rate": 0.0007646589210490556, + "loss": 3.5703, + "step": 27710 + }, + { + "epoch": 1.8830683516782172, + "grad_norm": 0.7703170776367188, + "learning_rate": 0.0007646164560402229, + "loss": 3.7251, + "step": 27715 + }, + { + "epoch": 1.883408071748879, + "grad_norm": 0.8097624778747559, + "learning_rate": 0.0007645739910313901, + "loss": 3.5606, + "step": 27720 + }, + { + "epoch": 1.8837477918195407, + "grad_norm": 0.7362086176872253, + "learning_rate": 0.0007645315260225574, + "loss": 3.578, + "step": 27725 + }, + { + "epoch": 1.8840875118902025, + "grad_norm": 1.0768812894821167, + "learning_rate": 0.0007644890610137247, + "loss": 3.5988, + "step": 27730 + }, + { + "epoch": 1.8844272319608644, + "grad_norm": 0.8561644554138184, + "learning_rate": 0.0007644465960048919, + "loss": 3.6768, + "step": 27735 + }, + { + "epoch": 1.884766952031526, + "grad_norm": 0.9110159277915955, + "learning_rate": 0.0007644041309960593, + "loss": 3.6808, + "step": 27740 + }, + { + "epoch": 1.8851066721021879, + "grad_norm": 0.7943991422653198, + "learning_rate": 0.0007643616659872266, + "loss": 3.4954, + "step": 27745 + }, + { + "epoch": 1.8854463921728497, + "grad_norm": 0.6693992614746094, + "learning_rate": 0.0007643192009783938, + "loss": 3.4861, + "step": 27750 + }, + { + "epoch": 1.8857861122435113, + "grad_norm": 0.7504506707191467, + "learning_rate": 0.000764276735969561, + "loss": 3.4306, + "step": 27755 + }, + { + "epoch": 1.886125832314173, + "grad_norm": 0.9541374444961548, + "learning_rate": 0.0007642342709607284, + "loss": 3.7398, + "step": 27760 + }, + { + "epoch": 1.886465552384835, + "grad_norm": 1.605481505393982, + "learning_rate": 0.0007641918059518956, + "loss": 3.6045, + "step": 27765 + }, + { + "epoch": 1.8868052724554967, + "grad_norm": 1.0158085823059082, + "learning_rate": 0.0007641493409430629, + "loss": 3.7272, + "step": 27770 + }, + { + "epoch": 1.8871449925261583, + "grad_norm": 0.8868354558944702, + "learning_rate": 0.0007641068759342303, + "loss": 3.396, + "step": 27775 + }, + { + "epoch": 1.8874847125968204, + "grad_norm": 0.6586178541183472, + "learning_rate": 0.0007640644109253975, + "loss": 3.4357, + "step": 27780 + }, + { + "epoch": 1.887824432667482, + "grad_norm": 1.1147302389144897, + "learning_rate": 0.0007640219459165648, + "loss": 3.3869, + "step": 27785 + }, + { + "epoch": 1.8881641527381436, + "grad_norm": 1.4214062690734863, + "learning_rate": 0.0007639794809077321, + "loss": 3.7593, + "step": 27790 + }, + { + "epoch": 1.8885038728088055, + "grad_norm": 0.7250053286552429, + "learning_rate": 0.0007639370158988993, + "loss": 3.5766, + "step": 27795 + }, + { + "epoch": 1.8888435928794673, + "grad_norm": 0.8354542851448059, + "learning_rate": 0.0007638945508900666, + "loss": 3.5954, + "step": 27800 + }, + { + "epoch": 1.889183312950129, + "grad_norm": 0.7880530953407288, + "learning_rate": 0.0007638520858812339, + "loss": 3.5222, + "step": 27805 + }, + { + "epoch": 1.8895230330207908, + "grad_norm": 2.096595287322998, + "learning_rate": 0.0007638096208724012, + "loss": 3.8444, + "step": 27810 + }, + { + "epoch": 1.8898627530914527, + "grad_norm": 1.105875849723816, + "learning_rate": 0.0007637671558635685, + "loss": 3.5127, + "step": 27815 + }, + { + "epoch": 1.8902024731621143, + "grad_norm": 0.8182263970375061, + "learning_rate": 0.0007637246908547357, + "loss": 3.5066, + "step": 27820 + }, + { + "epoch": 1.8905421932327762, + "grad_norm": 1.4389373064041138, + "learning_rate": 0.000763682225845903, + "loss": 3.5328, + "step": 27825 + }, + { + "epoch": 1.890881913303438, + "grad_norm": 0.9205893874168396, + "learning_rate": 0.0007636397608370703, + "loss": 3.9885, + "step": 27830 + }, + { + "epoch": 1.8912216333740997, + "grad_norm": 0.9334415793418884, + "learning_rate": 0.0007635972958282375, + "loss": 3.7796, + "step": 27835 + }, + { + "epoch": 1.8915613534447615, + "grad_norm": 0.7761504650115967, + "learning_rate": 0.0007635548308194048, + "loss": 3.389, + "step": 27840 + }, + { + "epoch": 1.8919010735154234, + "grad_norm": 0.9125584363937378, + "learning_rate": 0.0007635123658105722, + "loss": 3.5199, + "step": 27845 + }, + { + "epoch": 1.892240793586085, + "grad_norm": 1.073754072189331, + "learning_rate": 0.0007634699008017394, + "loss": 3.5102, + "step": 27850 + }, + { + "epoch": 1.8925805136567468, + "grad_norm": 0.9121300578117371, + "learning_rate": 0.0007634274357929067, + "loss": 3.9014, + "step": 27855 + }, + { + "epoch": 1.8929202337274087, + "grad_norm": 0.9460108876228333, + "learning_rate": 0.000763384970784074, + "loss": 3.8095, + "step": 27860 + }, + { + "epoch": 1.8932599537980703, + "grad_norm": 1.0225892066955566, + "learning_rate": 0.0007633425057752412, + "loss": 3.4162, + "step": 27865 + }, + { + "epoch": 1.8935996738687322, + "grad_norm": 0.9070990681648254, + "learning_rate": 0.0007633000407664084, + "loss": 3.8772, + "step": 27870 + }, + { + "epoch": 1.893939393939394, + "grad_norm": 1.0600470304489136, + "learning_rate": 0.0007632575757575758, + "loss": 3.7199, + "step": 27875 + }, + { + "epoch": 1.8942791140100557, + "grad_norm": 0.764045000076294, + "learning_rate": 0.0007632151107487431, + "loss": 3.6753, + "step": 27880 + }, + { + "epoch": 1.8946188340807175, + "grad_norm": 0.9605132341384888, + "learning_rate": 0.0007631726457399103, + "loss": 3.4173, + "step": 27885 + }, + { + "epoch": 1.8949585541513794, + "grad_norm": 0.6812487840652466, + "learning_rate": 0.0007631301807310777, + "loss": 3.5835, + "step": 27890 + }, + { + "epoch": 1.895298274222041, + "grad_norm": 1.0720773935317993, + "learning_rate": 0.0007630877157222449, + "loss": 3.7022, + "step": 27895 + }, + { + "epoch": 1.8956379942927029, + "grad_norm": 1.05709969997406, + "learning_rate": 0.0007630452507134121, + "loss": 3.7093, + "step": 27900 + }, + { + "epoch": 1.8959777143633647, + "grad_norm": 1.2263301610946655, + "learning_rate": 0.0007630027857045795, + "loss": 3.3946, + "step": 27905 + }, + { + "epoch": 1.8963174344340263, + "grad_norm": 0.8029810190200806, + "learning_rate": 0.0007629603206957467, + "loss": 3.6446, + "step": 27910 + }, + { + "epoch": 1.8966571545046882, + "grad_norm": 0.8597691655158997, + "learning_rate": 0.000762917855686914, + "loss": 3.5462, + "step": 27915 + }, + { + "epoch": 1.89699687457535, + "grad_norm": 0.8356927037239075, + "learning_rate": 0.0007628753906780813, + "loss": 3.5887, + "step": 27920 + }, + { + "epoch": 1.8973365946460117, + "grad_norm": 1.0175747871398926, + "learning_rate": 0.0007628329256692486, + "loss": 3.6338, + "step": 27925 + }, + { + "epoch": 1.8976763147166733, + "grad_norm": 0.7892065644264221, + "learning_rate": 0.0007627904606604158, + "loss": 3.6273, + "step": 27930 + }, + { + "epoch": 1.8980160347873354, + "grad_norm": 0.8458698391914368, + "learning_rate": 0.0007627479956515831, + "loss": 3.5968, + "step": 27935 + }, + { + "epoch": 1.898355754857997, + "grad_norm": 0.8552168607711792, + "learning_rate": 0.0007627055306427504, + "loss": 3.4676, + "step": 27940 + }, + { + "epoch": 1.8986954749286586, + "grad_norm": 1.0278396606445312, + "learning_rate": 0.0007626630656339176, + "loss": 3.4311, + "step": 27945 + }, + { + "epoch": 1.8990351949993207, + "grad_norm": 0.7547438144683838, + "learning_rate": 0.000762620600625085, + "loss": 3.5143, + "step": 27950 + }, + { + "epoch": 1.8993749150699824, + "grad_norm": 0.7855194807052612, + "learning_rate": 0.0007625781356162523, + "loss": 3.3154, + "step": 27955 + }, + { + "epoch": 1.899714635140644, + "grad_norm": 0.7551230192184448, + "learning_rate": 0.0007625356706074195, + "loss": 3.4993, + "step": 27960 + }, + { + "epoch": 1.9000543552113058, + "grad_norm": 1.033882737159729, + "learning_rate": 0.0007624932055985868, + "loss": 3.7465, + "step": 27965 + }, + { + "epoch": 1.9003940752819677, + "grad_norm": 0.7596902847290039, + "learning_rate": 0.000762450740589754, + "loss": 3.7756, + "step": 27970 + }, + { + "epoch": 1.9007337953526293, + "grad_norm": 0.8528746366500854, + "learning_rate": 0.0007624082755809213, + "loss": 3.5839, + "step": 27975 + }, + { + "epoch": 1.9010735154232912, + "grad_norm": 0.6484443545341492, + "learning_rate": 0.0007623658105720886, + "loss": 3.6108, + "step": 27980 + }, + { + "epoch": 1.901413235493953, + "grad_norm": 1.2080998420715332, + "learning_rate": 0.0007623233455632559, + "loss": 3.906, + "step": 27985 + }, + { + "epoch": 1.9017529555646147, + "grad_norm": 1.0229791402816772, + "learning_rate": 0.0007622808805544232, + "loss": 3.6065, + "step": 27990 + }, + { + "epoch": 1.9020926756352765, + "grad_norm": 1.0935660600662231, + "learning_rate": 0.0007622384155455905, + "loss": 3.4556, + "step": 27995 + }, + { + "epoch": 1.9024323957059384, + "grad_norm": 0.7671613097190857, + "learning_rate": 0.0007621959505367577, + "loss": 3.5871, + "step": 28000 + }, + { + "epoch": 1.9027721157766, + "grad_norm": 0.967738687992096, + "learning_rate": 0.0007621534855279249, + "loss": 3.6703, + "step": 28005 + }, + { + "epoch": 1.9031118358472618, + "grad_norm": 0.8478018641471863, + "learning_rate": 0.0007621110205190923, + "loss": 3.8198, + "step": 28010 + }, + { + "epoch": 1.9034515559179237, + "grad_norm": 0.9766073226928711, + "learning_rate": 0.0007620685555102595, + "loss": 3.4404, + "step": 28015 + }, + { + "epoch": 1.9037912759885853, + "grad_norm": 0.8470941781997681, + "learning_rate": 0.0007620260905014268, + "loss": 3.668, + "step": 28020 + }, + { + "epoch": 1.9041309960592472, + "grad_norm": 0.9228593707084656, + "learning_rate": 0.0007619836254925942, + "loss": 3.6934, + "step": 28025 + }, + { + "epoch": 1.904470716129909, + "grad_norm": 0.7718645930290222, + "learning_rate": 0.0007619411604837614, + "loss": 3.7298, + "step": 28030 + }, + { + "epoch": 1.9048104362005707, + "grad_norm": 3.500680446624756, + "learning_rate": 0.0007618986954749286, + "loss": 3.5958, + "step": 28035 + }, + { + "epoch": 1.9051501562712325, + "grad_norm": 1.4715577363967896, + "learning_rate": 0.000761856230466096, + "loss": 3.5392, + "step": 28040 + }, + { + "epoch": 1.9054898763418944, + "grad_norm": 0.7793188095092773, + "learning_rate": 0.0007618137654572632, + "loss": 3.7854, + "step": 28045 + }, + { + "epoch": 1.905829596412556, + "grad_norm": 0.9477581977844238, + "learning_rate": 0.0007617713004484304, + "loss": 3.4512, + "step": 28050 + }, + { + "epoch": 1.9061693164832179, + "grad_norm": 0.9024282693862915, + "learning_rate": 0.0007617288354395979, + "loss": 3.5512, + "step": 28055 + }, + { + "epoch": 1.9065090365538797, + "grad_norm": 0.7619509100914001, + "learning_rate": 0.0007616863704307651, + "loss": 3.7984, + "step": 28060 + }, + { + "epoch": 1.9068487566245413, + "grad_norm": 0.7706624865531921, + "learning_rate": 0.0007616439054219323, + "loss": 3.3561, + "step": 28065 + }, + { + "epoch": 1.9071884766952032, + "grad_norm": 0.742607831954956, + "learning_rate": 0.0007616014404130996, + "loss": 3.4511, + "step": 28070 + }, + { + "epoch": 1.907528196765865, + "grad_norm": 1.0033817291259766, + "learning_rate": 0.0007615589754042669, + "loss": 3.6443, + "step": 28075 + }, + { + "epoch": 1.9078679168365267, + "grad_norm": 0.7715677618980408, + "learning_rate": 0.0007615165103954341, + "loss": 3.5183, + "step": 28080 + }, + { + "epoch": 1.9082076369071885, + "grad_norm": 3.076587200164795, + "learning_rate": 0.0007614740453866014, + "loss": 3.5551, + "step": 28085 + }, + { + "epoch": 1.9085473569778504, + "grad_norm": 0.718163251876831, + "learning_rate": 0.0007614315803777688, + "loss": 3.8148, + "step": 28090 + }, + { + "epoch": 1.908887077048512, + "grad_norm": 0.8399987816810608, + "learning_rate": 0.000761389115368936, + "loss": 3.456, + "step": 28095 + }, + { + "epoch": 1.9092267971191736, + "grad_norm": 0.9357495307922363, + "learning_rate": 0.0007613466503601033, + "loss": 3.4895, + "step": 28100 + }, + { + "epoch": 1.9095665171898357, + "grad_norm": 0.7830958962440491, + "learning_rate": 0.0007613041853512705, + "loss": 3.504, + "step": 28105 + }, + { + "epoch": 1.9099062372604974, + "grad_norm": 0.9575322866439819, + "learning_rate": 0.0007612617203424379, + "loss": 3.6611, + "step": 28110 + }, + { + "epoch": 1.910245957331159, + "grad_norm": 0.9187508225440979, + "learning_rate": 0.0007612192553336051, + "loss": 3.6058, + "step": 28115 + }, + { + "epoch": 1.910585677401821, + "grad_norm": 0.7837285399436951, + "learning_rate": 0.0007611767903247723, + "loss": 3.7639, + "step": 28120 + }, + { + "epoch": 1.9109253974724827, + "grad_norm": 0.8744210600852966, + "learning_rate": 0.0007611343253159398, + "loss": 3.8573, + "step": 28125 + }, + { + "epoch": 1.9112651175431443, + "grad_norm": 1.0014907121658325, + "learning_rate": 0.000761091860307107, + "loss": 3.5586, + "step": 28130 + }, + { + "epoch": 1.9116048376138062, + "grad_norm": 1.105457067489624, + "learning_rate": 0.0007610493952982742, + "loss": 3.4191, + "step": 28135 + }, + { + "epoch": 1.911944557684468, + "grad_norm": 1.0035400390625, + "learning_rate": 0.0007610069302894416, + "loss": 3.9603, + "step": 28140 + }, + { + "epoch": 1.9122842777551297, + "grad_norm": 0.7402544617652893, + "learning_rate": 0.0007609644652806088, + "loss": 3.6728, + "step": 28145 + }, + { + "epoch": 1.9126239978257915, + "grad_norm": 0.7774442434310913, + "learning_rate": 0.000760922000271776, + "loss": 3.5254, + "step": 28150 + }, + { + "epoch": 1.9129637178964534, + "grad_norm": 0.9557531476020813, + "learning_rate": 0.0007608795352629433, + "loss": 3.695, + "step": 28155 + }, + { + "epoch": 1.913303437967115, + "grad_norm": 1.0287468433380127, + "learning_rate": 0.0007608370702541107, + "loss": 3.7225, + "step": 28160 + }, + { + "epoch": 1.9136431580377768, + "grad_norm": 0.9337747097015381, + "learning_rate": 0.0007607946052452779, + "loss": 3.8231, + "step": 28165 + }, + { + "epoch": 1.9139828781084387, + "grad_norm": 0.9472321271896362, + "learning_rate": 0.0007607521402364452, + "loss": 3.543, + "step": 28170 + }, + { + "epoch": 1.9143225981791003, + "grad_norm": 1.073047161102295, + "learning_rate": 0.0007607096752276125, + "loss": 3.393, + "step": 28175 + }, + { + "epoch": 1.9146623182497622, + "grad_norm": 0.8966087102890015, + "learning_rate": 0.0007606672102187797, + "loss": 3.7088, + "step": 28180 + }, + { + "epoch": 1.915002038320424, + "grad_norm": 1.0451486110687256, + "learning_rate": 0.000760624745209947, + "loss": 3.6303, + "step": 28185 + }, + { + "epoch": 1.9153417583910857, + "grad_norm": 0.9368476271629333, + "learning_rate": 0.0007605822802011143, + "loss": 3.5946, + "step": 28190 + }, + { + "epoch": 1.9156814784617475, + "grad_norm": 0.761813759803772, + "learning_rate": 0.0007605398151922816, + "loss": 3.4495, + "step": 28195 + }, + { + "epoch": 1.9160211985324094, + "grad_norm": 0.8372112512588501, + "learning_rate": 0.0007604973501834489, + "loss": 3.9955, + "step": 28200 + }, + { + "epoch": 1.916360918603071, + "grad_norm": 1.0954171419143677, + "learning_rate": 0.0007604548851746161, + "loss": 3.6138, + "step": 28205 + }, + { + "epoch": 1.9167006386737329, + "grad_norm": 0.7364549040794373, + "learning_rate": 0.0007604124201657834, + "loss": 3.6222, + "step": 28210 + }, + { + "epoch": 1.9170403587443947, + "grad_norm": 0.8073201775550842, + "learning_rate": 0.0007603699551569507, + "loss": 3.4635, + "step": 28215 + }, + { + "epoch": 1.9173800788150563, + "grad_norm": 0.8864088654518127, + "learning_rate": 0.0007603274901481179, + "loss": 3.5409, + "step": 28220 + }, + { + "epoch": 1.9177197988857182, + "grad_norm": 0.9455179572105408, + "learning_rate": 0.0007602850251392852, + "loss": 3.3552, + "step": 28225 + }, + { + "epoch": 1.91805951895638, + "grad_norm": 0.8603058457374573, + "learning_rate": 0.0007602425601304526, + "loss": 3.2582, + "step": 28230 + }, + { + "epoch": 1.9183992390270417, + "grad_norm": 0.8076804280281067, + "learning_rate": 0.0007602000951216198, + "loss": 3.6138, + "step": 28235 + }, + { + "epoch": 1.9187389590977035, + "grad_norm": 0.779769778251648, + "learning_rate": 0.0007601576301127871, + "loss": 3.6792, + "step": 28240 + }, + { + "epoch": 1.9190786791683654, + "grad_norm": 1.0759003162384033, + "learning_rate": 0.0007601151651039544, + "loss": 3.5209, + "step": 28245 + }, + { + "epoch": 1.919418399239027, + "grad_norm": 1.349061369895935, + "learning_rate": 0.0007600727000951216, + "loss": 3.6442, + "step": 28250 + }, + { + "epoch": 1.9197581193096889, + "grad_norm": 0.8423357605934143, + "learning_rate": 0.0007600302350862888, + "loss": 3.746, + "step": 28255 + }, + { + "epoch": 1.9200978393803507, + "grad_norm": 0.8904449343681335, + "learning_rate": 0.0007599877700774563, + "loss": 3.6733, + "step": 28260 + }, + { + "epoch": 1.9204375594510124, + "grad_norm": 1.0457406044006348, + "learning_rate": 0.0007599453050686235, + "loss": 3.7641, + "step": 28265 + }, + { + "epoch": 1.920777279521674, + "grad_norm": 1.1688271760940552, + "learning_rate": 0.0007599028400597907, + "loss": 3.721, + "step": 28270 + }, + { + "epoch": 1.921116999592336, + "grad_norm": 0.8149852156639099, + "learning_rate": 0.0007598603750509581, + "loss": 3.6576, + "step": 28275 + }, + { + "epoch": 1.9214567196629977, + "grad_norm": 0.6973206996917725, + "learning_rate": 0.0007598179100421253, + "loss": 3.4242, + "step": 28280 + }, + { + "epoch": 1.9217964397336593, + "grad_norm": 0.947292149066925, + "learning_rate": 0.0007597754450332925, + "loss": 3.6933, + "step": 28285 + }, + { + "epoch": 1.9221361598043214, + "grad_norm": 0.7033851146697998, + "learning_rate": 0.0007597329800244599, + "loss": 3.7296, + "step": 28290 + }, + { + "epoch": 1.922475879874983, + "grad_norm": 0.8169011473655701, + "learning_rate": 0.0007596905150156272, + "loss": 3.5704, + "step": 28295 + }, + { + "epoch": 1.9228155999456447, + "grad_norm": 0.811951756477356, + "learning_rate": 0.0007596480500067944, + "loss": 3.7214, + "step": 28300 + }, + { + "epoch": 1.9231553200163065, + "grad_norm": 0.6803563833236694, + "learning_rate": 0.0007596055849979618, + "loss": 3.6676, + "step": 28305 + }, + { + "epoch": 1.9234950400869684, + "grad_norm": 0.7913274168968201, + "learning_rate": 0.000759563119989129, + "loss": 3.7017, + "step": 28310 + }, + { + "epoch": 1.92383476015763, + "grad_norm": 0.7437870502471924, + "learning_rate": 0.0007595206549802962, + "loss": 3.846, + "step": 28315 + }, + { + "epoch": 1.9241744802282919, + "grad_norm": 0.8227376341819763, + "learning_rate": 0.0007594781899714635, + "loss": 3.6799, + "step": 28320 + }, + { + "epoch": 1.9245142002989537, + "grad_norm": 0.8750634789466858, + "learning_rate": 0.0007594357249626308, + "loss": 3.5371, + "step": 28325 + }, + { + "epoch": 1.9248539203696153, + "grad_norm": 4.627682209014893, + "learning_rate": 0.0007593932599537981, + "loss": 3.6174, + "step": 28330 + }, + { + "epoch": 1.9251936404402772, + "grad_norm": 1.0895839929580688, + "learning_rate": 0.0007593507949449654, + "loss": 3.5953, + "step": 28335 + }, + { + "epoch": 1.925533360510939, + "grad_norm": 0.8343336582183838, + "learning_rate": 0.0007593083299361327, + "loss": 3.6144, + "step": 28340 + }, + { + "epoch": 1.9258730805816007, + "grad_norm": 0.8747314810752869, + "learning_rate": 0.0007592658649272999, + "loss": 3.5115, + "step": 28345 + }, + { + "epoch": 1.9262128006522625, + "grad_norm": 0.9775891304016113, + "learning_rate": 0.0007592233999184672, + "loss": 3.516, + "step": 28350 + }, + { + "epoch": 1.9265525207229244, + "grad_norm": 0.9417952299118042, + "learning_rate": 0.0007591809349096344, + "loss": 3.438, + "step": 28355 + }, + { + "epoch": 1.926892240793586, + "grad_norm": 1.0056073665618896, + "learning_rate": 0.0007591384699008017, + "loss": 3.4382, + "step": 28360 + }, + { + "epoch": 1.9272319608642479, + "grad_norm": 0.9947229623794556, + "learning_rate": 0.0007590960048919691, + "loss": 3.6533, + "step": 28365 + }, + { + "epoch": 1.9275716809349097, + "grad_norm": 0.7735212445259094, + "learning_rate": 0.0007590535398831363, + "loss": 3.4921, + "step": 28370 + }, + { + "epoch": 1.9279114010055713, + "grad_norm": 0.8552477955818176, + "learning_rate": 0.0007590110748743036, + "loss": 3.5746, + "step": 28375 + }, + { + "epoch": 1.9282511210762332, + "grad_norm": 0.9084644913673401, + "learning_rate": 0.0007589686098654709, + "loss": 3.6247, + "step": 28380 + }, + { + "epoch": 1.928590841146895, + "grad_norm": 0.7161413431167603, + "learning_rate": 0.0007589261448566381, + "loss": 3.3702, + "step": 28385 + }, + { + "epoch": 1.9289305612175567, + "grad_norm": 2.041700839996338, + "learning_rate": 0.0007588836798478053, + "loss": 3.7149, + "step": 28390 + }, + { + "epoch": 1.9292702812882185, + "grad_norm": 0.7780112624168396, + "learning_rate": 0.0007588412148389727, + "loss": 3.6575, + "step": 28395 + }, + { + "epoch": 1.9296100013588804, + "grad_norm": 0.9971717596054077, + "learning_rate": 0.00075879874983014, + "loss": 3.7108, + "step": 28400 + }, + { + "epoch": 1.929949721429542, + "grad_norm": 0.8554769158363342, + "learning_rate": 0.0007587562848213072, + "loss": 3.6328, + "step": 28405 + }, + { + "epoch": 1.9302894415002039, + "grad_norm": 0.6781002879142761, + "learning_rate": 0.0007587138198124746, + "loss": 3.5295, + "step": 28410 + }, + { + "epoch": 1.9306291615708657, + "grad_norm": 1.050248384475708, + "learning_rate": 0.0007586713548036418, + "loss": 3.2926, + "step": 28415 + }, + { + "epoch": 1.9309688816415274, + "grad_norm": 0.7920815348625183, + "learning_rate": 0.000758628889794809, + "loss": 3.8175, + "step": 28420 + }, + { + "epoch": 1.9313086017121892, + "grad_norm": 0.818256676197052, + "learning_rate": 0.0007585864247859764, + "loss": 3.5565, + "step": 28425 + }, + { + "epoch": 1.931648321782851, + "grad_norm": 0.8688068389892578, + "learning_rate": 0.0007585439597771436, + "loss": 3.5658, + "step": 28430 + }, + { + "epoch": 1.9319880418535127, + "grad_norm": 1.0677967071533203, + "learning_rate": 0.0007585014947683109, + "loss": 3.6674, + "step": 28435 + }, + { + "epoch": 1.9323277619241743, + "grad_norm": 0.902638852596283, + "learning_rate": 0.0007584590297594783, + "loss": 3.5321, + "step": 28440 + }, + { + "epoch": 1.9326674819948364, + "grad_norm": 0.7627798914909363, + "learning_rate": 0.0007584165647506455, + "loss": 3.5257, + "step": 28445 + }, + { + "epoch": 1.933007202065498, + "grad_norm": 0.9267018437385559, + "learning_rate": 0.0007583740997418128, + "loss": 3.7782, + "step": 28450 + }, + { + "epoch": 1.9333469221361597, + "grad_norm": 0.9510639309883118, + "learning_rate": 0.00075833163473298, + "loss": 3.6926, + "step": 28455 + }, + { + "epoch": 1.9336866422068217, + "grad_norm": 0.6269004344940186, + "learning_rate": 0.0007582891697241473, + "loss": 3.7019, + "step": 28460 + }, + { + "epoch": 1.9340263622774834, + "grad_norm": 0.9256649017333984, + "learning_rate": 0.0007582467047153146, + "loss": 3.6565, + "step": 28465 + }, + { + "epoch": 1.934366082348145, + "grad_norm": 0.7546939849853516, + "learning_rate": 0.0007582042397064819, + "loss": 3.4047, + "step": 28470 + }, + { + "epoch": 1.9347058024188069, + "grad_norm": 1.9614346027374268, + "learning_rate": 0.0007581617746976492, + "loss": 3.8338, + "step": 28475 + }, + { + "epoch": 1.9350455224894687, + "grad_norm": 0.7255362272262573, + "learning_rate": 0.0007581193096888165, + "loss": 3.6765, + "step": 28480 + }, + { + "epoch": 1.9353852425601303, + "grad_norm": 0.8525300621986389, + "learning_rate": 0.0007580768446799837, + "loss": 3.5382, + "step": 28485 + }, + { + "epoch": 1.9357249626307922, + "grad_norm": 0.7200602889060974, + "learning_rate": 0.000758034379671151, + "loss": 3.3816, + "step": 28490 + }, + { + "epoch": 1.936064682701454, + "grad_norm": 0.9942429065704346, + "learning_rate": 0.0007579919146623183, + "loss": 3.6353, + "step": 28495 + }, + { + "epoch": 1.9364044027721157, + "grad_norm": 0.8485697507858276, + "learning_rate": 0.0007579494496534855, + "loss": 3.6562, + "step": 28500 + }, + { + "epoch": 1.9367441228427775, + "grad_norm": 0.8071969747543335, + "learning_rate": 0.0007579069846446528, + "loss": 3.6914, + "step": 28505 + }, + { + "epoch": 1.9370838429134394, + "grad_norm": 0.968704879283905, + "learning_rate": 0.0007578645196358202, + "loss": 3.4869, + "step": 28510 + }, + { + "epoch": 1.937423562984101, + "grad_norm": 0.8211969137191772, + "learning_rate": 0.0007578220546269874, + "loss": 3.6826, + "step": 28515 + }, + { + "epoch": 1.9377632830547629, + "grad_norm": 0.8874754309654236, + "learning_rate": 0.0007577795896181546, + "loss": 3.5212, + "step": 28520 + }, + { + "epoch": 1.9381030031254247, + "grad_norm": 0.7931695580482483, + "learning_rate": 0.000757737124609322, + "loss": 3.6002, + "step": 28525 + }, + { + "epoch": 1.9384427231960863, + "grad_norm": 0.7379376292228699, + "learning_rate": 0.0007576946596004892, + "loss": 3.6429, + "step": 28530 + }, + { + "epoch": 1.9387824432667482, + "grad_norm": 0.8322670459747314, + "learning_rate": 0.0007576521945916564, + "loss": 3.6122, + "step": 28535 + }, + { + "epoch": 1.93912216333741, + "grad_norm": 2.0350444316864014, + "learning_rate": 0.0007576097295828239, + "loss": 3.7357, + "step": 28540 + }, + { + "epoch": 1.9394618834080717, + "grad_norm": 0.8767146468162537, + "learning_rate": 0.0007575672645739911, + "loss": 3.6049, + "step": 28545 + }, + { + "epoch": 1.9398016034787335, + "grad_norm": 1.0677225589752197, + "learning_rate": 0.0007575247995651583, + "loss": 3.664, + "step": 28550 + }, + { + "epoch": 1.9401413235493954, + "grad_norm": 0.7669855952262878, + "learning_rate": 0.0007574823345563256, + "loss": 3.5136, + "step": 28555 + }, + { + "epoch": 1.940481043620057, + "grad_norm": 0.803428053855896, + "learning_rate": 0.0007574398695474929, + "loss": 3.7272, + "step": 28560 + }, + { + "epoch": 1.9408207636907189, + "grad_norm": 0.7286089658737183, + "learning_rate": 0.0007573974045386601, + "loss": 3.5448, + "step": 28565 + }, + { + "epoch": 1.9411604837613807, + "grad_norm": 0.9872887134552002, + "learning_rate": 0.0007573549395298274, + "loss": 3.6438, + "step": 28570 + }, + { + "epoch": 1.9415002038320424, + "grad_norm": 0.8057947754859924, + "learning_rate": 0.0007573124745209948, + "loss": 3.4015, + "step": 28575 + }, + { + "epoch": 1.9418399239027042, + "grad_norm": 1.7552804946899414, + "learning_rate": 0.000757270009512162, + "loss": 3.6373, + "step": 28580 + }, + { + "epoch": 1.942179643973366, + "grad_norm": 0.6760085225105286, + "learning_rate": 0.0007572275445033293, + "loss": 3.8561, + "step": 28585 + }, + { + "epoch": 1.9425193640440277, + "grad_norm": 0.9069851636886597, + "learning_rate": 0.0007571850794944966, + "loss": 3.638, + "step": 28590 + }, + { + "epoch": 1.9428590841146895, + "grad_norm": 0.7618457674980164, + "learning_rate": 0.0007571426144856638, + "loss": 3.5034, + "step": 28595 + }, + { + "epoch": 1.9431988041853514, + "grad_norm": 1.030716896057129, + "learning_rate": 0.0007571001494768311, + "loss": 3.2856, + "step": 28600 + }, + { + "epoch": 1.943538524256013, + "grad_norm": 0.914871096611023, + "learning_rate": 0.0007570576844679983, + "loss": 3.5833, + "step": 28605 + }, + { + "epoch": 1.9438782443266747, + "grad_norm": 1.038490891456604, + "learning_rate": 0.0007570152194591657, + "loss": 3.4763, + "step": 28610 + }, + { + "epoch": 1.9442179643973367, + "grad_norm": 0.8067642450332642, + "learning_rate": 0.000756972754450333, + "loss": 3.5549, + "step": 28615 + }, + { + "epoch": 1.9445576844679984, + "grad_norm": 1.0564671754837036, + "learning_rate": 0.0007569302894415002, + "loss": 3.1271, + "step": 28620 + }, + { + "epoch": 1.94489740453866, + "grad_norm": 0.8853978514671326, + "learning_rate": 0.0007568878244326675, + "loss": 3.4998, + "step": 28625 + }, + { + "epoch": 1.945237124609322, + "grad_norm": 0.7977197170257568, + "learning_rate": 0.0007568453594238348, + "loss": 3.5712, + "step": 28630 + }, + { + "epoch": 1.9455768446799837, + "grad_norm": 2.813751697540283, + "learning_rate": 0.000756802894415002, + "loss": 3.5143, + "step": 28635 + }, + { + "epoch": 1.9459165647506453, + "grad_norm": 0.7260422110557556, + "learning_rate": 0.0007567604294061692, + "loss": 3.4412, + "step": 28640 + }, + { + "epoch": 1.9462562848213072, + "grad_norm": 0.8211414813995361, + "learning_rate": 0.0007567179643973367, + "loss": 3.5394, + "step": 28645 + }, + { + "epoch": 1.946596004891969, + "grad_norm": 1.2254635095596313, + "learning_rate": 0.0007566754993885039, + "loss": 3.2922, + "step": 28650 + }, + { + "epoch": 1.9469357249626307, + "grad_norm": 0.8842107653617859, + "learning_rate": 0.0007566330343796711, + "loss": 3.7405, + "step": 28655 + }, + { + "epoch": 1.9472754450332925, + "grad_norm": 0.9992979168891907, + "learning_rate": 0.0007565905693708385, + "loss": 3.8473, + "step": 28660 + }, + { + "epoch": 1.9476151651039544, + "grad_norm": 1.2302290201187134, + "learning_rate": 0.0007565481043620057, + "loss": 3.4311, + "step": 28665 + }, + { + "epoch": 1.947954885174616, + "grad_norm": 1.0497682094573975, + "learning_rate": 0.0007565056393531729, + "loss": 3.5467, + "step": 28670 + }, + { + "epoch": 1.9482946052452779, + "grad_norm": 1.0691750049591064, + "learning_rate": 0.0007564631743443403, + "loss": 3.4072, + "step": 28675 + }, + { + "epoch": 1.9486343253159397, + "grad_norm": 0.7533948421478271, + "learning_rate": 0.0007564207093355076, + "loss": 3.792, + "step": 28680 + }, + { + "epoch": 1.9489740453866014, + "grad_norm": 1.0500904321670532, + "learning_rate": 0.0007563782443266748, + "loss": 3.7924, + "step": 28685 + }, + { + "epoch": 1.9493137654572632, + "grad_norm": 0.7273003458976746, + "learning_rate": 0.0007563357793178422, + "loss": 3.6146, + "step": 28690 + }, + { + "epoch": 1.949653485527925, + "grad_norm": 0.7840332388877869, + "learning_rate": 0.0007562933143090094, + "loss": 3.6988, + "step": 28695 + }, + { + "epoch": 1.9499932055985867, + "grad_norm": 0.9248551726341248, + "learning_rate": 0.0007562508493001766, + "loss": 3.6088, + "step": 28700 + }, + { + "epoch": 1.9503329256692485, + "grad_norm": 0.8148496747016907, + "learning_rate": 0.0007562083842913439, + "loss": 3.516, + "step": 28705 + }, + { + "epoch": 1.9506726457399104, + "grad_norm": 1.110105037689209, + "learning_rate": 0.0007561659192825112, + "loss": 3.8206, + "step": 28710 + }, + { + "epoch": 1.951012365810572, + "grad_norm": 0.8824129104614258, + "learning_rate": 0.0007561234542736785, + "loss": 3.6933, + "step": 28715 + }, + { + "epoch": 1.9513520858812339, + "grad_norm": 0.7973528504371643, + "learning_rate": 0.0007560809892648458, + "loss": 3.575, + "step": 28720 + }, + { + "epoch": 1.9516918059518957, + "grad_norm": 2.1247591972351074, + "learning_rate": 0.0007560385242560131, + "loss": 3.6658, + "step": 28725 + }, + { + "epoch": 1.9520315260225574, + "grad_norm": 1.136062502861023, + "learning_rate": 0.0007559960592471803, + "loss": 3.6825, + "step": 28730 + }, + { + "epoch": 1.9523712460932192, + "grad_norm": 0.7387412786483765, + "learning_rate": 0.0007559535942383476, + "loss": 3.7793, + "step": 28735 + }, + { + "epoch": 1.952710966163881, + "grad_norm": 0.7842884063720703, + "learning_rate": 0.0007559111292295148, + "loss": 3.7047, + "step": 28740 + }, + { + "epoch": 1.9530506862345427, + "grad_norm": 0.8975210189819336, + "learning_rate": 0.0007558686642206821, + "loss": 3.8107, + "step": 28745 + }, + { + "epoch": 1.9533904063052046, + "grad_norm": 0.7974251508712769, + "learning_rate": 0.0007558261992118495, + "loss": 3.5251, + "step": 28750 + }, + { + "epoch": 1.9537301263758664, + "grad_norm": 0.9227864742279053, + "learning_rate": 0.0007557837342030167, + "loss": 3.6562, + "step": 28755 + }, + { + "epoch": 1.954069846446528, + "grad_norm": 0.773226797580719, + "learning_rate": 0.000755741269194184, + "loss": 3.6523, + "step": 28760 + }, + { + "epoch": 1.95440956651719, + "grad_norm": 1.1268770694732666, + "learning_rate": 0.0007556988041853513, + "loss": 3.7385, + "step": 28765 + }, + { + "epoch": 1.9547492865878517, + "grad_norm": 1.0929476022720337, + "learning_rate": 0.0007556563391765185, + "loss": 3.7262, + "step": 28770 + }, + { + "epoch": 1.9550890066585134, + "grad_norm": 0.9068654775619507, + "learning_rate": 0.0007556138741676858, + "loss": 3.5628, + "step": 28775 + }, + { + "epoch": 1.955428726729175, + "grad_norm": 0.9386857748031616, + "learning_rate": 0.0007555714091588531, + "loss": 3.729, + "step": 28780 + }, + { + "epoch": 1.955768446799837, + "grad_norm": 1.0286078453063965, + "learning_rate": 0.0007555289441500204, + "loss": 3.5653, + "step": 28785 + }, + { + "epoch": 1.9561081668704987, + "grad_norm": 0.6781588792800903, + "learning_rate": 0.0007554864791411878, + "loss": 3.7909, + "step": 28790 + }, + { + "epoch": 1.9564478869411603, + "grad_norm": 2.163759469985962, + "learning_rate": 0.000755444014132355, + "loss": 3.5412, + "step": 28795 + }, + { + "epoch": 1.9567876070118224, + "grad_norm": 0.683273434638977, + "learning_rate": 0.0007554015491235222, + "loss": 3.5831, + "step": 28800 + }, + { + "epoch": 1.957127327082484, + "grad_norm": 0.9136443734169006, + "learning_rate": 0.0007553590841146895, + "loss": 3.3184, + "step": 28805 + }, + { + "epoch": 1.9574670471531457, + "grad_norm": 0.9115235209465027, + "learning_rate": 0.0007553166191058568, + "loss": 3.7052, + "step": 28810 + }, + { + "epoch": 1.9578067672238075, + "grad_norm": 0.9279776215553284, + "learning_rate": 0.000755274154097024, + "loss": 3.6018, + "step": 28815 + }, + { + "epoch": 1.9581464872944694, + "grad_norm": 0.8505581021308899, + "learning_rate": 0.0007552316890881914, + "loss": 3.6655, + "step": 28820 + }, + { + "epoch": 1.958486207365131, + "grad_norm": 1.2644245624542236, + "learning_rate": 0.0007551892240793587, + "loss": 3.5212, + "step": 28825 + }, + { + "epoch": 1.9588259274357929, + "grad_norm": 1.1435060501098633, + "learning_rate": 0.0007551467590705259, + "loss": 3.3801, + "step": 28830 + }, + { + "epoch": 1.9591656475064547, + "grad_norm": 1.210832118988037, + "learning_rate": 0.0007551042940616932, + "loss": 3.41, + "step": 28835 + }, + { + "epoch": 1.9595053675771164, + "grad_norm": 0.7638642191886902, + "learning_rate": 0.0007550618290528604, + "loss": 3.6881, + "step": 28840 + }, + { + "epoch": 1.9598450876477782, + "grad_norm": 0.8216562271118164, + "learning_rate": 0.0007550193640440277, + "loss": 3.694, + "step": 28845 + }, + { + "epoch": 1.96018480771844, + "grad_norm": 1.1245285272598267, + "learning_rate": 0.0007549768990351951, + "loss": 3.6368, + "step": 28850 + }, + { + "epoch": 1.9605245277891017, + "grad_norm": 0.706417441368103, + "learning_rate": 0.0007549344340263623, + "loss": 3.5578, + "step": 28855 + }, + { + "epoch": 1.9608642478597635, + "grad_norm": 0.7962143421173096, + "learning_rate": 0.0007548919690175296, + "loss": 3.7875, + "step": 28860 + }, + { + "epoch": 1.9612039679304254, + "grad_norm": 0.6766414642333984, + "learning_rate": 0.0007548495040086969, + "loss": 3.6219, + "step": 28865 + }, + { + "epoch": 1.961543688001087, + "grad_norm": 0.8412807583808899, + "learning_rate": 0.0007548070389998641, + "loss": 3.5257, + "step": 28870 + }, + { + "epoch": 1.9618834080717489, + "grad_norm": 0.8144760131835938, + "learning_rate": 0.0007547645739910314, + "loss": 3.7681, + "step": 28875 + }, + { + "epoch": 1.9622231281424107, + "grad_norm": 0.6655876636505127, + "learning_rate": 0.0007547221089821987, + "loss": 3.5578, + "step": 28880 + }, + { + "epoch": 1.9625628482130724, + "grad_norm": 0.9093195796012878, + "learning_rate": 0.000754679643973366, + "loss": 3.3821, + "step": 28885 + }, + { + "epoch": 1.9629025682837342, + "grad_norm": 0.6724187135696411, + "learning_rate": 0.0007546371789645332, + "loss": 3.5931, + "step": 28890 + }, + { + "epoch": 1.963242288354396, + "grad_norm": 1.607333779335022, + "learning_rate": 0.0007545947139557006, + "loss": 3.5183, + "step": 28895 + }, + { + "epoch": 1.9635820084250577, + "grad_norm": 0.9282097220420837, + "learning_rate": 0.0007545522489468678, + "loss": 3.5025, + "step": 28900 + }, + { + "epoch": 1.9639217284957196, + "grad_norm": 1.0963257551193237, + "learning_rate": 0.000754509783938035, + "loss": 3.7237, + "step": 28905 + }, + { + "epoch": 1.9642614485663814, + "grad_norm": 0.8336236476898193, + "learning_rate": 0.0007544673189292024, + "loss": 3.4077, + "step": 28910 + }, + { + "epoch": 1.964601168637043, + "grad_norm": 0.8237790465354919, + "learning_rate": 0.0007544248539203696, + "loss": 3.4558, + "step": 28915 + }, + { + "epoch": 1.964940888707705, + "grad_norm": 0.7438591718673706, + "learning_rate": 0.0007543823889115369, + "loss": 3.737, + "step": 28920 + }, + { + "epoch": 1.9652806087783667, + "grad_norm": 0.8561109304428101, + "learning_rate": 0.0007543399239027043, + "loss": 3.529, + "step": 28925 + }, + { + "epoch": 1.9656203288490284, + "grad_norm": 0.9789109826087952, + "learning_rate": 0.0007542974588938715, + "loss": 3.5817, + "step": 28930 + }, + { + "epoch": 1.9659600489196902, + "grad_norm": 0.8005071878433228, + "learning_rate": 0.0007542549938850387, + "loss": 3.4739, + "step": 28935 + }, + { + "epoch": 1.966299768990352, + "grad_norm": 0.8143114447593689, + "learning_rate": 0.000754212528876206, + "loss": 3.5272, + "step": 28940 + }, + { + "epoch": 1.9666394890610137, + "grad_norm": 0.7283206582069397, + "learning_rate": 0.0007541700638673733, + "loss": 3.7608, + "step": 28945 + }, + { + "epoch": 1.9669792091316753, + "grad_norm": 1.002816915512085, + "learning_rate": 0.0007541275988585405, + "loss": 3.8601, + "step": 28950 + }, + { + "epoch": 1.9673189292023374, + "grad_norm": 1.4885766506195068, + "learning_rate": 0.0007540851338497079, + "loss": 3.4029, + "step": 28955 + }, + { + "epoch": 1.967658649272999, + "grad_norm": 0.8344660997390747, + "learning_rate": 0.0007540426688408752, + "loss": 3.3417, + "step": 28960 + }, + { + "epoch": 1.9679983693436607, + "grad_norm": 1.0546966791152954, + "learning_rate": 0.0007540002038320424, + "loss": 3.7401, + "step": 28965 + }, + { + "epoch": 1.9683380894143228, + "grad_norm": 0.9657679200172424, + "learning_rate": 0.0007539577388232097, + "loss": 3.5961, + "step": 28970 + }, + { + "epoch": 1.9686778094849844, + "grad_norm": 0.9433134198188782, + "learning_rate": 0.000753915273814377, + "loss": 3.3656, + "step": 28975 + }, + { + "epoch": 1.969017529555646, + "grad_norm": 0.7718935608863831, + "learning_rate": 0.0007538728088055442, + "loss": 3.3715, + "step": 28980 + }, + { + "epoch": 1.9693572496263079, + "grad_norm": 1.0160068273544312, + "learning_rate": 0.0007538303437967115, + "loss": 3.5351, + "step": 28985 + }, + { + "epoch": 1.9696969696969697, + "grad_norm": 0.8678659200668335, + "learning_rate": 0.0007537878787878788, + "loss": 3.583, + "step": 28990 + }, + { + "epoch": 1.9700366897676314, + "grad_norm": 0.7335505485534668, + "learning_rate": 0.0007537454137790461, + "loss": 3.343, + "step": 28995 + }, + { + "epoch": 1.9703764098382932, + "grad_norm": 0.7890894412994385, + "learning_rate": 0.0007537029487702134, + "loss": 3.7078, + "step": 29000 + }, + { + "epoch": 1.970716129908955, + "grad_norm": 0.7053489089012146, + "learning_rate": 0.0007536604837613806, + "loss": 3.5596, + "step": 29005 + }, + { + "epoch": 1.9710558499796167, + "grad_norm": 0.7064149379730225, + "learning_rate": 0.0007536180187525479, + "loss": 3.6475, + "step": 29010 + }, + { + "epoch": 1.9713955700502785, + "grad_norm": 0.8108725547790527, + "learning_rate": 0.0007535755537437152, + "loss": 3.6204, + "step": 29015 + }, + { + "epoch": 1.9717352901209404, + "grad_norm": 0.8527393341064453, + "learning_rate": 0.0007535330887348824, + "loss": 3.6671, + "step": 29020 + }, + { + "epoch": 1.972075010191602, + "grad_norm": 0.9500375390052795, + "learning_rate": 0.0007534906237260498, + "loss": 3.6597, + "step": 29025 + }, + { + "epoch": 1.9724147302622639, + "grad_norm": 1.1384083032608032, + "learning_rate": 0.0007534481587172171, + "loss": 3.87, + "step": 29030 + }, + { + "epoch": 1.9727544503329257, + "grad_norm": 1.4571928977966309, + "learning_rate": 0.0007534056937083843, + "loss": 3.5968, + "step": 29035 + }, + { + "epoch": 1.9730941704035874, + "grad_norm": 0.7645004987716675, + "learning_rate": 0.0007533632286995515, + "loss": 3.5517, + "step": 29040 + }, + { + "epoch": 1.9734338904742492, + "grad_norm": 0.9181371927261353, + "learning_rate": 0.0007533207636907189, + "loss": 3.5608, + "step": 29045 + }, + { + "epoch": 1.973773610544911, + "grad_norm": 0.9836548566818237, + "learning_rate": 0.0007532782986818861, + "loss": 3.6286, + "step": 29050 + }, + { + "epoch": 1.9741133306155727, + "grad_norm": 1.0330171585083008, + "learning_rate": 0.0007532358336730533, + "loss": 3.8315, + "step": 29055 + }, + { + "epoch": 1.9744530506862346, + "grad_norm": 0.9947535395622253, + "learning_rate": 0.0007531933686642208, + "loss": 3.5824, + "step": 29060 + }, + { + "epoch": 1.9747927707568964, + "grad_norm": 0.9524598717689514, + "learning_rate": 0.000753150903655388, + "loss": 3.938, + "step": 29065 + }, + { + "epoch": 1.975132490827558, + "grad_norm": 0.8610302805900574, + "learning_rate": 0.0007531084386465552, + "loss": 3.5288, + "step": 29070 + }, + { + "epoch": 1.97547221089822, + "grad_norm": 0.837296724319458, + "learning_rate": 0.0007530659736377226, + "loss": 3.4799, + "step": 29075 + }, + { + "epoch": 1.9758119309688817, + "grad_norm": 0.9392138123512268, + "learning_rate": 0.0007530235086288898, + "loss": 3.4918, + "step": 29080 + }, + { + "epoch": 1.9761516510395434, + "grad_norm": 0.8305526375770569, + "learning_rate": 0.000752981043620057, + "loss": 3.7236, + "step": 29085 + }, + { + "epoch": 1.9764913711102052, + "grad_norm": 0.9105304479598999, + "learning_rate": 0.0007529385786112243, + "loss": 3.6312, + "step": 29090 + }, + { + "epoch": 1.976831091180867, + "grad_norm": 0.752221405506134, + "learning_rate": 0.0007528961136023917, + "loss": 3.649, + "step": 29095 + }, + { + "epoch": 1.9771708112515287, + "grad_norm": 7.136263370513916, + "learning_rate": 0.0007528536485935589, + "loss": 3.3271, + "step": 29100 + }, + { + "epoch": 1.9775105313221906, + "grad_norm": 0.7253371477127075, + "learning_rate": 0.0007528111835847262, + "loss": 3.668, + "step": 29105 + }, + { + "epoch": 1.9778502513928524, + "grad_norm": 0.9408615827560425, + "learning_rate": 0.0007527687185758935, + "loss": 3.372, + "step": 29110 + }, + { + "epoch": 1.978189971463514, + "grad_norm": 0.7048061490058899, + "learning_rate": 0.0007527262535670607, + "loss": 3.5793, + "step": 29115 + }, + { + "epoch": 1.9785296915341757, + "grad_norm": 0.8520619869232178, + "learning_rate": 0.000752683788558228, + "loss": 3.7079, + "step": 29120 + }, + { + "epoch": 1.9788694116048378, + "grad_norm": 0.8509448766708374, + "learning_rate": 0.0007526413235493952, + "loss": 3.8453, + "step": 29125 + }, + { + "epoch": 1.9792091316754994, + "grad_norm": 0.9482970237731934, + "learning_rate": 0.0007525988585405627, + "loss": 3.7508, + "step": 29130 + }, + { + "epoch": 1.979548851746161, + "grad_norm": 0.7908791303634644, + "learning_rate": 0.0007525563935317299, + "loss": 3.6798, + "step": 29135 + }, + { + "epoch": 1.979888571816823, + "grad_norm": 1.5976383686065674, + "learning_rate": 0.0007525139285228971, + "loss": 3.4236, + "step": 29140 + }, + { + "epoch": 1.9802282918874847, + "grad_norm": 0.6679471135139465, + "learning_rate": 0.0007524714635140645, + "loss": 3.5702, + "step": 29145 + }, + { + "epoch": 1.9805680119581464, + "grad_norm": 0.8123327493667603, + "learning_rate": 0.0007524289985052317, + "loss": 3.6634, + "step": 29150 + }, + { + "epoch": 1.9809077320288082, + "grad_norm": 0.7589690685272217, + "learning_rate": 0.0007523865334963989, + "loss": 3.4967, + "step": 29155 + }, + { + "epoch": 1.98124745209947, + "grad_norm": 1.0659916400909424, + "learning_rate": 0.0007523440684875663, + "loss": 3.4568, + "step": 29160 + }, + { + "epoch": 1.9815871721701317, + "grad_norm": 0.970257580280304, + "learning_rate": 0.0007523016034787336, + "loss": 3.4835, + "step": 29165 + }, + { + "epoch": 1.9819268922407935, + "grad_norm": 0.9269804358482361, + "learning_rate": 0.0007522591384699008, + "loss": 3.7128, + "step": 29170 + }, + { + "epoch": 1.9822666123114554, + "grad_norm": 0.742489755153656, + "learning_rate": 0.0007522166734610682, + "loss": 3.8559, + "step": 29175 + }, + { + "epoch": 1.982606332382117, + "grad_norm": 0.7392188906669617, + "learning_rate": 0.0007521742084522354, + "loss": 3.4747, + "step": 29180 + }, + { + "epoch": 1.9829460524527789, + "grad_norm": 0.6585854887962341, + "learning_rate": 0.0007521317434434026, + "loss": 3.7298, + "step": 29185 + }, + { + "epoch": 1.9832857725234407, + "grad_norm": 0.923235297203064, + "learning_rate": 0.0007520892784345699, + "loss": 3.7449, + "step": 29190 + }, + { + "epoch": 1.9836254925941024, + "grad_norm": 0.8680022358894348, + "learning_rate": 0.0007520468134257372, + "loss": 3.7723, + "step": 29195 + }, + { + "epoch": 1.9839652126647642, + "grad_norm": 0.8059386610984802, + "learning_rate": 0.0007520043484169045, + "loss": 3.8393, + "step": 29200 + }, + { + "epoch": 1.984304932735426, + "grad_norm": 0.7884226441383362, + "learning_rate": 0.0007519618834080718, + "loss": 3.6193, + "step": 29205 + }, + { + "epoch": 1.9846446528060877, + "grad_norm": 0.7910537123680115, + "learning_rate": 0.0007519194183992391, + "loss": 3.3136, + "step": 29210 + }, + { + "epoch": 1.9849843728767496, + "grad_norm": 0.9270262718200684, + "learning_rate": 0.0007518769533904063, + "loss": 3.7627, + "step": 29215 + }, + { + "epoch": 1.9853240929474114, + "grad_norm": 1.0346249341964722, + "learning_rate": 0.0007518344883815736, + "loss": 3.4026, + "step": 29220 + }, + { + "epoch": 1.985663813018073, + "grad_norm": 0.8513607978820801, + "learning_rate": 0.0007517920233727409, + "loss": 3.7147, + "step": 29225 + }, + { + "epoch": 1.986003533088735, + "grad_norm": 0.8393589854240417, + "learning_rate": 0.0007517495583639081, + "loss": 3.4863, + "step": 29230 + }, + { + "epoch": 1.9863432531593967, + "grad_norm": 0.8679572343826294, + "learning_rate": 0.0007517070933550755, + "loss": 3.5135, + "step": 29235 + }, + { + "epoch": 1.9866829732300584, + "grad_norm": 1.0132571458816528, + "learning_rate": 0.0007516646283462427, + "loss": 3.3759, + "step": 29240 + }, + { + "epoch": 1.9870226933007202, + "grad_norm": 0.9760110378265381, + "learning_rate": 0.00075162216333741, + "loss": 3.7447, + "step": 29245 + }, + { + "epoch": 1.987362413371382, + "grad_norm": 0.8186237812042236, + "learning_rate": 0.0007515796983285773, + "loss": 3.7374, + "step": 29250 + }, + { + "epoch": 1.9877021334420437, + "grad_norm": 1.1092513799667358, + "learning_rate": 0.0007515372333197445, + "loss": 3.5809, + "step": 29255 + }, + { + "epoch": 1.9880418535127056, + "grad_norm": 1.261192798614502, + "learning_rate": 0.0007514947683109118, + "loss": 3.8282, + "step": 29260 + }, + { + "epoch": 1.9883815735833674, + "grad_norm": 0.9189577102661133, + "learning_rate": 0.0007514523033020791, + "loss": 3.4139, + "step": 29265 + }, + { + "epoch": 1.988721293654029, + "grad_norm": 0.8337008357048035, + "learning_rate": 0.0007514098382932464, + "loss": 3.5481, + "step": 29270 + }, + { + "epoch": 1.989061013724691, + "grad_norm": 1.409749150276184, + "learning_rate": 0.0007513673732844137, + "loss": 3.3206, + "step": 29275 + }, + { + "epoch": 1.9894007337953528, + "grad_norm": 0.6449939608573914, + "learning_rate": 0.000751324908275581, + "loss": 3.5363, + "step": 29280 + }, + { + "epoch": 1.9897404538660144, + "grad_norm": 1.1269621849060059, + "learning_rate": 0.0007512824432667482, + "loss": 3.5596, + "step": 29285 + }, + { + "epoch": 1.990080173936676, + "grad_norm": 0.7379196286201477, + "learning_rate": 0.0007512399782579154, + "loss": 3.491, + "step": 29290 + }, + { + "epoch": 1.990419894007338, + "grad_norm": 0.7300626635551453, + "learning_rate": 0.0007511975132490828, + "loss": 3.6164, + "step": 29295 + }, + { + "epoch": 1.9907596140779997, + "grad_norm": 0.7448827624320984, + "learning_rate": 0.00075115504824025, + "loss": 3.6993, + "step": 29300 + }, + { + "epoch": 1.9910993341486614, + "grad_norm": 0.8180156946182251, + "learning_rate": 0.0007511125832314173, + "loss": 3.5765, + "step": 29305 + }, + { + "epoch": 1.9914390542193234, + "grad_norm": 0.6944757103919983, + "learning_rate": 0.0007510701182225847, + "loss": 3.6794, + "step": 29310 + }, + { + "epoch": 1.991778774289985, + "grad_norm": 0.7398365139961243, + "learning_rate": 0.0007510276532137519, + "loss": 3.5284, + "step": 29315 + }, + { + "epoch": 1.9921184943606467, + "grad_norm": 2.0357625484466553, + "learning_rate": 0.0007509851882049191, + "loss": 3.543, + "step": 29320 + }, + { + "epoch": 1.9924582144313085, + "grad_norm": 0.934351921081543, + "learning_rate": 0.0007509427231960865, + "loss": 3.7063, + "step": 29325 + }, + { + "epoch": 1.9927979345019704, + "grad_norm": 0.7739645838737488, + "learning_rate": 0.0007509002581872537, + "loss": 3.4061, + "step": 29330 + }, + { + "epoch": 1.993137654572632, + "grad_norm": 0.8704223036766052, + "learning_rate": 0.0007508577931784209, + "loss": 3.539, + "step": 29335 + }, + { + "epoch": 1.9934773746432939, + "grad_norm": 0.8656898736953735, + "learning_rate": 0.0007508153281695883, + "loss": 3.7759, + "step": 29340 + }, + { + "epoch": 1.9938170947139557, + "grad_norm": 0.9884647727012634, + "learning_rate": 0.0007507728631607556, + "loss": 3.3741, + "step": 29345 + }, + { + "epoch": 1.9941568147846174, + "grad_norm": 0.8359668850898743, + "learning_rate": 0.0007507303981519228, + "loss": 3.8857, + "step": 29350 + }, + { + "epoch": 1.9944965348552792, + "grad_norm": 0.7502020001411438, + "learning_rate": 0.0007506879331430901, + "loss": 3.3919, + "step": 29355 + }, + { + "epoch": 1.994836254925941, + "grad_norm": 1.1888049840927124, + "learning_rate": 0.0007506454681342574, + "loss": 3.7313, + "step": 29360 + }, + { + "epoch": 1.9951759749966027, + "grad_norm": 0.8564573526382446, + "learning_rate": 0.0007506030031254246, + "loss": 3.6416, + "step": 29365 + }, + { + "epoch": 1.9955156950672646, + "grad_norm": 0.8001554608345032, + "learning_rate": 0.000750560538116592, + "loss": 3.6559, + "step": 29370 + }, + { + "epoch": 1.9958554151379264, + "grad_norm": 0.8352419137954712, + "learning_rate": 0.0007505180731077593, + "loss": 3.7106, + "step": 29375 + }, + { + "epoch": 1.996195135208588, + "grad_norm": 0.6696701645851135, + "learning_rate": 0.0007504756080989265, + "loss": 3.8092, + "step": 29380 + }, + { + "epoch": 1.99653485527925, + "grad_norm": 0.9772933125495911, + "learning_rate": 0.0007504331430900938, + "loss": 3.6293, + "step": 29385 + }, + { + "epoch": 1.9968745753499118, + "grad_norm": 0.8547801375389099, + "learning_rate": 0.000750390678081261, + "loss": 3.4399, + "step": 29390 + }, + { + "epoch": 1.9972142954205734, + "grad_norm": 0.6507989168167114, + "learning_rate": 0.0007503482130724283, + "loss": 3.7233, + "step": 29395 + }, + { + "epoch": 1.9975540154912352, + "grad_norm": 0.8760092854499817, + "learning_rate": 0.0007503057480635956, + "loss": 3.7968, + "step": 29400 + }, + { + "epoch": 1.997893735561897, + "grad_norm": 0.8082702159881592, + "learning_rate": 0.0007502632830547629, + "loss": 3.5677, + "step": 29405 + }, + { + "epoch": 1.9982334556325587, + "grad_norm": 1.0632838010787964, + "learning_rate": 0.0007502208180459302, + "loss": 3.6244, + "step": 29410 + }, + { + "epoch": 1.9985731757032206, + "grad_norm": 1.1742656230926514, + "learning_rate": 0.0007501783530370975, + "loss": 3.5269, + "step": 29415 + }, + { + "epoch": 1.9989128957738824, + "grad_norm": 0.952635645866394, + "learning_rate": 0.0007501358880282647, + "loss": 3.67, + "step": 29420 + }, + { + "epoch": 1.999252615844544, + "grad_norm": 1.0246108770370483, + "learning_rate": 0.0007500934230194319, + "loss": 3.6288, + "step": 29425 + }, + { + "epoch": 1.999592335915206, + "grad_norm": 0.7565904855728149, + "learning_rate": 0.0007500509580105993, + "loss": 3.5684, + "step": 29430 + }, + { + "epoch": 1.9999320559858678, + "grad_norm": 0.7612989544868469, + "learning_rate": 0.0007500084930017665, + "loss": 3.7262, + "step": 29435 + }, + { + "epoch": 2.0, + "eval_bertscore": { + "f1": 0.8401453708687405, + "precision": 0.84065271906063, + "recall": 0.8405500507271559 + }, + "eval_bleu_4": 0.018990813977970547, + "eval_exact_match": 0.00019381723035177828, + "eval_loss": 3.4735400676727295, + "eval_meteor": 0.09309370200781929, + "eval_rouge": { + "rouge1": 0.12514640946339917, + "rouge2": 0.01804564127942626, + "rougeL": 0.10785387297955415, + "rougeLsum": 0.10789256108330744 + }, + "eval_runtime": 3205.431, + "eval_samples_per_second": 3.219, + "eval_steps_per_second": 0.402, + "step": 29436 + }, + { + "epoch": 2.0002717760565294, + "grad_norm": 0.907905638217926, + "learning_rate": 0.0007499660279929338, + "loss": 3.5941, + "step": 29440 + }, + { + "epoch": 2.000611496127191, + "grad_norm": 0.6962242722511292, + "learning_rate": 0.0007499235629841012, + "loss": 3.5878, + "step": 29445 + }, + { + "epoch": 2.000951216197853, + "grad_norm": 0.737462043762207, + "learning_rate": 0.0007498810979752684, + "loss": 3.5378, + "step": 29450 + }, + { + "epoch": 2.0012909362685147, + "grad_norm": 0.798230767250061, + "learning_rate": 0.0007498386329664356, + "loss": 3.5264, + "step": 29455 + }, + { + "epoch": 2.0016306563391764, + "grad_norm": 1.5187355279922485, + "learning_rate": 0.000749796167957603, + "loss": 3.4892, + "step": 29460 + }, + { + "epoch": 2.0019703764098384, + "grad_norm": 0.9855580925941467, + "learning_rate": 0.0007497537029487702, + "loss": 3.3585, + "step": 29465 + }, + { + "epoch": 2.0023100964805, + "grad_norm": 1.1158792972564697, + "learning_rate": 0.0007497112379399374, + "loss": 3.37, + "step": 29470 + }, + { + "epoch": 2.0026498165511617, + "grad_norm": 1.0571216344833374, + "learning_rate": 0.0007496687729311049, + "loss": 3.4063, + "step": 29475 + }, + { + "epoch": 2.0029895366218238, + "grad_norm": 1.1315616369247437, + "learning_rate": 0.0007496263079222721, + "loss": 3.4602, + "step": 29480 + }, + { + "epoch": 2.0033292566924854, + "grad_norm": 0.920049250125885, + "learning_rate": 0.0007495838429134394, + "loss": 3.3883, + "step": 29485 + }, + { + "epoch": 2.003668976763147, + "grad_norm": 0.8580490350723267, + "learning_rate": 0.0007495413779046066, + "loss": 3.6258, + "step": 29490 + }, + { + "epoch": 2.004008696833809, + "grad_norm": 0.6755834221839905, + "learning_rate": 0.0007494989128957739, + "loss": 3.7057, + "step": 29495 + }, + { + "epoch": 2.0043484169044707, + "grad_norm": 0.7505844831466675, + "learning_rate": 0.0007494564478869412, + "loss": 3.1538, + "step": 29500 + }, + { + "epoch": 2.0046881369751324, + "grad_norm": 2.153061628341675, + "learning_rate": 0.0007494139828781084, + "loss": 3.5878, + "step": 29505 + }, + { + "epoch": 2.0050278570457944, + "grad_norm": 0.9868181347846985, + "learning_rate": 0.0007493715178692758, + "loss": 3.8243, + "step": 29510 + }, + { + "epoch": 2.005367577116456, + "grad_norm": 0.8526576161384583, + "learning_rate": 0.0007493290528604431, + "loss": 3.5244, + "step": 29515 + }, + { + "epoch": 2.0057072971871177, + "grad_norm": 1.3826223611831665, + "learning_rate": 0.0007492865878516103, + "loss": 3.3284, + "step": 29520 + }, + { + "epoch": 2.00604701725778, + "grad_norm": 1.0314799547195435, + "learning_rate": 0.0007492441228427775, + "loss": 3.6559, + "step": 29525 + }, + { + "epoch": 2.0063867373284414, + "grad_norm": 0.8046008944511414, + "learning_rate": 0.0007492016578339449, + "loss": 3.6961, + "step": 29530 + }, + { + "epoch": 2.006726457399103, + "grad_norm": 1.0257865190505981, + "learning_rate": 0.0007491591928251121, + "loss": 3.7188, + "step": 29535 + }, + { + "epoch": 2.007066177469765, + "grad_norm": 0.7950069904327393, + "learning_rate": 0.0007491167278162793, + "loss": 3.6145, + "step": 29540 + }, + { + "epoch": 2.0074058975404268, + "grad_norm": 0.964979887008667, + "learning_rate": 0.0007490742628074468, + "loss": 3.8685, + "step": 29545 + }, + { + "epoch": 2.0077456176110884, + "grad_norm": 0.8541505932807922, + "learning_rate": 0.000749031797798614, + "loss": 3.1539, + "step": 29550 + }, + { + "epoch": 2.0080853376817505, + "grad_norm": 0.9248681664466858, + "learning_rate": 0.0007489893327897812, + "loss": 3.5821, + "step": 29555 + }, + { + "epoch": 2.008425057752412, + "grad_norm": 0.8942385911941528, + "learning_rate": 0.0007489468677809486, + "loss": 3.3341, + "step": 29560 + }, + { + "epoch": 2.0087647778230737, + "grad_norm": 1.231141448020935, + "learning_rate": 0.0007489044027721158, + "loss": 3.4626, + "step": 29565 + }, + { + "epoch": 2.0091044978937354, + "grad_norm": 0.8439732789993286, + "learning_rate": 0.000748861937763283, + "loss": 3.6587, + "step": 29570 + }, + { + "epoch": 2.0094442179643974, + "grad_norm": 0.7009230256080627, + "learning_rate": 0.0007488194727544503, + "loss": 3.551, + "step": 29575 + }, + { + "epoch": 2.009783938035059, + "grad_norm": 0.7958117127418518, + "learning_rate": 0.0007487770077456177, + "loss": 3.662, + "step": 29580 + }, + { + "epoch": 2.0101236581057207, + "grad_norm": 0.6516826152801514, + "learning_rate": 0.0007487345427367849, + "loss": 3.5244, + "step": 29585 + }, + { + "epoch": 2.0104633781763828, + "grad_norm": 0.9384609460830688, + "learning_rate": 0.0007486920777279522, + "loss": 3.5378, + "step": 29590 + }, + { + "epoch": 2.0108030982470444, + "grad_norm": 1.3192696571350098, + "learning_rate": 0.0007486496127191195, + "loss": 3.5726, + "step": 29595 + }, + { + "epoch": 2.011142818317706, + "grad_norm": 0.8813995122909546, + "learning_rate": 0.0007486071477102867, + "loss": 3.396, + "step": 29600 + }, + { + "epoch": 2.011482538388368, + "grad_norm": 0.9518397450447083, + "learning_rate": 0.000748564682701454, + "loss": 3.393, + "step": 29605 + }, + { + "epoch": 2.0118222584590297, + "grad_norm": 0.8820313811302185, + "learning_rate": 0.0007485222176926213, + "loss": 3.5008, + "step": 29610 + }, + { + "epoch": 2.0121619785296914, + "grad_norm": 0.9841651916503906, + "learning_rate": 0.0007484797526837886, + "loss": 3.6952, + "step": 29615 + }, + { + "epoch": 2.0125016986003534, + "grad_norm": 1.007918357849121, + "learning_rate": 0.0007484372876749559, + "loss": 3.6085, + "step": 29620 + }, + { + "epoch": 2.012841418671015, + "grad_norm": 0.9101667404174805, + "learning_rate": 0.0007483948226661231, + "loss": 3.4666, + "step": 29625 + }, + { + "epoch": 2.0131811387416767, + "grad_norm": 0.9427982568740845, + "learning_rate": 0.0007483523576572904, + "loss": 3.677, + "step": 29630 + }, + { + "epoch": 2.0135208588123388, + "grad_norm": 0.7218946218490601, + "learning_rate": 0.0007483098926484577, + "loss": 3.6277, + "step": 29635 + }, + { + "epoch": 2.0138605788830004, + "grad_norm": 0.8679907917976379, + "learning_rate": 0.0007482674276396249, + "loss": 3.3277, + "step": 29640 + }, + { + "epoch": 2.014200298953662, + "grad_norm": 0.8475306034088135, + "learning_rate": 0.0007482249626307922, + "loss": 3.4709, + "step": 29645 + }, + { + "epoch": 2.014540019024324, + "grad_norm": 0.7366526126861572, + "learning_rate": 0.0007481824976219596, + "loss": 3.5437, + "step": 29650 + }, + { + "epoch": 2.0148797390949857, + "grad_norm": 1.0132291316986084, + "learning_rate": 0.0007481400326131268, + "loss": 3.8658, + "step": 29655 + }, + { + "epoch": 2.0152194591656474, + "grad_norm": 0.7501122355461121, + "learning_rate": 0.0007480975676042941, + "loss": 3.7098, + "step": 29660 + }, + { + "epoch": 2.0155591792363095, + "grad_norm": 0.8191317319869995, + "learning_rate": 0.0007480551025954614, + "loss": 3.6435, + "step": 29665 + }, + { + "epoch": 2.015898899306971, + "grad_norm": 1.051823377609253, + "learning_rate": 0.0007480126375866286, + "loss": 3.4452, + "step": 29670 + }, + { + "epoch": 2.0162386193776327, + "grad_norm": 0.9380770921707153, + "learning_rate": 0.0007479701725777958, + "loss": 3.1687, + "step": 29675 + }, + { + "epoch": 2.016578339448295, + "grad_norm": 0.9178208708763123, + "learning_rate": 0.0007479277075689632, + "loss": 3.8156, + "step": 29680 + }, + { + "epoch": 2.0169180595189564, + "grad_norm": 1.190659999847412, + "learning_rate": 0.0007478852425601305, + "loss": 3.4968, + "step": 29685 + }, + { + "epoch": 2.017257779589618, + "grad_norm": 0.7861708998680115, + "learning_rate": 0.0007478427775512977, + "loss": 3.7399, + "step": 29690 + }, + { + "epoch": 2.01759749966028, + "grad_norm": 0.6784389615058899, + "learning_rate": 0.0007478003125424651, + "loss": 3.7482, + "step": 29695 + }, + { + "epoch": 2.0179372197309418, + "grad_norm": 0.993552029132843, + "learning_rate": 0.0007477578475336323, + "loss": 3.4869, + "step": 29700 + }, + { + "epoch": 2.0182769398016034, + "grad_norm": 0.7340433597564697, + "learning_rate": 0.0007477153825247995, + "loss": 3.4401, + "step": 29705 + }, + { + "epoch": 2.0186166598722655, + "grad_norm": 0.7984585165977478, + "learning_rate": 0.0007476729175159669, + "loss": 3.7574, + "step": 29710 + }, + { + "epoch": 2.018956379942927, + "grad_norm": 0.9595564603805542, + "learning_rate": 0.0007476304525071341, + "loss": 3.5996, + "step": 29715 + }, + { + "epoch": 2.0192961000135887, + "grad_norm": 0.8063091039657593, + "learning_rate": 0.0007475879874983014, + "loss": 3.378, + "step": 29720 + }, + { + "epoch": 2.0196358200842504, + "grad_norm": 0.8322507739067078, + "learning_rate": 0.0007475455224894687, + "loss": 3.5582, + "step": 29725 + }, + { + "epoch": 2.0199755401549124, + "grad_norm": 0.9427157640457153, + "learning_rate": 0.000747503057480636, + "loss": 3.2115, + "step": 29730 + }, + { + "epoch": 2.020315260225574, + "grad_norm": 1.0460054874420166, + "learning_rate": 0.0007474605924718032, + "loss": 3.3804, + "step": 29735 + }, + { + "epoch": 2.0206549802962357, + "grad_norm": 0.8137238621711731, + "learning_rate": 0.0007474181274629705, + "loss": 3.4971, + "step": 29740 + }, + { + "epoch": 2.0209947003668978, + "grad_norm": 1.254673719406128, + "learning_rate": 0.0007473756624541378, + "loss": 3.3985, + "step": 29745 + }, + { + "epoch": 2.0213344204375594, + "grad_norm": 0.778195858001709, + "learning_rate": 0.000747333197445305, + "loss": 3.4578, + "step": 29750 + }, + { + "epoch": 2.021674140508221, + "grad_norm": 0.7506092190742493, + "learning_rate": 0.0007472907324364724, + "loss": 3.6048, + "step": 29755 + }, + { + "epoch": 2.022013860578883, + "grad_norm": 0.7908632755279541, + "learning_rate": 0.0007472482674276397, + "loss": 3.701, + "step": 29760 + }, + { + "epoch": 2.0223535806495447, + "grad_norm": 0.9338049292564392, + "learning_rate": 0.0007472058024188069, + "loss": 3.5614, + "step": 29765 + }, + { + "epoch": 2.0226933007202064, + "grad_norm": 0.959753155708313, + "learning_rate": 0.0007471633374099742, + "loss": 3.4331, + "step": 29770 + }, + { + "epoch": 2.0230330207908684, + "grad_norm": 0.8978981971740723, + "learning_rate": 0.0007471208724011414, + "loss": 3.4794, + "step": 29775 + }, + { + "epoch": 2.02337274086153, + "grad_norm": 0.9128916263580322, + "learning_rate": 0.0007470784073923087, + "loss": 3.5545, + "step": 29780 + }, + { + "epoch": 2.0237124609321917, + "grad_norm": 0.8868361115455627, + "learning_rate": 0.000747035942383476, + "loss": 3.5537, + "step": 29785 + }, + { + "epoch": 2.024052181002854, + "grad_norm": 0.7643030881881714, + "learning_rate": 0.0007469934773746433, + "loss": 3.5983, + "step": 29790 + }, + { + "epoch": 2.0243919010735154, + "grad_norm": 0.8862969875335693, + "learning_rate": 0.0007469510123658106, + "loss": 3.5653, + "step": 29795 + }, + { + "epoch": 2.024731621144177, + "grad_norm": 0.6573250889778137, + "learning_rate": 0.0007469085473569779, + "loss": 3.912, + "step": 29800 + }, + { + "epoch": 2.025071341214839, + "grad_norm": 0.9341059923171997, + "learning_rate": 0.0007468660823481451, + "loss": 3.6593, + "step": 29805 + }, + { + "epoch": 2.0254110612855007, + "grad_norm": 3.6508257389068604, + "learning_rate": 0.0007468236173393123, + "loss": 3.6188, + "step": 29810 + }, + { + "epoch": 2.0257507813561624, + "grad_norm": 0.9624465107917786, + "learning_rate": 0.0007467811523304797, + "loss": 3.523, + "step": 29815 + }, + { + "epoch": 2.0260905014268245, + "grad_norm": 0.7539438605308533, + "learning_rate": 0.0007467386873216469, + "loss": 3.527, + "step": 29820 + }, + { + "epoch": 2.026430221497486, + "grad_norm": 0.9194149971008301, + "learning_rate": 0.0007466962223128144, + "loss": 3.5303, + "step": 29825 + }, + { + "epoch": 2.0267699415681477, + "grad_norm": 0.8312097191810608, + "learning_rate": 0.0007466537573039816, + "loss": 3.384, + "step": 29830 + }, + { + "epoch": 2.02710966163881, + "grad_norm": 0.8856743574142456, + "learning_rate": 0.0007466112922951488, + "loss": 3.36, + "step": 29835 + }, + { + "epoch": 2.0274493817094714, + "grad_norm": 1.442902684211731, + "learning_rate": 0.0007465688272863161, + "loss": 3.7349, + "step": 29840 + }, + { + "epoch": 2.027789101780133, + "grad_norm": 0.8912723660469055, + "learning_rate": 0.0007465263622774834, + "loss": 3.3736, + "step": 29845 + }, + { + "epoch": 2.028128821850795, + "grad_norm": 0.8849855065345764, + "learning_rate": 0.0007464838972686506, + "loss": 3.5301, + "step": 29850 + }, + { + "epoch": 2.0284685419214568, + "grad_norm": 0.9413118958473206, + "learning_rate": 0.0007464414322598179, + "loss": 3.7613, + "step": 29855 + }, + { + "epoch": 2.0288082619921184, + "grad_norm": 0.6635482311248779, + "learning_rate": 0.0007463989672509853, + "loss": 3.5251, + "step": 29860 + }, + { + "epoch": 2.0291479820627805, + "grad_norm": 1.132591962814331, + "learning_rate": 0.0007463565022421525, + "loss": 3.7151, + "step": 29865 + }, + { + "epoch": 2.029487702133442, + "grad_norm": 0.9874240756034851, + "learning_rate": 0.0007463140372333198, + "loss": 3.6424, + "step": 29870 + }, + { + "epoch": 2.0298274222041037, + "grad_norm": 0.818653404712677, + "learning_rate": 0.000746271572224487, + "loss": 3.719, + "step": 29875 + }, + { + "epoch": 2.030167142274766, + "grad_norm": 1.0717358589172363, + "learning_rate": 0.0007462291072156543, + "loss": 3.648, + "step": 29880 + }, + { + "epoch": 2.0305068623454274, + "grad_norm": 0.9440126419067383, + "learning_rate": 0.0007461866422068216, + "loss": 3.5654, + "step": 29885 + }, + { + "epoch": 2.030846582416089, + "grad_norm": 0.8054519891738892, + "learning_rate": 0.0007461441771979888, + "loss": 3.5355, + "step": 29890 + }, + { + "epoch": 2.031186302486751, + "grad_norm": 1.1522808074951172, + "learning_rate": 0.0007461017121891562, + "loss": 3.2367, + "step": 29895 + }, + { + "epoch": 2.0315260225574128, + "grad_norm": 1.3224278688430786, + "learning_rate": 0.0007460592471803235, + "loss": 3.2162, + "step": 29900 + }, + { + "epoch": 2.0318657426280744, + "grad_norm": 0.816796064376831, + "learning_rate": 0.0007460167821714907, + "loss": 3.8512, + "step": 29905 + }, + { + "epoch": 2.032205462698736, + "grad_norm": 0.7117137908935547, + "learning_rate": 0.000745974317162658, + "loss": 3.7982, + "step": 29910 + }, + { + "epoch": 2.032545182769398, + "grad_norm": 0.8739092350006104, + "learning_rate": 0.0007459318521538253, + "loss": 3.4923, + "step": 29915 + }, + { + "epoch": 2.0328849028400597, + "grad_norm": 0.8217427730560303, + "learning_rate": 0.0007458893871449925, + "loss": 3.4505, + "step": 29920 + }, + { + "epoch": 2.0332246229107214, + "grad_norm": 0.9769152998924255, + "learning_rate": 0.0007458469221361597, + "loss": 3.6454, + "step": 29925 + }, + { + "epoch": 2.0335643429813834, + "grad_norm": 0.7304091453552246, + "learning_rate": 0.0007458044571273272, + "loss": 3.6234, + "step": 29930 + }, + { + "epoch": 2.033904063052045, + "grad_norm": 1.0225117206573486, + "learning_rate": 0.0007457619921184944, + "loss": 3.4735, + "step": 29935 + }, + { + "epoch": 2.0342437831227067, + "grad_norm": 0.8672761917114258, + "learning_rate": 0.0007457195271096616, + "loss": 3.8095, + "step": 29940 + }, + { + "epoch": 2.034583503193369, + "grad_norm": 0.997830867767334, + "learning_rate": 0.000745677062100829, + "loss": 3.5252, + "step": 29945 + }, + { + "epoch": 2.0349232232640304, + "grad_norm": 0.6500917077064514, + "learning_rate": 0.0007456345970919962, + "loss": 3.4479, + "step": 29950 + }, + { + "epoch": 2.035262943334692, + "grad_norm": 0.9575283527374268, + "learning_rate": 0.0007455921320831634, + "loss": 3.5827, + "step": 29955 + }, + { + "epoch": 2.035602663405354, + "grad_norm": 0.7531545758247375, + "learning_rate": 0.0007455496670743309, + "loss": 3.5396, + "step": 29960 + }, + { + "epoch": 2.0359423834760157, + "grad_norm": 2.2459850311279297, + "learning_rate": 0.0007455072020654981, + "loss": 3.3888, + "step": 29965 + }, + { + "epoch": 2.0362821035466774, + "grad_norm": 1.1024680137634277, + "learning_rate": 0.0007454647370566653, + "loss": 3.6355, + "step": 29970 + }, + { + "epoch": 2.0366218236173395, + "grad_norm": 0.862695038318634, + "learning_rate": 0.0007454222720478326, + "loss": 3.7001, + "step": 29975 + }, + { + "epoch": 2.036961543688001, + "grad_norm": 0.9912177920341492, + "learning_rate": 0.0007453798070389999, + "loss": 3.341, + "step": 29980 + }, + { + "epoch": 2.0373012637586627, + "grad_norm": 0.8192928433418274, + "learning_rate": 0.0007453373420301671, + "loss": 3.5358, + "step": 29985 + }, + { + "epoch": 2.037640983829325, + "grad_norm": 0.8122713565826416, + "learning_rate": 0.0007452948770213344, + "loss": 3.7265, + "step": 29990 + }, + { + "epoch": 2.0379807038999864, + "grad_norm": 1.0422576665878296, + "learning_rate": 0.0007452524120125018, + "loss": 3.7656, + "step": 29995 + }, + { + "epoch": 2.038320423970648, + "grad_norm": 0.7285352349281311, + "learning_rate": 0.000745209947003669, + "loss": 3.726, + "step": 30000 + }, + { + "epoch": 2.03866014404131, + "grad_norm": 0.8428540825843811, + "learning_rate": 0.0007451674819948363, + "loss": 3.6522, + "step": 30005 + }, + { + "epoch": 2.0389998641119718, + "grad_norm": 0.8177939653396606, + "learning_rate": 0.0007451250169860036, + "loss": 3.8713, + "step": 30010 + }, + { + "epoch": 2.0393395841826334, + "grad_norm": 0.8508073687553406, + "learning_rate": 0.0007450825519771708, + "loss": 3.8155, + "step": 30015 + }, + { + "epoch": 2.0396793042532955, + "grad_norm": 0.7089604735374451, + "learning_rate": 0.0007450400869683381, + "loss": 3.6199, + "step": 30020 + }, + { + "epoch": 2.040019024323957, + "grad_norm": 0.8525442481040955, + "learning_rate": 0.0007449976219595053, + "loss": 3.5146, + "step": 30025 + }, + { + "epoch": 2.0403587443946187, + "grad_norm": 0.8071790933609009, + "learning_rate": 0.0007449551569506727, + "loss": 3.648, + "step": 30030 + }, + { + "epoch": 2.040698464465281, + "grad_norm": 0.9341453313827515, + "learning_rate": 0.00074491269194184, + "loss": 3.5942, + "step": 30035 + }, + { + "epoch": 2.0410381845359424, + "grad_norm": 1.0676578283309937, + "learning_rate": 0.0007448702269330072, + "loss": 3.5487, + "step": 30040 + }, + { + "epoch": 2.041377904606604, + "grad_norm": 0.7942273616790771, + "learning_rate": 0.0007448277619241745, + "loss": 3.4798, + "step": 30045 + }, + { + "epoch": 2.041717624677266, + "grad_norm": 1.496077060699463, + "learning_rate": 0.0007447852969153418, + "loss": 3.561, + "step": 30050 + }, + { + "epoch": 2.0420573447479278, + "grad_norm": 0.8735402226448059, + "learning_rate": 0.000744742831906509, + "loss": 3.3278, + "step": 30055 + }, + { + "epoch": 2.0423970648185894, + "grad_norm": 0.9960219264030457, + "learning_rate": 0.0007447003668976762, + "loss": 3.5946, + "step": 30060 + }, + { + "epoch": 2.042736784889251, + "grad_norm": 0.8716889023780823, + "learning_rate": 0.0007446579018888437, + "loss": 3.5035, + "step": 30065 + }, + { + "epoch": 2.043076504959913, + "grad_norm": 0.8317398428916931, + "learning_rate": 0.0007446154368800109, + "loss": 3.8043, + "step": 30070 + }, + { + "epoch": 2.0434162250305747, + "grad_norm": 0.8044674396514893, + "learning_rate": 0.0007445729718711781, + "loss": 3.2921, + "step": 30075 + }, + { + "epoch": 2.0437559451012364, + "grad_norm": 0.783662736415863, + "learning_rate": 0.0007445305068623455, + "loss": 3.4793, + "step": 30080 + }, + { + "epoch": 2.0440956651718984, + "grad_norm": 0.7997136116027832, + "learning_rate": 0.0007444880418535127, + "loss": 3.6485, + "step": 30085 + }, + { + "epoch": 2.04443538524256, + "grad_norm": 0.6637369990348816, + "learning_rate": 0.0007444455768446799, + "loss": 3.6048, + "step": 30090 + }, + { + "epoch": 2.0447751053132217, + "grad_norm": 0.7963462471961975, + "learning_rate": 0.0007444031118358473, + "loss": 3.4203, + "step": 30095 + }, + { + "epoch": 2.045114825383884, + "grad_norm": 1.0084829330444336, + "learning_rate": 0.0007443606468270146, + "loss": 3.4502, + "step": 30100 + }, + { + "epoch": 2.0454545454545454, + "grad_norm": 0.8999413847923279, + "learning_rate": 0.0007443181818181818, + "loss": 3.6533, + "step": 30105 + }, + { + "epoch": 2.045794265525207, + "grad_norm": 0.9683690071105957, + "learning_rate": 0.0007442757168093492, + "loss": 3.5035, + "step": 30110 + }, + { + "epoch": 2.046133985595869, + "grad_norm": 0.7637821435928345, + "learning_rate": 0.0007442332518005164, + "loss": 3.7422, + "step": 30115 + }, + { + "epoch": 2.0464737056665308, + "grad_norm": 0.7321568727493286, + "learning_rate": 0.0007441907867916836, + "loss": 3.5244, + "step": 30120 + }, + { + "epoch": 2.0468134257371924, + "grad_norm": 0.8551080226898193, + "learning_rate": 0.0007441483217828509, + "loss": 3.5163, + "step": 30125 + }, + { + "epoch": 2.0471531458078545, + "grad_norm": 0.7921773195266724, + "learning_rate": 0.0007441058567740182, + "loss": 3.4562, + "step": 30130 + }, + { + "epoch": 2.047492865878516, + "grad_norm": 0.7873597145080566, + "learning_rate": 0.0007440633917651855, + "loss": 3.5316, + "step": 30135 + }, + { + "epoch": 2.0478325859491777, + "grad_norm": 0.7279452681541443, + "learning_rate": 0.0007440209267563528, + "loss": 3.747, + "step": 30140 + }, + { + "epoch": 2.04817230601984, + "grad_norm": 1.2161774635314941, + "learning_rate": 0.0007439784617475201, + "loss": 3.6942, + "step": 30145 + }, + { + "epoch": 2.0485120260905014, + "grad_norm": 0.9279671907424927, + "learning_rate": 0.0007439359967386873, + "loss": 3.3763, + "step": 30150 + }, + { + "epoch": 2.048851746161163, + "grad_norm": 0.81589275598526, + "learning_rate": 0.0007438935317298546, + "loss": 3.3799, + "step": 30155 + }, + { + "epoch": 2.049191466231825, + "grad_norm": 0.759606122970581, + "learning_rate": 0.0007438510667210218, + "loss": 3.6719, + "step": 30160 + }, + { + "epoch": 2.0495311863024868, + "grad_norm": 0.7813960313796997, + "learning_rate": 0.0007438086017121892, + "loss": 3.501, + "step": 30165 + }, + { + "epoch": 2.0498709063731484, + "grad_norm": 0.862910807132721, + "learning_rate": 0.0007437661367033565, + "loss": 3.5236, + "step": 30170 + }, + { + "epoch": 2.0502106264438105, + "grad_norm": 0.9117962718009949, + "learning_rate": 0.0007437236716945237, + "loss": 3.4349, + "step": 30175 + }, + { + "epoch": 2.050550346514472, + "grad_norm": 0.762190043926239, + "learning_rate": 0.0007436812066856911, + "loss": 3.7909, + "step": 30180 + }, + { + "epoch": 2.0508900665851337, + "grad_norm": 1.0550425052642822, + "learning_rate": 0.0007436387416768583, + "loss": 3.3549, + "step": 30185 + }, + { + "epoch": 2.051229786655796, + "grad_norm": 0.7871007919311523, + "learning_rate": 0.0007435962766680255, + "loss": 3.3736, + "step": 30190 + }, + { + "epoch": 2.0515695067264574, + "grad_norm": 0.9111835360527039, + "learning_rate": 0.0007435538116591929, + "loss": 3.3158, + "step": 30195 + }, + { + "epoch": 2.051909226797119, + "grad_norm": 0.8797037601470947, + "learning_rate": 0.0007435113466503601, + "loss": 3.8605, + "step": 30200 + }, + { + "epoch": 2.052248946867781, + "grad_norm": 0.648908793926239, + "learning_rate": 0.0007434688816415274, + "loss": 3.5761, + "step": 30205 + }, + { + "epoch": 2.0525886669384428, + "grad_norm": 1.0742509365081787, + "learning_rate": 0.0007434264166326948, + "loss": 3.4852, + "step": 30210 + }, + { + "epoch": 2.0529283870091044, + "grad_norm": 0.9244308471679688, + "learning_rate": 0.000743383951623862, + "loss": 3.583, + "step": 30215 + }, + { + "epoch": 2.0532681070797665, + "grad_norm": 0.780324399471283, + "learning_rate": 0.0007433414866150292, + "loss": 3.7677, + "step": 30220 + }, + { + "epoch": 2.053607827150428, + "grad_norm": 1.364317536354065, + "learning_rate": 0.0007432990216061965, + "loss": 3.4176, + "step": 30225 + }, + { + "epoch": 2.0539475472210897, + "grad_norm": 1.0801634788513184, + "learning_rate": 0.0007432565565973638, + "loss": 3.315, + "step": 30230 + }, + { + "epoch": 2.054287267291752, + "grad_norm": 0.7172245979309082, + "learning_rate": 0.000743214091588531, + "loss": 3.5615, + "step": 30235 + }, + { + "epoch": 2.0546269873624134, + "grad_norm": 0.7711317539215088, + "learning_rate": 0.0007431716265796984, + "loss": 3.5654, + "step": 30240 + }, + { + "epoch": 2.054966707433075, + "grad_norm": 1.7499786615371704, + "learning_rate": 0.0007431291615708657, + "loss": 3.5924, + "step": 30245 + }, + { + "epoch": 2.0553064275037367, + "grad_norm": 0.8764796257019043, + "learning_rate": 0.0007430866965620329, + "loss": 3.4943, + "step": 30250 + }, + { + "epoch": 2.055646147574399, + "grad_norm": 0.8986364006996155, + "learning_rate": 0.0007430442315532002, + "loss": 3.779, + "step": 30255 + }, + { + "epoch": 2.0559858676450604, + "grad_norm": 0.7314947843551636, + "learning_rate": 0.0007430017665443674, + "loss": 3.7973, + "step": 30260 + }, + { + "epoch": 2.056325587715722, + "grad_norm": 0.882265031337738, + "learning_rate": 0.0007429593015355347, + "loss": 3.6973, + "step": 30265 + }, + { + "epoch": 2.056665307786384, + "grad_norm": 0.7728033065795898, + "learning_rate": 0.000742916836526702, + "loss": 3.3723, + "step": 30270 + }, + { + "epoch": 2.0570050278570458, + "grad_norm": 0.8112080097198486, + "learning_rate": 0.0007428743715178693, + "loss": 3.6802, + "step": 30275 + }, + { + "epoch": 2.0573447479277074, + "grad_norm": 0.797694742679596, + "learning_rate": 0.0007428319065090366, + "loss": 3.2253, + "step": 30280 + }, + { + "epoch": 2.0576844679983695, + "grad_norm": 0.8730963468551636, + "learning_rate": 0.0007427894415002039, + "loss": 3.5329, + "step": 30285 + }, + { + "epoch": 2.058024188069031, + "grad_norm": 0.7174362540245056, + "learning_rate": 0.0007427469764913711, + "loss": 3.6808, + "step": 30290 + }, + { + "epoch": 2.0583639081396927, + "grad_norm": 1.0370514392852783, + "learning_rate": 0.0007427045114825384, + "loss": 3.4656, + "step": 30295 + }, + { + "epoch": 2.058703628210355, + "grad_norm": 1.2400455474853516, + "learning_rate": 0.0007426620464737057, + "loss": 3.659, + "step": 30300 + }, + { + "epoch": 2.0590433482810164, + "grad_norm": 0.905286431312561, + "learning_rate": 0.0007426195814648729, + "loss": 3.6574, + "step": 30305 + }, + { + "epoch": 2.059383068351678, + "grad_norm": 0.9873246550559998, + "learning_rate": 0.0007425771164560402, + "loss": 3.5455, + "step": 30310 + }, + { + "epoch": 2.05972278842234, + "grad_norm": 0.75831139087677, + "learning_rate": 0.0007425346514472076, + "loss": 3.5152, + "step": 30315 + }, + { + "epoch": 2.0600625084930018, + "grad_norm": 0.7477074265480042, + "learning_rate": 0.0007424921864383748, + "loss": 3.6138, + "step": 30320 + }, + { + "epoch": 2.0604022285636634, + "grad_norm": 0.7729437947273254, + "learning_rate": 0.000742449721429542, + "loss": 3.6884, + "step": 30325 + }, + { + "epoch": 2.0607419486343255, + "grad_norm": 0.866472601890564, + "learning_rate": 0.0007424072564207094, + "loss": 3.3981, + "step": 30330 + }, + { + "epoch": 2.061081668704987, + "grad_norm": 0.9379312992095947, + "learning_rate": 0.0007423647914118766, + "loss": 3.4964, + "step": 30335 + }, + { + "epoch": 2.0614213887756487, + "grad_norm": 0.7489362359046936, + "learning_rate": 0.0007423223264030438, + "loss": 3.5504, + "step": 30340 + }, + { + "epoch": 2.061761108846311, + "grad_norm": 0.8190793395042419, + "learning_rate": 0.0007422798613942113, + "loss": 3.4219, + "step": 30345 + }, + { + "epoch": 2.0621008289169724, + "grad_norm": 0.8518457412719727, + "learning_rate": 0.0007422373963853785, + "loss": 3.6322, + "step": 30350 + }, + { + "epoch": 2.062440548987634, + "grad_norm": 0.7793540954589844, + "learning_rate": 0.0007421949313765457, + "loss": 3.5534, + "step": 30355 + }, + { + "epoch": 2.062780269058296, + "grad_norm": 0.9046018719673157, + "learning_rate": 0.000742152466367713, + "loss": 3.6531, + "step": 30360 + }, + { + "epoch": 2.0631199891289578, + "grad_norm": 0.8473179340362549, + "learning_rate": 0.0007421100013588803, + "loss": 3.3025, + "step": 30365 + }, + { + "epoch": 2.0634597091996194, + "grad_norm": 1.0683820247650146, + "learning_rate": 0.0007420675363500475, + "loss": 3.485, + "step": 30370 + }, + { + "epoch": 2.0637994292702815, + "grad_norm": 0.9478075504302979, + "learning_rate": 0.0007420250713412148, + "loss": 3.3582, + "step": 30375 + }, + { + "epoch": 2.064139149340943, + "grad_norm": 0.763405442237854, + "learning_rate": 0.0007419826063323822, + "loss": 3.4806, + "step": 30380 + }, + { + "epoch": 2.0644788694116047, + "grad_norm": 0.8223350644111633, + "learning_rate": 0.0007419401413235494, + "loss": 3.5762, + "step": 30385 + }, + { + "epoch": 2.064818589482267, + "grad_norm": 0.9755221009254456, + "learning_rate": 0.0007418976763147167, + "loss": 3.7148, + "step": 30390 + }, + { + "epoch": 2.0651583095529285, + "grad_norm": 1.0897862911224365, + "learning_rate": 0.000741855211305884, + "loss": 3.6498, + "step": 30395 + }, + { + "epoch": 2.06549802962359, + "grad_norm": 0.9741256237030029, + "learning_rate": 0.0007418127462970512, + "loss": 3.3485, + "step": 30400 + }, + { + "epoch": 2.0658377496942517, + "grad_norm": 0.8971060514450073, + "learning_rate": 0.0007417702812882185, + "loss": 3.4101, + "step": 30405 + }, + { + "epoch": 2.066177469764914, + "grad_norm": 0.7620330452919006, + "learning_rate": 0.0007417278162793857, + "loss": 3.6017, + "step": 30410 + }, + { + "epoch": 2.0665171898355754, + "grad_norm": 0.6156169176101685, + "learning_rate": 0.0007416853512705531, + "loss": 3.4944, + "step": 30415 + }, + { + "epoch": 2.066856909906237, + "grad_norm": 1.0007234811782837, + "learning_rate": 0.0007416428862617204, + "loss": 3.3781, + "step": 30420 + }, + { + "epoch": 2.067196629976899, + "grad_norm": 0.7318184971809387, + "learning_rate": 0.0007416004212528876, + "loss": 3.5418, + "step": 30425 + }, + { + "epoch": 2.0675363500475608, + "grad_norm": 0.7923423051834106, + "learning_rate": 0.0007415579562440549, + "loss": 3.5652, + "step": 30430 + }, + { + "epoch": 2.0678760701182224, + "grad_norm": 0.612704873085022, + "learning_rate": 0.0007415154912352222, + "loss": 3.5636, + "step": 30435 + }, + { + "epoch": 2.0682157901888845, + "grad_norm": 0.6869993805885315, + "learning_rate": 0.0007414730262263894, + "loss": 3.7997, + "step": 30440 + }, + { + "epoch": 2.068555510259546, + "grad_norm": 1.023064374923706, + "learning_rate": 0.0007414305612175566, + "loss": 3.6569, + "step": 30445 + }, + { + "epoch": 2.0688952303302077, + "grad_norm": 0.6793161630630493, + "learning_rate": 0.0007413880962087241, + "loss": 3.7399, + "step": 30450 + }, + { + "epoch": 2.06923495040087, + "grad_norm": 0.8024521470069885, + "learning_rate": 0.0007413456311998913, + "loss": 3.3655, + "step": 30455 + }, + { + "epoch": 2.0695746704715314, + "grad_norm": 0.9406973719596863, + "learning_rate": 0.0007413031661910585, + "loss": 3.6199, + "step": 30460 + }, + { + "epoch": 2.069914390542193, + "grad_norm": 0.8953339457511902, + "learning_rate": 0.0007412607011822259, + "loss": 3.616, + "step": 30465 + }, + { + "epoch": 2.070254110612855, + "grad_norm": 0.810542106628418, + "learning_rate": 0.0007412182361733931, + "loss": 3.5286, + "step": 30470 + }, + { + "epoch": 2.0705938306835168, + "grad_norm": 0.8588165044784546, + "learning_rate": 0.0007411757711645603, + "loss": 3.3497, + "step": 30475 + }, + { + "epoch": 2.0709335507541784, + "grad_norm": 2.339153289794922, + "learning_rate": 0.0007411333061557277, + "loss": 3.5623, + "step": 30480 + }, + { + "epoch": 2.0712732708248405, + "grad_norm": 0.9562128782272339, + "learning_rate": 0.000741090841146895, + "loss": 3.349, + "step": 30485 + }, + { + "epoch": 2.071612990895502, + "grad_norm": 0.7850987315177917, + "learning_rate": 0.0007410483761380622, + "loss": 3.5961, + "step": 30490 + }, + { + "epoch": 2.0719527109661637, + "grad_norm": 0.8163318037986755, + "learning_rate": 0.0007410059111292296, + "loss": 3.6388, + "step": 30495 + }, + { + "epoch": 2.072292431036826, + "grad_norm": 0.990432858467102, + "learning_rate": 0.0007409634461203968, + "loss": 3.2966, + "step": 30500 + }, + { + "epoch": 2.0726321511074874, + "grad_norm": 0.8055406212806702, + "learning_rate": 0.0007409209811115641, + "loss": 3.7438, + "step": 30505 + }, + { + "epoch": 2.072971871178149, + "grad_norm": 1.0218015909194946, + "learning_rate": 0.0007408785161027313, + "loss": 3.4548, + "step": 30510 + }, + { + "epoch": 2.073311591248811, + "grad_norm": 1.0589393377304077, + "learning_rate": 0.0007408360510938986, + "loss": 3.4523, + "step": 30515 + }, + { + "epoch": 2.073651311319473, + "grad_norm": 0.7979984283447266, + "learning_rate": 0.000740793586085066, + "loss": 3.5664, + "step": 30520 + }, + { + "epoch": 2.0739910313901344, + "grad_norm": 1.0792994499206543, + "learning_rate": 0.0007407511210762332, + "loss": 3.7811, + "step": 30525 + }, + { + "epoch": 2.0743307514607965, + "grad_norm": 0.8642266392707825, + "learning_rate": 0.0007407086560674005, + "loss": 3.5929, + "step": 30530 + }, + { + "epoch": 2.074670471531458, + "grad_norm": 0.9011669754981995, + "learning_rate": 0.0007406661910585678, + "loss": 3.3067, + "step": 30535 + }, + { + "epoch": 2.0750101916021197, + "grad_norm": 0.8261159062385559, + "learning_rate": 0.000740623726049735, + "loss": 3.3304, + "step": 30540 + }, + { + "epoch": 2.075349911672782, + "grad_norm": 1.800051212310791, + "learning_rate": 0.0007405812610409022, + "loss": 3.5425, + "step": 30545 + }, + { + "epoch": 2.0756896317434435, + "grad_norm": 0.7925270199775696, + "learning_rate": 0.0007405387960320697, + "loss": 3.5296, + "step": 30550 + }, + { + "epoch": 2.076029351814105, + "grad_norm": 3.5795559883117676, + "learning_rate": 0.0007404963310232369, + "loss": 3.5929, + "step": 30555 + }, + { + "epoch": 2.076369071884767, + "grad_norm": 0.9162275791168213, + "learning_rate": 0.0007404538660144041, + "loss": 3.5214, + "step": 30560 + }, + { + "epoch": 2.076708791955429, + "grad_norm": 1.4265096187591553, + "learning_rate": 0.0007404114010055715, + "loss": 3.5892, + "step": 30565 + }, + { + "epoch": 2.0770485120260904, + "grad_norm": 0.7997646927833557, + "learning_rate": 0.0007403689359967387, + "loss": 3.3387, + "step": 30570 + }, + { + "epoch": 2.0773882320967525, + "grad_norm": 1.1077004671096802, + "learning_rate": 0.0007403264709879059, + "loss": 3.5577, + "step": 30575 + }, + { + "epoch": 2.077727952167414, + "grad_norm": 0.9957609176635742, + "learning_rate": 0.0007402840059790733, + "loss": 3.613, + "step": 30580 + }, + { + "epoch": 2.0780676722380758, + "grad_norm": 0.846028745174408, + "learning_rate": 0.0007402415409702406, + "loss": 3.4101, + "step": 30585 + }, + { + "epoch": 2.0784073923087374, + "grad_norm": 0.8597121834754944, + "learning_rate": 0.0007401990759614078, + "loss": 3.415, + "step": 30590 + }, + { + "epoch": 2.0787471123793995, + "grad_norm": 0.9149390459060669, + "learning_rate": 0.0007401566109525752, + "loss": 3.4515, + "step": 30595 + }, + { + "epoch": 2.079086832450061, + "grad_norm": 0.9115296006202698, + "learning_rate": 0.0007401141459437424, + "loss": 3.6031, + "step": 30600 + }, + { + "epoch": 2.0794265525207227, + "grad_norm": 0.8619689345359802, + "learning_rate": 0.0007400716809349096, + "loss": 3.7249, + "step": 30605 + }, + { + "epoch": 2.079766272591385, + "grad_norm": 0.8290433287620544, + "learning_rate": 0.0007400292159260769, + "loss": 3.7367, + "step": 30610 + }, + { + "epoch": 2.0801059926620464, + "grad_norm": 1.5585417747497559, + "learning_rate": 0.0007399867509172442, + "loss": 3.5775, + "step": 30615 + }, + { + "epoch": 2.080445712732708, + "grad_norm": 0.9745950698852539, + "learning_rate": 0.0007399442859084115, + "loss": 3.3943, + "step": 30620 + }, + { + "epoch": 2.08078543280337, + "grad_norm": 0.7606522440910339, + "learning_rate": 0.0007399018208995788, + "loss": 3.6931, + "step": 30625 + }, + { + "epoch": 2.0811251528740318, + "grad_norm": 0.8307530283927917, + "learning_rate": 0.0007398593558907461, + "loss": 3.5827, + "step": 30630 + }, + { + "epoch": 2.0814648729446934, + "grad_norm": 0.9678487181663513, + "learning_rate": 0.0007398168908819133, + "loss": 3.5312, + "step": 30635 + }, + { + "epoch": 2.0818045930153555, + "grad_norm": 0.9765388369560242, + "learning_rate": 0.0007397744258730806, + "loss": 3.5896, + "step": 30640 + }, + { + "epoch": 2.082144313086017, + "grad_norm": 0.997481107711792, + "learning_rate": 0.0007397319608642479, + "loss": 3.9057, + "step": 30645 + }, + { + "epoch": 2.0824840331566787, + "grad_norm": 0.8829625248908997, + "learning_rate": 0.0007396894958554151, + "loss": 3.7211, + "step": 30650 + }, + { + "epoch": 2.082823753227341, + "grad_norm": 0.8595649003982544, + "learning_rate": 0.0007396470308465825, + "loss": 3.7974, + "step": 30655 + }, + { + "epoch": 2.0831634732980024, + "grad_norm": 1.0464352369308472, + "learning_rate": 0.0007396045658377497, + "loss": 3.6111, + "step": 30660 + }, + { + "epoch": 2.083503193368664, + "grad_norm": 0.8240364193916321, + "learning_rate": 0.000739562100828917, + "loss": 3.4898, + "step": 30665 + }, + { + "epoch": 2.083842913439326, + "grad_norm": 1.0037145614624023, + "learning_rate": 0.0007395196358200843, + "loss": 3.5247, + "step": 30670 + }, + { + "epoch": 2.084182633509988, + "grad_norm": 0.8251913785934448, + "learning_rate": 0.0007394771708112515, + "loss": 3.5309, + "step": 30675 + }, + { + "epoch": 2.0845223535806494, + "grad_norm": 0.8879213929176331, + "learning_rate": 0.0007394347058024188, + "loss": 3.4845, + "step": 30680 + }, + { + "epoch": 2.0848620736513115, + "grad_norm": 1.0858031511306763, + "learning_rate": 0.0007393922407935861, + "loss": 3.6329, + "step": 30685 + }, + { + "epoch": 2.085201793721973, + "grad_norm": 1.2666829824447632, + "learning_rate": 0.0007393497757847534, + "loss": 3.3307, + "step": 30690 + }, + { + "epoch": 2.0855415137926347, + "grad_norm": 0.9408236742019653, + "learning_rate": 0.0007393073107759207, + "loss": 3.5853, + "step": 30695 + }, + { + "epoch": 2.085881233863297, + "grad_norm": 0.919680655002594, + "learning_rate": 0.000739264845767088, + "loss": 3.3659, + "step": 30700 + }, + { + "epoch": 2.0862209539339585, + "grad_norm": 0.7047791481018066, + "learning_rate": 0.0007392223807582552, + "loss": 3.5284, + "step": 30705 + }, + { + "epoch": 2.08656067400462, + "grad_norm": 1.0470675230026245, + "learning_rate": 0.0007391799157494224, + "loss": 3.5477, + "step": 30710 + }, + { + "epoch": 2.086900394075282, + "grad_norm": 0.9449440240859985, + "learning_rate": 0.0007391374507405898, + "loss": 3.5362, + "step": 30715 + }, + { + "epoch": 2.087240114145944, + "grad_norm": 0.785474956035614, + "learning_rate": 0.000739094985731757, + "loss": 3.6538, + "step": 30720 + }, + { + "epoch": 2.0875798342166054, + "grad_norm": 0.7510668635368347, + "learning_rate": 0.0007390525207229243, + "loss": 3.8347, + "step": 30725 + }, + { + "epoch": 2.0879195542872675, + "grad_norm": 1.516573190689087, + "learning_rate": 0.0007390100557140917, + "loss": 3.6469, + "step": 30730 + }, + { + "epoch": 2.088259274357929, + "grad_norm": 1.1345758438110352, + "learning_rate": 0.0007389675907052589, + "loss": 3.5847, + "step": 30735 + }, + { + "epoch": 2.0885989944285908, + "grad_norm": 0.6628506779670715, + "learning_rate": 0.0007389251256964261, + "loss": 3.5002, + "step": 30740 + }, + { + "epoch": 2.0889387144992524, + "grad_norm": 0.8574951887130737, + "learning_rate": 0.0007388826606875935, + "loss": 3.5501, + "step": 30745 + }, + { + "epoch": 2.0892784345699145, + "grad_norm": 0.8082968592643738, + "learning_rate": 0.0007388401956787607, + "loss": 3.6739, + "step": 30750 + }, + { + "epoch": 2.089618154640576, + "grad_norm": 0.7693761587142944, + "learning_rate": 0.0007387977306699279, + "loss": 3.6925, + "step": 30755 + }, + { + "epoch": 2.0899578747112377, + "grad_norm": 0.894877016544342, + "learning_rate": 0.0007387552656610953, + "loss": 3.4466, + "step": 30760 + }, + { + "epoch": 2.0902975947819, + "grad_norm": 0.6918829679489136, + "learning_rate": 0.0007387128006522626, + "loss": 4.0783, + "step": 30765 + }, + { + "epoch": 2.0906373148525614, + "grad_norm": 0.6671911478042603, + "learning_rate": 0.0007386703356434298, + "loss": 3.6155, + "step": 30770 + }, + { + "epoch": 2.090977034923223, + "grad_norm": 0.7401109933853149, + "learning_rate": 0.0007386278706345971, + "loss": 3.455, + "step": 30775 + }, + { + "epoch": 2.091316754993885, + "grad_norm": 0.6841751933097839, + "learning_rate": 0.0007385854056257644, + "loss": 3.4967, + "step": 30780 + }, + { + "epoch": 2.0916564750645468, + "grad_norm": 0.9585409760475159, + "learning_rate": 0.0007385429406169316, + "loss": 3.6054, + "step": 30785 + }, + { + "epoch": 2.0919961951352084, + "grad_norm": 0.7005188465118408, + "learning_rate": 0.0007385004756080989, + "loss": 3.6842, + "step": 30790 + }, + { + "epoch": 2.0923359152058705, + "grad_norm": 1.0339319705963135, + "learning_rate": 0.0007384580105992663, + "loss": 3.6938, + "step": 30795 + }, + { + "epoch": 2.092675635276532, + "grad_norm": 0.67942214012146, + "learning_rate": 0.0007384155455904335, + "loss": 3.5447, + "step": 30800 + }, + { + "epoch": 2.0930153553471937, + "grad_norm": 1.498030185699463, + "learning_rate": 0.0007383730805816008, + "loss": 3.4657, + "step": 30805 + }, + { + "epoch": 2.093355075417856, + "grad_norm": 0.8941583037376404, + "learning_rate": 0.000738330615572768, + "loss": 3.5493, + "step": 30810 + }, + { + "epoch": 2.0936947954885174, + "grad_norm": 0.8852878212928772, + "learning_rate": 0.0007382881505639353, + "loss": 3.6949, + "step": 30815 + }, + { + "epoch": 2.094034515559179, + "grad_norm": 1.6694402694702148, + "learning_rate": 0.0007382456855551026, + "loss": 3.4862, + "step": 30820 + }, + { + "epoch": 2.094374235629841, + "grad_norm": 0.9616932272911072, + "learning_rate": 0.0007382032205462698, + "loss": 3.5881, + "step": 30825 + }, + { + "epoch": 2.094713955700503, + "grad_norm": 0.9270486831665039, + "learning_rate": 0.0007381607555374372, + "loss": 3.6069, + "step": 30830 + }, + { + "epoch": 2.0950536757711644, + "grad_norm": 0.8870041966438293, + "learning_rate": 0.0007381182905286045, + "loss": 3.5657, + "step": 30835 + }, + { + "epoch": 2.0953933958418265, + "grad_norm": 0.869133710861206, + "learning_rate": 0.0007380758255197717, + "loss": 3.5809, + "step": 30840 + }, + { + "epoch": 2.095733115912488, + "grad_norm": 1.1645175218582153, + "learning_rate": 0.000738033360510939, + "loss": 3.7026, + "step": 30845 + }, + { + "epoch": 2.0960728359831498, + "grad_norm": 0.6898667216300964, + "learning_rate": 0.0007379908955021063, + "loss": 3.825, + "step": 30850 + }, + { + "epoch": 2.096412556053812, + "grad_norm": 0.8482049107551575, + "learning_rate": 0.0007379484304932735, + "loss": 3.5534, + "step": 30855 + }, + { + "epoch": 2.0967522761244735, + "grad_norm": 0.8328823447227478, + "learning_rate": 0.0007379059654844408, + "loss": 3.6489, + "step": 30860 + }, + { + "epoch": 2.097091996195135, + "grad_norm": 1.1751227378845215, + "learning_rate": 0.0007378635004756082, + "loss": 3.2842, + "step": 30865 + }, + { + "epoch": 2.097431716265797, + "grad_norm": 0.8591613173484802, + "learning_rate": 0.0007378210354667754, + "loss": 3.7298, + "step": 30870 + }, + { + "epoch": 2.097771436336459, + "grad_norm": 0.8904399275779724, + "learning_rate": 0.0007377785704579427, + "loss": 3.5831, + "step": 30875 + }, + { + "epoch": 2.0981111564071204, + "grad_norm": 1.0145186185836792, + "learning_rate": 0.00073773610544911, + "loss": 3.4177, + "step": 30880 + }, + { + "epoch": 2.0984508764777825, + "grad_norm": 1.009313702583313, + "learning_rate": 0.0007376936404402772, + "loss": 3.7288, + "step": 30885 + }, + { + "epoch": 2.098790596548444, + "grad_norm": 0.8114494681358337, + "learning_rate": 0.0007376511754314445, + "loss": 3.7062, + "step": 30890 + }, + { + "epoch": 2.0991303166191058, + "grad_norm": 1.1027445793151855, + "learning_rate": 0.0007376087104226117, + "loss": 3.4822, + "step": 30895 + }, + { + "epoch": 2.099470036689768, + "grad_norm": 0.8629288673400879, + "learning_rate": 0.0007375662454137791, + "loss": 3.5784, + "step": 30900 + }, + { + "epoch": 2.0998097567604295, + "grad_norm": 2.1976678371429443, + "learning_rate": 0.0007375237804049464, + "loss": 3.8541, + "step": 30905 + }, + { + "epoch": 2.100149476831091, + "grad_norm": 1.0396888256072998, + "learning_rate": 0.0007374813153961136, + "loss": 3.6352, + "step": 30910 + }, + { + "epoch": 2.100489196901753, + "grad_norm": 0.8514304757118225, + "learning_rate": 0.0007374388503872809, + "loss": 3.6542, + "step": 30915 + }, + { + "epoch": 2.100828916972415, + "grad_norm": 0.9642804861068726, + "learning_rate": 0.0007373963853784482, + "loss": 3.2502, + "step": 30920 + }, + { + "epoch": 2.1011686370430764, + "grad_norm": 1.1004514694213867, + "learning_rate": 0.0007373539203696154, + "loss": 3.765, + "step": 30925 + }, + { + "epoch": 2.101508357113738, + "grad_norm": 0.7562880516052246, + "learning_rate": 0.0007373114553607827, + "loss": 3.4817, + "step": 30930 + }, + { + "epoch": 2.1018480771844, + "grad_norm": 0.8613396883010864, + "learning_rate": 0.0007372689903519501, + "loss": 3.6272, + "step": 30935 + }, + { + "epoch": 2.1021877972550618, + "grad_norm": 0.6694457530975342, + "learning_rate": 0.0007372265253431173, + "loss": 3.7453, + "step": 30940 + }, + { + "epoch": 2.1025275173257234, + "grad_norm": 0.8620917201042175, + "learning_rate": 0.0007371840603342845, + "loss": 3.2337, + "step": 30945 + }, + { + "epoch": 2.1028672373963855, + "grad_norm": 0.9174368381500244, + "learning_rate": 0.0007371415953254519, + "loss": 3.441, + "step": 30950 + }, + { + "epoch": 2.103206957467047, + "grad_norm": 1.4081940650939941, + "learning_rate": 0.0007370991303166191, + "loss": 3.6905, + "step": 30955 + }, + { + "epoch": 2.1035466775377087, + "grad_norm": 4.196097373962402, + "learning_rate": 0.0007370566653077863, + "loss": 3.913, + "step": 30960 + }, + { + "epoch": 2.103886397608371, + "grad_norm": 0.9279444217681885, + "learning_rate": 0.0007370142002989537, + "loss": 3.5284, + "step": 30965 + }, + { + "epoch": 2.1042261176790324, + "grad_norm": 0.8992620706558228, + "learning_rate": 0.000736971735290121, + "loss": 3.3375, + "step": 30970 + }, + { + "epoch": 2.104565837749694, + "grad_norm": 1.0974339246749878, + "learning_rate": 0.0007369292702812882, + "loss": 3.5736, + "step": 30975 + }, + { + "epoch": 2.104905557820356, + "grad_norm": 0.8205392360687256, + "learning_rate": 0.0007368868052724556, + "loss": 3.4748, + "step": 30980 + }, + { + "epoch": 2.105245277891018, + "grad_norm": 0.823762834072113, + "learning_rate": 0.0007368443402636228, + "loss": 3.4716, + "step": 30985 + }, + { + "epoch": 2.1055849979616794, + "grad_norm": 0.9685941934585571, + "learning_rate": 0.00073680187525479, + "loss": 3.5789, + "step": 30990 + }, + { + "epoch": 2.1059247180323415, + "grad_norm": 0.9597747325897217, + "learning_rate": 0.0007367594102459573, + "loss": 3.7585, + "step": 30995 + }, + { + "epoch": 2.106264438103003, + "grad_norm": 0.9118058085441589, + "learning_rate": 0.0007367169452371246, + "loss": 3.6869, + "step": 31000 + }, + { + "epoch": 2.1066041581736648, + "grad_norm": 0.7891155481338501, + "learning_rate": 0.0007366744802282919, + "loss": 3.4723, + "step": 31005 + }, + { + "epoch": 2.106943878244327, + "grad_norm": 0.8345896601676941, + "learning_rate": 0.0007366320152194592, + "loss": 3.5754, + "step": 31010 + }, + { + "epoch": 2.1072835983149885, + "grad_norm": 0.8935109376907349, + "learning_rate": 0.0007365895502106265, + "loss": 3.8115, + "step": 31015 + }, + { + "epoch": 2.10762331838565, + "grad_norm": 0.7765200138092041, + "learning_rate": 0.0007365470852017937, + "loss": 3.4921, + "step": 31020 + }, + { + "epoch": 2.107963038456312, + "grad_norm": 0.9236768484115601, + "learning_rate": 0.000736504620192961, + "loss": 3.5042, + "step": 31025 + }, + { + "epoch": 2.108302758526974, + "grad_norm": 0.7747576236724854, + "learning_rate": 0.0007364621551841283, + "loss": 3.6333, + "step": 31030 + }, + { + "epoch": 2.1086424785976354, + "grad_norm": 1.0654182434082031, + "learning_rate": 0.0007364196901752955, + "loss": 3.3361, + "step": 31035 + }, + { + "epoch": 2.1089821986682975, + "grad_norm": 1.389291524887085, + "learning_rate": 0.0007363772251664629, + "loss": 3.5247, + "step": 31040 + }, + { + "epoch": 2.109321918738959, + "grad_norm": 0.6754319667816162, + "learning_rate": 0.0007363347601576301, + "loss": 3.6945, + "step": 31045 + }, + { + "epoch": 2.1096616388096208, + "grad_norm": 1.1766279935836792, + "learning_rate": 0.0007362922951487974, + "loss": 3.5849, + "step": 31050 + }, + { + "epoch": 2.110001358880283, + "grad_norm": 0.8909016847610474, + "learning_rate": 0.0007362498301399647, + "loss": 3.8529, + "step": 31055 + }, + { + "epoch": 2.1103410789509445, + "grad_norm": 1.0372370481491089, + "learning_rate": 0.0007362073651311319, + "loss": 3.5731, + "step": 31060 + }, + { + "epoch": 2.110680799021606, + "grad_norm": 1.0577566623687744, + "learning_rate": 0.0007361649001222992, + "loss": 3.6801, + "step": 31065 + }, + { + "epoch": 2.111020519092268, + "grad_norm": 0.8992224931716919, + "learning_rate": 0.0007361224351134666, + "loss": 3.6454, + "step": 31070 + }, + { + "epoch": 2.11136023916293, + "grad_norm": 0.7491416931152344, + "learning_rate": 0.0007360799701046338, + "loss": 3.8411, + "step": 31075 + }, + { + "epoch": 2.1116999592335914, + "grad_norm": 0.98088139295578, + "learning_rate": 0.0007360375050958011, + "loss": 3.4313, + "step": 31080 + }, + { + "epoch": 2.112039679304253, + "grad_norm": 1.5353848934173584, + "learning_rate": 0.0007359950400869684, + "loss": 3.4704, + "step": 31085 + }, + { + "epoch": 2.112379399374915, + "grad_norm": 0.8137950301170349, + "learning_rate": 0.0007359525750781356, + "loss": 3.3954, + "step": 31090 + }, + { + "epoch": 2.1127191194455768, + "grad_norm": 0.7912168502807617, + "learning_rate": 0.0007359101100693028, + "loss": 3.5286, + "step": 31095 + }, + { + "epoch": 2.1130588395162384, + "grad_norm": 0.7438257932662964, + "learning_rate": 0.0007358676450604702, + "loss": 3.366, + "step": 31100 + }, + { + "epoch": 2.1133985595869005, + "grad_norm": 0.9867879152297974, + "learning_rate": 0.0007358251800516375, + "loss": 3.6732, + "step": 31105 + }, + { + "epoch": 2.113738279657562, + "grad_norm": 1.0709060430526733, + "learning_rate": 0.0007357827150428047, + "loss": 3.744, + "step": 31110 + }, + { + "epoch": 2.1140779997282237, + "grad_norm": 1.883725643157959, + "learning_rate": 0.0007357402500339721, + "loss": 3.3822, + "step": 31115 + }, + { + "epoch": 2.114417719798886, + "grad_norm": 0.7549963593482971, + "learning_rate": 0.0007356977850251393, + "loss": 3.5248, + "step": 31120 + }, + { + "epoch": 2.1147574398695475, + "grad_norm": 0.9590659737586975, + "learning_rate": 0.0007356553200163065, + "loss": 3.8197, + "step": 31125 + }, + { + "epoch": 2.115097159940209, + "grad_norm": 0.9155833721160889, + "learning_rate": 0.0007356128550074739, + "loss": 3.4443, + "step": 31130 + }, + { + "epoch": 2.115436880010871, + "grad_norm": 0.922195315361023, + "learning_rate": 0.0007355703899986411, + "loss": 3.227, + "step": 31135 + }, + { + "epoch": 2.115776600081533, + "grad_norm": 0.8331239223480225, + "learning_rate": 0.0007355279249898084, + "loss": 3.3328, + "step": 31140 + }, + { + "epoch": 2.1161163201521944, + "grad_norm": 0.8495606184005737, + "learning_rate": 0.0007354854599809757, + "loss": 3.8232, + "step": 31145 + }, + { + "epoch": 2.1164560402228565, + "grad_norm": 6.0059990882873535, + "learning_rate": 0.000735442994972143, + "loss": 3.9294, + "step": 31150 + }, + { + "epoch": 2.116795760293518, + "grad_norm": 0.8534340262413025, + "learning_rate": 0.0007354005299633102, + "loss": 3.3439, + "step": 31155 + }, + { + "epoch": 2.1171354803641798, + "grad_norm": 1.2177562713623047, + "learning_rate": 0.0007353580649544775, + "loss": 3.2805, + "step": 31160 + }, + { + "epoch": 2.117475200434842, + "grad_norm": 0.9601883292198181, + "learning_rate": 0.0007353155999456448, + "loss": 3.7246, + "step": 31165 + }, + { + "epoch": 2.1178149205055035, + "grad_norm": 1.1258084774017334, + "learning_rate": 0.000735273134936812, + "loss": 3.4845, + "step": 31170 + }, + { + "epoch": 2.118154640576165, + "grad_norm": 0.8835456967353821, + "learning_rate": 0.0007352306699279794, + "loss": 3.6241, + "step": 31175 + }, + { + "epoch": 2.118494360646827, + "grad_norm": 0.7710118889808655, + "learning_rate": 0.0007351882049191467, + "loss": 3.6438, + "step": 31180 + }, + { + "epoch": 2.118834080717489, + "grad_norm": 0.8302015066146851, + "learning_rate": 0.000735145739910314, + "loss": 3.474, + "step": 31185 + }, + { + "epoch": 2.1191738007881504, + "grad_norm": 0.879035472869873, + "learning_rate": 0.0007351032749014812, + "loss": 3.4775, + "step": 31190 + }, + { + "epoch": 2.1195135208588125, + "grad_norm": 1.0090245008468628, + "learning_rate": 0.0007350608098926484, + "loss": 3.4724, + "step": 31195 + }, + { + "epoch": 2.119853240929474, + "grad_norm": 0.8304756283760071, + "learning_rate": 0.0007350183448838158, + "loss": 3.6889, + "step": 31200 + }, + { + "epoch": 2.1201929610001358, + "grad_norm": 1.096398115158081, + "learning_rate": 0.000734975879874983, + "loss": 3.755, + "step": 31205 + }, + { + "epoch": 2.120532681070798, + "grad_norm": 1.7347301244735718, + "learning_rate": 0.0007349334148661503, + "loss": 3.4463, + "step": 31210 + }, + { + "epoch": 2.1208724011414595, + "grad_norm": 0.9564123153686523, + "learning_rate": 0.0007348909498573177, + "loss": 3.6471, + "step": 31215 + }, + { + "epoch": 2.121212121212121, + "grad_norm": 2.1438848972320557, + "learning_rate": 0.0007348484848484849, + "loss": 3.5666, + "step": 31220 + }, + { + "epoch": 2.121551841282783, + "grad_norm": 1.0155041217803955, + "learning_rate": 0.0007348060198396521, + "loss": 3.4862, + "step": 31225 + }, + { + "epoch": 2.121891561353445, + "grad_norm": 0.7622844576835632, + "learning_rate": 0.0007347635548308195, + "loss": 3.6398, + "step": 31230 + }, + { + "epoch": 2.1222312814241064, + "grad_norm": 0.9833621382713318, + "learning_rate": 0.0007347210898219867, + "loss": 3.5922, + "step": 31235 + }, + { + "epoch": 2.1225710014947685, + "grad_norm": 0.8327182531356812, + "learning_rate": 0.0007346786248131539, + "loss": 3.601, + "step": 31240 + }, + { + "epoch": 2.12291072156543, + "grad_norm": 0.7352569103240967, + "learning_rate": 0.0007346361598043214, + "loss": 3.6843, + "step": 31245 + }, + { + "epoch": 2.123250441636092, + "grad_norm": 2.2563624382019043, + "learning_rate": 0.0007345936947954886, + "loss": 3.6987, + "step": 31250 + }, + { + "epoch": 2.123590161706754, + "grad_norm": 1.4925892353057861, + "learning_rate": 0.0007345512297866558, + "loss": 3.6914, + "step": 31255 + }, + { + "epoch": 2.1239298817774155, + "grad_norm": 0.6803785562515259, + "learning_rate": 0.0007345087647778231, + "loss": 3.4913, + "step": 31260 + }, + { + "epoch": 2.124269601848077, + "grad_norm": 0.9997332096099854, + "learning_rate": 0.0007344662997689904, + "loss": 3.2839, + "step": 31265 + }, + { + "epoch": 2.1246093219187387, + "grad_norm": 0.9116175770759583, + "learning_rate": 0.0007344238347601576, + "loss": 3.491, + "step": 31270 + }, + { + "epoch": 2.124949041989401, + "grad_norm": 0.877632737159729, + "learning_rate": 0.0007343813697513249, + "loss": 3.2774, + "step": 31275 + }, + { + "epoch": 2.1252887620600625, + "grad_norm": 0.8134282231330872, + "learning_rate": 0.0007343389047424923, + "loss": 3.5925, + "step": 31280 + }, + { + "epoch": 2.125628482130724, + "grad_norm": 0.8061376810073853, + "learning_rate": 0.0007342964397336595, + "loss": 3.6866, + "step": 31285 + }, + { + "epoch": 2.125968202201386, + "grad_norm": 0.9105165004730225, + "learning_rate": 0.0007342539747248268, + "loss": 3.6668, + "step": 31290 + }, + { + "epoch": 2.126307922272048, + "grad_norm": 0.7287615537643433, + "learning_rate": 0.000734211509715994, + "loss": 3.102, + "step": 31295 + }, + { + "epoch": 2.1266476423427094, + "grad_norm": 0.8636132478713989, + "learning_rate": 0.0007341690447071613, + "loss": 3.3987, + "step": 31300 + }, + { + "epoch": 2.1269873624133715, + "grad_norm": 1.0694997310638428, + "learning_rate": 0.0007341265796983286, + "loss": 3.393, + "step": 31305 + }, + { + "epoch": 2.127327082484033, + "grad_norm": 0.7727710008621216, + "learning_rate": 0.0007340841146894958, + "loss": 3.4094, + "step": 31310 + }, + { + "epoch": 2.1276668025546948, + "grad_norm": 0.7723323702812195, + "learning_rate": 0.0007340416496806632, + "loss": 3.4469, + "step": 31315 + }, + { + "epoch": 2.128006522625357, + "grad_norm": 0.9947643280029297, + "learning_rate": 0.0007339991846718305, + "loss": 3.5012, + "step": 31320 + }, + { + "epoch": 2.1283462426960185, + "grad_norm": 1.4604175090789795, + "learning_rate": 0.0007339567196629977, + "loss": 3.5038, + "step": 31325 + }, + { + "epoch": 2.12868596276668, + "grad_norm": 1.0697214603424072, + "learning_rate": 0.000733914254654165, + "loss": 3.5198, + "step": 31330 + }, + { + "epoch": 2.129025682837342, + "grad_norm": 1.5371419191360474, + "learning_rate": 0.0007338717896453323, + "loss": 3.7354, + "step": 31335 + }, + { + "epoch": 2.129365402908004, + "grad_norm": 0.8357107043266296, + "learning_rate": 0.0007338293246364995, + "loss": 3.5535, + "step": 31340 + }, + { + "epoch": 2.1297051229786654, + "grad_norm": 1.016566514968872, + "learning_rate": 0.0007337868596276667, + "loss": 3.5474, + "step": 31345 + }, + { + "epoch": 2.1300448430493275, + "grad_norm": 0.8068065047264099, + "learning_rate": 0.0007337443946188342, + "loss": 3.6461, + "step": 31350 + }, + { + "epoch": 2.130384563119989, + "grad_norm": 0.9013103246688843, + "learning_rate": 0.0007337019296100014, + "loss": 3.5535, + "step": 31355 + }, + { + "epoch": 2.1307242831906508, + "grad_norm": 1.0109392404556274, + "learning_rate": 0.0007336594646011686, + "loss": 3.3527, + "step": 31360 + }, + { + "epoch": 2.131064003261313, + "grad_norm": 0.8996456265449524, + "learning_rate": 0.000733616999592336, + "loss": 3.5124, + "step": 31365 + }, + { + "epoch": 2.1314037233319745, + "grad_norm": 0.736445963382721, + "learning_rate": 0.0007335745345835032, + "loss": 3.5266, + "step": 31370 + }, + { + "epoch": 2.131743443402636, + "grad_norm": 0.9316646456718445, + "learning_rate": 0.0007335320695746704, + "loss": 3.5799, + "step": 31375 + }, + { + "epoch": 2.132083163473298, + "grad_norm": 1.0985591411590576, + "learning_rate": 0.0007334896045658378, + "loss": 3.4757, + "step": 31380 + }, + { + "epoch": 2.13242288354396, + "grad_norm": 0.7550480365753174, + "learning_rate": 0.0007334471395570051, + "loss": 3.6612, + "step": 31385 + }, + { + "epoch": 2.1327626036146214, + "grad_norm": 1.1711262464523315, + "learning_rate": 0.0007334046745481723, + "loss": 3.7815, + "step": 31390 + }, + { + "epoch": 2.1331023236852835, + "grad_norm": 0.9393274784088135, + "learning_rate": 0.0007333622095393396, + "loss": 3.3389, + "step": 31395 + }, + { + "epoch": 2.133442043755945, + "grad_norm": 0.9003846049308777, + "learning_rate": 0.0007333197445305069, + "loss": 3.6046, + "step": 31400 + }, + { + "epoch": 2.133781763826607, + "grad_norm": 0.8741315603256226, + "learning_rate": 0.0007332772795216741, + "loss": 3.3711, + "step": 31405 + }, + { + "epoch": 2.134121483897269, + "grad_norm": 0.8278862833976746, + "learning_rate": 0.0007332348145128414, + "loss": 3.6874, + "step": 31410 + }, + { + "epoch": 2.1344612039679305, + "grad_norm": 1.0153931379318237, + "learning_rate": 0.0007331923495040087, + "loss": 3.5025, + "step": 31415 + }, + { + "epoch": 2.134800924038592, + "grad_norm": 0.8533931374549866, + "learning_rate": 0.000733149884495176, + "loss": 3.5955, + "step": 31420 + }, + { + "epoch": 2.1351406441092537, + "grad_norm": 0.9252077341079712, + "learning_rate": 0.0007331074194863433, + "loss": 3.7191, + "step": 31425 + }, + { + "epoch": 2.135480364179916, + "grad_norm": 1.2899731397628784, + "learning_rate": 0.0007330649544775106, + "loss": 3.6253, + "step": 31430 + }, + { + "epoch": 2.1358200842505775, + "grad_norm": 0.9998446106910706, + "learning_rate": 0.0007330224894686778, + "loss": 3.4152, + "step": 31435 + }, + { + "epoch": 2.136159804321239, + "grad_norm": 0.8295803070068359, + "learning_rate": 0.0007329800244598451, + "loss": 3.5939, + "step": 31440 + }, + { + "epoch": 2.136499524391901, + "grad_norm": 0.852884829044342, + "learning_rate": 0.0007329375594510123, + "loss": 3.3085, + "step": 31445 + }, + { + "epoch": 2.136839244462563, + "grad_norm": 0.9598560929298401, + "learning_rate": 0.0007328950944421796, + "loss": 3.5181, + "step": 31450 + }, + { + "epoch": 2.1371789645332244, + "grad_norm": 1.0232887268066406, + "learning_rate": 0.000732852629433347, + "loss": 3.5802, + "step": 31455 + }, + { + "epoch": 2.1375186846038865, + "grad_norm": 0.7499103546142578, + "learning_rate": 0.0007328101644245142, + "loss": 3.5178, + "step": 31460 + }, + { + "epoch": 2.137858404674548, + "grad_norm": 1.0047237873077393, + "learning_rate": 0.0007327676994156815, + "loss": 3.3931, + "step": 31465 + }, + { + "epoch": 2.1381981247452098, + "grad_norm": 1.0767709016799927, + "learning_rate": 0.0007327252344068488, + "loss": 3.2317, + "step": 31470 + }, + { + "epoch": 2.138537844815872, + "grad_norm": 0.9548283815383911, + "learning_rate": 0.000732682769398016, + "loss": 3.3947, + "step": 31475 + }, + { + "epoch": 2.1388775648865335, + "grad_norm": 0.9128632545471191, + "learning_rate": 0.0007326403043891832, + "loss": 3.3272, + "step": 31480 + }, + { + "epoch": 2.139217284957195, + "grad_norm": 0.9172502160072327, + "learning_rate": 0.0007325978393803506, + "loss": 3.6509, + "step": 31485 + }, + { + "epoch": 2.139557005027857, + "grad_norm": 1.179800033569336, + "learning_rate": 0.0007325553743715179, + "loss": 3.6627, + "step": 31490 + }, + { + "epoch": 2.139896725098519, + "grad_norm": 1.347105622291565, + "learning_rate": 0.0007325129093626851, + "loss": 3.4968, + "step": 31495 + }, + { + "epoch": 2.1402364451691804, + "grad_norm": 0.8214390277862549, + "learning_rate": 0.0007324704443538525, + "loss": 3.3475, + "step": 31500 + }, + { + "epoch": 2.1405761652398425, + "grad_norm": 1.0101730823516846, + "learning_rate": 0.0007324279793450197, + "loss": 3.4742, + "step": 31505 + }, + { + "epoch": 2.140915885310504, + "grad_norm": 1.0550105571746826, + "learning_rate": 0.0007323855143361869, + "loss": 3.4472, + "step": 31510 + }, + { + "epoch": 2.1412556053811658, + "grad_norm": 0.9533572793006897, + "learning_rate": 0.0007323430493273543, + "loss": 3.4693, + "step": 31515 + }, + { + "epoch": 2.141595325451828, + "grad_norm": 0.8550202250480652, + "learning_rate": 0.0007323005843185215, + "loss": 3.5023, + "step": 31520 + }, + { + "epoch": 2.1419350455224895, + "grad_norm": 0.8884824514389038, + "learning_rate": 0.0007322581193096889, + "loss": 3.3, + "step": 31525 + }, + { + "epoch": 2.142274765593151, + "grad_norm": 0.9333264231681824, + "learning_rate": 0.0007322156543008562, + "loss": 3.6576, + "step": 31530 + }, + { + "epoch": 2.142614485663813, + "grad_norm": 0.853769838809967, + "learning_rate": 0.0007321731892920234, + "loss": 3.5697, + "step": 31535 + }, + { + "epoch": 2.142954205734475, + "grad_norm": 0.715916633605957, + "learning_rate": 0.0007321307242831907, + "loss": 3.6361, + "step": 31540 + }, + { + "epoch": 2.1432939258051364, + "grad_norm": 0.8997506499290466, + "learning_rate": 0.0007320882592743579, + "loss": 3.4186, + "step": 31545 + }, + { + "epoch": 2.1436336458757985, + "grad_norm": 0.7232168912887573, + "learning_rate": 0.0007320457942655252, + "loss": 3.8004, + "step": 31550 + }, + { + "epoch": 2.14397336594646, + "grad_norm": 1.2245670557022095, + "learning_rate": 0.0007320033292566925, + "loss": 3.6357, + "step": 31555 + }, + { + "epoch": 2.144313086017122, + "grad_norm": 0.9550303220748901, + "learning_rate": 0.0007319608642478598, + "loss": 3.8391, + "step": 31560 + }, + { + "epoch": 2.144652806087784, + "grad_norm": 1.229727864265442, + "learning_rate": 0.0007319183992390271, + "loss": 3.5964, + "step": 31565 + }, + { + "epoch": 2.1449925261584455, + "grad_norm": 0.8532562851905823, + "learning_rate": 0.0007318759342301944, + "loss": 3.4641, + "step": 31570 + }, + { + "epoch": 2.145332246229107, + "grad_norm": 0.9707767963409424, + "learning_rate": 0.0007318334692213616, + "loss": 3.6332, + "step": 31575 + }, + { + "epoch": 2.145671966299769, + "grad_norm": 1.135162591934204, + "learning_rate": 0.0007317910042125288, + "loss": 3.4967, + "step": 31580 + }, + { + "epoch": 2.146011686370431, + "grad_norm": 0.9108568429946899, + "learning_rate": 0.0007317485392036962, + "loss": 3.5378, + "step": 31585 + }, + { + "epoch": 2.1463514064410925, + "grad_norm": 0.9217504858970642, + "learning_rate": 0.0007317060741948634, + "loss": 3.5773, + "step": 31590 + }, + { + "epoch": 2.1466911265117545, + "grad_norm": 1.0232627391815186, + "learning_rate": 0.0007316636091860307, + "loss": 3.4168, + "step": 31595 + }, + { + "epoch": 2.147030846582416, + "grad_norm": 0.7857924699783325, + "learning_rate": 0.0007316211441771981, + "loss": 3.5239, + "step": 31600 + }, + { + "epoch": 2.147370566653078, + "grad_norm": 1.0647212266921997, + "learning_rate": 0.0007315786791683653, + "loss": 3.3871, + "step": 31605 + }, + { + "epoch": 2.14771028672374, + "grad_norm": 0.9916369318962097, + "learning_rate": 0.0007315362141595325, + "loss": 3.3874, + "step": 31610 + }, + { + "epoch": 2.1480500067944015, + "grad_norm": 0.7761172652244568, + "learning_rate": 0.0007314937491506999, + "loss": 3.6364, + "step": 31615 + }, + { + "epoch": 2.148389726865063, + "grad_norm": 0.8220212459564209, + "learning_rate": 0.0007314512841418671, + "loss": 3.5365, + "step": 31620 + }, + { + "epoch": 2.1487294469357248, + "grad_norm": 1.8060420751571655, + "learning_rate": 0.0007314088191330343, + "loss": 3.6777, + "step": 31625 + }, + { + "epoch": 2.149069167006387, + "grad_norm": 1.0996476411819458, + "learning_rate": 0.0007313663541242018, + "loss": 3.5047, + "step": 31630 + }, + { + "epoch": 2.1494088870770485, + "grad_norm": 1.1521632671356201, + "learning_rate": 0.000731323889115369, + "loss": 3.4793, + "step": 31635 + }, + { + "epoch": 2.14974860714771, + "grad_norm": 1.1348285675048828, + "learning_rate": 0.0007312814241065362, + "loss": 3.6641, + "step": 31640 + }, + { + "epoch": 2.150088327218372, + "grad_norm": 0.9098120331764221, + "learning_rate": 0.0007312389590977035, + "loss": 3.4719, + "step": 31645 + }, + { + "epoch": 2.150428047289034, + "grad_norm": 0.8423128724098206, + "learning_rate": 0.0007311964940888708, + "loss": 3.7904, + "step": 31650 + }, + { + "epoch": 2.1507677673596954, + "grad_norm": 0.7840149998664856, + "learning_rate": 0.000731154029080038, + "loss": 3.7199, + "step": 31655 + }, + { + "epoch": 2.1511074874303575, + "grad_norm": 0.7220330238342285, + "learning_rate": 0.0007311115640712054, + "loss": 3.7743, + "step": 31660 + }, + { + "epoch": 2.151447207501019, + "grad_norm": 0.7513092160224915, + "learning_rate": 0.0007310690990623727, + "loss": 3.7024, + "step": 31665 + }, + { + "epoch": 2.1517869275716808, + "grad_norm": 0.7922874689102173, + "learning_rate": 0.0007310266340535399, + "loss": 3.6589, + "step": 31670 + }, + { + "epoch": 2.152126647642343, + "grad_norm": 0.9872101545333862, + "learning_rate": 0.0007309841690447072, + "loss": 3.6992, + "step": 31675 + }, + { + "epoch": 2.1524663677130045, + "grad_norm": 1.164272665977478, + "learning_rate": 0.0007309417040358744, + "loss": 3.4242, + "step": 31680 + }, + { + "epoch": 2.152806087783666, + "grad_norm": 0.9403621554374695, + "learning_rate": 0.0007308992390270417, + "loss": 3.5247, + "step": 31685 + }, + { + "epoch": 2.153145807854328, + "grad_norm": 1.1007064580917358, + "learning_rate": 0.000730856774018209, + "loss": 3.8102, + "step": 31690 + }, + { + "epoch": 2.15348552792499, + "grad_norm": 0.7965973615646362, + "learning_rate": 0.0007308143090093763, + "loss": 3.1304, + "step": 31695 + }, + { + "epoch": 2.1538252479956514, + "grad_norm": 0.8228327035903931, + "learning_rate": 0.0007307718440005436, + "loss": 3.6233, + "step": 31700 + }, + { + "epoch": 2.1541649680663135, + "grad_norm": 1.12892484664917, + "learning_rate": 0.0007307293789917109, + "loss": 3.6274, + "step": 31705 + }, + { + "epoch": 2.154504688136975, + "grad_norm": 0.9403474926948547, + "learning_rate": 0.0007306869139828781, + "loss": 3.5683, + "step": 31710 + }, + { + "epoch": 2.154844408207637, + "grad_norm": 0.7245636582374573, + "learning_rate": 0.0007306444489740454, + "loss": 3.4514, + "step": 31715 + }, + { + "epoch": 2.155184128278299, + "grad_norm": 1.0180444717407227, + "learning_rate": 0.0007306019839652127, + "loss": 3.5083, + "step": 31720 + }, + { + "epoch": 2.1555238483489605, + "grad_norm": 0.8991326093673706, + "learning_rate": 0.0007305595189563799, + "loss": 3.4947, + "step": 31725 + }, + { + "epoch": 2.155863568419622, + "grad_norm": 0.7014590501785278, + "learning_rate": 0.0007305170539475472, + "loss": 3.7489, + "step": 31730 + }, + { + "epoch": 2.156203288490284, + "grad_norm": 0.9303785562515259, + "learning_rate": 0.0007304745889387146, + "loss": 3.2134, + "step": 31735 + }, + { + "epoch": 2.156543008560946, + "grad_norm": 0.9980180859565735, + "learning_rate": 0.0007304321239298818, + "loss": 3.7292, + "step": 31740 + }, + { + "epoch": 2.1568827286316075, + "grad_norm": 0.8129411339759827, + "learning_rate": 0.000730389658921049, + "loss": 3.4917, + "step": 31745 + }, + { + "epoch": 2.1572224487022695, + "grad_norm": 0.7390148043632507, + "learning_rate": 0.0007303471939122164, + "loss": 3.261, + "step": 31750 + }, + { + "epoch": 2.157562168772931, + "grad_norm": 0.7620751857757568, + "learning_rate": 0.0007303047289033836, + "loss": 3.5553, + "step": 31755 + }, + { + "epoch": 2.157901888843593, + "grad_norm": 1.3783831596374512, + "learning_rate": 0.0007302622638945508, + "loss": 3.5328, + "step": 31760 + }, + { + "epoch": 2.1582416089142544, + "grad_norm": 0.9731869697570801, + "learning_rate": 0.0007302197988857183, + "loss": 3.428, + "step": 31765 + }, + { + "epoch": 2.1585813289849165, + "grad_norm": 0.830479085445404, + "learning_rate": 0.0007301773338768855, + "loss": 3.6724, + "step": 31770 + }, + { + "epoch": 2.158921049055578, + "grad_norm": 1.0441572666168213, + "learning_rate": 0.0007301348688680527, + "loss": 3.2826, + "step": 31775 + }, + { + "epoch": 2.1592607691262398, + "grad_norm": 0.9143582582473755, + "learning_rate": 0.00073009240385922, + "loss": 3.4812, + "step": 31780 + }, + { + "epoch": 2.159600489196902, + "grad_norm": 0.7824612855911255, + "learning_rate": 0.0007300499388503873, + "loss": 3.5202, + "step": 31785 + }, + { + "epoch": 2.1599402092675635, + "grad_norm": 0.8735954761505127, + "learning_rate": 0.0007300074738415545, + "loss": 3.8512, + "step": 31790 + }, + { + "epoch": 2.160279929338225, + "grad_norm": 0.798884928226471, + "learning_rate": 0.0007299650088327218, + "loss": 3.6352, + "step": 31795 + }, + { + "epoch": 2.160619649408887, + "grad_norm": 0.7739851474761963, + "learning_rate": 0.0007299225438238892, + "loss": 3.3699, + "step": 31800 + }, + { + "epoch": 2.160959369479549, + "grad_norm": 1.1707955598831177, + "learning_rate": 0.0007298800788150564, + "loss": 3.6818, + "step": 31805 + }, + { + "epoch": 2.1612990895502104, + "grad_norm": 0.9298028945922852, + "learning_rate": 0.0007298376138062237, + "loss": 3.8862, + "step": 31810 + }, + { + "epoch": 2.1616388096208725, + "grad_norm": 0.7256051301956177, + "learning_rate": 0.000729795148797391, + "loss": 3.6736, + "step": 31815 + }, + { + "epoch": 2.161978529691534, + "grad_norm": 0.8461595773696899, + "learning_rate": 0.0007297526837885582, + "loss": 3.555, + "step": 31820 + }, + { + "epoch": 2.1623182497621958, + "grad_norm": 0.7672905921936035, + "learning_rate": 0.0007297102187797255, + "loss": 3.7106, + "step": 31825 + }, + { + "epoch": 2.162657969832858, + "grad_norm": 1.0254000425338745, + "learning_rate": 0.0007296677537708927, + "loss": 3.5311, + "step": 31830 + }, + { + "epoch": 2.1629976899035195, + "grad_norm": 0.8974380493164062, + "learning_rate": 0.0007296252887620601, + "loss": 3.4838, + "step": 31835 + }, + { + "epoch": 2.163337409974181, + "grad_norm": 0.8667131662368774, + "learning_rate": 0.0007295828237532274, + "loss": 3.8285, + "step": 31840 + }, + { + "epoch": 2.163677130044843, + "grad_norm": 0.8348760008811951, + "learning_rate": 0.0007295403587443946, + "loss": 3.8101, + "step": 31845 + }, + { + "epoch": 2.164016850115505, + "grad_norm": 0.9489355683326721, + "learning_rate": 0.0007294978937355619, + "loss": 3.5214, + "step": 31850 + }, + { + "epoch": 2.1643565701861665, + "grad_norm": 0.6789339184761047, + "learning_rate": 0.0007294554287267292, + "loss": 3.6683, + "step": 31855 + }, + { + "epoch": 2.1646962902568285, + "grad_norm": 0.8377249836921692, + "learning_rate": 0.0007294129637178964, + "loss": 3.6151, + "step": 31860 + }, + { + "epoch": 2.16503601032749, + "grad_norm": 0.7281380891799927, + "learning_rate": 0.0007293704987090638, + "loss": 3.7247, + "step": 31865 + }, + { + "epoch": 2.165375730398152, + "grad_norm": 0.8438053131103516, + "learning_rate": 0.0007293280337002311, + "loss": 3.6184, + "step": 31870 + }, + { + "epoch": 2.165715450468814, + "grad_norm": 0.9111326336860657, + "learning_rate": 0.0007292855686913983, + "loss": 3.5633, + "step": 31875 + }, + { + "epoch": 2.1660551705394755, + "grad_norm": 0.9842153191566467, + "learning_rate": 0.0007292431036825656, + "loss": 3.3378, + "step": 31880 + }, + { + "epoch": 2.166394890610137, + "grad_norm": 0.6745507121086121, + "learning_rate": 0.0007292006386737329, + "loss": 3.7712, + "step": 31885 + }, + { + "epoch": 2.166734610680799, + "grad_norm": 0.6508551836013794, + "learning_rate": 0.0007291581736649001, + "loss": 3.6096, + "step": 31890 + }, + { + "epoch": 2.167074330751461, + "grad_norm": 0.931950569152832, + "learning_rate": 0.0007291157086560674, + "loss": 3.5407, + "step": 31895 + }, + { + "epoch": 2.1674140508221225, + "grad_norm": 0.9310864210128784, + "learning_rate": 0.0007290732436472347, + "loss": 3.5262, + "step": 31900 + }, + { + "epoch": 2.1677537708927845, + "grad_norm": 0.7360971570014954, + "learning_rate": 0.000729030778638402, + "loss": 3.6069, + "step": 31905 + }, + { + "epoch": 2.168093490963446, + "grad_norm": 0.9720098376274109, + "learning_rate": 0.0007289883136295693, + "loss": 3.7101, + "step": 31910 + }, + { + "epoch": 2.168433211034108, + "grad_norm": 1.2836759090423584, + "learning_rate": 0.0007289458486207366, + "loss": 3.6137, + "step": 31915 + }, + { + "epoch": 2.16877293110477, + "grad_norm": 0.9214504957199097, + "learning_rate": 0.0007289033836119038, + "loss": 3.6009, + "step": 31920 + }, + { + "epoch": 2.1691126511754315, + "grad_norm": 0.8209858536720276, + "learning_rate": 0.0007288609186030711, + "loss": 3.5509, + "step": 31925 + }, + { + "epoch": 2.169452371246093, + "grad_norm": 0.8441799879074097, + "learning_rate": 0.0007288184535942383, + "loss": 3.3762, + "step": 31930 + }, + { + "epoch": 2.169792091316755, + "grad_norm": 0.8889607191085815, + "learning_rate": 0.0007287759885854056, + "loss": 3.6119, + "step": 31935 + }, + { + "epoch": 2.170131811387417, + "grad_norm": 0.7439685463905334, + "learning_rate": 0.000728733523576573, + "loss": 3.6788, + "step": 31940 + }, + { + "epoch": 2.1704715314580785, + "grad_norm": 0.881527841091156, + "learning_rate": 0.0007286910585677402, + "loss": 3.575, + "step": 31945 + }, + { + "epoch": 2.1708112515287405, + "grad_norm": 1.2256739139556885, + "learning_rate": 0.0007286485935589075, + "loss": 3.4081, + "step": 31950 + }, + { + "epoch": 2.171150971599402, + "grad_norm": 0.9902128577232361, + "learning_rate": 0.0007286061285500748, + "loss": 3.6815, + "step": 31955 + }, + { + "epoch": 2.171490691670064, + "grad_norm": 0.7661766409873962, + "learning_rate": 0.000728563663541242, + "loss": 3.6245, + "step": 31960 + }, + { + "epoch": 2.1718304117407254, + "grad_norm": 0.9393168687820435, + "learning_rate": 0.0007285211985324092, + "loss": 3.6549, + "step": 31965 + }, + { + "epoch": 2.1721701318113875, + "grad_norm": 0.7727407217025757, + "learning_rate": 0.0007284787335235766, + "loss": 3.605, + "step": 31970 + }, + { + "epoch": 2.172509851882049, + "grad_norm": 1.4381295442581177, + "learning_rate": 0.0007284362685147439, + "loss": 3.5801, + "step": 31975 + }, + { + "epoch": 2.172849571952711, + "grad_norm": 1.126309871673584, + "learning_rate": 0.0007283938035059111, + "loss": 3.4406, + "step": 31980 + }, + { + "epoch": 2.173189292023373, + "grad_norm": 1.1382906436920166, + "learning_rate": 0.0007283513384970785, + "loss": 3.601, + "step": 31985 + }, + { + "epoch": 2.1735290120940345, + "grad_norm": 0.8750792741775513, + "learning_rate": 0.0007283088734882457, + "loss": 3.4852, + "step": 31990 + }, + { + "epoch": 2.173868732164696, + "grad_norm": 0.9577144980430603, + "learning_rate": 0.0007282664084794129, + "loss": 3.4796, + "step": 31995 + }, + { + "epoch": 2.174208452235358, + "grad_norm": 0.791410505771637, + "learning_rate": 0.0007282239434705803, + "loss": 3.4427, + "step": 32000 + }, + { + "epoch": 2.17454817230602, + "grad_norm": 1.2013370990753174, + "learning_rate": 0.0007281814784617475, + "loss": 3.9451, + "step": 32005 + }, + { + "epoch": 2.1748878923766815, + "grad_norm": 1.6336698532104492, + "learning_rate": 0.0007281390134529148, + "loss": 3.5103, + "step": 32010 + }, + { + "epoch": 2.1752276124473435, + "grad_norm": 0.744517982006073, + "learning_rate": 0.0007280965484440822, + "loss": 3.6145, + "step": 32015 + }, + { + "epoch": 2.175567332518005, + "grad_norm": 1.6193718910217285, + "learning_rate": 0.0007280540834352494, + "loss": 3.4132, + "step": 32020 + }, + { + "epoch": 2.175907052588667, + "grad_norm": 0.9837519526481628, + "learning_rate": 0.0007280116184264166, + "loss": 3.5978, + "step": 32025 + }, + { + "epoch": 2.176246772659329, + "grad_norm": 0.8598582744598389, + "learning_rate": 0.0007279691534175839, + "loss": 3.588, + "step": 32030 + }, + { + "epoch": 2.1765864927299905, + "grad_norm": 1.020618200302124, + "learning_rate": 0.0007279266884087512, + "loss": 3.6982, + "step": 32035 + }, + { + "epoch": 2.176926212800652, + "grad_norm": 0.7065830826759338, + "learning_rate": 0.0007278842233999184, + "loss": 3.3536, + "step": 32040 + }, + { + "epoch": 2.177265932871314, + "grad_norm": 0.7865496873855591, + "learning_rate": 0.0007278417583910858, + "loss": 3.6723, + "step": 32045 + }, + { + "epoch": 2.177605652941976, + "grad_norm": 1.6191611289978027, + "learning_rate": 0.0007277992933822531, + "loss": 3.4818, + "step": 32050 + }, + { + "epoch": 2.1779453730126375, + "grad_norm": 1.0523301362991333, + "learning_rate": 0.0007277568283734203, + "loss": 3.4491, + "step": 32055 + }, + { + "epoch": 2.1782850930832995, + "grad_norm": 0.7991454005241394, + "learning_rate": 0.0007277143633645876, + "loss": 3.6163, + "step": 32060 + }, + { + "epoch": 2.178624813153961, + "grad_norm": 0.8753361701965332, + "learning_rate": 0.0007276718983557549, + "loss": 3.5957, + "step": 32065 + }, + { + "epoch": 2.178964533224623, + "grad_norm": 0.7132070064544678, + "learning_rate": 0.0007276294333469221, + "loss": 3.6721, + "step": 32070 + }, + { + "epoch": 2.179304253295285, + "grad_norm": 1.0106687545776367, + "learning_rate": 0.0007275869683380894, + "loss": 3.5634, + "step": 32075 + }, + { + "epoch": 2.1796439733659465, + "grad_norm": 0.8802247047424316, + "learning_rate": 0.0007275445033292567, + "loss": 3.6829, + "step": 32080 + }, + { + "epoch": 2.179983693436608, + "grad_norm": 0.8940814733505249, + "learning_rate": 0.000727502038320424, + "loss": 3.876, + "step": 32085 + }, + { + "epoch": 2.18032341350727, + "grad_norm": 0.7642830610275269, + "learning_rate": 0.0007274595733115913, + "loss": 3.2673, + "step": 32090 + }, + { + "epoch": 2.180663133577932, + "grad_norm": 0.9021485447883606, + "learning_rate": 0.0007274171083027585, + "loss": 3.5188, + "step": 32095 + }, + { + "epoch": 2.1810028536485935, + "grad_norm": 0.6848967671394348, + "learning_rate": 0.0007273746432939258, + "loss": 3.6715, + "step": 32100 + }, + { + "epoch": 2.181342573719255, + "grad_norm": 1.0565532445907593, + "learning_rate": 0.0007273321782850931, + "loss": 3.51, + "step": 32105 + }, + { + "epoch": 2.181682293789917, + "grad_norm": 0.8490297794342041, + "learning_rate": 0.0007272897132762603, + "loss": 3.5255, + "step": 32110 + }, + { + "epoch": 2.182022013860579, + "grad_norm": 1.003279447555542, + "learning_rate": 0.0007272472482674277, + "loss": 3.4336, + "step": 32115 + }, + { + "epoch": 2.1823617339312404, + "grad_norm": 1.5934319496154785, + "learning_rate": 0.000727204783258595, + "loss": 3.5803, + "step": 32120 + }, + { + "epoch": 2.1827014540019025, + "grad_norm": 0.7985337972640991, + "learning_rate": 0.0007271623182497622, + "loss": 3.7513, + "step": 32125 + }, + { + "epoch": 2.183041174072564, + "grad_norm": 0.7382664084434509, + "learning_rate": 0.0007271198532409294, + "loss": 3.4937, + "step": 32130 + }, + { + "epoch": 2.183380894143226, + "grad_norm": 1.0742384195327759, + "learning_rate": 0.0007270773882320968, + "loss": 3.6019, + "step": 32135 + }, + { + "epoch": 2.183720614213888, + "grad_norm": 0.774810254573822, + "learning_rate": 0.000727034923223264, + "loss": 3.5984, + "step": 32140 + }, + { + "epoch": 2.1840603342845495, + "grad_norm": 0.9531111121177673, + "learning_rate": 0.0007269924582144312, + "loss": 3.4163, + "step": 32145 + }, + { + "epoch": 2.184400054355211, + "grad_norm": 0.8622708320617676, + "learning_rate": 0.0007269499932055987, + "loss": 3.4777, + "step": 32150 + }, + { + "epoch": 2.184739774425873, + "grad_norm": 0.8366093039512634, + "learning_rate": 0.0007269075281967659, + "loss": 3.2415, + "step": 32155 + }, + { + "epoch": 2.185079494496535, + "grad_norm": 0.7602417469024658, + "learning_rate": 0.0007268650631879331, + "loss": 3.5645, + "step": 32160 + }, + { + "epoch": 2.1854192145671965, + "grad_norm": 0.8636565208435059, + "learning_rate": 0.0007268225981791005, + "loss": 3.6202, + "step": 32165 + }, + { + "epoch": 2.1857589346378585, + "grad_norm": 0.8804951906204224, + "learning_rate": 0.0007267801331702677, + "loss": 3.5869, + "step": 32170 + }, + { + "epoch": 2.18609865470852, + "grad_norm": 1.190540075302124, + "learning_rate": 0.0007267376681614349, + "loss": 3.3927, + "step": 32175 + }, + { + "epoch": 2.186438374779182, + "grad_norm": 0.881720781326294, + "learning_rate": 0.0007266952031526022, + "loss": 3.5997, + "step": 32180 + }, + { + "epoch": 2.186778094849844, + "grad_norm": 0.8046108484268188, + "learning_rate": 0.0007266527381437696, + "loss": 3.7171, + "step": 32185 + }, + { + "epoch": 2.1871178149205055, + "grad_norm": 1.0526396036148071, + "learning_rate": 0.0007266102731349368, + "loss": 3.7582, + "step": 32190 + }, + { + "epoch": 2.187457534991167, + "grad_norm": 0.7152811884880066, + "learning_rate": 0.0007265678081261041, + "loss": 3.7741, + "step": 32195 + }, + { + "epoch": 2.187797255061829, + "grad_norm": 1.1968938112258911, + "learning_rate": 0.0007265253431172714, + "loss": 3.3958, + "step": 32200 + }, + { + "epoch": 2.188136975132491, + "grad_norm": 0.8464360237121582, + "learning_rate": 0.0007264828781084387, + "loss": 3.6738, + "step": 32205 + }, + { + "epoch": 2.1884766952031525, + "grad_norm": 0.8990806341171265, + "learning_rate": 0.0007264404130996059, + "loss": 3.7505, + "step": 32210 + }, + { + "epoch": 2.1888164152738145, + "grad_norm": 0.9128430485725403, + "learning_rate": 0.0007263979480907731, + "loss": 3.5511, + "step": 32215 + }, + { + "epoch": 2.189156135344476, + "grad_norm": 1.0432490110397339, + "learning_rate": 0.0007263554830819406, + "loss": 3.5986, + "step": 32220 + }, + { + "epoch": 2.189495855415138, + "grad_norm": 0.8298383355140686, + "learning_rate": 0.0007263130180731078, + "loss": 3.3839, + "step": 32225 + }, + { + "epoch": 2.1898355754858, + "grad_norm": 1.0730490684509277, + "learning_rate": 0.000726270553064275, + "loss": 3.614, + "step": 32230 + }, + { + "epoch": 2.1901752955564615, + "grad_norm": 0.937078595161438, + "learning_rate": 0.0007262280880554424, + "loss": 3.7991, + "step": 32235 + }, + { + "epoch": 2.190515015627123, + "grad_norm": 0.7295710444450378, + "learning_rate": 0.0007261856230466096, + "loss": 3.6619, + "step": 32240 + }, + { + "epoch": 2.190854735697785, + "grad_norm": 0.9337127208709717, + "learning_rate": 0.0007261431580377768, + "loss": 3.2815, + "step": 32245 + }, + { + "epoch": 2.191194455768447, + "grad_norm": 0.7765060663223267, + "learning_rate": 0.0007261006930289443, + "loss": 3.5362, + "step": 32250 + }, + { + "epoch": 2.1915341758391085, + "grad_norm": 0.8145327568054199, + "learning_rate": 0.0007260582280201115, + "loss": 3.5171, + "step": 32255 + }, + { + "epoch": 2.1918738959097706, + "grad_norm": 0.7860673069953918, + "learning_rate": 0.0007260157630112787, + "loss": 3.3525, + "step": 32260 + }, + { + "epoch": 2.192213615980432, + "grad_norm": 0.9476709365844727, + "learning_rate": 0.000725973298002446, + "loss": 3.8501, + "step": 32265 + }, + { + "epoch": 2.192553336051094, + "grad_norm": 0.9440532326698303, + "learning_rate": 0.0007259308329936133, + "loss": 3.5489, + "step": 32270 + }, + { + "epoch": 2.192893056121756, + "grad_norm": 0.7818008065223694, + "learning_rate": 0.0007258883679847805, + "loss": 3.63, + "step": 32275 + }, + { + "epoch": 2.1932327761924175, + "grad_norm": 0.9779481291770935, + "learning_rate": 0.0007258459029759478, + "loss": 3.5989, + "step": 32280 + }, + { + "epoch": 2.193572496263079, + "grad_norm": 1.2730967998504639, + "learning_rate": 0.0007258034379671152, + "loss": 3.0607, + "step": 32285 + }, + { + "epoch": 2.1939122163337412, + "grad_norm": 0.908113420009613, + "learning_rate": 0.0007257609729582824, + "loss": 3.52, + "step": 32290 + }, + { + "epoch": 2.194251936404403, + "grad_norm": 0.9928163290023804, + "learning_rate": 0.0007257185079494497, + "loss": 3.6207, + "step": 32295 + }, + { + "epoch": 2.1945916564750645, + "grad_norm": 0.9725337624549866, + "learning_rate": 0.000725676042940617, + "loss": 3.5023, + "step": 32300 + }, + { + "epoch": 2.194931376545726, + "grad_norm": 1.0681414604187012, + "learning_rate": 0.0007256335779317842, + "loss": 3.5948, + "step": 32305 + }, + { + "epoch": 2.195271096616388, + "grad_norm": 1.376857876777649, + "learning_rate": 0.0007255911129229515, + "loss": 3.5413, + "step": 32310 + }, + { + "epoch": 2.19561081668705, + "grad_norm": 0.7144401669502258, + "learning_rate": 0.0007255486479141187, + "loss": 3.3796, + "step": 32315 + }, + { + "epoch": 2.1959505367577115, + "grad_norm": 1.0554355382919312, + "learning_rate": 0.0007255061829052861, + "loss": 3.2464, + "step": 32320 + }, + { + "epoch": 2.1962902568283735, + "grad_norm": 0.8161914944648743, + "learning_rate": 0.0007254637178964534, + "loss": 3.6515, + "step": 32325 + }, + { + "epoch": 2.196629976899035, + "grad_norm": 0.7045878171920776, + "learning_rate": 0.0007254212528876206, + "loss": 3.4592, + "step": 32330 + }, + { + "epoch": 2.196969696969697, + "grad_norm": 1.0574604272842407, + "learning_rate": 0.0007253787878787879, + "loss": 3.6202, + "step": 32335 + }, + { + "epoch": 2.197309417040359, + "grad_norm": 0.7589321732521057, + "learning_rate": 0.0007253363228699552, + "loss": 3.719, + "step": 32340 + }, + { + "epoch": 2.1976491371110205, + "grad_norm": 0.9884923100471497, + "learning_rate": 0.0007252938578611224, + "loss": 3.7307, + "step": 32345 + }, + { + "epoch": 2.197988857181682, + "grad_norm": 0.8472917079925537, + "learning_rate": 0.0007252513928522897, + "loss": 3.7195, + "step": 32350 + }, + { + "epoch": 2.198328577252344, + "grad_norm": 1.1842713356018066, + "learning_rate": 0.0007252089278434571, + "loss": 3.6106, + "step": 32355 + }, + { + "epoch": 2.198668297323006, + "grad_norm": 0.848441481590271, + "learning_rate": 0.0007251664628346243, + "loss": 3.5959, + "step": 32360 + }, + { + "epoch": 2.1990080173936675, + "grad_norm": 0.7582809925079346, + "learning_rate": 0.0007251239978257915, + "loss": 3.3248, + "step": 32365 + }, + { + "epoch": 2.1993477374643295, + "grad_norm": 1.0703539848327637, + "learning_rate": 0.0007250815328169589, + "loss": 3.2531, + "step": 32370 + }, + { + "epoch": 2.199687457534991, + "grad_norm": 0.7876978516578674, + "learning_rate": 0.0007250390678081261, + "loss": 3.3242, + "step": 32375 + }, + { + "epoch": 2.200027177605653, + "grad_norm": 0.8323952555656433, + "learning_rate": 0.0007249966027992933, + "loss": 3.5724, + "step": 32380 + }, + { + "epoch": 2.200366897676315, + "grad_norm": 0.9582597017288208, + "learning_rate": 0.0007249541377904607, + "loss": 3.5901, + "step": 32385 + }, + { + "epoch": 2.2007066177469765, + "grad_norm": 0.8199208974838257, + "learning_rate": 0.000724911672781628, + "loss": 3.8667, + "step": 32390 + }, + { + "epoch": 2.201046337817638, + "grad_norm": 0.902228057384491, + "learning_rate": 0.0007248692077727952, + "loss": 3.3433, + "step": 32395 + }, + { + "epoch": 2.2013860578883, + "grad_norm": 0.699664294719696, + "learning_rate": 0.0007248267427639626, + "loss": 3.6722, + "step": 32400 + }, + { + "epoch": 2.201725777958962, + "grad_norm": 0.7893538475036621, + "learning_rate": 0.0007247842777551298, + "loss": 3.6817, + "step": 32405 + }, + { + "epoch": 2.2020654980296235, + "grad_norm": 1.2912781238555908, + "learning_rate": 0.000724741812746297, + "loss": 3.5174, + "step": 32410 + }, + { + "epoch": 2.2024052181002856, + "grad_norm": 0.8231721520423889, + "learning_rate": 0.0007246993477374643, + "loss": 3.716, + "step": 32415 + }, + { + "epoch": 2.202744938170947, + "grad_norm": 1.0842039585113525, + "learning_rate": 0.0007246568827286316, + "loss": 3.5368, + "step": 32420 + }, + { + "epoch": 2.203084658241609, + "grad_norm": 1.0638808012008667, + "learning_rate": 0.0007246144177197989, + "loss": 3.6361, + "step": 32425 + }, + { + "epoch": 2.203424378312271, + "grad_norm": 0.9698588252067566, + "learning_rate": 0.0007245719527109662, + "loss": 3.2701, + "step": 32430 + }, + { + "epoch": 2.2037640983829325, + "grad_norm": 2.4767730236053467, + "learning_rate": 0.0007245294877021335, + "loss": 3.5859, + "step": 32435 + }, + { + "epoch": 2.204103818453594, + "grad_norm": 0.7966992855072021, + "learning_rate": 0.0007244870226933007, + "loss": 3.7905, + "step": 32440 + }, + { + "epoch": 2.204443538524256, + "grad_norm": 0.7904607057571411, + "learning_rate": 0.000724444557684468, + "loss": 3.8365, + "step": 32445 + }, + { + "epoch": 2.204783258594918, + "grad_norm": 0.9263324737548828, + "learning_rate": 0.0007244020926756353, + "loss": 3.3993, + "step": 32450 + }, + { + "epoch": 2.2051229786655795, + "grad_norm": 0.8912951946258545, + "learning_rate": 0.0007243596276668025, + "loss": 3.3937, + "step": 32455 + }, + { + "epoch": 2.205462698736241, + "grad_norm": 0.8640767931938171, + "learning_rate": 0.0007243171626579699, + "loss": 3.5114, + "step": 32460 + }, + { + "epoch": 2.205802418806903, + "grad_norm": 0.7920671105384827, + "learning_rate": 0.0007242746976491371, + "loss": 3.6954, + "step": 32465 + }, + { + "epoch": 2.206142138877565, + "grad_norm": 1.273353934288025, + "learning_rate": 0.0007242322326403044, + "loss": 3.6205, + "step": 32470 + }, + { + "epoch": 2.2064818589482265, + "grad_norm": 0.8244420886039734, + "learning_rate": 0.0007241897676314717, + "loss": 3.7842, + "step": 32475 + }, + { + "epoch": 2.2068215790188885, + "grad_norm": 0.9040079116821289, + "learning_rate": 0.0007241473026226389, + "loss": 3.1961, + "step": 32480 + }, + { + "epoch": 2.20716129908955, + "grad_norm": 0.8033730387687683, + "learning_rate": 0.0007241048376138062, + "loss": 3.834, + "step": 32485 + }, + { + "epoch": 2.207501019160212, + "grad_norm": 0.9443610310554504, + "learning_rate": 0.0007240623726049735, + "loss": 3.4635, + "step": 32490 + }, + { + "epoch": 2.207840739230874, + "grad_norm": 0.8103798031806946, + "learning_rate": 0.0007240199075961408, + "loss": 3.7182, + "step": 32495 + }, + { + "epoch": 2.2081804593015355, + "grad_norm": 0.7979253530502319, + "learning_rate": 0.0007239774425873081, + "loss": 3.4059, + "step": 32500 + }, + { + "epoch": 2.208520179372197, + "grad_norm": 0.8763925433158875, + "learning_rate": 0.0007239349775784754, + "loss": 3.5197, + "step": 32505 + }, + { + "epoch": 2.208859899442859, + "grad_norm": 0.9153876304626465, + "learning_rate": 0.0007238925125696426, + "loss": 3.6599, + "step": 32510 + }, + { + "epoch": 2.209199619513521, + "grad_norm": 1.0884946584701538, + "learning_rate": 0.0007238500475608098, + "loss": 3.7522, + "step": 32515 + }, + { + "epoch": 2.2095393395841825, + "grad_norm": 1.2856712341308594, + "learning_rate": 0.0007238075825519772, + "loss": 3.6226, + "step": 32520 + }, + { + "epoch": 2.2098790596548445, + "grad_norm": 0.9621086716651917, + "learning_rate": 0.0007237651175431444, + "loss": 3.7566, + "step": 32525 + }, + { + "epoch": 2.210218779725506, + "grad_norm": 0.8527933955192566, + "learning_rate": 0.0007237226525343117, + "loss": 3.6458, + "step": 32530 + }, + { + "epoch": 2.210558499796168, + "grad_norm": 0.8334488868713379, + "learning_rate": 0.0007236801875254791, + "loss": 3.6017, + "step": 32535 + }, + { + "epoch": 2.21089821986683, + "grad_norm": 0.75522780418396, + "learning_rate": 0.0007236377225166463, + "loss": 3.7891, + "step": 32540 + }, + { + "epoch": 2.2112379399374915, + "grad_norm": 0.922684371471405, + "learning_rate": 0.0007235952575078136, + "loss": 3.5874, + "step": 32545 + }, + { + "epoch": 2.211577660008153, + "grad_norm": 0.7306028604507446, + "learning_rate": 0.0007235527924989809, + "loss": 3.6804, + "step": 32550 + }, + { + "epoch": 2.211917380078815, + "grad_norm": 1.1049672365188599, + "learning_rate": 0.0007235103274901481, + "loss": 3.6485, + "step": 32555 + }, + { + "epoch": 2.212257100149477, + "grad_norm": 0.9597798585891724, + "learning_rate": 0.0007234678624813154, + "loss": 3.5013, + "step": 32560 + }, + { + "epoch": 2.2125968202201385, + "grad_norm": 0.7286372184753418, + "learning_rate": 0.0007234253974724827, + "loss": 3.6512, + "step": 32565 + }, + { + "epoch": 2.2129365402908006, + "grad_norm": 0.932775616645813, + "learning_rate": 0.00072338293246365, + "loss": 3.4226, + "step": 32570 + }, + { + "epoch": 2.213276260361462, + "grad_norm": 1.3865336179733276, + "learning_rate": 0.0007233404674548173, + "loss": 3.4846, + "step": 32575 + }, + { + "epoch": 2.213615980432124, + "grad_norm": 0.8341137766838074, + "learning_rate": 0.0007232980024459845, + "loss": 3.6311, + "step": 32580 + }, + { + "epoch": 2.213955700502786, + "grad_norm": 0.8115186095237732, + "learning_rate": 0.0007232555374371518, + "loss": 3.726, + "step": 32585 + }, + { + "epoch": 2.2142954205734475, + "grad_norm": 0.768846869468689, + "learning_rate": 0.0007232130724283191, + "loss": 3.5069, + "step": 32590 + }, + { + "epoch": 2.214635140644109, + "grad_norm": 0.9354526400566101, + "learning_rate": 0.0007231706074194863, + "loss": 3.3344, + "step": 32595 + }, + { + "epoch": 2.2149748607147712, + "grad_norm": 0.919762134552002, + "learning_rate": 0.0007231281424106537, + "loss": 3.6443, + "step": 32600 + }, + { + "epoch": 2.215314580785433, + "grad_norm": 0.7725939750671387, + "learning_rate": 0.000723085677401821, + "loss": 3.3866, + "step": 32605 + }, + { + "epoch": 2.2156543008560945, + "grad_norm": 0.942682147026062, + "learning_rate": 0.0007230432123929882, + "loss": 3.3307, + "step": 32610 + }, + { + "epoch": 2.2159940209267566, + "grad_norm": 0.8874782919883728, + "learning_rate": 0.0007230007473841554, + "loss": 3.7634, + "step": 32615 + }, + { + "epoch": 2.216333740997418, + "grad_norm": 0.8687102794647217, + "learning_rate": 0.0007229582823753228, + "loss": 3.6211, + "step": 32620 + }, + { + "epoch": 2.21667346106808, + "grad_norm": 0.964253306388855, + "learning_rate": 0.00072291581736649, + "loss": 3.602, + "step": 32625 + }, + { + "epoch": 2.217013181138742, + "grad_norm": 0.878630518913269, + "learning_rate": 0.0007228733523576572, + "loss": 3.4792, + "step": 32630 + }, + { + "epoch": 2.2173529012094035, + "grad_norm": 1.075579047203064, + "learning_rate": 0.0007228308873488247, + "loss": 3.5107, + "step": 32635 + }, + { + "epoch": 2.217692621280065, + "grad_norm": 1.0901236534118652, + "learning_rate": 0.0007227884223399919, + "loss": 3.5373, + "step": 32640 + }, + { + "epoch": 2.218032341350727, + "grad_norm": 0.9886935353279114, + "learning_rate": 0.0007227459573311591, + "loss": 3.569, + "step": 32645 + }, + { + "epoch": 2.218372061421389, + "grad_norm": 0.7487064003944397, + "learning_rate": 0.0007227034923223265, + "loss": 3.3605, + "step": 32650 + }, + { + "epoch": 2.2187117814920505, + "grad_norm": 1.1131982803344727, + "learning_rate": 0.0007226610273134937, + "loss": 3.5597, + "step": 32655 + }, + { + "epoch": 2.219051501562712, + "grad_norm": 1.1309337615966797, + "learning_rate": 0.0007226185623046609, + "loss": 3.503, + "step": 32660 + }, + { + "epoch": 2.219391221633374, + "grad_norm": 0.8000867366790771, + "learning_rate": 0.0007225760972958282, + "loss": 3.2924, + "step": 32665 + }, + { + "epoch": 2.219730941704036, + "grad_norm": 0.887381911277771, + "learning_rate": 0.0007225336322869956, + "loss": 3.6958, + "step": 32670 + }, + { + "epoch": 2.2200706617746975, + "grad_norm": 0.8937597274780273, + "learning_rate": 0.0007224911672781628, + "loss": 3.5569, + "step": 32675 + }, + { + "epoch": 2.2204103818453595, + "grad_norm": 1.062612533569336, + "learning_rate": 0.0007224487022693301, + "loss": 3.3807, + "step": 32680 + }, + { + "epoch": 2.220750101916021, + "grad_norm": 1.657966136932373, + "learning_rate": 0.0007224062372604974, + "loss": 3.541, + "step": 32685 + }, + { + "epoch": 2.221089821986683, + "grad_norm": 1.0473896265029907, + "learning_rate": 0.0007223637722516646, + "loss": 3.3144, + "step": 32690 + }, + { + "epoch": 2.221429542057345, + "grad_norm": 1.1639864444732666, + "learning_rate": 0.0007223213072428319, + "loss": 3.4216, + "step": 32695 + }, + { + "epoch": 2.2217692621280065, + "grad_norm": 0.7584365606307983, + "learning_rate": 0.0007222788422339991, + "loss": 3.5164, + "step": 32700 + }, + { + "epoch": 2.222108982198668, + "grad_norm": 0.7838539481163025, + "learning_rate": 0.0007222363772251665, + "loss": 3.5092, + "step": 32705 + }, + { + "epoch": 2.2224487022693302, + "grad_norm": 0.6943447589874268, + "learning_rate": 0.0007221939122163338, + "loss": 3.6687, + "step": 32710 + }, + { + "epoch": 2.222788422339992, + "grad_norm": 0.7911533117294312, + "learning_rate": 0.000722151447207501, + "loss": 3.6781, + "step": 32715 + }, + { + "epoch": 2.2231281424106535, + "grad_norm": 0.9920641183853149, + "learning_rate": 0.0007221089821986683, + "loss": 3.5283, + "step": 32720 + }, + { + "epoch": 2.2234678624813156, + "grad_norm": 0.8376407623291016, + "learning_rate": 0.0007220665171898356, + "loss": 3.4906, + "step": 32725 + }, + { + "epoch": 2.223807582551977, + "grad_norm": 0.9474460482597351, + "learning_rate": 0.0007220240521810028, + "loss": 3.6072, + "step": 32730 + }, + { + "epoch": 2.224147302622639, + "grad_norm": 0.8528099656105042, + "learning_rate": 0.0007219815871721701, + "loss": 3.649, + "step": 32735 + }, + { + "epoch": 2.224487022693301, + "grad_norm": 0.9030001163482666, + "learning_rate": 0.0007219391221633375, + "loss": 3.675, + "step": 32740 + }, + { + "epoch": 2.2248267427639625, + "grad_norm": 0.9056916832923889, + "learning_rate": 0.0007218966571545047, + "loss": 3.6448, + "step": 32745 + }, + { + "epoch": 2.225166462834624, + "grad_norm": 0.9632193446159363, + "learning_rate": 0.000721854192145672, + "loss": 3.2476, + "step": 32750 + }, + { + "epoch": 2.2255061829052862, + "grad_norm": 1.1059685945510864, + "learning_rate": 0.0007218117271368393, + "loss": 3.5806, + "step": 32755 + }, + { + "epoch": 2.225845902975948, + "grad_norm": 0.7420191764831543, + "learning_rate": 0.0007217692621280065, + "loss": 3.7043, + "step": 32760 + }, + { + "epoch": 2.2261856230466095, + "grad_norm": 0.8974022269248962, + "learning_rate": 0.0007217267971191737, + "loss": 3.4824, + "step": 32765 + }, + { + "epoch": 2.2265253431172716, + "grad_norm": 0.9585962891578674, + "learning_rate": 0.0007216843321103411, + "loss": 3.6685, + "step": 32770 + }, + { + "epoch": 2.226865063187933, + "grad_norm": 1.177390694618225, + "learning_rate": 0.0007216418671015084, + "loss": 3.5998, + "step": 32775 + }, + { + "epoch": 2.227204783258595, + "grad_norm": 0.8217247128486633, + "learning_rate": 0.0007215994020926756, + "loss": 3.641, + "step": 32780 + }, + { + "epoch": 2.2275445033292565, + "grad_norm": 0.9738537073135376, + "learning_rate": 0.000721556937083843, + "loss": 3.5254, + "step": 32785 + }, + { + "epoch": 2.2278842233999185, + "grad_norm": 0.8631089925765991, + "learning_rate": 0.0007215144720750102, + "loss": 3.4482, + "step": 32790 + }, + { + "epoch": 2.22822394347058, + "grad_norm": 0.867121160030365, + "learning_rate": 0.0007214720070661774, + "loss": 3.7201, + "step": 32795 + }, + { + "epoch": 2.228563663541242, + "grad_norm": 0.8319726586341858, + "learning_rate": 0.0007214295420573448, + "loss": 3.7933, + "step": 32800 + }, + { + "epoch": 2.228903383611904, + "grad_norm": 0.8822260499000549, + "learning_rate": 0.000721387077048512, + "loss": 3.5026, + "step": 32805 + }, + { + "epoch": 2.2292431036825655, + "grad_norm": 0.7983752489089966, + "learning_rate": 0.0007213446120396793, + "loss": 3.5965, + "step": 32810 + }, + { + "epoch": 2.229582823753227, + "grad_norm": 0.7111396789550781, + "learning_rate": 0.0007213021470308466, + "loss": 3.6576, + "step": 32815 + }, + { + "epoch": 2.229922543823889, + "grad_norm": 1.037390947341919, + "learning_rate": 0.0007212596820220139, + "loss": 3.6479, + "step": 32820 + }, + { + "epoch": 2.230262263894551, + "grad_norm": 0.7104963660240173, + "learning_rate": 0.0007212172170131811, + "loss": 3.6445, + "step": 32825 + }, + { + "epoch": 2.2306019839652125, + "grad_norm": 1.064030408859253, + "learning_rate": 0.0007211747520043484, + "loss": 3.8064, + "step": 32830 + }, + { + "epoch": 2.2309417040358746, + "grad_norm": 1.0181405544281006, + "learning_rate": 0.0007211322869955157, + "loss": 3.62, + "step": 32835 + }, + { + "epoch": 2.231281424106536, + "grad_norm": 1.0560693740844727, + "learning_rate": 0.0007210898219866829, + "loss": 3.5056, + "step": 32840 + }, + { + "epoch": 2.231621144177198, + "grad_norm": 1.155771255493164, + "learning_rate": 0.0007210473569778503, + "loss": 3.5113, + "step": 32845 + }, + { + "epoch": 2.23196086424786, + "grad_norm": 0.903671383857727, + "learning_rate": 0.0007210048919690176, + "loss": 3.5671, + "step": 32850 + }, + { + "epoch": 2.2323005843185215, + "grad_norm": 0.8087148070335388, + "learning_rate": 0.0007209624269601848, + "loss": 3.4991, + "step": 32855 + }, + { + "epoch": 2.232640304389183, + "grad_norm": 0.9262293577194214, + "learning_rate": 0.0007209199619513521, + "loss": 3.3508, + "step": 32860 + }, + { + "epoch": 2.2329800244598452, + "grad_norm": 1.0619874000549316, + "learning_rate": 0.0007208774969425193, + "loss": 3.288, + "step": 32865 + }, + { + "epoch": 2.233319744530507, + "grad_norm": 0.7767520546913147, + "learning_rate": 0.0007208350319336866, + "loss": 3.756, + "step": 32870 + }, + { + "epoch": 2.2336594646011685, + "grad_norm": 0.9351972937583923, + "learning_rate": 0.000720792566924854, + "loss": 3.7703, + "step": 32875 + }, + { + "epoch": 2.2339991846718306, + "grad_norm": 0.8207551836967468, + "learning_rate": 0.0007207501019160212, + "loss": 3.5417, + "step": 32880 + }, + { + "epoch": 2.234338904742492, + "grad_norm": 0.8272174596786499, + "learning_rate": 0.0007207076369071886, + "loss": 3.4901, + "step": 32885 + }, + { + "epoch": 2.234678624813154, + "grad_norm": 1.8387123346328735, + "learning_rate": 0.0007206651718983558, + "loss": 3.558, + "step": 32890 + }, + { + "epoch": 2.235018344883816, + "grad_norm": 1.307273507118225, + "learning_rate": 0.000720622706889523, + "loss": 3.5348, + "step": 32895 + }, + { + "epoch": 2.2353580649544775, + "grad_norm": 0.8822222352027893, + "learning_rate": 0.0007205802418806904, + "loss": 3.5816, + "step": 32900 + }, + { + "epoch": 2.235697785025139, + "grad_norm": 0.8231761455535889, + "learning_rate": 0.0007205377768718576, + "loss": 3.523, + "step": 32905 + }, + { + "epoch": 2.2360375050958012, + "grad_norm": 0.8112620711326599, + "learning_rate": 0.0007204953118630249, + "loss": 3.8385, + "step": 32910 + }, + { + "epoch": 2.236377225166463, + "grad_norm": 1.0755014419555664, + "learning_rate": 0.0007204528468541922, + "loss": 3.7934, + "step": 32915 + }, + { + "epoch": 2.2367169452371245, + "grad_norm": 0.8414238095283508, + "learning_rate": 0.0007204103818453595, + "loss": 3.464, + "step": 32920 + }, + { + "epoch": 2.2370566653077866, + "grad_norm": 0.8555359840393066, + "learning_rate": 0.0007203679168365267, + "loss": 3.6465, + "step": 32925 + }, + { + "epoch": 2.237396385378448, + "grad_norm": 0.8031982183456421, + "learning_rate": 0.000720325451827694, + "loss": 3.8341, + "step": 32930 + }, + { + "epoch": 2.23773610544911, + "grad_norm": 0.9278504252433777, + "learning_rate": 0.0007202829868188613, + "loss": 3.3356, + "step": 32935 + }, + { + "epoch": 2.238075825519772, + "grad_norm": 1.064771294593811, + "learning_rate": 0.0007202405218100285, + "loss": 3.3553, + "step": 32940 + }, + { + "epoch": 2.2384155455904335, + "grad_norm": 2.3802034854888916, + "learning_rate": 0.0007201980568011959, + "loss": 3.6065, + "step": 32945 + }, + { + "epoch": 2.238755265661095, + "grad_norm": 0.7531465291976929, + "learning_rate": 0.0007201555917923632, + "loss": 3.7289, + "step": 32950 + }, + { + "epoch": 2.2390949857317572, + "grad_norm": 0.7438278794288635, + "learning_rate": 0.0007201131267835304, + "loss": 3.5987, + "step": 32955 + }, + { + "epoch": 2.239434705802419, + "grad_norm": 0.7301222681999207, + "learning_rate": 0.0007200706617746977, + "loss": 3.5205, + "step": 32960 + }, + { + "epoch": 2.2397744258730805, + "grad_norm": 1.117942214012146, + "learning_rate": 0.0007200281967658649, + "loss": 3.6124, + "step": 32965 + }, + { + "epoch": 2.2401141459437426, + "grad_norm": 0.9310340285301208, + "learning_rate": 0.0007199857317570322, + "loss": 3.6187, + "step": 32970 + }, + { + "epoch": 2.240453866014404, + "grad_norm": 1.2055370807647705, + "learning_rate": 0.0007199432667481995, + "loss": 3.5309, + "step": 32975 + }, + { + "epoch": 2.240793586085066, + "grad_norm": 0.7803168892860413, + "learning_rate": 0.0007199008017393668, + "loss": 3.6, + "step": 32980 + }, + { + "epoch": 2.2411333061557275, + "grad_norm": 1.1834999322891235, + "learning_rate": 0.0007198583367305341, + "loss": 3.5526, + "step": 32985 + }, + { + "epoch": 2.2414730262263896, + "grad_norm": 0.957045316696167, + "learning_rate": 0.0007198158717217014, + "loss": 3.4927, + "step": 32990 + }, + { + "epoch": 2.241812746297051, + "grad_norm": 1.13032865524292, + "learning_rate": 0.0007197734067128686, + "loss": 3.5098, + "step": 32995 + }, + { + "epoch": 2.242152466367713, + "grad_norm": 0.8925689458847046, + "learning_rate": 0.0007197309417040358, + "loss": 3.484, + "step": 33000 + }, + { + "epoch": 2.242492186438375, + "grad_norm": 0.8652591705322266, + "learning_rate": 0.0007196884766952032, + "loss": 3.581, + "step": 33005 + }, + { + "epoch": 2.2428319065090365, + "grad_norm": 0.8590790629386902, + "learning_rate": 0.0007196460116863704, + "loss": 3.255, + "step": 33010 + }, + { + "epoch": 2.243171626579698, + "grad_norm": 1.02212655544281, + "learning_rate": 0.0007196035466775377, + "loss": 3.7149, + "step": 33015 + }, + { + "epoch": 2.2435113466503602, + "grad_norm": 0.8495359420776367, + "learning_rate": 0.0007195610816687051, + "loss": 3.5893, + "step": 33020 + }, + { + "epoch": 2.243851066721022, + "grad_norm": 0.7727569341659546, + "learning_rate": 0.0007195186166598723, + "loss": 3.5309, + "step": 33025 + }, + { + "epoch": 2.2441907867916835, + "grad_norm": 0.8089480400085449, + "learning_rate": 0.0007194761516510395, + "loss": 3.581, + "step": 33030 + }, + { + "epoch": 2.2445305068623456, + "grad_norm": 0.7776786684989929, + "learning_rate": 0.0007194336866422069, + "loss": 3.4994, + "step": 33035 + }, + { + "epoch": 2.244870226933007, + "grad_norm": 0.9157167673110962, + "learning_rate": 0.0007193912216333741, + "loss": 3.5078, + "step": 33040 + }, + { + "epoch": 2.245209947003669, + "grad_norm": 0.8186530470848083, + "learning_rate": 0.0007193487566245413, + "loss": 3.6267, + "step": 33045 + }, + { + "epoch": 2.245549667074331, + "grad_norm": 0.8314428329467773, + "learning_rate": 0.0007193062916157088, + "loss": 3.9037, + "step": 33050 + }, + { + "epoch": 2.2458893871449925, + "grad_norm": 0.992611825466156, + "learning_rate": 0.000719263826606876, + "loss": 3.8596, + "step": 33055 + }, + { + "epoch": 2.246229107215654, + "grad_norm": 0.7888532876968384, + "learning_rate": 0.0007192213615980432, + "loss": 3.5228, + "step": 33060 + }, + { + "epoch": 2.2465688272863162, + "grad_norm": 0.9443208575248718, + "learning_rate": 0.0007191788965892105, + "loss": 3.6179, + "step": 33065 + }, + { + "epoch": 2.246908547356978, + "grad_norm": 1.0666089057922363, + "learning_rate": 0.0007191364315803778, + "loss": 3.58, + "step": 33070 + }, + { + "epoch": 2.2472482674276395, + "grad_norm": 0.8038224577903748, + "learning_rate": 0.000719093966571545, + "loss": 3.7215, + "step": 33075 + }, + { + "epoch": 2.2475879874983016, + "grad_norm": 0.8206089735031128, + "learning_rate": 0.0007190515015627123, + "loss": 3.5763, + "step": 33080 + }, + { + "epoch": 2.247927707568963, + "grad_norm": 1.1060070991516113, + "learning_rate": 0.0007190090365538797, + "loss": 3.5541, + "step": 33085 + }, + { + "epoch": 2.248267427639625, + "grad_norm": 0.7552579045295715, + "learning_rate": 0.0007189665715450469, + "loss": 3.4304, + "step": 33090 + }, + { + "epoch": 2.248607147710287, + "grad_norm": 0.8153659701347351, + "learning_rate": 0.0007189241065362142, + "loss": 3.277, + "step": 33095 + }, + { + "epoch": 2.2489468677809485, + "grad_norm": 0.8223702907562256, + "learning_rate": 0.0007188816415273814, + "loss": 3.4822, + "step": 33100 + }, + { + "epoch": 2.24928658785161, + "grad_norm": 0.7493183016777039, + "learning_rate": 0.0007188391765185487, + "loss": 3.3604, + "step": 33105 + }, + { + "epoch": 2.2496263079222723, + "grad_norm": 0.8348582983016968, + "learning_rate": 0.000718796711509716, + "loss": 3.4184, + "step": 33110 + }, + { + "epoch": 2.249966027992934, + "grad_norm": 0.6842407584190369, + "learning_rate": 0.0007187542465008832, + "loss": 3.6876, + "step": 33115 + }, + { + "epoch": 2.2503057480635955, + "grad_norm": 0.9893853664398193, + "learning_rate": 0.0007187117814920506, + "loss": 3.5327, + "step": 33120 + }, + { + "epoch": 2.250645468134257, + "grad_norm": 0.932339072227478, + "learning_rate": 0.0007186693164832179, + "loss": 3.7372, + "step": 33125 + }, + { + "epoch": 2.250985188204919, + "grad_norm": 0.8852616548538208, + "learning_rate": 0.0007186268514743851, + "loss": 3.7989, + "step": 33130 + }, + { + "epoch": 2.251324908275581, + "grad_norm": 0.7991600036621094, + "learning_rate": 0.0007185843864655524, + "loss": 3.683, + "step": 33135 + }, + { + "epoch": 2.2516646283462425, + "grad_norm": 0.8995228409767151, + "learning_rate": 0.0007185419214567197, + "loss": 3.6304, + "step": 33140 + }, + { + "epoch": 2.2520043484169046, + "grad_norm": 0.9056110382080078, + "learning_rate": 0.0007184994564478869, + "loss": 3.5715, + "step": 33145 + }, + { + "epoch": 2.252344068487566, + "grad_norm": 1.0736966133117676, + "learning_rate": 0.0007184569914390541, + "loss": 3.3348, + "step": 33150 + }, + { + "epoch": 2.252683788558228, + "grad_norm": 0.8796764612197876, + "learning_rate": 0.0007184145264302216, + "loss": 3.8133, + "step": 33155 + }, + { + "epoch": 2.25302350862889, + "grad_norm": 0.7989653944969177, + "learning_rate": 0.0007183720614213888, + "loss": 3.6325, + "step": 33160 + }, + { + "epoch": 2.2533632286995515, + "grad_norm": 0.7202712297439575, + "learning_rate": 0.000718329596412556, + "loss": 3.7224, + "step": 33165 + }, + { + "epoch": 2.253702948770213, + "grad_norm": 1.2048949003219604, + "learning_rate": 0.0007182871314037234, + "loss": 3.4293, + "step": 33170 + }, + { + "epoch": 2.2540426688408752, + "grad_norm": 0.9079889059066772, + "learning_rate": 0.0007182446663948906, + "loss": 3.4945, + "step": 33175 + }, + { + "epoch": 2.254382388911537, + "grad_norm": 0.8293566703796387, + "learning_rate": 0.0007182022013860578, + "loss": 3.3822, + "step": 33180 + }, + { + "epoch": 2.2547221089821985, + "grad_norm": 0.8465235233306885, + "learning_rate": 0.0007181597363772252, + "loss": 3.6184, + "step": 33185 + }, + { + "epoch": 2.2550618290528606, + "grad_norm": 0.8670704960823059, + "learning_rate": 0.0007181172713683925, + "loss": 3.5337, + "step": 33190 + }, + { + "epoch": 2.255401549123522, + "grad_norm": 1.127526044845581, + "learning_rate": 0.0007180748063595597, + "loss": 3.7075, + "step": 33195 + }, + { + "epoch": 2.255741269194184, + "grad_norm": 0.7312422394752502, + "learning_rate": 0.000718032341350727, + "loss": 3.5978, + "step": 33200 + }, + { + "epoch": 2.256080989264846, + "grad_norm": 0.7384932637214661, + "learning_rate": 0.0007179898763418943, + "loss": 3.6336, + "step": 33205 + }, + { + "epoch": 2.2564207093355075, + "grad_norm": 0.7124952673912048, + "learning_rate": 0.0007179474113330615, + "loss": 3.6529, + "step": 33210 + }, + { + "epoch": 2.256760429406169, + "grad_norm": 0.8823535442352295, + "learning_rate": 0.0007179049463242288, + "loss": 3.4039, + "step": 33215 + }, + { + "epoch": 2.2571001494768312, + "grad_norm": 0.9871249794960022, + "learning_rate": 0.0007178624813153961, + "loss": 3.7726, + "step": 33220 + }, + { + "epoch": 2.257439869547493, + "grad_norm": 1.136189579963684, + "learning_rate": 0.0007178200163065635, + "loss": 3.3842, + "step": 33225 + }, + { + "epoch": 2.2577795896181545, + "grad_norm": 1.0106223821640015, + "learning_rate": 0.0007177775512977307, + "loss": 3.5458, + "step": 33230 + }, + { + "epoch": 2.2581193096888166, + "grad_norm": 0.8007284998893738, + "learning_rate": 0.000717735086288898, + "loss": 3.5947, + "step": 33235 + }, + { + "epoch": 2.258459029759478, + "grad_norm": 0.9360009431838989, + "learning_rate": 0.0007176926212800653, + "loss": 3.5269, + "step": 33240 + }, + { + "epoch": 2.25879874983014, + "grad_norm": 1.0695573091506958, + "learning_rate": 0.0007176501562712325, + "loss": 3.3501, + "step": 33245 + }, + { + "epoch": 2.259138469900802, + "grad_norm": 0.8601471185684204, + "learning_rate": 0.0007176076912623997, + "loss": 3.6174, + "step": 33250 + }, + { + "epoch": 2.2594781899714635, + "grad_norm": 0.7691159248352051, + "learning_rate": 0.0007175652262535671, + "loss": 3.398, + "step": 33255 + }, + { + "epoch": 2.259817910042125, + "grad_norm": 0.6757928729057312, + "learning_rate": 0.0007175227612447344, + "loss": 3.7447, + "step": 33260 + }, + { + "epoch": 2.2601576301127873, + "grad_norm": 1.0051745176315308, + "learning_rate": 0.0007174802962359016, + "loss": 3.7604, + "step": 33265 + }, + { + "epoch": 2.260497350183449, + "grad_norm": 0.8519276976585388, + "learning_rate": 0.000717437831227069, + "loss": 3.4593, + "step": 33270 + }, + { + "epoch": 2.2608370702541105, + "grad_norm": 0.9402484893798828, + "learning_rate": 0.0007173953662182362, + "loss": 3.5249, + "step": 33275 + }, + { + "epoch": 2.2611767903247726, + "grad_norm": 0.9661968350410461, + "learning_rate": 0.0007173529012094034, + "loss": 3.5901, + "step": 33280 + }, + { + "epoch": 2.261516510395434, + "grad_norm": 1.0839349031448364, + "learning_rate": 0.0007173104362005708, + "loss": 3.6345, + "step": 33285 + }, + { + "epoch": 2.261856230466096, + "grad_norm": 0.6857048869132996, + "learning_rate": 0.000717267971191738, + "loss": 3.5202, + "step": 33290 + }, + { + "epoch": 2.262195950536758, + "grad_norm": 0.9796864986419678, + "learning_rate": 0.0007172255061829053, + "loss": 3.625, + "step": 33295 + }, + { + "epoch": 2.2625356706074196, + "grad_norm": 0.7852342128753662, + "learning_rate": 0.0007171830411740726, + "loss": 3.5383, + "step": 33300 + }, + { + "epoch": 2.262875390678081, + "grad_norm": 0.8394941091537476, + "learning_rate": 0.0007171405761652399, + "loss": 3.4653, + "step": 33305 + }, + { + "epoch": 2.2632151107487433, + "grad_norm": 1.0016043186187744, + "learning_rate": 0.0007170981111564071, + "loss": 3.5829, + "step": 33310 + }, + { + "epoch": 2.263554830819405, + "grad_norm": 0.8145662546157837, + "learning_rate": 0.0007170556461475744, + "loss": 3.7781, + "step": 33315 + }, + { + "epoch": 2.2638945508900665, + "grad_norm": 0.8144311904907227, + "learning_rate": 0.0007170131811387417, + "loss": 3.5867, + "step": 33320 + }, + { + "epoch": 2.2642342709607286, + "grad_norm": 0.8240835666656494, + "learning_rate": 0.0007169707161299089, + "loss": 3.4374, + "step": 33325 + }, + { + "epoch": 2.2645739910313902, + "grad_norm": 1.0463414192199707, + "learning_rate": 0.0007169282511210763, + "loss": 3.5152, + "step": 33330 + }, + { + "epoch": 2.264913711102052, + "grad_norm": 0.7924541234970093, + "learning_rate": 0.0007168857861122436, + "loss": 3.5487, + "step": 33335 + }, + { + "epoch": 2.2652534311727135, + "grad_norm": 0.7810038328170776, + "learning_rate": 0.0007168433211034108, + "loss": 3.514, + "step": 33340 + }, + { + "epoch": 2.2655931512433756, + "grad_norm": 0.7996058464050293, + "learning_rate": 0.0007168008560945781, + "loss": 3.4942, + "step": 33345 + }, + { + "epoch": 2.265932871314037, + "grad_norm": 0.9138662815093994, + "learning_rate": 0.0007167583910857453, + "loss": 3.7177, + "step": 33350 + }, + { + "epoch": 2.266272591384699, + "grad_norm": 0.9390940070152283, + "learning_rate": 0.0007167159260769126, + "loss": 3.716, + "step": 33355 + }, + { + "epoch": 2.266612311455361, + "grad_norm": 0.6953398585319519, + "learning_rate": 0.00071667346106808, + "loss": 3.754, + "step": 33360 + }, + { + "epoch": 2.2669520315260225, + "grad_norm": 0.8761784434318542, + "learning_rate": 0.0007166309960592472, + "loss": 3.6092, + "step": 33365 + }, + { + "epoch": 2.267291751596684, + "grad_norm": 0.7568075656890869, + "learning_rate": 0.0007165885310504145, + "loss": 3.6632, + "step": 33370 + }, + { + "epoch": 2.2676314716673462, + "grad_norm": 0.901160717010498, + "learning_rate": 0.0007165460660415818, + "loss": 3.2737, + "step": 33375 + }, + { + "epoch": 2.267971191738008, + "grad_norm": 0.7893183827400208, + "learning_rate": 0.000716503601032749, + "loss": 3.744, + "step": 33380 + }, + { + "epoch": 2.2683109118086695, + "grad_norm": 0.7973024249076843, + "learning_rate": 0.0007164611360239162, + "loss": 3.3238, + "step": 33385 + }, + { + "epoch": 2.2686506318793316, + "grad_norm": 0.9344809651374817, + "learning_rate": 0.0007164186710150836, + "loss": 3.5552, + "step": 33390 + }, + { + "epoch": 2.268990351949993, + "grad_norm": 0.8530999422073364, + "learning_rate": 0.0007163762060062509, + "loss": 3.5628, + "step": 33395 + }, + { + "epoch": 2.269330072020655, + "grad_norm": 0.9041158556938171, + "learning_rate": 0.0007163337409974181, + "loss": 3.3898, + "step": 33400 + }, + { + "epoch": 2.269669792091317, + "grad_norm": 1.436303734779358, + "learning_rate": 0.0007162912759885855, + "loss": 3.3977, + "step": 33405 + }, + { + "epoch": 2.2700095121619785, + "grad_norm": 0.8674291968345642, + "learning_rate": 0.0007162488109797527, + "loss": 3.4566, + "step": 33410 + }, + { + "epoch": 2.27034923223264, + "grad_norm": 0.8038323521614075, + "learning_rate": 0.0007162063459709199, + "loss": 3.6008, + "step": 33415 + }, + { + "epoch": 2.2706889523033023, + "grad_norm": 0.8486367464065552, + "learning_rate": 0.0007161638809620873, + "loss": 3.6199, + "step": 33420 + }, + { + "epoch": 2.271028672373964, + "grad_norm": 4.3038740158081055, + "learning_rate": 0.0007161214159532545, + "loss": 3.3495, + "step": 33425 + }, + { + "epoch": 2.2713683924446255, + "grad_norm": 0.9377936720848083, + "learning_rate": 0.0007160789509444218, + "loss": 3.4689, + "step": 33430 + }, + { + "epoch": 2.2717081125152876, + "grad_norm": 0.6448661684989929, + "learning_rate": 0.0007160364859355892, + "loss": 3.3967, + "step": 33435 + }, + { + "epoch": 2.2720478325859492, + "grad_norm": 0.7574926018714905, + "learning_rate": 0.0007159940209267564, + "loss": 3.434, + "step": 33440 + }, + { + "epoch": 2.272387552656611, + "grad_norm": 0.7551472783088684, + "learning_rate": 0.0007159515559179236, + "loss": 3.5716, + "step": 33445 + }, + { + "epoch": 2.2727272727272725, + "grad_norm": 0.9378146529197693, + "learning_rate": 0.0007159090909090909, + "loss": 3.3339, + "step": 33450 + }, + { + "epoch": 2.2730669927979346, + "grad_norm": 1.1030871868133545, + "learning_rate": 0.0007158666259002582, + "loss": 3.6648, + "step": 33455 + }, + { + "epoch": 2.273406712868596, + "grad_norm": 1.0810635089874268, + "learning_rate": 0.0007158241608914254, + "loss": 3.5567, + "step": 33460 + }, + { + "epoch": 2.273746432939258, + "grad_norm": 0.928423285484314, + "learning_rate": 0.0007157816958825928, + "loss": 3.4463, + "step": 33465 + }, + { + "epoch": 2.27408615300992, + "grad_norm": 1.1333566904067993, + "learning_rate": 0.0007157392308737601, + "loss": 3.4522, + "step": 33470 + }, + { + "epoch": 2.2744258730805815, + "grad_norm": 0.7303792238235474, + "learning_rate": 0.0007156967658649273, + "loss": 3.6039, + "step": 33475 + }, + { + "epoch": 2.274765593151243, + "grad_norm": 0.9278403520584106, + "learning_rate": 0.0007156543008560946, + "loss": 3.5457, + "step": 33480 + }, + { + "epoch": 2.2751053132219052, + "grad_norm": 0.8834285736083984, + "learning_rate": 0.0007156118358472619, + "loss": 3.5495, + "step": 33485 + }, + { + "epoch": 2.275445033292567, + "grad_norm": 0.7755470275878906, + "learning_rate": 0.0007155693708384291, + "loss": 3.7187, + "step": 33490 + }, + { + "epoch": 2.2757847533632285, + "grad_norm": 0.9774354696273804, + "learning_rate": 0.0007155269058295964, + "loss": 3.6987, + "step": 33495 + }, + { + "epoch": 2.2761244734338906, + "grad_norm": 0.729327380657196, + "learning_rate": 0.0007154844408207637, + "loss": 3.7649, + "step": 33500 + }, + { + "epoch": 2.276464193504552, + "grad_norm": 0.7892672419548035, + "learning_rate": 0.000715441975811931, + "loss": 3.8116, + "step": 33505 + }, + { + "epoch": 2.276803913575214, + "grad_norm": 0.8903812766075134, + "learning_rate": 0.0007153995108030983, + "loss": 3.3989, + "step": 33510 + }, + { + "epoch": 2.277143633645876, + "grad_norm": 0.9278993606567383, + "learning_rate": 0.0007153570457942655, + "loss": 3.625, + "step": 33515 + }, + { + "epoch": 2.2774833537165375, + "grad_norm": 0.9745951890945435, + "learning_rate": 0.0007153145807854328, + "loss": 3.5104, + "step": 33520 + }, + { + "epoch": 2.277823073787199, + "grad_norm": 0.8301883339881897, + "learning_rate": 0.0007152721157766001, + "loss": 3.5483, + "step": 33525 + }, + { + "epoch": 2.2781627938578612, + "grad_norm": 0.8892806172370911, + "learning_rate": 0.0007152296507677673, + "loss": 3.5899, + "step": 33530 + }, + { + "epoch": 2.278502513928523, + "grad_norm": 0.8409011363983154, + "learning_rate": 0.0007151871857589347, + "loss": 3.4398, + "step": 33535 + }, + { + "epoch": 2.2788422339991845, + "grad_norm": 4.0718536376953125, + "learning_rate": 0.000715144720750102, + "loss": 3.8608, + "step": 33540 + }, + { + "epoch": 2.2791819540698466, + "grad_norm": 0.943730890750885, + "learning_rate": 0.0007151022557412692, + "loss": 3.4986, + "step": 33545 + }, + { + "epoch": 2.279521674140508, + "grad_norm": 0.8227397799491882, + "learning_rate": 0.0007150597907324364, + "loss": 3.5182, + "step": 33550 + }, + { + "epoch": 2.27986139421117, + "grad_norm": 1.0455341339111328, + "learning_rate": 0.0007150173257236038, + "loss": 3.4311, + "step": 33555 + }, + { + "epoch": 2.280201114281832, + "grad_norm": 0.9557324647903442, + "learning_rate": 0.000714974860714771, + "loss": 3.4792, + "step": 33560 + }, + { + "epoch": 2.2805408343524936, + "grad_norm": 0.746530294418335, + "learning_rate": 0.0007149323957059383, + "loss": 3.4188, + "step": 33565 + }, + { + "epoch": 2.280880554423155, + "grad_norm": 0.6280292868614197, + "learning_rate": 0.0007148899306971057, + "loss": 3.7707, + "step": 33570 + }, + { + "epoch": 2.2812202744938173, + "grad_norm": 0.8083510994911194, + "learning_rate": 0.0007148474656882729, + "loss": 3.4646, + "step": 33575 + }, + { + "epoch": 2.281559994564479, + "grad_norm": 1.2106924057006836, + "learning_rate": 0.0007148050006794402, + "loss": 3.3421, + "step": 33580 + }, + { + "epoch": 2.2818997146351405, + "grad_norm": 0.7883691787719727, + "learning_rate": 0.0007147625356706075, + "loss": 3.5916, + "step": 33585 + }, + { + "epoch": 2.2822394347058026, + "grad_norm": 0.948059618473053, + "learning_rate": 0.0007147200706617747, + "loss": 3.4732, + "step": 33590 + }, + { + "epoch": 2.2825791547764642, + "grad_norm": 1.2449692487716675, + "learning_rate": 0.000714677605652942, + "loss": 3.4399, + "step": 33595 + }, + { + "epoch": 2.282918874847126, + "grad_norm": 0.8142520785331726, + "learning_rate": 0.0007146351406441092, + "loss": 3.6643, + "step": 33600 + }, + { + "epoch": 2.283258594917788, + "grad_norm": 1.11040461063385, + "learning_rate": 0.0007145926756352766, + "loss": 3.4133, + "step": 33605 + }, + { + "epoch": 2.2835983149884496, + "grad_norm": 1.0723345279693604, + "learning_rate": 0.0007145502106264439, + "loss": 3.57, + "step": 33610 + }, + { + "epoch": 2.283938035059111, + "grad_norm": 0.8764246702194214, + "learning_rate": 0.0007145077456176111, + "loss": 3.4837, + "step": 33615 + }, + { + "epoch": 2.2842777551297733, + "grad_norm": 0.8795246481895447, + "learning_rate": 0.0007144652806087784, + "loss": 3.5893, + "step": 33620 + }, + { + "epoch": 2.284617475200435, + "grad_norm": 0.8362054824829102, + "learning_rate": 0.0007144228155999457, + "loss": 3.8411, + "step": 33625 + }, + { + "epoch": 2.2849571952710965, + "grad_norm": 0.9274553060531616, + "learning_rate": 0.0007143803505911129, + "loss": 3.4455, + "step": 33630 + }, + { + "epoch": 2.2852969153417586, + "grad_norm": 0.9331597089767456, + "learning_rate": 0.0007143378855822801, + "loss": 3.5365, + "step": 33635 + }, + { + "epoch": 2.2856366354124202, + "grad_norm": 0.8610716462135315, + "learning_rate": 0.0007142954205734476, + "loss": 3.8021, + "step": 33640 + }, + { + "epoch": 2.285976355483082, + "grad_norm": 1.0017797946929932, + "learning_rate": 0.0007142529555646148, + "loss": 3.6703, + "step": 33645 + }, + { + "epoch": 2.286316075553744, + "grad_norm": 0.9519618153572083, + "learning_rate": 0.000714210490555782, + "loss": 3.5385, + "step": 33650 + }, + { + "epoch": 2.2866557956244056, + "grad_norm": 0.9531429409980774, + "learning_rate": 0.0007141680255469494, + "loss": 3.5308, + "step": 33655 + }, + { + "epoch": 2.286995515695067, + "grad_norm": 0.7119864225387573, + "learning_rate": 0.0007141255605381166, + "loss": 3.6731, + "step": 33660 + }, + { + "epoch": 2.2873352357657293, + "grad_norm": 0.7924984097480774, + "learning_rate": 0.0007140830955292838, + "loss": 3.4853, + "step": 33665 + }, + { + "epoch": 2.287674955836391, + "grad_norm": 0.8567978143692017, + "learning_rate": 0.0007140406305204512, + "loss": 3.5347, + "step": 33670 + }, + { + "epoch": 2.2880146759070525, + "grad_norm": 0.9263978600502014, + "learning_rate": 0.0007139981655116185, + "loss": 3.3028, + "step": 33675 + }, + { + "epoch": 2.288354395977714, + "grad_norm": 0.9461192488670349, + "learning_rate": 0.0007139557005027857, + "loss": 3.5105, + "step": 33680 + }, + { + "epoch": 2.2886941160483762, + "grad_norm": 0.9066718816757202, + "learning_rate": 0.000713913235493953, + "loss": 3.5792, + "step": 33685 + }, + { + "epoch": 2.289033836119038, + "grad_norm": 0.8691078424453735, + "learning_rate": 0.0007138707704851203, + "loss": 3.6218, + "step": 33690 + }, + { + "epoch": 2.2893735561896995, + "grad_norm": 0.8015620708465576, + "learning_rate": 0.0007138283054762875, + "loss": 3.6117, + "step": 33695 + }, + { + "epoch": 2.2897132762603616, + "grad_norm": 0.8820092082023621, + "learning_rate": 0.0007137858404674548, + "loss": 3.7952, + "step": 33700 + }, + { + "epoch": 2.290052996331023, + "grad_norm": 0.7063429355621338, + "learning_rate": 0.0007137433754586221, + "loss": 3.5758, + "step": 33705 + }, + { + "epoch": 2.290392716401685, + "grad_norm": 1.047690510749817, + "learning_rate": 0.0007137009104497894, + "loss": 3.3326, + "step": 33710 + }, + { + "epoch": 2.290732436472347, + "grad_norm": 0.9483758211135864, + "learning_rate": 0.0007136584454409567, + "loss": 3.4372, + "step": 33715 + }, + { + "epoch": 2.2910721565430086, + "grad_norm": 0.7899272441864014, + "learning_rate": 0.000713615980432124, + "loss": 3.4505, + "step": 33720 + }, + { + "epoch": 2.29141187661367, + "grad_norm": 0.7726660966873169, + "learning_rate": 0.0007135735154232912, + "loss": 3.5856, + "step": 33725 + }, + { + "epoch": 2.2917515966843323, + "grad_norm": 0.9461966156959534, + "learning_rate": 0.0007135310504144585, + "loss": 3.8457, + "step": 33730 + }, + { + "epoch": 2.292091316754994, + "grad_norm": 0.8097044229507446, + "learning_rate": 0.0007134885854056257, + "loss": 3.4337, + "step": 33735 + }, + { + "epoch": 2.2924310368256555, + "grad_norm": 0.8588869571685791, + "learning_rate": 0.000713446120396793, + "loss": 3.5262, + "step": 33740 + }, + { + "epoch": 2.2927707568963176, + "grad_norm": 0.8851317763328552, + "learning_rate": 0.0007134036553879604, + "loss": 3.4683, + "step": 33745 + }, + { + "epoch": 2.2931104769669792, + "grad_norm": 0.8512869477272034, + "learning_rate": 0.0007133611903791276, + "loss": 3.669, + "step": 33750 + }, + { + "epoch": 2.293450197037641, + "grad_norm": 0.849254846572876, + "learning_rate": 0.0007133187253702949, + "loss": 3.6922, + "step": 33755 + }, + { + "epoch": 2.293789917108303, + "grad_norm": 0.6882972717285156, + "learning_rate": 0.0007132762603614622, + "loss": 3.3624, + "step": 33760 + }, + { + "epoch": 2.2941296371789646, + "grad_norm": 0.7374040484428406, + "learning_rate": 0.0007132337953526294, + "loss": 3.2656, + "step": 33765 + }, + { + "epoch": 2.294469357249626, + "grad_norm": 0.8655424118041992, + "learning_rate": 0.0007131913303437967, + "loss": 3.691, + "step": 33770 + }, + { + "epoch": 2.2948090773202883, + "grad_norm": 0.8615302443504333, + "learning_rate": 0.000713148865334964, + "loss": 3.5827, + "step": 33775 + }, + { + "epoch": 2.29514879739095, + "grad_norm": 0.8697820901870728, + "learning_rate": 0.0007131064003261313, + "loss": 3.5313, + "step": 33780 + }, + { + "epoch": 2.2954885174616115, + "grad_norm": 0.9026974439620972, + "learning_rate": 0.0007130639353172985, + "loss": 3.6302, + "step": 33785 + }, + { + "epoch": 2.295828237532273, + "grad_norm": 1.0698466300964355, + "learning_rate": 0.0007130214703084659, + "loss": 3.5132, + "step": 33790 + }, + { + "epoch": 2.2961679576029352, + "grad_norm": 1.1608260869979858, + "learning_rate": 0.0007129790052996331, + "loss": 3.608, + "step": 33795 + }, + { + "epoch": 2.296507677673597, + "grad_norm": 0.7514434456825256, + "learning_rate": 0.0007129365402908003, + "loss": 3.6538, + "step": 33800 + }, + { + "epoch": 2.2968473977442585, + "grad_norm": 1.1244678497314453, + "learning_rate": 0.0007128940752819677, + "loss": 3.5982, + "step": 33805 + }, + { + "epoch": 2.2971871178149206, + "grad_norm": 1.0733412504196167, + "learning_rate": 0.0007128516102731349, + "loss": 3.4568, + "step": 33810 + }, + { + "epoch": 2.297526837885582, + "grad_norm": 0.7152422666549683, + "learning_rate": 0.0007128091452643022, + "loss": 3.6435, + "step": 33815 + }, + { + "epoch": 2.297866557956244, + "grad_norm": 0.7616744637489319, + "learning_rate": 0.0007127666802554696, + "loss": 3.5654, + "step": 33820 + }, + { + "epoch": 2.298206278026906, + "grad_norm": 0.8126440644264221, + "learning_rate": 0.0007127242152466368, + "loss": 3.4232, + "step": 33825 + }, + { + "epoch": 2.2985459980975675, + "grad_norm": 0.8614236116409302, + "learning_rate": 0.000712681750237804, + "loss": 3.624, + "step": 33830 + }, + { + "epoch": 2.298885718168229, + "grad_norm": 0.9122059345245361, + "learning_rate": 0.0007126392852289713, + "loss": 3.4455, + "step": 33835 + }, + { + "epoch": 2.2992254382388913, + "grad_norm": 0.8862968683242798, + "learning_rate": 0.0007125968202201386, + "loss": 3.4778, + "step": 33840 + }, + { + "epoch": 2.299565158309553, + "grad_norm": 0.8370521664619446, + "learning_rate": 0.0007125543552113058, + "loss": 3.778, + "step": 33845 + }, + { + "epoch": 2.2999048783802145, + "grad_norm": 0.6971193552017212, + "learning_rate": 0.0007125118902024732, + "loss": 3.6383, + "step": 33850 + }, + { + "epoch": 2.3002445984508766, + "grad_norm": 0.7824249863624573, + "learning_rate": 0.0007124694251936405, + "loss": 3.5573, + "step": 33855 + }, + { + "epoch": 2.300584318521538, + "grad_norm": 0.7947288155555725, + "learning_rate": 0.0007124269601848077, + "loss": 3.2653, + "step": 33860 + }, + { + "epoch": 2.3009240385922, + "grad_norm": 0.8118384480476379, + "learning_rate": 0.000712384495175975, + "loss": 3.7857, + "step": 33865 + }, + { + "epoch": 2.301263758662862, + "grad_norm": 0.8397238850593567, + "learning_rate": 0.0007123420301671423, + "loss": 3.2688, + "step": 33870 + }, + { + "epoch": 2.3016034787335236, + "grad_norm": 0.6829814910888672, + "learning_rate": 0.0007122995651583095, + "loss": 3.34, + "step": 33875 + }, + { + "epoch": 2.301943198804185, + "grad_norm": 1.0572757720947266, + "learning_rate": 0.0007122571001494768, + "loss": 3.6329, + "step": 33880 + }, + { + "epoch": 2.3022829188748473, + "grad_norm": 0.8951638340950012, + "learning_rate": 0.0007122146351406441, + "loss": 3.4023, + "step": 33885 + }, + { + "epoch": 2.302622638945509, + "grad_norm": 0.7757227420806885, + "learning_rate": 0.0007121721701318114, + "loss": 3.5675, + "step": 33890 + }, + { + "epoch": 2.3029623590161705, + "grad_norm": 0.8249233365058899, + "learning_rate": 0.0007121297051229787, + "loss": 3.5986, + "step": 33895 + }, + { + "epoch": 2.3033020790868326, + "grad_norm": 0.8921540975570679, + "learning_rate": 0.0007120872401141459, + "loss": 3.3791, + "step": 33900 + }, + { + "epoch": 2.3036417991574942, + "grad_norm": 0.977718710899353, + "learning_rate": 0.0007120447751053133, + "loss": 3.6686, + "step": 33905 + }, + { + "epoch": 2.303981519228156, + "grad_norm": 0.8318045735359192, + "learning_rate": 0.0007120023100964805, + "loss": 3.5758, + "step": 33910 + }, + { + "epoch": 2.304321239298818, + "grad_norm": 0.9497118592262268, + "learning_rate": 0.0007119598450876477, + "loss": 3.7846, + "step": 33915 + }, + { + "epoch": 2.3046609593694796, + "grad_norm": 0.9017803072929382, + "learning_rate": 0.0007119173800788152, + "loss": 3.7131, + "step": 33920 + }, + { + "epoch": 2.305000679440141, + "grad_norm": 0.8985109329223633, + "learning_rate": 0.0007118749150699824, + "loss": 3.6689, + "step": 33925 + }, + { + "epoch": 2.3053403995108033, + "grad_norm": 0.7551419734954834, + "learning_rate": 0.0007118324500611496, + "loss": 3.3541, + "step": 33930 + }, + { + "epoch": 2.305680119581465, + "grad_norm": 1.7869906425476074, + "learning_rate": 0.000711789985052317, + "loss": 3.6841, + "step": 33935 + }, + { + "epoch": 2.3060198396521265, + "grad_norm": 0.8264340162277222, + "learning_rate": 0.0007117475200434842, + "loss": 3.6175, + "step": 33940 + }, + { + "epoch": 2.3063595597227886, + "grad_norm": 0.8188421130180359, + "learning_rate": 0.0007117050550346514, + "loss": 3.6617, + "step": 33945 + }, + { + "epoch": 2.3066992797934502, + "grad_norm": 0.8296565413475037, + "learning_rate": 0.0007116625900258188, + "loss": 3.5027, + "step": 33950 + }, + { + "epoch": 2.307038999864112, + "grad_norm": 0.824295163154602, + "learning_rate": 0.0007116201250169861, + "loss": 3.4179, + "step": 33955 + }, + { + "epoch": 2.307378719934774, + "grad_norm": 0.8817291855812073, + "learning_rate": 0.0007115776600081533, + "loss": 3.673, + "step": 33960 + }, + { + "epoch": 2.3077184400054356, + "grad_norm": 1.6192768812179565, + "learning_rate": 0.0007115351949993206, + "loss": 3.4492, + "step": 33965 + }, + { + "epoch": 2.308058160076097, + "grad_norm": 0.8845073580741882, + "learning_rate": 0.0007114927299904879, + "loss": 3.2858, + "step": 33970 + }, + { + "epoch": 2.3083978801467593, + "grad_norm": 0.6938125491142273, + "learning_rate": 0.0007114502649816551, + "loss": 3.7204, + "step": 33975 + }, + { + "epoch": 2.308737600217421, + "grad_norm": 0.9887435436248779, + "learning_rate": 0.0007114077999728224, + "loss": 3.3122, + "step": 33980 + }, + { + "epoch": 2.3090773202880825, + "grad_norm": 1.0122140645980835, + "learning_rate": 0.0007113653349639897, + "loss": 3.6489, + "step": 33985 + }, + { + "epoch": 2.3094170403587446, + "grad_norm": 0.9755750894546509, + "learning_rate": 0.000711322869955157, + "loss": 3.5054, + "step": 33990 + }, + { + "epoch": 2.3097567604294063, + "grad_norm": 1.0172529220581055, + "learning_rate": 0.0007112804049463243, + "loss": 3.7896, + "step": 33995 + }, + { + "epoch": 2.310096480500068, + "grad_norm": 1.064594030380249, + "learning_rate": 0.0007112379399374915, + "loss": 3.546, + "step": 34000 + }, + { + "epoch": 2.31043620057073, + "grad_norm": 1.0242984294891357, + "learning_rate": 0.0007111954749286588, + "loss": 3.6751, + "step": 34005 + }, + { + "epoch": 2.3107759206413916, + "grad_norm": 1.041520595550537, + "learning_rate": 0.0007111530099198261, + "loss": 3.7273, + "step": 34010 + }, + { + "epoch": 2.311115640712053, + "grad_norm": 0.8324095606803894, + "learning_rate": 0.0007111105449109933, + "loss": 3.6797, + "step": 34015 + }, + { + "epoch": 2.311455360782715, + "grad_norm": 0.7015892267227173, + "learning_rate": 0.0007110680799021607, + "loss": 3.6101, + "step": 34020 + }, + { + "epoch": 2.311795080853377, + "grad_norm": 0.803220808506012, + "learning_rate": 0.000711025614893328, + "loss": 3.5955, + "step": 34025 + }, + { + "epoch": 2.3121348009240386, + "grad_norm": 0.8305432200431824, + "learning_rate": 0.0007109831498844952, + "loss": 3.2995, + "step": 34030 + }, + { + "epoch": 2.3124745209947, + "grad_norm": 0.8131983876228333, + "learning_rate": 0.0007109406848756624, + "loss": 3.6584, + "step": 34035 + }, + { + "epoch": 2.3128142410653623, + "grad_norm": 0.7462334036827087, + "learning_rate": 0.0007108982198668298, + "loss": 3.7746, + "step": 34040 + }, + { + "epoch": 2.313153961136024, + "grad_norm": 0.7840363383293152, + "learning_rate": 0.000710855754857997, + "loss": 3.7043, + "step": 34045 + }, + { + "epoch": 2.3134936812066855, + "grad_norm": 1.8237195014953613, + "learning_rate": 0.0007108132898491642, + "loss": 3.4715, + "step": 34050 + }, + { + "epoch": 2.3138334012773476, + "grad_norm": 0.9244502782821655, + "learning_rate": 0.0007107708248403317, + "loss": 3.3044, + "step": 34055 + }, + { + "epoch": 2.3141731213480092, + "grad_norm": 0.7685158848762512, + "learning_rate": 0.0007107283598314989, + "loss": 3.3397, + "step": 34060 + }, + { + "epoch": 2.314512841418671, + "grad_norm": 0.7994508147239685, + "learning_rate": 0.0007106858948226661, + "loss": 3.6908, + "step": 34065 + }, + { + "epoch": 2.314852561489333, + "grad_norm": 1.0769963264465332, + "learning_rate": 0.0007106434298138335, + "loss": 3.5231, + "step": 34070 + }, + { + "epoch": 2.3151922815599946, + "grad_norm": 0.7900683283805847, + "learning_rate": 0.0007106009648050007, + "loss": 3.3827, + "step": 34075 + }, + { + "epoch": 2.315532001630656, + "grad_norm": 1.0778696537017822, + "learning_rate": 0.0007105584997961679, + "loss": 3.7908, + "step": 34080 + }, + { + "epoch": 2.3158717217013183, + "grad_norm": 0.760898768901825, + "learning_rate": 0.0007105160347873352, + "loss": 3.7848, + "step": 34085 + }, + { + "epoch": 2.31621144177198, + "grad_norm": 0.6816806793212891, + "learning_rate": 0.0007104735697785026, + "loss": 3.6774, + "step": 34090 + }, + { + "epoch": 2.3165511618426415, + "grad_norm": 0.8084598183631897, + "learning_rate": 0.0007104311047696698, + "loss": 3.813, + "step": 34095 + }, + { + "epoch": 2.3168908819133036, + "grad_norm": 1.1742398738861084, + "learning_rate": 0.0007103886397608371, + "loss": 3.697, + "step": 34100 + }, + { + "epoch": 2.3172306019839652, + "grad_norm": 0.846848726272583, + "learning_rate": 0.0007103461747520044, + "loss": 3.3247, + "step": 34105 + }, + { + "epoch": 2.317570322054627, + "grad_norm": 0.9724783301353455, + "learning_rate": 0.0007103037097431716, + "loss": 3.7151, + "step": 34110 + }, + { + "epoch": 2.317910042125289, + "grad_norm": 1.0472772121429443, + "learning_rate": 0.0007102612447343389, + "loss": 3.7369, + "step": 34115 + }, + { + "epoch": 2.3182497621959506, + "grad_norm": 1.0491966009140015, + "learning_rate": 0.0007102187797255061, + "loss": 3.6542, + "step": 34120 + }, + { + "epoch": 2.318589482266612, + "grad_norm": 0.9557170271873474, + "learning_rate": 0.0007101763147166735, + "loss": 3.6361, + "step": 34125 + }, + { + "epoch": 2.318929202337274, + "grad_norm": 0.9192716479301453, + "learning_rate": 0.0007101338497078408, + "loss": 3.6267, + "step": 34130 + }, + { + "epoch": 2.319268922407936, + "grad_norm": 0.9863161444664001, + "learning_rate": 0.000710091384699008, + "loss": 3.7624, + "step": 34135 + }, + { + "epoch": 2.3196086424785975, + "grad_norm": 0.7052904367446899, + "learning_rate": 0.0007100489196901753, + "loss": 3.7789, + "step": 34140 + }, + { + "epoch": 2.319948362549259, + "grad_norm": 1.1891669034957886, + "learning_rate": 0.0007100064546813426, + "loss": 3.4648, + "step": 34145 + }, + { + "epoch": 2.3202880826199213, + "grad_norm": 0.9277065396308899, + "learning_rate": 0.0007099639896725098, + "loss": 3.6379, + "step": 34150 + }, + { + "epoch": 2.320627802690583, + "grad_norm": 0.9203821420669556, + "learning_rate": 0.0007099215246636771, + "loss": 3.467, + "step": 34155 + }, + { + "epoch": 2.3209675227612445, + "grad_norm": 1.2827599048614502, + "learning_rate": 0.0007098790596548445, + "loss": 3.416, + "step": 34160 + }, + { + "epoch": 2.3213072428319066, + "grad_norm": 1.0372949838638306, + "learning_rate": 0.0007098365946460117, + "loss": 3.4516, + "step": 34165 + }, + { + "epoch": 2.3216469629025682, + "grad_norm": 0.757472813129425, + "learning_rate": 0.000709794129637179, + "loss": 3.5345, + "step": 34170 + }, + { + "epoch": 2.32198668297323, + "grad_norm": 0.8775049448013306, + "learning_rate": 0.0007097516646283463, + "loss": 3.3518, + "step": 34175 + }, + { + "epoch": 2.322326403043892, + "grad_norm": 0.9360730051994324, + "learning_rate": 0.0007097091996195135, + "loss": 3.3618, + "step": 34180 + }, + { + "epoch": 2.3226661231145536, + "grad_norm": 0.9197375774383545, + "learning_rate": 0.0007096667346106807, + "loss": 3.4241, + "step": 34185 + }, + { + "epoch": 2.323005843185215, + "grad_norm": 0.8105782270431519, + "learning_rate": 0.0007096242696018481, + "loss": 3.3682, + "step": 34190 + }, + { + "epoch": 2.3233455632558773, + "grad_norm": 0.9358724355697632, + "learning_rate": 0.0007095818045930154, + "loss": 3.6549, + "step": 34195 + }, + { + "epoch": 2.323685283326539, + "grad_norm": 1.0015637874603271, + "learning_rate": 0.0007095393395841826, + "loss": 3.4941, + "step": 34200 + }, + { + "epoch": 2.3240250033972005, + "grad_norm": 1.0032161474227905, + "learning_rate": 0.00070949687457535, + "loss": 3.4708, + "step": 34205 + }, + { + "epoch": 2.3243647234678626, + "grad_norm": 0.7717537879943848, + "learning_rate": 0.0007094544095665172, + "loss": 3.9169, + "step": 34210 + }, + { + "epoch": 2.3247044435385242, + "grad_norm": 0.7563071250915527, + "learning_rate": 0.0007094119445576844, + "loss": 3.4856, + "step": 34215 + }, + { + "epoch": 2.325044163609186, + "grad_norm": 0.9331188797950745, + "learning_rate": 0.0007093694795488518, + "loss": 3.8392, + "step": 34220 + }, + { + "epoch": 2.325383883679848, + "grad_norm": 1.0929288864135742, + "learning_rate": 0.000709327014540019, + "loss": 3.456, + "step": 34225 + }, + { + "epoch": 2.3257236037505096, + "grad_norm": 0.8375023603439331, + "learning_rate": 0.0007092845495311863, + "loss": 3.6963, + "step": 34230 + }, + { + "epoch": 2.326063323821171, + "grad_norm": 0.8193178176879883, + "learning_rate": 0.0007092420845223536, + "loss": 3.541, + "step": 34235 + }, + { + "epoch": 2.3264030438918333, + "grad_norm": 0.7918261289596558, + "learning_rate": 0.0007091996195135209, + "loss": 3.6571, + "step": 34240 + }, + { + "epoch": 2.326742763962495, + "grad_norm": 1.0284161567687988, + "learning_rate": 0.0007091571545046882, + "loss": 3.5652, + "step": 34245 + }, + { + "epoch": 2.3270824840331565, + "grad_norm": 0.9130529761314392, + "learning_rate": 0.0007091146894958554, + "loss": 3.6421, + "step": 34250 + }, + { + "epoch": 2.3274222041038186, + "grad_norm": 0.8366966843605042, + "learning_rate": 0.0007090722244870227, + "loss": 3.6082, + "step": 34255 + }, + { + "epoch": 2.3277619241744802, + "grad_norm": 1.1698920726776123, + "learning_rate": 0.00070902975947819, + "loss": 3.6491, + "step": 34260 + }, + { + "epoch": 2.328101644245142, + "grad_norm": 0.8507300019264221, + "learning_rate": 0.0007089872944693573, + "loss": 3.3246, + "step": 34265 + }, + { + "epoch": 2.328441364315804, + "grad_norm": 0.909733235836029, + "learning_rate": 0.0007089448294605246, + "loss": 3.7497, + "step": 34270 + }, + { + "epoch": 2.3287810843864656, + "grad_norm": 0.8864739537239075, + "learning_rate": 0.0007089023644516919, + "loss": 3.3475, + "step": 34275 + }, + { + "epoch": 2.329120804457127, + "grad_norm": 0.7572029829025269, + "learning_rate": 0.0007088598994428591, + "loss": 3.9518, + "step": 34280 + }, + { + "epoch": 2.3294605245277893, + "grad_norm": 0.7196884751319885, + "learning_rate": 0.0007088174344340263, + "loss": 3.4821, + "step": 34285 + }, + { + "epoch": 2.329800244598451, + "grad_norm": 1.327818751335144, + "learning_rate": 0.0007087749694251937, + "loss": 3.3762, + "step": 34290 + }, + { + "epoch": 2.3301399646691126, + "grad_norm": 0.7785714268684387, + "learning_rate": 0.0007087325044163609, + "loss": 3.6989, + "step": 34295 + }, + { + "epoch": 2.3304796847397746, + "grad_norm": 0.7618208527565002, + "learning_rate": 0.0007086900394075282, + "loss": 3.5059, + "step": 34300 + }, + { + "epoch": 2.3308194048104363, + "grad_norm": 0.9117652177810669, + "learning_rate": 0.0007086475743986956, + "loss": 3.762, + "step": 34305 + }, + { + "epoch": 2.331159124881098, + "grad_norm": 0.7800503969192505, + "learning_rate": 0.0007086051093898628, + "loss": 3.635, + "step": 34310 + }, + { + "epoch": 2.33149884495176, + "grad_norm": 0.969446063041687, + "learning_rate": 0.00070856264438103, + "loss": 3.5902, + "step": 34315 + }, + { + "epoch": 2.3318385650224216, + "grad_norm": 1.0706977844238281, + "learning_rate": 0.0007085201793721974, + "loss": 3.2806, + "step": 34320 + }, + { + "epoch": 2.3321782850930832, + "grad_norm": 0.8545742034912109, + "learning_rate": 0.0007084777143633646, + "loss": 3.3652, + "step": 34325 + }, + { + "epoch": 2.3325180051637453, + "grad_norm": 0.9091259837150574, + "learning_rate": 0.0007084352493545318, + "loss": 3.6613, + "step": 34330 + }, + { + "epoch": 2.332857725234407, + "grad_norm": 0.8166537284851074, + "learning_rate": 0.0007083927843456992, + "loss": 3.3974, + "step": 34335 + }, + { + "epoch": 2.3331974453050686, + "grad_norm": 0.86677086353302, + "learning_rate": 0.0007083503193368665, + "loss": 3.5774, + "step": 34340 + }, + { + "epoch": 2.3335371653757306, + "grad_norm": 0.847618818283081, + "learning_rate": 0.0007083078543280337, + "loss": 3.4753, + "step": 34345 + }, + { + "epoch": 2.3338768854463923, + "grad_norm": 0.903878390789032, + "learning_rate": 0.000708265389319201, + "loss": 3.6949, + "step": 34350 + }, + { + "epoch": 2.334216605517054, + "grad_norm": 1.0377613306045532, + "learning_rate": 0.0007082229243103683, + "loss": 3.5921, + "step": 34355 + }, + { + "epoch": 2.3345563255877155, + "grad_norm": 0.682405412197113, + "learning_rate": 0.0007081804593015355, + "loss": 3.7734, + "step": 34360 + }, + { + "epoch": 2.3348960456583776, + "grad_norm": 0.8002089262008667, + "learning_rate": 0.0007081379942927028, + "loss": 3.4987, + "step": 34365 + }, + { + "epoch": 2.3352357657290392, + "grad_norm": 1.0584214925765991, + "learning_rate": 0.0007080955292838702, + "loss": 3.6984, + "step": 34370 + }, + { + "epoch": 2.335575485799701, + "grad_norm": 0.7826297283172607, + "learning_rate": 0.0007080530642750374, + "loss": 3.632, + "step": 34375 + }, + { + "epoch": 2.335915205870363, + "grad_norm": 0.9048815369606018, + "learning_rate": 0.0007080105992662047, + "loss": 3.442, + "step": 34380 + }, + { + "epoch": 2.3362549259410246, + "grad_norm": 0.8867250084877014, + "learning_rate": 0.0007079681342573719, + "loss": 3.6201, + "step": 34385 + }, + { + "epoch": 2.336594646011686, + "grad_norm": 1.0725702047348022, + "learning_rate": 0.0007079256692485392, + "loss": 3.5104, + "step": 34390 + }, + { + "epoch": 2.3369343660823483, + "grad_norm": 1.0236880779266357, + "learning_rate": 0.0007078832042397065, + "loss": 3.7025, + "step": 34395 + }, + { + "epoch": 2.33727408615301, + "grad_norm": 0.9194849729537964, + "learning_rate": 0.0007078407392308737, + "loss": 3.7502, + "step": 34400 + }, + { + "epoch": 2.3376138062236715, + "grad_norm": 1.975117802619934, + "learning_rate": 0.0007077982742220411, + "loss": 3.6178, + "step": 34405 + }, + { + "epoch": 2.3379535262943336, + "grad_norm": 0.8757022023200989, + "learning_rate": 0.0007077558092132084, + "loss": 3.6882, + "step": 34410 + }, + { + "epoch": 2.3382932463649952, + "grad_norm": 1.0355061292648315, + "learning_rate": 0.0007077133442043756, + "loss": 3.6011, + "step": 34415 + }, + { + "epoch": 2.338632966435657, + "grad_norm": 0.9576653242111206, + "learning_rate": 0.0007076708791955428, + "loss": 3.3939, + "step": 34420 + }, + { + "epoch": 2.338972686506319, + "grad_norm": 0.7670703530311584, + "learning_rate": 0.0007076284141867102, + "loss": 3.2081, + "step": 34425 + }, + { + "epoch": 2.3393124065769806, + "grad_norm": 0.7585484385490417, + "learning_rate": 0.0007075859491778774, + "loss": 3.6672, + "step": 34430 + }, + { + "epoch": 2.339652126647642, + "grad_norm": 0.7069069147109985, + "learning_rate": 0.0007075434841690446, + "loss": 3.7496, + "step": 34435 + }, + { + "epoch": 2.3399918467183043, + "grad_norm": 1.0143811702728271, + "learning_rate": 0.0007075010191602121, + "loss": 3.3745, + "step": 34440 + }, + { + "epoch": 2.340331566788966, + "grad_norm": 0.8294817209243774, + "learning_rate": 0.0007074585541513793, + "loss": 3.5667, + "step": 34445 + }, + { + "epoch": 2.3406712868596276, + "grad_norm": 0.8429718613624573, + "learning_rate": 0.0007074160891425465, + "loss": 3.6083, + "step": 34450 + }, + { + "epoch": 2.3410110069302896, + "grad_norm": 0.8887079358100891, + "learning_rate": 0.0007073736241337139, + "loss": 3.784, + "step": 34455 + }, + { + "epoch": 2.3413507270009513, + "grad_norm": 1.3208699226379395, + "learning_rate": 0.0007073311591248811, + "loss": 3.676, + "step": 34460 + }, + { + "epoch": 2.341690447071613, + "grad_norm": 1.106257677078247, + "learning_rate": 0.0007072886941160483, + "loss": 3.6297, + "step": 34465 + }, + { + "epoch": 2.3420301671422745, + "grad_norm": 0.7275393009185791, + "learning_rate": 0.0007072462291072156, + "loss": 3.5263, + "step": 34470 + }, + { + "epoch": 2.3423698872129366, + "grad_norm": 0.9168503284454346, + "learning_rate": 0.000707203764098383, + "loss": 3.3449, + "step": 34475 + }, + { + "epoch": 2.3427096072835982, + "grad_norm": 0.8990557193756104, + "learning_rate": 0.0007071612990895502, + "loss": 3.6523, + "step": 34480 + }, + { + "epoch": 2.34304932735426, + "grad_norm": 0.9464854001998901, + "learning_rate": 0.0007071188340807175, + "loss": 3.5586, + "step": 34485 + }, + { + "epoch": 2.343389047424922, + "grad_norm": 1.0622926950454712, + "learning_rate": 0.0007070763690718848, + "loss": 3.6451, + "step": 34490 + }, + { + "epoch": 2.3437287674955836, + "grad_norm": 4.860132694244385, + "learning_rate": 0.000707033904063052, + "loss": 3.7712, + "step": 34495 + }, + { + "epoch": 2.344068487566245, + "grad_norm": 0.9055885076522827, + "learning_rate": 0.0007069914390542193, + "loss": 3.3985, + "step": 34500 + }, + { + "epoch": 2.3444082076369073, + "grad_norm": 1.129346489906311, + "learning_rate": 0.0007069489740453866, + "loss": 3.6186, + "step": 34505 + }, + { + "epoch": 2.344747927707569, + "grad_norm": 1.1102019548416138, + "learning_rate": 0.0007069065090365539, + "loss": 3.5941, + "step": 34510 + }, + { + "epoch": 2.3450876477782305, + "grad_norm": 1.0202596187591553, + "learning_rate": 0.0007068640440277212, + "loss": 3.4945, + "step": 34515 + }, + { + "epoch": 2.3454273678488926, + "grad_norm": 0.870394766330719, + "learning_rate": 0.0007068215790188884, + "loss": 3.4026, + "step": 34520 + }, + { + "epoch": 2.3457670879195542, + "grad_norm": 0.8878560066223145, + "learning_rate": 0.0007067791140100557, + "loss": 3.7226, + "step": 34525 + }, + { + "epoch": 2.346106807990216, + "grad_norm": 1.066633939743042, + "learning_rate": 0.000706736649001223, + "loss": 3.7562, + "step": 34530 + }, + { + "epoch": 2.346446528060878, + "grad_norm": 1.0387638807296753, + "learning_rate": 0.0007066941839923902, + "loss": 3.7419, + "step": 34535 + }, + { + "epoch": 2.3467862481315396, + "grad_norm": 0.9734619855880737, + "learning_rate": 0.0007066517189835575, + "loss": 3.3707, + "step": 34540 + }, + { + "epoch": 2.347125968202201, + "grad_norm": 0.8226066827774048, + "learning_rate": 0.0007066092539747249, + "loss": 3.5612, + "step": 34545 + }, + { + "epoch": 2.3474656882728633, + "grad_norm": 1.1062244176864624, + "learning_rate": 0.0007065667889658921, + "loss": 3.6997, + "step": 34550 + }, + { + "epoch": 2.347805408343525, + "grad_norm": 0.9129453897476196, + "learning_rate": 0.0007065243239570594, + "loss": 3.6935, + "step": 34555 + }, + { + "epoch": 2.3481451284141865, + "grad_norm": 0.8260941505432129, + "learning_rate": 0.0007064818589482267, + "loss": 3.7246, + "step": 34560 + }, + { + "epoch": 2.3484848484848486, + "grad_norm": 0.8649405241012573, + "learning_rate": 0.0007064393939393939, + "loss": 3.7534, + "step": 34565 + }, + { + "epoch": 2.3488245685555103, + "grad_norm": 0.7330026030540466, + "learning_rate": 0.0007063969289305611, + "loss": 3.6331, + "step": 34570 + }, + { + "epoch": 2.349164288626172, + "grad_norm": 1.1282317638397217, + "learning_rate": 0.0007063544639217286, + "loss": 3.5202, + "step": 34575 + }, + { + "epoch": 2.349504008696834, + "grad_norm": 0.9743011593818665, + "learning_rate": 0.0007063119989128958, + "loss": 3.6357, + "step": 34580 + }, + { + "epoch": 2.3498437287674956, + "grad_norm": 0.921506941318512, + "learning_rate": 0.0007062695339040631, + "loss": 3.5229, + "step": 34585 + }, + { + "epoch": 2.350183448838157, + "grad_norm": 0.7268308401107788, + "learning_rate": 0.0007062270688952304, + "loss": 3.7577, + "step": 34590 + }, + { + "epoch": 2.3505231689088193, + "grad_norm": 0.8359577059745789, + "learning_rate": 0.0007061846038863976, + "loss": 3.4376, + "step": 34595 + }, + { + "epoch": 2.350862888979481, + "grad_norm": 0.7796173095703125, + "learning_rate": 0.0007061421388775649, + "loss": 3.6381, + "step": 34600 + }, + { + "epoch": 2.3512026090501426, + "grad_norm": 1.031511902809143, + "learning_rate": 0.0007060996738687322, + "loss": 3.6355, + "step": 34605 + }, + { + "epoch": 2.3515423291208046, + "grad_norm": 0.8212525844573975, + "learning_rate": 0.0007060572088598995, + "loss": 3.7198, + "step": 34610 + }, + { + "epoch": 2.3518820491914663, + "grad_norm": 0.8388593196868896, + "learning_rate": 0.0007060147438510668, + "loss": 3.4942, + "step": 34615 + }, + { + "epoch": 2.352221769262128, + "grad_norm": 0.6859729290008545, + "learning_rate": 0.000705972278842234, + "loss": 3.4758, + "step": 34620 + }, + { + "epoch": 2.35256148933279, + "grad_norm": 0.7173684239387512, + "learning_rate": 0.0007059298138334013, + "loss": 3.6544, + "step": 34625 + }, + { + "epoch": 2.3529012094034516, + "grad_norm": 0.8629988431930542, + "learning_rate": 0.0007058873488245686, + "loss": 3.7, + "step": 34630 + }, + { + "epoch": 2.3532409294741132, + "grad_norm": 1.128861427307129, + "learning_rate": 0.0007058448838157358, + "loss": 3.8156, + "step": 34635 + }, + { + "epoch": 2.3535806495447753, + "grad_norm": 0.8494544625282288, + "learning_rate": 0.0007058024188069031, + "loss": 3.4643, + "step": 34640 + }, + { + "epoch": 2.353920369615437, + "grad_norm": 0.6636223793029785, + "learning_rate": 0.0007057599537980705, + "loss": 3.5783, + "step": 34645 + }, + { + "epoch": 2.3542600896860986, + "grad_norm": 0.7290956974029541, + "learning_rate": 0.0007057174887892377, + "loss": 3.5841, + "step": 34650 + }, + { + "epoch": 2.3545998097567606, + "grad_norm": 0.9678652882575989, + "learning_rate": 0.000705675023780405, + "loss": 3.629, + "step": 34655 + }, + { + "epoch": 2.3549395298274223, + "grad_norm": 0.7674130201339722, + "learning_rate": 0.0007056325587715723, + "loss": 3.6045, + "step": 34660 + }, + { + "epoch": 2.355279249898084, + "grad_norm": 0.902842104434967, + "learning_rate": 0.0007055900937627395, + "loss": 3.6456, + "step": 34665 + }, + { + "epoch": 2.355618969968746, + "grad_norm": 0.7511454820632935, + "learning_rate": 0.0007055476287539067, + "loss": 3.6316, + "step": 34670 + }, + { + "epoch": 2.3559586900394076, + "grad_norm": 0.6847243309020996, + "learning_rate": 0.0007055051637450741, + "loss": 3.6069, + "step": 34675 + }, + { + "epoch": 2.3562984101100692, + "grad_norm": 0.9734264016151428, + "learning_rate": 0.0007054626987362414, + "loss": 3.5082, + "step": 34680 + }, + { + "epoch": 2.3566381301807313, + "grad_norm": 0.8817508220672607, + "learning_rate": 0.0007054202337274086, + "loss": 3.5564, + "step": 34685 + }, + { + "epoch": 2.356977850251393, + "grad_norm": 0.7149307131767273, + "learning_rate": 0.000705377768718576, + "loss": 3.5211, + "step": 34690 + }, + { + "epoch": 2.3573175703220546, + "grad_norm": 0.8015739321708679, + "learning_rate": 0.0007053353037097432, + "loss": 3.4549, + "step": 34695 + }, + { + "epoch": 2.357657290392716, + "grad_norm": 0.9455263018608093, + "learning_rate": 0.0007052928387009104, + "loss": 3.9212, + "step": 34700 + }, + { + "epoch": 2.3579970104633783, + "grad_norm": 1.5807610750198364, + "learning_rate": 0.0007052503736920778, + "loss": 3.5016, + "step": 34705 + }, + { + "epoch": 2.35833673053404, + "grad_norm": 0.8543570637702942, + "learning_rate": 0.000705207908683245, + "loss": 3.4, + "step": 34710 + }, + { + "epoch": 2.3586764506047015, + "grad_norm": 0.9081307053565979, + "learning_rate": 0.0007051654436744123, + "loss": 3.3649, + "step": 34715 + }, + { + "epoch": 2.3590161706753636, + "grad_norm": 0.8815417289733887, + "learning_rate": 0.0007051229786655796, + "loss": 3.6709, + "step": 34720 + }, + { + "epoch": 2.3593558907460253, + "grad_norm": 0.8759732246398926, + "learning_rate": 0.0007050805136567469, + "loss": 3.6189, + "step": 34725 + }, + { + "epoch": 2.359695610816687, + "grad_norm": 0.9433268308639526, + "learning_rate": 0.0007050380486479141, + "loss": 3.4962, + "step": 34730 + }, + { + "epoch": 2.360035330887349, + "grad_norm": 1.0115960836410522, + "learning_rate": 0.0007049955836390814, + "loss": 3.5265, + "step": 34735 + }, + { + "epoch": 2.3603750509580106, + "grad_norm": 0.871773898601532, + "learning_rate": 0.0007049531186302487, + "loss": 3.2949, + "step": 34740 + }, + { + "epoch": 2.360714771028672, + "grad_norm": 0.8277033567428589, + "learning_rate": 0.0007049106536214159, + "loss": 3.5797, + "step": 34745 + }, + { + "epoch": 2.3610544910993343, + "grad_norm": 0.8331530094146729, + "learning_rate": 0.0007048681886125833, + "loss": 3.5718, + "step": 34750 + }, + { + "epoch": 2.361394211169996, + "grad_norm": 0.7831059694290161, + "learning_rate": 0.0007048257236037506, + "loss": 3.5988, + "step": 34755 + }, + { + "epoch": 2.3617339312406576, + "grad_norm": 0.7845896482467651, + "learning_rate": 0.0007047832585949178, + "loss": 3.6117, + "step": 34760 + }, + { + "epoch": 2.3620736513113196, + "grad_norm": 0.9187370538711548, + "learning_rate": 0.0007047407935860851, + "loss": 3.5545, + "step": 34765 + }, + { + "epoch": 2.3624133713819813, + "grad_norm": 0.9950335621833801, + "learning_rate": 0.0007046983285772523, + "loss": 3.2773, + "step": 34770 + }, + { + "epoch": 2.362753091452643, + "grad_norm": 1.020647406578064, + "learning_rate": 0.0007046558635684196, + "loss": 3.543, + "step": 34775 + }, + { + "epoch": 2.363092811523305, + "grad_norm": 0.8755601644515991, + "learning_rate": 0.0007046133985595869, + "loss": 3.5004, + "step": 34780 + }, + { + "epoch": 2.3634325315939666, + "grad_norm": 0.898391604423523, + "learning_rate": 0.0007045709335507542, + "loss": 3.5907, + "step": 34785 + }, + { + "epoch": 2.3637722516646282, + "grad_norm": 1.3889415264129639, + "learning_rate": 0.0007045284685419215, + "loss": 3.7194, + "step": 34790 + }, + { + "epoch": 2.3641119717352903, + "grad_norm": 0.6181111931800842, + "learning_rate": 0.0007044860035330888, + "loss": 3.5344, + "step": 34795 + }, + { + "epoch": 2.364451691805952, + "grad_norm": 0.9931139945983887, + "learning_rate": 0.000704443538524256, + "loss": 3.7138, + "step": 34800 + }, + { + "epoch": 2.3647914118766136, + "grad_norm": 0.7110602259635925, + "learning_rate": 0.0007044010735154232, + "loss": 3.6617, + "step": 34805 + }, + { + "epoch": 2.365131131947275, + "grad_norm": 4.998938083648682, + "learning_rate": 0.0007043586085065906, + "loss": 3.6314, + "step": 34810 + }, + { + "epoch": 2.3654708520179373, + "grad_norm": 0.7764558792114258, + "learning_rate": 0.0007043161434977578, + "loss": 3.3432, + "step": 34815 + }, + { + "epoch": 2.365810572088599, + "grad_norm": 0.9735468626022339, + "learning_rate": 0.0007042736784889251, + "loss": 3.3736, + "step": 34820 + }, + { + "epoch": 2.3661502921592605, + "grad_norm": 0.7165238261222839, + "learning_rate": 0.0007042312134800925, + "loss": 3.0851, + "step": 34825 + }, + { + "epoch": 2.3664900122299226, + "grad_norm": 0.8510459065437317, + "learning_rate": 0.0007041887484712597, + "loss": 3.1694, + "step": 34830 + }, + { + "epoch": 2.3668297323005842, + "grad_norm": 0.7846376299858093, + "learning_rate": 0.0007041462834624269, + "loss": 3.7947, + "step": 34835 + }, + { + "epoch": 2.367169452371246, + "grad_norm": 0.8124869465827942, + "learning_rate": 0.0007041038184535943, + "loss": 3.7984, + "step": 34840 + }, + { + "epoch": 2.367509172441908, + "grad_norm": 0.8498038053512573, + "learning_rate": 0.0007040613534447615, + "loss": 3.5317, + "step": 34845 + }, + { + "epoch": 2.3678488925125696, + "grad_norm": 0.7841882705688477, + "learning_rate": 0.0007040188884359287, + "loss": 3.6931, + "step": 34850 + }, + { + "epoch": 2.368188612583231, + "grad_norm": 0.7706218957901001, + "learning_rate": 0.0007039764234270962, + "loss": 3.5774, + "step": 34855 + }, + { + "epoch": 2.3685283326538933, + "grad_norm": 0.8382828831672668, + "learning_rate": 0.0007039339584182634, + "loss": 3.5038, + "step": 34860 + }, + { + "epoch": 2.368868052724555, + "grad_norm": 0.8014661073684692, + "learning_rate": 0.0007038914934094306, + "loss": 3.561, + "step": 34865 + }, + { + "epoch": 2.3692077727952165, + "grad_norm": 1.1246716976165771, + "learning_rate": 0.0007038490284005979, + "loss": 3.6765, + "step": 34870 + }, + { + "epoch": 2.3695474928658786, + "grad_norm": 0.8412127494812012, + "learning_rate": 0.0007038065633917652, + "loss": 3.7021, + "step": 34875 + }, + { + "epoch": 2.3698872129365403, + "grad_norm": 0.8694748878479004, + "learning_rate": 0.0007037640983829324, + "loss": 3.5457, + "step": 34880 + }, + { + "epoch": 2.370226933007202, + "grad_norm": 0.829232931137085, + "learning_rate": 0.0007037216333740997, + "loss": 3.6355, + "step": 34885 + }, + { + "epoch": 2.370566653077864, + "grad_norm": 1.800268292427063, + "learning_rate": 0.0007036791683652671, + "loss": 3.5249, + "step": 34890 + }, + { + "epoch": 2.3709063731485256, + "grad_norm": 0.8035861253738403, + "learning_rate": 0.0007036367033564343, + "loss": 3.6698, + "step": 34895 + }, + { + "epoch": 2.3712460932191872, + "grad_norm": 2.7226827144622803, + "learning_rate": 0.0007035942383476016, + "loss": 3.6353, + "step": 34900 + }, + { + "epoch": 2.3715858132898493, + "grad_norm": 0.9045145511627197, + "learning_rate": 0.0007035517733387688, + "loss": 3.5663, + "step": 34905 + }, + { + "epoch": 2.371925533360511, + "grad_norm": 1.0053484439849854, + "learning_rate": 0.0007035093083299361, + "loss": 3.3924, + "step": 34910 + }, + { + "epoch": 2.3722652534311726, + "grad_norm": 0.7027516961097717, + "learning_rate": 0.0007034668433211034, + "loss": 3.7158, + "step": 34915 + }, + { + "epoch": 2.3726049735018346, + "grad_norm": 0.878036379814148, + "learning_rate": 0.0007034243783122706, + "loss": 3.568, + "step": 34920 + }, + { + "epoch": 2.3729446935724963, + "grad_norm": 1.3156311511993408, + "learning_rate": 0.0007033819133034381, + "loss": 3.532, + "step": 34925 + }, + { + "epoch": 2.373284413643158, + "grad_norm": 0.885959267616272, + "learning_rate": 0.0007033394482946053, + "loss": 3.6234, + "step": 34930 + }, + { + "epoch": 2.37362413371382, + "grad_norm": 1.0305720567703247, + "learning_rate": 0.0007032969832857725, + "loss": 3.7252, + "step": 34935 + }, + { + "epoch": 2.3739638537844816, + "grad_norm": 0.8468777537345886, + "learning_rate": 0.0007032545182769399, + "loss": 3.7487, + "step": 34940 + }, + { + "epoch": 2.3743035738551432, + "grad_norm": 1.4521880149841309, + "learning_rate": 0.0007032120532681071, + "loss": 3.4531, + "step": 34945 + }, + { + "epoch": 2.3746432939258053, + "grad_norm": 0.9655662178993225, + "learning_rate": 0.0007031695882592743, + "loss": 3.6747, + "step": 34950 + }, + { + "epoch": 2.374983013996467, + "grad_norm": 0.9979805946350098, + "learning_rate": 0.0007031271232504417, + "loss": 3.5536, + "step": 34955 + }, + { + "epoch": 2.3753227340671286, + "grad_norm": 0.7736377120018005, + "learning_rate": 0.000703084658241609, + "loss": 3.5414, + "step": 34960 + }, + { + "epoch": 2.3756624541377906, + "grad_norm": 0.8296526074409485, + "learning_rate": 0.0007030421932327762, + "loss": 3.6541, + "step": 34965 + }, + { + "epoch": 2.3760021742084523, + "grad_norm": 0.8666050434112549, + "learning_rate": 0.0007029997282239435, + "loss": 3.6216, + "step": 34970 + }, + { + "epoch": 2.376341894279114, + "grad_norm": 0.8654815554618835, + "learning_rate": 0.0007029572632151108, + "loss": 3.6338, + "step": 34975 + }, + { + "epoch": 2.376681614349776, + "grad_norm": 0.9888479113578796, + "learning_rate": 0.000702914798206278, + "loss": 3.4786, + "step": 34980 + }, + { + "epoch": 2.3770213344204376, + "grad_norm": 1.0325915813446045, + "learning_rate": 0.0007028723331974453, + "loss": 3.4872, + "step": 34985 + }, + { + "epoch": 2.3773610544910992, + "grad_norm": 0.7684481739997864, + "learning_rate": 0.0007028298681886126, + "loss": 3.744, + "step": 34990 + }, + { + "epoch": 2.3777007745617613, + "grad_norm": 0.9074941873550415, + "learning_rate": 0.0007027874031797799, + "loss": 3.5123, + "step": 34995 + }, + { + "epoch": 2.378040494632423, + "grad_norm": 0.8340933322906494, + "learning_rate": 0.0007027449381709472, + "loss": 3.6753, + "step": 35000 + }, + { + "epoch": 2.3783802147030846, + "grad_norm": 0.9796059131622314, + "learning_rate": 0.0007027024731621145, + "loss": 3.4091, + "step": 35005 + }, + { + "epoch": 2.3787199347737467, + "grad_norm": 0.7915922999382019, + "learning_rate": 0.0007026600081532817, + "loss": 3.7104, + "step": 35010 + }, + { + "epoch": 2.3790596548444083, + "grad_norm": 0.8944940567016602, + "learning_rate": 0.000702617543144449, + "loss": 3.5503, + "step": 35015 + }, + { + "epoch": 2.37939937491507, + "grad_norm": 0.7971494793891907, + "learning_rate": 0.0007025750781356162, + "loss": 3.6392, + "step": 35020 + }, + { + "epoch": 2.379739094985732, + "grad_norm": 1.332970380783081, + "learning_rate": 0.0007025326131267835, + "loss": 3.7434, + "step": 35025 + }, + { + "epoch": 2.3800788150563936, + "grad_norm": 1.3748971223831177, + "learning_rate": 0.0007024901481179509, + "loss": 3.5831, + "step": 35030 + }, + { + "epoch": 2.3804185351270553, + "grad_norm": 0.8054746985435486, + "learning_rate": 0.0007024476831091181, + "loss": 3.6861, + "step": 35035 + }, + { + "epoch": 2.380758255197717, + "grad_norm": 1.1740444898605347, + "learning_rate": 0.0007024052181002854, + "loss": 3.7624, + "step": 35040 + }, + { + "epoch": 2.381097975268379, + "grad_norm": 1.0862822532653809, + "learning_rate": 0.0007023627530914527, + "loss": 3.2013, + "step": 35045 + }, + { + "epoch": 2.3814376953390406, + "grad_norm": 0.823153555393219, + "learning_rate": 0.0007023202880826199, + "loss": 3.2775, + "step": 35050 + }, + { + "epoch": 2.3817774154097022, + "grad_norm": 1.2217553853988647, + "learning_rate": 0.0007022778230737871, + "loss": 3.6296, + "step": 35055 + }, + { + "epoch": 2.3821171354803643, + "grad_norm": 0.8057252168655396, + "learning_rate": 0.0007022353580649546, + "loss": 3.5592, + "step": 35060 + }, + { + "epoch": 2.382456855551026, + "grad_norm": 0.8286086916923523, + "learning_rate": 0.0007021928930561218, + "loss": 3.6506, + "step": 35065 + }, + { + "epoch": 2.3827965756216876, + "grad_norm": 0.9963324069976807, + "learning_rate": 0.000702150428047289, + "loss": 3.5708, + "step": 35070 + }, + { + "epoch": 2.3831362956923496, + "grad_norm": 0.8757644295692444, + "learning_rate": 0.0007021079630384564, + "loss": 3.5466, + "step": 35075 + }, + { + "epoch": 2.3834760157630113, + "grad_norm": 1.0636075735092163, + "learning_rate": 0.0007020654980296236, + "loss": 3.5707, + "step": 35080 + }, + { + "epoch": 2.383815735833673, + "grad_norm": 1.0848828554153442, + "learning_rate": 0.0007020230330207908, + "loss": 3.422, + "step": 35085 + }, + { + "epoch": 2.384155455904335, + "grad_norm": 0.8940354585647583, + "learning_rate": 0.0007019805680119582, + "loss": 3.5293, + "step": 35090 + }, + { + "epoch": 2.3844951759749966, + "grad_norm": 0.8815000057220459, + "learning_rate": 0.0007019381030031255, + "loss": 3.8505, + "step": 35095 + }, + { + "epoch": 2.3848348960456582, + "grad_norm": 0.7502058744430542, + "learning_rate": 0.0007018956379942927, + "loss": 3.7819, + "step": 35100 + }, + { + "epoch": 2.3851746161163203, + "grad_norm": 1.832905888557434, + "learning_rate": 0.00070185317298546, + "loss": 3.8133, + "step": 35105 + }, + { + "epoch": 2.385514336186982, + "grad_norm": 0.9052382111549377, + "learning_rate": 0.0007018107079766273, + "loss": 3.7156, + "step": 35110 + }, + { + "epoch": 2.3858540562576436, + "grad_norm": 0.7821452021598816, + "learning_rate": 0.0007017682429677945, + "loss": 3.5069, + "step": 35115 + }, + { + "epoch": 2.3861937763283056, + "grad_norm": 0.8948304057121277, + "learning_rate": 0.0007017257779589618, + "loss": 3.812, + "step": 35120 + }, + { + "epoch": 2.3865334963989673, + "grad_norm": 0.8012406826019287, + "learning_rate": 0.0007016833129501291, + "loss": 3.6965, + "step": 35125 + }, + { + "epoch": 2.386873216469629, + "grad_norm": 0.762101948261261, + "learning_rate": 0.0007016408479412964, + "loss": 3.8838, + "step": 35130 + }, + { + "epoch": 2.387212936540291, + "grad_norm": 0.824510931968689, + "learning_rate": 0.0007015983829324637, + "loss": 3.5625, + "step": 35135 + }, + { + "epoch": 2.3875526566109526, + "grad_norm": 0.7094634175300598, + "learning_rate": 0.000701555917923631, + "loss": 3.5224, + "step": 35140 + }, + { + "epoch": 2.3878923766816142, + "grad_norm": 0.9212439060211182, + "learning_rate": 0.0007015134529147982, + "loss": 3.638, + "step": 35145 + }, + { + "epoch": 2.388232096752276, + "grad_norm": 1.0975406169891357, + "learning_rate": 0.0007014709879059655, + "loss": 3.6119, + "step": 35150 + }, + { + "epoch": 2.388571816822938, + "grad_norm": 0.7583838701248169, + "learning_rate": 0.0007014285228971327, + "loss": 3.6175, + "step": 35155 + }, + { + "epoch": 2.3889115368935996, + "grad_norm": 0.8665974140167236, + "learning_rate": 0.0007013860578883, + "loss": 3.75, + "step": 35160 + }, + { + "epoch": 2.389251256964261, + "grad_norm": 1.0681042671203613, + "learning_rate": 0.0007013435928794674, + "loss": 3.6799, + "step": 35165 + }, + { + "epoch": 2.3895909770349233, + "grad_norm": 1.0444397926330566, + "learning_rate": 0.0007013011278706346, + "loss": 3.4206, + "step": 35170 + }, + { + "epoch": 2.389930697105585, + "grad_norm": 0.9459745287895203, + "learning_rate": 0.0007012586628618019, + "loss": 3.6722, + "step": 35175 + }, + { + "epoch": 2.3902704171762466, + "grad_norm": 1.1398289203643799, + "learning_rate": 0.0007012161978529692, + "loss": 3.4389, + "step": 35180 + }, + { + "epoch": 2.3906101372469086, + "grad_norm": 0.9716097712516785, + "learning_rate": 0.0007011737328441364, + "loss": 3.5042, + "step": 35185 + }, + { + "epoch": 2.3909498573175703, + "grad_norm": 0.7929019927978516, + "learning_rate": 0.0007011312678353037, + "loss": 3.5095, + "step": 35190 + }, + { + "epoch": 2.391289577388232, + "grad_norm": 0.8883367776870728, + "learning_rate": 0.000701088802826471, + "loss": 3.7362, + "step": 35195 + }, + { + "epoch": 2.391629297458894, + "grad_norm": 0.8819423913955688, + "learning_rate": 0.0007010463378176383, + "loss": 3.8422, + "step": 35200 + }, + { + "epoch": 2.3919690175295556, + "grad_norm": 0.8674731254577637, + "learning_rate": 0.0007010038728088055, + "loss": 3.1445, + "step": 35205 + }, + { + "epoch": 2.3923087376002172, + "grad_norm": 0.854587972164154, + "learning_rate": 0.0007009614077999729, + "loss": 3.9388, + "step": 35210 + }, + { + "epoch": 2.3926484576708793, + "grad_norm": 0.9078401327133179, + "learning_rate": 0.0007009189427911401, + "loss": 3.6791, + "step": 35215 + }, + { + "epoch": 2.392988177741541, + "grad_norm": 0.6886181831359863, + "learning_rate": 0.0007008764777823073, + "loss": 3.6302, + "step": 35220 + }, + { + "epoch": 2.3933278978122026, + "grad_norm": 0.8518375754356384, + "learning_rate": 0.0007008340127734747, + "loss": 3.814, + "step": 35225 + }, + { + "epoch": 2.3936676178828646, + "grad_norm": 0.9054287672042847, + "learning_rate": 0.0007007915477646419, + "loss": 3.7222, + "step": 35230 + }, + { + "epoch": 2.3940073379535263, + "grad_norm": 0.9529050588607788, + "learning_rate": 0.0007007490827558092, + "loss": 3.4628, + "step": 35235 + }, + { + "epoch": 2.394347058024188, + "grad_norm": 0.9817489385604858, + "learning_rate": 0.0007007066177469766, + "loss": 3.7105, + "step": 35240 + }, + { + "epoch": 2.39468677809485, + "grad_norm": 0.8480692505836487, + "learning_rate": 0.0007006641527381438, + "loss": 3.4509, + "step": 35245 + }, + { + "epoch": 2.3950264981655116, + "grad_norm": 0.9525547027587891, + "learning_rate": 0.000700621687729311, + "loss": 3.4153, + "step": 35250 + }, + { + "epoch": 2.3953662182361732, + "grad_norm": 0.7224661111831665, + "learning_rate": 0.0007005792227204783, + "loss": 3.4337, + "step": 35255 + }, + { + "epoch": 2.3957059383068353, + "grad_norm": 1.3708618879318237, + "learning_rate": 0.0007005367577116456, + "loss": 3.5217, + "step": 35260 + }, + { + "epoch": 2.396045658377497, + "grad_norm": 0.7442063093185425, + "learning_rate": 0.0007004942927028129, + "loss": 3.8185, + "step": 35265 + }, + { + "epoch": 2.3963853784481586, + "grad_norm": 0.8112321496009827, + "learning_rate": 0.0007004518276939802, + "loss": 3.6568, + "step": 35270 + }, + { + "epoch": 2.3967250985188207, + "grad_norm": 0.7316697239875793, + "learning_rate": 0.0007004093626851475, + "loss": 3.5508, + "step": 35275 + }, + { + "epoch": 2.3970648185894823, + "grad_norm": 0.8168145418167114, + "learning_rate": 0.0007003668976763148, + "loss": 3.641, + "step": 35280 + }, + { + "epoch": 2.397404538660144, + "grad_norm": 0.7948176860809326, + "learning_rate": 0.000700324432667482, + "loss": 3.6508, + "step": 35285 + }, + { + "epoch": 2.397744258730806, + "grad_norm": 0.8953681588172913, + "learning_rate": 0.0007002819676586493, + "loss": 3.6532, + "step": 35290 + }, + { + "epoch": 2.3980839788014676, + "grad_norm": 0.7659631371498108, + "learning_rate": 0.0007002395026498166, + "loss": 3.5541, + "step": 35295 + }, + { + "epoch": 2.3984236988721293, + "grad_norm": 0.8775231838226318, + "learning_rate": 0.0007001970376409838, + "loss": 3.5622, + "step": 35300 + }, + { + "epoch": 2.3987634189427913, + "grad_norm": 0.832393229007721, + "learning_rate": 0.0007001545726321511, + "loss": 3.7594, + "step": 35305 + }, + { + "epoch": 2.399103139013453, + "grad_norm": 0.8624285459518433, + "learning_rate": 0.0007001121076233185, + "loss": 3.3551, + "step": 35310 + }, + { + "epoch": 2.3994428590841146, + "grad_norm": 1.0660191774368286, + "learning_rate": 0.0007000696426144857, + "loss": 3.4003, + "step": 35315 + }, + { + "epoch": 2.3997825791547767, + "grad_norm": 1.0410321950912476, + "learning_rate": 0.0007000271776056529, + "loss": 3.581, + "step": 35320 + }, + { + "epoch": 2.4001222992254383, + "grad_norm": 0.8879261016845703, + "learning_rate": 0.0006999847125968203, + "loss": 3.6809, + "step": 35325 + }, + { + "epoch": 2.4004620192961, + "grad_norm": 0.7513574957847595, + "learning_rate": 0.0006999422475879875, + "loss": 3.4495, + "step": 35330 + }, + { + "epoch": 2.400801739366762, + "grad_norm": 0.7090854048728943, + "learning_rate": 0.0006998997825791547, + "loss": 3.8296, + "step": 35335 + }, + { + "epoch": 2.4011414594374236, + "grad_norm": 0.7399373650550842, + "learning_rate": 0.0006998573175703222, + "loss": 3.5137, + "step": 35340 + }, + { + "epoch": 2.4014811795080853, + "grad_norm": 0.8372071981430054, + "learning_rate": 0.0006998148525614894, + "loss": 3.5607, + "step": 35345 + }, + { + "epoch": 2.4018208995787473, + "grad_norm": 1.0791149139404297, + "learning_rate": 0.0006997723875526566, + "loss": 3.3788, + "step": 35350 + }, + { + "epoch": 2.402160619649409, + "grad_norm": 0.9231259822845459, + "learning_rate": 0.000699729922543824, + "loss": 3.5509, + "step": 35355 + }, + { + "epoch": 2.4025003397200706, + "grad_norm": 1.2656488418579102, + "learning_rate": 0.0006996874575349912, + "loss": 3.5402, + "step": 35360 + }, + { + "epoch": 2.4028400597907327, + "grad_norm": 1.1341971158981323, + "learning_rate": 0.0006996449925261584, + "loss": 3.4829, + "step": 35365 + }, + { + "epoch": 2.4031797798613943, + "grad_norm": 0.8347205519676208, + "learning_rate": 0.0006996025275173257, + "loss": 3.7142, + "step": 35370 + }, + { + "epoch": 2.403519499932056, + "grad_norm": 0.7063478231430054, + "learning_rate": 0.0006995600625084931, + "loss": 3.4618, + "step": 35375 + }, + { + "epoch": 2.4038592200027176, + "grad_norm": 0.8571862578392029, + "learning_rate": 0.0006995175974996603, + "loss": 3.4671, + "step": 35380 + }, + { + "epoch": 2.4041989400733796, + "grad_norm": 0.8497893214225769, + "learning_rate": 0.0006994751324908276, + "loss": 3.6183, + "step": 35385 + }, + { + "epoch": 2.4045386601440413, + "grad_norm": 1.3127068281173706, + "learning_rate": 0.0006994326674819949, + "loss": 3.5464, + "step": 35390 + }, + { + "epoch": 2.404878380214703, + "grad_norm": 0.8803237676620483, + "learning_rate": 0.0006993902024731621, + "loss": 3.5213, + "step": 35395 + }, + { + "epoch": 2.405218100285365, + "grad_norm": 0.7964141368865967, + "learning_rate": 0.0006993477374643294, + "loss": 3.5652, + "step": 35400 + }, + { + "epoch": 2.4055578203560266, + "grad_norm": 0.8695953488349915, + "learning_rate": 0.0006993052724554966, + "loss": 3.7839, + "step": 35405 + }, + { + "epoch": 2.4058975404266882, + "grad_norm": 0.9421870112419128, + "learning_rate": 0.000699262807446664, + "loss": 3.4181, + "step": 35410 + }, + { + "epoch": 2.4062372604973503, + "grad_norm": 0.807289183139801, + "learning_rate": 0.0006992203424378313, + "loss": 3.451, + "step": 35415 + }, + { + "epoch": 2.406576980568012, + "grad_norm": 0.7639554142951965, + "learning_rate": 0.0006991778774289985, + "loss": 3.5529, + "step": 35420 + }, + { + "epoch": 2.4069167006386736, + "grad_norm": 1.0106925964355469, + "learning_rate": 0.0006991354124201658, + "loss": 3.6477, + "step": 35425 + }, + { + "epoch": 2.4072564207093357, + "grad_norm": 0.9396530389785767, + "learning_rate": 0.0006990929474113331, + "loss": 3.653, + "step": 35430 + }, + { + "epoch": 2.4075961407799973, + "grad_norm": 1.131540298461914, + "learning_rate": 0.0006990504824025003, + "loss": 3.4754, + "step": 35435 + }, + { + "epoch": 2.407935860850659, + "grad_norm": 0.7946423888206482, + "learning_rate": 0.0006990080173936675, + "loss": 3.5346, + "step": 35440 + }, + { + "epoch": 2.408275580921321, + "grad_norm": 0.8676256537437439, + "learning_rate": 0.000698965552384835, + "loss": 3.7595, + "step": 35445 + }, + { + "epoch": 2.4086153009919826, + "grad_norm": 0.9540257453918457, + "learning_rate": 0.0006989230873760022, + "loss": 3.8453, + "step": 35450 + }, + { + "epoch": 2.4089550210626443, + "grad_norm": 0.8447473645210266, + "learning_rate": 0.0006988806223671694, + "loss": 3.5738, + "step": 35455 + }, + { + "epoch": 2.4092947411333063, + "grad_norm": 0.8300119638442993, + "learning_rate": 0.0006988381573583368, + "loss": 3.755, + "step": 35460 + }, + { + "epoch": 2.409634461203968, + "grad_norm": 1.1039456129074097, + "learning_rate": 0.000698795692349504, + "loss": 3.4086, + "step": 35465 + }, + { + "epoch": 2.4099741812746296, + "grad_norm": 0.6949214339256287, + "learning_rate": 0.0006987532273406712, + "loss": 3.6585, + "step": 35470 + }, + { + "epoch": 2.4103139013452917, + "grad_norm": 0.8014123439788818, + "learning_rate": 0.0006987107623318386, + "loss": 3.5212, + "step": 35475 + }, + { + "epoch": 2.4106536214159533, + "grad_norm": 0.6742558479309082, + "learning_rate": 0.0006986682973230059, + "loss": 3.5719, + "step": 35480 + }, + { + "epoch": 2.410993341486615, + "grad_norm": 0.9560319185256958, + "learning_rate": 0.0006986258323141731, + "loss": 3.6049, + "step": 35485 + }, + { + "epoch": 2.4113330615572766, + "grad_norm": 0.7446780800819397, + "learning_rate": 0.0006985833673053405, + "loss": 3.7649, + "step": 35490 + }, + { + "epoch": 2.4116727816279386, + "grad_norm": 0.9310965538024902, + "learning_rate": 0.0006985409022965077, + "loss": 3.7156, + "step": 35495 + }, + { + "epoch": 2.4120125016986003, + "grad_norm": 0.9593939185142517, + "learning_rate": 0.0006984984372876749, + "loss": 3.7625, + "step": 35500 + }, + { + "epoch": 2.412352221769262, + "grad_norm": 0.8953796625137329, + "learning_rate": 0.0006984559722788422, + "loss": 3.291, + "step": 35505 + }, + { + "epoch": 2.412691941839924, + "grad_norm": 0.8510547280311584, + "learning_rate": 0.0006984135072700095, + "loss": 3.5676, + "step": 35510 + }, + { + "epoch": 2.4130316619105856, + "grad_norm": 0.7069025635719299, + "learning_rate": 0.0006983710422611768, + "loss": 3.4771, + "step": 35515 + }, + { + "epoch": 2.4133713819812472, + "grad_norm": 1.0682461261749268, + "learning_rate": 0.0006983285772523441, + "loss": 3.8241, + "step": 35520 + }, + { + "epoch": 2.4137111020519093, + "grad_norm": 1.0669869184494019, + "learning_rate": 0.0006982861122435114, + "loss": 3.5641, + "step": 35525 + }, + { + "epoch": 2.414050822122571, + "grad_norm": 0.9785784482955933, + "learning_rate": 0.0006982436472346786, + "loss": 3.7611, + "step": 35530 + }, + { + "epoch": 2.4143905421932326, + "grad_norm": 0.653236985206604, + "learning_rate": 0.0006982011822258459, + "loss": 3.706, + "step": 35535 + }, + { + "epoch": 2.4147302622638946, + "grad_norm": 0.8143099546432495, + "learning_rate": 0.0006981587172170131, + "loss": 3.6114, + "step": 35540 + }, + { + "epoch": 2.4150699823345563, + "grad_norm": 0.8603134751319885, + "learning_rate": 0.0006981162522081804, + "loss": 3.4839, + "step": 35545 + }, + { + "epoch": 2.415409702405218, + "grad_norm": 0.932726263999939, + "learning_rate": 0.0006980737871993478, + "loss": 3.397, + "step": 35550 + }, + { + "epoch": 2.41574942247588, + "grad_norm": 0.9290340542793274, + "learning_rate": 0.000698031322190515, + "loss": 3.7343, + "step": 35555 + }, + { + "epoch": 2.4160891425465416, + "grad_norm": 0.7325130105018616, + "learning_rate": 0.0006979888571816823, + "loss": 3.6894, + "step": 35560 + }, + { + "epoch": 2.4164288626172032, + "grad_norm": 0.7920358777046204, + "learning_rate": 0.0006979463921728496, + "loss": 3.4905, + "step": 35565 + }, + { + "epoch": 2.4167685826878653, + "grad_norm": 0.9487249851226807, + "learning_rate": 0.0006979039271640168, + "loss": 3.4996, + "step": 35570 + }, + { + "epoch": 2.417108302758527, + "grad_norm": 0.6337518095970154, + "learning_rate": 0.0006978614621551841, + "loss": 3.4349, + "step": 35575 + }, + { + "epoch": 2.4174480228291886, + "grad_norm": 1.0120391845703125, + "learning_rate": 0.0006978189971463514, + "loss": 3.5619, + "step": 35580 + }, + { + "epoch": 2.4177877428998507, + "grad_norm": 0.7809683680534363, + "learning_rate": 0.0006977765321375187, + "loss": 3.7061, + "step": 35585 + }, + { + "epoch": 2.4181274629705123, + "grad_norm": 0.7638624310493469, + "learning_rate": 0.000697734067128686, + "loss": 3.5365, + "step": 35590 + }, + { + "epoch": 2.418467183041174, + "grad_norm": 1.234602928161621, + "learning_rate": 0.0006976916021198533, + "loss": 3.7469, + "step": 35595 + }, + { + "epoch": 2.418806903111836, + "grad_norm": 0.982877790927887, + "learning_rate": 0.0006976491371110205, + "loss": 3.6337, + "step": 35600 + }, + { + "epoch": 2.4191466231824976, + "grad_norm": 0.6958208084106445, + "learning_rate": 0.0006976066721021878, + "loss": 3.651, + "step": 35605 + }, + { + "epoch": 2.4194863432531593, + "grad_norm": 0.7958633303642273, + "learning_rate": 0.0006975642070933551, + "loss": 3.5346, + "step": 35610 + }, + { + "epoch": 2.4198260633238213, + "grad_norm": 0.9256461262702942, + "learning_rate": 0.0006975217420845223, + "loss": 3.696, + "step": 35615 + }, + { + "epoch": 2.420165783394483, + "grad_norm": 0.7455571889877319, + "learning_rate": 0.0006974792770756897, + "loss": 3.653, + "step": 35620 + }, + { + "epoch": 2.4205055034651446, + "grad_norm": 0.8943120837211609, + "learning_rate": 0.000697436812066857, + "loss": 3.4327, + "step": 35625 + }, + { + "epoch": 2.4208452235358067, + "grad_norm": 1.0734084844589233, + "learning_rate": 0.0006973943470580242, + "loss": 3.56, + "step": 35630 + }, + { + "epoch": 2.4211849436064683, + "grad_norm": 0.7935028076171875, + "learning_rate": 0.0006973518820491915, + "loss": 3.7503, + "step": 35635 + }, + { + "epoch": 2.42152466367713, + "grad_norm": 0.8489923477172852, + "learning_rate": 0.0006973094170403588, + "loss": 3.4397, + "step": 35640 + }, + { + "epoch": 2.421864383747792, + "grad_norm": 0.6729536056518555, + "learning_rate": 0.000697266952031526, + "loss": 3.4819, + "step": 35645 + }, + { + "epoch": 2.4222041038184536, + "grad_norm": 0.845465362071991, + "learning_rate": 0.0006972244870226934, + "loss": 3.4502, + "step": 35650 + }, + { + "epoch": 2.4225438238891153, + "grad_norm": 0.7607055306434631, + "learning_rate": 0.0006971820220138606, + "loss": 3.35, + "step": 35655 + }, + { + "epoch": 2.4228835439597773, + "grad_norm": 1.330264687538147, + "learning_rate": 0.0006971395570050279, + "loss": 3.7882, + "step": 35660 + }, + { + "epoch": 2.423223264030439, + "grad_norm": 0.7368998527526855, + "learning_rate": 0.0006970970919961952, + "loss": 3.6089, + "step": 35665 + }, + { + "epoch": 2.4235629841011006, + "grad_norm": 1.0352815389633179, + "learning_rate": 0.0006970546269873624, + "loss": 3.7088, + "step": 35670 + }, + { + "epoch": 2.4239027041717627, + "grad_norm": 1.2250632047653198, + "learning_rate": 0.0006970121619785297, + "loss": 3.499, + "step": 35675 + }, + { + "epoch": 2.4242424242424243, + "grad_norm": 0.9267789125442505, + "learning_rate": 0.000696969696969697, + "loss": 3.5728, + "step": 35680 + }, + { + "epoch": 2.424582144313086, + "grad_norm": 1.1931495666503906, + "learning_rate": 0.0006969272319608643, + "loss": 3.546, + "step": 35685 + }, + { + "epoch": 2.424921864383748, + "grad_norm": 0.9957852959632874, + "learning_rate": 0.0006968847669520316, + "loss": 3.5427, + "step": 35690 + }, + { + "epoch": 2.4252615844544096, + "grad_norm": 0.9141325950622559, + "learning_rate": 0.0006968423019431989, + "loss": 3.4356, + "step": 35695 + }, + { + "epoch": 2.4256013045250713, + "grad_norm": 0.9408026337623596, + "learning_rate": 0.0006967998369343661, + "loss": 3.5648, + "step": 35700 + }, + { + "epoch": 2.4259410245957334, + "grad_norm": 0.9678852558135986, + "learning_rate": 0.0006967573719255333, + "loss": 3.103, + "step": 35705 + }, + { + "epoch": 2.426280744666395, + "grad_norm": 0.8087712526321411, + "learning_rate": 0.0006967149069167007, + "loss": 3.8023, + "step": 35710 + }, + { + "epoch": 2.4266204647370566, + "grad_norm": 0.732836127281189, + "learning_rate": 0.0006966724419078679, + "loss": 3.4035, + "step": 35715 + }, + { + "epoch": 2.4269601848077182, + "grad_norm": 0.8847925662994385, + "learning_rate": 0.0006966299768990352, + "loss": 3.6221, + "step": 35720 + }, + { + "epoch": 2.4272999048783803, + "grad_norm": 0.7148637771606445, + "learning_rate": 0.0006965875118902026, + "loss": 3.3436, + "step": 35725 + }, + { + "epoch": 2.427639624949042, + "grad_norm": 0.869138240814209, + "learning_rate": 0.0006965450468813698, + "loss": 3.7851, + "step": 35730 + }, + { + "epoch": 2.4279793450197036, + "grad_norm": 0.785347580909729, + "learning_rate": 0.000696502581872537, + "loss": 3.5752, + "step": 35735 + }, + { + "epoch": 2.4283190650903657, + "grad_norm": 0.7496178150177002, + "learning_rate": 0.0006964601168637044, + "loss": 3.5744, + "step": 35740 + }, + { + "epoch": 2.4286587851610273, + "grad_norm": 0.8821709156036377, + "learning_rate": 0.0006964176518548716, + "loss": 3.339, + "step": 35745 + }, + { + "epoch": 2.428998505231689, + "grad_norm": 0.8500383496284485, + "learning_rate": 0.0006963751868460388, + "loss": 3.6413, + "step": 35750 + }, + { + "epoch": 2.429338225302351, + "grad_norm": 0.9530502557754517, + "learning_rate": 0.0006963327218372062, + "loss": 3.4856, + "step": 35755 + }, + { + "epoch": 2.4296779453730126, + "grad_norm": 0.887913703918457, + "learning_rate": 0.0006962902568283735, + "loss": 3.8364, + "step": 35760 + }, + { + "epoch": 2.4300176654436743, + "grad_norm": 1.0631946325302124, + "learning_rate": 0.0006962477918195407, + "loss": 3.6963, + "step": 35765 + }, + { + "epoch": 2.4303573855143363, + "grad_norm": 0.7874326705932617, + "learning_rate": 0.000696205326810708, + "loss": 3.6464, + "step": 35770 + }, + { + "epoch": 2.430697105584998, + "grad_norm": 0.8797521591186523, + "learning_rate": 0.0006961628618018753, + "loss": 3.6598, + "step": 35775 + }, + { + "epoch": 2.4310368256556596, + "grad_norm": 0.6685046553611755, + "learning_rate": 0.0006961203967930425, + "loss": 3.1654, + "step": 35780 + }, + { + "epoch": 2.4313765457263217, + "grad_norm": 0.9128331542015076, + "learning_rate": 0.0006960779317842098, + "loss": 3.5792, + "step": 35785 + }, + { + "epoch": 2.4317162657969833, + "grad_norm": 0.8362668752670288, + "learning_rate": 0.0006960354667753772, + "loss": 3.5145, + "step": 35790 + }, + { + "epoch": 2.432055985867645, + "grad_norm": 0.8993712067604065, + "learning_rate": 0.0006959930017665444, + "loss": 3.69, + "step": 35795 + }, + { + "epoch": 2.432395705938307, + "grad_norm": 0.8033528923988342, + "learning_rate": 0.0006959505367577117, + "loss": 3.547, + "step": 35800 + }, + { + "epoch": 2.4327354260089686, + "grad_norm": 1.0936760902404785, + "learning_rate": 0.0006959080717488789, + "loss": 3.5311, + "step": 35805 + }, + { + "epoch": 2.4330751460796303, + "grad_norm": 0.7865379452705383, + "learning_rate": 0.0006958656067400462, + "loss": 3.7302, + "step": 35810 + }, + { + "epoch": 2.4334148661502923, + "grad_norm": 0.8467683792114258, + "learning_rate": 0.0006958231417312135, + "loss": 3.4578, + "step": 35815 + }, + { + "epoch": 2.433754586220954, + "grad_norm": 0.720098614692688, + "learning_rate": 0.0006957806767223807, + "loss": 3.7984, + "step": 35820 + }, + { + "epoch": 2.4340943062916156, + "grad_norm": 0.9408493638038635, + "learning_rate": 0.0006957382117135481, + "loss": 3.6219, + "step": 35825 + }, + { + "epoch": 2.4344340263622772, + "grad_norm": 0.7647580504417419, + "learning_rate": 0.0006956957467047154, + "loss": 3.5149, + "step": 35830 + }, + { + "epoch": 2.4347737464329393, + "grad_norm": 0.7991424798965454, + "learning_rate": 0.0006956532816958826, + "loss": 3.3739, + "step": 35835 + }, + { + "epoch": 2.435113466503601, + "grad_norm": 0.8955696225166321, + "learning_rate": 0.0006956108166870498, + "loss": 3.6552, + "step": 35840 + }, + { + "epoch": 2.4354531865742626, + "grad_norm": 1.013723611831665, + "learning_rate": 0.0006955683516782172, + "loss": 3.4633, + "step": 35845 + }, + { + "epoch": 2.4357929066449246, + "grad_norm": 1.0584102869033813, + "learning_rate": 0.0006955258866693844, + "loss": 3.4916, + "step": 35850 + }, + { + "epoch": 2.4361326267155863, + "grad_norm": 0.8208675980567932, + "learning_rate": 0.0006954834216605516, + "loss": 3.6064, + "step": 35855 + }, + { + "epoch": 2.436472346786248, + "grad_norm": 1.0219154357910156, + "learning_rate": 0.0006954409566517191, + "loss": 3.7318, + "step": 35860 + }, + { + "epoch": 2.43681206685691, + "grad_norm": 0.8373352885246277, + "learning_rate": 0.0006953984916428863, + "loss": 3.464, + "step": 35865 + }, + { + "epoch": 2.4371517869275716, + "grad_norm": 0.8732852935791016, + "learning_rate": 0.0006953560266340535, + "loss": 3.5093, + "step": 35870 + }, + { + "epoch": 2.4374915069982332, + "grad_norm": 0.8502609133720398, + "learning_rate": 0.0006953135616252209, + "loss": 3.6487, + "step": 35875 + }, + { + "epoch": 2.4378312270688953, + "grad_norm": 0.8389297723770142, + "learning_rate": 0.0006952710966163881, + "loss": 3.5812, + "step": 35880 + }, + { + "epoch": 2.438170947139557, + "grad_norm": 0.9552708864212036, + "learning_rate": 0.0006952286316075553, + "loss": 3.3851, + "step": 35885 + }, + { + "epoch": 2.4385106672102186, + "grad_norm": 1.0844228267669678, + "learning_rate": 0.0006951861665987226, + "loss": 3.5467, + "step": 35890 + }, + { + "epoch": 2.4388503872808807, + "grad_norm": 0.8239562511444092, + "learning_rate": 0.00069514370158989, + "loss": 3.5824, + "step": 35895 + }, + { + "epoch": 2.4391901073515423, + "grad_norm": 0.8297695517539978, + "learning_rate": 0.0006951012365810572, + "loss": 3.4096, + "step": 35900 + }, + { + "epoch": 2.439529827422204, + "grad_norm": 0.9136099815368652, + "learning_rate": 0.0006950587715722245, + "loss": 3.4727, + "step": 35905 + }, + { + "epoch": 2.439869547492866, + "grad_norm": 0.657521665096283, + "learning_rate": 0.0006950163065633918, + "loss": 3.7835, + "step": 35910 + }, + { + "epoch": 2.4402092675635276, + "grad_norm": 0.933928370475769, + "learning_rate": 0.000694973841554559, + "loss": 3.478, + "step": 35915 + }, + { + "epoch": 2.4405489876341893, + "grad_norm": 0.932222306728363, + "learning_rate": 0.0006949313765457263, + "loss": 3.4499, + "step": 35920 + }, + { + "epoch": 2.4408887077048513, + "grad_norm": 0.8237214684486389, + "learning_rate": 0.0006948889115368936, + "loss": 3.3677, + "step": 35925 + }, + { + "epoch": 2.441228427775513, + "grad_norm": 0.888521671295166, + "learning_rate": 0.0006948464465280609, + "loss": 3.771, + "step": 35930 + }, + { + "epoch": 2.4415681478461746, + "grad_norm": 0.9662396311759949, + "learning_rate": 0.0006948039815192282, + "loss": 3.8254, + "step": 35935 + }, + { + "epoch": 2.4419078679168367, + "grad_norm": 1.012626051902771, + "learning_rate": 0.0006947615165103954, + "loss": 3.4814, + "step": 35940 + }, + { + "epoch": 2.4422475879874983, + "grad_norm": 0.9767229557037354, + "learning_rate": 0.0006947190515015628, + "loss": 3.5873, + "step": 35945 + }, + { + "epoch": 2.44258730805816, + "grad_norm": 1.0281695127487183, + "learning_rate": 0.00069467658649273, + "loss": 3.5043, + "step": 35950 + }, + { + "epoch": 2.442927028128822, + "grad_norm": 0.7039209604263306, + "learning_rate": 0.0006946341214838972, + "loss": 3.7777, + "step": 35955 + }, + { + "epoch": 2.4432667481994836, + "grad_norm": 0.6964161992073059, + "learning_rate": 0.0006945916564750646, + "loss": 3.3807, + "step": 35960 + }, + { + "epoch": 2.4436064682701453, + "grad_norm": 0.7231311798095703, + "learning_rate": 0.0006945491914662319, + "loss": 3.6047, + "step": 35965 + }, + { + "epoch": 2.4439461883408073, + "grad_norm": 0.9180313944816589, + "learning_rate": 0.0006945067264573991, + "loss": 3.4751, + "step": 35970 + }, + { + "epoch": 2.444285908411469, + "grad_norm": 0.6846402287483215, + "learning_rate": 0.0006944642614485665, + "loss": 3.4168, + "step": 35975 + }, + { + "epoch": 2.4446256284821306, + "grad_norm": 0.8182682991027832, + "learning_rate": 0.0006944217964397337, + "loss": 3.2754, + "step": 35980 + }, + { + "epoch": 2.4449653485527927, + "grad_norm": 0.847457766532898, + "learning_rate": 0.0006943793314309009, + "loss": 3.1851, + "step": 35985 + }, + { + "epoch": 2.4453050686234543, + "grad_norm": 1.0363085269927979, + "learning_rate": 0.0006943368664220682, + "loss": 3.7026, + "step": 35990 + }, + { + "epoch": 2.445644788694116, + "grad_norm": 0.9380670189857483, + "learning_rate": 0.0006942944014132355, + "loss": 3.5107, + "step": 35995 + }, + { + "epoch": 2.445984508764778, + "grad_norm": 1.0116405487060547, + "learning_rate": 0.0006942519364044028, + "loss": 3.6311, + "step": 36000 + }, + { + "epoch": 2.4463242288354397, + "grad_norm": 1.0885958671569824, + "learning_rate": 0.0006942094713955701, + "loss": 3.6876, + "step": 36005 + }, + { + "epoch": 2.4466639489061013, + "grad_norm": 0.8997837901115417, + "learning_rate": 0.0006941670063867374, + "loss": 3.6208, + "step": 36010 + }, + { + "epoch": 2.4470036689767634, + "grad_norm": 0.9498965740203857, + "learning_rate": 0.0006941245413779046, + "loss": 3.5735, + "step": 36015 + }, + { + "epoch": 2.447343389047425, + "grad_norm": 0.7127948999404907, + "learning_rate": 0.0006940820763690719, + "loss": 3.3544, + "step": 36020 + }, + { + "epoch": 2.4476831091180866, + "grad_norm": 1.0788519382476807, + "learning_rate": 0.0006940396113602392, + "loss": 3.6056, + "step": 36025 + }, + { + "epoch": 2.4480228291887487, + "grad_norm": 0.8992520570755005, + "learning_rate": 0.0006939971463514064, + "loss": 3.4717, + "step": 36030 + }, + { + "epoch": 2.4483625492594103, + "grad_norm": 0.8551177382469177, + "learning_rate": 0.0006939546813425738, + "loss": 3.5759, + "step": 36035 + }, + { + "epoch": 2.448702269330072, + "grad_norm": 0.8903460502624512, + "learning_rate": 0.000693912216333741, + "loss": 3.4482, + "step": 36040 + }, + { + "epoch": 2.449041989400734, + "grad_norm": 0.7458207011222839, + "learning_rate": 0.0006938697513249083, + "loss": 3.7496, + "step": 36045 + }, + { + "epoch": 2.4493817094713957, + "grad_norm": 0.8646011352539062, + "learning_rate": 0.0006938272863160756, + "loss": 3.6397, + "step": 36050 + }, + { + "epoch": 2.4497214295420573, + "grad_norm": 0.9198480844497681, + "learning_rate": 0.0006937848213072428, + "loss": 3.6274, + "step": 36055 + }, + { + "epoch": 2.4500611496127194, + "grad_norm": 0.9103124737739563, + "learning_rate": 0.0006937423562984101, + "loss": 3.5538, + "step": 36060 + }, + { + "epoch": 2.450400869683381, + "grad_norm": 1.149709939956665, + "learning_rate": 0.0006936998912895774, + "loss": 3.5288, + "step": 36065 + }, + { + "epoch": 2.4507405897540426, + "grad_norm": 0.7911571860313416, + "learning_rate": 0.0006936574262807447, + "loss": 3.5274, + "step": 36070 + }, + { + "epoch": 2.4510803098247043, + "grad_norm": 0.9497578740119934, + "learning_rate": 0.000693614961271912, + "loss": 3.5979, + "step": 36075 + }, + { + "epoch": 2.4514200298953663, + "grad_norm": 0.8061007857322693, + "learning_rate": 0.0006935724962630793, + "loss": 3.721, + "step": 36080 + }, + { + "epoch": 2.451759749966028, + "grad_norm": 0.833150327205658, + "learning_rate": 0.0006935300312542465, + "loss": 3.7713, + "step": 36085 + }, + { + "epoch": 2.4520994700366896, + "grad_norm": 0.76576828956604, + "learning_rate": 0.0006934875662454137, + "loss": 3.3959, + "step": 36090 + }, + { + "epoch": 2.4524391901073517, + "grad_norm": 0.9413695931434631, + "learning_rate": 0.0006934451012365811, + "loss": 3.44, + "step": 36095 + }, + { + "epoch": 2.4527789101780133, + "grad_norm": 0.9613192677497864, + "learning_rate": 0.0006934026362277483, + "loss": 3.4142, + "step": 36100 + }, + { + "epoch": 2.453118630248675, + "grad_norm": 0.7438663840293884, + "learning_rate": 0.0006933601712189156, + "loss": 3.3793, + "step": 36105 + }, + { + "epoch": 2.453458350319337, + "grad_norm": 0.7741310000419617, + "learning_rate": 0.000693317706210083, + "loss": 3.7323, + "step": 36110 + }, + { + "epoch": 2.4537980703899986, + "grad_norm": 0.7437799572944641, + "learning_rate": 0.0006932752412012502, + "loss": 3.7139, + "step": 36115 + }, + { + "epoch": 2.4541377904606603, + "grad_norm": 0.9267513155937195, + "learning_rate": 0.0006932327761924174, + "loss": 3.6203, + "step": 36120 + }, + { + "epoch": 2.4544775105313223, + "grad_norm": 0.9539667367935181, + "learning_rate": 0.0006931903111835848, + "loss": 3.5079, + "step": 36125 + }, + { + "epoch": 2.454817230601984, + "grad_norm": 0.8683731555938721, + "learning_rate": 0.000693147846174752, + "loss": 3.6364, + "step": 36130 + }, + { + "epoch": 2.4551569506726456, + "grad_norm": 0.8454530835151672, + "learning_rate": 0.0006931053811659192, + "loss": 3.6136, + "step": 36135 + }, + { + "epoch": 2.4554966707433077, + "grad_norm": 0.7789711356163025, + "learning_rate": 0.0006930629161570866, + "loss": 3.2908, + "step": 36140 + }, + { + "epoch": 2.4558363908139693, + "grad_norm": 0.9197741150856018, + "learning_rate": 0.0006930204511482539, + "loss": 3.7711, + "step": 36145 + }, + { + "epoch": 2.456176110884631, + "grad_norm": 0.7262634038925171, + "learning_rate": 0.0006929779861394211, + "loss": 3.5818, + "step": 36150 + }, + { + "epoch": 2.456515830955293, + "grad_norm": 0.889327883720398, + "learning_rate": 0.0006929355211305884, + "loss": 3.7181, + "step": 36155 + }, + { + "epoch": 2.4568555510259547, + "grad_norm": 1.0491082668304443, + "learning_rate": 0.0006928930561217557, + "loss": 3.6124, + "step": 36160 + }, + { + "epoch": 2.4571952710966163, + "grad_norm": 0.8459656834602356, + "learning_rate": 0.0006928505911129229, + "loss": 3.8499, + "step": 36165 + }, + { + "epoch": 2.457534991167278, + "grad_norm": 0.8021054863929749, + "learning_rate": 0.0006928081261040902, + "loss": 3.3295, + "step": 36170 + }, + { + "epoch": 2.45787471123794, + "grad_norm": 0.7558298707008362, + "learning_rate": 0.0006927656610952576, + "loss": 3.5904, + "step": 36175 + }, + { + "epoch": 2.4582144313086016, + "grad_norm": 0.8450889587402344, + "learning_rate": 0.0006927231960864248, + "loss": 3.548, + "step": 36180 + }, + { + "epoch": 2.4585541513792633, + "grad_norm": 0.8992059230804443, + "learning_rate": 0.0006926807310775921, + "loss": 3.6591, + "step": 36185 + }, + { + "epoch": 2.4588938714499253, + "grad_norm": 0.6901722550392151, + "learning_rate": 0.0006926382660687593, + "loss": 3.6373, + "step": 36190 + }, + { + "epoch": 2.459233591520587, + "grad_norm": 0.9824018478393555, + "learning_rate": 0.0006925958010599266, + "loss": 3.6866, + "step": 36195 + }, + { + "epoch": 2.4595733115912486, + "grad_norm": 0.9984708428382874, + "learning_rate": 0.0006925533360510939, + "loss": 3.4464, + "step": 36200 + }, + { + "epoch": 2.4599130316619107, + "grad_norm": 0.8927099108695984, + "learning_rate": 0.0006925108710422611, + "loss": 3.909, + "step": 36205 + }, + { + "epoch": 2.4602527517325723, + "grad_norm": 0.7059174180030823, + "learning_rate": 0.0006924684060334285, + "loss": 3.6478, + "step": 36210 + }, + { + "epoch": 2.460592471803234, + "grad_norm": 0.8820169568061829, + "learning_rate": 0.0006924259410245958, + "loss": 3.3821, + "step": 36215 + }, + { + "epoch": 2.460932191873896, + "grad_norm": 0.906986653804779, + "learning_rate": 0.000692383476015763, + "loss": 3.8304, + "step": 36220 + }, + { + "epoch": 2.4612719119445576, + "grad_norm": 0.8624225854873657, + "learning_rate": 0.0006923410110069302, + "loss": 3.3044, + "step": 36225 + }, + { + "epoch": 2.4616116320152193, + "grad_norm": 0.7914595007896423, + "learning_rate": 0.0006922985459980976, + "loss": 3.7538, + "step": 36230 + }, + { + "epoch": 2.4619513520858813, + "grad_norm": 0.6971786618232727, + "learning_rate": 0.0006922560809892648, + "loss": 3.2757, + "step": 36235 + }, + { + "epoch": 2.462291072156543, + "grad_norm": 0.8726166486740112, + "learning_rate": 0.000692213615980432, + "loss": 3.497, + "step": 36240 + }, + { + "epoch": 2.4626307922272046, + "grad_norm": 0.9129696488380432, + "learning_rate": 0.0006921711509715995, + "loss": 3.6021, + "step": 36245 + }, + { + "epoch": 2.4629705122978667, + "grad_norm": 0.9540472030639648, + "learning_rate": 0.0006921286859627667, + "loss": 3.4729, + "step": 36250 + }, + { + "epoch": 2.4633102323685283, + "grad_norm": 1.0409308671951294, + "learning_rate": 0.0006920862209539339, + "loss": 3.4031, + "step": 36255 + }, + { + "epoch": 2.46364995243919, + "grad_norm": 0.8868533968925476, + "learning_rate": 0.0006920437559451013, + "loss": 3.4979, + "step": 36260 + }, + { + "epoch": 2.463989672509852, + "grad_norm": 0.9985135793685913, + "learning_rate": 0.0006920012909362685, + "loss": 3.5361, + "step": 36265 + }, + { + "epoch": 2.4643293925805136, + "grad_norm": 0.865886926651001, + "learning_rate": 0.0006919588259274357, + "loss": 3.7461, + "step": 36270 + }, + { + "epoch": 2.4646691126511753, + "grad_norm": 0.9465520977973938, + "learning_rate": 0.0006919163609186032, + "loss": 3.5689, + "step": 36275 + }, + { + "epoch": 2.4650088327218374, + "grad_norm": 0.9770486950874329, + "learning_rate": 0.0006918738959097704, + "loss": 3.5385, + "step": 36280 + }, + { + "epoch": 2.465348552792499, + "grad_norm": 0.7647587060928345, + "learning_rate": 0.0006918314309009377, + "loss": 3.5755, + "step": 36285 + }, + { + "epoch": 2.4656882728631606, + "grad_norm": 1.102231502532959, + "learning_rate": 0.0006917889658921049, + "loss": 3.6893, + "step": 36290 + }, + { + "epoch": 2.4660279929338227, + "grad_norm": 0.7716435790061951, + "learning_rate": 0.0006917465008832722, + "loss": 3.2765, + "step": 36295 + }, + { + "epoch": 2.4663677130044843, + "grad_norm": 0.928592324256897, + "learning_rate": 0.0006917040358744395, + "loss": 3.5901, + "step": 36300 + }, + { + "epoch": 2.466707433075146, + "grad_norm": 0.67704176902771, + "learning_rate": 0.0006916615708656067, + "loss": 3.4815, + "step": 36305 + }, + { + "epoch": 2.467047153145808, + "grad_norm": 1.0540499687194824, + "learning_rate": 0.0006916191058567741, + "loss": 3.6412, + "step": 36310 + }, + { + "epoch": 2.4673868732164697, + "grad_norm": 0.7599851489067078, + "learning_rate": 0.0006915766408479414, + "loss": 3.8823, + "step": 36315 + }, + { + "epoch": 2.4677265932871313, + "grad_norm": 0.9563323259353638, + "learning_rate": 0.0006915341758391086, + "loss": 3.7295, + "step": 36320 + }, + { + "epoch": 2.4680663133577934, + "grad_norm": 0.899450421333313, + "learning_rate": 0.0006914917108302758, + "loss": 3.6357, + "step": 36325 + }, + { + "epoch": 2.468406033428455, + "grad_norm": 0.8754828572273254, + "learning_rate": 0.0006914492458214432, + "loss": 3.5001, + "step": 36330 + }, + { + "epoch": 2.4687457534991166, + "grad_norm": 0.856401801109314, + "learning_rate": 0.0006914067808126104, + "loss": 3.7415, + "step": 36335 + }, + { + "epoch": 2.4690854735697787, + "grad_norm": 0.8319867849349976, + "learning_rate": 0.0006913643158037776, + "loss": 3.259, + "step": 36340 + }, + { + "epoch": 2.4694251936404403, + "grad_norm": 0.954565703868866, + "learning_rate": 0.0006913218507949451, + "loss": 3.621, + "step": 36345 + }, + { + "epoch": 2.469764913711102, + "grad_norm": 0.9378578662872314, + "learning_rate": 0.0006912793857861123, + "loss": 3.6769, + "step": 36350 + }, + { + "epoch": 2.470104633781764, + "grad_norm": 1.0436452627182007, + "learning_rate": 0.0006912369207772795, + "loss": 3.3329, + "step": 36355 + }, + { + "epoch": 2.4704443538524257, + "grad_norm": 0.8781203031539917, + "learning_rate": 0.0006911944557684469, + "loss": 3.4577, + "step": 36360 + }, + { + "epoch": 2.4707840739230873, + "grad_norm": 0.8413339853286743, + "learning_rate": 0.0006911519907596141, + "loss": 3.7904, + "step": 36365 + }, + { + "epoch": 2.4711237939937494, + "grad_norm": 0.8315607905387878, + "learning_rate": 0.0006911095257507813, + "loss": 3.5522, + "step": 36370 + }, + { + "epoch": 2.471463514064411, + "grad_norm": 0.9372484087944031, + "learning_rate": 0.0006910670607419487, + "loss": 3.4998, + "step": 36375 + }, + { + "epoch": 2.4718032341350726, + "grad_norm": 1.043075442314148, + "learning_rate": 0.000691024595733116, + "loss": 3.6908, + "step": 36380 + }, + { + "epoch": 2.4721429542057347, + "grad_norm": 0.8879507184028625, + "learning_rate": 0.0006909821307242832, + "loss": 3.6829, + "step": 36385 + }, + { + "epoch": 2.4724826742763963, + "grad_norm": 0.8010611534118652, + "learning_rate": 0.0006909396657154505, + "loss": 3.4733, + "step": 36390 + }, + { + "epoch": 2.472822394347058, + "grad_norm": 0.8769936561584473, + "learning_rate": 0.0006908972007066178, + "loss": 3.6313, + "step": 36395 + }, + { + "epoch": 2.47316211441772, + "grad_norm": 0.7876511216163635, + "learning_rate": 0.000690854735697785, + "loss": 3.4161, + "step": 36400 + }, + { + "epoch": 2.4735018344883817, + "grad_norm": 0.8814708590507507, + "learning_rate": 0.0006908122706889523, + "loss": 3.6457, + "step": 36405 + }, + { + "epoch": 2.4738415545590433, + "grad_norm": 0.871041476726532, + "learning_rate": 0.0006907698056801196, + "loss": 3.6814, + "step": 36410 + }, + { + "epoch": 2.474181274629705, + "grad_norm": 1.1853734254837036, + "learning_rate": 0.0006907273406712869, + "loss": 3.5313, + "step": 36415 + }, + { + "epoch": 2.474520994700367, + "grad_norm": 0.8447781205177307, + "learning_rate": 0.0006906848756624542, + "loss": 3.5422, + "step": 36420 + }, + { + "epoch": 2.4748607147710286, + "grad_norm": 0.8156052231788635, + "learning_rate": 0.0006906424106536215, + "loss": 3.4339, + "step": 36425 + }, + { + "epoch": 2.4752004348416903, + "grad_norm": 0.7409082055091858, + "learning_rate": 0.0006905999456447887, + "loss": 3.5654, + "step": 36430 + }, + { + "epoch": 2.4755401549123524, + "grad_norm": 0.7991155385971069, + "learning_rate": 0.000690557480635956, + "loss": 3.6512, + "step": 36435 + }, + { + "epoch": 2.475879874983014, + "grad_norm": 0.8262657523155212, + "learning_rate": 0.0006905150156271232, + "loss": 3.4858, + "step": 36440 + }, + { + "epoch": 2.4762195950536756, + "grad_norm": 0.8796854615211487, + "learning_rate": 0.0006904725506182905, + "loss": 3.7319, + "step": 36445 + }, + { + "epoch": 2.4765593151243377, + "grad_norm": 0.6967483162879944, + "learning_rate": 0.0006904300856094579, + "loss": 3.7002, + "step": 36450 + }, + { + "epoch": 2.4768990351949993, + "grad_norm": 1.10507071018219, + "learning_rate": 0.0006903876206006251, + "loss": 3.7097, + "step": 36455 + }, + { + "epoch": 2.477238755265661, + "grad_norm": 0.8796741962432861, + "learning_rate": 0.0006903451555917924, + "loss": 3.5756, + "step": 36460 + }, + { + "epoch": 2.477578475336323, + "grad_norm": 0.8185814023017883, + "learning_rate": 0.0006903026905829597, + "loss": 3.4423, + "step": 36465 + }, + { + "epoch": 2.4779181954069847, + "grad_norm": 0.8333466649055481, + "learning_rate": 0.0006902602255741269, + "loss": 3.4952, + "step": 36470 + }, + { + "epoch": 2.4782579154776463, + "grad_norm": 0.8334892392158508, + "learning_rate": 0.0006902177605652941, + "loss": 3.6537, + "step": 36475 + }, + { + "epoch": 2.4785976355483084, + "grad_norm": 0.9300302863121033, + "learning_rate": 0.0006901752955564615, + "loss": 3.4466, + "step": 36480 + }, + { + "epoch": 2.47893735561897, + "grad_norm": 0.9987243413925171, + "learning_rate": 0.0006901328305476288, + "loss": 3.4529, + "step": 36485 + }, + { + "epoch": 2.4792770756896316, + "grad_norm": 1.0005249977111816, + "learning_rate": 0.000690090365538796, + "loss": 3.3849, + "step": 36490 + }, + { + "epoch": 2.4796167957602937, + "grad_norm": 1.2294481992721558, + "learning_rate": 0.0006900479005299634, + "loss": 3.6634, + "step": 36495 + }, + { + "epoch": 2.4799565158309553, + "grad_norm": 0.8743250966072083, + "learning_rate": 0.0006900054355211306, + "loss": 3.3456, + "step": 36500 + }, + { + "epoch": 2.480296235901617, + "grad_norm": 0.8085228800773621, + "learning_rate": 0.0006899629705122978, + "loss": 3.6539, + "step": 36505 + }, + { + "epoch": 2.4806359559722786, + "grad_norm": 0.7113310098648071, + "learning_rate": 0.0006899205055034652, + "loss": 3.5481, + "step": 36510 + }, + { + "epoch": 2.4809756760429407, + "grad_norm": 1.0010929107666016, + "learning_rate": 0.0006898780404946324, + "loss": 3.4007, + "step": 36515 + }, + { + "epoch": 2.4813153961136023, + "grad_norm": 0.7953394055366516, + "learning_rate": 0.0006898355754857997, + "loss": 3.4604, + "step": 36520 + }, + { + "epoch": 2.481655116184264, + "grad_norm": 0.8871358036994934, + "learning_rate": 0.000689793110476967, + "loss": 3.648, + "step": 36525 + }, + { + "epoch": 2.481994836254926, + "grad_norm": 0.976520836353302, + "learning_rate": 0.0006897506454681343, + "loss": 3.6012, + "step": 36530 + }, + { + "epoch": 2.4823345563255876, + "grad_norm": 0.7046443819999695, + "learning_rate": 0.0006897081804593015, + "loss": 3.6521, + "step": 36535 + }, + { + "epoch": 2.4826742763962493, + "grad_norm": 0.7388847470283508, + "learning_rate": 0.0006896657154504688, + "loss": 3.3058, + "step": 36540 + }, + { + "epoch": 2.4830139964669113, + "grad_norm": 0.8523498773574829, + "learning_rate": 0.0006896232504416361, + "loss": 3.5752, + "step": 36545 + }, + { + "epoch": 2.483353716537573, + "grad_norm": 0.938737154006958, + "learning_rate": 0.0006895807854328033, + "loss": 3.459, + "step": 36550 + }, + { + "epoch": 2.4836934366082346, + "grad_norm": 0.8756615519523621, + "learning_rate": 0.0006895383204239707, + "loss": 3.4447, + "step": 36555 + }, + { + "epoch": 2.4840331566788967, + "grad_norm": 0.8922083377838135, + "learning_rate": 0.000689495855415138, + "loss": 3.3873, + "step": 36560 + }, + { + "epoch": 2.4843728767495583, + "grad_norm": 0.9193309545516968, + "learning_rate": 0.0006894533904063052, + "loss": 3.3683, + "step": 36565 + }, + { + "epoch": 2.48471259682022, + "grad_norm": 0.8712266087532043, + "learning_rate": 0.0006894109253974725, + "loss": 3.3698, + "step": 36570 + }, + { + "epoch": 2.485052316890882, + "grad_norm": 0.853512704372406, + "learning_rate": 0.0006893684603886397, + "loss": 3.7681, + "step": 36575 + }, + { + "epoch": 2.4853920369615436, + "grad_norm": 0.9410825967788696, + "learning_rate": 0.000689325995379807, + "loss": 3.7854, + "step": 36580 + }, + { + "epoch": 2.4857317570322053, + "grad_norm": 0.708361029624939, + "learning_rate": 0.0006892835303709743, + "loss": 3.4626, + "step": 36585 + }, + { + "epoch": 2.4860714771028674, + "grad_norm": 0.8282793164253235, + "learning_rate": 0.0006892410653621416, + "loss": 3.6188, + "step": 36590 + }, + { + "epoch": 2.486411197173529, + "grad_norm": 0.9518354535102844, + "learning_rate": 0.0006891986003533089, + "loss": 3.2638, + "step": 36595 + }, + { + "epoch": 2.4867509172441906, + "grad_norm": 0.942774772644043, + "learning_rate": 0.0006891561353444762, + "loss": 3.6533, + "step": 36600 + }, + { + "epoch": 2.4870906373148527, + "grad_norm": 0.7877184152603149, + "learning_rate": 0.0006891136703356434, + "loss": 3.4383, + "step": 36605 + }, + { + "epoch": 2.4874303573855143, + "grad_norm": 0.8960031867027283, + "learning_rate": 0.0006890712053268107, + "loss": 3.7348, + "step": 36610 + }, + { + "epoch": 2.487770077456176, + "grad_norm": 0.8156405091285706, + "learning_rate": 0.000689028740317978, + "loss": 3.7056, + "step": 36615 + }, + { + "epoch": 2.488109797526838, + "grad_norm": 1.0419365167617798, + "learning_rate": 0.0006889862753091452, + "loss": 3.5797, + "step": 36620 + }, + { + "epoch": 2.4884495175974997, + "grad_norm": 0.7346763610839844, + "learning_rate": 0.0006889438103003127, + "loss": 3.6299, + "step": 36625 + }, + { + "epoch": 2.4887892376681613, + "grad_norm": 0.9340589642524719, + "learning_rate": 0.0006889013452914799, + "loss": 3.4286, + "step": 36630 + }, + { + "epoch": 2.4891289577388234, + "grad_norm": 0.9242081046104431, + "learning_rate": 0.0006888588802826471, + "loss": 3.6487, + "step": 36635 + }, + { + "epoch": 2.489468677809485, + "grad_norm": 1.1686097383499146, + "learning_rate": 0.0006888164152738144, + "loss": 3.5239, + "step": 36640 + }, + { + "epoch": 2.4898083978801466, + "grad_norm": 0.7603017091751099, + "learning_rate": 0.0006887739502649817, + "loss": 3.5929, + "step": 36645 + }, + { + "epoch": 2.4901481179508087, + "grad_norm": 0.8958903551101685, + "learning_rate": 0.0006887314852561489, + "loss": 3.4947, + "step": 36650 + }, + { + "epoch": 2.4904878380214703, + "grad_norm": 0.86673504114151, + "learning_rate": 0.0006886890202473162, + "loss": 3.5726, + "step": 36655 + }, + { + "epoch": 2.490827558092132, + "grad_norm": 1.0016387701034546, + "learning_rate": 0.0006886465552384836, + "loss": 3.3194, + "step": 36660 + }, + { + "epoch": 2.491167278162794, + "grad_norm": 0.9451571106910706, + "learning_rate": 0.0006886040902296508, + "loss": 3.3076, + "step": 36665 + }, + { + "epoch": 2.4915069982334557, + "grad_norm": 1.0089694261550903, + "learning_rate": 0.0006885616252208181, + "loss": 3.4043, + "step": 36670 + }, + { + "epoch": 2.4918467183041173, + "grad_norm": 0.8287688493728638, + "learning_rate": 0.0006885191602119853, + "loss": 3.4409, + "step": 36675 + }, + { + "epoch": 2.4921864383747794, + "grad_norm": 0.975384533405304, + "learning_rate": 0.0006884766952031526, + "loss": 3.4429, + "step": 36680 + }, + { + "epoch": 2.492526158445441, + "grad_norm": 0.8510820269584656, + "learning_rate": 0.0006884342301943199, + "loss": 3.4138, + "step": 36685 + }, + { + "epoch": 2.4928658785161026, + "grad_norm": 1.0971518754959106, + "learning_rate": 0.0006883917651854871, + "loss": 3.3162, + "step": 36690 + }, + { + "epoch": 2.4932055985867647, + "grad_norm": 0.7318975925445557, + "learning_rate": 0.0006883493001766545, + "loss": 3.6157, + "step": 36695 + }, + { + "epoch": 2.4935453186574263, + "grad_norm": 0.8650673627853394, + "learning_rate": 0.0006883068351678218, + "loss": 3.5632, + "step": 36700 + }, + { + "epoch": 2.493885038728088, + "grad_norm": 1.056884765625, + "learning_rate": 0.000688264370158989, + "loss": 3.3088, + "step": 36705 + }, + { + "epoch": 2.49422475879875, + "grad_norm": 0.9452702403068542, + "learning_rate": 0.0006882219051501563, + "loss": 3.6457, + "step": 36710 + }, + { + "epoch": 2.4945644788694117, + "grad_norm": 0.7401376366615295, + "learning_rate": 0.0006881794401413236, + "loss": 3.7734, + "step": 36715 + }, + { + "epoch": 2.4949041989400733, + "grad_norm": 0.832613468170166, + "learning_rate": 0.0006881369751324908, + "loss": 3.3839, + "step": 36720 + }, + { + "epoch": 2.4952439190107354, + "grad_norm": 1.0046380758285522, + "learning_rate": 0.000688094510123658, + "loss": 3.2739, + "step": 36725 + }, + { + "epoch": 2.495583639081397, + "grad_norm": 0.7986001372337341, + "learning_rate": 0.0006880520451148255, + "loss": 3.6226, + "step": 36730 + }, + { + "epoch": 2.4959233591520587, + "grad_norm": 1.242040991783142, + "learning_rate": 0.0006880095801059927, + "loss": 3.8644, + "step": 36735 + }, + { + "epoch": 2.4962630792227207, + "grad_norm": 0.9378507733345032, + "learning_rate": 0.0006879671150971599, + "loss": 3.4524, + "step": 36740 + }, + { + "epoch": 2.4966027992933824, + "grad_norm": 0.9926351308822632, + "learning_rate": 0.0006879246500883273, + "loss": 3.4312, + "step": 36745 + }, + { + "epoch": 2.496942519364044, + "grad_norm": 0.7548143863677979, + "learning_rate": 0.0006878821850794945, + "loss": 3.7, + "step": 36750 + }, + { + "epoch": 2.4972822394347056, + "grad_norm": 0.8410499691963196, + "learning_rate": 0.0006878397200706617, + "loss": 3.3711, + "step": 36755 + }, + { + "epoch": 2.4976219595053677, + "grad_norm": 0.6862537860870361, + "learning_rate": 0.0006877972550618292, + "loss": 3.4216, + "step": 36760 + }, + { + "epoch": 2.4979616795760293, + "grad_norm": 0.722764253616333, + "learning_rate": 0.0006877547900529964, + "loss": 3.4782, + "step": 36765 + }, + { + "epoch": 2.498301399646691, + "grad_norm": 0.6903204917907715, + "learning_rate": 0.0006877123250441636, + "loss": 3.5483, + "step": 36770 + }, + { + "epoch": 2.498641119717353, + "grad_norm": 0.8590174317359924, + "learning_rate": 0.000687669860035331, + "loss": 3.3938, + "step": 36775 + }, + { + "epoch": 2.4989808397880147, + "grad_norm": 0.7185794115066528, + "learning_rate": 0.0006876273950264982, + "loss": 3.3916, + "step": 36780 + }, + { + "epoch": 2.4993205598586763, + "grad_norm": 0.9417985081672668, + "learning_rate": 0.0006875849300176654, + "loss": 3.5408, + "step": 36785 + }, + { + "epoch": 2.4996602799293384, + "grad_norm": 0.9269102811813354, + "learning_rate": 0.0006875424650088327, + "loss": 3.4685, + "step": 36790 + }, + { + "epoch": 2.5, + "grad_norm": 0.8657393455505371, + "learning_rate": 0.0006875, + "loss": 3.48, + "step": 36795 + }, + { + "epoch": 2.5003397200706616, + "grad_norm": 1.022359013557434, + "learning_rate": 0.0006874575349911673, + "loss": 3.6529, + "step": 36800 + }, + { + "epoch": 2.5006794401413237, + "grad_norm": 0.8003290295600891, + "learning_rate": 0.0006874150699823346, + "loss": 3.7177, + "step": 36805 + }, + { + "epoch": 2.5010191602119853, + "grad_norm": 0.8137874007225037, + "learning_rate": 0.0006873726049735019, + "loss": 3.6766, + "step": 36810 + }, + { + "epoch": 2.501358880282647, + "grad_norm": 0.9338594675064087, + "learning_rate": 0.0006873301399646691, + "loss": 3.4237, + "step": 36815 + }, + { + "epoch": 2.5016986003533086, + "grad_norm": 0.8676313161849976, + "learning_rate": 0.0006872876749558364, + "loss": 3.6658, + "step": 36820 + }, + { + "epoch": 2.5020383204239707, + "grad_norm": 0.7515265345573425, + "learning_rate": 0.0006872452099470036, + "loss": 3.5861, + "step": 36825 + }, + { + "epoch": 2.5023780404946323, + "grad_norm": 1.1176272630691528, + "learning_rate": 0.0006872027449381709, + "loss": 3.5769, + "step": 36830 + }, + { + "epoch": 2.502717760565294, + "grad_norm": 0.8808847069740295, + "learning_rate": 0.0006871602799293383, + "loss": 3.5341, + "step": 36835 + }, + { + "epoch": 2.503057480635956, + "grad_norm": 0.6433759927749634, + "learning_rate": 0.0006871178149205055, + "loss": 3.6638, + "step": 36840 + }, + { + "epoch": 2.5033972007066176, + "grad_norm": 1.1460437774658203, + "learning_rate": 0.0006870753499116728, + "loss": 3.6047, + "step": 36845 + }, + { + "epoch": 2.5037369207772793, + "grad_norm": 0.7882977724075317, + "learning_rate": 0.0006870328849028401, + "loss": 3.546, + "step": 36850 + }, + { + "epoch": 2.5040766408479413, + "grad_norm": 0.9900385737419128, + "learning_rate": 0.0006869904198940073, + "loss": 3.3811, + "step": 36855 + }, + { + "epoch": 2.504416360918603, + "grad_norm": 0.8787567019462585, + "learning_rate": 0.0006869479548851745, + "loss": 3.7839, + "step": 36860 + }, + { + "epoch": 2.5047560809892646, + "grad_norm": 0.8779926300048828, + "learning_rate": 0.000686905489876342, + "loss": 3.6486, + "step": 36865 + }, + { + "epoch": 2.5050958010599267, + "grad_norm": 0.7147447466850281, + "learning_rate": 0.0006868630248675092, + "loss": 3.2132, + "step": 36870 + }, + { + "epoch": 2.5054355211305883, + "grad_norm": 0.9452587366104126, + "learning_rate": 0.0006868205598586764, + "loss": 3.3614, + "step": 36875 + }, + { + "epoch": 2.50577524120125, + "grad_norm": 0.7602297067642212, + "learning_rate": 0.0006867780948498438, + "loss": 3.4661, + "step": 36880 + }, + { + "epoch": 2.506114961271912, + "grad_norm": 0.8211740851402283, + "learning_rate": 0.000686735629841011, + "loss": 3.5463, + "step": 36885 + }, + { + "epoch": 2.5064546813425737, + "grad_norm": 0.9510506987571716, + "learning_rate": 0.0006866931648321782, + "loss": 3.4778, + "step": 36890 + }, + { + "epoch": 2.5067944014132353, + "grad_norm": 0.8949901461601257, + "learning_rate": 0.0006866506998233456, + "loss": 3.3598, + "step": 36895 + }, + { + "epoch": 2.5071341214838974, + "grad_norm": 1.0402981042861938, + "learning_rate": 0.0006866082348145129, + "loss": 3.3759, + "step": 36900 + }, + { + "epoch": 2.507473841554559, + "grad_norm": 0.8432304859161377, + "learning_rate": 0.0006865657698056801, + "loss": 3.6168, + "step": 36905 + }, + { + "epoch": 2.5078135616252206, + "grad_norm": 0.6661829352378845, + "learning_rate": 0.0006865233047968475, + "loss": 3.4575, + "step": 36910 + }, + { + "epoch": 2.5081532816958827, + "grad_norm": 1.1519736051559448, + "learning_rate": 0.0006864808397880147, + "loss": 3.5699, + "step": 36915 + }, + { + "epoch": 2.5084930017665443, + "grad_norm": 0.7281892895698547, + "learning_rate": 0.0006864383747791819, + "loss": 3.3514, + "step": 36920 + }, + { + "epoch": 2.508832721837206, + "grad_norm": 0.8577595949172974, + "learning_rate": 0.0006863959097703492, + "loss": 3.7121, + "step": 36925 + }, + { + "epoch": 2.509172441907868, + "grad_norm": 0.7311611175537109, + "learning_rate": 0.0006863534447615165, + "loss": 3.3516, + "step": 36930 + }, + { + "epoch": 2.5095121619785297, + "grad_norm": 1.0527610778808594, + "learning_rate": 0.0006863109797526838, + "loss": 3.7124, + "step": 36935 + }, + { + "epoch": 2.5098518820491913, + "grad_norm": 0.7159387469291687, + "learning_rate": 0.0006862685147438511, + "loss": 3.6336, + "step": 36940 + }, + { + "epoch": 2.5101916021198534, + "grad_norm": 0.9110502004623413, + "learning_rate": 0.0006862260497350184, + "loss": 3.6723, + "step": 36945 + }, + { + "epoch": 2.510531322190515, + "grad_norm": 1.5016603469848633, + "learning_rate": 0.0006861835847261856, + "loss": 3.4908, + "step": 36950 + }, + { + "epoch": 2.5108710422611766, + "grad_norm": 1.0306766033172607, + "learning_rate": 0.0006861411197173529, + "loss": 3.6121, + "step": 36955 + }, + { + "epoch": 2.5112107623318387, + "grad_norm": 0.5805943608283997, + "learning_rate": 0.0006860986547085201, + "loss": 3.5461, + "step": 36960 + }, + { + "epoch": 2.5115504824025003, + "grad_norm": 0.9453427791595459, + "learning_rate": 0.0006860561896996874, + "loss": 3.401, + "step": 36965 + }, + { + "epoch": 2.511890202473162, + "grad_norm": 0.7779866456985474, + "learning_rate": 0.0006860137246908548, + "loss": 3.5209, + "step": 36970 + }, + { + "epoch": 2.512229922543824, + "grad_norm": 0.7268861532211304, + "learning_rate": 0.000685971259682022, + "loss": 3.744, + "step": 36975 + }, + { + "epoch": 2.5125696426144857, + "grad_norm": 0.7437829971313477, + "learning_rate": 0.0006859287946731894, + "loss": 3.4672, + "step": 36980 + }, + { + "epoch": 2.5129093626851473, + "grad_norm": 0.9142969846725464, + "learning_rate": 0.0006858863296643566, + "loss": 3.588, + "step": 36985 + }, + { + "epoch": 2.5132490827558094, + "grad_norm": 0.8910629153251648, + "learning_rate": 0.0006858438646555238, + "loss": 3.896, + "step": 36990 + }, + { + "epoch": 2.513588802826471, + "grad_norm": 0.7418274879455566, + "learning_rate": 0.0006858013996466912, + "loss": 3.8323, + "step": 36995 + }, + { + "epoch": 2.5139285228971326, + "grad_norm": 0.8523584008216858, + "learning_rate": 0.0006857589346378584, + "loss": 3.517, + "step": 37000 + }, + { + "epoch": 2.5142682429677947, + "grad_norm": 0.7802397608757019, + "learning_rate": 0.0006857164696290257, + "loss": 3.9053, + "step": 37005 + }, + { + "epoch": 2.5146079630384564, + "grad_norm": 0.6620455384254456, + "learning_rate": 0.0006856740046201931, + "loss": 3.3909, + "step": 37010 + }, + { + "epoch": 2.514947683109118, + "grad_norm": 0.7681835889816284, + "learning_rate": 0.0006856315396113603, + "loss": 3.5391, + "step": 37015 + }, + { + "epoch": 2.51528740317978, + "grad_norm": 0.7635166049003601, + "learning_rate": 0.0006855890746025275, + "loss": 3.4221, + "step": 37020 + }, + { + "epoch": 2.5156271232504417, + "grad_norm": 0.865240216255188, + "learning_rate": 0.0006855466095936948, + "loss": 3.4297, + "step": 37025 + }, + { + "epoch": 2.5159668433211033, + "grad_norm": 0.6793370246887207, + "learning_rate": 0.0006855041445848621, + "loss": 3.4611, + "step": 37030 + }, + { + "epoch": 2.5163065633917654, + "grad_norm": 0.8803624510765076, + "learning_rate": 0.0006854616795760293, + "loss": 3.4582, + "step": 37035 + }, + { + "epoch": 2.516646283462427, + "grad_norm": 0.7430471777915955, + "learning_rate": 0.0006854192145671967, + "loss": 3.5829, + "step": 37040 + }, + { + "epoch": 2.5169860035330887, + "grad_norm": 0.9532375931739807, + "learning_rate": 0.000685376749558364, + "loss": 3.4732, + "step": 37045 + }, + { + "epoch": 2.5173257236037507, + "grad_norm": 1.0152339935302734, + "learning_rate": 0.0006853342845495312, + "loss": 3.6815, + "step": 37050 + }, + { + "epoch": 2.5176654436744124, + "grad_norm": 1.0876001119613647, + "learning_rate": 0.0006852918195406985, + "loss": 3.6239, + "step": 37055 + }, + { + "epoch": 2.518005163745074, + "grad_norm": 0.9778195023536682, + "learning_rate": 0.0006852493545318657, + "loss": 3.3743, + "step": 37060 + }, + { + "epoch": 2.518344883815736, + "grad_norm": 0.7446442246437073, + "learning_rate": 0.000685206889523033, + "loss": 3.6971, + "step": 37065 + }, + { + "epoch": 2.5186846038863977, + "grad_norm": 0.823059618473053, + "learning_rate": 0.0006851644245142003, + "loss": 3.5565, + "step": 37070 + }, + { + "epoch": 2.5190243239570593, + "grad_norm": 0.7140182256698608, + "learning_rate": 0.0006851219595053676, + "loss": 3.241, + "step": 37075 + }, + { + "epoch": 2.5193640440277214, + "grad_norm": 1.0930211544036865, + "learning_rate": 0.0006850794944965349, + "loss": 3.6895, + "step": 37080 + }, + { + "epoch": 2.519703764098383, + "grad_norm": 0.733505368232727, + "learning_rate": 0.0006850370294877022, + "loss": 3.4622, + "step": 37085 + }, + { + "epoch": 2.5200434841690447, + "grad_norm": 0.7258961200714111, + "learning_rate": 0.0006849945644788694, + "loss": 3.5916, + "step": 37090 + }, + { + "epoch": 2.5203832042397067, + "grad_norm": 1.1903328895568848, + "learning_rate": 0.0006849520994700367, + "loss": 3.7739, + "step": 37095 + }, + { + "epoch": 2.5207229243103684, + "grad_norm": 0.8272022604942322, + "learning_rate": 0.000684909634461204, + "loss": 3.455, + "step": 37100 + }, + { + "epoch": 2.52106264438103, + "grad_norm": 0.9182109832763672, + "learning_rate": 0.0006848671694523712, + "loss": 3.3313, + "step": 37105 + }, + { + "epoch": 2.521402364451692, + "grad_norm": 1.2223981618881226, + "learning_rate": 0.0006848247044435386, + "loss": 3.4623, + "step": 37110 + }, + { + "epoch": 2.5217420845223537, + "grad_norm": 1.0115257501602173, + "learning_rate": 0.0006847822394347059, + "loss": 3.5466, + "step": 37115 + }, + { + "epoch": 2.5220818045930153, + "grad_norm": 0.860014021396637, + "learning_rate": 0.0006847397744258731, + "loss": 3.6982, + "step": 37120 + }, + { + "epoch": 2.522421524663677, + "grad_norm": 0.8742210865020752, + "learning_rate": 0.0006846973094170403, + "loss": 3.626, + "step": 37125 + }, + { + "epoch": 2.522761244734339, + "grad_norm": 0.8782237768173218, + "learning_rate": 0.0006846548444082077, + "loss": 3.5397, + "step": 37130 + }, + { + "epoch": 2.5231009648050007, + "grad_norm": 0.996138870716095, + "learning_rate": 0.0006846123793993749, + "loss": 3.7882, + "step": 37135 + }, + { + "epoch": 2.5234406848756623, + "grad_norm": 0.598161518573761, + "learning_rate": 0.0006845699143905421, + "loss": 3.6471, + "step": 37140 + }, + { + "epoch": 2.5237804049463244, + "grad_norm": 0.8116862773895264, + "learning_rate": 0.0006845274493817096, + "loss": 3.4462, + "step": 37145 + }, + { + "epoch": 2.524120125016986, + "grad_norm": 0.9528845548629761, + "learning_rate": 0.0006844849843728768, + "loss": 3.746, + "step": 37150 + }, + { + "epoch": 2.5244598450876476, + "grad_norm": 0.9268794059753418, + "learning_rate": 0.000684442519364044, + "loss": 3.5414, + "step": 37155 + }, + { + "epoch": 2.5247995651583093, + "grad_norm": 0.8700067400932312, + "learning_rate": 0.0006844000543552114, + "loss": 3.4506, + "step": 37160 + }, + { + "epoch": 2.5251392852289714, + "grad_norm": 0.9306234121322632, + "learning_rate": 0.0006843575893463786, + "loss": 3.7126, + "step": 37165 + }, + { + "epoch": 2.525479005299633, + "grad_norm": 0.9183580875396729, + "learning_rate": 0.0006843151243375458, + "loss": 3.4843, + "step": 37170 + }, + { + "epoch": 2.5258187253702946, + "grad_norm": 0.9226963520050049, + "learning_rate": 0.0006842726593287131, + "loss": 3.3709, + "step": 37175 + }, + { + "epoch": 2.5261584454409567, + "grad_norm": 1.014102578163147, + "learning_rate": 0.0006842301943198805, + "loss": 3.5456, + "step": 37180 + }, + { + "epoch": 2.5264981655116183, + "grad_norm": 0.6991927623748779, + "learning_rate": 0.0006841877293110477, + "loss": 3.4699, + "step": 37185 + }, + { + "epoch": 2.52683788558228, + "grad_norm": 0.8245767951011658, + "learning_rate": 0.000684145264302215, + "loss": 3.4473, + "step": 37190 + }, + { + "epoch": 2.527177605652942, + "grad_norm": 0.9326229095458984, + "learning_rate": 0.0006841027992933823, + "loss": 3.7456, + "step": 37195 + }, + { + "epoch": 2.5275173257236037, + "grad_norm": 0.8568160533905029, + "learning_rate": 0.0006840603342845495, + "loss": 3.4264, + "step": 37200 + }, + { + "epoch": 2.5278570457942653, + "grad_norm": 0.7080419659614563, + "learning_rate": 0.0006840178692757168, + "loss": 3.6784, + "step": 37205 + }, + { + "epoch": 2.5281967658649274, + "grad_norm": 0.8823555111885071, + "learning_rate": 0.000683975404266884, + "loss": 3.5315, + "step": 37210 + }, + { + "epoch": 2.528536485935589, + "grad_norm": 0.9758599400520325, + "learning_rate": 0.0006839329392580514, + "loss": 3.4048, + "step": 37215 + }, + { + "epoch": 2.5288762060062506, + "grad_norm": 0.6911731958389282, + "learning_rate": 0.0006838904742492187, + "loss": 3.665, + "step": 37220 + }, + { + "epoch": 2.5292159260769127, + "grad_norm": 0.8030714988708496, + "learning_rate": 0.0006838480092403859, + "loss": 3.4843, + "step": 37225 + }, + { + "epoch": 2.5295556461475743, + "grad_norm": 0.730897843837738, + "learning_rate": 0.0006838055442315532, + "loss": 3.4001, + "step": 37230 + }, + { + "epoch": 2.529895366218236, + "grad_norm": 1.1770083904266357, + "learning_rate": 0.0006837630792227205, + "loss": 3.4259, + "step": 37235 + }, + { + "epoch": 2.530235086288898, + "grad_norm": 1.0021699666976929, + "learning_rate": 0.0006837206142138877, + "loss": 3.575, + "step": 37240 + }, + { + "epoch": 2.5305748063595597, + "grad_norm": 1.0874969959259033, + "learning_rate": 0.000683678149205055, + "loss": 3.3868, + "step": 37245 + }, + { + "epoch": 2.5309145264302213, + "grad_norm": 0.8256418704986572, + "learning_rate": 0.0006836356841962224, + "loss": 3.5362, + "step": 37250 + }, + { + "epoch": 2.5312542465008834, + "grad_norm": 0.8466256260871887, + "learning_rate": 0.0006835932191873896, + "loss": 3.6982, + "step": 37255 + }, + { + "epoch": 2.531593966571545, + "grad_norm": 0.8407155275344849, + "learning_rate": 0.0006835507541785568, + "loss": 3.6027, + "step": 37260 + }, + { + "epoch": 2.5319336866422066, + "grad_norm": 0.9318026900291443, + "learning_rate": 0.0006835082891697242, + "loss": 3.4181, + "step": 37265 + }, + { + "epoch": 2.5322734067128687, + "grad_norm": 0.8569952249526978, + "learning_rate": 0.0006834658241608914, + "loss": 3.4468, + "step": 37270 + }, + { + "epoch": 2.5326131267835303, + "grad_norm": 0.7275272011756897, + "learning_rate": 0.0006834233591520586, + "loss": 3.55, + "step": 37275 + }, + { + "epoch": 2.532952846854192, + "grad_norm": 1.004207968711853, + "learning_rate": 0.000683380894143226, + "loss": 3.5624, + "step": 37280 + }, + { + "epoch": 2.533292566924854, + "grad_norm": 2.2090513706207275, + "learning_rate": 0.0006833384291343933, + "loss": 3.4734, + "step": 37285 + }, + { + "epoch": 2.5336322869955157, + "grad_norm": 0.7665367126464844, + "learning_rate": 0.0006832959641255605, + "loss": 3.5937, + "step": 37290 + }, + { + "epoch": 2.5339720070661773, + "grad_norm": 0.857237696647644, + "learning_rate": 0.0006832534991167279, + "loss": 3.4068, + "step": 37295 + }, + { + "epoch": 2.5343117271368394, + "grad_norm": 0.9589405655860901, + "learning_rate": 0.0006832110341078951, + "loss": 3.4182, + "step": 37300 + }, + { + "epoch": 2.534651447207501, + "grad_norm": 0.8746241927146912, + "learning_rate": 0.0006831685690990623, + "loss": 3.6515, + "step": 37305 + }, + { + "epoch": 2.5349911672781626, + "grad_norm": 0.7131111025810242, + "learning_rate": 0.0006831261040902296, + "loss": 3.398, + "step": 37310 + }, + { + "epoch": 2.5353308873488247, + "grad_norm": 1.3584858179092407, + "learning_rate": 0.0006830836390813969, + "loss": 3.2563, + "step": 37315 + }, + { + "epoch": 2.5356706074194864, + "grad_norm": 0.7180082201957703, + "learning_rate": 0.0006830411740725643, + "loss": 3.5323, + "step": 37320 + }, + { + "epoch": 2.536010327490148, + "grad_norm": 0.8817541599273682, + "learning_rate": 0.0006829987090637315, + "loss": 3.4727, + "step": 37325 + }, + { + "epoch": 2.53635004756081, + "grad_norm": 0.848643958568573, + "learning_rate": 0.0006829562440548988, + "loss": 3.6531, + "step": 37330 + }, + { + "epoch": 2.5366897676314717, + "grad_norm": 0.7705969214439392, + "learning_rate": 0.0006829137790460661, + "loss": 3.5679, + "step": 37335 + }, + { + "epoch": 2.5370294877021333, + "grad_norm": 0.6676970720291138, + "learning_rate": 0.0006828713140372333, + "loss": 3.6485, + "step": 37340 + }, + { + "epoch": 2.5373692077727954, + "grad_norm": 0.967232346534729, + "learning_rate": 0.0006828288490284006, + "loss": 3.8227, + "step": 37345 + }, + { + "epoch": 2.537708927843457, + "grad_norm": 0.8471551537513733, + "learning_rate": 0.000682786384019568, + "loss": 3.3796, + "step": 37350 + }, + { + "epoch": 2.5380486479141187, + "grad_norm": 0.6741301417350769, + "learning_rate": 0.0006827439190107352, + "loss": 3.7094, + "step": 37355 + }, + { + "epoch": 2.5383883679847807, + "grad_norm": 0.8954216241836548, + "learning_rate": 0.0006827014540019024, + "loss": 3.628, + "step": 37360 + }, + { + "epoch": 2.5387280880554424, + "grad_norm": 0.7996103167533875, + "learning_rate": 0.0006826589889930698, + "loss": 3.3768, + "step": 37365 + }, + { + "epoch": 2.539067808126104, + "grad_norm": 0.9981301426887512, + "learning_rate": 0.000682616523984237, + "loss": 3.5798, + "step": 37370 + }, + { + "epoch": 2.539407528196766, + "grad_norm": 0.7760396003723145, + "learning_rate": 0.0006825740589754042, + "loss": 3.6395, + "step": 37375 + }, + { + "epoch": 2.5397472482674277, + "grad_norm": 0.7435452938079834, + "learning_rate": 0.0006825315939665716, + "loss": 3.5326, + "step": 37380 + }, + { + "epoch": 2.5400869683380893, + "grad_norm": 0.763060450553894, + "learning_rate": 0.0006824891289577389, + "loss": 3.6156, + "step": 37385 + }, + { + "epoch": 2.5404266884087514, + "grad_norm": 0.9341066479682922, + "learning_rate": 0.0006824466639489061, + "loss": 3.6775, + "step": 37390 + }, + { + "epoch": 2.540766408479413, + "grad_norm": 0.704109251499176, + "learning_rate": 0.0006824041989400735, + "loss": 3.6327, + "step": 37395 + }, + { + "epoch": 2.5411061285500747, + "grad_norm": 0.7881059050559998, + "learning_rate": 0.0006823617339312407, + "loss": 3.8377, + "step": 37400 + }, + { + "epoch": 2.5414458486207367, + "grad_norm": 0.7162206172943115, + "learning_rate": 0.0006823192689224079, + "loss": 3.2856, + "step": 37405 + }, + { + "epoch": 2.5417855686913984, + "grad_norm": 0.8092921376228333, + "learning_rate": 0.0006822768039135752, + "loss": 3.5832, + "step": 37410 + }, + { + "epoch": 2.54212528876206, + "grad_norm": 0.8755738139152527, + "learning_rate": 0.0006822343389047425, + "loss": 3.8799, + "step": 37415 + }, + { + "epoch": 2.542465008832722, + "grad_norm": 0.8400694727897644, + "learning_rate": 0.0006821918738959098, + "loss": 3.8144, + "step": 37420 + }, + { + "epoch": 2.5428047289033837, + "grad_norm": 0.8859412670135498, + "learning_rate": 0.0006821494088870771, + "loss": 3.6268, + "step": 37425 + }, + { + "epoch": 2.5431444489740453, + "grad_norm": 0.7265583872795105, + "learning_rate": 0.0006821069438782444, + "loss": 3.6034, + "step": 37430 + }, + { + "epoch": 2.5434841690447074, + "grad_norm": 1.0732241868972778, + "learning_rate": 0.0006820644788694116, + "loss": 3.5784, + "step": 37435 + }, + { + "epoch": 2.543823889115369, + "grad_norm": 0.9312456846237183, + "learning_rate": 0.0006820220138605789, + "loss": 3.4733, + "step": 37440 + }, + { + "epoch": 2.5441636091860307, + "grad_norm": 0.8461014628410339, + "learning_rate": 0.0006819795488517462, + "loss": 3.3997, + "step": 37445 + }, + { + "epoch": 2.5445033292566928, + "grad_norm": 0.8358914256095886, + "learning_rate": 0.0006819370838429134, + "loss": 3.5168, + "step": 37450 + }, + { + "epoch": 2.5448430493273544, + "grad_norm": 1.2743711471557617, + "learning_rate": 0.0006818946188340808, + "loss": 3.4885, + "step": 37455 + }, + { + "epoch": 2.545182769398016, + "grad_norm": 0.8078498840332031, + "learning_rate": 0.000681852153825248, + "loss": 3.619, + "step": 37460 + }, + { + "epoch": 2.5455224894686777, + "grad_norm": 1.0167410373687744, + "learning_rate": 0.0006818096888164153, + "loss": 3.5456, + "step": 37465 + }, + { + "epoch": 2.5458622095393397, + "grad_norm": 0.8658894300460815, + "learning_rate": 0.0006817672238075826, + "loss": 3.6247, + "step": 37470 + }, + { + "epoch": 2.5462019296100014, + "grad_norm": 0.8564377427101135, + "learning_rate": 0.0006817247587987498, + "loss": 3.4208, + "step": 37475 + }, + { + "epoch": 2.546541649680663, + "grad_norm": 0.9480738043785095, + "learning_rate": 0.0006816822937899171, + "loss": 3.4077, + "step": 37480 + }, + { + "epoch": 2.546881369751325, + "grad_norm": 0.7453556656837463, + "learning_rate": 0.0006816398287810844, + "loss": 3.776, + "step": 37485 + }, + { + "epoch": 2.5472210898219867, + "grad_norm": 0.8725706934928894, + "learning_rate": 0.0006815973637722517, + "loss": 3.817, + "step": 37490 + }, + { + "epoch": 2.5475608098926483, + "grad_norm": 0.7744187712669373, + "learning_rate": 0.000681554898763419, + "loss": 3.3928, + "step": 37495 + }, + { + "epoch": 2.54790052996331, + "grad_norm": 0.9024271965026855, + "learning_rate": 0.0006815124337545863, + "loss": 3.5952, + "step": 37500 + }, + { + "epoch": 2.548240250033972, + "grad_norm": 0.7473392486572266, + "learning_rate": 0.0006814699687457535, + "loss": 3.7007, + "step": 37505 + }, + { + "epoch": 2.5485799701046337, + "grad_norm": 1.1972899436950684, + "learning_rate": 0.0006814275037369207, + "loss": 3.6493, + "step": 37510 + }, + { + "epoch": 2.5489196901752953, + "grad_norm": 0.7993201017379761, + "learning_rate": 0.0006813850387280881, + "loss": 3.5256, + "step": 37515 + }, + { + "epoch": 2.5492594102459574, + "grad_norm": 0.6989744901657104, + "learning_rate": 0.0006813425737192553, + "loss": 3.4708, + "step": 37520 + }, + { + "epoch": 2.549599130316619, + "grad_norm": 1.0735396146774292, + "learning_rate": 0.0006813001087104226, + "loss": 3.7403, + "step": 37525 + }, + { + "epoch": 2.5499388503872806, + "grad_norm": 0.8045423030853271, + "learning_rate": 0.00068125764370159, + "loss": 3.5402, + "step": 37530 + }, + { + "epoch": 2.5502785704579427, + "grad_norm": 0.7121123671531677, + "learning_rate": 0.0006812151786927572, + "loss": 3.7096, + "step": 37535 + }, + { + "epoch": 2.5506182905286043, + "grad_norm": 0.8172414898872375, + "learning_rate": 0.0006811727136839244, + "loss": 3.6282, + "step": 37540 + }, + { + "epoch": 2.550958010599266, + "grad_norm": 0.8707689642906189, + "learning_rate": 0.0006811302486750918, + "loss": 3.5806, + "step": 37545 + }, + { + "epoch": 2.551297730669928, + "grad_norm": 0.8858196139335632, + "learning_rate": 0.000681087783666259, + "loss": 3.3122, + "step": 37550 + }, + { + "epoch": 2.5516374507405897, + "grad_norm": 0.8651095628738403, + "learning_rate": 0.0006810453186574262, + "loss": 3.269, + "step": 37555 + }, + { + "epoch": 2.5519771708112513, + "grad_norm": 0.7171977162361145, + "learning_rate": 0.0006810028536485936, + "loss": 3.2971, + "step": 37560 + }, + { + "epoch": 2.5523168908819134, + "grad_norm": 0.736710250377655, + "learning_rate": 0.0006809603886397609, + "loss": 3.4751, + "step": 37565 + }, + { + "epoch": 2.552656610952575, + "grad_norm": 1.1818926334381104, + "learning_rate": 0.0006809179236309281, + "loss": 3.4499, + "step": 37570 + }, + { + "epoch": 2.5529963310232366, + "grad_norm": 0.8949518203735352, + "learning_rate": 0.0006808754586220954, + "loss": 3.4745, + "step": 37575 + }, + { + "epoch": 2.5533360510938987, + "grad_norm": 0.760863184928894, + "learning_rate": 0.0006808329936132627, + "loss": 3.828, + "step": 37580 + }, + { + "epoch": 2.5536757711645603, + "grad_norm": 0.6896313428878784, + "learning_rate": 0.0006807905286044299, + "loss": 3.1893, + "step": 37585 + }, + { + "epoch": 2.554015491235222, + "grad_norm": 0.8468573689460754, + "learning_rate": 0.0006807480635955972, + "loss": 3.5887, + "step": 37590 + }, + { + "epoch": 2.554355211305884, + "grad_norm": 0.9476075768470764, + "learning_rate": 0.0006807055985867646, + "loss": 3.6029, + "step": 37595 + }, + { + "epoch": 2.5546949313765457, + "grad_norm": 0.8353786468505859, + "learning_rate": 0.0006806631335779318, + "loss": 3.2934, + "step": 37600 + }, + { + "epoch": 2.5550346514472073, + "grad_norm": 0.705713152885437, + "learning_rate": 0.0006806206685690991, + "loss": 3.6285, + "step": 37605 + }, + { + "epoch": 2.5553743715178694, + "grad_norm": 0.9897499680519104, + "learning_rate": 0.0006805782035602663, + "loss": 3.657, + "step": 37610 + }, + { + "epoch": 2.555714091588531, + "grad_norm": 1.1788630485534668, + "learning_rate": 0.0006805357385514336, + "loss": 3.3612, + "step": 37615 + }, + { + "epoch": 2.5560538116591927, + "grad_norm": 0.8140487670898438, + "learning_rate": 0.0006804932735426009, + "loss": 3.6974, + "step": 37620 + }, + { + "epoch": 2.5563935317298547, + "grad_norm": 1.4328302145004272, + "learning_rate": 0.0006804508085337681, + "loss": 3.7755, + "step": 37625 + }, + { + "epoch": 2.5567332518005164, + "grad_norm": 0.9705634713172913, + "learning_rate": 0.0006804083435249355, + "loss": 3.4662, + "step": 37630 + }, + { + "epoch": 2.557072971871178, + "grad_norm": 0.860119640827179, + "learning_rate": 0.0006803658785161028, + "loss": 3.717, + "step": 37635 + }, + { + "epoch": 2.55741269194184, + "grad_norm": 0.9648481607437134, + "learning_rate": 0.00068032341350727, + "loss": 3.3909, + "step": 37640 + }, + { + "epoch": 2.5577524120125017, + "grad_norm": 0.8376898765563965, + "learning_rate": 0.0006802809484984372, + "loss": 3.5654, + "step": 37645 + }, + { + "epoch": 2.5580921320831633, + "grad_norm": 0.7412816286087036, + "learning_rate": 0.0006802384834896046, + "loss": 3.4574, + "step": 37650 + }, + { + "epoch": 2.5584318521538254, + "grad_norm": 0.7815935611724854, + "learning_rate": 0.0006801960184807718, + "loss": 3.6062, + "step": 37655 + }, + { + "epoch": 2.558771572224487, + "grad_norm": 0.9171143174171448, + "learning_rate": 0.0006801535534719391, + "loss": 3.5079, + "step": 37660 + }, + { + "epoch": 2.5591112922951487, + "grad_norm": 0.7003941535949707, + "learning_rate": 0.0006801110884631065, + "loss": 3.2882, + "step": 37665 + }, + { + "epoch": 2.5594510123658107, + "grad_norm": 0.7786941528320312, + "learning_rate": 0.0006800686234542737, + "loss": 3.5496, + "step": 37670 + }, + { + "epoch": 2.5597907324364724, + "grad_norm": 0.7433000206947327, + "learning_rate": 0.000680026158445441, + "loss": 3.6646, + "step": 37675 + }, + { + "epoch": 2.560130452507134, + "grad_norm": 0.9838359951972961, + "learning_rate": 0.0006799836934366083, + "loss": 3.6243, + "step": 37680 + }, + { + "epoch": 2.560470172577796, + "grad_norm": 0.8476574420928955, + "learning_rate": 0.0006799412284277755, + "loss": 3.6368, + "step": 37685 + }, + { + "epoch": 2.5608098926484577, + "grad_norm": 1.1009407043457031, + "learning_rate": 0.0006798987634189428, + "loss": 3.6687, + "step": 37690 + }, + { + "epoch": 2.5611496127191193, + "grad_norm": 0.7650176882743835, + "learning_rate": 0.00067985629841011, + "loss": 3.5522, + "step": 37695 + }, + { + "epoch": 2.5614893327897814, + "grad_norm": 0.9319393038749695, + "learning_rate": 0.0006798138334012774, + "loss": 3.5043, + "step": 37700 + }, + { + "epoch": 2.561829052860443, + "grad_norm": 0.9441753029823303, + "learning_rate": 0.0006797713683924447, + "loss": 3.5682, + "step": 37705 + }, + { + "epoch": 2.5621687729311047, + "grad_norm": 0.8322954773902893, + "learning_rate": 0.0006797289033836119, + "loss": 3.4793, + "step": 37710 + }, + { + "epoch": 2.5625084930017668, + "grad_norm": 0.7383533716201782, + "learning_rate": 0.0006796864383747792, + "loss": 3.6736, + "step": 37715 + }, + { + "epoch": 2.5628482130724284, + "grad_norm": 0.9590554237365723, + "learning_rate": 0.0006796439733659465, + "loss": 3.5587, + "step": 37720 + }, + { + "epoch": 2.56318793314309, + "grad_norm": 1.002205491065979, + "learning_rate": 0.0006796015083571137, + "loss": 3.3145, + "step": 37725 + }, + { + "epoch": 2.563527653213752, + "grad_norm": 0.7824024558067322, + "learning_rate": 0.000679559043348281, + "loss": 3.7085, + "step": 37730 + }, + { + "epoch": 2.5638673732844137, + "grad_norm": 0.8643611669540405, + "learning_rate": 0.0006795165783394484, + "loss": 3.6592, + "step": 37735 + }, + { + "epoch": 2.5642070933550754, + "grad_norm": 0.7834507822990417, + "learning_rate": 0.0006794741133306156, + "loss": 3.6691, + "step": 37740 + }, + { + "epoch": 2.5645468134257374, + "grad_norm": 0.6759098768234253, + "learning_rate": 0.0006794316483217828, + "loss": 3.6876, + "step": 37745 + }, + { + "epoch": 2.564886533496399, + "grad_norm": 0.8073316812515259, + "learning_rate": 0.0006793891833129502, + "loss": 3.3911, + "step": 37750 + }, + { + "epoch": 2.5652262535670607, + "grad_norm": 0.796397864818573, + "learning_rate": 0.0006793467183041174, + "loss": 3.5533, + "step": 37755 + }, + { + "epoch": 2.5655659736377228, + "grad_norm": 0.7748994827270508, + "learning_rate": 0.0006793042532952846, + "loss": 3.5981, + "step": 37760 + }, + { + "epoch": 2.5659056937083844, + "grad_norm": 0.837766170501709, + "learning_rate": 0.000679261788286452, + "loss": 3.6042, + "step": 37765 + }, + { + "epoch": 2.566245413779046, + "grad_norm": 0.9411345720291138, + "learning_rate": 0.0006792193232776193, + "loss": 3.4886, + "step": 37770 + }, + { + "epoch": 2.566585133849708, + "grad_norm": 0.9730944037437439, + "learning_rate": 0.0006791768582687865, + "loss": 3.7639, + "step": 37775 + }, + { + "epoch": 2.5669248539203697, + "grad_norm": 0.7867385149002075, + "learning_rate": 0.0006791343932599539, + "loss": 3.7003, + "step": 37780 + }, + { + "epoch": 2.5672645739910314, + "grad_norm": 0.7268301248550415, + "learning_rate": 0.0006790919282511211, + "loss": 3.4639, + "step": 37785 + }, + { + "epoch": 2.5676042940616934, + "grad_norm": 0.754845917224884, + "learning_rate": 0.0006790494632422883, + "loss": 3.8566, + "step": 37790 + }, + { + "epoch": 2.567944014132355, + "grad_norm": 0.8006362915039062, + "learning_rate": 0.0006790069982334557, + "loss": 3.5566, + "step": 37795 + }, + { + "epoch": 2.5682837342030167, + "grad_norm": 1.0977997779846191, + "learning_rate": 0.0006789645332246229, + "loss": 3.7898, + "step": 37800 + }, + { + "epoch": 2.5686234542736783, + "grad_norm": 0.8390315771102905, + "learning_rate": 0.0006789220682157902, + "loss": 3.5081, + "step": 37805 + }, + { + "epoch": 2.5689631743443404, + "grad_norm": 0.7826142311096191, + "learning_rate": 0.0006788796032069575, + "loss": 3.5884, + "step": 37810 + }, + { + "epoch": 2.569302894415002, + "grad_norm": 0.77899169921875, + "learning_rate": 0.0006788371381981248, + "loss": 3.6451, + "step": 37815 + }, + { + "epoch": 2.5696426144856637, + "grad_norm": 0.9358908534049988, + "learning_rate": 0.000678794673189292, + "loss": 3.6303, + "step": 37820 + }, + { + "epoch": 2.5699823345563257, + "grad_norm": 0.8480795621871948, + "learning_rate": 0.0006787522081804593, + "loss": 3.5702, + "step": 37825 + }, + { + "epoch": 2.5703220546269874, + "grad_norm": 0.8820386528968811, + "learning_rate": 0.0006787097431716266, + "loss": 3.2465, + "step": 37830 + }, + { + "epoch": 2.570661774697649, + "grad_norm": 0.846269428730011, + "learning_rate": 0.0006786672781627938, + "loss": 3.406, + "step": 37835 + }, + { + "epoch": 2.5710014947683106, + "grad_norm": 0.8806630373001099, + "learning_rate": 0.0006786248131539612, + "loss": 3.482, + "step": 37840 + }, + { + "epoch": 2.5713412148389727, + "grad_norm": 0.9678062200546265, + "learning_rate": 0.0006785823481451285, + "loss": 3.4212, + "step": 37845 + }, + { + "epoch": 2.5716809349096343, + "grad_norm": 0.9641690850257874, + "learning_rate": 0.0006785398831362957, + "loss": 3.7284, + "step": 37850 + }, + { + "epoch": 2.572020654980296, + "grad_norm": 0.86585533618927, + "learning_rate": 0.000678497418127463, + "loss": 3.7102, + "step": 37855 + }, + { + "epoch": 2.572360375050958, + "grad_norm": 0.6904707551002502, + "learning_rate": 0.0006784549531186302, + "loss": 3.4816, + "step": 37860 + }, + { + "epoch": 2.5727000951216197, + "grad_norm": 0.9663662910461426, + "learning_rate": 0.0006784124881097975, + "loss": 3.7681, + "step": 37865 + }, + { + "epoch": 2.5730398151922813, + "grad_norm": 0.8621360063552856, + "learning_rate": 0.0006783700231009648, + "loss": 3.5581, + "step": 37870 + }, + { + "epoch": 2.5733795352629434, + "grad_norm": 0.8445187211036682, + "learning_rate": 0.0006783275580921321, + "loss": 3.6319, + "step": 37875 + }, + { + "epoch": 2.573719255333605, + "grad_norm": 1.0300123691558838, + "learning_rate": 0.0006782850930832994, + "loss": 3.4343, + "step": 37880 + }, + { + "epoch": 2.5740589754042666, + "grad_norm": 0.8465536236763, + "learning_rate": 0.0006782426280744667, + "loss": 3.5243, + "step": 37885 + }, + { + "epoch": 2.5743986954749287, + "grad_norm": 0.7552660703659058, + "learning_rate": 0.0006782001630656339, + "loss": 3.6014, + "step": 37890 + }, + { + "epoch": 2.5747384155455904, + "grad_norm": 1.1084339618682861, + "learning_rate": 0.0006781576980568011, + "loss": 3.634, + "step": 37895 + }, + { + "epoch": 2.575078135616252, + "grad_norm": 0.8992234468460083, + "learning_rate": 0.0006781152330479685, + "loss": 3.8, + "step": 37900 + }, + { + "epoch": 2.575417855686914, + "grad_norm": 0.7705115675926208, + "learning_rate": 0.0006780727680391357, + "loss": 3.7682, + "step": 37905 + }, + { + "epoch": 2.5757575757575757, + "grad_norm": 1.2505043745040894, + "learning_rate": 0.000678030303030303, + "loss": 3.6418, + "step": 37910 + }, + { + "epoch": 2.5760972958282373, + "grad_norm": 0.9458591938018799, + "learning_rate": 0.0006779878380214704, + "loss": 3.608, + "step": 37915 + }, + { + "epoch": 2.5764370158988994, + "grad_norm": 0.7450234293937683, + "learning_rate": 0.0006779453730126376, + "loss": 3.4988, + "step": 37920 + }, + { + "epoch": 2.576776735969561, + "grad_norm": 0.9315587282180786, + "learning_rate": 0.0006779029080038048, + "loss": 3.6405, + "step": 37925 + }, + { + "epoch": 2.5771164560402227, + "grad_norm": 0.8893061876296997, + "learning_rate": 0.0006778604429949722, + "loss": 3.3913, + "step": 37930 + }, + { + "epoch": 2.5774561761108847, + "grad_norm": 0.9576724767684937, + "learning_rate": 0.0006778179779861394, + "loss": 3.3531, + "step": 37935 + }, + { + "epoch": 2.5777958961815464, + "grad_norm": 0.8206709027290344, + "learning_rate": 0.0006777755129773066, + "loss": 3.508, + "step": 37940 + }, + { + "epoch": 2.578135616252208, + "grad_norm": 0.8342892527580261, + "learning_rate": 0.000677733047968474, + "loss": 3.4397, + "step": 37945 + }, + { + "epoch": 2.57847533632287, + "grad_norm": 1.346496343612671, + "learning_rate": 0.0006776905829596413, + "loss": 3.4765, + "step": 37950 + }, + { + "epoch": 2.5788150563935317, + "grad_norm": 0.7249206304550171, + "learning_rate": 0.0006776481179508085, + "loss": 3.45, + "step": 37955 + }, + { + "epoch": 2.5791547764641933, + "grad_norm": 0.8977625370025635, + "learning_rate": 0.0006776056529419758, + "loss": 3.2067, + "step": 37960 + }, + { + "epoch": 2.5794944965348554, + "grad_norm": 0.8230484127998352, + "learning_rate": 0.0006775631879331431, + "loss": 3.7792, + "step": 37965 + }, + { + "epoch": 2.579834216605517, + "grad_norm": 0.8896654844284058, + "learning_rate": 0.0006775207229243103, + "loss": 3.3746, + "step": 37970 + }, + { + "epoch": 2.5801739366761787, + "grad_norm": 0.8910081386566162, + "learning_rate": 0.0006774782579154777, + "loss": 3.4598, + "step": 37975 + }, + { + "epoch": 2.5805136567468407, + "grad_norm": 0.8321078419685364, + "learning_rate": 0.000677435792906645, + "loss": 3.6167, + "step": 37980 + }, + { + "epoch": 2.5808533768175024, + "grad_norm": 0.9067173004150391, + "learning_rate": 0.0006773933278978122, + "loss": 3.374, + "step": 37985 + }, + { + "epoch": 2.581193096888164, + "grad_norm": 0.9578613638877869, + "learning_rate": 0.0006773508628889795, + "loss": 3.651, + "step": 37990 + }, + { + "epoch": 2.581532816958826, + "grad_norm": 0.7611261010169983, + "learning_rate": 0.0006773083978801467, + "loss": 3.7245, + "step": 37995 + }, + { + "epoch": 2.5818725370294877, + "grad_norm": 1.0454410314559937, + "learning_rate": 0.0006772659328713141, + "loss": 3.4177, + "step": 38000 + }, + { + "epoch": 2.5822122571001493, + "grad_norm": 1.0576915740966797, + "learning_rate": 0.0006772234678624813, + "loss": 3.4216, + "step": 38005 + }, + { + "epoch": 2.5825519771708114, + "grad_norm": 0.6573306322097778, + "learning_rate": 0.0006771810028536486, + "loss": 3.8496, + "step": 38010 + }, + { + "epoch": 2.582891697241473, + "grad_norm": 0.9440877437591553, + "learning_rate": 0.000677138537844816, + "loss": 3.6634, + "step": 38015 + }, + { + "epoch": 2.5832314173121347, + "grad_norm": 0.8111201524734497, + "learning_rate": 0.0006770960728359832, + "loss": 3.7969, + "step": 38020 + }, + { + "epoch": 2.5835711373827968, + "grad_norm": 0.6250476241111755, + "learning_rate": 0.0006770536078271504, + "loss": 3.547, + "step": 38025 + }, + { + "epoch": 2.5839108574534584, + "grad_norm": 0.8649698495864868, + "learning_rate": 0.0006770111428183178, + "loss": 3.4831, + "step": 38030 + }, + { + "epoch": 2.58425057752412, + "grad_norm": 1.282321810722351, + "learning_rate": 0.000676968677809485, + "loss": 3.592, + "step": 38035 + }, + { + "epoch": 2.584590297594782, + "grad_norm": 0.939784824848175, + "learning_rate": 0.0006769262128006522, + "loss": 3.7098, + "step": 38040 + }, + { + "epoch": 2.5849300176654437, + "grad_norm": 0.8066626191139221, + "learning_rate": 0.0006768837477918197, + "loss": 3.5911, + "step": 38045 + }, + { + "epoch": 2.5852697377361054, + "grad_norm": 0.901806652545929, + "learning_rate": 0.0006768412827829869, + "loss": 3.5094, + "step": 38050 + }, + { + "epoch": 2.5856094578067674, + "grad_norm": 1.9045625925064087, + "learning_rate": 0.0006767988177741541, + "loss": 3.7184, + "step": 38055 + }, + { + "epoch": 2.585949177877429, + "grad_norm": 0.8009050488471985, + "learning_rate": 0.0006767563527653214, + "loss": 3.5618, + "step": 38060 + }, + { + "epoch": 2.5862888979480907, + "grad_norm": 1.0326038599014282, + "learning_rate": 0.0006767138877564887, + "loss": 3.7053, + "step": 38065 + }, + { + "epoch": 2.5866286180187528, + "grad_norm": 0.6755421161651611, + "learning_rate": 0.0006766714227476559, + "loss": 3.745, + "step": 38070 + }, + { + "epoch": 2.5869683380894144, + "grad_norm": 0.8439539074897766, + "learning_rate": 0.0006766289577388232, + "loss": 3.6241, + "step": 38075 + }, + { + "epoch": 2.587308058160076, + "grad_norm": 0.8789964318275452, + "learning_rate": 0.0006765864927299906, + "loss": 3.3594, + "step": 38080 + }, + { + "epoch": 2.587647778230738, + "grad_norm": 0.5861936211585999, + "learning_rate": 0.0006765440277211578, + "loss": 3.5959, + "step": 38085 + }, + { + "epoch": 2.5879874983013997, + "grad_norm": 0.8412250876426697, + "learning_rate": 0.0006765015627123251, + "loss": 3.5651, + "step": 38090 + }, + { + "epoch": 2.5883272183720614, + "grad_norm": 0.8585387468338013, + "learning_rate": 0.0006764590977034923, + "loss": 3.5619, + "step": 38095 + }, + { + "epoch": 2.5886669384427234, + "grad_norm": 0.7352014780044556, + "learning_rate": 0.0006764166326946596, + "loss": 3.5438, + "step": 38100 + }, + { + "epoch": 2.589006658513385, + "grad_norm": 0.884844958782196, + "learning_rate": 0.0006763741676858269, + "loss": 3.8177, + "step": 38105 + }, + { + "epoch": 2.5893463785840467, + "grad_norm": 0.9844726324081421, + "learning_rate": 0.0006763317026769941, + "loss": 3.771, + "step": 38110 + }, + { + "epoch": 2.589686098654709, + "grad_norm": 0.9863755106925964, + "learning_rate": 0.0006762892376681615, + "loss": 3.551, + "step": 38115 + }, + { + "epoch": 2.5900258187253704, + "grad_norm": 0.9825115203857422, + "learning_rate": 0.0006762467726593288, + "loss": 3.4705, + "step": 38120 + }, + { + "epoch": 2.590365538796032, + "grad_norm": 0.9233160018920898, + "learning_rate": 0.000676204307650496, + "loss": 3.7045, + "step": 38125 + }, + { + "epoch": 2.590705258866694, + "grad_norm": 0.9878427386283875, + "learning_rate": 0.0006761618426416633, + "loss": 3.6086, + "step": 38130 + }, + { + "epoch": 2.5910449789373557, + "grad_norm": 0.9156361222267151, + "learning_rate": 0.0006761193776328306, + "loss": 3.6003, + "step": 38135 + }, + { + "epoch": 2.5913846990080174, + "grad_norm": 0.9913665056228638, + "learning_rate": 0.0006760769126239978, + "loss": 3.4969, + "step": 38140 + }, + { + "epoch": 2.591724419078679, + "grad_norm": 0.8464643955230713, + "learning_rate": 0.000676034447615165, + "loss": 3.7677, + "step": 38145 + }, + { + "epoch": 2.592064139149341, + "grad_norm": 1.024053692817688, + "learning_rate": 0.0006759919826063325, + "loss": 3.507, + "step": 38150 + }, + { + "epoch": 2.5924038592200027, + "grad_norm": 0.7304444909095764, + "learning_rate": 0.0006759495175974997, + "loss": 3.556, + "step": 38155 + }, + { + "epoch": 2.5927435792906643, + "grad_norm": 0.8833540081977844, + "learning_rate": 0.0006759070525886669, + "loss": 3.6095, + "step": 38160 + }, + { + "epoch": 2.5930832993613264, + "grad_norm": 0.9241132140159607, + "learning_rate": 0.0006758645875798343, + "loss": 3.5069, + "step": 38165 + }, + { + "epoch": 2.593423019431988, + "grad_norm": 0.8799516558647156, + "learning_rate": 0.0006758221225710015, + "loss": 3.4759, + "step": 38170 + }, + { + "epoch": 2.5937627395026497, + "grad_norm": 1.264857292175293, + "learning_rate": 0.0006757796575621687, + "loss": 3.3816, + "step": 38175 + }, + { + "epoch": 2.5941024595733113, + "grad_norm": 0.7831274271011353, + "learning_rate": 0.000675737192553336, + "loss": 3.6048, + "step": 38180 + }, + { + "epoch": 2.5944421796439734, + "grad_norm": 0.8285607099533081, + "learning_rate": 0.0006756947275445034, + "loss": 3.3295, + "step": 38185 + }, + { + "epoch": 2.594781899714635, + "grad_norm": 0.8769218325614929, + "learning_rate": 0.0006756522625356706, + "loss": 3.5181, + "step": 38190 + }, + { + "epoch": 2.5951216197852967, + "grad_norm": 1.0275450944900513, + "learning_rate": 0.000675609797526838, + "loss": 3.7917, + "step": 38195 + }, + { + "epoch": 2.5954613398559587, + "grad_norm": 0.7965763211250305, + "learning_rate": 0.0006755673325180052, + "loss": 3.3955, + "step": 38200 + }, + { + "epoch": 2.5958010599266204, + "grad_norm": 0.7361631989479065, + "learning_rate": 0.0006755248675091724, + "loss": 3.4683, + "step": 38205 + }, + { + "epoch": 2.596140779997282, + "grad_norm": 0.6477368474006653, + "learning_rate": 0.0006754824025003397, + "loss": 3.6982, + "step": 38210 + }, + { + "epoch": 2.596480500067944, + "grad_norm": 0.8972387313842773, + "learning_rate": 0.000675439937491507, + "loss": 3.4413, + "step": 38215 + }, + { + "epoch": 2.5968202201386057, + "grad_norm": 0.7857843637466431, + "learning_rate": 0.0006753974724826743, + "loss": 3.5718, + "step": 38220 + }, + { + "epoch": 2.5971599402092673, + "grad_norm": 0.8097231984138489, + "learning_rate": 0.0006753550074738416, + "loss": 3.5958, + "step": 38225 + }, + { + "epoch": 2.5974996602799294, + "grad_norm": 0.7015044689178467, + "learning_rate": 0.0006753125424650089, + "loss": 3.6646, + "step": 38230 + }, + { + "epoch": 2.597839380350591, + "grad_norm": 0.927734911441803, + "learning_rate": 0.0006752700774561761, + "loss": 3.5838, + "step": 38235 + }, + { + "epoch": 2.5981791004212527, + "grad_norm": 0.831920325756073, + "learning_rate": 0.0006752276124473434, + "loss": 3.5729, + "step": 38240 + }, + { + "epoch": 2.5985188204919147, + "grad_norm": 0.7816717028617859, + "learning_rate": 0.0006751851474385106, + "loss": 3.6579, + "step": 38245 + }, + { + "epoch": 2.5988585405625764, + "grad_norm": 1.7639280557632446, + "learning_rate": 0.0006751426824296779, + "loss": 3.7849, + "step": 38250 + }, + { + "epoch": 2.599198260633238, + "grad_norm": 0.8847293257713318, + "learning_rate": 0.0006751002174208453, + "loss": 3.6717, + "step": 38255 + }, + { + "epoch": 2.5995379807039, + "grad_norm": 0.7849623560905457, + "learning_rate": 0.0006750577524120125, + "loss": 3.6197, + "step": 38260 + }, + { + "epoch": 2.5998777007745617, + "grad_norm": 0.8404530882835388, + "learning_rate": 0.0006750152874031798, + "loss": 3.4211, + "step": 38265 + }, + { + "epoch": 2.6002174208452233, + "grad_norm": 0.9761221408843994, + "learning_rate": 0.0006749728223943471, + "loss": 3.8696, + "step": 38270 + }, + { + "epoch": 2.6005571409158854, + "grad_norm": 0.7586433291435242, + "learning_rate": 0.0006749303573855143, + "loss": 3.3688, + "step": 38275 + }, + { + "epoch": 2.600896860986547, + "grad_norm": 0.8199018239974976, + "learning_rate": 0.0006748878923766815, + "loss": 3.5302, + "step": 38280 + }, + { + "epoch": 2.6012365810572087, + "grad_norm": 0.7461830377578735, + "learning_rate": 0.0006748454273678489, + "loss": 3.7262, + "step": 38285 + }, + { + "epoch": 2.6015763011278707, + "grad_norm": 1.0927621126174927, + "learning_rate": 0.0006748029623590162, + "loss": 3.3659, + "step": 38290 + }, + { + "epoch": 2.6019160211985324, + "grad_norm": 0.9734758138656616, + "learning_rate": 0.0006747604973501834, + "loss": 3.5388, + "step": 38295 + }, + { + "epoch": 2.602255741269194, + "grad_norm": 1.051708459854126, + "learning_rate": 0.0006747180323413508, + "loss": 3.7233, + "step": 38300 + }, + { + "epoch": 2.602595461339856, + "grad_norm": 1.4222198724746704, + "learning_rate": 0.000674675567332518, + "loss": 3.4036, + "step": 38305 + }, + { + "epoch": 2.6029351814105177, + "grad_norm": 0.7939115762710571, + "learning_rate": 0.0006746331023236852, + "loss": 3.4957, + "step": 38310 + }, + { + "epoch": 2.6032749014811793, + "grad_norm": 1.2306339740753174, + "learning_rate": 0.0006745906373148526, + "loss": 3.5228, + "step": 38315 + }, + { + "epoch": 2.6036146215518414, + "grad_norm": 0.8348496556282043, + "learning_rate": 0.0006745481723060198, + "loss": 3.6153, + "step": 38320 + }, + { + "epoch": 2.603954341622503, + "grad_norm": 1.0470563173294067, + "learning_rate": 0.0006745057072971871, + "loss": 3.7779, + "step": 38325 + }, + { + "epoch": 2.6042940616931647, + "grad_norm": 0.6838955283164978, + "learning_rate": 0.0006744632422883545, + "loss": 3.7124, + "step": 38330 + }, + { + "epoch": 2.6046337817638268, + "grad_norm": 0.8200861215591431, + "learning_rate": 0.0006744207772795217, + "loss": 3.6478, + "step": 38335 + }, + { + "epoch": 2.6049735018344884, + "grad_norm": 0.9785839915275574, + "learning_rate": 0.000674378312270689, + "loss": 3.6081, + "step": 38340 + }, + { + "epoch": 2.60531322190515, + "grad_norm": 0.7723677754402161, + "learning_rate": 0.0006743358472618562, + "loss": 3.5961, + "step": 38345 + }, + { + "epoch": 2.605652941975812, + "grad_norm": 1.0171928405761719, + "learning_rate": 0.0006742933822530235, + "loss": 3.4563, + "step": 38350 + }, + { + "epoch": 2.6059926620464737, + "grad_norm": 1.064400553703308, + "learning_rate": 0.0006742509172441908, + "loss": 3.4759, + "step": 38355 + }, + { + "epoch": 2.6063323821171354, + "grad_norm": 0.9959901571273804, + "learning_rate": 0.0006742084522353581, + "loss": 3.5084, + "step": 38360 + }, + { + "epoch": 2.6066721021877974, + "grad_norm": 0.9633559584617615, + "learning_rate": 0.0006741659872265254, + "loss": 3.3502, + "step": 38365 + }, + { + "epoch": 2.607011822258459, + "grad_norm": 0.6645146012306213, + "learning_rate": 0.0006741235222176927, + "loss": 3.7727, + "step": 38370 + }, + { + "epoch": 2.6073515423291207, + "grad_norm": 0.7200229167938232, + "learning_rate": 0.0006740810572088599, + "loss": 3.9959, + "step": 38375 + }, + { + "epoch": 2.6076912623997828, + "grad_norm": 0.7621169686317444, + "learning_rate": 0.0006740385922000271, + "loss": 3.6673, + "step": 38380 + }, + { + "epoch": 2.6080309824704444, + "grad_norm": 1.0611259937286377, + "learning_rate": 0.0006739961271911945, + "loss": 3.6608, + "step": 38385 + }, + { + "epoch": 2.608370702541106, + "grad_norm": 0.7128794193267822, + "learning_rate": 0.0006739536621823617, + "loss": 3.4661, + "step": 38390 + }, + { + "epoch": 2.608710422611768, + "grad_norm": 0.8369954824447632, + "learning_rate": 0.000673911197173529, + "loss": 3.8463, + "step": 38395 + }, + { + "epoch": 2.6090501426824297, + "grad_norm": 0.6605397462844849, + "learning_rate": 0.0006738687321646964, + "loss": 3.6906, + "step": 38400 + }, + { + "epoch": 2.6093898627530914, + "grad_norm": 0.7816190123558044, + "learning_rate": 0.0006738262671558636, + "loss": 3.5596, + "step": 38405 + }, + { + "epoch": 2.6097295828237534, + "grad_norm": 0.7149060368537903, + "learning_rate": 0.0006737838021470308, + "loss": 3.6477, + "step": 38410 + }, + { + "epoch": 2.610069302894415, + "grad_norm": 0.830822229385376, + "learning_rate": 0.0006737413371381982, + "loss": 3.6696, + "step": 38415 + }, + { + "epoch": 2.6104090229650767, + "grad_norm": 1.726894736289978, + "learning_rate": 0.0006736988721293654, + "loss": 3.4857, + "step": 38420 + }, + { + "epoch": 2.610748743035739, + "grad_norm": 0.841387927532196, + "learning_rate": 0.0006736564071205326, + "loss": 3.5815, + "step": 38425 + }, + { + "epoch": 2.6110884631064004, + "grad_norm": 0.7283778190612793, + "learning_rate": 0.0006736139421117001, + "loss": 3.269, + "step": 38430 + }, + { + "epoch": 2.611428183177062, + "grad_norm": 0.8097230792045593, + "learning_rate": 0.0006735714771028673, + "loss": 3.3093, + "step": 38435 + }, + { + "epoch": 2.611767903247724, + "grad_norm": 1.0492138862609863, + "learning_rate": 0.0006735290120940345, + "loss": 3.5646, + "step": 38440 + }, + { + "epoch": 2.6121076233183858, + "grad_norm": 1.0851428508758545, + "learning_rate": 0.0006734865470852018, + "loss": 3.3848, + "step": 38445 + }, + { + "epoch": 2.6124473433890474, + "grad_norm": 0.7617583274841309, + "learning_rate": 0.0006734440820763691, + "loss": 3.5219, + "step": 38450 + }, + { + "epoch": 2.6127870634597095, + "grad_norm": 0.7774679660797119, + "learning_rate": 0.0006734016170675363, + "loss": 3.5262, + "step": 38455 + }, + { + "epoch": 2.613126783530371, + "grad_norm": 0.8226400017738342, + "learning_rate": 0.0006733591520587036, + "loss": 3.6576, + "step": 38460 + }, + { + "epoch": 2.6134665036010327, + "grad_norm": 0.7864985466003418, + "learning_rate": 0.000673316687049871, + "loss": 3.6255, + "step": 38465 + }, + { + "epoch": 2.613806223671695, + "grad_norm": 1.197155237197876, + "learning_rate": 0.0006732742220410382, + "loss": 3.566, + "step": 38470 + }, + { + "epoch": 2.6141459437423564, + "grad_norm": 0.927047848701477, + "learning_rate": 0.0006732317570322055, + "loss": 3.4633, + "step": 38475 + }, + { + "epoch": 2.614485663813018, + "grad_norm": 1.0048460960388184, + "learning_rate": 0.0006731892920233727, + "loss": 3.7472, + "step": 38480 + }, + { + "epoch": 2.6148253838836797, + "grad_norm": 0.7877074480056763, + "learning_rate": 0.00067314682701454, + "loss": 3.4778, + "step": 38485 + }, + { + "epoch": 2.6151651039543418, + "grad_norm": 1.1517329216003418, + "learning_rate": 0.0006731043620057073, + "loss": 3.3867, + "step": 38490 + }, + { + "epoch": 2.6155048240250034, + "grad_norm": 0.884728193283081, + "learning_rate": 0.0006730618969968745, + "loss": 3.5884, + "step": 38495 + }, + { + "epoch": 2.615844544095665, + "grad_norm": 1.135661244392395, + "learning_rate": 0.0006730194319880419, + "loss": 3.4396, + "step": 38500 + }, + { + "epoch": 2.616184264166327, + "grad_norm": 0.9094513058662415, + "learning_rate": 0.0006729769669792092, + "loss": 3.1697, + "step": 38505 + }, + { + "epoch": 2.6165239842369887, + "grad_norm": 1.1300067901611328, + "learning_rate": 0.0006729345019703764, + "loss": 3.5583, + "step": 38510 + }, + { + "epoch": 2.6168637043076504, + "grad_norm": 4.949653625488281, + "learning_rate": 0.0006728920369615437, + "loss": 3.3656, + "step": 38515 + }, + { + "epoch": 2.617203424378312, + "grad_norm": 0.8903859257698059, + "learning_rate": 0.000672849571952711, + "loss": 3.5788, + "step": 38520 + }, + { + "epoch": 2.617543144448974, + "grad_norm": 0.8498395085334778, + "learning_rate": 0.0006728071069438782, + "loss": 3.4334, + "step": 38525 + }, + { + "epoch": 2.6178828645196357, + "grad_norm": 1.0390172004699707, + "learning_rate": 0.0006727646419350454, + "loss": 3.4975, + "step": 38530 + }, + { + "epoch": 2.6182225845902973, + "grad_norm": 0.6246967315673828, + "learning_rate": 0.0006727221769262129, + "loss": 3.2806, + "step": 38535 + }, + { + "epoch": 2.6185623046609594, + "grad_norm": 0.6392166018486023, + "learning_rate": 0.0006726797119173801, + "loss": 3.6332, + "step": 38540 + }, + { + "epoch": 2.618902024731621, + "grad_norm": 0.8629065752029419, + "learning_rate": 0.0006726372469085473, + "loss": 3.5773, + "step": 38545 + }, + { + "epoch": 2.6192417448022827, + "grad_norm": 0.6214338541030884, + "learning_rate": 0.0006725947818997147, + "loss": 3.4667, + "step": 38550 + }, + { + "epoch": 2.6195814648729447, + "grad_norm": 0.9043987989425659, + "learning_rate": 0.0006725523168908819, + "loss": 3.6781, + "step": 38555 + }, + { + "epoch": 2.6199211849436064, + "grad_norm": 0.8652035593986511, + "learning_rate": 0.0006725098518820491, + "loss": 3.5256, + "step": 38560 + }, + { + "epoch": 2.620260905014268, + "grad_norm": 0.9345974922180176, + "learning_rate": 0.0006724673868732166, + "loss": 3.5936, + "step": 38565 + }, + { + "epoch": 2.62060062508493, + "grad_norm": 0.8196595311164856, + "learning_rate": 0.0006724249218643838, + "loss": 3.8122, + "step": 38570 + }, + { + "epoch": 2.6209403451555917, + "grad_norm": 0.9962665438652039, + "learning_rate": 0.000672382456855551, + "loss": 3.5679, + "step": 38575 + }, + { + "epoch": 2.6212800652262533, + "grad_norm": 1.7092270851135254, + "learning_rate": 0.0006723399918467184, + "loss": 3.7124, + "step": 38580 + }, + { + "epoch": 2.6216197852969154, + "grad_norm": 0.903668224811554, + "learning_rate": 0.0006722975268378856, + "loss": 3.6283, + "step": 38585 + }, + { + "epoch": 2.621959505367577, + "grad_norm": 0.8718007206916809, + "learning_rate": 0.0006722550618290528, + "loss": 3.5193, + "step": 38590 + }, + { + "epoch": 2.6222992254382387, + "grad_norm": 1.2119618654251099, + "learning_rate": 0.0006722125968202201, + "loss": 3.6052, + "step": 38595 + }, + { + "epoch": 2.6226389455089008, + "grad_norm": 0.8903998136520386, + "learning_rate": 0.0006721701318113875, + "loss": 3.6537, + "step": 38600 + }, + { + "epoch": 2.6229786655795624, + "grad_norm": 0.8654888868331909, + "learning_rate": 0.0006721276668025547, + "loss": 3.7145, + "step": 38605 + }, + { + "epoch": 2.623318385650224, + "grad_norm": 0.8859283924102783, + "learning_rate": 0.000672085201793722, + "loss": 3.6125, + "step": 38610 + }, + { + "epoch": 2.623658105720886, + "grad_norm": 1.5566061735153198, + "learning_rate": 0.0006720427367848893, + "loss": 3.5142, + "step": 38615 + }, + { + "epoch": 2.6239978257915477, + "grad_norm": 0.9676165580749512, + "learning_rate": 0.0006720002717760565, + "loss": 3.5339, + "step": 38620 + }, + { + "epoch": 2.6243375458622094, + "grad_norm": 0.9105085134506226, + "learning_rate": 0.0006719578067672238, + "loss": 3.8566, + "step": 38625 + }, + { + "epoch": 2.6246772659328714, + "grad_norm": 0.8054528832435608, + "learning_rate": 0.000671915341758391, + "loss": 3.7914, + "step": 38630 + }, + { + "epoch": 2.625016986003533, + "grad_norm": 1.1351432800292969, + "learning_rate": 0.0006718728767495584, + "loss": 3.9082, + "step": 38635 + }, + { + "epoch": 2.6253567060741947, + "grad_norm": 1.2112489938735962, + "learning_rate": 0.0006718304117407257, + "loss": 3.7037, + "step": 38640 + }, + { + "epoch": 2.6256964261448568, + "grad_norm": 0.7334005832672119, + "learning_rate": 0.0006717879467318929, + "loss": 3.6933, + "step": 38645 + }, + { + "epoch": 2.6260361462155184, + "grad_norm": 0.817758321762085, + "learning_rate": 0.0006717454817230602, + "loss": 3.7132, + "step": 38650 + }, + { + "epoch": 2.62637586628618, + "grad_norm": 0.9176595211029053, + "learning_rate": 0.0006717030167142275, + "loss": 3.2833, + "step": 38655 + }, + { + "epoch": 2.626715586356842, + "grad_norm": 0.941903293132782, + "learning_rate": 0.0006716605517053947, + "loss": 3.5427, + "step": 38660 + }, + { + "epoch": 2.6270553064275037, + "grad_norm": 0.8159829378128052, + "learning_rate": 0.000671618086696562, + "loss": 3.3996, + "step": 38665 + }, + { + "epoch": 2.6273950264981654, + "grad_norm": 1.0653454065322876, + "learning_rate": 0.0006715756216877294, + "loss": 3.6255, + "step": 38670 + }, + { + "epoch": 2.6277347465688274, + "grad_norm": 1.0850754976272583, + "learning_rate": 0.0006715331566788966, + "loss": 3.4204, + "step": 38675 + }, + { + "epoch": 2.628074466639489, + "grad_norm": 0.9883540868759155, + "learning_rate": 0.000671490691670064, + "loss": 3.6716, + "step": 38680 + }, + { + "epoch": 2.6284141867101507, + "grad_norm": 0.6990470886230469, + "learning_rate": 0.0006714482266612312, + "loss": 3.7462, + "step": 38685 + }, + { + "epoch": 2.6287539067808128, + "grad_norm": 0.7547423839569092, + "learning_rate": 0.0006714057616523984, + "loss": 3.6905, + "step": 38690 + }, + { + "epoch": 2.6290936268514744, + "grad_norm": 1.020064115524292, + "learning_rate": 0.0006713632966435657, + "loss": 3.4894, + "step": 38695 + }, + { + "epoch": 2.629433346922136, + "grad_norm": 0.7413952946662903, + "learning_rate": 0.000671320831634733, + "loss": 3.5132, + "step": 38700 + }, + { + "epoch": 2.629773066992798, + "grad_norm": 0.7968048453330994, + "learning_rate": 0.0006712783666259003, + "loss": 3.7338, + "step": 38705 + }, + { + "epoch": 2.6301127870634597, + "grad_norm": 0.6279461979866028, + "learning_rate": 0.0006712359016170676, + "loss": 3.7586, + "step": 38710 + }, + { + "epoch": 2.6304525071341214, + "grad_norm": 1.3969736099243164, + "learning_rate": 0.0006711934366082349, + "loss": 3.6384, + "step": 38715 + }, + { + "epoch": 2.6307922272047835, + "grad_norm": 0.8817132711410522, + "learning_rate": 0.0006711509715994021, + "loss": 3.562, + "step": 38720 + }, + { + "epoch": 2.631131947275445, + "grad_norm": 0.7038147449493408, + "learning_rate": 0.0006711085065905694, + "loss": 3.5306, + "step": 38725 + }, + { + "epoch": 2.6314716673461067, + "grad_norm": 1.0569236278533936, + "learning_rate": 0.0006710660415817366, + "loss": 3.4945, + "step": 38730 + }, + { + "epoch": 2.631811387416769, + "grad_norm": 0.9933321475982666, + "learning_rate": 0.0006710235765729039, + "loss": 3.5395, + "step": 38735 + }, + { + "epoch": 2.6321511074874304, + "grad_norm": 0.9027732014656067, + "learning_rate": 0.0006709811115640713, + "loss": 3.583, + "step": 38740 + }, + { + "epoch": 2.632490827558092, + "grad_norm": 0.8695491552352905, + "learning_rate": 0.0006709386465552385, + "loss": 3.6478, + "step": 38745 + }, + { + "epoch": 2.632830547628754, + "grad_norm": 0.8274911046028137, + "learning_rate": 0.0006708961815464058, + "loss": 3.5436, + "step": 38750 + }, + { + "epoch": 2.6331702676994158, + "grad_norm": 1.0487310886383057, + "learning_rate": 0.0006708537165375731, + "loss": 3.2929, + "step": 38755 + }, + { + "epoch": 2.6335099877700774, + "grad_norm": 0.7929543852806091, + "learning_rate": 0.0006708112515287403, + "loss": 3.5429, + "step": 38760 + }, + { + "epoch": 2.6338497078407395, + "grad_norm": 0.7827832102775574, + "learning_rate": 0.0006707687865199076, + "loss": 3.4065, + "step": 38765 + }, + { + "epoch": 2.634189427911401, + "grad_norm": 1.0113779306411743, + "learning_rate": 0.0006707263215110749, + "loss": 3.6951, + "step": 38770 + }, + { + "epoch": 2.6345291479820627, + "grad_norm": 0.8027418255805969, + "learning_rate": 0.0006706838565022422, + "loss": 3.5684, + "step": 38775 + }, + { + "epoch": 2.634868868052725, + "grad_norm": 0.8337321877479553, + "learning_rate": 0.0006706413914934094, + "loss": 3.7862, + "step": 38780 + }, + { + "epoch": 2.6352085881233864, + "grad_norm": 0.9075456261634827, + "learning_rate": 0.0006705989264845768, + "loss": 3.8197, + "step": 38785 + }, + { + "epoch": 2.635548308194048, + "grad_norm": 1.1221232414245605, + "learning_rate": 0.000670556461475744, + "loss": 3.7347, + "step": 38790 + }, + { + "epoch": 2.63588802826471, + "grad_norm": 0.9430304765701294, + "learning_rate": 0.0006705139964669112, + "loss": 3.5805, + "step": 38795 + }, + { + "epoch": 2.6362277483353718, + "grad_norm": 0.8529255390167236, + "learning_rate": 0.0006704715314580786, + "loss": 3.6401, + "step": 38800 + }, + { + "epoch": 2.6365674684060334, + "grad_norm": 0.8448794484138489, + "learning_rate": 0.0006704290664492458, + "loss": 3.48, + "step": 38805 + }, + { + "epoch": 2.6369071884766955, + "grad_norm": 1.2087451219558716, + "learning_rate": 0.0006703866014404131, + "loss": 3.4545, + "step": 38810 + }, + { + "epoch": 2.637246908547357, + "grad_norm": 0.8385733366012573, + "learning_rate": 0.0006703441364315805, + "loss": 3.3509, + "step": 38815 + }, + { + "epoch": 2.6375866286180187, + "grad_norm": 0.9531806111335754, + "learning_rate": 0.0006703016714227477, + "loss": 3.8468, + "step": 38820 + }, + { + "epoch": 2.6379263486886804, + "grad_norm": 0.9492085576057434, + "learning_rate": 0.0006702592064139149, + "loss": 3.6461, + "step": 38825 + }, + { + "epoch": 2.6382660687593424, + "grad_norm": 0.7910784482955933, + "learning_rate": 0.0006702167414050822, + "loss": 3.5708, + "step": 38830 + }, + { + "epoch": 2.638605788830004, + "grad_norm": 0.7317068576812744, + "learning_rate": 0.0006701742763962495, + "loss": 3.7697, + "step": 38835 + }, + { + "epoch": 2.6389455089006657, + "grad_norm": 0.8129504919052124, + "learning_rate": 0.0006701318113874167, + "loss": 3.4037, + "step": 38840 + }, + { + "epoch": 2.639285228971328, + "grad_norm": 0.6434752345085144, + "learning_rate": 0.0006700893463785841, + "loss": 3.5381, + "step": 38845 + }, + { + "epoch": 2.6396249490419894, + "grad_norm": 1.0738664865493774, + "learning_rate": 0.0006700468813697514, + "loss": 3.4779, + "step": 38850 + }, + { + "epoch": 2.639964669112651, + "grad_norm": 0.9184576869010925, + "learning_rate": 0.0006700044163609186, + "loss": 3.5823, + "step": 38855 + }, + { + "epoch": 2.640304389183313, + "grad_norm": 0.8333171606063843, + "learning_rate": 0.0006699619513520859, + "loss": 3.535, + "step": 38860 + }, + { + "epoch": 2.6406441092539747, + "grad_norm": 0.8588457107543945, + "learning_rate": 0.0006699194863432532, + "loss": 3.579, + "step": 38865 + }, + { + "epoch": 2.6409838293246364, + "grad_norm": 0.7158852815628052, + "learning_rate": 0.0006698770213344204, + "loss": 3.4853, + "step": 38870 + }, + { + "epoch": 2.641323549395298, + "grad_norm": 0.9022689461708069, + "learning_rate": 0.0006698345563255877, + "loss": 3.7643, + "step": 38875 + }, + { + "epoch": 2.64166326946596, + "grad_norm": 25.67415428161621, + "learning_rate": 0.000669792091316755, + "loss": 3.6054, + "step": 38880 + }, + { + "epoch": 2.6420029895366217, + "grad_norm": 0.7990885972976685, + "learning_rate": 0.0006697496263079223, + "loss": 3.7794, + "step": 38885 + }, + { + "epoch": 2.6423427096072833, + "grad_norm": 0.7637774348258972, + "learning_rate": 0.0006697071612990896, + "loss": 3.6493, + "step": 38890 + }, + { + "epoch": 2.6426824296779454, + "grad_norm": 0.9356614947319031, + "learning_rate": 0.0006696646962902568, + "loss": 3.5832, + "step": 38895 + }, + { + "epoch": 2.643022149748607, + "grad_norm": 0.9159225225448608, + "learning_rate": 0.0006696222312814241, + "loss": 3.2256, + "step": 38900 + }, + { + "epoch": 2.6433618698192687, + "grad_norm": 0.966389000415802, + "learning_rate": 0.0006695797662725914, + "loss": 3.4058, + "step": 38905 + }, + { + "epoch": 2.6437015898899308, + "grad_norm": 0.7136698961257935, + "learning_rate": 0.0006695373012637586, + "loss": 3.4978, + "step": 38910 + }, + { + "epoch": 2.6440413099605924, + "grad_norm": 0.8356807827949524, + "learning_rate": 0.000669494836254926, + "loss": 3.5399, + "step": 38915 + }, + { + "epoch": 2.644381030031254, + "grad_norm": 0.8544842600822449, + "learning_rate": 0.0006694523712460933, + "loss": 3.7321, + "step": 38920 + }, + { + "epoch": 2.644720750101916, + "grad_norm": 0.8490853905677795, + "learning_rate": 0.0006694099062372605, + "loss": 3.3927, + "step": 38925 + }, + { + "epoch": 2.6450604701725777, + "grad_norm": 0.8461875319480896, + "learning_rate": 0.0006693674412284277, + "loss": 3.3326, + "step": 38930 + }, + { + "epoch": 2.6454001902432394, + "grad_norm": 0.8575462102890015, + "learning_rate": 0.0006693249762195951, + "loss": 3.3817, + "step": 38935 + }, + { + "epoch": 2.6457399103139014, + "grad_norm": 0.858337938785553, + "learning_rate": 0.0006692825112107623, + "loss": 3.4148, + "step": 38940 + }, + { + "epoch": 2.646079630384563, + "grad_norm": 0.7775519490242004, + "learning_rate": 0.0006692400462019295, + "loss": 3.6192, + "step": 38945 + }, + { + "epoch": 2.6464193504552247, + "grad_norm": 1.0075974464416504, + "learning_rate": 0.000669197581193097, + "loss": 3.4872, + "step": 38950 + }, + { + "epoch": 2.6467590705258868, + "grad_norm": 0.8635708689689636, + "learning_rate": 0.0006691551161842642, + "loss": 3.5131, + "step": 38955 + }, + { + "epoch": 2.6470987905965484, + "grad_norm": 0.9143396615982056, + "learning_rate": 0.0006691126511754314, + "loss": 3.2962, + "step": 38960 + }, + { + "epoch": 2.64743851066721, + "grad_norm": 0.8970433473587036, + "learning_rate": 0.0006690701861665988, + "loss": 3.4906, + "step": 38965 + }, + { + "epoch": 2.647778230737872, + "grad_norm": 0.838435173034668, + "learning_rate": 0.000669027721157766, + "loss": 3.5363, + "step": 38970 + }, + { + "epoch": 2.6481179508085337, + "grad_norm": 0.7684668302536011, + "learning_rate": 0.0006689852561489332, + "loss": 3.5185, + "step": 38975 + }, + { + "epoch": 2.6484576708791954, + "grad_norm": 0.9158735275268555, + "learning_rate": 0.0006689427911401005, + "loss": 3.719, + "step": 38980 + }, + { + "epoch": 2.6487973909498574, + "grad_norm": 0.908362090587616, + "learning_rate": 0.0006689003261312679, + "loss": 3.4921, + "step": 38985 + }, + { + "epoch": 2.649137111020519, + "grad_norm": 0.8787447214126587, + "learning_rate": 0.0006688578611224351, + "loss": 3.5257, + "step": 38990 + }, + { + "epoch": 2.6494768310911807, + "grad_norm": 0.9223270416259766, + "learning_rate": 0.0006688153961136024, + "loss": 3.3577, + "step": 38995 + }, + { + "epoch": 2.649816551161843, + "grad_norm": 0.9562312960624695, + "learning_rate": 0.0006687729311047697, + "loss": 3.695, + "step": 39000 + }, + { + "epoch": 2.6501562712325044, + "grad_norm": 1.0691815614700317, + "learning_rate": 0.0006687304660959369, + "loss": 3.4451, + "step": 39005 + }, + { + "epoch": 2.650495991303166, + "grad_norm": 0.8978657722473145, + "learning_rate": 0.0006686880010871042, + "loss": 3.6595, + "step": 39010 + }, + { + "epoch": 2.650835711373828, + "grad_norm": 0.9087507724761963, + "learning_rate": 0.0006686455360782714, + "loss": 3.6368, + "step": 39015 + }, + { + "epoch": 2.6511754314444897, + "grad_norm": 0.9738811254501343, + "learning_rate": 0.0006686030710694389, + "loss": 3.3309, + "step": 39020 + }, + { + "epoch": 2.6515151515151514, + "grad_norm": 0.7656195759773254, + "learning_rate": 0.0006685606060606061, + "loss": 3.6944, + "step": 39025 + }, + { + "epoch": 2.6518548715858135, + "grad_norm": 0.8604965806007385, + "learning_rate": 0.0006685181410517733, + "loss": 3.7169, + "step": 39030 + }, + { + "epoch": 2.652194591656475, + "grad_norm": 0.8998962044715881, + "learning_rate": 0.0006684756760429407, + "loss": 3.7117, + "step": 39035 + }, + { + "epoch": 2.6525343117271367, + "grad_norm": 0.9582512974739075, + "learning_rate": 0.0006684332110341079, + "loss": 3.646, + "step": 39040 + }, + { + "epoch": 2.652874031797799, + "grad_norm": 0.8939792513847351, + "learning_rate": 0.0006683907460252751, + "loss": 3.4811, + "step": 39045 + }, + { + "epoch": 2.6532137518684604, + "grad_norm": 0.901568591594696, + "learning_rate": 0.0006683482810164426, + "loss": 3.4898, + "step": 39050 + }, + { + "epoch": 2.653553471939122, + "grad_norm": 1.177497386932373, + "learning_rate": 0.0006683058160076098, + "loss": 3.1865, + "step": 39055 + }, + { + "epoch": 2.653893192009784, + "grad_norm": 1.246700406074524, + "learning_rate": 0.000668263350998777, + "loss": 3.7413, + "step": 39060 + }, + { + "epoch": 2.6542329120804458, + "grad_norm": 0.8685649037361145, + "learning_rate": 0.0006682208859899444, + "loss": 3.2871, + "step": 39065 + }, + { + "epoch": 2.6545726321511074, + "grad_norm": 0.7652031779289246, + "learning_rate": 0.0006681784209811116, + "loss": 3.8107, + "step": 39070 + }, + { + "epoch": 2.6549123522217695, + "grad_norm": 1.0441315174102783, + "learning_rate": 0.0006681359559722788, + "loss": 3.4456, + "step": 39075 + }, + { + "epoch": 2.655252072292431, + "grad_norm": 0.8567469120025635, + "learning_rate": 0.0006680934909634461, + "loss": 3.5856, + "step": 39080 + }, + { + "epoch": 2.6555917923630927, + "grad_norm": 0.7644731402397156, + "learning_rate": 0.0006680510259546135, + "loss": 3.6342, + "step": 39085 + }, + { + "epoch": 2.655931512433755, + "grad_norm": 0.6943089962005615, + "learning_rate": 0.0006680085609457807, + "loss": 3.5026, + "step": 39090 + }, + { + "epoch": 2.6562712325044164, + "grad_norm": 0.7880464196205139, + "learning_rate": 0.000667966095936948, + "loss": 3.5199, + "step": 39095 + }, + { + "epoch": 2.656610952575078, + "grad_norm": 1.070072889328003, + "learning_rate": 0.0006679236309281153, + "loss": 3.5301, + "step": 39100 + }, + { + "epoch": 2.65695067264574, + "grad_norm": 0.7754001617431641, + "learning_rate": 0.0006678811659192825, + "loss": 3.4531, + "step": 39105 + }, + { + "epoch": 2.6572903927164018, + "grad_norm": 0.9749760031700134, + "learning_rate": 0.0006678387009104498, + "loss": 3.2918, + "step": 39110 + }, + { + "epoch": 2.6576301127870634, + "grad_norm": 0.9373928904533386, + "learning_rate": 0.000667796235901617, + "loss": 3.4242, + "step": 39115 + }, + { + "epoch": 2.6579698328577255, + "grad_norm": 0.8050748109817505, + "learning_rate": 0.0006677537708927844, + "loss": 3.7281, + "step": 39120 + }, + { + "epoch": 2.658309552928387, + "grad_norm": 0.8123527765274048, + "learning_rate": 0.0006677113058839517, + "loss": 3.2858, + "step": 39125 + }, + { + "epoch": 2.6586492729990487, + "grad_norm": 0.9150041937828064, + "learning_rate": 0.0006676688408751189, + "loss": 3.4277, + "step": 39130 + }, + { + "epoch": 2.658988993069711, + "grad_norm": 0.8619819283485413, + "learning_rate": 0.0006676263758662862, + "loss": 3.6669, + "step": 39135 + }, + { + "epoch": 2.6593287131403724, + "grad_norm": 0.7923856973648071, + "learning_rate": 0.0006675839108574535, + "loss": 3.4763, + "step": 39140 + }, + { + "epoch": 2.659668433211034, + "grad_norm": 0.816271960735321, + "learning_rate": 0.0006675414458486207, + "loss": 3.3649, + "step": 39145 + }, + { + "epoch": 2.660008153281696, + "grad_norm": 0.7809050679206848, + "learning_rate": 0.000667498980839788, + "loss": 3.26, + "step": 39150 + }, + { + "epoch": 2.660347873352358, + "grad_norm": 0.7251778841018677, + "learning_rate": 0.0006674565158309554, + "loss": 3.9465, + "step": 39155 + }, + { + "epoch": 2.6606875934230194, + "grad_norm": 0.9981110692024231, + "learning_rate": 0.0006674140508221226, + "loss": 3.4462, + "step": 39160 + }, + { + "epoch": 2.661027313493681, + "grad_norm": 1.0442508459091187, + "learning_rate": 0.0006673715858132898, + "loss": 3.7769, + "step": 39165 + }, + { + "epoch": 2.661367033564343, + "grad_norm": 0.9589933156967163, + "learning_rate": 0.0006673291208044572, + "loss": 3.5334, + "step": 39170 + }, + { + "epoch": 2.6617067536350048, + "grad_norm": 0.8386595845222473, + "learning_rate": 0.0006672866557956244, + "loss": 3.4662, + "step": 39175 + }, + { + "epoch": 2.6620464737056664, + "grad_norm": 0.823900580406189, + "learning_rate": 0.0006672441907867916, + "loss": 3.4005, + "step": 39180 + }, + { + "epoch": 2.6623861937763285, + "grad_norm": 0.8166152834892273, + "learning_rate": 0.000667201725777959, + "loss": 3.284, + "step": 39185 + }, + { + "epoch": 2.66272591384699, + "grad_norm": 1.0028407573699951, + "learning_rate": 0.0006671592607691263, + "loss": 3.4659, + "step": 39190 + }, + { + "epoch": 2.6630656339176517, + "grad_norm": 0.8185228109359741, + "learning_rate": 0.0006671167957602935, + "loss": 3.6145, + "step": 39195 + }, + { + "epoch": 2.663405353988314, + "grad_norm": 0.9407854080200195, + "learning_rate": 0.0006670743307514609, + "loss": 3.7606, + "step": 39200 + }, + { + "epoch": 2.6637450740589754, + "grad_norm": 1.0144007205963135, + "learning_rate": 0.0006670318657426281, + "loss": 3.4303, + "step": 39205 + }, + { + "epoch": 2.664084794129637, + "grad_norm": 0.9002841114997864, + "learning_rate": 0.0006669894007337953, + "loss": 3.5891, + "step": 39210 + }, + { + "epoch": 2.6644245142002987, + "grad_norm": 0.8895481824874878, + "learning_rate": 0.0006669469357249627, + "loss": 3.4951, + "step": 39215 + }, + { + "epoch": 2.6647642342709608, + "grad_norm": 0.8655336499214172, + "learning_rate": 0.0006669044707161299, + "loss": 3.3381, + "step": 39220 + }, + { + "epoch": 2.6651039543416224, + "grad_norm": 0.7400758266448975, + "learning_rate": 0.0006668620057072972, + "loss": 3.7898, + "step": 39225 + }, + { + "epoch": 2.665443674412284, + "grad_norm": 0.9576754570007324, + "learning_rate": 0.0006668195406984645, + "loss": 3.4518, + "step": 39230 + }, + { + "epoch": 2.665783394482946, + "grad_norm": 0.9104257225990295, + "learning_rate": 0.0006667770756896318, + "loss": 3.4648, + "step": 39235 + }, + { + "epoch": 2.6661231145536077, + "grad_norm": 0.9224026799201965, + "learning_rate": 0.000666734610680799, + "loss": 3.4527, + "step": 39240 + }, + { + "epoch": 2.6664628346242694, + "grad_norm": 0.7793877124786377, + "learning_rate": 0.0006666921456719663, + "loss": 3.53, + "step": 39245 + }, + { + "epoch": 2.6668025546949314, + "grad_norm": 0.7612549066543579, + "learning_rate": 0.0006666496806631336, + "loss": 3.6605, + "step": 39250 + }, + { + "epoch": 2.667142274765593, + "grad_norm": 0.685893177986145, + "learning_rate": 0.0006666072156543008, + "loss": 3.3682, + "step": 39255 + }, + { + "epoch": 2.6674819948362547, + "grad_norm": 1.1559594869613647, + "learning_rate": 0.0006665647506454682, + "loss": 3.7983, + "step": 39260 + }, + { + "epoch": 2.6678217149069168, + "grad_norm": 0.9116805195808411, + "learning_rate": 0.0006665222856366355, + "loss": 3.3838, + "step": 39265 + }, + { + "epoch": 2.6681614349775784, + "grad_norm": 0.8146975636482239, + "learning_rate": 0.0006664798206278027, + "loss": 3.4497, + "step": 39270 + }, + { + "epoch": 2.66850115504824, + "grad_norm": 0.7309738993644714, + "learning_rate": 0.00066643735561897, + "loss": 3.7783, + "step": 39275 + }, + { + "epoch": 2.668840875118902, + "grad_norm": 1.1440026760101318, + "learning_rate": 0.0006663948906101372, + "loss": 3.624, + "step": 39280 + }, + { + "epoch": 2.6691805951895637, + "grad_norm": 1.3363008499145508, + "learning_rate": 0.0006663524256013045, + "loss": 3.8108, + "step": 39285 + }, + { + "epoch": 2.6695203152602254, + "grad_norm": 0.9720861315727234, + "learning_rate": 0.0006663099605924718, + "loss": 3.5352, + "step": 39290 + }, + { + "epoch": 2.6698600353308874, + "grad_norm": 0.8297094106674194, + "learning_rate": 0.0006662674955836391, + "loss": 3.7056, + "step": 39295 + }, + { + "epoch": 2.670199755401549, + "grad_norm": 0.8896045684814453, + "learning_rate": 0.0006662250305748064, + "loss": 3.561, + "step": 39300 + }, + { + "epoch": 2.6705394754722107, + "grad_norm": 0.8745763301849365, + "learning_rate": 0.0006661825655659737, + "loss": 3.4967, + "step": 39305 + }, + { + "epoch": 2.670879195542873, + "grad_norm": 0.8838697671890259, + "learning_rate": 0.0006661401005571409, + "loss": 3.5506, + "step": 39310 + }, + { + "epoch": 2.6712189156135344, + "grad_norm": 0.9359268546104431, + "learning_rate": 0.0006660976355483081, + "loss": 3.6081, + "step": 39315 + }, + { + "epoch": 2.671558635684196, + "grad_norm": 0.9610401391983032, + "learning_rate": 0.0006660551705394755, + "loss": 3.8735, + "step": 39320 + }, + { + "epoch": 2.671898355754858, + "grad_norm": 0.7464884519577026, + "learning_rate": 0.0006660127055306427, + "loss": 3.6112, + "step": 39325 + }, + { + "epoch": 2.6722380758255198, + "grad_norm": 0.7213006615638733, + "learning_rate": 0.00066597024052181, + "loss": 3.6616, + "step": 39330 + }, + { + "epoch": 2.6725777958961814, + "grad_norm": 0.8091405034065247, + "learning_rate": 0.0006659277755129774, + "loss": 3.4883, + "step": 39335 + }, + { + "epoch": 2.6729175159668435, + "grad_norm": 0.9522737264633179, + "learning_rate": 0.0006658853105041446, + "loss": 3.7208, + "step": 39340 + }, + { + "epoch": 2.673257236037505, + "grad_norm": 1.0663467645645142, + "learning_rate": 0.0006658428454953118, + "loss": 3.4942, + "step": 39345 + }, + { + "epoch": 2.6735969561081667, + "grad_norm": 0.8880877494812012, + "learning_rate": 0.0006658003804864792, + "loss": 3.5162, + "step": 39350 + }, + { + "epoch": 2.673936676178829, + "grad_norm": 0.7805944085121155, + "learning_rate": 0.0006657579154776464, + "loss": 3.6987, + "step": 39355 + }, + { + "epoch": 2.6742763962494904, + "grad_norm": 0.7099069952964783, + "learning_rate": 0.0006657154504688137, + "loss": 3.3643, + "step": 39360 + }, + { + "epoch": 2.674616116320152, + "grad_norm": 0.7927857041358948, + "learning_rate": 0.000665672985459981, + "loss": 3.4248, + "step": 39365 + }, + { + "epoch": 2.674955836390814, + "grad_norm": 0.8989330530166626, + "learning_rate": 0.0006656305204511483, + "loss": 3.5511, + "step": 39370 + }, + { + "epoch": 2.6752955564614758, + "grad_norm": 1.0804892778396606, + "learning_rate": 0.0006655880554423156, + "loss": 3.6852, + "step": 39375 + }, + { + "epoch": 2.6756352765321374, + "grad_norm": 0.7740371227264404, + "learning_rate": 0.0006655455904334828, + "loss": 3.4718, + "step": 39380 + }, + { + "epoch": 2.6759749966027995, + "grad_norm": 0.7535105347633362, + "learning_rate": 0.0006655031254246501, + "loss": 3.4063, + "step": 39385 + }, + { + "epoch": 2.676314716673461, + "grad_norm": 0.8613489270210266, + "learning_rate": 0.0006654606604158174, + "loss": 3.716, + "step": 39390 + }, + { + "epoch": 2.6766544367441227, + "grad_norm": 0.9653170704841614, + "learning_rate": 0.0006654181954069846, + "loss": 3.7586, + "step": 39395 + }, + { + "epoch": 2.676994156814785, + "grad_norm": 0.7744818329811096, + "learning_rate": 0.000665375730398152, + "loss": 3.7027, + "step": 39400 + }, + { + "epoch": 2.6773338768854464, + "grad_norm": 0.8145372867584229, + "learning_rate": 0.0006653332653893193, + "loss": 3.4915, + "step": 39405 + }, + { + "epoch": 2.677673596956108, + "grad_norm": 0.9243924021720886, + "learning_rate": 0.0006652908003804865, + "loss": 3.3064, + "step": 39410 + }, + { + "epoch": 2.67801331702677, + "grad_norm": 0.7611715197563171, + "learning_rate": 0.0006652483353716537, + "loss": 3.6374, + "step": 39415 + }, + { + "epoch": 2.6783530370974318, + "grad_norm": 0.9148041605949402, + "learning_rate": 0.0006652058703628211, + "loss": 3.6513, + "step": 39420 + }, + { + "epoch": 2.6786927571680934, + "grad_norm": 0.954589307308197, + "learning_rate": 0.0006651634053539883, + "loss": 3.6613, + "step": 39425 + }, + { + "epoch": 2.6790324772387555, + "grad_norm": 1.2141176462173462, + "learning_rate": 0.0006651209403451555, + "loss": 3.6164, + "step": 39430 + }, + { + "epoch": 2.679372197309417, + "grad_norm": 0.7941005229949951, + "learning_rate": 0.000665078475336323, + "loss": 3.5494, + "step": 39435 + }, + { + "epoch": 2.6797119173800787, + "grad_norm": 0.7552561163902283, + "learning_rate": 0.0006650360103274902, + "loss": 3.5026, + "step": 39440 + }, + { + "epoch": 2.680051637450741, + "grad_norm": 0.8646746873855591, + "learning_rate": 0.0006649935453186574, + "loss": 3.6857, + "step": 39445 + }, + { + "epoch": 2.6803913575214025, + "grad_norm": 0.719590425491333, + "learning_rate": 0.0006649510803098248, + "loss": 3.7175, + "step": 39450 + }, + { + "epoch": 2.680731077592064, + "grad_norm": 0.8356661200523376, + "learning_rate": 0.000664908615300992, + "loss": 3.5311, + "step": 39455 + }, + { + "epoch": 2.681070797662726, + "grad_norm": 0.8313671350479126, + "learning_rate": 0.0006648661502921592, + "loss": 3.5614, + "step": 39460 + }, + { + "epoch": 2.681410517733388, + "grad_norm": 0.8496318459510803, + "learning_rate": 0.0006648236852833265, + "loss": 3.4067, + "step": 39465 + }, + { + "epoch": 2.6817502378040494, + "grad_norm": 0.7232120037078857, + "learning_rate": 0.0006647812202744939, + "loss": 3.6396, + "step": 39470 + }, + { + "epoch": 2.6820899578747115, + "grad_norm": 0.7688809037208557, + "learning_rate": 0.0006647387552656611, + "loss": 3.4844, + "step": 39475 + }, + { + "epoch": 2.682429677945373, + "grad_norm": 0.7387804388999939, + "learning_rate": 0.0006646962902568284, + "loss": 3.4331, + "step": 39480 + }, + { + "epoch": 2.6827693980160348, + "grad_norm": 1.0792258977890015, + "learning_rate": 0.0006646538252479957, + "loss": 3.6457, + "step": 39485 + }, + { + "epoch": 2.683109118086697, + "grad_norm": 0.8235520720481873, + "learning_rate": 0.0006646113602391629, + "loss": 3.7389, + "step": 39490 + }, + { + "epoch": 2.6834488381573585, + "grad_norm": 0.9406498074531555, + "learning_rate": 0.0006645688952303302, + "loss": 3.4926, + "step": 39495 + }, + { + "epoch": 2.68378855822802, + "grad_norm": 0.8052217960357666, + "learning_rate": 0.0006645264302214975, + "loss": 3.4535, + "step": 39500 + }, + { + "epoch": 2.6841282782986817, + "grad_norm": 1.107268214225769, + "learning_rate": 0.0006644839652126648, + "loss": 3.5002, + "step": 39505 + }, + { + "epoch": 2.684467998369344, + "grad_norm": 0.6901599168777466, + "learning_rate": 0.0006644415002038321, + "loss": 3.6099, + "step": 39510 + }, + { + "epoch": 2.6848077184400054, + "grad_norm": 0.9755265712738037, + "learning_rate": 0.0006643990351949993, + "loss": 3.4136, + "step": 39515 + }, + { + "epoch": 2.685147438510667, + "grad_norm": 1.1893476247787476, + "learning_rate": 0.0006643565701861666, + "loss": 3.3938, + "step": 39520 + }, + { + "epoch": 2.685487158581329, + "grad_norm": 1.077042818069458, + "learning_rate": 0.0006643141051773339, + "loss": 3.4097, + "step": 39525 + }, + { + "epoch": 2.6858268786519908, + "grad_norm": 0.9407221078872681, + "learning_rate": 0.0006642716401685011, + "loss": 3.5794, + "step": 39530 + }, + { + "epoch": 2.6861665987226524, + "grad_norm": 0.814281165599823, + "learning_rate": 0.0006642291751596684, + "loss": 3.7497, + "step": 39535 + }, + { + "epoch": 2.6865063187933145, + "grad_norm": 0.9309245347976685, + "learning_rate": 0.0006641867101508358, + "loss": 3.5242, + "step": 39540 + }, + { + "epoch": 2.686846038863976, + "grad_norm": 1.1620299816131592, + "learning_rate": 0.000664144245142003, + "loss": 3.3924, + "step": 39545 + }, + { + "epoch": 2.6871857589346377, + "grad_norm": 1.1696833372116089, + "learning_rate": 0.0006641017801331703, + "loss": 3.5454, + "step": 39550 + }, + { + "epoch": 2.6875254790052994, + "grad_norm": 1.1456794738769531, + "learning_rate": 0.0006640593151243376, + "loss": 3.6153, + "step": 39555 + }, + { + "epoch": 2.6878651990759614, + "grad_norm": 0.8372112512588501, + "learning_rate": 0.0006640168501155048, + "loss": 3.4789, + "step": 39560 + }, + { + "epoch": 2.688204919146623, + "grad_norm": 0.7948510646820068, + "learning_rate": 0.000663974385106672, + "loss": 3.6489, + "step": 39565 + }, + { + "epoch": 2.6885446392172847, + "grad_norm": 1.6276192665100098, + "learning_rate": 0.0006639319200978394, + "loss": 3.5652, + "step": 39570 + }, + { + "epoch": 2.688884359287947, + "grad_norm": 0.7661606669425964, + "learning_rate": 0.0006638894550890067, + "loss": 3.5418, + "step": 39575 + }, + { + "epoch": 2.6892240793586084, + "grad_norm": 0.8404340744018555, + "learning_rate": 0.0006638469900801739, + "loss": 3.4948, + "step": 39580 + }, + { + "epoch": 2.68956379942927, + "grad_norm": 1.0753207206726074, + "learning_rate": 0.0006638045250713413, + "loss": 3.854, + "step": 39585 + }, + { + "epoch": 2.689903519499932, + "grad_norm": 0.9387730360031128, + "learning_rate": 0.0006637620600625085, + "loss": 3.6577, + "step": 39590 + }, + { + "epoch": 2.6902432395705937, + "grad_norm": 0.7888733148574829, + "learning_rate": 0.0006637195950536757, + "loss": 3.6716, + "step": 39595 + }, + { + "epoch": 2.6905829596412554, + "grad_norm": 0.977247953414917, + "learning_rate": 0.000663677130044843, + "loss": 3.461, + "step": 39600 + }, + { + "epoch": 2.6909226797119175, + "grad_norm": 1.0005854368209839, + "learning_rate": 0.0006636346650360103, + "loss": 3.559, + "step": 39605 + }, + { + "epoch": 2.691262399782579, + "grad_norm": 0.9771857857704163, + "learning_rate": 0.0006635922000271776, + "loss": 3.5618, + "step": 39610 + }, + { + "epoch": 2.6916021198532407, + "grad_norm": 0.7702210545539856, + "learning_rate": 0.000663549735018345, + "loss": 3.7666, + "step": 39615 + }, + { + "epoch": 2.691941839923903, + "grad_norm": 0.7041729092597961, + "learning_rate": 0.0006635072700095122, + "loss": 3.6083, + "step": 39620 + }, + { + "epoch": 2.6922815599945644, + "grad_norm": 0.9051448106765747, + "learning_rate": 0.0006634648050006794, + "loss": 3.5937, + "step": 39625 + }, + { + "epoch": 2.692621280065226, + "grad_norm": 0.9443694353103638, + "learning_rate": 0.0006634223399918467, + "loss": 3.521, + "step": 39630 + }, + { + "epoch": 2.692961000135888, + "grad_norm": 0.9342772960662842, + "learning_rate": 0.000663379874983014, + "loss": 3.4778, + "step": 39635 + }, + { + "epoch": 2.6933007202065498, + "grad_norm": 0.9229065775871277, + "learning_rate": 0.0006633374099741812, + "loss": 3.4587, + "step": 39640 + }, + { + "epoch": 2.6936404402772114, + "grad_norm": 0.9385390281677246, + "learning_rate": 0.0006632949449653486, + "loss": 3.3121, + "step": 39645 + }, + { + "epoch": 2.6939801603478735, + "grad_norm": 0.8475236892700195, + "learning_rate": 0.0006632524799565159, + "loss": 3.5958, + "step": 39650 + }, + { + "epoch": 2.694319880418535, + "grad_norm": 0.8443287014961243, + "learning_rate": 0.0006632100149476831, + "loss": 3.4901, + "step": 39655 + }, + { + "epoch": 2.6946596004891967, + "grad_norm": 1.015817642211914, + "learning_rate": 0.0006631675499388504, + "loss": 3.6771, + "step": 39660 + }, + { + "epoch": 2.694999320559859, + "grad_norm": 0.7274404168128967, + "learning_rate": 0.0006631250849300176, + "loss": 3.6129, + "step": 39665 + }, + { + "epoch": 2.6953390406305204, + "grad_norm": 0.7245690822601318, + "learning_rate": 0.0006630826199211849, + "loss": 3.4489, + "step": 39670 + }, + { + "epoch": 2.695678760701182, + "grad_norm": 1.0628036260604858, + "learning_rate": 0.0006630401549123523, + "loss": 3.3562, + "step": 39675 + }, + { + "epoch": 2.696018480771844, + "grad_norm": 0.8156663179397583, + "learning_rate": 0.0006629976899035195, + "loss": 3.4015, + "step": 39680 + }, + { + "epoch": 2.6963582008425058, + "grad_norm": 0.961750328540802, + "learning_rate": 0.0006629552248946868, + "loss": 3.4847, + "step": 39685 + }, + { + "epoch": 2.6966979209131674, + "grad_norm": 0.8796713948249817, + "learning_rate": 0.0006629127598858541, + "loss": 3.4862, + "step": 39690 + }, + { + "epoch": 2.6970376409838295, + "grad_norm": 0.7819221615791321, + "learning_rate": 0.0006628702948770213, + "loss": 3.5789, + "step": 39695 + }, + { + "epoch": 2.697377361054491, + "grad_norm": 0.7031370401382446, + "learning_rate": 0.0006628278298681887, + "loss": 3.6516, + "step": 39700 + }, + { + "epoch": 2.6977170811251527, + "grad_norm": 0.7477797865867615, + "learning_rate": 0.0006627853648593559, + "loss": 3.4195, + "step": 39705 + }, + { + "epoch": 2.698056801195815, + "grad_norm": 0.8148168325424194, + "learning_rate": 0.0006627428998505232, + "loss": 3.5894, + "step": 39710 + }, + { + "epoch": 2.6983965212664764, + "grad_norm": 1.1020723581314087, + "learning_rate": 0.0006627004348416905, + "loss": 3.4635, + "step": 39715 + }, + { + "epoch": 2.698736241337138, + "grad_norm": 0.9621317982673645, + "learning_rate": 0.0006626579698328578, + "loss": 3.752, + "step": 39720 + }, + { + "epoch": 2.6990759614078, + "grad_norm": 0.6750820279121399, + "learning_rate": 0.000662615504824025, + "loss": 3.5458, + "step": 39725 + }, + { + "epoch": 2.699415681478462, + "grad_norm": 0.6667659282684326, + "learning_rate": 0.0006625730398151923, + "loss": 3.4146, + "step": 39730 + }, + { + "epoch": 2.6997554015491234, + "grad_norm": 1.2838027477264404, + "learning_rate": 0.0006625305748063596, + "loss": 3.563, + "step": 39735 + }, + { + "epoch": 2.7000951216197855, + "grad_norm": 0.7612210512161255, + "learning_rate": 0.0006624881097975268, + "loss": 3.2695, + "step": 39740 + }, + { + "epoch": 2.700434841690447, + "grad_norm": 0.7487524747848511, + "learning_rate": 0.0006624456447886942, + "loss": 3.6091, + "step": 39745 + }, + { + "epoch": 2.7007745617611087, + "grad_norm": 0.9047382473945618, + "learning_rate": 0.0006624031797798615, + "loss": 3.6756, + "step": 39750 + }, + { + "epoch": 2.701114281831771, + "grad_norm": 0.7021672129631042, + "learning_rate": 0.0006623607147710287, + "loss": 3.5139, + "step": 39755 + }, + { + "epoch": 2.7014540019024325, + "grad_norm": 1.0043939352035522, + "learning_rate": 0.000662318249762196, + "loss": 3.6311, + "step": 39760 + }, + { + "epoch": 2.701793721973094, + "grad_norm": 0.7244371771812439, + "learning_rate": 0.0006622757847533632, + "loss": 3.5967, + "step": 39765 + }, + { + "epoch": 2.702133442043756, + "grad_norm": 0.9074171185493469, + "learning_rate": 0.0006622333197445305, + "loss": 3.47, + "step": 39770 + }, + { + "epoch": 2.702473162114418, + "grad_norm": 0.8315871953964233, + "learning_rate": 0.0006621908547356978, + "loss": 3.3909, + "step": 39775 + }, + { + "epoch": 2.7028128821850794, + "grad_norm": 1.0892366170883179, + "learning_rate": 0.0006621483897268651, + "loss": 3.3348, + "step": 39780 + }, + { + "epoch": 2.7031526022557415, + "grad_norm": 0.9583085179328918, + "learning_rate": 0.0006621059247180324, + "loss": 3.4619, + "step": 39785 + }, + { + "epoch": 2.703492322326403, + "grad_norm": 0.9405403137207031, + "learning_rate": 0.0006620634597091997, + "loss": 3.7066, + "step": 39790 + }, + { + "epoch": 2.7038320423970648, + "grad_norm": 1.1698927879333496, + "learning_rate": 0.0006620209947003669, + "loss": 3.5143, + "step": 39795 + }, + { + "epoch": 2.704171762467727, + "grad_norm": 1.123669147491455, + "learning_rate": 0.0006619785296915341, + "loss": 3.5656, + "step": 39800 + }, + { + "epoch": 2.7045114825383885, + "grad_norm": 0.8114689588546753, + "learning_rate": 0.0006619360646827015, + "loss": 3.4688, + "step": 39805 + }, + { + "epoch": 2.70485120260905, + "grad_norm": 0.8444017767906189, + "learning_rate": 0.0006618935996738687, + "loss": 3.5631, + "step": 39810 + }, + { + "epoch": 2.705190922679712, + "grad_norm": 0.8161699771881104, + "learning_rate": 0.000661851134665036, + "loss": 3.4663, + "step": 39815 + }, + { + "epoch": 2.705530642750374, + "grad_norm": 0.8161278367042542, + "learning_rate": 0.0006618086696562034, + "loss": 3.6632, + "step": 39820 + }, + { + "epoch": 2.7058703628210354, + "grad_norm": 0.8805455565452576, + "learning_rate": 0.0006617662046473706, + "loss": 3.195, + "step": 39825 + }, + { + "epoch": 2.7062100828916975, + "grad_norm": 0.8140993714332581, + "learning_rate": 0.0006617237396385378, + "loss": 3.4359, + "step": 39830 + }, + { + "epoch": 2.706549802962359, + "grad_norm": 0.9666776061058044, + "learning_rate": 0.0006616812746297052, + "loss": 3.5496, + "step": 39835 + }, + { + "epoch": 2.7068895230330208, + "grad_norm": 0.9289587736129761, + "learning_rate": 0.0006616388096208724, + "loss": 3.3168, + "step": 39840 + }, + { + "epoch": 2.7072292431036824, + "grad_norm": 0.7441180944442749, + "learning_rate": 0.0006615963446120396, + "loss": 3.4691, + "step": 39845 + }, + { + "epoch": 2.7075689631743445, + "grad_norm": 0.8969364166259766, + "learning_rate": 0.0006615538796032071, + "loss": 3.6486, + "step": 39850 + }, + { + "epoch": 2.707908683245006, + "grad_norm": 0.8045446872711182, + "learning_rate": 0.0006615114145943743, + "loss": 3.4694, + "step": 39855 + }, + { + "epoch": 2.7082484033156677, + "grad_norm": 0.8576960563659668, + "learning_rate": 0.0006614689495855415, + "loss": 3.5353, + "step": 39860 + }, + { + "epoch": 2.70858812338633, + "grad_norm": 0.8782486319541931, + "learning_rate": 0.0006614264845767088, + "loss": 3.6126, + "step": 39865 + }, + { + "epoch": 2.7089278434569914, + "grad_norm": 1.0336476564407349, + "learning_rate": 0.0006613840195678761, + "loss": 3.6594, + "step": 39870 + }, + { + "epoch": 2.709267563527653, + "grad_norm": 0.8430631756782532, + "learning_rate": 0.0006613415545590433, + "loss": 3.48, + "step": 39875 + }, + { + "epoch": 2.709607283598315, + "grad_norm": 1.1418321132659912, + "learning_rate": 0.0006612990895502106, + "loss": 3.5315, + "step": 39880 + }, + { + "epoch": 2.709947003668977, + "grad_norm": 0.7815362811088562, + "learning_rate": 0.000661256624541378, + "loss": 3.68, + "step": 39885 + }, + { + "epoch": 2.7102867237396384, + "grad_norm": 0.7009844183921814, + "learning_rate": 0.0006612141595325452, + "loss": 3.6887, + "step": 39890 + }, + { + "epoch": 2.7106264438103, + "grad_norm": 0.7809376120567322, + "learning_rate": 0.0006611716945237125, + "loss": 3.5866, + "step": 39895 + }, + { + "epoch": 2.710966163880962, + "grad_norm": 0.8068056702613831, + "learning_rate": 0.0006611292295148797, + "loss": 3.3631, + "step": 39900 + }, + { + "epoch": 2.7113058839516238, + "grad_norm": 1.1175657510757446, + "learning_rate": 0.000661086764506047, + "loss": 3.6421, + "step": 39905 + }, + { + "epoch": 2.7116456040222854, + "grad_norm": 0.8177772760391235, + "learning_rate": 0.0006610442994972143, + "loss": 3.5571, + "step": 39910 + }, + { + "epoch": 2.7119853240929475, + "grad_norm": 0.8902101516723633, + "learning_rate": 0.0006610018344883815, + "loss": 3.5564, + "step": 39915 + }, + { + "epoch": 2.712325044163609, + "grad_norm": 0.736126720905304, + "learning_rate": 0.0006609593694795489, + "loss": 3.7114, + "step": 39920 + }, + { + "epoch": 2.7126647642342707, + "grad_norm": 0.7346240878105164, + "learning_rate": 0.0006609169044707162, + "loss": 3.7131, + "step": 39925 + }, + { + "epoch": 2.713004484304933, + "grad_norm": 1.0551403760910034, + "learning_rate": 0.0006608744394618834, + "loss": 3.7793, + "step": 39930 + }, + { + "epoch": 2.7133442043755944, + "grad_norm": 1.1357228755950928, + "learning_rate": 0.0006608319744530507, + "loss": 3.674, + "step": 39935 + }, + { + "epoch": 2.713683924446256, + "grad_norm": 0.8803273439407349, + "learning_rate": 0.000660789509444218, + "loss": 3.5907, + "step": 39940 + }, + { + "epoch": 2.714023644516918, + "grad_norm": 0.8253132104873657, + "learning_rate": 0.0006607470444353852, + "loss": 3.2503, + "step": 39945 + }, + { + "epoch": 2.7143633645875798, + "grad_norm": 1.1099398136138916, + "learning_rate": 0.0006607045794265524, + "loss": 3.4453, + "step": 39950 + }, + { + "epoch": 2.7147030846582414, + "grad_norm": 0.9525774717330933, + "learning_rate": 0.0006606621144177199, + "loss": 3.5244, + "step": 39955 + }, + { + "epoch": 2.7150428047289035, + "grad_norm": 0.7008033394813538, + "learning_rate": 0.0006606196494088871, + "loss": 3.7403, + "step": 39960 + }, + { + "epoch": 2.715382524799565, + "grad_norm": 0.8683610558509827, + "learning_rate": 0.0006605771844000543, + "loss": 3.5985, + "step": 39965 + }, + { + "epoch": 2.7157222448702267, + "grad_norm": 1.175252079963684, + "learning_rate": 0.0006605347193912217, + "loss": 3.2575, + "step": 39970 + }, + { + "epoch": 2.716061964940889, + "grad_norm": 0.888146698474884, + "learning_rate": 0.0006604922543823889, + "loss": 3.6604, + "step": 39975 + }, + { + "epoch": 2.7164016850115504, + "grad_norm": 0.9928892254829407, + "learning_rate": 0.0006604497893735561, + "loss": 3.4821, + "step": 39980 + }, + { + "epoch": 2.716741405082212, + "grad_norm": 0.8842623829841614, + "learning_rate": 0.0006604073243647235, + "loss": 3.4429, + "step": 39985 + }, + { + "epoch": 2.717081125152874, + "grad_norm": 0.7981084585189819, + "learning_rate": 0.0006603648593558908, + "loss": 3.5934, + "step": 39990 + }, + { + "epoch": 2.7174208452235358, + "grad_norm": 0.841220498085022, + "learning_rate": 0.000660322394347058, + "loss": 3.5876, + "step": 39995 + }, + { + "epoch": 2.7177605652941974, + "grad_norm": 0.7617934942245483, + "learning_rate": 0.0006602799293382254, + "loss": 3.6606, + "step": 40000 + }, + { + "epoch": 2.7181002853648595, + "grad_norm": 1.173885464668274, + "learning_rate": 0.0006602374643293926, + "loss": 3.5908, + "step": 40005 + }, + { + "epoch": 2.718440005435521, + "grad_norm": 0.8232343196868896, + "learning_rate": 0.0006601949993205598, + "loss": 3.4696, + "step": 40010 + }, + { + "epoch": 2.7187797255061827, + "grad_norm": 0.7001729011535645, + "learning_rate": 0.0006601525343117271, + "loss": 3.4657, + "step": 40015 + }, + { + "epoch": 2.719119445576845, + "grad_norm": 0.9046725630760193, + "learning_rate": 0.0006601100693028944, + "loss": 3.6355, + "step": 40020 + }, + { + "epoch": 2.7194591656475064, + "grad_norm": 0.6880229711532593, + "learning_rate": 0.0006600676042940617, + "loss": 3.8121, + "step": 40025 + }, + { + "epoch": 2.719798885718168, + "grad_norm": 1.1041297912597656, + "learning_rate": 0.000660025139285229, + "loss": 3.6441, + "step": 40030 + }, + { + "epoch": 2.72013860578883, + "grad_norm": 0.7319654226303101, + "learning_rate": 0.0006599826742763963, + "loss": 3.3554, + "step": 40035 + }, + { + "epoch": 2.720478325859492, + "grad_norm": 1.0054905414581299, + "learning_rate": 0.0006599402092675636, + "loss": 3.6667, + "step": 40040 + }, + { + "epoch": 2.7208180459301534, + "grad_norm": 0.8685222864151001, + "learning_rate": 0.0006598977442587308, + "loss": 3.4287, + "step": 40045 + }, + { + "epoch": 2.7211577660008155, + "grad_norm": 0.8062014579772949, + "learning_rate": 0.000659855279249898, + "loss": 3.4739, + "step": 40050 + }, + { + "epoch": 2.721497486071477, + "grad_norm": 0.9211556315422058, + "learning_rate": 0.0006598128142410654, + "loss": 3.58, + "step": 40055 + }, + { + "epoch": 2.7218372061421388, + "grad_norm": 1.118658185005188, + "learning_rate": 0.0006597703492322327, + "loss": 3.3311, + "step": 40060 + }, + { + "epoch": 2.722176926212801, + "grad_norm": 0.9092445969581604, + "learning_rate": 0.0006597278842233999, + "loss": 3.6422, + "step": 40065 + }, + { + "epoch": 2.7225166462834625, + "grad_norm": 0.8178219795227051, + "learning_rate": 0.0006596854192145673, + "loss": 3.6058, + "step": 40070 + }, + { + "epoch": 2.722856366354124, + "grad_norm": 0.9425751566886902, + "learning_rate": 0.0006596429542057345, + "loss": 3.677, + "step": 40075 + }, + { + "epoch": 2.723196086424786, + "grad_norm": 1.1031423807144165, + "learning_rate": 0.0006596004891969017, + "loss": 3.4428, + "step": 40080 + }, + { + "epoch": 2.723535806495448, + "grad_norm": 0.7799941897392273, + "learning_rate": 0.0006595580241880691, + "loss": 3.7331, + "step": 40085 + }, + { + "epoch": 2.7238755265661094, + "grad_norm": 0.9952499866485596, + "learning_rate": 0.0006595155591792363, + "loss": 3.6353, + "step": 40090 + }, + { + "epoch": 2.7242152466367715, + "grad_norm": 0.7231821417808533, + "learning_rate": 0.0006594730941704036, + "loss": 3.6645, + "step": 40095 + }, + { + "epoch": 2.724554966707433, + "grad_norm": 0.885179340839386, + "learning_rate": 0.000659430629161571, + "loss": 3.5133, + "step": 40100 + }, + { + "epoch": 2.7248946867780948, + "grad_norm": 0.8990301489830017, + "learning_rate": 0.0006593881641527382, + "loss": 3.4832, + "step": 40105 + }, + { + "epoch": 2.725234406848757, + "grad_norm": 2.263291835784912, + "learning_rate": 0.0006593456991439054, + "loss": 3.6553, + "step": 40110 + }, + { + "epoch": 2.7255741269194185, + "grad_norm": 0.8630979061126709, + "learning_rate": 0.0006593032341350727, + "loss": 3.634, + "step": 40115 + }, + { + "epoch": 2.72591384699008, + "grad_norm": 0.9513310790061951, + "learning_rate": 0.00065926076912624, + "loss": 3.4474, + "step": 40120 + }, + { + "epoch": 2.726253567060742, + "grad_norm": 0.8380131721496582, + "learning_rate": 0.0006592183041174072, + "loss": 3.3844, + "step": 40125 + }, + { + "epoch": 2.726593287131404, + "grad_norm": 0.7741430401802063, + "learning_rate": 0.0006591758391085746, + "loss": 3.7444, + "step": 40130 + }, + { + "epoch": 2.7269330072020654, + "grad_norm": 0.8665879964828491, + "learning_rate": 0.0006591333740997419, + "loss": 3.5391, + "step": 40135 + }, + { + "epoch": 2.7272727272727275, + "grad_norm": 0.920905590057373, + "learning_rate": 0.0006590909090909091, + "loss": 3.4289, + "step": 40140 + }, + { + "epoch": 2.727612447343389, + "grad_norm": 0.9568038582801819, + "learning_rate": 0.0006590484440820764, + "loss": 3.5123, + "step": 40145 + }, + { + "epoch": 2.7279521674140508, + "grad_norm": 1.0718151330947876, + "learning_rate": 0.0006590059790732436, + "loss": 3.6251, + "step": 40150 + }, + { + "epoch": 2.728291887484713, + "grad_norm": 0.8815311789512634, + "learning_rate": 0.0006589635140644109, + "loss": 3.5821, + "step": 40155 + }, + { + "epoch": 2.7286316075553745, + "grad_norm": 0.8185691833496094, + "learning_rate": 0.0006589210490555782, + "loss": 3.5243, + "step": 40160 + }, + { + "epoch": 2.728971327626036, + "grad_norm": 1.332939624786377, + "learning_rate": 0.0006588785840467455, + "loss": 3.7498, + "step": 40165 + }, + { + "epoch": 2.729311047696698, + "grad_norm": 1.276260256767273, + "learning_rate": 0.0006588361190379128, + "loss": 3.5576, + "step": 40170 + }, + { + "epoch": 2.72965076776736, + "grad_norm": 0.7643992900848389, + "learning_rate": 0.0006587936540290801, + "loss": 3.6814, + "step": 40175 + }, + { + "epoch": 2.7299904878380215, + "grad_norm": 1.1474686861038208, + "learning_rate": 0.0006587511890202473, + "loss": 3.7066, + "step": 40180 + }, + { + "epoch": 2.730330207908683, + "grad_norm": 0.9082356691360474, + "learning_rate": 0.0006587087240114146, + "loss": 3.6036, + "step": 40185 + }, + { + "epoch": 2.730669927979345, + "grad_norm": 0.751785397529602, + "learning_rate": 0.0006586662590025819, + "loss": 3.7263, + "step": 40190 + }, + { + "epoch": 2.731009648050007, + "grad_norm": 0.76153165102005, + "learning_rate": 0.0006586237939937491, + "loss": 3.5915, + "step": 40195 + }, + { + "epoch": 2.7313493681206684, + "grad_norm": 0.9065309166908264, + "learning_rate": 0.0006585813289849164, + "loss": 3.7779, + "step": 40200 + }, + { + "epoch": 2.7316890881913305, + "grad_norm": 1.0192867517471313, + "learning_rate": 0.0006585388639760838, + "loss": 3.385, + "step": 40205 + }, + { + "epoch": 2.732028808261992, + "grad_norm": 0.6835248470306396, + "learning_rate": 0.000658496398967251, + "loss": 3.6086, + "step": 40210 + }, + { + "epoch": 2.7323685283326538, + "grad_norm": 0.9173712134361267, + "learning_rate": 0.0006584539339584182, + "loss": 3.5931, + "step": 40215 + }, + { + "epoch": 2.732708248403316, + "grad_norm": 0.8721328377723694, + "learning_rate": 0.0006584114689495856, + "loss": 3.6052, + "step": 40220 + }, + { + "epoch": 2.7330479684739775, + "grad_norm": 0.9276031255722046, + "learning_rate": 0.0006583690039407528, + "loss": 3.5505, + "step": 40225 + }, + { + "epoch": 2.733387688544639, + "grad_norm": 1.031188726425171, + "learning_rate": 0.00065832653893192, + "loss": 3.4408, + "step": 40230 + }, + { + "epoch": 2.7337274086153007, + "grad_norm": 1.0801630020141602, + "learning_rate": 0.0006582840739230875, + "loss": 3.5817, + "step": 40235 + }, + { + "epoch": 2.734067128685963, + "grad_norm": 0.8701567053794861, + "learning_rate": 0.0006582416089142547, + "loss": 3.625, + "step": 40240 + }, + { + "epoch": 2.7344068487566244, + "grad_norm": 0.8715426921844482, + "learning_rate": 0.0006581991439054219, + "loss": 3.4749, + "step": 40245 + }, + { + "epoch": 2.734746568827286, + "grad_norm": 1.1303719282150269, + "learning_rate": 0.0006581566788965892, + "loss": 3.7009, + "step": 40250 + }, + { + "epoch": 2.735086288897948, + "grad_norm": 0.9289172887802124, + "learning_rate": 0.0006581142138877565, + "loss": 3.7178, + "step": 40255 + }, + { + "epoch": 2.7354260089686098, + "grad_norm": 0.6357284784317017, + "learning_rate": 0.0006580717488789237, + "loss": 3.5377, + "step": 40260 + }, + { + "epoch": 2.7357657290392714, + "grad_norm": 0.8401117920875549, + "learning_rate": 0.0006580292838700911, + "loss": 3.4947, + "step": 40265 + }, + { + "epoch": 2.7361054491099335, + "grad_norm": 0.7872085571289062, + "learning_rate": 0.0006579868188612584, + "loss": 3.5856, + "step": 40270 + }, + { + "epoch": 2.736445169180595, + "grad_norm": 0.9084161520004272, + "learning_rate": 0.0006579443538524256, + "loss": 3.2861, + "step": 40275 + }, + { + "epoch": 2.7367848892512567, + "grad_norm": 0.9328500628471375, + "learning_rate": 0.0006579018888435929, + "loss": 3.7289, + "step": 40280 + }, + { + "epoch": 2.737124609321919, + "grad_norm": 0.8494167327880859, + "learning_rate": 0.0006578594238347602, + "loss": 3.4918, + "step": 40285 + }, + { + "epoch": 2.7374643293925804, + "grad_norm": 0.7552316188812256, + "learning_rate": 0.0006578169588259274, + "loss": 3.6883, + "step": 40290 + }, + { + "epoch": 2.737804049463242, + "grad_norm": 0.6937495470046997, + "learning_rate": 0.0006577744938170947, + "loss": 3.3879, + "step": 40295 + }, + { + "epoch": 2.738143769533904, + "grad_norm": 0.8413161039352417, + "learning_rate": 0.000657732028808262, + "loss": 3.427, + "step": 40300 + }, + { + "epoch": 2.738483489604566, + "grad_norm": 0.8789914846420288, + "learning_rate": 0.0006576895637994293, + "loss": 3.4447, + "step": 40305 + }, + { + "epoch": 2.7388232096752274, + "grad_norm": 0.8155306577682495, + "learning_rate": 0.0006576470987905966, + "loss": 3.7293, + "step": 40310 + }, + { + "epoch": 2.7391629297458895, + "grad_norm": 0.9186259508132935, + "learning_rate": 0.0006576046337817638, + "loss": 3.5566, + "step": 40315 + }, + { + "epoch": 2.739502649816551, + "grad_norm": 0.9676461815834045, + "learning_rate": 0.0006575621687729311, + "loss": 3.2586, + "step": 40320 + }, + { + "epoch": 2.7398423698872127, + "grad_norm": 0.84682297706604, + "learning_rate": 0.0006575197037640984, + "loss": 3.6196, + "step": 40325 + }, + { + "epoch": 2.740182089957875, + "grad_norm": 0.8325427770614624, + "learning_rate": 0.0006574772387552656, + "loss": 3.6705, + "step": 40330 + }, + { + "epoch": 2.7405218100285365, + "grad_norm": 0.6998556852340698, + "learning_rate": 0.000657434773746433, + "loss": 3.6178, + "step": 40335 + }, + { + "epoch": 2.740861530099198, + "grad_norm": 0.8425354957580566, + "learning_rate": 0.0006573923087376003, + "loss": 3.4702, + "step": 40340 + }, + { + "epoch": 2.74120125016986, + "grad_norm": 0.9224645495414734, + "learning_rate": 0.0006573498437287675, + "loss": 3.6163, + "step": 40345 + }, + { + "epoch": 2.741540970240522, + "grad_norm": 1.0837113857269287, + "learning_rate": 0.0006573073787199347, + "loss": 3.0764, + "step": 40350 + }, + { + "epoch": 2.7418806903111834, + "grad_norm": 0.7782437801361084, + "learning_rate": 0.0006572649137111021, + "loss": 3.3457, + "step": 40355 + }, + { + "epoch": 2.7422204103818455, + "grad_norm": 0.8085513114929199, + "learning_rate": 0.0006572224487022693, + "loss": 3.638, + "step": 40360 + }, + { + "epoch": 2.742560130452507, + "grad_norm": 1.6547608375549316, + "learning_rate": 0.0006571799836934365, + "loss": 3.638, + "step": 40365 + }, + { + "epoch": 2.7428998505231688, + "grad_norm": 0.7194579839706421, + "learning_rate": 0.000657137518684604, + "loss": 3.6916, + "step": 40370 + }, + { + "epoch": 2.743239570593831, + "grad_norm": 1.110898733139038, + "learning_rate": 0.0006570950536757712, + "loss": 3.5896, + "step": 40375 + }, + { + "epoch": 2.7435792906644925, + "grad_norm": 0.8973047733306885, + "learning_rate": 0.0006570525886669385, + "loss": 3.7063, + "step": 40380 + }, + { + "epoch": 2.743919010735154, + "grad_norm": 1.019471526145935, + "learning_rate": 0.0006570101236581058, + "loss": 3.5558, + "step": 40385 + }, + { + "epoch": 2.744258730805816, + "grad_norm": 0.795246422290802, + "learning_rate": 0.000656967658649273, + "loss": 3.7662, + "step": 40390 + }, + { + "epoch": 2.744598450876478, + "grad_norm": 0.8133932948112488, + "learning_rate": 0.0006569251936404403, + "loss": 3.5625, + "step": 40395 + }, + { + "epoch": 2.7449381709471394, + "grad_norm": 0.8875547647476196, + "learning_rate": 0.0006568827286316075, + "loss": 3.2861, + "step": 40400 + }, + { + "epoch": 2.7452778910178015, + "grad_norm": 0.7816322445869446, + "learning_rate": 0.0006568402636227749, + "loss": 3.5642, + "step": 40405 + }, + { + "epoch": 2.745617611088463, + "grad_norm": 0.7662747502326965, + "learning_rate": 0.0006567977986139422, + "loss": 3.4831, + "step": 40410 + }, + { + "epoch": 2.7459573311591248, + "grad_norm": 0.7646679282188416, + "learning_rate": 0.0006567553336051094, + "loss": 3.5403, + "step": 40415 + }, + { + "epoch": 2.746297051229787, + "grad_norm": 0.6875748634338379, + "learning_rate": 0.0006567128685962767, + "loss": 3.6506, + "step": 40420 + }, + { + "epoch": 2.7466367713004485, + "grad_norm": 0.7331896424293518, + "learning_rate": 0.000656670403587444, + "loss": 3.6811, + "step": 40425 + }, + { + "epoch": 2.74697649137111, + "grad_norm": 0.7540739178657532, + "learning_rate": 0.0006566279385786112, + "loss": 3.3072, + "step": 40430 + }, + { + "epoch": 2.747316211441772, + "grad_norm": 0.788153350353241, + "learning_rate": 0.0006565854735697784, + "loss": 3.5628, + "step": 40435 + }, + { + "epoch": 2.747655931512434, + "grad_norm": 0.8513244390487671, + "learning_rate": 0.0006565430085609459, + "loss": 3.7153, + "step": 40440 + }, + { + "epoch": 2.7479956515830954, + "grad_norm": 0.8342524170875549, + "learning_rate": 0.0006565005435521131, + "loss": 3.6279, + "step": 40445 + }, + { + "epoch": 2.7483353716537575, + "grad_norm": 1.0505750179290771, + "learning_rate": 0.0006564580785432803, + "loss": 3.636, + "step": 40450 + }, + { + "epoch": 2.748675091724419, + "grad_norm": 0.9458935260772705, + "learning_rate": 0.0006564156135344477, + "loss": 3.5757, + "step": 40455 + }, + { + "epoch": 2.749014811795081, + "grad_norm": 1.0131213665008545, + "learning_rate": 0.0006563731485256149, + "loss": 3.7699, + "step": 40460 + }, + { + "epoch": 2.749354531865743, + "grad_norm": 1.017067790031433, + "learning_rate": 0.0006563306835167821, + "loss": 3.4812, + "step": 40465 + }, + { + "epoch": 2.7496942519364045, + "grad_norm": 0.8894461393356323, + "learning_rate": 0.0006562882185079495, + "loss": 3.2047, + "step": 40470 + }, + { + "epoch": 2.750033972007066, + "grad_norm": 0.8313674330711365, + "learning_rate": 0.0006562457534991168, + "loss": 3.6895, + "step": 40475 + }, + { + "epoch": 2.750373692077728, + "grad_norm": 0.9415770769119263, + "learning_rate": 0.000656203288490284, + "loss": 3.6684, + "step": 40480 + }, + { + "epoch": 2.75071341214839, + "grad_norm": 0.6898900270462036, + "learning_rate": 0.0006561608234814514, + "loss": 3.6185, + "step": 40485 + }, + { + "epoch": 2.7510531322190515, + "grad_norm": 0.7222219109535217, + "learning_rate": 0.0006561183584726186, + "loss": 3.5232, + "step": 40490 + }, + { + "epoch": 2.7513928522897135, + "grad_norm": 0.8290750980377197, + "learning_rate": 0.0006560758934637858, + "loss": 3.5704, + "step": 40495 + }, + { + "epoch": 2.751732572360375, + "grad_norm": 0.8512561321258545, + "learning_rate": 0.0006560334284549531, + "loss": 3.5888, + "step": 40500 + }, + { + "epoch": 2.752072292431037, + "grad_norm": 0.8354743123054504, + "learning_rate": 0.0006559909634461204, + "loss": 3.4185, + "step": 40505 + }, + { + "epoch": 2.752412012501699, + "grad_norm": 0.8362182974815369, + "learning_rate": 0.0006559484984372877, + "loss": 3.5621, + "step": 40510 + }, + { + "epoch": 2.7527517325723605, + "grad_norm": 0.8976656198501587, + "learning_rate": 0.000655906033428455, + "loss": 3.57, + "step": 40515 + }, + { + "epoch": 2.753091452643022, + "grad_norm": 0.8194878697395325, + "learning_rate": 0.0006558635684196223, + "loss": 3.4224, + "step": 40520 + }, + { + "epoch": 2.7534311727136838, + "grad_norm": 1.195380687713623, + "learning_rate": 0.0006558211034107895, + "loss": 3.5948, + "step": 40525 + }, + { + "epoch": 2.753770892784346, + "grad_norm": 0.8025371432304382, + "learning_rate": 0.0006557786384019568, + "loss": 3.6011, + "step": 40530 + }, + { + "epoch": 2.7541106128550075, + "grad_norm": 0.6247618198394775, + "learning_rate": 0.000655736173393124, + "loss": 3.7089, + "step": 40535 + }, + { + "epoch": 2.754450332925669, + "grad_norm": 1.003104567527771, + "learning_rate": 0.0006556937083842913, + "loss": 3.5178, + "step": 40540 + }, + { + "epoch": 2.754790052996331, + "grad_norm": 0.8687087893486023, + "learning_rate": 0.0006556512433754587, + "loss": 3.509, + "step": 40545 + }, + { + "epoch": 2.755129773066993, + "grad_norm": 0.958437442779541, + "learning_rate": 0.0006556087783666259, + "loss": 3.7245, + "step": 40550 + }, + { + "epoch": 2.7554694931376544, + "grad_norm": 1.0121073722839355, + "learning_rate": 0.0006555663133577932, + "loss": 3.5874, + "step": 40555 + }, + { + "epoch": 2.7558092132083165, + "grad_norm": 1.3760305643081665, + "learning_rate": 0.0006555238483489605, + "loss": 3.5112, + "step": 40560 + }, + { + "epoch": 2.756148933278978, + "grad_norm": 0.7471091151237488, + "learning_rate": 0.0006554813833401277, + "loss": 3.5137, + "step": 40565 + }, + { + "epoch": 2.7564886533496398, + "grad_norm": 0.6749697327613831, + "learning_rate": 0.000655438918331295, + "loss": 3.3283, + "step": 40570 + }, + { + "epoch": 2.7568283734203014, + "grad_norm": 0.9833710789680481, + "learning_rate": 0.0006553964533224623, + "loss": 3.5585, + "step": 40575 + }, + { + "epoch": 2.7571680934909635, + "grad_norm": 1.0973074436187744, + "learning_rate": 0.0006553539883136296, + "loss": 3.4875, + "step": 40580 + }, + { + "epoch": 2.757507813561625, + "grad_norm": 1.0187256336212158, + "learning_rate": 0.0006553115233047968, + "loss": 3.4893, + "step": 40585 + }, + { + "epoch": 2.7578475336322867, + "grad_norm": 0.9151908755302429, + "learning_rate": 0.0006552690582959642, + "loss": 3.4982, + "step": 40590 + }, + { + "epoch": 2.758187253702949, + "grad_norm": 0.9743619561195374, + "learning_rate": 0.0006552265932871314, + "loss": 3.5472, + "step": 40595 + }, + { + "epoch": 2.7585269737736104, + "grad_norm": 0.8172758221626282, + "learning_rate": 0.0006551841282782986, + "loss": 3.6355, + "step": 40600 + }, + { + "epoch": 2.758866693844272, + "grad_norm": 0.8158865571022034, + "learning_rate": 0.000655141663269466, + "loss": 3.6191, + "step": 40605 + }, + { + "epoch": 2.759206413914934, + "grad_norm": 0.9475234746932983, + "learning_rate": 0.0006550991982606332, + "loss": 3.5825, + "step": 40610 + }, + { + "epoch": 2.759546133985596, + "grad_norm": 0.9995924234390259, + "learning_rate": 0.0006550567332518005, + "loss": 3.1282, + "step": 40615 + }, + { + "epoch": 2.7598858540562574, + "grad_norm": 0.925312876701355, + "learning_rate": 0.0006550142682429679, + "loss": 3.7246, + "step": 40620 + }, + { + "epoch": 2.7602255741269195, + "grad_norm": 0.855171799659729, + "learning_rate": 0.0006549718032341351, + "loss": 3.4692, + "step": 40625 + }, + { + "epoch": 2.760565294197581, + "grad_norm": 1.0137652158737183, + "learning_rate": 0.0006549293382253023, + "loss": 3.3815, + "step": 40630 + }, + { + "epoch": 2.7609050142682428, + "grad_norm": 0.8715655207633972, + "learning_rate": 0.0006548868732164696, + "loss": 3.5159, + "step": 40635 + }, + { + "epoch": 2.761244734338905, + "grad_norm": 0.6796324253082275, + "learning_rate": 0.0006548444082076369, + "loss": 3.1392, + "step": 40640 + }, + { + "epoch": 2.7615844544095665, + "grad_norm": 0.8689772486686707, + "learning_rate": 0.0006548019431988041, + "loss": 3.7051, + "step": 40645 + }, + { + "epoch": 2.761924174480228, + "grad_norm": 0.9542728066444397, + "learning_rate": 0.0006547594781899715, + "loss": 3.4893, + "step": 40650 + }, + { + "epoch": 2.76226389455089, + "grad_norm": 0.7278113961219788, + "learning_rate": 0.0006547170131811388, + "loss": 3.5894, + "step": 40655 + }, + { + "epoch": 2.762603614621552, + "grad_norm": 0.8851398229598999, + "learning_rate": 0.000654674548172306, + "loss": 3.5513, + "step": 40660 + }, + { + "epoch": 2.7629433346922134, + "grad_norm": 0.7643380761146545, + "learning_rate": 0.0006546320831634733, + "loss": 3.5327, + "step": 40665 + }, + { + "epoch": 2.7632830547628755, + "grad_norm": 0.7789206504821777, + "learning_rate": 0.0006545896181546406, + "loss": 3.5304, + "step": 40670 + }, + { + "epoch": 2.763622774833537, + "grad_norm": 0.851741373538971, + "learning_rate": 0.0006545471531458078, + "loss": 3.3978, + "step": 40675 + }, + { + "epoch": 2.7639624949041988, + "grad_norm": 0.7207466959953308, + "learning_rate": 0.0006545046881369751, + "loss": 3.4144, + "step": 40680 + }, + { + "epoch": 2.764302214974861, + "grad_norm": 1.1106083393096924, + "learning_rate": 0.0006544622231281425, + "loss": 3.7285, + "step": 40685 + }, + { + "epoch": 2.7646419350455225, + "grad_norm": 0.7861536741256714, + "learning_rate": 0.0006544197581193097, + "loss": 3.6548, + "step": 40690 + }, + { + "epoch": 2.764981655116184, + "grad_norm": 0.853843629360199, + "learning_rate": 0.000654377293110477, + "loss": 3.5457, + "step": 40695 + }, + { + "epoch": 2.765321375186846, + "grad_norm": 1.037405252456665, + "learning_rate": 0.0006543348281016442, + "loss": 3.4414, + "step": 40700 + }, + { + "epoch": 2.765661095257508, + "grad_norm": 0.806270182132721, + "learning_rate": 0.0006542923630928115, + "loss": 3.6084, + "step": 40705 + }, + { + "epoch": 2.7660008153281694, + "grad_norm": 0.971283495426178, + "learning_rate": 0.0006542498980839788, + "loss": 3.5334, + "step": 40710 + }, + { + "epoch": 2.7663405353988315, + "grad_norm": 0.7653524875640869, + "learning_rate": 0.000654207433075146, + "loss": 3.5272, + "step": 40715 + }, + { + "epoch": 2.766680255469493, + "grad_norm": 1.3915772438049316, + "learning_rate": 0.0006541649680663135, + "loss": 3.4962, + "step": 40720 + }, + { + "epoch": 2.7670199755401548, + "grad_norm": 0.9580897092819214, + "learning_rate": 0.0006541225030574807, + "loss": 3.3549, + "step": 40725 + }, + { + "epoch": 2.767359695610817, + "grad_norm": 1.2455593347549438, + "learning_rate": 0.0006540800380486479, + "loss": 3.457, + "step": 40730 + }, + { + "epoch": 2.7676994156814785, + "grad_norm": 0.7226529121398926, + "learning_rate": 0.0006540375730398153, + "loss": 3.5481, + "step": 40735 + }, + { + "epoch": 2.76803913575214, + "grad_norm": 0.7970854043960571, + "learning_rate": 0.0006539951080309825, + "loss": 3.5547, + "step": 40740 + }, + { + "epoch": 2.768378855822802, + "grad_norm": 0.7825415134429932, + "learning_rate": 0.0006539526430221497, + "loss": 3.641, + "step": 40745 + }, + { + "epoch": 2.768718575893464, + "grad_norm": 0.9059216380119324, + "learning_rate": 0.0006539101780133171, + "loss": 3.6653, + "step": 40750 + }, + { + "epoch": 2.7690582959641254, + "grad_norm": 0.903237521648407, + "learning_rate": 0.0006538677130044844, + "loss": 3.5349, + "step": 40755 + }, + { + "epoch": 2.7693980160347875, + "grad_norm": 0.8359914422035217, + "learning_rate": 0.0006538252479956516, + "loss": 3.5534, + "step": 40760 + }, + { + "epoch": 2.769737736105449, + "grad_norm": 1.1874862909317017, + "learning_rate": 0.0006537827829868189, + "loss": 3.5867, + "step": 40765 + }, + { + "epoch": 2.770077456176111, + "grad_norm": 0.7669093012809753, + "learning_rate": 0.0006537403179779862, + "loss": 3.6542, + "step": 40770 + }, + { + "epoch": 2.770417176246773, + "grad_norm": 0.949467658996582, + "learning_rate": 0.0006536978529691534, + "loss": 3.7316, + "step": 40775 + }, + { + "epoch": 2.7707568963174345, + "grad_norm": 1.008885383605957, + "learning_rate": 0.0006536553879603207, + "loss": 3.5683, + "step": 40780 + }, + { + "epoch": 2.771096616388096, + "grad_norm": 0.8699371218681335, + "learning_rate": 0.000653612922951488, + "loss": 3.4793, + "step": 40785 + }, + { + "epoch": 2.771436336458758, + "grad_norm": 0.7293240427970886, + "learning_rate": 0.0006535704579426553, + "loss": 3.7199, + "step": 40790 + }, + { + "epoch": 2.77177605652942, + "grad_norm": 0.9556752443313599, + "learning_rate": 0.0006535279929338226, + "loss": 3.7293, + "step": 40795 + }, + { + "epoch": 2.7721157766000815, + "grad_norm": 0.859874427318573, + "learning_rate": 0.0006534855279249898, + "loss": 3.4217, + "step": 40800 + }, + { + "epoch": 2.7724554966707435, + "grad_norm": 0.877276599407196, + "learning_rate": 0.0006534430629161571, + "loss": 3.4427, + "step": 40805 + }, + { + "epoch": 2.772795216741405, + "grad_norm": 0.8203192353248596, + "learning_rate": 0.0006534005979073244, + "loss": 3.3766, + "step": 40810 + }, + { + "epoch": 2.773134936812067, + "grad_norm": 0.9335929751396179, + "learning_rate": 0.0006533581328984916, + "loss": 3.6727, + "step": 40815 + }, + { + "epoch": 2.773474656882729, + "grad_norm": 0.7954270243644714, + "learning_rate": 0.000653315667889659, + "loss": 3.6444, + "step": 40820 + }, + { + "epoch": 2.7738143769533905, + "grad_norm": 0.9236946105957031, + "learning_rate": 0.0006532732028808263, + "loss": 3.39, + "step": 40825 + }, + { + "epoch": 2.774154097024052, + "grad_norm": 1.0424996614456177, + "learning_rate": 0.0006532307378719935, + "loss": 3.4322, + "step": 40830 + }, + { + "epoch": 2.774493817094714, + "grad_norm": 0.8398603796958923, + "learning_rate": 0.0006531882728631607, + "loss": 3.6222, + "step": 40835 + }, + { + "epoch": 2.774833537165376, + "grad_norm": 0.7191680669784546, + "learning_rate": 0.0006531458078543281, + "loss": 3.7335, + "step": 40840 + }, + { + "epoch": 2.7751732572360375, + "grad_norm": 0.8853099942207336, + "learning_rate": 0.0006531033428454953, + "loss": 3.8024, + "step": 40845 + }, + { + "epoch": 2.7755129773066995, + "grad_norm": 1.0183979272842407, + "learning_rate": 0.0006530608778366625, + "loss": 3.3684, + "step": 40850 + }, + { + "epoch": 2.775852697377361, + "grad_norm": 1.303118348121643, + "learning_rate": 0.00065301841282783, + "loss": 3.4345, + "step": 40855 + }, + { + "epoch": 2.776192417448023, + "grad_norm": 0.9715933799743652, + "learning_rate": 0.0006529759478189972, + "loss": 3.4978, + "step": 40860 + }, + { + "epoch": 2.7765321375186844, + "grad_norm": 0.9406045079231262, + "learning_rate": 0.0006529334828101644, + "loss": 3.7486, + "step": 40865 + }, + { + "epoch": 2.7768718575893465, + "grad_norm": 0.7973609566688538, + "learning_rate": 0.0006528910178013318, + "loss": 3.544, + "step": 40870 + }, + { + "epoch": 2.777211577660008, + "grad_norm": 1.0448392629623413, + "learning_rate": 0.000652848552792499, + "loss": 3.5673, + "step": 40875 + }, + { + "epoch": 2.7775512977306698, + "grad_norm": 0.9311828017234802, + "learning_rate": 0.0006528060877836662, + "loss": 3.6106, + "step": 40880 + }, + { + "epoch": 2.777891017801332, + "grad_norm": 2.4790256023406982, + "learning_rate": 0.0006527636227748335, + "loss": 3.301, + "step": 40885 + }, + { + "epoch": 2.7782307378719935, + "grad_norm": 1.1215834617614746, + "learning_rate": 0.0006527211577660009, + "loss": 3.6576, + "step": 40890 + }, + { + "epoch": 2.778570457942655, + "grad_norm": 0.7589175701141357, + "learning_rate": 0.0006526786927571681, + "loss": 3.7616, + "step": 40895 + }, + { + "epoch": 2.778910178013317, + "grad_norm": 0.8938860893249512, + "learning_rate": 0.0006526362277483354, + "loss": 3.1507, + "step": 40900 + }, + { + "epoch": 2.779249898083979, + "grad_norm": 0.9298586845397949, + "learning_rate": 0.0006525937627395027, + "loss": 3.7021, + "step": 40905 + }, + { + "epoch": 2.7795896181546405, + "grad_norm": 0.7274421453475952, + "learning_rate": 0.0006525512977306699, + "loss": 3.5557, + "step": 40910 + }, + { + "epoch": 2.779929338225302, + "grad_norm": 1.043718934059143, + "learning_rate": 0.0006525088327218372, + "loss": 3.6564, + "step": 40915 + }, + { + "epoch": 2.780269058295964, + "grad_norm": 1.0683753490447998, + "learning_rate": 0.0006524663677130045, + "loss": 3.1886, + "step": 40920 + }, + { + "epoch": 2.780608778366626, + "grad_norm": 0.7982569336891174, + "learning_rate": 0.0006524239027041718, + "loss": 3.7112, + "step": 40925 + }, + { + "epoch": 2.7809484984372874, + "grad_norm": 0.7379443645477295, + "learning_rate": 0.0006523814376953391, + "loss": 3.6198, + "step": 40930 + }, + { + "epoch": 2.7812882185079495, + "grad_norm": 0.7897406816482544, + "learning_rate": 0.0006523389726865063, + "loss": 3.6165, + "step": 40935 + }, + { + "epoch": 2.781627938578611, + "grad_norm": 0.6179049015045166, + "learning_rate": 0.0006522965076776736, + "loss": 3.6887, + "step": 40940 + }, + { + "epoch": 2.7819676586492728, + "grad_norm": 0.8219224214553833, + "learning_rate": 0.0006522540426688409, + "loss": 3.3352, + "step": 40945 + }, + { + "epoch": 2.782307378719935, + "grad_norm": 1.139896273612976, + "learning_rate": 0.0006522115776600081, + "loss": 3.6948, + "step": 40950 + }, + { + "epoch": 2.7826470987905965, + "grad_norm": 0.8027321696281433, + "learning_rate": 0.0006521691126511754, + "loss": 3.5342, + "step": 40955 + }, + { + "epoch": 2.782986818861258, + "grad_norm": 0.7882165908813477, + "learning_rate": 0.0006521266476423428, + "loss": 3.6057, + "step": 40960 + }, + { + "epoch": 2.78332653893192, + "grad_norm": 0.8381488919258118, + "learning_rate": 0.00065208418263351, + "loss": 3.5081, + "step": 40965 + }, + { + "epoch": 2.783666259002582, + "grad_norm": 0.6658279895782471, + "learning_rate": 0.0006520417176246773, + "loss": 3.4365, + "step": 40970 + }, + { + "epoch": 2.7840059790732434, + "grad_norm": 0.7792349457740784, + "learning_rate": 0.0006519992526158446, + "loss": 3.6053, + "step": 40975 + }, + { + "epoch": 2.7843456991439055, + "grad_norm": 0.850094199180603, + "learning_rate": 0.0006519567876070118, + "loss": 3.322, + "step": 40980 + }, + { + "epoch": 2.784685419214567, + "grad_norm": 0.8450062274932861, + "learning_rate": 0.000651914322598179, + "loss": 3.627, + "step": 40985 + }, + { + "epoch": 2.7850251392852288, + "grad_norm": 0.9091132879257202, + "learning_rate": 0.0006518718575893464, + "loss": 3.4873, + "step": 40990 + }, + { + "epoch": 2.785364859355891, + "grad_norm": 0.9887285232543945, + "learning_rate": 0.0006518293925805137, + "loss": 3.7454, + "step": 40995 + }, + { + "epoch": 2.7857045794265525, + "grad_norm": 1.0718141794204712, + "learning_rate": 0.0006517869275716809, + "loss": 3.3801, + "step": 41000 + }, + { + "epoch": 2.786044299497214, + "grad_norm": 0.6924198269844055, + "learning_rate": 0.0006517444625628483, + "loss": 3.8928, + "step": 41005 + }, + { + "epoch": 2.786384019567876, + "grad_norm": 0.8262925744056702, + "learning_rate": 0.0006517019975540155, + "loss": 3.7331, + "step": 41010 + }, + { + "epoch": 2.786723739638538, + "grad_norm": 0.7325571179389954, + "learning_rate": 0.0006516595325451827, + "loss": 3.7014, + "step": 41015 + }, + { + "epoch": 2.7870634597091994, + "grad_norm": 0.8553724884986877, + "learning_rate": 0.00065161706753635, + "loss": 3.4751, + "step": 41020 + }, + { + "epoch": 2.7874031797798615, + "grad_norm": 0.8223779201507568, + "learning_rate": 0.0006515746025275173, + "loss": 3.7135, + "step": 41025 + }, + { + "epoch": 2.787742899850523, + "grad_norm": 1.0360852479934692, + "learning_rate": 0.0006515321375186846, + "loss": 3.2543, + "step": 41030 + }, + { + "epoch": 2.788082619921185, + "grad_norm": 0.8779467344284058, + "learning_rate": 0.000651489672509852, + "loss": 3.7755, + "step": 41035 + }, + { + "epoch": 2.788422339991847, + "grad_norm": 0.8353963494300842, + "learning_rate": 0.0006514472075010192, + "loss": 3.5913, + "step": 41040 + }, + { + "epoch": 2.7887620600625085, + "grad_norm": 0.7837998270988464, + "learning_rate": 0.0006514047424921864, + "loss": 3.1349, + "step": 41045 + }, + { + "epoch": 2.78910178013317, + "grad_norm": 0.9109930396080017, + "learning_rate": 0.0006513622774833537, + "loss": 3.2446, + "step": 41050 + }, + { + "epoch": 2.789441500203832, + "grad_norm": 0.640507698059082, + "learning_rate": 0.000651319812474521, + "loss": 3.7637, + "step": 41055 + }, + { + "epoch": 2.789781220274494, + "grad_norm": 0.807341456413269, + "learning_rate": 0.0006512773474656883, + "loss": 3.7173, + "step": 41060 + }, + { + "epoch": 2.7901209403451555, + "grad_norm": 0.7839886546134949, + "learning_rate": 0.0006512348824568556, + "loss": 3.5668, + "step": 41065 + }, + { + "epoch": 2.7904606604158175, + "grad_norm": 0.7389859557151794, + "learning_rate": 0.0006511924174480229, + "loss": 3.5327, + "step": 41070 + }, + { + "epoch": 2.790800380486479, + "grad_norm": 0.8813762068748474, + "learning_rate": 0.0006511499524391902, + "loss": 3.3964, + "step": 41075 + }, + { + "epoch": 2.791140100557141, + "grad_norm": 0.8989459276199341, + "learning_rate": 0.0006511074874303574, + "loss": 3.547, + "step": 41080 + }, + { + "epoch": 2.791479820627803, + "grad_norm": 1.0809416770935059, + "learning_rate": 0.0006510650224215246, + "loss": 3.5185, + "step": 41085 + }, + { + "epoch": 2.7918195406984645, + "grad_norm": 0.739500880241394, + "learning_rate": 0.000651022557412692, + "loss": 3.5512, + "step": 41090 + }, + { + "epoch": 2.792159260769126, + "grad_norm": 0.8862757682800293, + "learning_rate": 0.0006509800924038592, + "loss": 3.6412, + "step": 41095 + }, + { + "epoch": 2.792498980839788, + "grad_norm": 0.922012448310852, + "learning_rate": 0.0006509376273950265, + "loss": 3.4056, + "step": 41100 + }, + { + "epoch": 2.79283870091045, + "grad_norm": 0.7956035733222961, + "learning_rate": 0.0006508951623861939, + "loss": 3.6471, + "step": 41105 + }, + { + "epoch": 2.7931784209811115, + "grad_norm": 0.749803900718689, + "learning_rate": 0.0006508526973773611, + "loss": 3.6001, + "step": 41110 + }, + { + "epoch": 2.7935181410517735, + "grad_norm": 0.8472878336906433, + "learning_rate": 0.0006508102323685283, + "loss": 3.7048, + "step": 41115 + }, + { + "epoch": 2.793857861122435, + "grad_norm": 0.8235501050949097, + "learning_rate": 0.0006507677673596957, + "loss": 3.4751, + "step": 41120 + }, + { + "epoch": 2.794197581193097, + "grad_norm": 0.7599331140518188, + "learning_rate": 0.0006507253023508629, + "loss": 3.6153, + "step": 41125 + }, + { + "epoch": 2.794537301263759, + "grad_norm": 0.8559428453445435, + "learning_rate": 0.0006506828373420301, + "loss": 3.2042, + "step": 41130 + }, + { + "epoch": 2.7948770213344205, + "grad_norm": 0.7228989601135254, + "learning_rate": 0.0006506403723331975, + "loss": 3.4054, + "step": 41135 + }, + { + "epoch": 2.795216741405082, + "grad_norm": 0.8533180356025696, + "learning_rate": 0.0006505979073243648, + "loss": 3.4009, + "step": 41140 + }, + { + "epoch": 2.795556461475744, + "grad_norm": 1.193645715713501, + "learning_rate": 0.000650555442315532, + "loss": 3.3261, + "step": 41145 + }, + { + "epoch": 2.795896181546406, + "grad_norm": 0.8603028655052185, + "learning_rate": 0.0006505129773066993, + "loss": 3.4895, + "step": 41150 + }, + { + "epoch": 2.7962359016170675, + "grad_norm": 0.8908913135528564, + "learning_rate": 0.0006504705122978666, + "loss": 3.4189, + "step": 41155 + }, + { + "epoch": 2.7965756216877296, + "grad_norm": 0.9131405353546143, + "learning_rate": 0.0006504280472890338, + "loss": 3.4005, + "step": 41160 + }, + { + "epoch": 2.796915341758391, + "grad_norm": 1.7250792980194092, + "learning_rate": 0.0006503855822802011, + "loss": 3.5472, + "step": 41165 + }, + { + "epoch": 2.797255061829053, + "grad_norm": 0.7639291882514954, + "learning_rate": 0.0006503431172713685, + "loss": 3.5514, + "step": 41170 + }, + { + "epoch": 2.797594781899715, + "grad_norm": 1.2182351350784302, + "learning_rate": 0.0006503006522625357, + "loss": 3.4675, + "step": 41175 + }, + { + "epoch": 2.7979345019703765, + "grad_norm": 0.8525551557540894, + "learning_rate": 0.000650258187253703, + "loss": 3.6614, + "step": 41180 + }, + { + "epoch": 2.798274222041038, + "grad_norm": 0.849395751953125, + "learning_rate": 0.0006502157222448702, + "loss": 3.425, + "step": 41185 + }, + { + "epoch": 2.7986139421117002, + "grad_norm": 0.7365756034851074, + "learning_rate": 0.0006501732572360375, + "loss": 3.637, + "step": 41190 + }, + { + "epoch": 2.798953662182362, + "grad_norm": 0.7442958950996399, + "learning_rate": 0.0006501307922272048, + "loss": 3.8011, + "step": 41195 + }, + { + "epoch": 2.7992933822530235, + "grad_norm": 0.8995551466941833, + "learning_rate": 0.000650088327218372, + "loss": 3.5012, + "step": 41200 + }, + { + "epoch": 2.799633102323685, + "grad_norm": 0.7885043621063232, + "learning_rate": 0.0006500458622095394, + "loss": 3.5857, + "step": 41205 + }, + { + "epoch": 2.799972822394347, + "grad_norm": 0.8332861661911011, + "learning_rate": 0.0006500033972007067, + "loss": 3.6609, + "step": 41210 + }, + { + "epoch": 2.800312542465009, + "grad_norm": 0.7752822041511536, + "learning_rate": 0.0006499609321918739, + "loss": 3.384, + "step": 41215 + }, + { + "epoch": 2.8006522625356705, + "grad_norm": 0.9063897132873535, + "learning_rate": 0.0006499184671830411, + "loss": 3.3559, + "step": 41220 + }, + { + "epoch": 2.8009919826063325, + "grad_norm": 0.7396212816238403, + "learning_rate": 0.0006498760021742085, + "loss": 3.5052, + "step": 41225 + }, + { + "epoch": 2.801331702676994, + "grad_norm": 0.9083309769630432, + "learning_rate": 0.0006498335371653757, + "loss": 3.5104, + "step": 41230 + }, + { + "epoch": 2.801671422747656, + "grad_norm": 0.8172621726989746, + "learning_rate": 0.0006497910721565429, + "loss": 3.4756, + "step": 41235 + }, + { + "epoch": 2.802011142818318, + "grad_norm": 1.024101972579956, + "learning_rate": 0.0006497486071477104, + "loss": 3.5534, + "step": 41240 + }, + { + "epoch": 2.8023508628889795, + "grad_norm": 0.8846065402030945, + "learning_rate": 0.0006497061421388776, + "loss": 3.3963, + "step": 41245 + }, + { + "epoch": 2.802690582959641, + "grad_norm": 0.8324682116508484, + "learning_rate": 0.0006496636771300448, + "loss": 3.4444, + "step": 41250 + }, + { + "epoch": 2.8030303030303028, + "grad_norm": 0.888038694858551, + "learning_rate": 0.0006496212121212122, + "loss": 3.5378, + "step": 41255 + }, + { + "epoch": 2.803370023100965, + "grad_norm": 1.1987608671188354, + "learning_rate": 0.0006495787471123794, + "loss": 3.9997, + "step": 41260 + }, + { + "epoch": 2.8037097431716265, + "grad_norm": 0.7669888138771057, + "learning_rate": 0.0006495362821035466, + "loss": 3.509, + "step": 41265 + }, + { + "epoch": 2.804049463242288, + "grad_norm": 0.9518877863883972, + "learning_rate": 0.000649493817094714, + "loss": 3.3775, + "step": 41270 + }, + { + "epoch": 2.80438918331295, + "grad_norm": 2.1852855682373047, + "learning_rate": 0.0006494513520858813, + "loss": 3.5079, + "step": 41275 + }, + { + "epoch": 2.804728903383612, + "grad_norm": 0.8351412415504456, + "learning_rate": 0.0006494088870770485, + "loss": 3.6044, + "step": 41280 + }, + { + "epoch": 2.8050686234542734, + "grad_norm": 0.8314439058303833, + "learning_rate": 0.0006493664220682158, + "loss": 3.4677, + "step": 41285 + }, + { + "epoch": 2.8054083435249355, + "grad_norm": 0.9713733792304993, + "learning_rate": 0.0006493239570593831, + "loss": 3.6681, + "step": 41290 + }, + { + "epoch": 2.805748063595597, + "grad_norm": 1.1723856925964355, + "learning_rate": 0.0006492814920505503, + "loss": 3.1873, + "step": 41295 + }, + { + "epoch": 2.8060877836662588, + "grad_norm": 0.8414456844329834, + "learning_rate": 0.0006492390270417176, + "loss": 3.6145, + "step": 41300 + }, + { + "epoch": 2.806427503736921, + "grad_norm": 0.7658439874649048, + "learning_rate": 0.0006491965620328849, + "loss": 3.6697, + "step": 41305 + }, + { + "epoch": 2.8067672238075825, + "grad_norm": 0.8715805411338806, + "learning_rate": 0.0006491540970240522, + "loss": 3.4449, + "step": 41310 + }, + { + "epoch": 2.807106943878244, + "grad_norm": 0.8142425417900085, + "learning_rate": 0.0006491116320152195, + "loss": 3.4652, + "step": 41315 + }, + { + "epoch": 2.807446663948906, + "grad_norm": 0.9371588230133057, + "learning_rate": 0.0006490691670063867, + "loss": 3.4198, + "step": 41320 + }, + { + "epoch": 2.807786384019568, + "grad_norm": 0.7339078187942505, + "learning_rate": 0.000649026701997554, + "loss": 3.6215, + "step": 41325 + }, + { + "epoch": 2.8081261040902294, + "grad_norm": 0.8386957049369812, + "learning_rate": 0.0006489842369887213, + "loss": 3.4108, + "step": 41330 + }, + { + "epoch": 2.8084658241608915, + "grad_norm": 1.1156625747680664, + "learning_rate": 0.0006489417719798885, + "loss": 3.5202, + "step": 41335 + }, + { + "epoch": 2.808805544231553, + "grad_norm": 0.8288564682006836, + "learning_rate": 0.0006488993069710558, + "loss": 3.5625, + "step": 41340 + }, + { + "epoch": 2.809145264302215, + "grad_norm": 0.7618797421455383, + "learning_rate": 0.0006488568419622232, + "loss": 3.6131, + "step": 41345 + }, + { + "epoch": 2.809484984372877, + "grad_norm": 1.0272284746170044, + "learning_rate": 0.0006488143769533904, + "loss": 3.3034, + "step": 41350 + }, + { + "epoch": 2.8098247044435385, + "grad_norm": 0.7986567616462708, + "learning_rate": 0.0006487719119445577, + "loss": 3.6474, + "step": 41355 + }, + { + "epoch": 2.8101644245142, + "grad_norm": 0.9640021920204163, + "learning_rate": 0.000648729446935725, + "loss": 3.4959, + "step": 41360 + }, + { + "epoch": 2.810504144584862, + "grad_norm": 1.209046721458435, + "learning_rate": 0.0006486869819268922, + "loss": 3.512, + "step": 41365 + }, + { + "epoch": 2.810843864655524, + "grad_norm": 0.9461942315101624, + "learning_rate": 0.0006486445169180594, + "loss": 3.403, + "step": 41370 + }, + { + "epoch": 2.8111835847261855, + "grad_norm": 0.9588611125946045, + "learning_rate": 0.0006486020519092269, + "loss": 3.495, + "step": 41375 + }, + { + "epoch": 2.8115233047968475, + "grad_norm": 0.7493300437927246, + "learning_rate": 0.0006485595869003941, + "loss": 3.6506, + "step": 41380 + }, + { + "epoch": 2.811863024867509, + "grad_norm": 0.8535450100898743, + "learning_rate": 0.0006485171218915613, + "loss": 3.6834, + "step": 41385 + }, + { + "epoch": 2.812202744938171, + "grad_norm": 0.8242064118385315, + "learning_rate": 0.0006484746568827287, + "loss": 3.7053, + "step": 41390 + }, + { + "epoch": 2.812542465008833, + "grad_norm": 0.8100264668464661, + "learning_rate": 0.0006484321918738959, + "loss": 3.6562, + "step": 41395 + }, + { + "epoch": 2.8128821850794945, + "grad_norm": 0.859417200088501, + "learning_rate": 0.0006483897268650632, + "loss": 3.6942, + "step": 41400 + }, + { + "epoch": 2.813221905150156, + "grad_norm": 0.8369675278663635, + "learning_rate": 0.0006483472618562305, + "loss": 3.4838, + "step": 41405 + }, + { + "epoch": 2.813561625220818, + "grad_norm": 0.7112186551094055, + "learning_rate": 0.0006483047968473978, + "loss": 3.6894, + "step": 41410 + }, + { + "epoch": 2.81390134529148, + "grad_norm": 0.9024016261100769, + "learning_rate": 0.0006482623318385651, + "loss": 3.6715, + "step": 41415 + }, + { + "epoch": 2.8142410653621415, + "grad_norm": 1.2939252853393555, + "learning_rate": 0.0006482198668297324, + "loss": 3.8274, + "step": 41420 + }, + { + "epoch": 2.8145807854328035, + "grad_norm": 1.1902871131896973, + "learning_rate": 0.0006481774018208996, + "loss": 3.4898, + "step": 41425 + }, + { + "epoch": 2.814920505503465, + "grad_norm": 1.1284643411636353, + "learning_rate": 0.0006481349368120669, + "loss": 3.4472, + "step": 41430 + }, + { + "epoch": 2.815260225574127, + "grad_norm": 0.8643841743469238, + "learning_rate": 0.0006480924718032341, + "loss": 3.4466, + "step": 41435 + }, + { + "epoch": 2.815599945644789, + "grad_norm": 0.9510395526885986, + "learning_rate": 0.0006480500067944014, + "loss": 3.5191, + "step": 41440 + }, + { + "epoch": 2.8159396657154505, + "grad_norm": 0.954417884349823, + "learning_rate": 0.0006480075417855688, + "loss": 3.6143, + "step": 41445 + }, + { + "epoch": 2.816279385786112, + "grad_norm": 0.8017139434814453, + "learning_rate": 0.000647965076776736, + "loss": 3.8417, + "step": 41450 + }, + { + "epoch": 2.816619105856774, + "grad_norm": 0.8395540714263916, + "learning_rate": 0.0006479226117679033, + "loss": 3.5647, + "step": 41455 + }, + { + "epoch": 2.816958825927436, + "grad_norm": 0.9345386028289795, + "learning_rate": 0.0006478801467590706, + "loss": 3.7737, + "step": 41460 + }, + { + "epoch": 2.8172985459980975, + "grad_norm": 0.8753471374511719, + "learning_rate": 0.0006478376817502378, + "loss": 3.4784, + "step": 41465 + }, + { + "epoch": 2.8176382660687596, + "grad_norm": 0.722888171672821, + "learning_rate": 0.000647795216741405, + "loss": 3.5069, + "step": 41470 + }, + { + "epoch": 2.817977986139421, + "grad_norm": 0.6951571106910706, + "learning_rate": 0.0006477527517325724, + "loss": 3.7842, + "step": 41475 + }, + { + "epoch": 2.818317706210083, + "grad_norm": 0.6592711806297302, + "learning_rate": 0.0006477102867237397, + "loss": 3.604, + "step": 41480 + }, + { + "epoch": 2.818657426280745, + "grad_norm": 0.8487776517868042, + "learning_rate": 0.0006476678217149069, + "loss": 3.7043, + "step": 41485 + }, + { + "epoch": 2.8189971463514065, + "grad_norm": 0.9052252173423767, + "learning_rate": 0.0006476253567060743, + "loss": 3.5509, + "step": 41490 + }, + { + "epoch": 2.819336866422068, + "grad_norm": 0.835721492767334, + "learning_rate": 0.0006475828916972415, + "loss": 3.6987, + "step": 41495 + }, + { + "epoch": 2.8196765864927302, + "grad_norm": 1.0919418334960938, + "learning_rate": 0.0006475404266884087, + "loss": 3.4681, + "step": 41500 + }, + { + "epoch": 2.820016306563392, + "grad_norm": 0.8099721074104309, + "learning_rate": 0.0006474979616795761, + "loss": 3.65, + "step": 41505 + }, + { + "epoch": 2.8203560266340535, + "grad_norm": 0.8060786724090576, + "learning_rate": 0.0006474554966707433, + "loss": 3.6157, + "step": 41510 + }, + { + "epoch": 2.8206957467047156, + "grad_norm": 0.8671443462371826, + "learning_rate": 0.0006474130316619106, + "loss": 3.5806, + "step": 41515 + }, + { + "epoch": 2.821035466775377, + "grad_norm": 0.8679184317588806, + "learning_rate": 0.000647370566653078, + "loss": 3.6527, + "step": 41520 + }, + { + "epoch": 2.821375186846039, + "grad_norm": 0.7229198217391968, + "learning_rate": 0.0006473281016442452, + "loss": 3.5314, + "step": 41525 + }, + { + "epoch": 2.821714906916701, + "grad_norm": 0.6718996167182922, + "learning_rate": 0.0006472856366354124, + "loss": 3.6724, + "step": 41530 + }, + { + "epoch": 2.8220546269873625, + "grad_norm": 0.641571581363678, + "learning_rate": 0.0006472431716265797, + "loss": 3.697, + "step": 41535 + }, + { + "epoch": 2.822394347058024, + "grad_norm": 0.9195499420166016, + "learning_rate": 0.000647200706617747, + "loss": 3.489, + "step": 41540 + }, + { + "epoch": 2.822734067128686, + "grad_norm": 0.6031072735786438, + "learning_rate": 0.0006471582416089142, + "loss": 3.3711, + "step": 41545 + }, + { + "epoch": 2.823073787199348, + "grad_norm": 0.7756946086883545, + "learning_rate": 0.0006471157766000816, + "loss": 3.5265, + "step": 41550 + }, + { + "epoch": 2.8234135072700095, + "grad_norm": 0.9457759857177734, + "learning_rate": 0.0006470733115912489, + "loss": 3.6576, + "step": 41555 + }, + { + "epoch": 2.823753227340671, + "grad_norm": 0.7569836378097534, + "learning_rate": 0.0006470308465824161, + "loss": 3.6099, + "step": 41560 + }, + { + "epoch": 2.824092947411333, + "grad_norm": 0.9484497308731079, + "learning_rate": 0.0006469883815735834, + "loss": 3.9439, + "step": 41565 + }, + { + "epoch": 2.824432667481995, + "grad_norm": 0.9010297060012817, + "learning_rate": 0.0006469459165647506, + "loss": 3.4099, + "step": 41570 + }, + { + "epoch": 2.8247723875526565, + "grad_norm": 1.157332181930542, + "learning_rate": 0.0006469034515559179, + "loss": 3.58, + "step": 41575 + }, + { + "epoch": 2.8251121076233185, + "grad_norm": 0.6770252585411072, + "learning_rate": 0.0006468609865470852, + "loss": 3.5915, + "step": 41580 + }, + { + "epoch": 2.82545182769398, + "grad_norm": 1.2056981325149536, + "learning_rate": 0.0006468185215382525, + "loss": 3.4417, + "step": 41585 + }, + { + "epoch": 2.825791547764642, + "grad_norm": 1.1649360656738281, + "learning_rate": 0.0006467760565294198, + "loss": 3.7079, + "step": 41590 + }, + { + "epoch": 2.8261312678353034, + "grad_norm": 0.8605906963348389, + "learning_rate": 0.0006467335915205871, + "loss": 3.4819, + "step": 41595 + }, + { + "epoch": 2.8264709879059655, + "grad_norm": 0.8827203512191772, + "learning_rate": 0.0006466911265117543, + "loss": 3.7741, + "step": 41600 + }, + { + "epoch": 2.826810707976627, + "grad_norm": 0.7515257596969604, + "learning_rate": 0.0006466486615029216, + "loss": 3.6435, + "step": 41605 + }, + { + "epoch": 2.8271504280472888, + "grad_norm": 0.7814648151397705, + "learning_rate": 0.0006466061964940889, + "loss": 3.5114, + "step": 41610 + }, + { + "epoch": 2.827490148117951, + "grad_norm": 0.7758761644363403, + "learning_rate": 0.0006465637314852561, + "loss": 3.618, + "step": 41615 + }, + { + "epoch": 2.8278298681886125, + "grad_norm": 0.7390299439430237, + "learning_rate": 0.0006465212664764234, + "loss": 3.5086, + "step": 41620 + }, + { + "epoch": 2.828169588259274, + "grad_norm": 0.9217628836631775, + "learning_rate": 0.0006464788014675908, + "loss": 3.3841, + "step": 41625 + }, + { + "epoch": 2.828509308329936, + "grad_norm": 0.7836593389511108, + "learning_rate": 0.000646436336458758, + "loss": 3.7769, + "step": 41630 + }, + { + "epoch": 2.828849028400598, + "grad_norm": 0.7080562710762024, + "learning_rate": 0.0006463938714499252, + "loss": 3.4178, + "step": 41635 + }, + { + "epoch": 2.8291887484712595, + "grad_norm": 0.7039355039596558, + "learning_rate": 0.0006463514064410926, + "loss": 3.563, + "step": 41640 + }, + { + "epoch": 2.8295284685419215, + "grad_norm": 0.8547645211219788, + "learning_rate": 0.0006463089414322598, + "loss": 3.6418, + "step": 41645 + }, + { + "epoch": 2.829868188612583, + "grad_norm": 0.9048673510551453, + "learning_rate": 0.000646266476423427, + "loss": 3.676, + "step": 41650 + }, + { + "epoch": 2.830207908683245, + "grad_norm": 0.9365345239639282, + "learning_rate": 0.0006462240114145945, + "loss": 3.5881, + "step": 41655 + }, + { + "epoch": 2.830547628753907, + "grad_norm": 0.939250111579895, + "learning_rate": 0.0006461815464057617, + "loss": 3.4857, + "step": 41660 + }, + { + "epoch": 2.8308873488245685, + "grad_norm": 0.9623307585716248, + "learning_rate": 0.0006461390813969289, + "loss": 3.3108, + "step": 41665 + }, + { + "epoch": 2.83122706889523, + "grad_norm": 1.0015333890914917, + "learning_rate": 0.0006460966163880962, + "loss": 3.7509, + "step": 41670 + }, + { + "epoch": 2.831566788965892, + "grad_norm": 0.9369686841964722, + "learning_rate": 0.0006460541513792635, + "loss": 3.5462, + "step": 41675 + }, + { + "epoch": 2.831906509036554, + "grad_norm": 0.8528543710708618, + "learning_rate": 0.0006460116863704307, + "loss": 3.6148, + "step": 41680 + }, + { + "epoch": 2.8322462291072155, + "grad_norm": 0.8709750771522522, + "learning_rate": 0.000645969221361598, + "loss": 3.6066, + "step": 41685 + }, + { + "epoch": 2.8325859491778775, + "grad_norm": 0.7227299213409424, + "learning_rate": 0.0006459267563527654, + "loss": 3.6639, + "step": 41690 + }, + { + "epoch": 2.832925669248539, + "grad_norm": 0.9132999181747437, + "learning_rate": 0.0006458842913439326, + "loss": 3.7505, + "step": 41695 + }, + { + "epoch": 2.833265389319201, + "grad_norm": 1.0032469034194946, + "learning_rate": 0.0006458418263350999, + "loss": 3.5422, + "step": 41700 + }, + { + "epoch": 2.833605109389863, + "grad_norm": 0.7329068779945374, + "learning_rate": 0.0006457993613262672, + "loss": 3.6211, + "step": 41705 + }, + { + "epoch": 2.8339448294605245, + "grad_norm": 1.0247801542282104, + "learning_rate": 0.0006457568963174344, + "loss": 3.6968, + "step": 41710 + }, + { + "epoch": 2.834284549531186, + "grad_norm": 0.9737585186958313, + "learning_rate": 0.0006457144313086017, + "loss": 3.4537, + "step": 41715 + }, + { + "epoch": 2.834624269601848, + "grad_norm": 0.7124726176261902, + "learning_rate": 0.0006456719662997689, + "loss": 3.5022, + "step": 41720 + }, + { + "epoch": 2.83496398967251, + "grad_norm": 0.7406711578369141, + "learning_rate": 0.0006456295012909363, + "loss": 3.6671, + "step": 41725 + }, + { + "epoch": 2.8353037097431715, + "grad_norm": 0.8273912668228149, + "learning_rate": 0.0006455870362821036, + "loss": 3.7653, + "step": 41730 + }, + { + "epoch": 2.8356434298138335, + "grad_norm": 1.0623247623443604, + "learning_rate": 0.0006455445712732708, + "loss": 3.3684, + "step": 41735 + }, + { + "epoch": 2.835983149884495, + "grad_norm": 0.6887548565864563, + "learning_rate": 0.0006455021062644382, + "loss": 3.5987, + "step": 41740 + }, + { + "epoch": 2.836322869955157, + "grad_norm": 0.9533951282501221, + "learning_rate": 0.0006454596412556054, + "loss": 3.7599, + "step": 41745 + }, + { + "epoch": 2.836662590025819, + "grad_norm": 0.8202416300773621, + "learning_rate": 0.0006454171762467726, + "loss": 3.5961, + "step": 41750 + }, + { + "epoch": 2.8370023100964805, + "grad_norm": 0.7496135830879211, + "learning_rate": 0.00064537471123794, + "loss": 3.4107, + "step": 41755 + }, + { + "epoch": 2.837342030167142, + "grad_norm": 1.181650161743164, + "learning_rate": 0.0006453322462291073, + "loss": 3.8101, + "step": 41760 + }, + { + "epoch": 2.8376817502378042, + "grad_norm": 0.7017676830291748, + "learning_rate": 0.0006452897812202745, + "loss": 3.5105, + "step": 41765 + }, + { + "epoch": 2.838021470308466, + "grad_norm": 1.0137461423873901, + "learning_rate": 0.0006452473162114418, + "loss": 3.7202, + "step": 41770 + }, + { + "epoch": 2.8383611903791275, + "grad_norm": 0.8122454285621643, + "learning_rate": 0.0006452048512026091, + "loss": 3.6935, + "step": 41775 + }, + { + "epoch": 2.8387009104497896, + "grad_norm": 0.7557841539382935, + "learning_rate": 0.0006451623861937763, + "loss": 3.6296, + "step": 41780 + }, + { + "epoch": 2.839040630520451, + "grad_norm": 0.7402377128601074, + "learning_rate": 0.0006451199211849436, + "loss": 3.708, + "step": 41785 + }, + { + "epoch": 2.839380350591113, + "grad_norm": 0.7256947159767151, + "learning_rate": 0.0006450774561761109, + "loss": 3.5295, + "step": 41790 + }, + { + "epoch": 2.839720070661775, + "grad_norm": 0.8619740605354309, + "learning_rate": 0.0006450349911672782, + "loss": 3.6633, + "step": 41795 + }, + { + "epoch": 2.8400597907324365, + "grad_norm": 0.8553181290626526, + "learning_rate": 0.0006449925261584455, + "loss": 3.2943, + "step": 41800 + }, + { + "epoch": 2.840399510803098, + "grad_norm": 0.8918444514274597, + "learning_rate": 0.0006449500611496128, + "loss": 3.7773, + "step": 41805 + }, + { + "epoch": 2.8407392308737602, + "grad_norm": 0.9942034482955933, + "learning_rate": 0.00064490759614078, + "loss": 3.7759, + "step": 41810 + }, + { + "epoch": 2.841078950944422, + "grad_norm": 0.674082338809967, + "learning_rate": 0.0006448651311319473, + "loss": 3.7917, + "step": 41815 + }, + { + "epoch": 2.8414186710150835, + "grad_norm": 0.8962203860282898, + "learning_rate": 0.0006448226661231145, + "loss": 3.2653, + "step": 41820 + }, + { + "epoch": 2.8417583910857456, + "grad_norm": 0.881939172744751, + "learning_rate": 0.0006447802011142818, + "loss": 3.5808, + "step": 41825 + }, + { + "epoch": 2.842098111156407, + "grad_norm": 0.8823063969612122, + "learning_rate": 0.0006447377361054492, + "loss": 3.3298, + "step": 41830 + }, + { + "epoch": 2.842437831227069, + "grad_norm": 0.7699676752090454, + "learning_rate": 0.0006446952710966164, + "loss": 3.4607, + "step": 41835 + }, + { + "epoch": 2.842777551297731, + "grad_norm": 0.8576290011405945, + "learning_rate": 0.0006446528060877837, + "loss": 3.6437, + "step": 41840 + }, + { + "epoch": 2.8431172713683925, + "grad_norm": 0.9803963899612427, + "learning_rate": 0.000644610341078951, + "loss": 3.7071, + "step": 41845 + }, + { + "epoch": 2.843456991439054, + "grad_norm": 1.2538588047027588, + "learning_rate": 0.0006445678760701182, + "loss": 3.378, + "step": 41850 + }, + { + "epoch": 2.8437967115097162, + "grad_norm": 0.9848114252090454, + "learning_rate": 0.0006445254110612854, + "loss": 3.6173, + "step": 41855 + }, + { + "epoch": 2.844136431580378, + "grad_norm": 0.8595824837684631, + "learning_rate": 0.0006444829460524528, + "loss": 3.1623, + "step": 41860 + }, + { + "epoch": 2.8444761516510395, + "grad_norm": 0.8997084498405457, + "learning_rate": 0.0006444404810436201, + "loss": 3.335, + "step": 41865 + }, + { + "epoch": 2.8448158717217016, + "grad_norm": 0.9591615796089172, + "learning_rate": 0.0006443980160347873, + "loss": 3.5138, + "step": 41870 + }, + { + "epoch": 2.845155591792363, + "grad_norm": 0.8541892766952515, + "learning_rate": 0.0006443555510259547, + "loss": 3.4273, + "step": 41875 + }, + { + "epoch": 2.845495311863025, + "grad_norm": 0.8266332745552063, + "learning_rate": 0.0006443130860171219, + "loss": 3.384, + "step": 41880 + }, + { + "epoch": 2.8458350319336865, + "grad_norm": 0.8416678309440613, + "learning_rate": 0.0006442706210082891, + "loss": 3.4545, + "step": 41885 + }, + { + "epoch": 2.8461747520043486, + "grad_norm": 0.8054022192955017, + "learning_rate": 0.0006442281559994565, + "loss": 3.6852, + "step": 41890 + }, + { + "epoch": 2.84651447207501, + "grad_norm": 1.0356812477111816, + "learning_rate": 0.0006441856909906237, + "loss": 3.6916, + "step": 41895 + }, + { + "epoch": 2.846854192145672, + "grad_norm": 1.0687991380691528, + "learning_rate": 0.000644143225981791, + "loss": 3.5356, + "step": 41900 + }, + { + "epoch": 2.847193912216334, + "grad_norm": 0.7813867926597595, + "learning_rate": 0.0006441007609729584, + "loss": 3.5088, + "step": 41905 + }, + { + "epoch": 2.8475336322869955, + "grad_norm": 1.1813397407531738, + "learning_rate": 0.0006440582959641256, + "loss": 3.5169, + "step": 41910 + }, + { + "epoch": 2.847873352357657, + "grad_norm": 0.7769020199775696, + "learning_rate": 0.0006440158309552928, + "loss": 3.7573, + "step": 41915 + }, + { + "epoch": 2.8482130724283192, + "grad_norm": 0.8390051126480103, + "learning_rate": 0.0006439733659464601, + "loss": 3.8694, + "step": 41920 + }, + { + "epoch": 2.848552792498981, + "grad_norm": 0.8402785062789917, + "learning_rate": 0.0006439309009376274, + "loss": 3.5078, + "step": 41925 + }, + { + "epoch": 2.8488925125696425, + "grad_norm": 1.0050921440124512, + "learning_rate": 0.0006438884359287946, + "loss": 3.6184, + "step": 41930 + }, + { + "epoch": 2.849232232640304, + "grad_norm": 0.9718076586723328, + "learning_rate": 0.000643845970919962, + "loss": 3.4694, + "step": 41935 + }, + { + "epoch": 2.849571952710966, + "grad_norm": 1.109322428703308, + "learning_rate": 0.0006438035059111293, + "loss": 3.4884, + "step": 41940 + }, + { + "epoch": 2.849911672781628, + "grad_norm": 0.7805372476577759, + "learning_rate": 0.0006437610409022965, + "loss": 3.4998, + "step": 41945 + }, + { + "epoch": 2.8502513928522895, + "grad_norm": 1.2811968326568604, + "learning_rate": 0.0006437185758934638, + "loss": 3.4424, + "step": 41950 + }, + { + "epoch": 2.8505911129229515, + "grad_norm": 0.7912372946739197, + "learning_rate": 0.000643676110884631, + "loss": 3.5649, + "step": 41955 + }, + { + "epoch": 2.850930832993613, + "grad_norm": 0.7288410663604736, + "learning_rate": 0.0006436336458757983, + "loss": 3.6666, + "step": 41960 + }, + { + "epoch": 2.851270553064275, + "grad_norm": 0.8947926759719849, + "learning_rate": 0.0006435911808669657, + "loss": 3.4914, + "step": 41965 + }, + { + "epoch": 2.851610273134937, + "grad_norm": 0.6313999891281128, + "learning_rate": 0.0006435487158581329, + "loss": 3.6639, + "step": 41970 + }, + { + "epoch": 2.8519499932055985, + "grad_norm": 0.8949326872825623, + "learning_rate": 0.0006435062508493002, + "loss": 3.6489, + "step": 41975 + }, + { + "epoch": 2.85228971327626, + "grad_norm": 0.7926784753799438, + "learning_rate": 0.0006434637858404675, + "loss": 3.3057, + "step": 41980 + }, + { + "epoch": 2.852629433346922, + "grad_norm": 0.9715356826782227, + "learning_rate": 0.0006434213208316347, + "loss": 3.7546, + "step": 41985 + }, + { + "epoch": 2.852969153417584, + "grad_norm": 1.4090635776519775, + "learning_rate": 0.000643378855822802, + "loss": 3.4772, + "step": 41990 + }, + { + "epoch": 2.8533088734882455, + "grad_norm": 0.9378007650375366, + "learning_rate": 0.0006433363908139693, + "loss": 3.6404, + "step": 41995 + }, + { + "epoch": 2.8536485935589075, + "grad_norm": 0.6559866666793823, + "learning_rate": 0.0006432939258051366, + "loss": 3.725, + "step": 42000 + }, + { + "epoch": 2.853988313629569, + "grad_norm": 0.685090959072113, + "learning_rate": 0.0006432514607963038, + "loss": 3.6177, + "step": 42005 + }, + { + "epoch": 2.854328033700231, + "grad_norm": 1.0530961751937866, + "learning_rate": 0.0006432089957874712, + "loss": 3.2503, + "step": 42010 + }, + { + "epoch": 2.854667753770893, + "grad_norm": 0.7915587425231934, + "learning_rate": 0.0006431665307786384, + "loss": 3.5532, + "step": 42015 + }, + { + "epoch": 2.8550074738415545, + "grad_norm": 0.8528828620910645, + "learning_rate": 0.0006431240657698056, + "loss": 3.6983, + "step": 42020 + }, + { + "epoch": 2.855347193912216, + "grad_norm": 0.8784139156341553, + "learning_rate": 0.000643081600760973, + "loss": 3.844, + "step": 42025 + }, + { + "epoch": 2.855686913982878, + "grad_norm": 0.8080273270606995, + "learning_rate": 0.0006430391357521402, + "loss": 3.3913, + "step": 42030 + }, + { + "epoch": 2.85602663405354, + "grad_norm": 0.969186544418335, + "learning_rate": 0.0006429966707433075, + "loss": 3.2613, + "step": 42035 + }, + { + "epoch": 2.8563663541242015, + "grad_norm": 0.8907771706581116, + "learning_rate": 0.0006429542057344749, + "loss": 3.5874, + "step": 42040 + }, + { + "epoch": 2.8567060741948636, + "grad_norm": 0.9256178736686707, + "learning_rate": 0.0006429117407256421, + "loss": 3.7473, + "step": 42045 + }, + { + "epoch": 2.857045794265525, + "grad_norm": 0.8172616362571716, + "learning_rate": 0.0006428692757168093, + "loss": 3.4484, + "step": 42050 + }, + { + "epoch": 2.857385514336187, + "grad_norm": 0.8367016315460205, + "learning_rate": 0.0006428268107079766, + "loss": 3.6257, + "step": 42055 + }, + { + "epoch": 2.857725234406849, + "grad_norm": 0.7967295050621033, + "learning_rate": 0.0006427843456991439, + "loss": 3.4888, + "step": 42060 + }, + { + "epoch": 2.8580649544775105, + "grad_norm": 0.8925996422767639, + "learning_rate": 0.0006427418806903111, + "loss": 3.6407, + "step": 42065 + }, + { + "epoch": 2.858404674548172, + "grad_norm": 0.8647528290748596, + "learning_rate": 0.0006426994156814785, + "loss": 3.5231, + "step": 42070 + }, + { + "epoch": 2.8587443946188342, + "grad_norm": 0.9978330731391907, + "learning_rate": 0.0006426569506726458, + "loss": 3.5665, + "step": 42075 + }, + { + "epoch": 2.859084114689496, + "grad_norm": 0.9879549741744995, + "learning_rate": 0.0006426144856638131, + "loss": 3.5062, + "step": 42080 + }, + { + "epoch": 2.8594238347601575, + "grad_norm": 0.7880136966705322, + "learning_rate": 0.0006425720206549803, + "loss": 3.4226, + "step": 42085 + }, + { + "epoch": 2.8597635548308196, + "grad_norm": 0.7803234457969666, + "learning_rate": 0.0006425295556461476, + "loss": 3.6137, + "step": 42090 + }, + { + "epoch": 2.860103274901481, + "grad_norm": 1.073130488395691, + "learning_rate": 0.0006424870906373149, + "loss": 3.5561, + "step": 42095 + }, + { + "epoch": 2.860442994972143, + "grad_norm": 0.9077723622322083, + "learning_rate": 0.0006424446256284821, + "loss": 3.663, + "step": 42100 + }, + { + "epoch": 2.860782715042805, + "grad_norm": 0.6709428429603577, + "learning_rate": 0.0006424021606196495, + "loss": 3.2919, + "step": 42105 + }, + { + "epoch": 2.8611224351134665, + "grad_norm": 0.9544896483421326, + "learning_rate": 0.0006423596956108168, + "loss": 3.8351, + "step": 42110 + }, + { + "epoch": 2.861462155184128, + "grad_norm": 0.9154482483863831, + "learning_rate": 0.000642317230601984, + "loss": 3.4774, + "step": 42115 + }, + { + "epoch": 2.8618018752547902, + "grad_norm": 0.8827584385871887, + "learning_rate": 0.0006422747655931512, + "loss": 3.6207, + "step": 42120 + }, + { + "epoch": 2.862141595325452, + "grad_norm": 0.8696451187133789, + "learning_rate": 0.0006422323005843186, + "loss": 3.6162, + "step": 42125 + }, + { + "epoch": 2.8624813153961135, + "grad_norm": 0.9808441400527954, + "learning_rate": 0.0006421898355754858, + "loss": 3.5638, + "step": 42130 + }, + { + "epoch": 2.8628210354667756, + "grad_norm": 1.0594931840896606, + "learning_rate": 0.000642147370566653, + "loss": 3.5955, + "step": 42135 + }, + { + "epoch": 2.863160755537437, + "grad_norm": 0.8295979499816895, + "learning_rate": 0.0006421049055578205, + "loss": 3.2135, + "step": 42140 + }, + { + "epoch": 2.863500475608099, + "grad_norm": 1.2168315649032593, + "learning_rate": 0.0006420624405489877, + "loss": 3.5142, + "step": 42145 + }, + { + "epoch": 2.863840195678761, + "grad_norm": 0.9181390404701233, + "learning_rate": 0.0006420199755401549, + "loss": 3.4905, + "step": 42150 + }, + { + "epoch": 2.8641799157494225, + "grad_norm": 0.8940221667289734, + "learning_rate": 0.0006419775105313223, + "loss": 3.6417, + "step": 42155 + }, + { + "epoch": 2.864519635820084, + "grad_norm": 0.6922504901885986, + "learning_rate": 0.0006419350455224895, + "loss": 3.4489, + "step": 42160 + }, + { + "epoch": 2.8648593558907463, + "grad_norm": 0.7633323073387146, + "learning_rate": 0.0006418925805136567, + "loss": 3.3769, + "step": 42165 + }, + { + "epoch": 2.865199075961408, + "grad_norm": 0.910102367401123, + "learning_rate": 0.000641850115504824, + "loss": 3.5851, + "step": 42170 + }, + { + "epoch": 2.8655387960320695, + "grad_norm": 0.8751612305641174, + "learning_rate": 0.0006418076504959914, + "loss": 3.5135, + "step": 42175 + }, + { + "epoch": 2.8658785161027316, + "grad_norm": 0.8148460984230042, + "learning_rate": 0.0006417651854871586, + "loss": 3.6535, + "step": 42180 + }, + { + "epoch": 2.866218236173393, + "grad_norm": 0.8185792565345764, + "learning_rate": 0.0006417227204783259, + "loss": 3.2767, + "step": 42185 + }, + { + "epoch": 2.866557956244055, + "grad_norm": 0.7995314598083496, + "learning_rate": 0.0006416802554694932, + "loss": 3.547, + "step": 42190 + }, + { + "epoch": 2.866897676314717, + "grad_norm": 1.3204153776168823, + "learning_rate": 0.0006416377904606604, + "loss": 3.5845, + "step": 42195 + }, + { + "epoch": 2.8672373963853786, + "grad_norm": 6.705195903778076, + "learning_rate": 0.0006415953254518277, + "loss": 3.5688, + "step": 42200 + }, + { + "epoch": 2.86757711645604, + "grad_norm": 0.9194809198379517, + "learning_rate": 0.0006415528604429949, + "loss": 3.4978, + "step": 42205 + }, + { + "epoch": 2.8679168365267023, + "grad_norm": 1.4958008527755737, + "learning_rate": 0.0006415103954341623, + "loss": 3.6392, + "step": 42210 + }, + { + "epoch": 2.868256556597364, + "grad_norm": 0.7360958456993103, + "learning_rate": 0.0006414679304253296, + "loss": 3.5015, + "step": 42215 + }, + { + "epoch": 2.8685962766680255, + "grad_norm": 0.9407960772514343, + "learning_rate": 0.0006414254654164968, + "loss": 3.2406, + "step": 42220 + }, + { + "epoch": 2.868935996738687, + "grad_norm": 0.8606539964675903, + "learning_rate": 0.0006413830004076641, + "loss": 3.7269, + "step": 42225 + }, + { + "epoch": 2.8692757168093492, + "grad_norm": 1.269915223121643, + "learning_rate": 0.0006413405353988314, + "loss": 3.3868, + "step": 42230 + }, + { + "epoch": 2.869615436880011, + "grad_norm": 0.8137083053588867, + "learning_rate": 0.0006412980703899986, + "loss": 3.5869, + "step": 42235 + }, + { + "epoch": 2.8699551569506725, + "grad_norm": 0.8669810891151428, + "learning_rate": 0.0006412556053811658, + "loss": 3.6903, + "step": 42240 + }, + { + "epoch": 2.8702948770213346, + "grad_norm": 0.9795706272125244, + "learning_rate": 0.0006412131403723333, + "loss": 3.6634, + "step": 42245 + }, + { + "epoch": 2.870634597091996, + "grad_norm": 1.6822649240493774, + "learning_rate": 0.0006411706753635005, + "loss": 3.5108, + "step": 42250 + }, + { + "epoch": 2.870974317162658, + "grad_norm": 0.968251645565033, + "learning_rate": 0.0006411282103546677, + "loss": 3.6763, + "step": 42255 + }, + { + "epoch": 2.87131403723332, + "grad_norm": 0.6832924485206604, + "learning_rate": 0.0006410857453458351, + "loss": 3.4391, + "step": 42260 + }, + { + "epoch": 2.8716537573039815, + "grad_norm": 0.8875095844268799, + "learning_rate": 0.0006410432803370023, + "loss": 3.4593, + "step": 42265 + }, + { + "epoch": 2.871993477374643, + "grad_norm": 1.0385401248931885, + "learning_rate": 0.0006410008153281695, + "loss": 3.4715, + "step": 42270 + }, + { + "epoch": 2.872333197445305, + "grad_norm": 0.9741016626358032, + "learning_rate": 0.0006409583503193369, + "loss": 3.7553, + "step": 42275 + }, + { + "epoch": 2.872672917515967, + "grad_norm": 0.7240934371948242, + "learning_rate": 0.0006409158853105042, + "loss": 3.6119, + "step": 42280 + }, + { + "epoch": 2.8730126375866285, + "grad_norm": 0.9612627029418945, + "learning_rate": 0.0006408734203016714, + "loss": 3.4585, + "step": 42285 + }, + { + "epoch": 2.87335235765729, + "grad_norm": 1.1005984544754028, + "learning_rate": 0.0006408309552928388, + "loss": 3.4789, + "step": 42290 + }, + { + "epoch": 2.873692077727952, + "grad_norm": 0.8549172878265381, + "learning_rate": 0.000640788490284006, + "loss": 3.5615, + "step": 42295 + }, + { + "epoch": 2.874031797798614, + "grad_norm": 0.7889883518218994, + "learning_rate": 0.0006407460252751732, + "loss": 3.7247, + "step": 42300 + }, + { + "epoch": 2.8743715178692755, + "grad_norm": 0.8491808176040649, + "learning_rate": 0.0006407035602663405, + "loss": 3.5565, + "step": 42305 + }, + { + "epoch": 2.8747112379399375, + "grad_norm": 1.0010915994644165, + "learning_rate": 0.0006406610952575078, + "loss": 3.548, + "step": 42310 + }, + { + "epoch": 2.875050958010599, + "grad_norm": 0.90715491771698, + "learning_rate": 0.0006406186302486751, + "loss": 3.585, + "step": 42315 + }, + { + "epoch": 2.875390678081261, + "grad_norm": 0.8811260461807251, + "learning_rate": 0.0006405761652398424, + "loss": 3.5687, + "step": 42320 + }, + { + "epoch": 2.875730398151923, + "grad_norm": 0.8841518759727478, + "learning_rate": 0.0006405337002310097, + "loss": 3.5043, + "step": 42325 + }, + { + "epoch": 2.8760701182225845, + "grad_norm": 0.9076136946678162, + "learning_rate": 0.0006404912352221769, + "loss": 3.5579, + "step": 42330 + }, + { + "epoch": 2.876409838293246, + "grad_norm": 1.0862607955932617, + "learning_rate": 0.0006404487702133442, + "loss": 3.4655, + "step": 42335 + }, + { + "epoch": 2.876749558363908, + "grad_norm": 0.7681044340133667, + "learning_rate": 0.0006404063052045115, + "loss": 3.6666, + "step": 42340 + }, + { + "epoch": 2.87708927843457, + "grad_norm": 1.1122334003448486, + "learning_rate": 0.0006403638401956787, + "loss": 3.4034, + "step": 42345 + }, + { + "epoch": 2.8774289985052315, + "grad_norm": 0.76408851146698, + "learning_rate": 0.0006403213751868461, + "loss": 3.7245, + "step": 42350 + }, + { + "epoch": 2.8777687185758936, + "grad_norm": 0.800439178943634, + "learning_rate": 0.0006402789101780133, + "loss": 3.5469, + "step": 42355 + }, + { + "epoch": 2.878108438646555, + "grad_norm": 0.7807614803314209, + "learning_rate": 0.0006402364451691806, + "loss": 3.6776, + "step": 42360 + }, + { + "epoch": 2.878448158717217, + "grad_norm": 0.7730633020401001, + "learning_rate": 0.0006401939801603479, + "loss": 3.5172, + "step": 42365 + }, + { + "epoch": 2.878787878787879, + "grad_norm": 0.946793258190155, + "learning_rate": 0.0006401515151515151, + "loss": 3.6086, + "step": 42370 + }, + { + "epoch": 2.8791275988585405, + "grad_norm": 0.8278414607048035, + "learning_rate": 0.0006401090501426824, + "loss": 3.8165, + "step": 42375 + }, + { + "epoch": 2.879467318929202, + "grad_norm": 0.8552778959274292, + "learning_rate": 0.0006400665851338497, + "loss": 3.6534, + "step": 42380 + }, + { + "epoch": 2.8798070389998642, + "grad_norm": 0.6798197031021118, + "learning_rate": 0.000640024120125017, + "loss": 3.4707, + "step": 42385 + }, + { + "epoch": 2.880146759070526, + "grad_norm": 0.7311868071556091, + "learning_rate": 0.0006399816551161843, + "loss": 3.504, + "step": 42390 + }, + { + "epoch": 2.8804864791411875, + "grad_norm": 0.8453450202941895, + "learning_rate": 0.0006399391901073516, + "loss": 3.5748, + "step": 42395 + }, + { + "epoch": 2.8808261992118496, + "grad_norm": 0.897329568862915, + "learning_rate": 0.0006398967250985188, + "loss": 3.7554, + "step": 42400 + }, + { + "epoch": 2.881165919282511, + "grad_norm": 0.9080669283866882, + "learning_rate": 0.000639854260089686, + "loss": 3.4393, + "step": 42405 + }, + { + "epoch": 2.881505639353173, + "grad_norm": 0.892851710319519, + "learning_rate": 0.0006398117950808534, + "loss": 3.6194, + "step": 42410 + }, + { + "epoch": 2.881845359423835, + "grad_norm": 0.9177255034446716, + "learning_rate": 0.0006397693300720206, + "loss": 3.6608, + "step": 42415 + }, + { + "epoch": 2.8821850794944965, + "grad_norm": 1.098097562789917, + "learning_rate": 0.000639726865063188, + "loss": 3.827, + "step": 42420 + }, + { + "epoch": 2.882524799565158, + "grad_norm": 1.0553370714187622, + "learning_rate": 0.0006396844000543553, + "loss": 3.3986, + "step": 42425 + }, + { + "epoch": 2.8828645196358202, + "grad_norm": 0.9975004196166992, + "learning_rate": 0.0006396419350455225, + "loss": 3.6784, + "step": 42430 + }, + { + "epoch": 2.883204239706482, + "grad_norm": 0.7348812818527222, + "learning_rate": 0.0006395994700366898, + "loss": 3.5928, + "step": 42435 + }, + { + "epoch": 2.8835439597771435, + "grad_norm": 0.9415812492370605, + "learning_rate": 0.000639557005027857, + "loss": 3.5491, + "step": 42440 + }, + { + "epoch": 2.8838836798478056, + "grad_norm": 0.7492969036102295, + "learning_rate": 0.0006395145400190243, + "loss": 3.8582, + "step": 42445 + }, + { + "epoch": 2.884223399918467, + "grad_norm": 0.9638067483901978, + "learning_rate": 0.0006394720750101916, + "loss": 3.3706, + "step": 42450 + }, + { + "epoch": 2.884563119989129, + "grad_norm": 0.8698820471763611, + "learning_rate": 0.000639429610001359, + "loss": 3.5963, + "step": 42455 + }, + { + "epoch": 2.884902840059791, + "grad_norm": 0.7817463874816895, + "learning_rate": 0.0006393871449925262, + "loss": 3.6897, + "step": 42460 + }, + { + "epoch": 2.8852425601304525, + "grad_norm": 0.6357342004776001, + "learning_rate": 0.0006393446799836935, + "loss": 3.631, + "step": 42465 + }, + { + "epoch": 2.885582280201114, + "grad_norm": 0.9060226678848267, + "learning_rate": 0.0006393022149748607, + "loss": 3.4765, + "step": 42470 + }, + { + "epoch": 2.8859220002717763, + "grad_norm": 1.1503204107284546, + "learning_rate": 0.000639259749966028, + "loss": 3.7089, + "step": 42475 + }, + { + "epoch": 2.886261720342438, + "grad_norm": 0.8642039895057678, + "learning_rate": 0.0006392172849571953, + "loss": 3.7822, + "step": 42480 + }, + { + "epoch": 2.8866014404130995, + "grad_norm": 0.7597491145133972, + "learning_rate": 0.0006391748199483625, + "loss": 3.5902, + "step": 42485 + }, + { + "epoch": 2.8869411604837616, + "grad_norm": 0.7859035134315491, + "learning_rate": 0.0006391323549395299, + "loss": 3.6041, + "step": 42490 + }, + { + "epoch": 2.8872808805544232, + "grad_norm": 0.851076066493988, + "learning_rate": 0.0006390898899306972, + "loss": 3.6319, + "step": 42495 + }, + { + "epoch": 2.887620600625085, + "grad_norm": 0.8372633457183838, + "learning_rate": 0.0006390474249218644, + "loss": 3.5286, + "step": 42500 + }, + { + "epoch": 2.887960320695747, + "grad_norm": 0.796132504940033, + "learning_rate": 0.0006390049599130316, + "loss": 3.5349, + "step": 42505 + }, + { + "epoch": 2.8883000407664086, + "grad_norm": 1.2908692359924316, + "learning_rate": 0.000638962494904199, + "loss": 3.4829, + "step": 42510 + }, + { + "epoch": 2.88863976083707, + "grad_norm": 0.8849920034408569, + "learning_rate": 0.0006389200298953662, + "loss": 3.2734, + "step": 42515 + }, + { + "epoch": 2.8889794809077323, + "grad_norm": 0.9792762398719788, + "learning_rate": 0.0006388775648865334, + "loss": 3.7267, + "step": 42520 + }, + { + "epoch": 2.889319200978394, + "grad_norm": 1.0214260816574097, + "learning_rate": 0.0006388350998777009, + "loss": 3.5204, + "step": 42525 + }, + { + "epoch": 2.8896589210490555, + "grad_norm": 0.9677252769470215, + "learning_rate": 0.0006387926348688681, + "loss": 3.5951, + "step": 42530 + }, + { + "epoch": 2.8899986411197176, + "grad_norm": 0.9316784143447876, + "learning_rate": 0.0006387501698600353, + "loss": 3.2639, + "step": 42535 + }, + { + "epoch": 2.8903383611903792, + "grad_norm": 0.988141655921936, + "learning_rate": 0.0006387077048512027, + "loss": 3.7225, + "step": 42540 + }, + { + "epoch": 2.890678081261041, + "grad_norm": 0.8708176612854004, + "learning_rate": 0.0006386652398423699, + "loss": 3.4157, + "step": 42545 + }, + { + "epoch": 2.891017801331703, + "grad_norm": 1.3555513620376587, + "learning_rate": 0.0006386227748335371, + "loss": 3.3983, + "step": 42550 + }, + { + "epoch": 2.8913575214023646, + "grad_norm": 1.010156273841858, + "learning_rate": 0.0006385803098247045, + "loss": 3.4824, + "step": 42555 + }, + { + "epoch": 2.891697241473026, + "grad_norm": 0.8730334043502808, + "learning_rate": 0.0006385378448158718, + "loss": 3.6738, + "step": 42560 + }, + { + "epoch": 2.8920369615436883, + "grad_norm": 0.9817825555801392, + "learning_rate": 0.000638495379807039, + "loss": 3.6883, + "step": 42565 + }, + { + "epoch": 2.89237668161435, + "grad_norm": 0.7821949124336243, + "learning_rate": 0.0006384529147982063, + "loss": 3.5876, + "step": 42570 + }, + { + "epoch": 2.8927164016850115, + "grad_norm": 0.9443778991699219, + "learning_rate": 0.0006384104497893736, + "loss": 3.5109, + "step": 42575 + }, + { + "epoch": 2.893056121755673, + "grad_norm": 0.8248337507247925, + "learning_rate": 0.0006383679847805408, + "loss": 3.4168, + "step": 42580 + }, + { + "epoch": 2.8933958418263352, + "grad_norm": 0.7502477765083313, + "learning_rate": 0.0006383255197717081, + "loss": 3.5999, + "step": 42585 + }, + { + "epoch": 2.893735561896997, + "grad_norm": 1.1297409534454346, + "learning_rate": 0.0006382830547628755, + "loss": 3.5377, + "step": 42590 + }, + { + "epoch": 2.8940752819676585, + "grad_norm": 1.4060485363006592, + "learning_rate": 0.0006382405897540427, + "loss": 3.626, + "step": 42595 + }, + { + "epoch": 2.8944150020383206, + "grad_norm": 0.9989623427391052, + "learning_rate": 0.00063819812474521, + "loss": 3.6374, + "step": 42600 + }, + { + "epoch": 2.894754722108982, + "grad_norm": 0.9649537801742554, + "learning_rate": 0.0006381556597363772, + "loss": 3.2851, + "step": 42605 + }, + { + "epoch": 2.895094442179644, + "grad_norm": 0.8392823338508606, + "learning_rate": 0.0006381131947275445, + "loss": 3.4792, + "step": 42610 + }, + { + "epoch": 2.8954341622503055, + "grad_norm": 0.9658290147781372, + "learning_rate": 0.0006380707297187118, + "loss": 3.3912, + "step": 42615 + }, + { + "epoch": 2.8957738823209676, + "grad_norm": 0.8299267292022705, + "learning_rate": 0.000638028264709879, + "loss": 3.5084, + "step": 42620 + }, + { + "epoch": 2.896113602391629, + "grad_norm": 0.9467092156410217, + "learning_rate": 0.0006379857997010464, + "loss": 3.414, + "step": 42625 + }, + { + "epoch": 2.896453322462291, + "grad_norm": 0.9420976638793945, + "learning_rate": 0.0006379433346922137, + "loss": 3.5193, + "step": 42630 + }, + { + "epoch": 2.896793042532953, + "grad_norm": 0.9384837746620178, + "learning_rate": 0.0006379008696833809, + "loss": 3.6153, + "step": 42635 + }, + { + "epoch": 2.8971327626036145, + "grad_norm": 0.933294951915741, + "learning_rate": 0.0006378584046745481, + "loss": 3.5723, + "step": 42640 + }, + { + "epoch": 2.897472482674276, + "grad_norm": 0.8453119993209839, + "learning_rate": 0.0006378159396657155, + "loss": 3.8758, + "step": 42645 + }, + { + "epoch": 2.8978122027449382, + "grad_norm": 0.6300462484359741, + "learning_rate": 0.0006377734746568827, + "loss": 3.5315, + "step": 42650 + }, + { + "epoch": 2.8981519228156, + "grad_norm": 0.8902947306632996, + "learning_rate": 0.0006377310096480499, + "loss": 3.4045, + "step": 42655 + }, + { + "epoch": 2.8984916428862615, + "grad_norm": 0.9883577823638916, + "learning_rate": 0.0006376885446392174, + "loss": 3.5604, + "step": 42660 + }, + { + "epoch": 2.8988313629569236, + "grad_norm": 0.9162705540657043, + "learning_rate": 0.0006376460796303846, + "loss": 3.627, + "step": 42665 + }, + { + "epoch": 2.899171083027585, + "grad_norm": 0.9432220458984375, + "learning_rate": 0.0006376036146215518, + "loss": 3.4278, + "step": 42670 + }, + { + "epoch": 2.899510803098247, + "grad_norm": 0.8451558947563171, + "learning_rate": 0.0006375611496127192, + "loss": 3.5946, + "step": 42675 + }, + { + "epoch": 2.899850523168909, + "grad_norm": 0.8453450798988342, + "learning_rate": 0.0006375186846038864, + "loss": 3.4331, + "step": 42680 + }, + { + "epoch": 2.9001902432395705, + "grad_norm": 1.061569333076477, + "learning_rate": 0.0006374762195950536, + "loss": 3.5746, + "step": 42685 + }, + { + "epoch": 2.900529963310232, + "grad_norm": 0.9832978844642639, + "learning_rate": 0.000637433754586221, + "loss": 3.4752, + "step": 42690 + }, + { + "epoch": 2.9008696833808942, + "grad_norm": 1.009527325630188, + "learning_rate": 0.0006373912895773883, + "loss": 3.6149, + "step": 42695 + }, + { + "epoch": 2.901209403451556, + "grad_norm": 0.787018895149231, + "learning_rate": 0.0006373488245685555, + "loss": 3.6025, + "step": 42700 + }, + { + "epoch": 2.9015491235222175, + "grad_norm": 1.1298191547393799, + "learning_rate": 0.0006373063595597228, + "loss": 3.8568, + "step": 42705 + }, + { + "epoch": 2.9018888435928796, + "grad_norm": 0.9390883445739746, + "learning_rate": 0.0006372638945508901, + "loss": 3.6322, + "step": 42710 + }, + { + "epoch": 2.902228563663541, + "grad_norm": 0.8891267776489258, + "learning_rate": 0.0006372214295420573, + "loss": 3.4913, + "step": 42715 + }, + { + "epoch": 2.902568283734203, + "grad_norm": 0.7465917468070984, + "learning_rate": 0.0006371789645332246, + "loss": 3.4973, + "step": 42720 + }, + { + "epoch": 2.902908003804865, + "grad_norm": 0.897925853729248, + "learning_rate": 0.0006371364995243919, + "loss": 3.5614, + "step": 42725 + }, + { + "epoch": 2.9032477238755265, + "grad_norm": 1.2403651475906372, + "learning_rate": 0.0006370940345155592, + "loss": 3.6058, + "step": 42730 + }, + { + "epoch": 2.903587443946188, + "grad_norm": 0.6772661805152893, + "learning_rate": 0.0006370515695067265, + "loss": 3.6434, + "step": 42735 + }, + { + "epoch": 2.9039271640168502, + "grad_norm": 1.0289537906646729, + "learning_rate": 0.0006370091044978937, + "loss": 3.5642, + "step": 42740 + }, + { + "epoch": 2.904266884087512, + "grad_norm": 0.84995037317276, + "learning_rate": 0.000636966639489061, + "loss": 3.3625, + "step": 42745 + }, + { + "epoch": 2.9046066041581735, + "grad_norm": 0.8925580382347107, + "learning_rate": 0.0006369241744802283, + "loss": 3.4654, + "step": 42750 + }, + { + "epoch": 2.9049463242288356, + "grad_norm": 0.859586775302887, + "learning_rate": 0.0006368817094713955, + "loss": 3.5087, + "step": 42755 + }, + { + "epoch": 2.905286044299497, + "grad_norm": 0.9112608432769775, + "learning_rate": 0.0006368392444625629, + "loss": 3.5589, + "step": 42760 + }, + { + "epoch": 2.905625764370159, + "grad_norm": 0.8865873217582703, + "learning_rate": 0.0006367967794537302, + "loss": 3.5169, + "step": 42765 + }, + { + "epoch": 2.905965484440821, + "grad_norm": 0.6465127468109131, + "learning_rate": 0.0006367543144448974, + "loss": 3.5673, + "step": 42770 + }, + { + "epoch": 2.9063052045114826, + "grad_norm": 0.7113096714019775, + "learning_rate": 0.0006367118494360648, + "loss": 3.734, + "step": 42775 + }, + { + "epoch": 2.906644924582144, + "grad_norm": 1.079235553741455, + "learning_rate": 0.000636669384427232, + "loss": 3.6069, + "step": 42780 + }, + { + "epoch": 2.9069846446528063, + "grad_norm": 0.8976610898971558, + "learning_rate": 0.0006366269194183992, + "loss": 3.2982, + "step": 42785 + }, + { + "epoch": 2.907324364723468, + "grad_norm": 0.872797429561615, + "learning_rate": 0.0006365844544095666, + "loss": 3.5512, + "step": 42790 + }, + { + "epoch": 2.9076640847941295, + "grad_norm": 0.8457685708999634, + "learning_rate": 0.0006365419894007338, + "loss": 3.3455, + "step": 42795 + }, + { + "epoch": 2.9080038048647916, + "grad_norm": 0.7257461547851562, + "learning_rate": 0.0006364995243919011, + "loss": 3.6599, + "step": 42800 + }, + { + "epoch": 2.9083435249354532, + "grad_norm": 1.4832161664962769, + "learning_rate": 0.0006364570593830684, + "loss": 3.4785, + "step": 42805 + }, + { + "epoch": 2.908683245006115, + "grad_norm": 0.7436572909355164, + "learning_rate": 0.0006364145943742357, + "loss": 3.7304, + "step": 42810 + }, + { + "epoch": 2.909022965076777, + "grad_norm": 0.8569587469100952, + "learning_rate": 0.0006363721293654029, + "loss": 3.6556, + "step": 42815 + }, + { + "epoch": 2.9093626851474386, + "grad_norm": 0.6778474450111389, + "learning_rate": 0.0006363296643565702, + "loss": 3.5356, + "step": 42820 + }, + { + "epoch": 2.9097024052181, + "grad_norm": 0.7488446235656738, + "learning_rate": 0.0006362871993477375, + "loss": 3.5043, + "step": 42825 + }, + { + "epoch": 2.9100421252887623, + "grad_norm": 0.6841427683830261, + "learning_rate": 0.0006362447343389047, + "loss": 3.4015, + "step": 42830 + }, + { + "epoch": 2.910381845359424, + "grad_norm": 0.947187066078186, + "learning_rate": 0.0006362022693300721, + "loss": 3.7585, + "step": 42835 + }, + { + "epoch": 2.9107215654300855, + "grad_norm": 0.817452609539032, + "learning_rate": 0.0006361598043212394, + "loss": 3.6684, + "step": 42840 + }, + { + "epoch": 2.9110612855007476, + "grad_norm": 0.6989634037017822, + "learning_rate": 0.0006361173393124066, + "loss": 3.4543, + "step": 42845 + }, + { + "epoch": 2.9114010055714092, + "grad_norm": 0.9328689575195312, + "learning_rate": 0.0006360748743035739, + "loss": 3.5661, + "step": 42850 + }, + { + "epoch": 2.911740725642071, + "grad_norm": 1.1119638681411743, + "learning_rate": 0.0006360324092947411, + "loss": 3.5179, + "step": 42855 + }, + { + "epoch": 2.912080445712733, + "grad_norm": 0.8440215587615967, + "learning_rate": 0.0006359899442859084, + "loss": 3.6756, + "step": 42860 + }, + { + "epoch": 2.9124201657833946, + "grad_norm": 0.9562031626701355, + "learning_rate": 0.0006359474792770757, + "loss": 3.5806, + "step": 42865 + }, + { + "epoch": 2.912759885854056, + "grad_norm": 0.827194333076477, + "learning_rate": 0.000635905014268243, + "loss": 3.688, + "step": 42870 + }, + { + "epoch": 2.9130996059247183, + "grad_norm": 0.8765027523040771, + "learning_rate": 0.0006358625492594103, + "loss": 3.5777, + "step": 42875 + }, + { + "epoch": 2.91343932599538, + "grad_norm": 0.7576547861099243, + "learning_rate": 0.0006358200842505776, + "loss": 3.5804, + "step": 42880 + }, + { + "epoch": 2.9137790460660415, + "grad_norm": 0.7906240224838257, + "learning_rate": 0.0006357776192417448, + "loss": 3.8544, + "step": 42885 + }, + { + "epoch": 2.9141187661367036, + "grad_norm": 0.810924768447876, + "learning_rate": 0.000635735154232912, + "loss": 3.5058, + "step": 42890 + }, + { + "epoch": 2.9144584862073653, + "grad_norm": 0.8478354215621948, + "learning_rate": 0.0006356926892240794, + "loss": 3.5976, + "step": 42895 + }, + { + "epoch": 2.914798206278027, + "grad_norm": 0.8921616077423096, + "learning_rate": 0.0006356502242152466, + "loss": 3.3183, + "step": 42900 + }, + { + "epoch": 2.915137926348689, + "grad_norm": 0.7466316223144531, + "learning_rate": 0.0006356077592064139, + "loss": 3.684, + "step": 42905 + }, + { + "epoch": 2.9154776464193506, + "grad_norm": 0.8609406352043152, + "learning_rate": 0.0006355652941975813, + "loss": 3.4325, + "step": 42910 + }, + { + "epoch": 2.915817366490012, + "grad_norm": 0.6886776685714722, + "learning_rate": 0.0006355228291887485, + "loss": 3.6301, + "step": 42915 + }, + { + "epoch": 2.916157086560674, + "grad_norm": 1.0339919328689575, + "learning_rate": 0.0006354803641799157, + "loss": 3.5278, + "step": 42920 + }, + { + "epoch": 2.916496806631336, + "grad_norm": 0.8952237367630005, + "learning_rate": 0.0006354378991710831, + "loss": 3.6656, + "step": 42925 + }, + { + "epoch": 2.9168365267019976, + "grad_norm": 0.8915913701057434, + "learning_rate": 0.0006353954341622503, + "loss": 3.442, + "step": 42930 + }, + { + "epoch": 2.917176246772659, + "grad_norm": 0.9781518578529358, + "learning_rate": 0.0006353529691534175, + "loss": 3.6552, + "step": 42935 + }, + { + "epoch": 2.9175159668433213, + "grad_norm": 0.6803771257400513, + "learning_rate": 0.000635310504144585, + "loss": 3.7985, + "step": 42940 + }, + { + "epoch": 2.917855686913983, + "grad_norm": 0.7576701045036316, + "learning_rate": 0.0006352680391357522, + "loss": 3.7594, + "step": 42945 + }, + { + "epoch": 2.9181954069846445, + "grad_norm": 0.850638747215271, + "learning_rate": 0.0006352255741269194, + "loss": 3.4467, + "step": 42950 + }, + { + "epoch": 2.918535127055306, + "grad_norm": 1.1481186151504517, + "learning_rate": 0.0006351831091180867, + "loss": 3.429, + "step": 42955 + }, + { + "epoch": 2.9188748471259682, + "grad_norm": 0.7756760120391846, + "learning_rate": 0.000635140644109254, + "loss": 3.4955, + "step": 42960 + }, + { + "epoch": 2.91921456719663, + "grad_norm": 0.7409411072731018, + "learning_rate": 0.0006350981791004212, + "loss": 3.6201, + "step": 42965 + }, + { + "epoch": 2.9195542872672915, + "grad_norm": 0.9981425404548645, + "learning_rate": 0.0006350557140915885, + "loss": 3.6949, + "step": 42970 + }, + { + "epoch": 2.9198940073379536, + "grad_norm": 0.6571142077445984, + "learning_rate": 0.0006350132490827559, + "loss": 3.4643, + "step": 42975 + }, + { + "epoch": 2.920233727408615, + "grad_norm": 0.7857457995414734, + "learning_rate": 0.0006349707840739231, + "loss": 3.708, + "step": 42980 + }, + { + "epoch": 2.920573447479277, + "grad_norm": 1.0771361589431763, + "learning_rate": 0.0006349283190650904, + "loss": 3.3373, + "step": 42985 + }, + { + "epoch": 2.920913167549939, + "grad_norm": 0.9270572066307068, + "learning_rate": 0.0006348858540562576, + "loss": 3.5306, + "step": 42990 + }, + { + "epoch": 2.9212528876206005, + "grad_norm": 0.9011585116386414, + "learning_rate": 0.0006348433890474249, + "loss": 3.5624, + "step": 42995 + }, + { + "epoch": 2.921592607691262, + "grad_norm": 0.8225682973861694, + "learning_rate": 0.0006348009240385922, + "loss": 3.5298, + "step": 43000 + }, + { + "epoch": 2.9219323277619242, + "grad_norm": 0.8072992563247681, + "learning_rate": 0.0006347584590297594, + "loss": 3.7955, + "step": 43005 + }, + { + "epoch": 2.922272047832586, + "grad_norm": 0.8372965455055237, + "learning_rate": 0.0006347159940209268, + "loss": 3.5219, + "step": 43010 + }, + { + "epoch": 2.9226117679032475, + "grad_norm": 0.890300989151001, + "learning_rate": 0.0006346735290120941, + "loss": 3.2631, + "step": 43015 + }, + { + "epoch": 2.9229514879739096, + "grad_norm": 0.9603749513626099, + "learning_rate": 0.0006346310640032613, + "loss": 3.6234, + "step": 43020 + }, + { + "epoch": 2.923291208044571, + "grad_norm": 0.8113959431648254, + "learning_rate": 0.0006345885989944286, + "loss": 3.7105, + "step": 43025 + }, + { + "epoch": 2.923630928115233, + "grad_norm": 0.9562439918518066, + "learning_rate": 0.0006345461339855959, + "loss": 3.6015, + "step": 43030 + }, + { + "epoch": 2.923970648185895, + "grad_norm": 0.6846343874931335, + "learning_rate": 0.0006345036689767631, + "loss": 3.5917, + "step": 43035 + }, + { + "epoch": 2.9243103682565565, + "grad_norm": 0.7247750163078308, + "learning_rate": 0.0006344612039679303, + "loss": 3.5819, + "step": 43040 + }, + { + "epoch": 2.924650088327218, + "grad_norm": 0.7040160298347473, + "learning_rate": 0.0006344187389590978, + "loss": 3.6997, + "step": 43045 + }, + { + "epoch": 2.9249898083978803, + "grad_norm": 0.8270775675773621, + "learning_rate": 0.000634376273950265, + "loss": 3.7975, + "step": 43050 + }, + { + "epoch": 2.925329528468542, + "grad_norm": 0.9861984252929688, + "learning_rate": 0.0006343338089414322, + "loss": 3.4485, + "step": 43055 + }, + { + "epoch": 2.9256692485392035, + "grad_norm": 0.9797326922416687, + "learning_rate": 0.0006342913439325996, + "loss": 3.5411, + "step": 43060 + }, + { + "epoch": 2.9260089686098656, + "grad_norm": 0.8767942786216736, + "learning_rate": 0.0006342488789237668, + "loss": 3.7338, + "step": 43065 + }, + { + "epoch": 2.926348688680527, + "grad_norm": 0.8385321497917175, + "learning_rate": 0.000634206413914934, + "loss": 3.3743, + "step": 43070 + }, + { + "epoch": 2.926688408751189, + "grad_norm": 0.8651174902915955, + "learning_rate": 0.0006341639489061015, + "loss": 3.569, + "step": 43075 + }, + { + "epoch": 2.927028128821851, + "grad_norm": 1.0775728225708008, + "learning_rate": 0.0006341214838972687, + "loss": 3.6486, + "step": 43080 + }, + { + "epoch": 2.9273678488925126, + "grad_norm": 0.8994227647781372, + "learning_rate": 0.0006340790188884359, + "loss": 3.6417, + "step": 43085 + }, + { + "epoch": 2.927707568963174, + "grad_norm": 0.785743772983551, + "learning_rate": 0.0006340365538796032, + "loss": 3.3074, + "step": 43090 + }, + { + "epoch": 2.9280472890338363, + "grad_norm": 0.9374095797538757, + "learning_rate": 0.0006339940888707705, + "loss": 3.3838, + "step": 43095 + }, + { + "epoch": 2.928387009104498, + "grad_norm": 0.8134713768959045, + "learning_rate": 0.0006339516238619378, + "loss": 3.4064, + "step": 43100 + }, + { + "epoch": 2.9287267291751595, + "grad_norm": 1.0632164478302002, + "learning_rate": 0.000633909158853105, + "loss": 3.4691, + "step": 43105 + }, + { + "epoch": 2.9290664492458216, + "grad_norm": 0.678580641746521, + "learning_rate": 0.0006338666938442724, + "loss": 3.4899, + "step": 43110 + }, + { + "epoch": 2.9294061693164832, + "grad_norm": 0.6749963760375977, + "learning_rate": 0.0006338242288354397, + "loss": 3.6664, + "step": 43115 + }, + { + "epoch": 2.929745889387145, + "grad_norm": 1.077505111694336, + "learning_rate": 0.0006337817638266069, + "loss": 3.2458, + "step": 43120 + }, + { + "epoch": 2.930085609457807, + "grad_norm": 0.745525598526001, + "learning_rate": 0.0006337392988177742, + "loss": 3.6514, + "step": 43125 + }, + { + "epoch": 2.9304253295284686, + "grad_norm": 0.808663010597229, + "learning_rate": 0.0006336968338089415, + "loss": 3.5602, + "step": 43130 + }, + { + "epoch": 2.93076504959913, + "grad_norm": 0.94858318567276, + "learning_rate": 0.0006336543688001087, + "loss": 3.1466, + "step": 43135 + }, + { + "epoch": 2.9311047696697923, + "grad_norm": 0.8936737775802612, + "learning_rate": 0.0006336119037912759, + "loss": 3.527, + "step": 43140 + }, + { + "epoch": 2.931444489740454, + "grad_norm": 0.6836514472961426, + "learning_rate": 0.0006335694387824434, + "loss": 3.7765, + "step": 43145 + }, + { + "epoch": 2.9317842098111155, + "grad_norm": 0.7926419377326965, + "learning_rate": 0.0006335269737736106, + "loss": 3.8441, + "step": 43150 + }, + { + "epoch": 2.9321239298817776, + "grad_norm": 0.9745867252349854, + "learning_rate": 0.0006334845087647778, + "loss": 3.6393, + "step": 43155 + }, + { + "epoch": 2.9324636499524392, + "grad_norm": 0.9688714146614075, + "learning_rate": 0.0006334420437559452, + "loss": 3.5635, + "step": 43160 + }, + { + "epoch": 2.932803370023101, + "grad_norm": 0.7379640936851501, + "learning_rate": 0.0006333995787471124, + "loss": 3.6406, + "step": 43165 + }, + { + "epoch": 2.933143090093763, + "grad_norm": 1.6365584135055542, + "learning_rate": 0.0006333571137382796, + "loss": 3.2963, + "step": 43170 + }, + { + "epoch": 2.9334828101644246, + "grad_norm": 0.9489404559135437, + "learning_rate": 0.000633314648729447, + "loss": 3.5189, + "step": 43175 + }, + { + "epoch": 2.933822530235086, + "grad_norm": 1.082284688949585, + "learning_rate": 0.0006332721837206143, + "loss": 3.6844, + "step": 43180 + }, + { + "epoch": 2.9341622503057483, + "grad_norm": 0.6399793028831482, + "learning_rate": 0.0006332297187117815, + "loss": 3.4025, + "step": 43185 + }, + { + "epoch": 2.93450197037641, + "grad_norm": 0.8884056210517883, + "learning_rate": 0.0006331872537029488, + "loss": 3.5939, + "step": 43190 + }, + { + "epoch": 2.9348416904470715, + "grad_norm": 1.1194565296173096, + "learning_rate": 0.0006331447886941161, + "loss": 3.6128, + "step": 43195 + }, + { + "epoch": 2.9351814105177336, + "grad_norm": 1.199790596961975, + "learning_rate": 0.0006331023236852833, + "loss": 3.6194, + "step": 43200 + }, + { + "epoch": 2.9355211305883953, + "grad_norm": 0.9104600548744202, + "learning_rate": 0.0006330598586764506, + "loss": 3.6906, + "step": 43205 + }, + { + "epoch": 2.935860850659057, + "grad_norm": 0.7115840911865234, + "learning_rate": 0.0006330173936676179, + "loss": 3.6423, + "step": 43210 + }, + { + "epoch": 2.936200570729719, + "grad_norm": 0.7971370816230774, + "learning_rate": 0.0006329749286587852, + "loss": 3.5575, + "step": 43215 + }, + { + "epoch": 2.9365402908003806, + "grad_norm": 0.7459076046943665, + "learning_rate": 0.0006329324636499525, + "loss": 3.4956, + "step": 43220 + }, + { + "epoch": 2.9368800108710422, + "grad_norm": 0.8045793771743774, + "learning_rate": 0.0006328899986411198, + "loss": 3.4865, + "step": 43225 + }, + { + "epoch": 2.9372197309417043, + "grad_norm": 0.8700524568557739, + "learning_rate": 0.000632847533632287, + "loss": 3.6058, + "step": 43230 + }, + { + "epoch": 2.937559451012366, + "grad_norm": 1.0913031101226807, + "learning_rate": 0.0006328050686234543, + "loss": 3.4015, + "step": 43235 + }, + { + "epoch": 2.9378991710830276, + "grad_norm": 0.9350839853286743, + "learning_rate": 0.0006327626036146215, + "loss": 3.5714, + "step": 43240 + }, + { + "epoch": 2.9382388911536896, + "grad_norm": 0.8130491971969604, + "learning_rate": 0.0006327201386057888, + "loss": 3.3993, + "step": 43245 + }, + { + "epoch": 2.9385786112243513, + "grad_norm": 0.7840045690536499, + "learning_rate": 0.0006326776735969562, + "loss": 3.4859, + "step": 43250 + }, + { + "epoch": 2.938918331295013, + "grad_norm": 0.8669129014015198, + "learning_rate": 0.0006326352085881234, + "loss": 3.7256, + "step": 43255 + }, + { + "epoch": 2.9392580513656745, + "grad_norm": 0.9097908735275269, + "learning_rate": 0.0006325927435792907, + "loss": 3.4386, + "step": 43260 + }, + { + "epoch": 2.9395977714363366, + "grad_norm": 0.9442394971847534, + "learning_rate": 0.000632550278570458, + "loss": 3.4204, + "step": 43265 + }, + { + "epoch": 2.9399374915069982, + "grad_norm": 0.7750453352928162, + "learning_rate": 0.0006325078135616252, + "loss": 3.5287, + "step": 43270 + }, + { + "epoch": 2.94027721157766, + "grad_norm": 0.7245613932609558, + "learning_rate": 0.0006324653485527924, + "loss": 3.4228, + "step": 43275 + }, + { + "epoch": 2.940616931648322, + "grad_norm": 0.9449433088302612, + "learning_rate": 0.0006324228835439598, + "loss": 3.6402, + "step": 43280 + }, + { + "epoch": 2.9409566517189836, + "grad_norm": 0.9120093584060669, + "learning_rate": 0.0006323804185351271, + "loss": 3.7134, + "step": 43285 + }, + { + "epoch": 2.941296371789645, + "grad_norm": 0.6628315448760986, + "learning_rate": 0.0006323379535262943, + "loss": 3.5407, + "step": 43290 + }, + { + "epoch": 2.941636091860307, + "grad_norm": 0.729255735874176, + "learning_rate": 0.0006322954885174617, + "loss": 3.7114, + "step": 43295 + }, + { + "epoch": 2.941975811930969, + "grad_norm": 0.9126265645027161, + "learning_rate": 0.0006322530235086289, + "loss": 3.7376, + "step": 43300 + }, + { + "epoch": 2.9423155320016305, + "grad_norm": 0.7074252367019653, + "learning_rate": 0.0006322105584997961, + "loss": 3.329, + "step": 43305 + }, + { + "epoch": 2.942655252072292, + "grad_norm": 0.8255859613418579, + "learning_rate": 0.0006321680934909635, + "loss": 3.5726, + "step": 43310 + }, + { + "epoch": 2.9429949721429542, + "grad_norm": 1.0203677415847778, + "learning_rate": 0.0006321256284821307, + "loss": 3.3668, + "step": 43315 + }, + { + "epoch": 2.943334692213616, + "grad_norm": 1.047762393951416, + "learning_rate": 0.000632083163473298, + "loss": 3.7313, + "step": 43320 + }, + { + "epoch": 2.9436744122842775, + "grad_norm": 0.9265753030776978, + "learning_rate": 0.0006320406984644654, + "loss": 3.3701, + "step": 43325 + }, + { + "epoch": 2.9440141323549396, + "grad_norm": 0.9602901339530945, + "learning_rate": 0.0006319982334556326, + "loss": 3.4059, + "step": 43330 + }, + { + "epoch": 2.944353852425601, + "grad_norm": 0.7473616003990173, + "learning_rate": 0.0006319557684467998, + "loss": 3.5092, + "step": 43335 + }, + { + "epoch": 2.944693572496263, + "grad_norm": 0.6957234740257263, + "learning_rate": 0.0006319133034379671, + "loss": 3.853, + "step": 43340 + }, + { + "epoch": 2.945033292566925, + "grad_norm": 1.0074269771575928, + "learning_rate": 0.0006318708384291344, + "loss": 3.5768, + "step": 43345 + }, + { + "epoch": 2.9453730126375866, + "grad_norm": 0.769325315952301, + "learning_rate": 0.0006318283734203016, + "loss": 3.5487, + "step": 43350 + }, + { + "epoch": 2.945712732708248, + "grad_norm": 0.835279107093811, + "learning_rate": 0.000631785908411469, + "loss": 3.7238, + "step": 43355 + }, + { + "epoch": 2.9460524527789103, + "grad_norm": 0.748982310295105, + "learning_rate": 0.0006317434434026363, + "loss": 3.5139, + "step": 43360 + }, + { + "epoch": 2.946392172849572, + "grad_norm": 0.8109654784202576, + "learning_rate": 0.0006317009783938035, + "loss": 3.4353, + "step": 43365 + }, + { + "epoch": 2.9467318929202335, + "grad_norm": 1.2224628925323486, + "learning_rate": 0.0006316585133849708, + "loss": 3.7055, + "step": 43370 + }, + { + "epoch": 2.9470716129908956, + "grad_norm": 0.8102120161056519, + "learning_rate": 0.000631616048376138, + "loss": 3.8043, + "step": 43375 + }, + { + "epoch": 2.9474113330615572, + "grad_norm": 0.8882004618644714, + "learning_rate": 0.0006315735833673053, + "loss": 3.3616, + "step": 43380 + }, + { + "epoch": 2.947751053132219, + "grad_norm": 0.9587160348892212, + "learning_rate": 0.0006315311183584726, + "loss": 3.4486, + "step": 43385 + }, + { + "epoch": 2.948090773202881, + "grad_norm": 0.8002419471740723, + "learning_rate": 0.0006314886533496399, + "loss": 3.7185, + "step": 43390 + }, + { + "epoch": 2.9484304932735426, + "grad_norm": 0.8356196284294128, + "learning_rate": 0.0006314461883408072, + "loss": 3.555, + "step": 43395 + }, + { + "epoch": 2.948770213344204, + "grad_norm": 0.9570494890213013, + "learning_rate": 0.0006314037233319745, + "loss": 3.5738, + "step": 43400 + }, + { + "epoch": 2.9491099334148663, + "grad_norm": 0.7428749203681946, + "learning_rate": 0.0006313612583231417, + "loss": 3.3779, + "step": 43405 + }, + { + "epoch": 2.949449653485528, + "grad_norm": 1.1080286502838135, + "learning_rate": 0.000631318793314309, + "loss": 3.28, + "step": 43410 + }, + { + "epoch": 2.9497893735561895, + "grad_norm": 0.8242689967155457, + "learning_rate": 0.0006312763283054763, + "loss": 3.5895, + "step": 43415 + }, + { + "epoch": 2.9501290936268516, + "grad_norm": 0.8062947988510132, + "learning_rate": 0.0006312338632966435, + "loss": 3.4197, + "step": 43420 + }, + { + "epoch": 2.9504688136975132, + "grad_norm": 0.7636348009109497, + "learning_rate": 0.0006311913982878108, + "loss": 3.7236, + "step": 43425 + }, + { + "epoch": 2.950808533768175, + "grad_norm": 1.1278858184814453, + "learning_rate": 0.0006311489332789782, + "loss": 3.8628, + "step": 43430 + }, + { + "epoch": 2.951148253838837, + "grad_norm": 1.1522893905639648, + "learning_rate": 0.0006311064682701454, + "loss": 3.7794, + "step": 43435 + }, + { + "epoch": 2.9514879739094986, + "grad_norm": 0.8423879742622375, + "learning_rate": 0.0006310640032613127, + "loss": 3.6291, + "step": 43440 + }, + { + "epoch": 2.95182769398016, + "grad_norm": 0.7841156125068665, + "learning_rate": 0.00063102153825248, + "loss": 3.7412, + "step": 43445 + }, + { + "epoch": 2.9521674140508223, + "grad_norm": 0.9093571305274963, + "learning_rate": 0.0006309790732436472, + "loss": 3.4045, + "step": 43450 + }, + { + "epoch": 2.952507134121484, + "grad_norm": 1.0720208883285522, + "learning_rate": 0.0006309366082348145, + "loss": 3.4602, + "step": 43455 + }, + { + "epoch": 2.9528468541921455, + "grad_norm": 0.7797314524650574, + "learning_rate": 0.0006308941432259819, + "loss": 3.5052, + "step": 43460 + }, + { + "epoch": 2.9531865742628076, + "grad_norm": 0.9134397506713867, + "learning_rate": 0.0006308516782171491, + "loss": 3.524, + "step": 43465 + }, + { + "epoch": 2.9535262943334692, + "grad_norm": 0.9190880656242371, + "learning_rate": 0.0006308092132083164, + "loss": 3.7669, + "step": 43470 + }, + { + "epoch": 2.953866014404131, + "grad_norm": 0.9022619128227234, + "learning_rate": 0.0006307667481994836, + "loss": 3.4313, + "step": 43475 + }, + { + "epoch": 2.954205734474793, + "grad_norm": 1.0890593528747559, + "learning_rate": 0.0006307242831906509, + "loss": 3.6823, + "step": 43480 + }, + { + "epoch": 2.9545454545454546, + "grad_norm": 0.6224093437194824, + "learning_rate": 0.0006306818181818182, + "loss": 3.6398, + "step": 43485 + }, + { + "epoch": 2.954885174616116, + "grad_norm": 0.7774902582168579, + "learning_rate": 0.0006306393531729854, + "loss": 3.7479, + "step": 43490 + }, + { + "epoch": 2.9552248946867783, + "grad_norm": 0.7682737112045288, + "learning_rate": 0.0006305968881641528, + "loss": 3.3711, + "step": 43495 + }, + { + "epoch": 2.95556461475744, + "grad_norm": 0.9445845484733582, + "learning_rate": 0.0006305544231553201, + "loss": 3.5381, + "step": 43500 + }, + { + "epoch": 2.9559043348281016, + "grad_norm": 0.8026590347290039, + "learning_rate": 0.0006305119581464873, + "loss": 3.2857, + "step": 43505 + }, + { + "epoch": 2.9562440548987636, + "grad_norm": 0.8016337156295776, + "learning_rate": 0.0006304694931376546, + "loss": 3.6065, + "step": 43510 + }, + { + "epoch": 2.9565837749694253, + "grad_norm": 0.8848589062690735, + "learning_rate": 0.0006304270281288219, + "loss": 3.2338, + "step": 43515 + }, + { + "epoch": 2.956923495040087, + "grad_norm": 0.6999162435531616, + "learning_rate": 0.0006303845631199891, + "loss": 3.4057, + "step": 43520 + }, + { + "epoch": 2.957263215110749, + "grad_norm": 0.8389745354652405, + "learning_rate": 0.0006303420981111563, + "loss": 3.5429, + "step": 43525 + }, + { + "epoch": 2.9576029351814106, + "grad_norm": 0.8223007917404175, + "learning_rate": 0.0006302996331023238, + "loss": 3.7101, + "step": 43530 + }, + { + "epoch": 2.9579426552520722, + "grad_norm": 1.1538479328155518, + "learning_rate": 0.000630257168093491, + "loss": 3.4019, + "step": 43535 + }, + { + "epoch": 2.9582823753227343, + "grad_norm": 0.8646745085716248, + "learning_rate": 0.0006302147030846582, + "loss": 3.4104, + "step": 43540 + }, + { + "epoch": 2.958622095393396, + "grad_norm": 0.8348680734634399, + "learning_rate": 0.0006301722380758256, + "loss": 3.6123, + "step": 43545 + }, + { + "epoch": 2.9589618154640576, + "grad_norm": 0.8467800617218018, + "learning_rate": 0.0006301297730669928, + "loss": 3.6381, + "step": 43550 + }, + { + "epoch": 2.9593015355347196, + "grad_norm": 0.9674025177955627, + "learning_rate": 0.00063008730805816, + "loss": 3.2129, + "step": 43555 + }, + { + "epoch": 2.9596412556053813, + "grad_norm": 0.8191869258880615, + "learning_rate": 0.0006300448430493274, + "loss": 3.5761, + "step": 43560 + }, + { + "epoch": 2.959980975676043, + "grad_norm": 0.7879136800765991, + "learning_rate": 0.0006300023780404947, + "loss": 3.3012, + "step": 43565 + }, + { + "epoch": 2.960320695746705, + "grad_norm": 0.98897784948349, + "learning_rate": 0.0006299599130316619, + "loss": 3.4116, + "step": 43570 + }, + { + "epoch": 2.9606604158173666, + "grad_norm": 0.9606609344482422, + "learning_rate": 0.0006299174480228293, + "loss": 3.7332, + "step": 43575 + }, + { + "epoch": 2.9610001358880282, + "grad_norm": 0.8893470764160156, + "learning_rate": 0.0006298749830139965, + "loss": 3.6162, + "step": 43580 + }, + { + "epoch": 2.9613398559586903, + "grad_norm": 0.8940730094909668, + "learning_rate": 0.0006298325180051637, + "loss": 3.6049, + "step": 43585 + }, + { + "epoch": 2.961679576029352, + "grad_norm": 0.6304094791412354, + "learning_rate": 0.000629790052996331, + "loss": 3.7876, + "step": 43590 + }, + { + "epoch": 2.9620192961000136, + "grad_norm": 0.8289564251899719, + "learning_rate": 0.0006297475879874983, + "loss": 3.5264, + "step": 43595 + }, + { + "epoch": 2.962359016170675, + "grad_norm": 0.7271533608436584, + "learning_rate": 0.0006297051229786656, + "loss": 3.7018, + "step": 43600 + }, + { + "epoch": 2.9626987362413373, + "grad_norm": 0.6673784852027893, + "learning_rate": 0.0006296626579698329, + "loss": 3.6634, + "step": 43605 + }, + { + "epoch": 2.963038456311999, + "grad_norm": 0.8496295213699341, + "learning_rate": 0.0006296201929610002, + "loss": 3.6089, + "step": 43610 + }, + { + "epoch": 2.9633781763826605, + "grad_norm": 0.9031476974487305, + "learning_rate": 0.0006295777279521674, + "loss": 3.3457, + "step": 43615 + }, + { + "epoch": 2.9637178964533226, + "grad_norm": 0.8204229474067688, + "learning_rate": 0.0006295352629433347, + "loss": 3.6942, + "step": 43620 + }, + { + "epoch": 2.9640576165239843, + "grad_norm": 0.8116554617881775, + "learning_rate": 0.0006294927979345019, + "loss": 3.6558, + "step": 43625 + }, + { + "epoch": 2.964397336594646, + "grad_norm": 1.00860595703125, + "learning_rate": 0.0006294503329256692, + "loss": 3.3386, + "step": 43630 + }, + { + "epoch": 2.9647370566653075, + "grad_norm": 0.9073499441146851, + "learning_rate": 0.0006294078679168366, + "loss": 3.5925, + "step": 43635 + }, + { + "epoch": 2.9650767767359696, + "grad_norm": 0.9822063446044922, + "learning_rate": 0.0006293654029080038, + "loss": 3.5747, + "step": 43640 + }, + { + "epoch": 2.965416496806631, + "grad_norm": 0.7479739785194397, + "learning_rate": 0.0006293229378991711, + "loss": 3.6546, + "step": 43645 + }, + { + "epoch": 2.965756216877293, + "grad_norm": 1.19801926612854, + "learning_rate": 0.0006292804728903384, + "loss": 3.581, + "step": 43650 + }, + { + "epoch": 2.966095936947955, + "grad_norm": 0.680182158946991, + "learning_rate": 0.0006292380078815056, + "loss": 3.7808, + "step": 43655 + }, + { + "epoch": 2.9664356570186166, + "grad_norm": 0.9670102596282959, + "learning_rate": 0.0006291955428726728, + "loss": 3.711, + "step": 43660 + }, + { + "epoch": 2.966775377089278, + "grad_norm": 0.8214612007141113, + "learning_rate": 0.0006291530778638403, + "loss": 3.5466, + "step": 43665 + }, + { + "epoch": 2.9671150971599403, + "grad_norm": 0.775115966796875, + "learning_rate": 0.0006291106128550075, + "loss": 3.7463, + "step": 43670 + }, + { + "epoch": 2.967454817230602, + "grad_norm": 0.7591709494590759, + "learning_rate": 0.0006290681478461747, + "loss": 3.5534, + "step": 43675 + }, + { + "epoch": 2.9677945373012635, + "grad_norm": 0.9011418223381042, + "learning_rate": 0.0006290256828373421, + "loss": 3.6353, + "step": 43680 + }, + { + "epoch": 2.9681342573719256, + "grad_norm": 0.836404025554657, + "learning_rate": 0.0006289832178285093, + "loss": 3.625, + "step": 43685 + }, + { + "epoch": 2.9684739774425872, + "grad_norm": 0.7148934006690979, + "learning_rate": 0.0006289407528196765, + "loss": 3.4055, + "step": 43690 + }, + { + "epoch": 2.968813697513249, + "grad_norm": 1.003218650817871, + "learning_rate": 0.0006288982878108439, + "loss": 3.7591, + "step": 43695 + }, + { + "epoch": 2.969153417583911, + "grad_norm": 0.9518901109695435, + "learning_rate": 0.0006288558228020112, + "loss": 3.5449, + "step": 43700 + }, + { + "epoch": 2.9694931376545726, + "grad_norm": 1.126798391342163, + "learning_rate": 0.0006288133577931784, + "loss": 3.5021, + "step": 43705 + }, + { + "epoch": 2.969832857725234, + "grad_norm": 0.8072933554649353, + "learning_rate": 0.0006287708927843458, + "loss": 3.5632, + "step": 43710 + }, + { + "epoch": 2.9701725777958963, + "grad_norm": 1.0812022686004639, + "learning_rate": 0.000628728427775513, + "loss": 3.3619, + "step": 43715 + }, + { + "epoch": 2.970512297866558, + "grad_norm": 0.7577311396598816, + "learning_rate": 0.0006286859627666802, + "loss": 3.3946, + "step": 43720 + }, + { + "epoch": 2.9708520179372195, + "grad_norm": 0.9825789332389832, + "learning_rate": 0.0006286434977578475, + "loss": 3.4622, + "step": 43725 + }, + { + "epoch": 2.9711917380078816, + "grad_norm": 1.282747507095337, + "learning_rate": 0.0006286010327490148, + "loss": 3.6503, + "step": 43730 + }, + { + "epoch": 2.9715314580785432, + "grad_norm": 0.8374155759811401, + "learning_rate": 0.0006285585677401821, + "loss": 3.545, + "step": 43735 + }, + { + "epoch": 2.971871178149205, + "grad_norm": 0.8747990727424622, + "learning_rate": 0.0006285161027313494, + "loss": 3.4923, + "step": 43740 + }, + { + "epoch": 2.972210898219867, + "grad_norm": 0.8463892936706543, + "learning_rate": 0.0006284736377225167, + "loss": 3.7538, + "step": 43745 + }, + { + "epoch": 2.9725506182905286, + "grad_norm": 0.7155724167823792, + "learning_rate": 0.0006284311727136839, + "loss": 3.6562, + "step": 43750 + }, + { + "epoch": 2.97289033836119, + "grad_norm": 1.040749192237854, + "learning_rate": 0.0006283887077048512, + "loss": 3.4812, + "step": 43755 + }, + { + "epoch": 2.9732300584318523, + "grad_norm": 1.0344443321228027, + "learning_rate": 0.0006283462426960185, + "loss": 3.8361, + "step": 43760 + }, + { + "epoch": 2.973569778502514, + "grad_norm": 1.138869047164917, + "learning_rate": 0.0006283037776871857, + "loss": 3.369, + "step": 43765 + }, + { + "epoch": 2.9739094985731755, + "grad_norm": 0.701641857624054, + "learning_rate": 0.0006282613126783531, + "loss": 3.6449, + "step": 43770 + }, + { + "epoch": 2.9742492186438376, + "grad_norm": 0.9684972763061523, + "learning_rate": 0.0006282188476695203, + "loss": 3.3405, + "step": 43775 + }, + { + "epoch": 2.9745889387144993, + "grad_norm": 0.9837685823440552, + "learning_rate": 0.0006281763826606877, + "loss": 3.679, + "step": 43780 + }, + { + "epoch": 2.974928658785161, + "grad_norm": 0.7664158940315247, + "learning_rate": 0.0006281339176518549, + "loss": 3.5599, + "step": 43785 + }, + { + "epoch": 2.975268378855823, + "grad_norm": 0.7252779603004456, + "learning_rate": 0.0006280914526430221, + "loss": 3.6912, + "step": 43790 + }, + { + "epoch": 2.9756080989264846, + "grad_norm": 1.0869415998458862, + "learning_rate": 0.0006280489876341895, + "loss": 3.5824, + "step": 43795 + }, + { + "epoch": 2.975947818997146, + "grad_norm": 0.7292291522026062, + "learning_rate": 0.0006280065226253567, + "loss": 3.5245, + "step": 43800 + }, + { + "epoch": 2.9762875390678083, + "grad_norm": 0.7518904209136963, + "learning_rate": 0.000627964057616524, + "loss": 3.5599, + "step": 43805 + }, + { + "epoch": 2.97662725913847, + "grad_norm": 0.9415013790130615, + "learning_rate": 0.0006279215926076914, + "loss": 3.4884, + "step": 43810 + }, + { + "epoch": 2.9769669792091316, + "grad_norm": 0.8391397595405579, + "learning_rate": 0.0006278791275988586, + "loss": 3.6852, + "step": 43815 + }, + { + "epoch": 2.9773066992797936, + "grad_norm": 1.0144782066345215, + "learning_rate": 0.0006278366625900258, + "loss": 3.8149, + "step": 43820 + }, + { + "epoch": 2.9776464193504553, + "grad_norm": 0.6430363059043884, + "learning_rate": 0.0006277941975811931, + "loss": 3.506, + "step": 43825 + }, + { + "epoch": 2.977986139421117, + "grad_norm": 0.7965458035469055, + "learning_rate": 0.0006277517325723604, + "loss": 3.4789, + "step": 43830 + }, + { + "epoch": 2.978325859491779, + "grad_norm": 0.9330983757972717, + "learning_rate": 0.0006277092675635276, + "loss": 3.1209, + "step": 43835 + }, + { + "epoch": 2.9786655795624406, + "grad_norm": 0.7246177792549133, + "learning_rate": 0.000627666802554695, + "loss": 3.6196, + "step": 43840 + }, + { + "epoch": 2.9790052996331022, + "grad_norm": 0.7585707902908325, + "learning_rate": 0.0006276243375458623, + "loss": 3.6361, + "step": 43845 + }, + { + "epoch": 2.9793450197037643, + "grad_norm": 0.9207903742790222, + "learning_rate": 0.0006275818725370295, + "loss": 3.6421, + "step": 43850 + }, + { + "epoch": 2.979684739774426, + "grad_norm": 0.6925026774406433, + "learning_rate": 0.0006275394075281968, + "loss": 3.3962, + "step": 43855 + }, + { + "epoch": 2.9800244598450876, + "grad_norm": 1.051146149635315, + "learning_rate": 0.000627496942519364, + "loss": 3.5245, + "step": 43860 + }, + { + "epoch": 2.9803641799157496, + "grad_norm": 0.8130782842636108, + "learning_rate": 0.0006274544775105313, + "loss": 3.4172, + "step": 43865 + }, + { + "epoch": 2.9807038999864113, + "grad_norm": 0.8268676996231079, + "learning_rate": 0.0006274120125016986, + "loss": 3.5554, + "step": 43870 + }, + { + "epoch": 2.981043620057073, + "grad_norm": 0.9722691774368286, + "learning_rate": 0.000627369547492866, + "loss": 3.3905, + "step": 43875 + }, + { + "epoch": 2.981383340127735, + "grad_norm": 0.8878602981567383, + "learning_rate": 0.0006273270824840332, + "loss": 3.7108, + "step": 43880 + }, + { + "epoch": 2.9817230601983966, + "grad_norm": 0.9585939049720764, + "learning_rate": 0.0006272846174752005, + "loss": 3.4167, + "step": 43885 + }, + { + "epoch": 2.9820627802690582, + "grad_norm": 0.9558597207069397, + "learning_rate": 0.0006272421524663677, + "loss": 3.3112, + "step": 43890 + }, + { + "epoch": 2.9824025003397203, + "grad_norm": 0.8682817220687866, + "learning_rate": 0.000627199687457535, + "loss": 3.9436, + "step": 43895 + }, + { + "epoch": 2.982742220410382, + "grad_norm": 1.224189043045044, + "learning_rate": 0.0006271572224487023, + "loss": 3.5899, + "step": 43900 + }, + { + "epoch": 2.9830819404810436, + "grad_norm": 0.8706996440887451, + "learning_rate": 0.0006271147574398695, + "loss": 3.544, + "step": 43905 + }, + { + "epoch": 2.9834216605517057, + "grad_norm": 1.1097068786621094, + "learning_rate": 0.0006270722924310369, + "loss": 3.5458, + "step": 43910 + }, + { + "epoch": 2.9837613806223673, + "grad_norm": 0.7391234636306763, + "learning_rate": 0.0006270298274222042, + "loss": 3.6481, + "step": 43915 + }, + { + "epoch": 2.984101100693029, + "grad_norm": 0.8417186141014099, + "learning_rate": 0.0006269873624133714, + "loss": 3.5111, + "step": 43920 + }, + { + "epoch": 2.984440820763691, + "grad_norm": 0.7816815376281738, + "learning_rate": 0.0006269448974045386, + "loss": 3.582, + "step": 43925 + }, + { + "epoch": 2.9847805408343526, + "grad_norm": 0.7922887802124023, + "learning_rate": 0.000626902432395706, + "loss": 3.3545, + "step": 43930 + }, + { + "epoch": 2.9851202609050143, + "grad_norm": 1.2046717405319214, + "learning_rate": 0.0006268599673868732, + "loss": 3.2082, + "step": 43935 + }, + { + "epoch": 2.985459980975676, + "grad_norm": 0.8437392711639404, + "learning_rate": 0.0006268175023780404, + "loss": 3.4684, + "step": 43940 + }, + { + "epoch": 2.985799701046338, + "grad_norm": 0.8711814880371094, + "learning_rate": 0.0006267750373692079, + "loss": 3.493, + "step": 43945 + }, + { + "epoch": 2.9861394211169996, + "grad_norm": 0.7585463523864746, + "learning_rate": 0.0006267325723603751, + "loss": 3.5958, + "step": 43950 + }, + { + "epoch": 2.9864791411876612, + "grad_norm": 0.9516391158103943, + "learning_rate": 0.0006266901073515423, + "loss": 3.5987, + "step": 43955 + }, + { + "epoch": 2.9868188612583233, + "grad_norm": 1.0616626739501953, + "learning_rate": 0.0006266476423427097, + "loss": 3.734, + "step": 43960 + }, + { + "epoch": 2.987158581328985, + "grad_norm": 0.9060606360435486, + "learning_rate": 0.0006266051773338769, + "loss": 3.5885, + "step": 43965 + }, + { + "epoch": 2.9874983013996466, + "grad_norm": 0.8912849426269531, + "learning_rate": 0.0006265627123250441, + "loss": 3.7553, + "step": 43970 + }, + { + "epoch": 2.987838021470308, + "grad_norm": 0.6663211584091187, + "learning_rate": 0.0006265202473162114, + "loss": 3.4388, + "step": 43975 + }, + { + "epoch": 2.9881777415409703, + "grad_norm": 0.9011244177818298, + "learning_rate": 0.0006264777823073788, + "loss": 3.5546, + "step": 43980 + }, + { + "epoch": 2.988517461611632, + "grad_norm": 0.7722564935684204, + "learning_rate": 0.000626435317298546, + "loss": 3.403, + "step": 43985 + }, + { + "epoch": 2.9888571816822935, + "grad_norm": 1.4766318798065186, + "learning_rate": 0.0006263928522897133, + "loss": 3.7331, + "step": 43990 + }, + { + "epoch": 2.9891969017529556, + "grad_norm": 0.9727010130882263, + "learning_rate": 0.0006263503872808806, + "loss": 3.786, + "step": 43995 + }, + { + "epoch": 2.9895366218236172, + "grad_norm": 0.8336761593818665, + "learning_rate": 0.0006263079222720478, + "loss": 3.4902, + "step": 44000 + }, + { + "epoch": 2.989876341894279, + "grad_norm": 0.9994022846221924, + "learning_rate": 0.0006262654572632151, + "loss": 3.555, + "step": 44005 + }, + { + "epoch": 2.990216061964941, + "grad_norm": 1.0250581502914429, + "learning_rate": 0.0006262229922543823, + "loss": 3.345, + "step": 44010 + }, + { + "epoch": 2.9905557820356026, + "grad_norm": 0.913670003414154, + "learning_rate": 0.0006261805272455497, + "loss": 3.7434, + "step": 44015 + }, + { + "epoch": 2.990895502106264, + "grad_norm": 0.9201777577400208, + "learning_rate": 0.000626138062236717, + "loss": 3.5661, + "step": 44020 + }, + { + "epoch": 2.9912352221769263, + "grad_norm": 0.8307444453239441, + "learning_rate": 0.0006260955972278842, + "loss": 3.528, + "step": 44025 + }, + { + "epoch": 2.991574942247588, + "grad_norm": 0.9045605063438416, + "learning_rate": 0.0006260531322190515, + "loss": 4.1383, + "step": 44030 + }, + { + "epoch": 2.9919146623182495, + "grad_norm": 0.7937968969345093, + "learning_rate": 0.0006260106672102188, + "loss": 3.4655, + "step": 44035 + }, + { + "epoch": 2.9922543823889116, + "grad_norm": 0.7280620336532593, + "learning_rate": 0.000625968202201386, + "loss": 3.4974, + "step": 44040 + }, + { + "epoch": 2.9925941024595732, + "grad_norm": 0.6923540830612183, + "learning_rate": 0.0006259257371925533, + "loss": 3.6847, + "step": 44045 + }, + { + "epoch": 2.992933822530235, + "grad_norm": 0.8300819993019104, + "learning_rate": 0.0006258832721837207, + "loss": 3.6558, + "step": 44050 + }, + { + "epoch": 2.993273542600897, + "grad_norm": 1.0668988227844238, + "learning_rate": 0.0006258408071748879, + "loss": 3.7649, + "step": 44055 + }, + { + "epoch": 2.9936132626715586, + "grad_norm": 0.9185171127319336, + "learning_rate": 0.0006257983421660551, + "loss": 3.6659, + "step": 44060 + }, + { + "epoch": 2.99395298274222, + "grad_norm": 1.0113152265548706, + "learning_rate": 0.0006257558771572225, + "loss": 3.5088, + "step": 44065 + }, + { + "epoch": 2.9942927028128823, + "grad_norm": 1.111271858215332, + "learning_rate": 0.0006257134121483897, + "loss": 3.6718, + "step": 44070 + }, + { + "epoch": 2.994632422883544, + "grad_norm": 0.8769662380218506, + "learning_rate": 0.0006256709471395569, + "loss": 3.5358, + "step": 44075 + }, + { + "epoch": 2.9949721429542056, + "grad_norm": 1.0024797916412354, + "learning_rate": 0.0006256284821307243, + "loss": 3.3745, + "step": 44080 + }, + { + "epoch": 2.9953118630248676, + "grad_norm": 1.2478286027908325, + "learning_rate": 0.0006255860171218916, + "loss": 3.6312, + "step": 44085 + }, + { + "epoch": 2.9956515830955293, + "grad_norm": 0.8599347472190857, + "learning_rate": 0.0006255435521130588, + "loss": 3.4607, + "step": 44090 + }, + { + "epoch": 2.995991303166191, + "grad_norm": 0.7776811122894287, + "learning_rate": 0.0006255010871042262, + "loss": 3.7538, + "step": 44095 + }, + { + "epoch": 2.996331023236853, + "grad_norm": 0.8967260122299194, + "learning_rate": 0.0006254586220953934, + "loss": 3.648, + "step": 44100 + }, + { + "epoch": 2.9966707433075146, + "grad_norm": 0.7885599136352539, + "learning_rate": 0.0006254161570865606, + "loss": 3.754, + "step": 44105 + }, + { + "epoch": 2.9970104633781762, + "grad_norm": 0.8298906683921814, + "learning_rate": 0.000625373692077728, + "loss": 3.5024, + "step": 44110 + }, + { + "epoch": 2.9973501834488383, + "grad_norm": 0.9467365741729736, + "learning_rate": 0.0006253312270688952, + "loss": 3.3715, + "step": 44115 + }, + { + "epoch": 2.9976899035195, + "grad_norm": 0.8111266493797302, + "learning_rate": 0.0006252887620600626, + "loss": 3.686, + "step": 44120 + }, + { + "epoch": 2.9980296235901616, + "grad_norm": 0.7424034476280212, + "learning_rate": 0.0006252462970512298, + "loss": 3.7784, + "step": 44125 + }, + { + "epoch": 2.9983693436608236, + "grad_norm": 1.104880452156067, + "learning_rate": 0.0006252038320423971, + "loss": 3.72, + "step": 44130 + }, + { + "epoch": 2.9987090637314853, + "grad_norm": 0.9103474020957947, + "learning_rate": 0.0006251613670335644, + "loss": 3.657, + "step": 44135 + }, + { + "epoch": 2.999048783802147, + "grad_norm": 0.8004710078239441, + "learning_rate": 0.0006251189020247316, + "loss": 3.2919, + "step": 44140 + }, + { + "epoch": 2.999388503872809, + "grad_norm": 0.8446131348609924, + "learning_rate": 0.0006250764370158989, + "loss": 3.5441, + "step": 44145 + }, + { + "epoch": 2.9997282239434706, + "grad_norm": 0.9181636571884155, + "learning_rate": 0.0006250339720070662, + "loss": 3.5484, + "step": 44150 + }, + { + "epoch": 3.0, + "eval_bertscore": { + "f1": 0.8418314553806632, + "precision": 0.8450208992641071, + "recall": 0.8394055590690574 + }, + "eval_bleu_4": 0.016728719458788406, + "eval_exact_match": 0.00048454307587944567, + "eval_loss": 3.4487621784210205, + "eval_meteor": 0.08368355167368755, + "eval_rouge": { + "rouge1": 0.1103591106566483, + "rouge2": 0.015271913591662111, + "rougeL": 0.09662826870291116, + "rougeLsum": 0.09671776163818707 + }, + "eval_runtime": 2871.8923, + "eval_samples_per_second": 3.593, + "eval_steps_per_second": 0.449, + "step": 44154 + }, + { + "epoch": 3.0000679440141322, + "grad_norm": 0.7587734460830688, + "learning_rate": 0.0006249915069982335, + "loss": 3.4551, + "step": 44155 + }, + { + "epoch": 3.0004076640847943, + "grad_norm": 0.8537425994873047, + "learning_rate": 0.0006249490419894007, + "loss": 3.6393, + "step": 44160 + }, + { + "epoch": 3.000747384155456, + "grad_norm": 0.9793977737426758, + "learning_rate": 0.0006249065769805681, + "loss": 3.5345, + "step": 44165 + }, + { + "epoch": 3.0010871042261176, + "grad_norm": 0.8857905268669128, + "learning_rate": 0.0006248641119717353, + "loss": 3.3382, + "step": 44170 + }, + { + "epoch": 3.0014268242967796, + "grad_norm": 1.116859793663025, + "learning_rate": 0.0006248216469629025, + "loss": 3.5566, + "step": 44175 + }, + { + "epoch": 3.0017665443674413, + "grad_norm": 0.8154410123825073, + "learning_rate": 0.0006247791819540699, + "loss": 3.4416, + "step": 44180 + }, + { + "epoch": 3.002106264438103, + "grad_norm": 0.9563935399055481, + "learning_rate": 0.0006247367169452371, + "loss": 3.3575, + "step": 44185 + }, + { + "epoch": 3.002445984508765, + "grad_norm": 0.7413280606269836, + "learning_rate": 0.0006246942519364044, + "loss": 3.7411, + "step": 44190 + }, + { + "epoch": 3.0027857045794266, + "grad_norm": 0.9423155188560486, + "learning_rate": 0.0006246517869275718, + "loss": 3.2664, + "step": 44195 + }, + { + "epoch": 3.0031254246500882, + "grad_norm": 0.8401862382888794, + "learning_rate": 0.000624609321918739, + "loss": 3.3992, + "step": 44200 + }, + { + "epoch": 3.0034651447207503, + "grad_norm": 0.9264494776725769, + "learning_rate": 0.0006245668569099062, + "loss": 3.5958, + "step": 44205 + }, + { + "epoch": 3.003804864791412, + "grad_norm": 0.7941858172416687, + "learning_rate": 0.0006245243919010735, + "loss": 3.5103, + "step": 44210 + }, + { + "epoch": 3.0041445848620736, + "grad_norm": 0.7986721992492676, + "learning_rate": 0.0006244819268922408, + "loss": 3.5613, + "step": 44215 + }, + { + "epoch": 3.004484304932735, + "grad_norm": 1.05695641040802, + "learning_rate": 0.000624439461883408, + "loss": 3.4487, + "step": 44220 + }, + { + "epoch": 3.0048240250033973, + "grad_norm": 1.0200462341308594, + "learning_rate": 0.0006243969968745754, + "loss": 3.5782, + "step": 44225 + }, + { + "epoch": 3.005163745074059, + "grad_norm": 0.9126034379005432, + "learning_rate": 0.0006243545318657427, + "loss": 3.353, + "step": 44230 + }, + { + "epoch": 3.0055034651447206, + "grad_norm": 1.036638617515564, + "learning_rate": 0.0006243120668569099, + "loss": 3.3516, + "step": 44235 + }, + { + "epoch": 3.0058431852153826, + "grad_norm": 0.8657658696174622, + "learning_rate": 0.0006242696018480772, + "loss": 3.5713, + "step": 44240 + }, + { + "epoch": 3.0061829052860443, + "grad_norm": 0.8542855381965637, + "learning_rate": 0.0006242271368392445, + "loss": 3.6116, + "step": 44245 + }, + { + "epoch": 3.006522625356706, + "grad_norm": 0.9780679941177368, + "learning_rate": 0.0006241846718304117, + "loss": 3.5583, + "step": 44250 + }, + { + "epoch": 3.006862345427368, + "grad_norm": 0.9350326061248779, + "learning_rate": 0.0006241422068215791, + "loss": 3.4318, + "step": 44255 + }, + { + "epoch": 3.0072020654980296, + "grad_norm": 0.910668671131134, + "learning_rate": 0.0006240997418127464, + "loss": 3.8419, + "step": 44260 + }, + { + "epoch": 3.0075417855686912, + "grad_norm": 0.9529253244400024, + "learning_rate": 0.0006240572768039136, + "loss": 3.6944, + "step": 44265 + }, + { + "epoch": 3.0078815056393533, + "grad_norm": 0.9596734642982483, + "learning_rate": 0.0006240148117950809, + "loss": 3.1237, + "step": 44270 + }, + { + "epoch": 3.008221225710015, + "grad_norm": 0.9969460368156433, + "learning_rate": 0.0006239723467862481, + "loss": 3.4173, + "step": 44275 + }, + { + "epoch": 3.0085609457806766, + "grad_norm": 0.7946327924728394, + "learning_rate": 0.0006239298817774154, + "loss": 3.721, + "step": 44280 + }, + { + "epoch": 3.0089006658513386, + "grad_norm": 0.9425920248031616, + "learning_rate": 0.0006238874167685827, + "loss": 3.4195, + "step": 44285 + }, + { + "epoch": 3.0092403859220003, + "grad_norm": 0.9533142447471619, + "learning_rate": 0.00062384495175975, + "loss": 3.4596, + "step": 44290 + }, + { + "epoch": 3.009580105992662, + "grad_norm": 0.7369396686553955, + "learning_rate": 0.0006238024867509173, + "loss": 3.5571, + "step": 44295 + }, + { + "epoch": 3.009919826063324, + "grad_norm": 0.743521511554718, + "learning_rate": 0.0006237600217420846, + "loss": 3.4584, + "step": 44300 + }, + { + "epoch": 3.0102595461339856, + "grad_norm": 0.8170987367630005, + "learning_rate": 0.0006237175567332518, + "loss": 3.4312, + "step": 44305 + }, + { + "epoch": 3.0105992662046472, + "grad_norm": 0.8228139877319336, + "learning_rate": 0.000623675091724419, + "loss": 3.7053, + "step": 44310 + }, + { + "epoch": 3.0109389862753093, + "grad_norm": 1.0465143918991089, + "learning_rate": 0.0006236326267155864, + "loss": 3.5666, + "step": 44315 + }, + { + "epoch": 3.011278706345971, + "grad_norm": 0.7851721048355103, + "learning_rate": 0.0006235901617067536, + "loss": 3.4488, + "step": 44320 + }, + { + "epoch": 3.0116184264166326, + "grad_norm": 1.0019341707229614, + "learning_rate": 0.0006235476966979209, + "loss": 3.5466, + "step": 44325 + }, + { + "epoch": 3.0119581464872947, + "grad_norm": 0.8480458855628967, + "learning_rate": 0.0006235052316890883, + "loss": 3.7905, + "step": 44330 + }, + { + "epoch": 3.0122978665579563, + "grad_norm": 0.7701349258422852, + "learning_rate": 0.0006234627666802555, + "loss": 3.4941, + "step": 44335 + }, + { + "epoch": 3.012637586628618, + "grad_norm": 0.814443051815033, + "learning_rate": 0.0006234203016714227, + "loss": 3.4202, + "step": 44340 + }, + { + "epoch": 3.01297730669928, + "grad_norm": 1.0867106914520264, + "learning_rate": 0.0006233778366625901, + "loss": 3.6937, + "step": 44345 + }, + { + "epoch": 3.0133170267699416, + "grad_norm": 1.02762770652771, + "learning_rate": 0.0006233353716537573, + "loss": 3.5081, + "step": 44350 + }, + { + "epoch": 3.0136567468406033, + "grad_norm": 0.9240368008613586, + "learning_rate": 0.0006232929066449245, + "loss": 3.6061, + "step": 44355 + }, + { + "epoch": 3.0139964669112653, + "grad_norm": 0.7748308181762695, + "learning_rate": 0.000623250441636092, + "loss": 3.5716, + "step": 44360 + }, + { + "epoch": 3.014336186981927, + "grad_norm": 0.8593391180038452, + "learning_rate": 0.0006232079766272592, + "loss": 3.4778, + "step": 44365 + }, + { + "epoch": 3.0146759070525886, + "grad_norm": 0.8019334077835083, + "learning_rate": 0.0006231655116184264, + "loss": 3.532, + "step": 44370 + }, + { + "epoch": 3.01501562712325, + "grad_norm": 0.7535484433174133, + "learning_rate": 0.0006231230466095937, + "loss": 3.6166, + "step": 44375 + }, + { + "epoch": 3.0153553471939123, + "grad_norm": 0.9959530234336853, + "learning_rate": 0.000623080581600761, + "loss": 3.3402, + "step": 44380 + }, + { + "epoch": 3.015695067264574, + "grad_norm": 1.0316100120544434, + "learning_rate": 0.0006230381165919282, + "loss": 3.6328, + "step": 44385 + }, + { + "epoch": 3.0160347873352356, + "grad_norm": 1.038744568824768, + "learning_rate": 0.0006229956515830955, + "loss": 3.5862, + "step": 44390 + }, + { + "epoch": 3.0163745074058976, + "grad_norm": 0.7308980226516724, + "learning_rate": 0.0006229531865742629, + "loss": 3.3996, + "step": 44395 + }, + { + "epoch": 3.0167142274765593, + "grad_norm": 0.9256865978240967, + "learning_rate": 0.0006229107215654301, + "loss": 3.7534, + "step": 44400 + }, + { + "epoch": 3.017053947547221, + "grad_norm": 0.7495373487472534, + "learning_rate": 0.0006228682565565974, + "loss": 3.4057, + "step": 44405 + }, + { + "epoch": 3.017393667617883, + "grad_norm": 0.9175406098365784, + "learning_rate": 0.0006228257915477646, + "loss": 3.4197, + "step": 44410 + }, + { + "epoch": 3.0177333876885446, + "grad_norm": 0.9846746921539307, + "learning_rate": 0.0006227833265389319, + "loss": 3.7763, + "step": 44415 + }, + { + "epoch": 3.0180731077592062, + "grad_norm": 0.9462878108024597, + "learning_rate": 0.0006227408615300992, + "loss": 3.65, + "step": 44420 + }, + { + "epoch": 3.0184128278298683, + "grad_norm": 0.9233075380325317, + "learning_rate": 0.0006226983965212664, + "loss": 3.7797, + "step": 44425 + }, + { + "epoch": 3.01875254790053, + "grad_norm": 0.6709443926811218, + "learning_rate": 0.0006226559315124338, + "loss": 3.6611, + "step": 44430 + }, + { + "epoch": 3.0190922679711916, + "grad_norm": 0.8499719500541687, + "learning_rate": 0.0006226134665036011, + "loss": 3.2718, + "step": 44435 + }, + { + "epoch": 3.0194319880418536, + "grad_norm": 0.7731772661209106, + "learning_rate": 0.0006225710014947683, + "loss": 3.824, + "step": 44440 + }, + { + "epoch": 3.0197717081125153, + "grad_norm": 0.8898109197616577, + "learning_rate": 0.0006225285364859356, + "loss": 3.5272, + "step": 44445 + }, + { + "epoch": 3.020111428183177, + "grad_norm": 0.8469430208206177, + "learning_rate": 0.0006224860714771029, + "loss": 3.5453, + "step": 44450 + }, + { + "epoch": 3.020451148253839, + "grad_norm": 1.038163661956787, + "learning_rate": 0.0006224436064682701, + "loss": 3.4011, + "step": 44455 + }, + { + "epoch": 3.0207908683245006, + "grad_norm": 0.850507915019989, + "learning_rate": 0.0006224011414594373, + "loss": 3.4436, + "step": 44460 + }, + { + "epoch": 3.0211305883951622, + "grad_norm": 1.0557423830032349, + "learning_rate": 0.0006223586764506048, + "loss": 3.5964, + "step": 44465 + }, + { + "epoch": 3.0214703084658243, + "grad_norm": 1.007834792137146, + "learning_rate": 0.000622316211441772, + "loss": 3.374, + "step": 44470 + }, + { + "epoch": 3.021810028536486, + "grad_norm": 0.9377970695495605, + "learning_rate": 0.0006222737464329393, + "loss": 3.5325, + "step": 44475 + }, + { + "epoch": 3.0221497486071476, + "grad_norm": 0.9997595548629761, + "learning_rate": 0.0006222312814241066, + "loss": 3.662, + "step": 44480 + }, + { + "epoch": 3.0224894686778097, + "grad_norm": 0.9916901588439941, + "learning_rate": 0.0006221888164152738, + "loss": 3.7871, + "step": 44485 + }, + { + "epoch": 3.0228291887484713, + "grad_norm": 0.7340381145477295, + "learning_rate": 0.0006221463514064411, + "loss": 3.2203, + "step": 44490 + }, + { + "epoch": 3.023168908819133, + "grad_norm": 0.9175915122032166, + "learning_rate": 0.0006221038863976084, + "loss": 3.5158, + "step": 44495 + }, + { + "epoch": 3.023508628889795, + "grad_norm": 0.8127618432044983, + "learning_rate": 0.0006220614213887757, + "loss": 3.5503, + "step": 44500 + }, + { + "epoch": 3.0238483489604566, + "grad_norm": 1.0173509120941162, + "learning_rate": 0.000622018956379943, + "loss": 3.37, + "step": 44505 + }, + { + "epoch": 3.0241880690311183, + "grad_norm": 0.8375577926635742, + "learning_rate": 0.0006219764913711102, + "loss": 3.5834, + "step": 44510 + }, + { + "epoch": 3.0245277891017803, + "grad_norm": 0.9129244685173035, + "learning_rate": 0.0006219340263622775, + "loss": 3.4842, + "step": 44515 + }, + { + "epoch": 3.024867509172442, + "grad_norm": 0.8662765622138977, + "learning_rate": 0.0006218915613534448, + "loss": 3.4254, + "step": 44520 + }, + { + "epoch": 3.0252072292431036, + "grad_norm": 0.9503839612007141, + "learning_rate": 0.000621849096344612, + "loss": 3.4832, + "step": 44525 + }, + { + "epoch": 3.0255469493137657, + "grad_norm": 1.7057368755340576, + "learning_rate": 0.0006218066313357793, + "loss": 3.6556, + "step": 44530 + }, + { + "epoch": 3.0258866693844273, + "grad_norm": 0.8857652544975281, + "learning_rate": 0.0006217641663269467, + "loss": 3.4592, + "step": 44535 + }, + { + "epoch": 3.026226389455089, + "grad_norm": 0.8442726135253906, + "learning_rate": 0.0006217217013181139, + "loss": 4.0455, + "step": 44540 + }, + { + "epoch": 3.026566109525751, + "grad_norm": 0.8663811087608337, + "learning_rate": 0.0006216792363092812, + "loss": 3.1774, + "step": 44545 + }, + { + "epoch": 3.0269058295964126, + "grad_norm": 1.2360427379608154, + "learning_rate": 0.0006216367713004485, + "loss": 3.5238, + "step": 44550 + }, + { + "epoch": 3.0272455496670743, + "grad_norm": 0.6989754438400269, + "learning_rate": 0.0006215943062916157, + "loss": 3.5719, + "step": 44555 + }, + { + "epoch": 3.027585269737736, + "grad_norm": 1.040585994720459, + "learning_rate": 0.0006215518412827829, + "loss": 3.4666, + "step": 44560 + }, + { + "epoch": 3.027924989808398, + "grad_norm": 0.7921519875526428, + "learning_rate": 0.0006215093762739503, + "loss": 3.6157, + "step": 44565 + }, + { + "epoch": 3.0282647098790596, + "grad_norm": 0.7128247022628784, + "learning_rate": 0.0006214669112651176, + "loss": 3.5776, + "step": 44570 + }, + { + "epoch": 3.0286044299497212, + "grad_norm": 0.9293463230133057, + "learning_rate": 0.0006214244462562848, + "loss": 3.5391, + "step": 44575 + }, + { + "epoch": 3.0289441500203833, + "grad_norm": 1.028183102607727, + "learning_rate": 0.0006213819812474522, + "loss": 3.3579, + "step": 44580 + }, + { + "epoch": 3.029283870091045, + "grad_norm": 1.0128626823425293, + "learning_rate": 0.0006213395162386194, + "loss": 3.5641, + "step": 44585 + }, + { + "epoch": 3.0296235901617066, + "grad_norm": 0.9808063507080078, + "learning_rate": 0.0006212970512297866, + "loss": 3.3048, + "step": 44590 + }, + { + "epoch": 3.0299633102323686, + "grad_norm": 1.132185459136963, + "learning_rate": 0.000621254586220954, + "loss": 3.5066, + "step": 44595 + }, + { + "epoch": 3.0303030303030303, + "grad_norm": 0.8683560490608215, + "learning_rate": 0.0006212121212121212, + "loss": 3.5271, + "step": 44600 + }, + { + "epoch": 3.030642750373692, + "grad_norm": 0.9837058782577515, + "learning_rate": 0.0006211696562032885, + "loss": 3.4809, + "step": 44605 + }, + { + "epoch": 3.030982470444354, + "grad_norm": 0.7621415257453918, + "learning_rate": 0.0006211271911944558, + "loss": 3.445, + "step": 44610 + }, + { + "epoch": 3.0313221905150156, + "grad_norm": 0.7445390820503235, + "learning_rate": 0.0006210847261856231, + "loss": 3.4419, + "step": 44615 + }, + { + "epoch": 3.0316619105856772, + "grad_norm": 0.7593072056770325, + "learning_rate": 0.0006210422611767903, + "loss": 3.3905, + "step": 44620 + }, + { + "epoch": 3.0320016306563393, + "grad_norm": 1.0786997079849243, + "learning_rate": 0.0006209997961679576, + "loss": 3.548, + "step": 44625 + }, + { + "epoch": 3.032341350727001, + "grad_norm": 0.9297082424163818, + "learning_rate": 0.0006209573311591249, + "loss": 3.7538, + "step": 44630 + }, + { + "epoch": 3.0326810707976626, + "grad_norm": 0.9149011969566345, + "learning_rate": 0.0006209148661502921, + "loss": 3.4998, + "step": 44635 + }, + { + "epoch": 3.0330207908683247, + "grad_norm": 0.6716964840888977, + "learning_rate": 0.0006208724011414595, + "loss": 3.2348, + "step": 44640 + }, + { + "epoch": 3.0333605109389863, + "grad_norm": 0.9719027280807495, + "learning_rate": 0.0006208299361326268, + "loss": 3.6899, + "step": 44645 + }, + { + "epoch": 3.033700231009648, + "grad_norm": 0.7154822945594788, + "learning_rate": 0.000620787471123794, + "loss": 3.1278, + "step": 44650 + }, + { + "epoch": 3.03403995108031, + "grad_norm": 0.9694234132766724, + "learning_rate": 0.0006207450061149613, + "loss": 3.5738, + "step": 44655 + }, + { + "epoch": 3.0343796711509716, + "grad_norm": 0.9827795028686523, + "learning_rate": 0.0006207025411061285, + "loss": 3.4599, + "step": 44660 + }, + { + "epoch": 3.0347193912216333, + "grad_norm": 0.8312759399414062, + "learning_rate": 0.0006206600760972958, + "loss": 3.7417, + "step": 44665 + }, + { + "epoch": 3.0350591112922953, + "grad_norm": 1.3861132860183716, + "learning_rate": 0.0006206176110884631, + "loss": 3.2752, + "step": 44670 + }, + { + "epoch": 3.035398831362957, + "grad_norm": 0.7960460186004639, + "learning_rate": 0.0006205751460796304, + "loss": 3.348, + "step": 44675 + }, + { + "epoch": 3.0357385514336186, + "grad_norm": 1.1006197929382324, + "learning_rate": 0.0006205326810707977, + "loss": 3.5713, + "step": 44680 + }, + { + "epoch": 3.0360782715042807, + "grad_norm": 0.8786075115203857, + "learning_rate": 0.000620490216061965, + "loss": 3.6173, + "step": 44685 + }, + { + "epoch": 3.0364179915749423, + "grad_norm": 0.7931172847747803, + "learning_rate": 0.0006204477510531322, + "loss": 3.483, + "step": 44690 + }, + { + "epoch": 3.036757711645604, + "grad_norm": 1.0337727069854736, + "learning_rate": 0.0006204052860442994, + "loss": 3.3107, + "step": 44695 + }, + { + "epoch": 3.037097431716266, + "grad_norm": 0.8015468120574951, + "learning_rate": 0.0006203628210354668, + "loss": 3.5872, + "step": 44700 + }, + { + "epoch": 3.0374371517869276, + "grad_norm": 0.9476364254951477, + "learning_rate": 0.000620320356026634, + "loss": 3.3736, + "step": 44705 + }, + { + "epoch": 3.0377768718575893, + "grad_norm": 0.7108371257781982, + "learning_rate": 0.0006202778910178013, + "loss": 3.7601, + "step": 44710 + }, + { + "epoch": 3.038116591928251, + "grad_norm": 0.9984156489372253, + "learning_rate": 0.0006202354260089687, + "loss": 3.237, + "step": 44715 + }, + { + "epoch": 3.038456311998913, + "grad_norm": 0.7695490121841431, + "learning_rate": 0.0006201929610001359, + "loss": 3.4121, + "step": 44720 + }, + { + "epoch": 3.0387960320695746, + "grad_norm": 0.7655357122421265, + "learning_rate": 0.0006201504959913031, + "loss": 3.4028, + "step": 44725 + }, + { + "epoch": 3.0391357521402362, + "grad_norm": 1.0114009380340576, + "learning_rate": 0.0006201080309824705, + "loss": 3.5339, + "step": 44730 + }, + { + "epoch": 3.0394754722108983, + "grad_norm": 0.9525306820869446, + "learning_rate": 0.0006200655659736377, + "loss": 3.6553, + "step": 44735 + }, + { + "epoch": 3.03981519228156, + "grad_norm": 1.0761363506317139, + "learning_rate": 0.0006200231009648049, + "loss": 3.178, + "step": 44740 + }, + { + "epoch": 3.0401549123522216, + "grad_norm": 0.9626399278640747, + "learning_rate": 0.0006199806359559724, + "loss": 3.6965, + "step": 44745 + }, + { + "epoch": 3.0404946324228836, + "grad_norm": 0.7250663042068481, + "learning_rate": 0.0006199381709471396, + "loss": 3.4678, + "step": 44750 + }, + { + "epoch": 3.0408343524935453, + "grad_norm": 0.7076701521873474, + "learning_rate": 0.0006198957059383068, + "loss": 3.1853, + "step": 44755 + }, + { + "epoch": 3.041174072564207, + "grad_norm": 0.9442560076713562, + "learning_rate": 0.0006198532409294741, + "loss": 3.3991, + "step": 44760 + }, + { + "epoch": 3.041513792634869, + "grad_norm": 0.7688484191894531, + "learning_rate": 0.0006198107759206414, + "loss": 3.3604, + "step": 44765 + }, + { + "epoch": 3.0418535127055306, + "grad_norm": 0.9175731539726257, + "learning_rate": 0.0006197683109118086, + "loss": 3.1811, + "step": 44770 + }, + { + "epoch": 3.0421932327761922, + "grad_norm": 1.1826454401016235, + "learning_rate": 0.000619725845902976, + "loss": 3.5752, + "step": 44775 + }, + { + "epoch": 3.0425329528468543, + "grad_norm": 1.0257331132888794, + "learning_rate": 0.0006196833808941433, + "loss": 3.4285, + "step": 44780 + }, + { + "epoch": 3.042872672917516, + "grad_norm": 0.8988550901412964, + "learning_rate": 0.0006196409158853105, + "loss": 3.4096, + "step": 44785 + }, + { + "epoch": 3.0432123929881776, + "grad_norm": 0.9561294913291931, + "learning_rate": 0.0006195984508764778, + "loss": 3.2367, + "step": 44790 + }, + { + "epoch": 3.0435521130588397, + "grad_norm": 0.9324541091918945, + "learning_rate": 0.000619555985867645, + "loss": 3.4837, + "step": 44795 + }, + { + "epoch": 3.0438918331295013, + "grad_norm": 0.7877375483512878, + "learning_rate": 0.0006195135208588123, + "loss": 3.5689, + "step": 44800 + }, + { + "epoch": 3.044231553200163, + "grad_norm": 1.051604151725769, + "learning_rate": 0.0006194710558499796, + "loss": 3.4706, + "step": 44805 + }, + { + "epoch": 3.044571273270825, + "grad_norm": 0.761008083820343, + "learning_rate": 0.0006194285908411469, + "loss": 3.2802, + "step": 44810 + }, + { + "epoch": 3.0449109933414866, + "grad_norm": 0.7640891075134277, + "learning_rate": 0.0006193861258323143, + "loss": 3.7687, + "step": 44815 + }, + { + "epoch": 3.0452507134121483, + "grad_norm": 1.0195801258087158, + "learning_rate": 0.0006193436608234815, + "loss": 3.5593, + "step": 44820 + }, + { + "epoch": 3.0455904334828103, + "grad_norm": 0.6773487329483032, + "learning_rate": 0.0006193011958146487, + "loss": 3.5398, + "step": 44825 + }, + { + "epoch": 3.045930153553472, + "grad_norm": 0.7669513821601868, + "learning_rate": 0.0006192587308058161, + "loss": 3.5649, + "step": 44830 + }, + { + "epoch": 3.0462698736241336, + "grad_norm": 0.8625152111053467, + "learning_rate": 0.0006192162657969833, + "loss": 3.2235, + "step": 44835 + }, + { + "epoch": 3.0466095936947957, + "grad_norm": 0.8469083905220032, + "learning_rate": 0.0006191738007881505, + "loss": 3.3657, + "step": 44840 + }, + { + "epoch": 3.0469493137654573, + "grad_norm": 0.8547325134277344, + "learning_rate": 0.000619131335779318, + "loss": 3.3513, + "step": 44845 + }, + { + "epoch": 3.047289033836119, + "grad_norm": 0.8656178116798401, + "learning_rate": 0.0006190888707704852, + "loss": 3.6527, + "step": 44850 + }, + { + "epoch": 3.047628753906781, + "grad_norm": 0.7473421692848206, + "learning_rate": 0.0006190464057616524, + "loss": 3.5475, + "step": 44855 + }, + { + "epoch": 3.0479684739774426, + "grad_norm": 0.8366896510124207, + "learning_rate": 0.0006190039407528197, + "loss": 3.3631, + "step": 44860 + }, + { + "epoch": 3.0483081940481043, + "grad_norm": 1.0271514654159546, + "learning_rate": 0.000618961475743987, + "loss": 3.47, + "step": 44865 + }, + { + "epoch": 3.0486479141187663, + "grad_norm": 0.8512331247329712, + "learning_rate": 0.0006189190107351542, + "loss": 3.4168, + "step": 44870 + }, + { + "epoch": 3.048987634189428, + "grad_norm": 0.9701826572418213, + "learning_rate": 0.0006188765457263215, + "loss": 3.4582, + "step": 44875 + }, + { + "epoch": 3.0493273542600896, + "grad_norm": 0.9623094797134399, + "learning_rate": 0.0006188340807174889, + "loss": 3.3999, + "step": 44880 + }, + { + "epoch": 3.0496670743307517, + "grad_norm": 1.053808331489563, + "learning_rate": 0.0006187916157086561, + "loss": 3.5541, + "step": 44885 + }, + { + "epoch": 3.0500067944014133, + "grad_norm": 0.8328718543052673, + "learning_rate": 0.0006187491506998234, + "loss": 3.6691, + "step": 44890 + }, + { + "epoch": 3.050346514472075, + "grad_norm": 0.88230299949646, + "learning_rate": 0.0006187066856909906, + "loss": 3.4575, + "step": 44895 + }, + { + "epoch": 3.0506862345427366, + "grad_norm": 1.1323318481445312, + "learning_rate": 0.0006186642206821579, + "loss": 3.5895, + "step": 44900 + }, + { + "epoch": 3.0510259546133986, + "grad_norm": 0.7941529750823975, + "learning_rate": 0.0006186217556733252, + "loss": 3.2563, + "step": 44905 + }, + { + "epoch": 3.0513656746840603, + "grad_norm": 1.0307403802871704, + "learning_rate": 0.0006185792906644924, + "loss": 3.4624, + "step": 44910 + }, + { + "epoch": 3.051705394754722, + "grad_norm": 0.8485110998153687, + "learning_rate": 0.0006185368256556598, + "loss": 3.7035, + "step": 44915 + }, + { + "epoch": 3.052045114825384, + "grad_norm": 0.7861713171005249, + "learning_rate": 0.0006184943606468271, + "loss": 3.6012, + "step": 44920 + }, + { + "epoch": 3.0523848348960456, + "grad_norm": 0.9029018878936768, + "learning_rate": 0.0006184518956379943, + "loss": 3.5309, + "step": 44925 + }, + { + "epoch": 3.0527245549667072, + "grad_norm": 1.1556962728500366, + "learning_rate": 0.0006184094306291616, + "loss": 3.5723, + "step": 44930 + }, + { + "epoch": 3.0530642750373693, + "grad_norm": 1.5281010866165161, + "learning_rate": 0.0006183669656203289, + "loss": 3.2678, + "step": 44935 + }, + { + "epoch": 3.053403995108031, + "grad_norm": 0.921668529510498, + "learning_rate": 0.0006183245006114961, + "loss": 3.464, + "step": 44940 + }, + { + "epoch": 3.0537437151786926, + "grad_norm": 0.7961602807044983, + "learning_rate": 0.0006182820356026633, + "loss": 3.6467, + "step": 44945 + }, + { + "epoch": 3.0540834352493547, + "grad_norm": 0.8791989684104919, + "learning_rate": 0.0006182395705938308, + "loss": 3.5686, + "step": 44950 + }, + { + "epoch": 3.0544231553200163, + "grad_norm": 0.7829093337059021, + "learning_rate": 0.000618197105584998, + "loss": 3.3888, + "step": 44955 + }, + { + "epoch": 3.054762875390678, + "grad_norm": 0.6417112350463867, + "learning_rate": 0.0006181546405761652, + "loss": 3.5382, + "step": 44960 + }, + { + "epoch": 3.05510259546134, + "grad_norm": 0.9146685600280762, + "learning_rate": 0.0006181121755673326, + "loss": 3.4807, + "step": 44965 + }, + { + "epoch": 3.0554423155320016, + "grad_norm": 0.8432337045669556, + "learning_rate": 0.0006180697105584998, + "loss": 3.3869, + "step": 44970 + }, + { + "epoch": 3.0557820356026633, + "grad_norm": 0.9609442353248596, + "learning_rate": 0.000618027245549667, + "loss": 3.5435, + "step": 44975 + }, + { + "epoch": 3.0561217556733253, + "grad_norm": 0.9016304016113281, + "learning_rate": 0.0006179847805408344, + "loss": 3.5718, + "step": 44980 + }, + { + "epoch": 3.056461475743987, + "grad_norm": 0.7896712422370911, + "learning_rate": 0.0006179423155320017, + "loss": 3.5571, + "step": 44985 + }, + { + "epoch": 3.0568011958146486, + "grad_norm": 0.8888314366340637, + "learning_rate": 0.0006178998505231689, + "loss": 3.5939, + "step": 44990 + }, + { + "epoch": 3.0571409158853107, + "grad_norm": 0.8216575384140015, + "learning_rate": 0.0006178573855143363, + "loss": 3.237, + "step": 44995 + }, + { + "epoch": 3.0574806359559723, + "grad_norm": 0.8404461145401001, + "learning_rate": 0.0006178149205055035, + "loss": 3.5068, + "step": 45000 + }, + { + "epoch": 3.057820356026634, + "grad_norm": 0.8509763479232788, + "learning_rate": 0.0006177724554966707, + "loss": 3.3418, + "step": 45005 + }, + { + "epoch": 3.058160076097296, + "grad_norm": 0.8443201184272766, + "learning_rate": 0.000617729990487838, + "loss": 3.2105, + "step": 45010 + }, + { + "epoch": 3.0584997961679576, + "grad_norm": 0.8886515498161316, + "learning_rate": 0.0006176875254790053, + "loss": 3.6226, + "step": 45015 + }, + { + "epoch": 3.0588395162386193, + "grad_norm": 0.9703799486160278, + "learning_rate": 0.0006176450604701726, + "loss": 3.2671, + "step": 45020 + }, + { + "epoch": 3.0591792363092813, + "grad_norm": 0.7728033661842346, + "learning_rate": 0.0006176025954613399, + "loss": 3.5969, + "step": 45025 + }, + { + "epoch": 3.059518956379943, + "grad_norm": 0.8360174894332886, + "learning_rate": 0.0006175601304525072, + "loss": 3.6248, + "step": 45030 + }, + { + "epoch": 3.0598586764506046, + "grad_norm": 1.0041223764419556, + "learning_rate": 0.0006175176654436744, + "loss": 3.4198, + "step": 45035 + }, + { + "epoch": 3.0601983965212667, + "grad_norm": 0.9259546995162964, + "learning_rate": 0.0006174752004348417, + "loss": 3.6459, + "step": 45040 + }, + { + "epoch": 3.0605381165919283, + "grad_norm": 0.8048959374427795, + "learning_rate": 0.0006174327354260089, + "loss": 3.3272, + "step": 45045 + }, + { + "epoch": 3.06087783666259, + "grad_norm": 0.9225819706916809, + "learning_rate": 0.0006173902704171762, + "loss": 3.4202, + "step": 45050 + }, + { + "epoch": 3.0612175567332516, + "grad_norm": 0.8617943525314331, + "learning_rate": 0.0006173478054083436, + "loss": 3.4648, + "step": 45055 + }, + { + "epoch": 3.0615572768039137, + "grad_norm": 0.7335138916969299, + "learning_rate": 0.0006173053403995108, + "loss": 3.5822, + "step": 45060 + }, + { + "epoch": 3.0618969968745753, + "grad_norm": 1.068841814994812, + "learning_rate": 0.0006172628753906781, + "loss": 3.5074, + "step": 45065 + }, + { + "epoch": 3.062236716945237, + "grad_norm": 1.107946515083313, + "learning_rate": 0.0006172204103818454, + "loss": 3.3981, + "step": 45070 + }, + { + "epoch": 3.062576437015899, + "grad_norm": 0.8038787245750427, + "learning_rate": 0.0006171779453730126, + "loss": 3.8388, + "step": 45075 + }, + { + "epoch": 3.0629161570865606, + "grad_norm": 0.6807218790054321, + "learning_rate": 0.0006171354803641798, + "loss": 3.5894, + "step": 45080 + }, + { + "epoch": 3.0632558771572223, + "grad_norm": 0.8259105682373047, + "learning_rate": 0.0006170930153553472, + "loss": 3.7078, + "step": 45085 + }, + { + "epoch": 3.0635955972278843, + "grad_norm": 0.7785531878471375, + "learning_rate": 0.0006170505503465145, + "loss": 3.4476, + "step": 45090 + }, + { + "epoch": 3.063935317298546, + "grad_norm": 0.7262071967124939, + "learning_rate": 0.0006170080853376817, + "loss": 3.4877, + "step": 45095 + }, + { + "epoch": 3.0642750373692076, + "grad_norm": 0.9480195045471191, + "learning_rate": 0.0006169656203288491, + "loss": 3.4708, + "step": 45100 + }, + { + "epoch": 3.0646147574398697, + "grad_norm": 1.0308184623718262, + "learning_rate": 0.0006169231553200163, + "loss": 3.6492, + "step": 45105 + }, + { + "epoch": 3.0649544775105313, + "grad_norm": 0.7712064981460571, + "learning_rate": 0.0006168806903111835, + "loss": 3.7732, + "step": 45110 + }, + { + "epoch": 3.065294197581193, + "grad_norm": 0.9747133851051331, + "learning_rate": 0.0006168382253023509, + "loss": 3.4881, + "step": 45115 + }, + { + "epoch": 3.065633917651855, + "grad_norm": 0.7935348749160767, + "learning_rate": 0.0006167957602935181, + "loss": 3.5671, + "step": 45120 + }, + { + "epoch": 3.0659736377225166, + "grad_norm": 0.9908781051635742, + "learning_rate": 0.0006167532952846854, + "loss": 3.4903, + "step": 45125 + }, + { + "epoch": 3.0663133577931783, + "grad_norm": 0.9632322192192078, + "learning_rate": 0.0006167108302758528, + "loss": 3.481, + "step": 45130 + }, + { + "epoch": 3.0666530778638403, + "grad_norm": 0.8346583843231201, + "learning_rate": 0.00061666836526702, + "loss": 3.4845, + "step": 45135 + }, + { + "epoch": 3.066992797934502, + "grad_norm": 1.064267635345459, + "learning_rate": 0.0006166259002581872, + "loss": 3.3632, + "step": 45140 + }, + { + "epoch": 3.0673325180051636, + "grad_norm": 0.9516655802726746, + "learning_rate": 0.0006165834352493545, + "loss": 3.3366, + "step": 45145 + }, + { + "epoch": 3.0676722380758257, + "grad_norm": 0.9191951155662537, + "learning_rate": 0.0006165409702405218, + "loss": 3.5985, + "step": 45150 + }, + { + "epoch": 3.0680119581464873, + "grad_norm": 1.1472865343093872, + "learning_rate": 0.0006164985052316891, + "loss": 3.4945, + "step": 45155 + }, + { + "epoch": 3.068351678217149, + "grad_norm": 0.7719565033912659, + "learning_rate": 0.0006164560402228564, + "loss": 3.449, + "step": 45160 + }, + { + "epoch": 3.068691398287811, + "grad_norm": 0.7928420305252075, + "learning_rate": 0.0006164135752140237, + "loss": 3.6857, + "step": 45165 + }, + { + "epoch": 3.0690311183584726, + "grad_norm": 0.8342414498329163, + "learning_rate": 0.000616371110205191, + "loss": 3.531, + "step": 45170 + }, + { + "epoch": 3.0693708384291343, + "grad_norm": 1.037482500076294, + "learning_rate": 0.0006163286451963582, + "loss": 3.4296, + "step": 45175 + }, + { + "epoch": 3.0697105584997963, + "grad_norm": 5.440883159637451, + "learning_rate": 0.0006162861801875255, + "loss": 3.4942, + "step": 45180 + }, + { + "epoch": 3.070050278570458, + "grad_norm": 0.9358147382736206, + "learning_rate": 0.0006162437151786928, + "loss": 3.5077, + "step": 45185 + }, + { + "epoch": 3.0703899986411196, + "grad_norm": 0.8425736427307129, + "learning_rate": 0.00061620125016986, + "loss": 3.6558, + "step": 45190 + }, + { + "epoch": 3.0707297187117817, + "grad_norm": 0.9440665245056152, + "learning_rate": 0.0006161587851610273, + "loss": 3.3705, + "step": 45195 + }, + { + "epoch": 3.0710694387824433, + "grad_norm": 0.8631463050842285, + "learning_rate": 0.0006161163201521947, + "loss": 3.4484, + "step": 45200 + }, + { + "epoch": 3.071409158853105, + "grad_norm": 0.8975459933280945, + "learning_rate": 0.0006160738551433619, + "loss": 3.1227, + "step": 45205 + }, + { + "epoch": 3.071748878923767, + "grad_norm": 1.028167486190796, + "learning_rate": 0.0006160313901345291, + "loss": 3.5941, + "step": 45210 + }, + { + "epoch": 3.0720885989944287, + "grad_norm": 0.9432021379470825, + "learning_rate": 0.0006159889251256965, + "loss": 3.5403, + "step": 45215 + }, + { + "epoch": 3.0724283190650903, + "grad_norm": 0.8812975883483887, + "learning_rate": 0.0006159464601168637, + "loss": 3.529, + "step": 45220 + }, + { + "epoch": 3.0727680391357524, + "grad_norm": 0.9139883518218994, + "learning_rate": 0.0006159039951080309, + "loss": 3.5489, + "step": 45225 + }, + { + "epoch": 3.073107759206414, + "grad_norm": 0.7881546020507812, + "learning_rate": 0.0006158615300991984, + "loss": 3.3667, + "step": 45230 + }, + { + "epoch": 3.0734474792770756, + "grad_norm": 0.9506675004959106, + "learning_rate": 0.0006158190650903656, + "loss": 3.1087, + "step": 45235 + }, + { + "epoch": 3.0737871993477373, + "grad_norm": 0.8200773000717163, + "learning_rate": 0.0006157766000815328, + "loss": 3.5376, + "step": 45240 + }, + { + "epoch": 3.0741269194183993, + "grad_norm": 0.9803846478462219, + "learning_rate": 0.0006157341350727001, + "loss": 3.759, + "step": 45245 + }, + { + "epoch": 3.074466639489061, + "grad_norm": 1.1714228391647339, + "learning_rate": 0.0006156916700638674, + "loss": 3.8305, + "step": 45250 + }, + { + "epoch": 3.0748063595597226, + "grad_norm": 0.7777289748191833, + "learning_rate": 0.0006156492050550346, + "loss": 3.5223, + "step": 45255 + }, + { + "epoch": 3.0751460796303847, + "grad_norm": 1.1189665794372559, + "learning_rate": 0.0006156067400462019, + "loss": 3.2781, + "step": 45260 + }, + { + "epoch": 3.0754857997010463, + "grad_norm": 1.408313512802124, + "learning_rate": 0.0006155642750373693, + "loss": 3.4674, + "step": 45265 + }, + { + "epoch": 3.075825519771708, + "grad_norm": 1.066909670829773, + "learning_rate": 0.0006155218100285365, + "loss": 3.2782, + "step": 45270 + }, + { + "epoch": 3.07616523984237, + "grad_norm": 0.8093449473381042, + "learning_rate": 0.0006154793450197038, + "loss": 3.2382, + "step": 45275 + }, + { + "epoch": 3.0765049599130316, + "grad_norm": 0.7433966398239136, + "learning_rate": 0.000615436880010871, + "loss": 3.346, + "step": 45280 + }, + { + "epoch": 3.0768446799836933, + "grad_norm": 0.7567658424377441, + "learning_rate": 0.0006153944150020383, + "loss": 3.726, + "step": 45285 + }, + { + "epoch": 3.0771844000543553, + "grad_norm": 0.6373672485351562, + "learning_rate": 0.0006153519499932056, + "loss": 3.4846, + "step": 45290 + }, + { + "epoch": 3.077524120125017, + "grad_norm": 1.0138905048370361, + "learning_rate": 0.0006153094849843728, + "loss": 3.5036, + "step": 45295 + }, + { + "epoch": 3.0778638401956786, + "grad_norm": 0.9293628931045532, + "learning_rate": 0.0006152670199755402, + "loss": 3.6881, + "step": 45300 + }, + { + "epoch": 3.0782035602663407, + "grad_norm": 0.6633815765380859, + "learning_rate": 0.0006152245549667075, + "loss": 3.4591, + "step": 45305 + }, + { + "epoch": 3.0785432803370023, + "grad_norm": 1.136802315711975, + "learning_rate": 0.0006151820899578747, + "loss": 3.5295, + "step": 45310 + }, + { + "epoch": 3.078883000407664, + "grad_norm": 0.6393746733665466, + "learning_rate": 0.000615139624949042, + "loss": 3.6466, + "step": 45315 + }, + { + "epoch": 3.079222720478326, + "grad_norm": 0.9364288449287415, + "learning_rate": 0.0006150971599402093, + "loss": 3.4708, + "step": 45320 + }, + { + "epoch": 3.0795624405489876, + "grad_norm": 0.7833111882209778, + "learning_rate": 0.0006150546949313765, + "loss": 3.4949, + "step": 45325 + }, + { + "epoch": 3.0799021606196493, + "grad_norm": 0.864564061164856, + "learning_rate": 0.0006150122299225437, + "loss": 3.516, + "step": 45330 + }, + { + "epoch": 3.0802418806903114, + "grad_norm": 0.9636841416358948, + "learning_rate": 0.0006149697649137112, + "loss": 3.629, + "step": 45335 + }, + { + "epoch": 3.080581600760973, + "grad_norm": 0.8058896660804749, + "learning_rate": 0.0006149272999048784, + "loss": 3.4274, + "step": 45340 + }, + { + "epoch": 3.0809213208316346, + "grad_norm": 0.8056949973106384, + "learning_rate": 0.0006148848348960456, + "loss": 3.7142, + "step": 45345 + }, + { + "epoch": 3.0812610409022967, + "grad_norm": 0.9493510127067566, + "learning_rate": 0.000614842369887213, + "loss": 3.5336, + "step": 45350 + }, + { + "epoch": 3.0816007609729583, + "grad_norm": 1.0686918497085571, + "learning_rate": 0.0006147999048783802, + "loss": 3.4349, + "step": 45355 + }, + { + "epoch": 3.08194048104362, + "grad_norm": 0.9191470742225647, + "learning_rate": 0.0006147574398695474, + "loss": 3.6214, + "step": 45360 + }, + { + "epoch": 3.082280201114282, + "grad_norm": 0.9779872298240662, + "learning_rate": 0.0006147149748607149, + "loss": 3.4851, + "step": 45365 + }, + { + "epoch": 3.0826199211849437, + "grad_norm": 0.9575986266136169, + "learning_rate": 0.0006146725098518821, + "loss": 3.5434, + "step": 45370 + }, + { + "epoch": 3.0829596412556053, + "grad_norm": 0.8877205848693848, + "learning_rate": 0.0006146300448430493, + "loss": 3.0983, + "step": 45375 + }, + { + "epoch": 3.0832993613262674, + "grad_norm": 0.849320113658905, + "learning_rate": 0.0006145875798342167, + "loss": 3.5135, + "step": 45380 + }, + { + "epoch": 3.083639081396929, + "grad_norm": 0.857899010181427, + "learning_rate": 0.0006145451148253839, + "loss": 3.5719, + "step": 45385 + }, + { + "epoch": 3.0839788014675906, + "grad_norm": 1.0104761123657227, + "learning_rate": 0.0006145026498165511, + "loss": 3.5767, + "step": 45390 + }, + { + "epoch": 3.0843185215382523, + "grad_norm": 0.7928483486175537, + "learning_rate": 0.0006144601848077184, + "loss": 3.5407, + "step": 45395 + }, + { + "epoch": 3.0846582416089143, + "grad_norm": 1.0758967399597168, + "learning_rate": 0.0006144177197988858, + "loss": 3.5006, + "step": 45400 + }, + { + "epoch": 3.084997961679576, + "grad_norm": 1.118409276008606, + "learning_rate": 0.000614375254790053, + "loss": 3.4327, + "step": 45405 + }, + { + "epoch": 3.0853376817502376, + "grad_norm": 0.9949268698692322, + "learning_rate": 0.0006143327897812203, + "loss": 3.3565, + "step": 45410 + }, + { + "epoch": 3.0856774018208997, + "grad_norm": 0.956587553024292, + "learning_rate": 0.0006142903247723876, + "loss": 3.3409, + "step": 45415 + }, + { + "epoch": 3.0860171218915613, + "grad_norm": 1.0234997272491455, + "learning_rate": 0.0006142478597635548, + "loss": 3.4243, + "step": 45420 + }, + { + "epoch": 3.086356841962223, + "grad_norm": 1.0361543893814087, + "learning_rate": 0.0006142053947547221, + "loss": 3.5781, + "step": 45425 + }, + { + "epoch": 3.086696562032885, + "grad_norm": 0.8830130696296692, + "learning_rate": 0.0006141629297458893, + "loss": 3.5968, + "step": 45430 + }, + { + "epoch": 3.0870362821035466, + "grad_norm": 0.8168507218360901, + "learning_rate": 0.0006141204647370567, + "loss": 3.6369, + "step": 45435 + }, + { + "epoch": 3.0873760021742083, + "grad_norm": 0.866477906703949, + "learning_rate": 0.000614077999728224, + "loss": 3.4407, + "step": 45440 + }, + { + "epoch": 3.0877157222448703, + "grad_norm": 0.9973773956298828, + "learning_rate": 0.0006140355347193912, + "loss": 3.4483, + "step": 45445 + }, + { + "epoch": 3.088055442315532, + "grad_norm": 0.9263100028038025, + "learning_rate": 0.0006139930697105585, + "loss": 3.8532, + "step": 45450 + }, + { + "epoch": 3.0883951623861936, + "grad_norm": 0.827989935874939, + "learning_rate": 0.0006139506047017258, + "loss": 3.5145, + "step": 45455 + }, + { + "epoch": 3.0887348824568557, + "grad_norm": 0.887234628200531, + "learning_rate": 0.000613908139692893, + "loss": 3.5755, + "step": 45460 + }, + { + "epoch": 3.0890746025275173, + "grad_norm": 0.909027636051178, + "learning_rate": 0.0006138656746840603, + "loss": 3.3482, + "step": 45465 + }, + { + "epoch": 3.089414322598179, + "grad_norm": 0.7998340725898743, + "learning_rate": 0.0006138232096752277, + "loss": 3.3738, + "step": 45470 + }, + { + "epoch": 3.089754042668841, + "grad_norm": 0.8746528625488281, + "learning_rate": 0.0006137807446663949, + "loss": 3.282, + "step": 45475 + }, + { + "epoch": 3.0900937627395026, + "grad_norm": 0.8293395042419434, + "learning_rate": 0.0006137382796575621, + "loss": 3.2905, + "step": 45480 + }, + { + "epoch": 3.0904334828101643, + "grad_norm": 0.7804747223854065, + "learning_rate": 0.0006136958146487295, + "loss": 3.5026, + "step": 45485 + }, + { + "epoch": 3.0907732028808264, + "grad_norm": 0.7489993572235107, + "learning_rate": 0.0006136533496398967, + "loss": 3.6158, + "step": 45490 + }, + { + "epoch": 3.091112922951488, + "grad_norm": 0.9320643544197083, + "learning_rate": 0.000613610884631064, + "loss": 3.5148, + "step": 45495 + }, + { + "epoch": 3.0914526430221496, + "grad_norm": 1.1442352533340454, + "learning_rate": 0.0006135684196222313, + "loss": 3.5386, + "step": 45500 + }, + { + "epoch": 3.0917923630928117, + "grad_norm": 1.1635397672653198, + "learning_rate": 0.0006135259546133986, + "loss": 3.3985, + "step": 45505 + }, + { + "epoch": 3.0921320831634733, + "grad_norm": 0.9988688826560974, + "learning_rate": 0.0006134834896045659, + "loss": 3.6746, + "step": 45510 + }, + { + "epoch": 3.092471803234135, + "grad_norm": 1.1425533294677734, + "learning_rate": 0.0006134410245957332, + "loss": 3.6726, + "step": 45515 + }, + { + "epoch": 3.092811523304797, + "grad_norm": 0.5993607044219971, + "learning_rate": 0.0006133985595869004, + "loss": 3.6583, + "step": 45520 + }, + { + "epoch": 3.0931512433754587, + "grad_norm": 1.0234966278076172, + "learning_rate": 0.0006133560945780677, + "loss": 3.5852, + "step": 45525 + }, + { + "epoch": 3.0934909634461203, + "grad_norm": 0.9165229797363281, + "learning_rate": 0.000613313629569235, + "loss": 3.6828, + "step": 45530 + }, + { + "epoch": 3.0938306835167824, + "grad_norm": 1.1545401811599731, + "learning_rate": 0.0006132711645604022, + "loss": 3.4452, + "step": 45535 + }, + { + "epoch": 3.094170403587444, + "grad_norm": 0.6971798539161682, + "learning_rate": 0.0006132286995515696, + "loss": 3.5657, + "step": 45540 + }, + { + "epoch": 3.0945101236581056, + "grad_norm": 0.8443059325218201, + "learning_rate": 0.0006131862345427368, + "loss": 3.7617, + "step": 45545 + }, + { + "epoch": 3.0948498437287677, + "grad_norm": 0.9233357310295105, + "learning_rate": 0.0006131437695339041, + "loss": 3.5182, + "step": 45550 + }, + { + "epoch": 3.0951895637994293, + "grad_norm": 0.827242374420166, + "learning_rate": 0.0006131013045250714, + "loss": 3.6712, + "step": 45555 + }, + { + "epoch": 3.095529283870091, + "grad_norm": 0.8903420567512512, + "learning_rate": 0.0006130588395162386, + "loss": 3.5544, + "step": 45560 + }, + { + "epoch": 3.095869003940753, + "grad_norm": 0.952860414981842, + "learning_rate": 0.0006130163745074059, + "loss": 3.5172, + "step": 45565 + }, + { + "epoch": 3.0962087240114147, + "grad_norm": 0.9107601046562195, + "learning_rate": 0.0006129739094985732, + "loss": 3.5609, + "step": 45570 + }, + { + "epoch": 3.0965484440820763, + "grad_norm": 0.7514654994010925, + "learning_rate": 0.0006129314444897405, + "loss": 3.5451, + "step": 45575 + }, + { + "epoch": 3.096888164152738, + "grad_norm": 0.8514201641082764, + "learning_rate": 0.0006128889794809077, + "loss": 3.3305, + "step": 45580 + }, + { + "epoch": 3.0972278842234, + "grad_norm": 0.7876724004745483, + "learning_rate": 0.0006128465144720751, + "loss": 3.6559, + "step": 45585 + }, + { + "epoch": 3.0975676042940616, + "grad_norm": 1.1478502750396729, + "learning_rate": 0.0006128040494632423, + "loss": 3.4563, + "step": 45590 + }, + { + "epoch": 3.0979073243647233, + "grad_norm": 0.6941608190536499, + "learning_rate": 0.0006127615844544095, + "loss": 3.3395, + "step": 45595 + }, + { + "epoch": 3.0982470444353853, + "grad_norm": 0.9790254235267639, + "learning_rate": 0.0006127191194455769, + "loss": 3.4416, + "step": 45600 + }, + { + "epoch": 3.098586764506047, + "grad_norm": 0.6391562819480896, + "learning_rate": 0.0006126766544367441, + "loss": 3.394, + "step": 45605 + }, + { + "epoch": 3.0989264845767086, + "grad_norm": 0.7648009657859802, + "learning_rate": 0.0006126341894279114, + "loss": 3.3705, + "step": 45610 + }, + { + "epoch": 3.0992662046473707, + "grad_norm": 0.9162180423736572, + "learning_rate": 0.0006125917244190788, + "loss": 3.2821, + "step": 45615 + }, + { + "epoch": 3.0996059247180323, + "grad_norm": 0.7310871481895447, + "learning_rate": 0.000612549259410246, + "loss": 3.5377, + "step": 45620 + }, + { + "epoch": 3.099945644788694, + "grad_norm": 1.382814884185791, + "learning_rate": 0.0006125067944014132, + "loss": 3.2878, + "step": 45625 + }, + { + "epoch": 3.100285364859356, + "grad_norm": 0.8591187000274658, + "learning_rate": 0.0006124643293925805, + "loss": 3.487, + "step": 45630 + }, + { + "epoch": 3.1006250849300176, + "grad_norm": 1.040737271308899, + "learning_rate": 0.0006124218643837478, + "loss": 3.5998, + "step": 45635 + }, + { + "epoch": 3.1009648050006793, + "grad_norm": 0.8227304220199585, + "learning_rate": 0.000612379399374915, + "loss": 3.2118, + "step": 45640 + }, + { + "epoch": 3.1013045250713414, + "grad_norm": 1.0637708902359009, + "learning_rate": 0.0006123369343660824, + "loss": 3.592, + "step": 45645 + }, + { + "epoch": 3.101644245142003, + "grad_norm": 0.8912635445594788, + "learning_rate": 0.0006122944693572497, + "loss": 3.4011, + "step": 45650 + }, + { + "epoch": 3.1019839652126646, + "grad_norm": 0.8216953873634338, + "learning_rate": 0.0006122520043484169, + "loss": 3.5048, + "step": 45655 + }, + { + "epoch": 3.1023236852833267, + "grad_norm": 1.0509198904037476, + "learning_rate": 0.0006122095393395842, + "loss": 3.7288, + "step": 45660 + }, + { + "epoch": 3.1026634053539883, + "grad_norm": 1.0647850036621094, + "learning_rate": 0.0006121670743307515, + "loss": 3.3626, + "step": 45665 + }, + { + "epoch": 3.10300312542465, + "grad_norm": 1.048405647277832, + "learning_rate": 0.0006121246093219187, + "loss": 3.3651, + "step": 45670 + }, + { + "epoch": 3.103342845495312, + "grad_norm": 0.8326548933982849, + "learning_rate": 0.000612082144313086, + "loss": 3.126, + "step": 45675 + }, + { + "epoch": 3.1036825655659737, + "grad_norm": 1.2499514818191528, + "learning_rate": 0.0006120396793042534, + "loss": 3.6108, + "step": 45680 + }, + { + "epoch": 3.1040222856366353, + "grad_norm": 0.7635319828987122, + "learning_rate": 0.0006119972142954206, + "loss": 3.709, + "step": 45685 + }, + { + "epoch": 3.1043620057072974, + "grad_norm": 1.0022763013839722, + "learning_rate": 0.0006119547492865879, + "loss": 3.4879, + "step": 45690 + }, + { + "epoch": 3.104701725777959, + "grad_norm": 1.2274727821350098, + "learning_rate": 0.0006119122842777551, + "loss": 3.4572, + "step": 45695 + }, + { + "epoch": 3.1050414458486206, + "grad_norm": 1.105048418045044, + "learning_rate": 0.0006118698192689224, + "loss": 3.7056, + "step": 45700 + }, + { + "epoch": 3.1053811659192827, + "grad_norm": 1.2544480562210083, + "learning_rate": 0.0006118273542600897, + "loss": 3.5978, + "step": 45705 + }, + { + "epoch": 3.1057208859899443, + "grad_norm": 0.956779956817627, + "learning_rate": 0.0006117848892512569, + "loss": 3.7558, + "step": 45710 + }, + { + "epoch": 3.106060606060606, + "grad_norm": 1.2315346002578735, + "learning_rate": 0.0006117424242424243, + "loss": 3.3869, + "step": 45715 + }, + { + "epoch": 3.106400326131268, + "grad_norm": 0.772576093673706, + "learning_rate": 0.0006116999592335916, + "loss": 3.4572, + "step": 45720 + }, + { + "epoch": 3.1067400462019297, + "grad_norm": 0.7742746472358704, + "learning_rate": 0.0006116574942247588, + "loss": 3.4549, + "step": 45725 + }, + { + "epoch": 3.1070797662725913, + "grad_norm": 0.7628903985023499, + "learning_rate": 0.000611615029215926, + "loss": 3.3794, + "step": 45730 + }, + { + "epoch": 3.107419486343253, + "grad_norm": 0.9067809581756592, + "learning_rate": 0.0006115725642070934, + "loss": 3.4851, + "step": 45735 + }, + { + "epoch": 3.107759206413915, + "grad_norm": 1.1830251216888428, + "learning_rate": 0.0006115300991982606, + "loss": 3.4939, + "step": 45740 + }, + { + "epoch": 3.1080989264845766, + "grad_norm": 0.9798082709312439, + "learning_rate": 0.0006114876341894278, + "loss": 3.6278, + "step": 45745 + }, + { + "epoch": 3.1084386465552383, + "grad_norm": 1.1009111404418945, + "learning_rate": 0.0006114451691805953, + "loss": 3.665, + "step": 45750 + }, + { + "epoch": 3.1087783666259003, + "grad_norm": 1.0659257173538208, + "learning_rate": 0.0006114027041717625, + "loss": 3.4826, + "step": 45755 + }, + { + "epoch": 3.109118086696562, + "grad_norm": 0.9402558207511902, + "learning_rate": 0.0006113602391629297, + "loss": 3.5599, + "step": 45760 + }, + { + "epoch": 3.1094578067672236, + "grad_norm": 0.9770012497901917, + "learning_rate": 0.0006113177741540971, + "loss": 3.2657, + "step": 45765 + }, + { + "epoch": 3.1097975268378857, + "grad_norm": 1.1731699705123901, + "learning_rate": 0.0006112753091452643, + "loss": 3.4244, + "step": 45770 + }, + { + "epoch": 3.1101372469085473, + "grad_norm": 0.7399146556854248, + "learning_rate": 0.0006112328441364315, + "loss": 3.4765, + "step": 45775 + }, + { + "epoch": 3.110476966979209, + "grad_norm": 0.8070579171180725, + "learning_rate": 0.0006111903791275988, + "loss": 3.6267, + "step": 45780 + }, + { + "epoch": 3.110816687049871, + "grad_norm": 0.8490104675292969, + "learning_rate": 0.0006111479141187662, + "loss": 3.577, + "step": 45785 + }, + { + "epoch": 3.1111564071205327, + "grad_norm": 0.879550039768219, + "learning_rate": 0.0006111054491099334, + "loss": 3.788, + "step": 45790 + }, + { + "epoch": 3.1114961271911943, + "grad_norm": 1.026548981666565, + "learning_rate": 0.0006110629841011007, + "loss": 3.6224, + "step": 45795 + }, + { + "epoch": 3.1118358472618564, + "grad_norm": 0.884977400302887, + "learning_rate": 0.000611020519092268, + "loss": 3.7323, + "step": 45800 + }, + { + "epoch": 3.112175567332518, + "grad_norm": 0.9593492746353149, + "learning_rate": 0.0006109780540834352, + "loss": 3.4578, + "step": 45805 + }, + { + "epoch": 3.1125152874031796, + "grad_norm": 0.7511401176452637, + "learning_rate": 0.0006109355890746025, + "loss": 3.3578, + "step": 45810 + }, + { + "epoch": 3.1128550074738417, + "grad_norm": 0.7133017778396606, + "learning_rate": 0.0006108931240657697, + "loss": 3.4851, + "step": 45815 + }, + { + "epoch": 3.1131947275445033, + "grad_norm": 0.8954891562461853, + "learning_rate": 0.0006108506590569371, + "loss": 3.749, + "step": 45820 + }, + { + "epoch": 3.113534447615165, + "grad_norm": 0.8989901542663574, + "learning_rate": 0.0006108081940481044, + "loss": 3.3813, + "step": 45825 + }, + { + "epoch": 3.113874167685827, + "grad_norm": 0.687868058681488, + "learning_rate": 0.0006107657290392716, + "loss": 3.6056, + "step": 45830 + }, + { + "epoch": 3.1142138877564887, + "grad_norm": 0.8543491363525391, + "learning_rate": 0.000610723264030439, + "loss": 3.5718, + "step": 45835 + }, + { + "epoch": 3.1145536078271503, + "grad_norm": 0.8791754245758057, + "learning_rate": 0.0006106807990216062, + "loss": 3.5718, + "step": 45840 + }, + { + "epoch": 3.1148933278978124, + "grad_norm": 0.822270393371582, + "learning_rate": 0.0006106383340127734, + "loss": 3.5301, + "step": 45845 + }, + { + "epoch": 3.115233047968474, + "grad_norm": 0.8403809070587158, + "learning_rate": 0.0006105958690039408, + "loss": 3.5613, + "step": 45850 + }, + { + "epoch": 3.1155727680391356, + "grad_norm": 1.1512744426727295, + "learning_rate": 0.0006105534039951081, + "loss": 3.466, + "step": 45855 + }, + { + "epoch": 3.1159124881097977, + "grad_norm": 0.8929910063743591, + "learning_rate": 0.0006105109389862753, + "loss": 3.3816, + "step": 45860 + }, + { + "epoch": 3.1162522081804593, + "grad_norm": 0.7213155031204224, + "learning_rate": 0.0006104684739774427, + "loss": 3.5908, + "step": 45865 + }, + { + "epoch": 3.116591928251121, + "grad_norm": 0.9248718619346619, + "learning_rate": 0.0006104260089686099, + "loss": 3.4261, + "step": 45870 + }, + { + "epoch": 3.116931648321783, + "grad_norm": 0.8567730188369751, + "learning_rate": 0.0006103835439597771, + "loss": 3.5589, + "step": 45875 + }, + { + "epoch": 3.1172713683924447, + "grad_norm": 0.7617050409317017, + "learning_rate": 0.0006103410789509444, + "loss": 3.7152, + "step": 45880 + }, + { + "epoch": 3.1176110884631063, + "grad_norm": 0.8798654079437256, + "learning_rate": 0.0006102986139421117, + "loss": 3.477, + "step": 45885 + }, + { + "epoch": 3.1179508085337684, + "grad_norm": 1.1022757291793823, + "learning_rate": 0.000610256148933279, + "loss": 3.5538, + "step": 45890 + }, + { + "epoch": 3.11829052860443, + "grad_norm": 1.0818129777908325, + "learning_rate": 0.0006102136839244463, + "loss": 3.3807, + "step": 45895 + }, + { + "epoch": 3.1186302486750916, + "grad_norm": 1.0450913906097412, + "learning_rate": 0.0006101712189156136, + "loss": 3.7009, + "step": 45900 + }, + { + "epoch": 3.1189699687457537, + "grad_norm": 0.7491384148597717, + "learning_rate": 0.0006101287539067808, + "loss": 3.7027, + "step": 45905 + }, + { + "epoch": 3.1193096888164153, + "grad_norm": 1.2240276336669922, + "learning_rate": 0.0006100862888979481, + "loss": 3.7465, + "step": 45910 + }, + { + "epoch": 3.119649408887077, + "grad_norm": 0.9320614337921143, + "learning_rate": 0.0006100438238891154, + "loss": 3.5809, + "step": 45915 + }, + { + "epoch": 3.1199891289577386, + "grad_norm": 0.833908200263977, + "learning_rate": 0.0006100013588802826, + "loss": 3.4359, + "step": 45920 + }, + { + "epoch": 3.1203288490284007, + "grad_norm": 0.8332535028457642, + "learning_rate": 0.00060995889387145, + "loss": 3.3515, + "step": 45925 + }, + { + "epoch": 3.1206685690990623, + "grad_norm": 0.8754799962043762, + "learning_rate": 0.0006099164288626172, + "loss": 3.5041, + "step": 45930 + }, + { + "epoch": 3.121008289169724, + "grad_norm": 0.9463571310043335, + "learning_rate": 0.0006098739638537845, + "loss": 3.4861, + "step": 45935 + }, + { + "epoch": 3.121348009240386, + "grad_norm": 0.8511512279510498, + "learning_rate": 0.0006098314988449518, + "loss": 3.2937, + "step": 45940 + }, + { + "epoch": 3.1216877293110477, + "grad_norm": 1.0166258811950684, + "learning_rate": 0.000609789033836119, + "loss": 3.5524, + "step": 45945 + }, + { + "epoch": 3.1220274493817093, + "grad_norm": 0.8162824511528015, + "learning_rate": 0.0006097465688272863, + "loss": 3.6029, + "step": 45950 + }, + { + "epoch": 3.1223671694523714, + "grad_norm": 0.8140731453895569, + "learning_rate": 0.0006097041038184537, + "loss": 3.2698, + "step": 45955 + }, + { + "epoch": 3.122706889523033, + "grad_norm": 0.929046630859375, + "learning_rate": 0.0006096616388096209, + "loss": 3.5996, + "step": 45960 + }, + { + "epoch": 3.1230466095936946, + "grad_norm": 0.7766615152359009, + "learning_rate": 0.0006096191738007882, + "loss": 3.2856, + "step": 45965 + }, + { + "epoch": 3.1233863296643567, + "grad_norm": 1.1617099046707153, + "learning_rate": 0.0006095767087919555, + "loss": 3.5837, + "step": 45970 + }, + { + "epoch": 3.1237260497350183, + "grad_norm": 1.1074469089508057, + "learning_rate": 0.0006095342437831227, + "loss": 3.6085, + "step": 45975 + }, + { + "epoch": 3.12406576980568, + "grad_norm": 0.878135621547699, + "learning_rate": 0.0006094917787742899, + "loss": 3.3347, + "step": 45980 + }, + { + "epoch": 3.124405489876342, + "grad_norm": 0.7617248296737671, + "learning_rate": 0.0006094493137654573, + "loss": 3.7015, + "step": 45985 + }, + { + "epoch": 3.1247452099470037, + "grad_norm": 0.7472463250160217, + "learning_rate": 0.0006094068487566246, + "loss": 3.2382, + "step": 45990 + }, + { + "epoch": 3.1250849300176653, + "grad_norm": 0.7606608867645264, + "learning_rate": 0.0006093643837477918, + "loss": 3.5554, + "step": 45995 + }, + { + "epoch": 3.1254246500883274, + "grad_norm": 0.87429279088974, + "learning_rate": 0.0006093219187389592, + "loss": 3.7741, + "step": 46000 + }, + { + "epoch": 3.125764370158989, + "grad_norm": 0.8357203602790833, + "learning_rate": 0.0006092794537301264, + "loss": 3.4156, + "step": 46005 + }, + { + "epoch": 3.1261040902296506, + "grad_norm": 0.8043299317359924, + "learning_rate": 0.0006092369887212936, + "loss": 3.6339, + "step": 46010 + }, + { + "epoch": 3.1264438103003127, + "grad_norm": 0.9077935814857483, + "learning_rate": 0.000609194523712461, + "loss": 3.4129, + "step": 46015 + }, + { + "epoch": 3.1267835303709743, + "grad_norm": 0.9562990069389343, + "learning_rate": 0.0006091520587036282, + "loss": 3.4865, + "step": 46020 + }, + { + "epoch": 3.127123250441636, + "grad_norm": 0.7083468437194824, + "learning_rate": 0.0006091095936947955, + "loss": 3.4868, + "step": 46025 + }, + { + "epoch": 3.127462970512298, + "grad_norm": 0.7912203669548035, + "learning_rate": 0.0006090671286859628, + "loss": 3.5057, + "step": 46030 + }, + { + "epoch": 3.1278026905829597, + "grad_norm": 0.9937058091163635, + "learning_rate": 0.0006090246636771301, + "loss": 3.2897, + "step": 46035 + }, + { + "epoch": 3.1281424106536213, + "grad_norm": 1.2250436544418335, + "learning_rate": 0.0006089821986682973, + "loss": 3.5508, + "step": 46040 + }, + { + "epoch": 3.1284821307242834, + "grad_norm": 0.8452025651931763, + "learning_rate": 0.0006089397336594646, + "loss": 3.3022, + "step": 46045 + }, + { + "epoch": 3.128821850794945, + "grad_norm": 1.0547435283660889, + "learning_rate": 0.0006088972686506319, + "loss": 3.5506, + "step": 46050 + }, + { + "epoch": 3.1291615708656066, + "grad_norm": 5.125038146972656, + "learning_rate": 0.0006088548036417991, + "loss": 3.3906, + "step": 46055 + }, + { + "epoch": 3.1295012909362687, + "grad_norm": 0.8804455399513245, + "learning_rate": 0.0006088123386329665, + "loss": 3.5465, + "step": 46060 + }, + { + "epoch": 3.1298410110069304, + "grad_norm": 1.298341989517212, + "learning_rate": 0.0006087698736241338, + "loss": 3.5442, + "step": 46065 + }, + { + "epoch": 3.130180731077592, + "grad_norm": 0.8561143279075623, + "learning_rate": 0.000608727408615301, + "loss": 3.6698, + "step": 46070 + }, + { + "epoch": 3.1305204511482536, + "grad_norm": 1.1191811561584473, + "learning_rate": 0.0006086849436064683, + "loss": 3.4043, + "step": 46075 + }, + { + "epoch": 3.1308601712189157, + "grad_norm": 1.1392382383346558, + "learning_rate": 0.0006086424785976355, + "loss": 3.5465, + "step": 46080 + }, + { + "epoch": 3.1311998912895773, + "grad_norm": 1.0232793092727661, + "learning_rate": 0.0006086000135888028, + "loss": 3.3369, + "step": 46085 + }, + { + "epoch": 3.131539611360239, + "grad_norm": 0.8181219100952148, + "learning_rate": 0.0006085575485799701, + "loss": 3.7747, + "step": 46090 + }, + { + "epoch": 3.131879331430901, + "grad_norm": 0.9445599317550659, + "learning_rate": 0.0006085150835711374, + "loss": 3.6791, + "step": 46095 + }, + { + "epoch": 3.1322190515015627, + "grad_norm": 0.8215139508247375, + "learning_rate": 0.0006084726185623047, + "loss": 3.6856, + "step": 46100 + }, + { + "epoch": 3.1325587715722243, + "grad_norm": 1.0595476627349854, + "learning_rate": 0.000608430153553472, + "loss": 3.82, + "step": 46105 + }, + { + "epoch": 3.1328984916428864, + "grad_norm": 0.8977946639060974, + "learning_rate": 0.0006083876885446392, + "loss": 3.5782, + "step": 46110 + }, + { + "epoch": 3.133238211713548, + "grad_norm": 1.048440933227539, + "learning_rate": 0.0006083452235358064, + "loss": 3.3939, + "step": 46115 + }, + { + "epoch": 3.1335779317842096, + "grad_norm": 0.8835228681564331, + "learning_rate": 0.0006083027585269738, + "loss": 3.4892, + "step": 46120 + }, + { + "epoch": 3.1339176518548717, + "grad_norm": 0.9006416201591492, + "learning_rate": 0.000608260293518141, + "loss": 3.758, + "step": 46125 + }, + { + "epoch": 3.1342573719255333, + "grad_norm": 0.8280627727508545, + "learning_rate": 0.0006082178285093083, + "loss": 3.1895, + "step": 46130 + }, + { + "epoch": 3.134597091996195, + "grad_norm": 1.0463484525680542, + "learning_rate": 0.0006081753635004757, + "loss": 3.4391, + "step": 46135 + }, + { + "epoch": 3.134936812066857, + "grad_norm": 0.7998212575912476, + "learning_rate": 0.0006081328984916429, + "loss": 3.3238, + "step": 46140 + }, + { + "epoch": 3.1352765321375187, + "grad_norm": 1.1017510890960693, + "learning_rate": 0.0006080904334828101, + "loss": 3.5266, + "step": 46145 + }, + { + "epoch": 3.1356162522081803, + "grad_norm": 1.0030502080917358, + "learning_rate": 0.0006080479684739775, + "loss": 3.5316, + "step": 46150 + }, + { + "epoch": 3.1359559722788424, + "grad_norm": 0.7428054809570312, + "learning_rate": 0.0006080055034651447, + "loss": 3.5298, + "step": 46155 + }, + { + "epoch": 3.136295692349504, + "grad_norm": 0.9752693176269531, + "learning_rate": 0.0006079630384563119, + "loss": 3.4351, + "step": 46160 + }, + { + "epoch": 3.1366354124201656, + "grad_norm": 1.1509345769882202, + "learning_rate": 0.0006079205734474794, + "loss": 3.5448, + "step": 46165 + }, + { + "epoch": 3.1369751324908277, + "grad_norm": 0.7178778648376465, + "learning_rate": 0.0006078781084386466, + "loss": 3.6176, + "step": 46170 + }, + { + "epoch": 3.1373148525614893, + "grad_norm": 0.9408407211303711, + "learning_rate": 0.0006078356434298139, + "loss": 3.6918, + "step": 46175 + }, + { + "epoch": 3.137654572632151, + "grad_norm": 0.8537754416465759, + "learning_rate": 0.0006077931784209811, + "loss": 3.5029, + "step": 46180 + }, + { + "epoch": 3.137994292702813, + "grad_norm": 0.9286121726036072, + "learning_rate": 0.0006077507134121484, + "loss": 3.5665, + "step": 46185 + }, + { + "epoch": 3.1383340127734747, + "grad_norm": 1.01081383228302, + "learning_rate": 0.0006077082484033157, + "loss": 3.6354, + "step": 46190 + }, + { + "epoch": 3.1386737328441363, + "grad_norm": 0.7832593321800232, + "learning_rate": 0.0006076657833944829, + "loss": 3.6438, + "step": 46195 + }, + { + "epoch": 3.1390134529147984, + "grad_norm": 1.0232324600219727, + "learning_rate": 0.0006076233183856503, + "loss": 3.5045, + "step": 46200 + }, + { + "epoch": 3.13935317298546, + "grad_norm": 0.9431591033935547, + "learning_rate": 0.0006075808533768176, + "loss": 3.4027, + "step": 46205 + }, + { + "epoch": 3.1396928930561216, + "grad_norm": 0.930108368396759, + "learning_rate": 0.0006075383883679848, + "loss": 3.4724, + "step": 46210 + }, + { + "epoch": 3.1400326131267837, + "grad_norm": 0.7250892519950867, + "learning_rate": 0.000607495923359152, + "loss": 3.8049, + "step": 46215 + }, + { + "epoch": 3.1403723331974454, + "grad_norm": 0.74888676404953, + "learning_rate": 0.0006074534583503194, + "loss": 3.5611, + "step": 46220 + }, + { + "epoch": 3.140712053268107, + "grad_norm": 0.9732818007469177, + "learning_rate": 0.0006074109933414866, + "loss": 3.4848, + "step": 46225 + }, + { + "epoch": 3.141051773338769, + "grad_norm": 0.9598218202590942, + "learning_rate": 0.0006073685283326538, + "loss": 3.3701, + "step": 46230 + }, + { + "epoch": 3.1413914934094307, + "grad_norm": 1.0637699365615845, + "learning_rate": 0.0006073260633238213, + "loss": 3.544, + "step": 46235 + }, + { + "epoch": 3.1417312134800923, + "grad_norm": 0.7196893692016602, + "learning_rate": 0.0006072835983149885, + "loss": 3.4909, + "step": 46240 + }, + { + "epoch": 3.1420709335507544, + "grad_norm": 1.0989863872528076, + "learning_rate": 0.0006072411333061557, + "loss": 3.411, + "step": 46245 + }, + { + "epoch": 3.142410653621416, + "grad_norm": 0.8415891528129578, + "learning_rate": 0.0006071986682973231, + "loss": 3.5753, + "step": 46250 + }, + { + "epoch": 3.1427503736920777, + "grad_norm": 1.1494345664978027, + "learning_rate": 0.0006071562032884903, + "loss": 3.4642, + "step": 46255 + }, + { + "epoch": 3.1430900937627397, + "grad_norm": 1.1215225458145142, + "learning_rate": 0.0006071137382796575, + "loss": 3.5905, + "step": 46260 + }, + { + "epoch": 3.1434298138334014, + "grad_norm": 1.0338077545166016, + "learning_rate": 0.0006070712732708248, + "loss": 3.5728, + "step": 46265 + }, + { + "epoch": 3.143769533904063, + "grad_norm": 1.2264279127120972, + "learning_rate": 0.0006070288082619922, + "loss": 3.6859, + "step": 46270 + }, + { + "epoch": 3.1441092539747246, + "grad_norm": 0.9312006831169128, + "learning_rate": 0.0006069863432531594, + "loss": 3.6933, + "step": 46275 + }, + { + "epoch": 3.1444489740453867, + "grad_norm": 0.9472574591636658, + "learning_rate": 0.0006069438782443267, + "loss": 3.575, + "step": 46280 + }, + { + "epoch": 3.1447886941160483, + "grad_norm": 0.8425294756889343, + "learning_rate": 0.000606901413235494, + "loss": 3.5546, + "step": 46285 + }, + { + "epoch": 3.14512841418671, + "grad_norm": 0.7385185360908508, + "learning_rate": 0.0006068589482266612, + "loss": 3.6576, + "step": 46290 + }, + { + "epoch": 3.145468134257372, + "grad_norm": 0.9139735698699951, + "learning_rate": 0.0006068164832178285, + "loss": 3.3996, + "step": 46295 + }, + { + "epoch": 3.1458078543280337, + "grad_norm": 0.8780132532119751, + "learning_rate": 0.0006067740182089958, + "loss": 3.7442, + "step": 46300 + }, + { + "epoch": 3.1461475743986953, + "grad_norm": 0.9901491403579712, + "learning_rate": 0.0006067315532001631, + "loss": 3.7015, + "step": 46305 + }, + { + "epoch": 3.1464872944693574, + "grad_norm": 0.9189825654029846, + "learning_rate": 0.0006066890881913304, + "loss": 3.3848, + "step": 46310 + }, + { + "epoch": 3.146827014540019, + "grad_norm": 0.7744531035423279, + "learning_rate": 0.0006066466231824976, + "loss": 3.3679, + "step": 46315 + }, + { + "epoch": 3.1471667346106806, + "grad_norm": 0.9594285488128662, + "learning_rate": 0.0006066041581736649, + "loss": 3.3517, + "step": 46320 + }, + { + "epoch": 3.1475064546813427, + "grad_norm": 1.1310527324676514, + "learning_rate": 0.0006065616931648322, + "loss": 3.7523, + "step": 46325 + }, + { + "epoch": 3.1478461747520043, + "grad_norm": 0.8826537728309631, + "learning_rate": 0.0006065192281559994, + "loss": 3.6548, + "step": 46330 + }, + { + "epoch": 3.148185894822666, + "grad_norm": 0.9914456009864807, + "learning_rate": 0.0006064767631471667, + "loss": 3.5523, + "step": 46335 + }, + { + "epoch": 3.148525614893328, + "grad_norm": 0.9719395637512207, + "learning_rate": 0.0006064342981383341, + "loss": 3.3755, + "step": 46340 + }, + { + "epoch": 3.1488653349639897, + "grad_norm": 1.081804633140564, + "learning_rate": 0.0006063918331295013, + "loss": 3.6586, + "step": 46345 + }, + { + "epoch": 3.1492050550346513, + "grad_norm": 0.7768450975418091, + "learning_rate": 0.0006063493681206686, + "loss": 3.4705, + "step": 46350 + }, + { + "epoch": 3.1495447751053134, + "grad_norm": 0.9053600430488586, + "learning_rate": 0.0006063069031118359, + "loss": 3.3648, + "step": 46355 + }, + { + "epoch": 3.149884495175975, + "grad_norm": 0.9429015517234802, + "learning_rate": 0.0006062644381030031, + "loss": 3.6208, + "step": 46360 + }, + { + "epoch": 3.1502242152466366, + "grad_norm": 0.9059349298477173, + "learning_rate": 0.0006062219730941703, + "loss": 3.725, + "step": 46365 + }, + { + "epoch": 3.1505639353172987, + "grad_norm": 1.1260673999786377, + "learning_rate": 0.0006061795080853377, + "loss": 3.3564, + "step": 46370 + }, + { + "epoch": 3.1509036553879604, + "grad_norm": 0.7665789127349854, + "learning_rate": 0.000606137043076505, + "loss": 3.3842, + "step": 46375 + }, + { + "epoch": 3.151243375458622, + "grad_norm": 0.9315463900566101, + "learning_rate": 0.0006060945780676722, + "loss": 3.4877, + "step": 46380 + }, + { + "epoch": 3.151583095529284, + "grad_norm": 0.8784552812576294, + "learning_rate": 0.0006060521130588396, + "loss": 3.6311, + "step": 46385 + }, + { + "epoch": 3.1519228155999457, + "grad_norm": 0.7249850630760193, + "learning_rate": 0.0006060096480500068, + "loss": 3.4901, + "step": 46390 + }, + { + "epoch": 3.1522625356706073, + "grad_norm": 0.8721307516098022, + "learning_rate": 0.000605967183041174, + "loss": 3.5998, + "step": 46395 + }, + { + "epoch": 3.1526022557412694, + "grad_norm": 0.92719966173172, + "learning_rate": 0.0006059247180323414, + "loss": 3.7339, + "step": 46400 + }, + { + "epoch": 3.152941975811931, + "grad_norm": 1.2453628778457642, + "learning_rate": 0.0006058822530235086, + "loss": 3.4902, + "step": 46405 + }, + { + "epoch": 3.1532816958825927, + "grad_norm": 0.8595046401023865, + "learning_rate": 0.0006058397880146759, + "loss": 3.5458, + "step": 46410 + }, + { + "epoch": 3.1536214159532543, + "grad_norm": 0.7776077389717102, + "learning_rate": 0.0006057973230058433, + "loss": 3.3416, + "step": 46415 + }, + { + "epoch": 3.1539611360239164, + "grad_norm": 0.9832218885421753, + "learning_rate": 0.0006057548579970105, + "loss": 3.5176, + "step": 46420 + }, + { + "epoch": 3.154300856094578, + "grad_norm": 0.9439659118652344, + "learning_rate": 0.0006057123929881777, + "loss": 3.7549, + "step": 46425 + }, + { + "epoch": 3.1546405761652396, + "grad_norm": 0.8922276496887207, + "learning_rate": 0.000605669927979345, + "loss": 3.5491, + "step": 46430 + }, + { + "epoch": 3.1549802962359017, + "grad_norm": 1.005068302154541, + "learning_rate": 0.0006056274629705123, + "loss": 3.4233, + "step": 46435 + }, + { + "epoch": 3.1553200163065633, + "grad_norm": 0.9090473055839539, + "learning_rate": 0.0006055849979616795, + "loss": 3.56, + "step": 46440 + }, + { + "epoch": 3.155659736377225, + "grad_norm": 0.9250444769859314, + "learning_rate": 0.0006055425329528469, + "loss": 3.6162, + "step": 46445 + }, + { + "epoch": 3.155999456447887, + "grad_norm": 0.7810159921646118, + "learning_rate": 0.0006055000679440142, + "loss": 3.7067, + "step": 46450 + }, + { + "epoch": 3.1563391765185487, + "grad_norm": 0.8447842597961426, + "learning_rate": 0.0006054576029351814, + "loss": 3.7874, + "step": 46455 + }, + { + "epoch": 3.1566788965892103, + "grad_norm": 0.935393214225769, + "learning_rate": 0.0006054151379263487, + "loss": 3.4429, + "step": 46460 + }, + { + "epoch": 3.1570186166598724, + "grad_norm": 0.6645131707191467, + "learning_rate": 0.0006053726729175159, + "loss": 3.4582, + "step": 46465 + }, + { + "epoch": 3.157358336730534, + "grad_norm": 1.0213868618011475, + "learning_rate": 0.0006053302079086832, + "loss": 3.6913, + "step": 46470 + }, + { + "epoch": 3.1576980568011956, + "grad_norm": 1.238684058189392, + "learning_rate": 0.0006052877428998506, + "loss": 3.2454, + "step": 46475 + }, + { + "epoch": 3.1580377768718577, + "grad_norm": 0.8169494271278381, + "learning_rate": 0.0006052452778910178, + "loss": 3.3102, + "step": 46480 + }, + { + "epoch": 3.1583774969425193, + "grad_norm": 0.8555706739425659, + "learning_rate": 0.0006052028128821851, + "loss": 3.4806, + "step": 46485 + }, + { + "epoch": 3.158717217013181, + "grad_norm": 0.9451950788497925, + "learning_rate": 0.0006051603478733524, + "loss": 3.4342, + "step": 46490 + }, + { + "epoch": 3.159056937083843, + "grad_norm": 0.9904257655143738, + "learning_rate": 0.0006051178828645196, + "loss": 3.5535, + "step": 46495 + }, + { + "epoch": 3.1593966571545047, + "grad_norm": 0.6905050277709961, + "learning_rate": 0.0006050754178556868, + "loss": 3.041, + "step": 46500 + }, + { + "epoch": 3.1597363772251663, + "grad_norm": 0.9902119040489197, + "learning_rate": 0.0006050329528468542, + "loss": 3.6276, + "step": 46505 + }, + { + "epoch": 3.1600760972958284, + "grad_norm": 1.1795539855957031, + "learning_rate": 0.0006049904878380215, + "loss": 3.6598, + "step": 46510 + }, + { + "epoch": 3.16041581736649, + "grad_norm": 0.9729307293891907, + "learning_rate": 0.0006049480228291889, + "loss": 3.7301, + "step": 46515 + }, + { + "epoch": 3.1607555374371517, + "grad_norm": 0.9324309825897217, + "learning_rate": 0.0006049055578203561, + "loss": 3.6959, + "step": 46520 + }, + { + "epoch": 3.1610952575078137, + "grad_norm": 1.0121296644210815, + "learning_rate": 0.0006048630928115233, + "loss": 3.4641, + "step": 46525 + }, + { + "epoch": 3.1614349775784754, + "grad_norm": 0.9177004098892212, + "learning_rate": 0.0006048206278026906, + "loss": 3.5441, + "step": 46530 + }, + { + "epoch": 3.161774697649137, + "grad_norm": 0.7553184628486633, + "learning_rate": 0.0006047781627938579, + "loss": 3.6605, + "step": 46535 + }, + { + "epoch": 3.162114417719799, + "grad_norm": 1.1297813653945923, + "learning_rate": 0.0006047356977850251, + "loss": 3.4095, + "step": 46540 + }, + { + "epoch": 3.1624541377904607, + "grad_norm": 1.040817141532898, + "learning_rate": 0.0006046932327761925, + "loss": 3.4169, + "step": 46545 + }, + { + "epoch": 3.1627938578611223, + "grad_norm": 0.6253998279571533, + "learning_rate": 0.0006046507677673598, + "loss": 3.6725, + "step": 46550 + }, + { + "epoch": 3.1631335779317844, + "grad_norm": 1.7424874305725098, + "learning_rate": 0.000604608302758527, + "loss": 3.6132, + "step": 46555 + }, + { + "epoch": 3.163473298002446, + "grad_norm": 1.038636565208435, + "learning_rate": 0.0006045658377496943, + "loss": 3.5153, + "step": 46560 + }, + { + "epoch": 3.1638130180731077, + "grad_norm": 0.6953262686729431, + "learning_rate": 0.0006045233727408615, + "loss": 3.6992, + "step": 46565 + }, + { + "epoch": 3.1641527381437697, + "grad_norm": 0.9195459485054016, + "learning_rate": 0.0006044809077320288, + "loss": 3.4486, + "step": 46570 + }, + { + "epoch": 3.1644924582144314, + "grad_norm": 0.9000627994537354, + "learning_rate": 0.0006044384427231961, + "loss": 3.7722, + "step": 46575 + }, + { + "epoch": 3.164832178285093, + "grad_norm": 0.7390528917312622, + "learning_rate": 0.0006043959777143634, + "loss": 3.754, + "step": 46580 + }, + { + "epoch": 3.165171898355755, + "grad_norm": 1.1858088970184326, + "learning_rate": 0.0006043535127055307, + "loss": 3.3887, + "step": 46585 + }, + { + "epoch": 3.1655116184264167, + "grad_norm": 1.1932612657546997, + "learning_rate": 0.000604311047696698, + "loss": 3.302, + "step": 46590 + }, + { + "epoch": 3.1658513384970783, + "grad_norm": 0.8553206324577332, + "learning_rate": 0.0006042685826878652, + "loss": 3.7434, + "step": 46595 + }, + { + "epoch": 3.1661910585677404, + "grad_norm": 0.9183288812637329, + "learning_rate": 0.0006042261176790325, + "loss": 3.5441, + "step": 46600 + }, + { + "epoch": 3.166530778638402, + "grad_norm": 0.9383103251457214, + "learning_rate": 0.0006041836526701998, + "loss": 3.6497, + "step": 46605 + }, + { + "epoch": 3.1668704987090637, + "grad_norm": 0.9877063632011414, + "learning_rate": 0.000604141187661367, + "loss": 3.4506, + "step": 46610 + }, + { + "epoch": 3.1672102187797253, + "grad_norm": 0.9763429760932922, + "learning_rate": 0.0006040987226525343, + "loss": 3.4583, + "step": 46615 + }, + { + "epoch": 3.1675499388503874, + "grad_norm": 0.9898585081100464, + "learning_rate": 0.0006040562576437017, + "loss": 3.6222, + "step": 46620 + }, + { + "epoch": 3.167889658921049, + "grad_norm": 0.8736823797225952, + "learning_rate": 0.0006040137926348689, + "loss": 3.8507, + "step": 46625 + }, + { + "epoch": 3.1682293789917106, + "grad_norm": 0.7745020389556885, + "learning_rate": 0.0006039713276260361, + "loss": 3.2142, + "step": 46630 + }, + { + "epoch": 3.1685690990623727, + "grad_norm": 0.8240602612495422, + "learning_rate": 0.0006039288626172035, + "loss": 3.6573, + "step": 46635 + }, + { + "epoch": 3.1689088191330343, + "grad_norm": 0.756883978843689, + "learning_rate": 0.0006038863976083707, + "loss": 3.5787, + "step": 46640 + }, + { + "epoch": 3.169248539203696, + "grad_norm": 1.2295364141464233, + "learning_rate": 0.0006038439325995379, + "loss": 3.6238, + "step": 46645 + }, + { + "epoch": 3.169588259274358, + "grad_norm": 1.1505990028381348, + "learning_rate": 0.0006038014675907054, + "loss": 3.6274, + "step": 46650 + }, + { + "epoch": 3.1699279793450197, + "grad_norm": 0.9199177026748657, + "learning_rate": 0.0006037590025818726, + "loss": 3.5468, + "step": 46655 + }, + { + "epoch": 3.1702676994156813, + "grad_norm": 0.9085102677345276, + "learning_rate": 0.0006037165375730398, + "loss": 3.6878, + "step": 46660 + }, + { + "epoch": 3.1706074194863434, + "grad_norm": 0.6793073415756226, + "learning_rate": 0.0006036740725642071, + "loss": 3.5002, + "step": 46665 + }, + { + "epoch": 3.170947139557005, + "grad_norm": 0.9024698138237, + "learning_rate": 0.0006036316075553744, + "loss": 3.5565, + "step": 46670 + }, + { + "epoch": 3.1712868596276667, + "grad_norm": 0.9943620562553406, + "learning_rate": 0.0006035891425465416, + "loss": 3.669, + "step": 46675 + }, + { + "epoch": 3.1716265796983287, + "grad_norm": 0.8209461569786072, + "learning_rate": 0.0006035466775377089, + "loss": 3.6526, + "step": 46680 + }, + { + "epoch": 3.1719662997689904, + "grad_norm": 1.0510286092758179, + "learning_rate": 0.0006035042125288763, + "loss": 3.3729, + "step": 46685 + }, + { + "epoch": 3.172306019839652, + "grad_norm": 1.8927001953125, + "learning_rate": 0.0006034617475200435, + "loss": 3.5641, + "step": 46690 + }, + { + "epoch": 3.172645739910314, + "grad_norm": 1.046531081199646, + "learning_rate": 0.0006034192825112108, + "loss": 3.771, + "step": 46695 + }, + { + "epoch": 3.1729854599809757, + "grad_norm": 0.9411276578903198, + "learning_rate": 0.000603376817502378, + "loss": 3.3155, + "step": 46700 + }, + { + "epoch": 3.1733251800516373, + "grad_norm": 0.9345074892044067, + "learning_rate": 0.0006033343524935453, + "loss": 3.4648, + "step": 46705 + }, + { + "epoch": 3.1736649001222994, + "grad_norm": 0.9831835627555847, + "learning_rate": 0.0006032918874847126, + "loss": 3.5108, + "step": 46710 + }, + { + "epoch": 3.174004620192961, + "grad_norm": 1.223319411277771, + "learning_rate": 0.0006032494224758798, + "loss": 3.713, + "step": 46715 + }, + { + "epoch": 3.1743443402636227, + "grad_norm": 0.8318305611610413, + "learning_rate": 0.0006032069574670472, + "loss": 3.5504, + "step": 46720 + }, + { + "epoch": 3.1746840603342847, + "grad_norm": 0.8005041480064392, + "learning_rate": 0.0006031644924582145, + "loss": 3.4345, + "step": 46725 + }, + { + "epoch": 3.1750237804049464, + "grad_norm": 0.7797107100486755, + "learning_rate": 0.0006031220274493817, + "loss": 3.5251, + "step": 46730 + }, + { + "epoch": 3.175363500475608, + "grad_norm": 0.889167845249176, + "learning_rate": 0.000603079562440549, + "loss": 3.47, + "step": 46735 + }, + { + "epoch": 3.17570322054627, + "grad_norm": 0.9055701494216919, + "learning_rate": 0.0006030370974317163, + "loss": 3.5907, + "step": 46740 + }, + { + "epoch": 3.1760429406169317, + "grad_norm": 0.942388653755188, + "learning_rate": 0.0006029946324228835, + "loss": 3.7345, + "step": 46745 + }, + { + "epoch": 3.1763826606875933, + "grad_norm": 0.8008630275726318, + "learning_rate": 0.0006029521674140507, + "loss": 3.4447, + "step": 46750 + }, + { + "epoch": 3.176722380758255, + "grad_norm": 0.7601468563079834, + "learning_rate": 0.0006029097024052182, + "loss": 3.1252, + "step": 46755 + }, + { + "epoch": 3.177062100828917, + "grad_norm": 1.0411310195922852, + "learning_rate": 0.0006028672373963854, + "loss": 3.5155, + "step": 46760 + }, + { + "epoch": 3.1774018208995787, + "grad_norm": 0.7391448616981506, + "learning_rate": 0.0006028247723875526, + "loss": 3.389, + "step": 46765 + }, + { + "epoch": 3.1777415409702403, + "grad_norm": 0.8726449608802795, + "learning_rate": 0.00060278230737872, + "loss": 3.5627, + "step": 46770 + }, + { + "epoch": 3.1780812610409024, + "grad_norm": 1.0630723237991333, + "learning_rate": 0.0006027398423698872, + "loss": 3.3517, + "step": 46775 + }, + { + "epoch": 3.178420981111564, + "grad_norm": 0.9158639311790466, + "learning_rate": 0.0006026973773610544, + "loss": 3.5598, + "step": 46780 + }, + { + "epoch": 3.1787607011822256, + "grad_norm": 0.8446755409240723, + "learning_rate": 0.0006026549123522218, + "loss": 3.4093, + "step": 46785 + }, + { + "epoch": 3.1791004212528877, + "grad_norm": 0.836820125579834, + "learning_rate": 0.0006026124473433891, + "loss": 3.814, + "step": 46790 + }, + { + "epoch": 3.1794401413235494, + "grad_norm": 0.769568145275116, + "learning_rate": 0.0006025699823345563, + "loss": 3.6058, + "step": 46795 + }, + { + "epoch": 3.179779861394211, + "grad_norm": 1.3094507455825806, + "learning_rate": 0.0006025275173257237, + "loss": 3.3002, + "step": 46800 + }, + { + "epoch": 3.180119581464873, + "grad_norm": 0.7064046859741211, + "learning_rate": 0.0006024850523168909, + "loss": 3.352, + "step": 46805 + }, + { + "epoch": 3.1804593015355347, + "grad_norm": 1.1609294414520264, + "learning_rate": 0.0006024425873080581, + "loss": 3.5385, + "step": 46810 + }, + { + "epoch": 3.1807990216061963, + "grad_norm": 0.8500900268554688, + "learning_rate": 0.0006024001222992254, + "loss": 3.4063, + "step": 46815 + }, + { + "epoch": 3.1811387416768584, + "grad_norm": 0.8350246548652649, + "learning_rate": 0.0006023576572903927, + "loss": 3.4049, + "step": 46820 + }, + { + "epoch": 3.18147846174752, + "grad_norm": 0.7162759900093079, + "learning_rate": 0.00060231519228156, + "loss": 3.5785, + "step": 46825 + }, + { + "epoch": 3.1818181818181817, + "grad_norm": 0.7550603151321411, + "learning_rate": 0.0006022727272727273, + "loss": 3.5449, + "step": 46830 + }, + { + "epoch": 3.1821579018888437, + "grad_norm": 0.8434705138206482, + "learning_rate": 0.0006022302622638946, + "loss": 3.5738, + "step": 46835 + }, + { + "epoch": 3.1824976219595054, + "grad_norm": 1.0039403438568115, + "learning_rate": 0.0006021877972550618, + "loss": 3.3661, + "step": 46840 + }, + { + "epoch": 3.182837342030167, + "grad_norm": 0.8391554951667786, + "learning_rate": 0.0006021453322462291, + "loss": 3.5292, + "step": 46845 + }, + { + "epoch": 3.183177062100829, + "grad_norm": 0.9582727551460266, + "learning_rate": 0.0006021028672373963, + "loss": 3.5577, + "step": 46850 + }, + { + "epoch": 3.1835167821714907, + "grad_norm": 0.8422686457633972, + "learning_rate": 0.0006020604022285637, + "loss": 3.5484, + "step": 46855 + }, + { + "epoch": 3.1838565022421523, + "grad_norm": 1.2003206014633179, + "learning_rate": 0.000602017937219731, + "loss": 3.6119, + "step": 46860 + }, + { + "epoch": 3.1841962223128144, + "grad_norm": 1.0818209648132324, + "learning_rate": 0.0006019754722108982, + "loss": 3.4672, + "step": 46865 + }, + { + "epoch": 3.184535942383476, + "grad_norm": 0.8151135444641113, + "learning_rate": 0.0006019330072020656, + "loss": 3.676, + "step": 46870 + }, + { + "epoch": 3.1848756624541377, + "grad_norm": 0.9601738452911377, + "learning_rate": 0.0006018905421932328, + "loss": 3.6208, + "step": 46875 + }, + { + "epoch": 3.1852153825247997, + "grad_norm": 0.7576903700828552, + "learning_rate": 0.0006018480771844, + "loss": 3.5858, + "step": 46880 + }, + { + "epoch": 3.1855551025954614, + "grad_norm": 0.8233585953712463, + "learning_rate": 0.0006018056121755674, + "loss": 3.748, + "step": 46885 + }, + { + "epoch": 3.185894822666123, + "grad_norm": 0.9840953946113586, + "learning_rate": 0.0006017631471667346, + "loss": 3.488, + "step": 46890 + }, + { + "epoch": 3.186234542736785, + "grad_norm": 1.1098753213882446, + "learning_rate": 0.0006017206821579019, + "loss": 3.4388, + "step": 46895 + }, + { + "epoch": 3.1865742628074467, + "grad_norm": 1.1684784889221191, + "learning_rate": 0.0006016782171490693, + "loss": 3.6886, + "step": 46900 + }, + { + "epoch": 3.1869139828781083, + "grad_norm": 0.8859899044036865, + "learning_rate": 0.0006016357521402365, + "loss": 3.3408, + "step": 46905 + }, + { + "epoch": 3.1872537029487704, + "grad_norm": 0.8561869859695435, + "learning_rate": 0.0006015932871314037, + "loss": 3.6252, + "step": 46910 + }, + { + "epoch": 3.187593423019432, + "grad_norm": 0.8533803224563599, + "learning_rate": 0.000601550822122571, + "loss": 3.4019, + "step": 46915 + }, + { + "epoch": 3.1879331430900937, + "grad_norm": 1.049674153327942, + "learning_rate": 0.0006015083571137383, + "loss": 3.6391, + "step": 46920 + }, + { + "epoch": 3.1882728631607558, + "grad_norm": 0.8303041458129883, + "learning_rate": 0.0006014658921049055, + "loss": 3.6925, + "step": 46925 + }, + { + "epoch": 3.1886125832314174, + "grad_norm": 0.8091197609901428, + "learning_rate": 0.0006014234270960729, + "loss": 3.5285, + "step": 46930 + }, + { + "epoch": 3.188952303302079, + "grad_norm": 0.9443579316139221, + "learning_rate": 0.0006013809620872402, + "loss": 3.5063, + "step": 46935 + }, + { + "epoch": 3.189292023372741, + "grad_norm": 1.035934329032898, + "learning_rate": 0.0006013384970784074, + "loss": 3.6002, + "step": 46940 + }, + { + "epoch": 3.1896317434434027, + "grad_norm": 0.6452840566635132, + "learning_rate": 0.0006012960320695747, + "loss": 3.4307, + "step": 46945 + }, + { + "epoch": 3.1899714635140644, + "grad_norm": 0.7430636286735535, + "learning_rate": 0.000601253567060742, + "loss": 3.5211, + "step": 46950 + }, + { + "epoch": 3.190311183584726, + "grad_norm": 1.0702970027923584, + "learning_rate": 0.0006012111020519092, + "loss": 3.3982, + "step": 46955 + }, + { + "epoch": 3.190650903655388, + "grad_norm": 0.9080306887626648, + "learning_rate": 0.0006011686370430765, + "loss": 3.324, + "step": 46960 + }, + { + "epoch": 3.1909906237260497, + "grad_norm": 1.035197138786316, + "learning_rate": 0.0006011261720342438, + "loss": 3.4461, + "step": 46965 + }, + { + "epoch": 3.1913303437967113, + "grad_norm": 1.2540777921676636, + "learning_rate": 0.0006010837070254111, + "loss": 3.3924, + "step": 46970 + }, + { + "epoch": 3.1916700638673734, + "grad_norm": 0.9723957777023315, + "learning_rate": 0.0006010412420165784, + "loss": 3.576, + "step": 46975 + }, + { + "epoch": 3.192009783938035, + "grad_norm": 1.0530803203582764, + "learning_rate": 0.0006009987770077456, + "loss": 3.4289, + "step": 46980 + }, + { + "epoch": 3.1923495040086967, + "grad_norm": 0.9055414795875549, + "learning_rate": 0.0006009563119989129, + "loss": 3.4464, + "step": 46985 + }, + { + "epoch": 3.1926892240793587, + "grad_norm": 0.9550784230232239, + "learning_rate": 0.0006009138469900802, + "loss": 3.5365, + "step": 46990 + }, + { + "epoch": 3.1930289441500204, + "grad_norm": 0.7817603349685669, + "learning_rate": 0.0006008713819812474, + "loss": 3.4384, + "step": 46995 + }, + { + "epoch": 3.193368664220682, + "grad_norm": 0.8422337174415588, + "learning_rate": 0.0006008289169724147, + "loss": 3.4229, + "step": 47000 + }, + { + "epoch": 3.193708384291344, + "grad_norm": 0.7332444190979004, + "learning_rate": 0.0006007864519635821, + "loss": 3.7348, + "step": 47005 + }, + { + "epoch": 3.1940481043620057, + "grad_norm": 0.8244227170944214, + "learning_rate": 0.0006007439869547493, + "loss": 3.3134, + "step": 47010 + }, + { + "epoch": 3.1943878244326673, + "grad_norm": 0.9194483757019043, + "learning_rate": 0.0006007015219459165, + "loss": 3.5259, + "step": 47015 + }, + { + "epoch": 3.1947275445033294, + "grad_norm": 0.820408821105957, + "learning_rate": 0.0006006590569370839, + "loss": 3.5827, + "step": 47020 + }, + { + "epoch": 3.195067264573991, + "grad_norm": 0.8288946151733398, + "learning_rate": 0.0006006165919282511, + "loss": 3.6084, + "step": 47025 + }, + { + "epoch": 3.1954069846446527, + "grad_norm": 0.8105474710464478, + "learning_rate": 0.0006005741269194183, + "loss": 3.4301, + "step": 47030 + }, + { + "epoch": 3.1957467047153147, + "grad_norm": 0.7377923727035522, + "learning_rate": 0.0006005316619105858, + "loss": 3.4609, + "step": 47035 + }, + { + "epoch": 3.1960864247859764, + "grad_norm": 1.0504072904586792, + "learning_rate": 0.000600489196901753, + "loss": 3.5594, + "step": 47040 + }, + { + "epoch": 3.196426144856638, + "grad_norm": 0.7716879844665527, + "learning_rate": 0.0006004467318929202, + "loss": 3.2592, + "step": 47045 + }, + { + "epoch": 3.1967658649273, + "grad_norm": 1.0174421072006226, + "learning_rate": 0.0006004042668840875, + "loss": 3.3935, + "step": 47050 + }, + { + "epoch": 3.1971055849979617, + "grad_norm": 0.7798956632614136, + "learning_rate": 0.0006003618018752548, + "loss": 3.5128, + "step": 47055 + }, + { + "epoch": 3.1974453050686233, + "grad_norm": 0.9326068162918091, + "learning_rate": 0.000600319336866422, + "loss": 3.2722, + "step": 47060 + }, + { + "epoch": 3.1977850251392854, + "grad_norm": 0.8306412696838379, + "learning_rate": 0.0006002768718575894, + "loss": 3.7013, + "step": 47065 + }, + { + "epoch": 3.198124745209947, + "grad_norm": 0.9422821998596191, + "learning_rate": 0.0006002344068487567, + "loss": 3.5461, + "step": 47070 + }, + { + "epoch": 3.1984644652806087, + "grad_norm": 0.7829583287239075, + "learning_rate": 0.0006001919418399239, + "loss": 3.7749, + "step": 47075 + }, + { + "epoch": 3.1988041853512708, + "grad_norm": 0.8829296231269836, + "learning_rate": 0.0006001494768310912, + "loss": 3.3466, + "step": 47080 + }, + { + "epoch": 3.1991439054219324, + "grad_norm": 1.0024772882461548, + "learning_rate": 0.0006001070118222585, + "loss": 3.6069, + "step": 47085 + }, + { + "epoch": 3.199483625492594, + "grad_norm": 0.8746809363365173, + "learning_rate": 0.0006000645468134257, + "loss": 3.6644, + "step": 47090 + }, + { + "epoch": 3.1998233455632556, + "grad_norm": 0.8248069286346436, + "learning_rate": 0.000600022081804593, + "loss": 3.3991, + "step": 47095 + }, + { + "epoch": 3.2001630656339177, + "grad_norm": 0.7885226011276245, + "learning_rate": 0.0005999796167957604, + "loss": 3.4272, + "step": 47100 + }, + { + "epoch": 3.2005027857045794, + "grad_norm": 0.9073771834373474, + "learning_rate": 0.0005999371517869276, + "loss": 3.4027, + "step": 47105 + }, + { + "epoch": 3.200842505775241, + "grad_norm": 0.7669292688369751, + "learning_rate": 0.0005998946867780949, + "loss": 3.8691, + "step": 47110 + }, + { + "epoch": 3.201182225845903, + "grad_norm": 0.8867843151092529, + "learning_rate": 0.0005998522217692621, + "loss": 3.5143, + "step": 47115 + }, + { + "epoch": 3.2015219459165647, + "grad_norm": 0.8381679654121399, + "learning_rate": 0.0005998097567604294, + "loss": 3.5298, + "step": 47120 + }, + { + "epoch": 3.2018616659872263, + "grad_norm": 0.8222712874412537, + "learning_rate": 0.0005997672917515967, + "loss": 3.1479, + "step": 47125 + }, + { + "epoch": 3.2022013860578884, + "grad_norm": 0.8474553823471069, + "learning_rate": 0.0005997248267427639, + "loss": 3.6948, + "step": 47130 + }, + { + "epoch": 3.20254110612855, + "grad_norm": 1.05476975440979, + "learning_rate": 0.0005996823617339313, + "loss": 3.6951, + "step": 47135 + }, + { + "epoch": 3.2028808261992117, + "grad_norm": 1.0135157108306885, + "learning_rate": 0.0005996398967250986, + "loss": 3.5089, + "step": 47140 + }, + { + "epoch": 3.2032205462698737, + "grad_norm": 0.82040935754776, + "learning_rate": 0.0005995974317162658, + "loss": 3.5641, + "step": 47145 + }, + { + "epoch": 3.2035602663405354, + "grad_norm": 0.8681055903434753, + "learning_rate": 0.000599554966707433, + "loss": 3.5092, + "step": 47150 + }, + { + "epoch": 3.203899986411197, + "grad_norm": 0.7641761302947998, + "learning_rate": 0.0005995125016986004, + "loss": 3.6961, + "step": 47155 + }, + { + "epoch": 3.204239706481859, + "grad_norm": 0.9352573752403259, + "learning_rate": 0.0005994700366897676, + "loss": 3.598, + "step": 47160 + }, + { + "epoch": 3.2045794265525207, + "grad_norm": 1.0525802373886108, + "learning_rate": 0.0005994275716809348, + "loss": 3.5342, + "step": 47165 + }, + { + "epoch": 3.2049191466231823, + "grad_norm": 0.8156723976135254, + "learning_rate": 0.0005993851066721023, + "loss": 3.7183, + "step": 47170 + }, + { + "epoch": 3.2052588666938444, + "grad_norm": 0.7460827231407166, + "learning_rate": 0.0005993426416632695, + "loss": 3.3779, + "step": 47175 + }, + { + "epoch": 3.205598586764506, + "grad_norm": 0.7945560216903687, + "learning_rate": 0.0005993001766544367, + "loss": 3.6999, + "step": 47180 + }, + { + "epoch": 3.2059383068351677, + "grad_norm": 0.8378802537918091, + "learning_rate": 0.0005992577116456041, + "loss": 3.3785, + "step": 47185 + }, + { + "epoch": 3.2062780269058297, + "grad_norm": 0.9115859866142273, + "learning_rate": 0.0005992152466367713, + "loss": 3.3966, + "step": 47190 + }, + { + "epoch": 3.2066177469764914, + "grad_norm": 0.9437305927276611, + "learning_rate": 0.0005991727816279386, + "loss": 3.4793, + "step": 47195 + }, + { + "epoch": 3.206957467047153, + "grad_norm": 0.7891079187393188, + "learning_rate": 0.0005991303166191058, + "loss": 3.2899, + "step": 47200 + }, + { + "epoch": 3.207297187117815, + "grad_norm": 1.0026376247406006, + "learning_rate": 0.0005990878516102732, + "loss": 3.3531, + "step": 47205 + }, + { + "epoch": 3.2076369071884767, + "grad_norm": 0.752679169178009, + "learning_rate": 0.0005990453866014405, + "loss": 3.6478, + "step": 47210 + }, + { + "epoch": 3.2079766272591383, + "grad_norm": 0.9419549107551575, + "learning_rate": 0.0005990029215926077, + "loss": 3.222, + "step": 47215 + }, + { + "epoch": 3.2083163473298004, + "grad_norm": 1.06291925907135, + "learning_rate": 0.000598960456583775, + "loss": 3.3734, + "step": 47220 + }, + { + "epoch": 3.208656067400462, + "grad_norm": 0.7885763645172119, + "learning_rate": 0.0005989179915749423, + "loss": 3.4591, + "step": 47225 + }, + { + "epoch": 3.2089957874711237, + "grad_norm": 0.804499626159668, + "learning_rate": 0.0005988755265661095, + "loss": 3.1605, + "step": 47230 + }, + { + "epoch": 3.2093355075417858, + "grad_norm": 1.008347988128662, + "learning_rate": 0.0005988330615572767, + "loss": 3.8184, + "step": 47235 + }, + { + "epoch": 3.2096752276124474, + "grad_norm": 0.9152479767799377, + "learning_rate": 0.0005987905965484442, + "loss": 3.5769, + "step": 47240 + }, + { + "epoch": 3.210014947683109, + "grad_norm": 0.908177375793457, + "learning_rate": 0.0005987481315396114, + "loss": 3.6251, + "step": 47245 + }, + { + "epoch": 3.210354667753771, + "grad_norm": 0.9669256806373596, + "learning_rate": 0.0005987056665307786, + "loss": 3.7742, + "step": 47250 + }, + { + "epoch": 3.2106943878244327, + "grad_norm": 0.8498631715774536, + "learning_rate": 0.000598663201521946, + "loss": 3.4866, + "step": 47255 + }, + { + "epoch": 3.2110341078950944, + "grad_norm": 0.8783431053161621, + "learning_rate": 0.0005986207365131132, + "loss": 3.5807, + "step": 47260 + }, + { + "epoch": 3.2113738279657564, + "grad_norm": 1.0012271404266357, + "learning_rate": 0.0005985782715042804, + "loss": 3.2675, + "step": 47265 + }, + { + "epoch": 3.211713548036418, + "grad_norm": 0.6955933570861816, + "learning_rate": 0.0005985358064954478, + "loss": 3.59, + "step": 47270 + }, + { + "epoch": 3.2120532681070797, + "grad_norm": 1.4229991436004639, + "learning_rate": 0.0005984933414866151, + "loss": 3.4889, + "step": 47275 + }, + { + "epoch": 3.2123929881777418, + "grad_norm": 1.0758135318756104, + "learning_rate": 0.0005984508764777823, + "loss": 3.6058, + "step": 47280 + }, + { + "epoch": 3.2127327082484034, + "grad_norm": 0.7102364897727966, + "learning_rate": 0.0005984084114689497, + "loss": 3.4476, + "step": 47285 + }, + { + "epoch": 3.213072428319065, + "grad_norm": 1.1684587001800537, + "learning_rate": 0.0005983659464601169, + "loss": 3.5806, + "step": 47290 + }, + { + "epoch": 3.2134121483897267, + "grad_norm": 0.9978742003440857, + "learning_rate": 0.0005983234814512841, + "loss": 3.3936, + "step": 47295 + }, + { + "epoch": 3.2137518684603887, + "grad_norm": 1.1747196912765503, + "learning_rate": 0.0005982810164424514, + "loss": 3.7275, + "step": 47300 + }, + { + "epoch": 3.2140915885310504, + "grad_norm": 0.713520884513855, + "learning_rate": 0.0005982385514336187, + "loss": 3.1827, + "step": 47305 + }, + { + "epoch": 3.214431308601712, + "grad_norm": 0.760844349861145, + "learning_rate": 0.000598196086424786, + "loss": 3.2023, + "step": 47310 + }, + { + "epoch": 3.214771028672374, + "grad_norm": 0.7904848456382751, + "learning_rate": 0.0005981536214159533, + "loss": 3.5108, + "step": 47315 + }, + { + "epoch": 3.2151107487430357, + "grad_norm": 0.7856009602546692, + "learning_rate": 0.0005981111564071206, + "loss": 3.673, + "step": 47320 + }, + { + "epoch": 3.2154504688136973, + "grad_norm": 0.7806258201599121, + "learning_rate": 0.0005980686913982878, + "loss": 3.6654, + "step": 47325 + }, + { + "epoch": 3.2157901888843594, + "grad_norm": 1.455320119857788, + "learning_rate": 0.0005980262263894551, + "loss": 3.3671, + "step": 47330 + }, + { + "epoch": 3.216129908955021, + "grad_norm": 0.9934332370758057, + "learning_rate": 0.0005979837613806224, + "loss": 3.4671, + "step": 47335 + }, + { + "epoch": 3.2164696290256827, + "grad_norm": 0.7893030643463135, + "learning_rate": 0.0005979412963717896, + "loss": 3.4871, + "step": 47340 + }, + { + "epoch": 3.2168093490963448, + "grad_norm": 0.9301550984382629, + "learning_rate": 0.000597898831362957, + "loss": 3.2587, + "step": 47345 + }, + { + "epoch": 3.2171490691670064, + "grad_norm": 0.8590857982635498, + "learning_rate": 0.0005978563663541242, + "loss": 3.5407, + "step": 47350 + }, + { + "epoch": 3.217488789237668, + "grad_norm": 0.7223317623138428, + "learning_rate": 0.0005978139013452915, + "loss": 3.3956, + "step": 47355 + }, + { + "epoch": 3.21782850930833, + "grad_norm": 0.9383039474487305, + "learning_rate": 0.0005977714363364588, + "loss": 3.7168, + "step": 47360 + }, + { + "epoch": 3.2181682293789917, + "grad_norm": 1.3038867712020874, + "learning_rate": 0.000597728971327626, + "loss": 3.3965, + "step": 47365 + }, + { + "epoch": 3.2185079494496533, + "grad_norm": 0.7732632160186768, + "learning_rate": 0.0005976865063187933, + "loss": 3.6644, + "step": 47370 + }, + { + "epoch": 3.2188476695203154, + "grad_norm": 0.8270350694656372, + "learning_rate": 0.0005976440413099606, + "loss": 3.383, + "step": 47375 + }, + { + "epoch": 3.219187389590977, + "grad_norm": 0.7388333678245544, + "learning_rate": 0.0005976015763011279, + "loss": 3.7236, + "step": 47380 + }, + { + "epoch": 3.2195271096616387, + "grad_norm": 0.7632237672805786, + "learning_rate": 0.0005975591112922952, + "loss": 3.1926, + "step": 47385 + }, + { + "epoch": 3.2198668297323008, + "grad_norm": 0.7406578660011292, + "learning_rate": 0.0005975166462834625, + "loss": 3.6323, + "step": 47390 + }, + { + "epoch": 3.2202065498029624, + "grad_norm": 1.0568454265594482, + "learning_rate": 0.0005974741812746297, + "loss": 3.5822, + "step": 47395 + }, + { + "epoch": 3.220546269873624, + "grad_norm": 0.7848625779151917, + "learning_rate": 0.0005974317162657969, + "loss": 3.4369, + "step": 47400 + }, + { + "epoch": 3.220885989944286, + "grad_norm": 1.4528298377990723, + "learning_rate": 0.0005973892512569643, + "loss": 3.4255, + "step": 47405 + }, + { + "epoch": 3.2212257100149477, + "grad_norm": 0.819679319858551, + "learning_rate": 0.0005973467862481315, + "loss": 3.4595, + "step": 47410 + }, + { + "epoch": 3.2215654300856094, + "grad_norm": 1.0311625003814697, + "learning_rate": 0.0005973043212392988, + "loss": 3.6325, + "step": 47415 + }, + { + "epoch": 3.2219051501562714, + "grad_norm": 1.0439918041229248, + "learning_rate": 0.0005972618562304662, + "loss": 3.4917, + "step": 47420 + }, + { + "epoch": 3.222244870226933, + "grad_norm": 1.1609846353530884, + "learning_rate": 0.0005972193912216334, + "loss": 3.4052, + "step": 47425 + }, + { + "epoch": 3.2225845902975947, + "grad_norm": 1.8750985860824585, + "learning_rate": 0.0005971769262128006, + "loss": 3.3674, + "step": 47430 + }, + { + "epoch": 3.2229243103682563, + "grad_norm": 1.1035839319229126, + "learning_rate": 0.000597134461203968, + "loss": 3.527, + "step": 47435 + }, + { + "epoch": 3.2232640304389184, + "grad_norm": 1.4744371175765991, + "learning_rate": 0.0005970919961951352, + "loss": 3.4153, + "step": 47440 + }, + { + "epoch": 3.22360375050958, + "grad_norm": 1.0677400827407837, + "learning_rate": 0.0005970495311863024, + "loss": 3.6763, + "step": 47445 + }, + { + "epoch": 3.2239434705802417, + "grad_norm": 1.136963129043579, + "learning_rate": 0.0005970070661774698, + "loss": 3.4439, + "step": 47450 + }, + { + "epoch": 3.2242831906509037, + "grad_norm": 0.8413625359535217, + "learning_rate": 0.0005969646011686371, + "loss": 3.6596, + "step": 47455 + }, + { + "epoch": 3.2246229107215654, + "grad_norm": 0.9899761080741882, + "learning_rate": 0.0005969221361598043, + "loss": 3.5833, + "step": 47460 + }, + { + "epoch": 3.224962630792227, + "grad_norm": 0.8009094595909119, + "learning_rate": 0.0005968796711509716, + "loss": 3.5576, + "step": 47465 + }, + { + "epoch": 3.225302350862889, + "grad_norm": 0.8916170597076416, + "learning_rate": 0.0005968372061421389, + "loss": 3.4667, + "step": 47470 + }, + { + "epoch": 3.2256420709335507, + "grad_norm": 0.7639058828353882, + "learning_rate": 0.0005967947411333061, + "loss": 3.5045, + "step": 47475 + }, + { + "epoch": 3.2259817910042123, + "grad_norm": 0.7537293434143066, + "learning_rate": 0.0005967522761244734, + "loss": 3.713, + "step": 47480 + }, + { + "epoch": 3.2263215110748744, + "grad_norm": 0.7929649353027344, + "learning_rate": 0.0005967098111156408, + "loss": 3.7627, + "step": 47485 + }, + { + "epoch": 3.226661231145536, + "grad_norm": 0.8976273536682129, + "learning_rate": 0.000596667346106808, + "loss": 3.5703, + "step": 47490 + }, + { + "epoch": 3.2270009512161977, + "grad_norm": 1.0552699565887451, + "learning_rate": 0.0005966248810979753, + "loss": 3.2194, + "step": 47495 + }, + { + "epoch": 3.2273406712868598, + "grad_norm": 0.9746345281600952, + "learning_rate": 0.0005965824160891425, + "loss": 3.2445, + "step": 47500 + }, + { + "epoch": 3.2276803913575214, + "grad_norm": 0.8705211281776428, + "learning_rate": 0.0005965399510803098, + "loss": 3.7214, + "step": 47505 + }, + { + "epoch": 3.228020111428183, + "grad_norm": 0.8772838711738586, + "learning_rate": 0.0005964974860714771, + "loss": 3.6214, + "step": 47510 + }, + { + "epoch": 3.228359831498845, + "grad_norm": 0.8401340842247009, + "learning_rate": 0.0005964550210626443, + "loss": 3.8197, + "step": 47515 + }, + { + "epoch": 3.2286995515695067, + "grad_norm": 0.7370373010635376, + "learning_rate": 0.0005964125560538117, + "loss": 3.2784, + "step": 47520 + }, + { + "epoch": 3.2290392716401684, + "grad_norm": 0.7231046557426453, + "learning_rate": 0.000596370091044979, + "loss": 3.5432, + "step": 47525 + }, + { + "epoch": 3.2293789917108304, + "grad_norm": 0.8223814964294434, + "learning_rate": 0.0005963276260361462, + "loss": 3.6303, + "step": 47530 + }, + { + "epoch": 3.229718711781492, + "grad_norm": 0.9232456088066101, + "learning_rate": 0.0005962851610273136, + "loss": 3.1634, + "step": 47535 + }, + { + "epoch": 3.2300584318521537, + "grad_norm": 0.9933381080627441, + "learning_rate": 0.0005962426960184808, + "loss": 3.4103, + "step": 47540 + }, + { + "epoch": 3.2303981519228158, + "grad_norm": 0.7173306345939636, + "learning_rate": 0.000596200231009648, + "loss": 3.4353, + "step": 47545 + }, + { + "epoch": 3.2307378719934774, + "grad_norm": 0.9653517007827759, + "learning_rate": 0.0005961577660008153, + "loss": 3.1973, + "step": 47550 + }, + { + "epoch": 3.231077592064139, + "grad_norm": 1.0191166400909424, + "learning_rate": 0.0005961153009919827, + "loss": 3.5182, + "step": 47555 + }, + { + "epoch": 3.231417312134801, + "grad_norm": 0.8312004804611206, + "learning_rate": 0.0005960728359831499, + "loss": 3.273, + "step": 47560 + }, + { + "epoch": 3.2317570322054627, + "grad_norm": 0.7020907402038574, + "learning_rate": 0.0005960303709743172, + "loss": 3.2968, + "step": 47565 + }, + { + "epoch": 3.2320967522761244, + "grad_norm": 0.7730855941772461, + "learning_rate": 0.0005959879059654845, + "loss": 3.467, + "step": 47570 + }, + { + "epoch": 3.2324364723467864, + "grad_norm": 0.8848169445991516, + "learning_rate": 0.0005959454409566517, + "loss": 3.5819, + "step": 47575 + }, + { + "epoch": 3.232776192417448, + "grad_norm": 0.9852948188781738, + "learning_rate": 0.000595902975947819, + "loss": 3.5139, + "step": 47580 + }, + { + "epoch": 3.2331159124881097, + "grad_norm": 0.971278190612793, + "learning_rate": 0.0005958605109389862, + "loss": 3.745, + "step": 47585 + }, + { + "epoch": 3.2334556325587718, + "grad_norm": 0.7775343656539917, + "learning_rate": 0.0005958180459301536, + "loss": 3.6919, + "step": 47590 + }, + { + "epoch": 3.2337953526294334, + "grad_norm": 0.8719583749771118, + "learning_rate": 0.0005957755809213209, + "loss": 3.3408, + "step": 47595 + }, + { + "epoch": 3.234135072700095, + "grad_norm": 0.8771272897720337, + "learning_rate": 0.0005957331159124881, + "loss": 3.8052, + "step": 47600 + }, + { + "epoch": 3.234474792770757, + "grad_norm": 1.0511424541473389, + "learning_rate": 0.0005956906509036554, + "loss": 3.669, + "step": 47605 + }, + { + "epoch": 3.2348145128414187, + "grad_norm": 0.7543144822120667, + "learning_rate": 0.0005956481858948227, + "loss": 3.3067, + "step": 47610 + }, + { + "epoch": 3.2351542329120804, + "grad_norm": 0.9410807490348816, + "learning_rate": 0.0005956057208859899, + "loss": 3.6167, + "step": 47615 + }, + { + "epoch": 3.2354939529827424, + "grad_norm": 0.7431885004043579, + "learning_rate": 0.0005955632558771572, + "loss": 3.7184, + "step": 47620 + }, + { + "epoch": 3.235833673053404, + "grad_norm": 1.0892412662506104, + "learning_rate": 0.0005955207908683246, + "loss": 3.3993, + "step": 47625 + }, + { + "epoch": 3.2361733931240657, + "grad_norm": 0.7989431619644165, + "learning_rate": 0.0005954783258594918, + "loss": 3.5031, + "step": 47630 + }, + { + "epoch": 3.2365131131947273, + "grad_norm": 0.8208248615264893, + "learning_rate": 0.000595435860850659, + "loss": 3.508, + "step": 47635 + }, + { + "epoch": 3.2368528332653894, + "grad_norm": 0.6892132759094238, + "learning_rate": 0.0005953933958418264, + "loss": 3.6028, + "step": 47640 + }, + { + "epoch": 3.237192553336051, + "grad_norm": 0.869514524936676, + "learning_rate": 0.0005953509308329936, + "loss": 3.5991, + "step": 47645 + }, + { + "epoch": 3.2375322734067127, + "grad_norm": 0.8583365678787231, + "learning_rate": 0.0005953084658241608, + "loss": 3.3453, + "step": 47650 + }, + { + "epoch": 3.2378719934773748, + "grad_norm": 1.061503291130066, + "learning_rate": 0.0005952660008153283, + "loss": 3.4741, + "step": 47655 + }, + { + "epoch": 3.2382117135480364, + "grad_norm": 0.8906923532485962, + "learning_rate": 0.0005952235358064955, + "loss": 3.7001, + "step": 47660 + }, + { + "epoch": 3.238551433618698, + "grad_norm": 0.9585073590278625, + "learning_rate": 0.0005951810707976627, + "loss": 3.5161, + "step": 47665 + }, + { + "epoch": 3.23889115368936, + "grad_norm": 0.6733388304710388, + "learning_rate": 0.0005951386057888301, + "loss": 3.5677, + "step": 47670 + }, + { + "epoch": 3.2392308737600217, + "grad_norm": 1.0495548248291016, + "learning_rate": 0.0005950961407799973, + "loss": 3.541, + "step": 47675 + }, + { + "epoch": 3.2395705938306834, + "grad_norm": 0.7816379070281982, + "learning_rate": 0.0005950536757711645, + "loss": 3.4436, + "step": 47680 + }, + { + "epoch": 3.2399103139013454, + "grad_norm": 0.9670003652572632, + "learning_rate": 0.0005950112107623318, + "loss": 3.8514, + "step": 47685 + }, + { + "epoch": 3.240250033972007, + "grad_norm": 0.8924698233604431, + "learning_rate": 0.0005949687457534992, + "loss": 3.4297, + "step": 47690 + }, + { + "epoch": 3.2405897540426687, + "grad_norm": 0.7472886443138123, + "learning_rate": 0.0005949262807446664, + "loss": 3.8485, + "step": 47695 + }, + { + "epoch": 3.2409294741133308, + "grad_norm": 0.7940961718559265, + "learning_rate": 0.0005948838157358337, + "loss": 3.4862, + "step": 47700 + }, + { + "epoch": 3.2412691941839924, + "grad_norm": 0.7485201358795166, + "learning_rate": 0.000594841350727001, + "loss": 3.6177, + "step": 47705 + }, + { + "epoch": 3.241608914254654, + "grad_norm": 1.0085467100143433, + "learning_rate": 0.0005947988857181682, + "loss": 3.6709, + "step": 47710 + }, + { + "epoch": 3.241948634325316, + "grad_norm": 1.01823091506958, + "learning_rate": 0.0005947564207093355, + "loss": 3.5782, + "step": 47715 + }, + { + "epoch": 3.2422883543959777, + "grad_norm": 1.1326749324798584, + "learning_rate": 0.0005947139557005028, + "loss": 3.5941, + "step": 47720 + }, + { + "epoch": 3.2426280744666394, + "grad_norm": 0.9730358123779297, + "learning_rate": 0.0005946714906916701, + "loss": 3.5624, + "step": 47725 + }, + { + "epoch": 3.2429677945373014, + "grad_norm": 1.0460444688796997, + "learning_rate": 0.0005946290256828374, + "loss": 3.4828, + "step": 47730 + }, + { + "epoch": 3.243307514607963, + "grad_norm": 0.9467706680297852, + "learning_rate": 0.0005945865606740046, + "loss": 3.5464, + "step": 47735 + }, + { + "epoch": 3.2436472346786247, + "grad_norm": 0.9326621890068054, + "learning_rate": 0.0005945440956651719, + "loss": 3.6374, + "step": 47740 + }, + { + "epoch": 3.2439869547492868, + "grad_norm": 0.7779879570007324, + "learning_rate": 0.0005945016306563392, + "loss": 3.5103, + "step": 47745 + }, + { + "epoch": 3.2443266748199484, + "grad_norm": 0.9772298336029053, + "learning_rate": 0.0005944591656475064, + "loss": 3.447, + "step": 47750 + }, + { + "epoch": 3.24466639489061, + "grad_norm": 0.9856424331665039, + "learning_rate": 0.0005944167006386737, + "loss": 3.711, + "step": 47755 + }, + { + "epoch": 3.245006114961272, + "grad_norm": 1.587141752243042, + "learning_rate": 0.0005943742356298411, + "loss": 3.5275, + "step": 47760 + }, + { + "epoch": 3.2453458350319337, + "grad_norm": 0.801163911819458, + "learning_rate": 0.0005943317706210083, + "loss": 3.663, + "step": 47765 + }, + { + "epoch": 3.2456855551025954, + "grad_norm": 0.8324598670005798, + "learning_rate": 0.0005942893056121756, + "loss": 3.5451, + "step": 47770 + }, + { + "epoch": 3.246025275173257, + "grad_norm": 0.856397807598114, + "learning_rate": 0.0005942468406033429, + "loss": 3.4862, + "step": 47775 + }, + { + "epoch": 3.246364995243919, + "grad_norm": 0.852536141872406, + "learning_rate": 0.0005942043755945101, + "loss": 3.5828, + "step": 47780 + }, + { + "epoch": 3.2467047153145807, + "grad_norm": 0.8401280641555786, + "learning_rate": 0.0005941619105856773, + "loss": 3.1456, + "step": 47785 + }, + { + "epoch": 3.2470444353852423, + "grad_norm": 0.9983728528022766, + "learning_rate": 0.0005941194455768447, + "loss": 3.4288, + "step": 47790 + }, + { + "epoch": 3.2473841554559044, + "grad_norm": 1.0214403867721558, + "learning_rate": 0.000594076980568012, + "loss": 3.4486, + "step": 47795 + }, + { + "epoch": 3.247723875526566, + "grad_norm": 0.8960226774215698, + "learning_rate": 0.0005940345155591792, + "loss": 3.4927, + "step": 47800 + }, + { + "epoch": 3.2480635955972277, + "grad_norm": 0.9636126160621643, + "learning_rate": 0.0005939920505503466, + "loss": 3.3291, + "step": 47805 + }, + { + "epoch": 3.2484033156678898, + "grad_norm": 0.7597578763961792, + "learning_rate": 0.0005939495855415138, + "loss": 3.6918, + "step": 47810 + }, + { + "epoch": 3.2487430357385514, + "grad_norm": 0.8202013373374939, + "learning_rate": 0.000593907120532681, + "loss": 3.2596, + "step": 47815 + }, + { + "epoch": 3.249082755809213, + "grad_norm": 0.7173140048980713, + "learning_rate": 0.0005938646555238484, + "loss": 3.7939, + "step": 47820 + }, + { + "epoch": 3.249422475879875, + "grad_norm": 1.1240214109420776, + "learning_rate": 0.0005938221905150156, + "loss": 3.5145, + "step": 47825 + }, + { + "epoch": 3.2497621959505367, + "grad_norm": 0.8800768256187439, + "learning_rate": 0.0005937797255061829, + "loss": 3.4274, + "step": 47830 + }, + { + "epoch": 3.2501019160211984, + "grad_norm": 0.8554653525352478, + "learning_rate": 0.0005937372604973503, + "loss": 3.611, + "step": 47835 + }, + { + "epoch": 3.2504416360918604, + "grad_norm": 0.9862136840820312, + "learning_rate": 0.0005936947954885175, + "loss": 3.7074, + "step": 47840 + }, + { + "epoch": 3.250781356162522, + "grad_norm": 0.7944707274436951, + "learning_rate": 0.0005936523304796847, + "loss": 3.5169, + "step": 47845 + }, + { + "epoch": 3.2511210762331837, + "grad_norm": 0.8265283703804016, + "learning_rate": 0.000593609865470852, + "loss": 3.7306, + "step": 47850 + }, + { + "epoch": 3.2514607963038458, + "grad_norm": 1.030619740486145, + "learning_rate": 0.0005935674004620193, + "loss": 3.6605, + "step": 47855 + }, + { + "epoch": 3.2518005163745074, + "grad_norm": 0.6259447932243347, + "learning_rate": 0.0005935249354531865, + "loss": 3.575, + "step": 47860 + }, + { + "epoch": 3.252140236445169, + "grad_norm": 0.8588775992393494, + "learning_rate": 0.0005934824704443539, + "loss": 3.319, + "step": 47865 + }, + { + "epoch": 3.252479956515831, + "grad_norm": 0.8911780714988708, + "learning_rate": 0.0005934400054355212, + "loss": 3.4639, + "step": 47870 + }, + { + "epoch": 3.2528196765864927, + "grad_norm": 0.9665262699127197, + "learning_rate": 0.0005933975404266885, + "loss": 3.297, + "step": 47875 + }, + { + "epoch": 3.2531593966571544, + "grad_norm": 0.9011222720146179, + "learning_rate": 0.0005933550754178557, + "loss": 3.6044, + "step": 47880 + }, + { + "epoch": 3.2534991167278164, + "grad_norm": 0.869540810585022, + "learning_rate": 0.0005933126104090229, + "loss": 3.5505, + "step": 47885 + }, + { + "epoch": 3.253838836798478, + "grad_norm": 0.9042954444885254, + "learning_rate": 0.0005932701454001903, + "loss": 3.6101, + "step": 47890 + }, + { + "epoch": 3.2541785568691397, + "grad_norm": 1.0394214391708374, + "learning_rate": 0.0005932276803913575, + "loss": 3.558, + "step": 47895 + }, + { + "epoch": 3.254518276939802, + "grad_norm": 0.8977407813072205, + "learning_rate": 0.0005931852153825248, + "loss": 3.367, + "step": 47900 + }, + { + "epoch": 3.2548579970104634, + "grad_norm": 1.0113568305969238, + "learning_rate": 0.0005931427503736922, + "loss": 3.4587, + "step": 47905 + }, + { + "epoch": 3.255197717081125, + "grad_norm": 0.8451798558235168, + "learning_rate": 0.0005931002853648594, + "loss": 3.4611, + "step": 47910 + }, + { + "epoch": 3.255537437151787, + "grad_norm": 1.0124350786209106, + "learning_rate": 0.0005930578203560266, + "loss": 3.3668, + "step": 47915 + }, + { + "epoch": 3.2558771572224487, + "grad_norm": 0.8732821345329285, + "learning_rate": 0.000593015355347194, + "loss": 3.8955, + "step": 47920 + }, + { + "epoch": 3.2562168772931104, + "grad_norm": 0.6830423474311829, + "learning_rate": 0.0005929728903383612, + "loss": 3.5467, + "step": 47925 + }, + { + "epoch": 3.2565565973637725, + "grad_norm": 0.90553879737854, + "learning_rate": 0.0005929304253295284, + "loss": 3.6289, + "step": 47930 + }, + { + "epoch": 3.256896317434434, + "grad_norm": 0.9308068156242371, + "learning_rate": 0.0005928879603206959, + "loss": 3.8293, + "step": 47935 + }, + { + "epoch": 3.2572360375050957, + "grad_norm": 1.0142700672149658, + "learning_rate": 0.0005928454953118631, + "loss": 3.4183, + "step": 47940 + }, + { + "epoch": 3.257575757575758, + "grad_norm": 0.826784074306488, + "learning_rate": 0.0005928030303030303, + "loss": 3.5251, + "step": 47945 + }, + { + "epoch": 3.2579154776464194, + "grad_norm": 0.9642004370689392, + "learning_rate": 0.0005927605652941976, + "loss": 3.5878, + "step": 47950 + }, + { + "epoch": 3.258255197717081, + "grad_norm": 0.7544974088668823, + "learning_rate": 0.0005927181002853649, + "loss": 3.6442, + "step": 47955 + }, + { + "epoch": 3.258594917787743, + "grad_norm": 0.8166085481643677, + "learning_rate": 0.0005926756352765321, + "loss": 3.7246, + "step": 47960 + }, + { + "epoch": 3.2589346378584048, + "grad_norm": 0.9310194253921509, + "learning_rate": 0.0005926331702676994, + "loss": 3.5062, + "step": 47965 + }, + { + "epoch": 3.2592743579290664, + "grad_norm": 0.8698039650917053, + "learning_rate": 0.0005925907052588668, + "loss": 3.6505, + "step": 47970 + }, + { + "epoch": 3.2596140779997285, + "grad_norm": 1.0251821279525757, + "learning_rate": 0.000592548240250034, + "loss": 3.3749, + "step": 47975 + }, + { + "epoch": 3.25995379807039, + "grad_norm": 0.8241837620735168, + "learning_rate": 0.0005925057752412013, + "loss": 3.3739, + "step": 47980 + }, + { + "epoch": 3.2602935181410517, + "grad_norm": 0.7880303263664246, + "learning_rate": 0.0005924633102323685, + "loss": 3.6489, + "step": 47985 + }, + { + "epoch": 3.2606332382117134, + "grad_norm": 0.8381921648979187, + "learning_rate": 0.0005924208452235358, + "loss": 3.5077, + "step": 47990 + }, + { + "epoch": 3.2609729582823754, + "grad_norm": 0.9544795751571655, + "learning_rate": 0.0005923783802147031, + "loss": 3.7188, + "step": 47995 + }, + { + "epoch": 3.261312678353037, + "grad_norm": 0.7482354044914246, + "learning_rate": 0.0005923359152058703, + "loss": 3.6199, + "step": 48000 + }, + { + "epoch": 3.2616523984236987, + "grad_norm": 0.9611053466796875, + "learning_rate": 0.0005922934501970377, + "loss": 3.5024, + "step": 48005 + }, + { + "epoch": 3.2619921184943608, + "grad_norm": 0.7979375720024109, + "learning_rate": 0.000592250985188205, + "loss": 3.7431, + "step": 48010 + }, + { + "epoch": 3.2623318385650224, + "grad_norm": 0.9857934713363647, + "learning_rate": 0.0005922085201793722, + "loss": 3.5621, + "step": 48015 + }, + { + "epoch": 3.262671558635684, + "grad_norm": 1.0658526420593262, + "learning_rate": 0.0005921660551705395, + "loss": 3.4895, + "step": 48020 + }, + { + "epoch": 3.263011278706346, + "grad_norm": 0.8188419342041016, + "learning_rate": 0.0005921235901617068, + "loss": 3.7967, + "step": 48025 + }, + { + "epoch": 3.2633509987770077, + "grad_norm": 0.6816413998603821, + "learning_rate": 0.000592081125152874, + "loss": 3.5721, + "step": 48030 + }, + { + "epoch": 3.2636907188476694, + "grad_norm": 0.8533735871315002, + "learning_rate": 0.0005920386601440412, + "loss": 3.4847, + "step": 48035 + }, + { + "epoch": 3.2640304389183314, + "grad_norm": 0.6705483794212341, + "learning_rate": 0.0005919961951352087, + "loss": 3.5714, + "step": 48040 + }, + { + "epoch": 3.264370158988993, + "grad_norm": 1.1123254299163818, + "learning_rate": 0.0005919537301263759, + "loss": 3.3895, + "step": 48045 + }, + { + "epoch": 3.2647098790596547, + "grad_norm": 0.9214597940444946, + "learning_rate": 0.0005919112651175431, + "loss": 3.7081, + "step": 48050 + }, + { + "epoch": 3.265049599130317, + "grad_norm": 0.9863991141319275, + "learning_rate": 0.0005918688001087105, + "loss": 3.2857, + "step": 48055 + }, + { + "epoch": 3.2653893192009784, + "grad_norm": 1.1323281526565552, + "learning_rate": 0.0005918263350998777, + "loss": 3.2503, + "step": 48060 + }, + { + "epoch": 3.26572903927164, + "grad_norm": 0.9129437208175659, + "learning_rate": 0.0005917838700910449, + "loss": 3.4975, + "step": 48065 + }, + { + "epoch": 3.266068759342302, + "grad_norm": 0.8707402348518372, + "learning_rate": 0.0005917414050822123, + "loss": 3.5394, + "step": 48070 + }, + { + "epoch": 3.2664084794129638, + "grad_norm": 0.8104270100593567, + "learning_rate": 0.0005916989400733796, + "loss": 3.557, + "step": 48075 + }, + { + "epoch": 3.2667481994836254, + "grad_norm": 0.7981496453285217, + "learning_rate": 0.0005916564750645468, + "loss": 3.3854, + "step": 48080 + }, + { + "epoch": 3.2670879195542875, + "grad_norm": 1.0782002210617065, + "learning_rate": 0.0005916140100557141, + "loss": 3.3512, + "step": 48085 + }, + { + "epoch": 3.267427639624949, + "grad_norm": 1.2251445055007935, + "learning_rate": 0.0005915715450468814, + "loss": 3.5906, + "step": 48090 + }, + { + "epoch": 3.2677673596956107, + "grad_norm": 1.0092848539352417, + "learning_rate": 0.0005915290800380486, + "loss": 3.5482, + "step": 48095 + }, + { + "epoch": 3.2681070797662723, + "grad_norm": 0.7547249794006348, + "learning_rate": 0.0005914866150292159, + "loss": 3.579, + "step": 48100 + }, + { + "epoch": 3.2684467998369344, + "grad_norm": 0.799691379070282, + "learning_rate": 0.0005914441500203832, + "loss": 3.5736, + "step": 48105 + }, + { + "epoch": 3.268786519907596, + "grad_norm": 1.1969653367996216, + "learning_rate": 0.0005914016850115505, + "loss": 3.5798, + "step": 48110 + }, + { + "epoch": 3.2691262399782577, + "grad_norm": 0.9269769191741943, + "learning_rate": 0.0005913592200027178, + "loss": 3.528, + "step": 48115 + }, + { + "epoch": 3.2694659600489198, + "grad_norm": 0.6286087036132812, + "learning_rate": 0.000591316754993885, + "loss": 3.5958, + "step": 48120 + }, + { + "epoch": 3.2698056801195814, + "grad_norm": 0.9626367688179016, + "learning_rate": 0.0005912742899850523, + "loss": 3.6998, + "step": 48125 + }, + { + "epoch": 3.270145400190243, + "grad_norm": 0.8893303871154785, + "learning_rate": 0.0005912318249762196, + "loss": 3.3395, + "step": 48130 + }, + { + "epoch": 3.270485120260905, + "grad_norm": 1.2342472076416016, + "learning_rate": 0.0005911893599673868, + "loss": 3.3006, + "step": 48135 + }, + { + "epoch": 3.2708248403315667, + "grad_norm": 1.2438536882400513, + "learning_rate": 0.0005911468949585541, + "loss": 3.508, + "step": 48140 + }, + { + "epoch": 3.2711645604022284, + "grad_norm": 1.0036423206329346, + "learning_rate": 0.0005911044299497215, + "loss": 3.6891, + "step": 48145 + }, + { + "epoch": 3.2715042804728904, + "grad_norm": 0.7225595712661743, + "learning_rate": 0.0005910619649408887, + "loss": 3.7253, + "step": 48150 + }, + { + "epoch": 3.271844000543552, + "grad_norm": 1.049621343612671, + "learning_rate": 0.000591019499932056, + "loss": 3.6803, + "step": 48155 + }, + { + "epoch": 3.2721837206142137, + "grad_norm": 0.9969396591186523, + "learning_rate": 0.0005909770349232233, + "loss": 3.0746, + "step": 48160 + }, + { + "epoch": 3.2725234406848758, + "grad_norm": 1.0230343341827393, + "learning_rate": 0.0005909345699143905, + "loss": 3.4988, + "step": 48165 + }, + { + "epoch": 3.2728631607555374, + "grad_norm": 1.0106868743896484, + "learning_rate": 0.0005908921049055577, + "loss": 3.2487, + "step": 48170 + }, + { + "epoch": 3.273202880826199, + "grad_norm": 0.988264262676239, + "learning_rate": 0.0005908496398967251, + "loss": 3.4339, + "step": 48175 + }, + { + "epoch": 3.273542600896861, + "grad_norm": 0.8124376535415649, + "learning_rate": 0.0005908071748878924, + "loss": 3.5168, + "step": 48180 + }, + { + "epoch": 3.2738823209675227, + "grad_norm": 1.035197377204895, + "learning_rate": 0.0005907647098790596, + "loss": 3.5886, + "step": 48185 + }, + { + "epoch": 3.2742220410381844, + "grad_norm": 1.0544496774673462, + "learning_rate": 0.000590722244870227, + "loss": 3.5541, + "step": 48190 + }, + { + "epoch": 3.2745617611088464, + "grad_norm": 0.9685846567153931, + "learning_rate": 0.0005906797798613942, + "loss": 3.6486, + "step": 48195 + }, + { + "epoch": 3.274901481179508, + "grad_norm": 0.9912694692611694, + "learning_rate": 0.0005906373148525614, + "loss": 3.5064, + "step": 48200 + }, + { + "epoch": 3.2752412012501697, + "grad_norm": 1.1322304010391235, + "learning_rate": 0.0005905948498437288, + "loss": 3.4654, + "step": 48205 + }, + { + "epoch": 3.275580921320832, + "grad_norm": 0.982255220413208, + "learning_rate": 0.000590552384834896, + "loss": 3.6384, + "step": 48210 + }, + { + "epoch": 3.2759206413914934, + "grad_norm": 0.859764575958252, + "learning_rate": 0.0005905099198260634, + "loss": 3.4107, + "step": 48215 + }, + { + "epoch": 3.276260361462155, + "grad_norm": 0.6648565530776978, + "learning_rate": 0.0005904674548172307, + "loss": 3.5586, + "step": 48220 + }, + { + "epoch": 3.276600081532817, + "grad_norm": 0.8247596025466919, + "learning_rate": 0.0005904249898083979, + "loss": 3.5166, + "step": 48225 + }, + { + "epoch": 3.2769398016034788, + "grad_norm": 1.1701301336288452, + "learning_rate": 0.0005903825247995652, + "loss": 3.5289, + "step": 48230 + }, + { + "epoch": 3.2772795216741404, + "grad_norm": 1.1855719089508057, + "learning_rate": 0.0005903400597907324, + "loss": 3.5381, + "step": 48235 + }, + { + "epoch": 3.2776192417448025, + "grad_norm": 0.9332132339477539, + "learning_rate": 0.0005902975947818997, + "loss": 3.5439, + "step": 48240 + }, + { + "epoch": 3.277958961815464, + "grad_norm": 0.7846083641052246, + "learning_rate": 0.0005902551297730671, + "loss": 3.7167, + "step": 48245 + }, + { + "epoch": 3.2782986818861257, + "grad_norm": 0.8178573846817017, + "learning_rate": 0.0005902126647642343, + "loss": 3.6252, + "step": 48250 + }, + { + "epoch": 3.278638401956788, + "grad_norm": 0.866978645324707, + "learning_rate": 0.0005901701997554016, + "loss": 3.3862, + "step": 48255 + }, + { + "epoch": 3.2789781220274494, + "grad_norm": 1.1222516298294067, + "learning_rate": 0.0005901277347465689, + "loss": 3.4811, + "step": 48260 + }, + { + "epoch": 3.279317842098111, + "grad_norm": 0.8810768723487854, + "learning_rate": 0.0005900852697377361, + "loss": 3.5017, + "step": 48265 + }, + { + "epoch": 3.279657562168773, + "grad_norm": 0.7866187691688538, + "learning_rate": 0.0005900428047289033, + "loss": 3.4903, + "step": 48270 + }, + { + "epoch": 3.2799972822394348, + "grad_norm": 1.19028902053833, + "learning_rate": 0.0005900003397200707, + "loss": 3.4481, + "step": 48275 + }, + { + "epoch": 3.2803370023100964, + "grad_norm": 3.2851321697235107, + "learning_rate": 0.000589957874711238, + "loss": 3.5865, + "step": 48280 + }, + { + "epoch": 3.2806767223807585, + "grad_norm": 0.8071588277816772, + "learning_rate": 0.0005899154097024052, + "loss": 3.5645, + "step": 48285 + }, + { + "epoch": 3.28101644245142, + "grad_norm": 0.916167140007019, + "learning_rate": 0.0005898729446935726, + "loss": 3.3681, + "step": 48290 + }, + { + "epoch": 3.2813561625220817, + "grad_norm": 0.7808096408843994, + "learning_rate": 0.0005898304796847398, + "loss": 3.6605, + "step": 48295 + }, + { + "epoch": 3.281695882592744, + "grad_norm": 0.9146581888198853, + "learning_rate": 0.000589788014675907, + "loss": 3.4266, + "step": 48300 + }, + { + "epoch": 3.2820356026634054, + "grad_norm": 1.1291333436965942, + "learning_rate": 0.0005897455496670744, + "loss": 3.5305, + "step": 48305 + }, + { + "epoch": 3.282375322734067, + "grad_norm": 1.1976310014724731, + "learning_rate": 0.0005897030846582416, + "loss": 3.7032, + "step": 48310 + }, + { + "epoch": 3.282715042804729, + "grad_norm": 0.9564703702926636, + "learning_rate": 0.0005896606196494089, + "loss": 3.5537, + "step": 48315 + }, + { + "epoch": 3.2830547628753908, + "grad_norm": 0.9660830497741699, + "learning_rate": 0.0005896181546405763, + "loss": 3.5807, + "step": 48320 + }, + { + "epoch": 3.2833944829460524, + "grad_norm": 1.0186069011688232, + "learning_rate": 0.0005895756896317435, + "loss": 3.5968, + "step": 48325 + }, + { + "epoch": 3.283734203016714, + "grad_norm": 0.880481481552124, + "learning_rate": 0.0005895332246229107, + "loss": 3.5364, + "step": 48330 + }, + { + "epoch": 3.284073923087376, + "grad_norm": 0.9628724455833435, + "learning_rate": 0.000589490759614078, + "loss": 3.4647, + "step": 48335 + }, + { + "epoch": 3.2844136431580377, + "grad_norm": 0.8436428308486938, + "learning_rate": 0.0005894482946052453, + "loss": 3.4925, + "step": 48340 + }, + { + "epoch": 3.2847533632286994, + "grad_norm": 0.7321258187294006, + "learning_rate": 0.0005894058295964125, + "loss": 3.1267, + "step": 48345 + }, + { + "epoch": 3.2850930832993614, + "grad_norm": 1.0233181715011597, + "learning_rate": 0.0005893633645875799, + "loss": 3.5618, + "step": 48350 + }, + { + "epoch": 3.285432803370023, + "grad_norm": 0.7987353205680847, + "learning_rate": 0.0005893208995787472, + "loss": 3.4351, + "step": 48355 + }, + { + "epoch": 3.2857725234406847, + "grad_norm": 1.008324384689331, + "learning_rate": 0.0005892784345699144, + "loss": 3.4012, + "step": 48360 + }, + { + "epoch": 3.286112243511347, + "grad_norm": 0.8408792614936829, + "learning_rate": 0.0005892359695610817, + "loss": 3.5417, + "step": 48365 + }, + { + "epoch": 3.2864519635820084, + "grad_norm": 2.9090662002563477, + "learning_rate": 0.000589193504552249, + "loss": 3.5862, + "step": 48370 + }, + { + "epoch": 3.28679168365267, + "grad_norm": 1.4570045471191406, + "learning_rate": 0.0005891510395434162, + "loss": 3.6157, + "step": 48375 + }, + { + "epoch": 3.287131403723332, + "grad_norm": 0.8945032954216003, + "learning_rate": 0.0005891085745345835, + "loss": 3.6317, + "step": 48380 + }, + { + "epoch": 3.2874711237939938, + "grad_norm": 0.7966932058334351, + "learning_rate": 0.0005890661095257508, + "loss": 3.4614, + "step": 48385 + }, + { + "epoch": 3.2878108438646554, + "grad_norm": 0.9876346588134766, + "learning_rate": 0.0005890236445169181, + "loss": 3.6183, + "step": 48390 + }, + { + "epoch": 3.2881505639353175, + "grad_norm": 0.8333478569984436, + "learning_rate": 0.0005889811795080854, + "loss": 3.5255, + "step": 48395 + }, + { + "epoch": 3.288490284005979, + "grad_norm": 0.7316277623176575, + "learning_rate": 0.0005889387144992526, + "loss": 3.6183, + "step": 48400 + }, + { + "epoch": 3.2888300040766407, + "grad_norm": 0.900107204914093, + "learning_rate": 0.0005888962494904199, + "loss": 3.861, + "step": 48405 + }, + { + "epoch": 3.289169724147303, + "grad_norm": 0.7224419713020325, + "learning_rate": 0.0005888537844815872, + "loss": 3.668, + "step": 48410 + }, + { + "epoch": 3.2895094442179644, + "grad_norm": 1.0782051086425781, + "learning_rate": 0.0005888113194727544, + "loss": 3.4122, + "step": 48415 + }, + { + "epoch": 3.289849164288626, + "grad_norm": 0.9715402126312256, + "learning_rate": 0.0005887688544639217, + "loss": 3.3546, + "step": 48420 + }, + { + "epoch": 3.290188884359288, + "grad_norm": 0.8950231075286865, + "learning_rate": 0.0005887263894550891, + "loss": 3.4969, + "step": 48425 + }, + { + "epoch": 3.2905286044299498, + "grad_norm": 1.0084583759307861, + "learning_rate": 0.0005886839244462563, + "loss": 3.6045, + "step": 48430 + }, + { + "epoch": 3.2908683245006114, + "grad_norm": 0.931252658367157, + "learning_rate": 0.0005886414594374235, + "loss": 3.3763, + "step": 48435 + }, + { + "epoch": 3.291208044571273, + "grad_norm": 1.1343733072280884, + "learning_rate": 0.0005885989944285909, + "loss": 3.8747, + "step": 48440 + }, + { + "epoch": 3.291547764641935, + "grad_norm": 0.8119016289710999, + "learning_rate": 0.0005885565294197581, + "loss": 3.8975, + "step": 48445 + }, + { + "epoch": 3.2918874847125967, + "grad_norm": 0.708626925945282, + "learning_rate": 0.0005885140644109253, + "loss": 3.7945, + "step": 48450 + }, + { + "epoch": 3.2922272047832584, + "grad_norm": 0.7924911379814148, + "learning_rate": 0.0005884715994020928, + "loss": 3.1025, + "step": 48455 + }, + { + "epoch": 3.2925669248539204, + "grad_norm": 0.815510630607605, + "learning_rate": 0.00058842913439326, + "loss": 3.2525, + "step": 48460 + }, + { + "epoch": 3.292906644924582, + "grad_norm": 0.875045120716095, + "learning_rate": 0.0005883866693844272, + "loss": 3.5468, + "step": 48465 + }, + { + "epoch": 3.2932463649952437, + "grad_norm": 1.8612422943115234, + "learning_rate": 0.0005883442043755945, + "loss": 3.6783, + "step": 48470 + }, + { + "epoch": 3.2935860850659058, + "grad_norm": 1.1833158731460571, + "learning_rate": 0.0005883017393667618, + "loss": 3.4675, + "step": 48475 + }, + { + "epoch": 3.2939258051365674, + "grad_norm": 1.1374233961105347, + "learning_rate": 0.000588259274357929, + "loss": 3.7298, + "step": 48480 + }, + { + "epoch": 3.294265525207229, + "grad_norm": 0.8647292852401733, + "learning_rate": 0.0005882168093490963, + "loss": 3.4268, + "step": 48485 + }, + { + "epoch": 3.294605245277891, + "grad_norm": 0.9696346521377563, + "learning_rate": 0.0005881743443402637, + "loss": 3.6202, + "step": 48490 + }, + { + "epoch": 3.2949449653485527, + "grad_norm": 1.0303493738174438, + "learning_rate": 0.0005881318793314309, + "loss": 3.4976, + "step": 48495 + }, + { + "epoch": 3.2952846854192144, + "grad_norm": 0.8287239670753479, + "learning_rate": 0.0005880894143225982, + "loss": 3.4932, + "step": 48500 + }, + { + "epoch": 3.2956244054898765, + "grad_norm": 0.7907270789146423, + "learning_rate": 0.0005880469493137655, + "loss": 3.6521, + "step": 48505 + }, + { + "epoch": 3.295964125560538, + "grad_norm": 0.8741850256919861, + "learning_rate": 0.0005880044843049327, + "loss": 3.3853, + "step": 48510 + }, + { + "epoch": 3.2963038456311997, + "grad_norm": 0.8967990875244141, + "learning_rate": 0.0005879620192961, + "loss": 3.5932, + "step": 48515 + }, + { + "epoch": 3.296643565701862, + "grad_norm": 0.79707932472229, + "learning_rate": 0.0005879195542872672, + "loss": 3.392, + "step": 48520 + }, + { + "epoch": 3.2969832857725234, + "grad_norm": 0.9180774092674255, + "learning_rate": 0.0005878770892784346, + "loss": 3.346, + "step": 48525 + }, + { + "epoch": 3.297323005843185, + "grad_norm": 2.717019557952881, + "learning_rate": 0.0005878346242696019, + "loss": 3.4252, + "step": 48530 + }, + { + "epoch": 3.297662725913847, + "grad_norm": 0.8125959634780884, + "learning_rate": 0.0005877921592607691, + "loss": 3.2545, + "step": 48535 + }, + { + "epoch": 3.2980024459845088, + "grad_norm": 0.921142041683197, + "learning_rate": 0.0005877496942519364, + "loss": 3.2168, + "step": 48540 + }, + { + "epoch": 3.2983421660551704, + "grad_norm": 0.9875665307044983, + "learning_rate": 0.0005877072292431037, + "loss": 3.6412, + "step": 48545 + }, + { + "epoch": 3.2986818861258325, + "grad_norm": 1.061631202697754, + "learning_rate": 0.0005876647642342709, + "loss": 3.4933, + "step": 48550 + }, + { + "epoch": 3.299021606196494, + "grad_norm": 0.8779887557029724, + "learning_rate": 0.0005876222992254383, + "loss": 3.6356, + "step": 48555 + }, + { + "epoch": 3.2993613262671557, + "grad_norm": 1.2114033699035645, + "learning_rate": 0.0005875798342166056, + "loss": 3.5949, + "step": 48560 + }, + { + "epoch": 3.299701046337818, + "grad_norm": 0.8092819452285767, + "learning_rate": 0.0005875373692077728, + "loss": 3.3185, + "step": 48565 + }, + { + "epoch": 3.3000407664084794, + "grad_norm": 0.7472546100616455, + "learning_rate": 0.0005874949041989402, + "loss": 3.6334, + "step": 48570 + }, + { + "epoch": 3.300380486479141, + "grad_norm": 0.9886726140975952, + "learning_rate": 0.0005874524391901074, + "loss": 3.5992, + "step": 48575 + }, + { + "epoch": 3.300720206549803, + "grad_norm": 0.8509883284568787, + "learning_rate": 0.0005874099741812746, + "loss": 3.6199, + "step": 48580 + }, + { + "epoch": 3.3010599266204648, + "grad_norm": 0.7618286609649658, + "learning_rate": 0.0005873675091724419, + "loss": 3.6605, + "step": 48585 + }, + { + "epoch": 3.3013996466911264, + "grad_norm": 0.9415149092674255, + "learning_rate": 0.0005873250441636092, + "loss": 3.363, + "step": 48590 + }, + { + "epoch": 3.3017393667617885, + "grad_norm": 0.8691444993019104, + "learning_rate": 0.0005872825791547765, + "loss": 3.6451, + "step": 48595 + }, + { + "epoch": 3.30207908683245, + "grad_norm": 0.7674158215522766, + "learning_rate": 0.0005872401141459438, + "loss": 3.4965, + "step": 48600 + }, + { + "epoch": 3.3024188069031117, + "grad_norm": 0.978895366191864, + "learning_rate": 0.0005871976491371111, + "loss": 3.3084, + "step": 48605 + }, + { + "epoch": 3.302758526973774, + "grad_norm": 0.9119603037834167, + "learning_rate": 0.0005871551841282783, + "loss": 3.5772, + "step": 48610 + }, + { + "epoch": 3.3030982470444354, + "grad_norm": 0.9546215534210205, + "learning_rate": 0.0005871127191194456, + "loss": 3.8092, + "step": 48615 + }, + { + "epoch": 3.303437967115097, + "grad_norm": 0.8560947775840759, + "learning_rate": 0.0005870702541106128, + "loss": 3.5182, + "step": 48620 + }, + { + "epoch": 3.303777687185759, + "grad_norm": 1.0189309120178223, + "learning_rate": 0.0005870277891017801, + "loss": 3.3749, + "step": 48625 + }, + { + "epoch": 3.304117407256421, + "grad_norm": 0.9473175406455994, + "learning_rate": 0.0005869853240929475, + "loss": 3.3947, + "step": 48630 + }, + { + "epoch": 3.3044571273270824, + "grad_norm": 0.9259925484657288, + "learning_rate": 0.0005869428590841147, + "loss": 3.7293, + "step": 48635 + }, + { + "epoch": 3.3047968473977445, + "grad_norm": 0.9136019349098206, + "learning_rate": 0.000586900394075282, + "loss": 3.7313, + "step": 48640 + }, + { + "epoch": 3.305136567468406, + "grad_norm": 1.0143972635269165, + "learning_rate": 0.0005868579290664493, + "loss": 3.609, + "step": 48645 + }, + { + "epoch": 3.3054762875390677, + "grad_norm": 0.757497251033783, + "learning_rate": 0.0005868154640576165, + "loss": 3.6623, + "step": 48650 + }, + { + "epoch": 3.30581600760973, + "grad_norm": 0.7379955649375916, + "learning_rate": 0.0005867729990487837, + "loss": 3.4996, + "step": 48655 + }, + { + "epoch": 3.3061557276803915, + "grad_norm": 0.7923427820205688, + "learning_rate": 0.0005867305340399511, + "loss": 3.7755, + "step": 48660 + }, + { + "epoch": 3.306495447751053, + "grad_norm": 0.7765454649925232, + "learning_rate": 0.0005866880690311184, + "loss": 3.4495, + "step": 48665 + }, + { + "epoch": 3.3068351678217147, + "grad_norm": 0.9614738821983337, + "learning_rate": 0.0005866456040222856, + "loss": 3.5063, + "step": 48670 + }, + { + "epoch": 3.307174887892377, + "grad_norm": 0.8087530732154846, + "learning_rate": 0.000586603139013453, + "loss": 3.4781, + "step": 48675 + }, + { + "epoch": 3.3075146079630384, + "grad_norm": 0.8304346203804016, + "learning_rate": 0.0005865606740046202, + "loss": 3.3196, + "step": 48680 + }, + { + "epoch": 3.3078543280337, + "grad_norm": 1.115853190422058, + "learning_rate": 0.0005865182089957874, + "loss": 3.2299, + "step": 48685 + }, + { + "epoch": 3.308194048104362, + "grad_norm": 0.9241434931755066, + "learning_rate": 0.0005864757439869548, + "loss": 3.3211, + "step": 48690 + }, + { + "epoch": 3.3085337681750238, + "grad_norm": 0.9389660358428955, + "learning_rate": 0.000586433278978122, + "loss": 3.5039, + "step": 48695 + }, + { + "epoch": 3.3088734882456854, + "grad_norm": 0.8963150978088379, + "learning_rate": 0.0005863908139692893, + "loss": 3.7154, + "step": 48700 + }, + { + "epoch": 3.3092132083163475, + "grad_norm": 0.7866663336753845, + "learning_rate": 0.0005863483489604567, + "loss": 3.4228, + "step": 48705 + }, + { + "epoch": 3.309552928387009, + "grad_norm": 0.9301358461380005, + "learning_rate": 0.0005863058839516239, + "loss": 3.3827, + "step": 48710 + }, + { + "epoch": 3.3098926484576707, + "grad_norm": 0.8656550645828247, + "learning_rate": 0.0005862634189427911, + "loss": 3.4802, + "step": 48715 + }, + { + "epoch": 3.310232368528333, + "grad_norm": 0.8601676225662231, + "learning_rate": 0.0005862209539339584, + "loss": 3.316, + "step": 48720 + }, + { + "epoch": 3.3105720885989944, + "grad_norm": 1.144139289855957, + "learning_rate": 0.0005861784889251257, + "loss": 3.582, + "step": 48725 + }, + { + "epoch": 3.310911808669656, + "grad_norm": 0.9279145002365112, + "learning_rate": 0.0005861360239162929, + "loss": 3.4837, + "step": 48730 + }, + { + "epoch": 3.311251528740318, + "grad_norm": 0.8956047296524048, + "learning_rate": 0.0005860935589074603, + "loss": 3.3379, + "step": 48735 + }, + { + "epoch": 3.3115912488109798, + "grad_norm": 0.8557221293449402, + "learning_rate": 0.0005860510938986276, + "loss": 3.6164, + "step": 48740 + }, + { + "epoch": 3.3119309688816414, + "grad_norm": 1.053959608078003, + "learning_rate": 0.0005860086288897948, + "loss": 3.3273, + "step": 48745 + }, + { + "epoch": 3.3122706889523035, + "grad_norm": 0.8266658782958984, + "learning_rate": 0.0005859661638809621, + "loss": 3.7829, + "step": 48750 + }, + { + "epoch": 3.312610409022965, + "grad_norm": 0.7742833495140076, + "learning_rate": 0.0005859236988721294, + "loss": 3.2963, + "step": 48755 + }, + { + "epoch": 3.3129501290936267, + "grad_norm": 0.7619365453720093, + "learning_rate": 0.0005858812338632966, + "loss": 3.399, + "step": 48760 + }, + { + "epoch": 3.313289849164289, + "grad_norm": 0.8914700150489807, + "learning_rate": 0.000585838768854464, + "loss": 3.5502, + "step": 48765 + }, + { + "epoch": 3.3136295692349504, + "grad_norm": 1.0287597179412842, + "learning_rate": 0.0005857963038456312, + "loss": 3.371, + "step": 48770 + }, + { + "epoch": 3.313969289305612, + "grad_norm": 0.7766013741493225, + "learning_rate": 0.0005857538388367985, + "loss": 3.4013, + "step": 48775 + }, + { + "epoch": 3.3143090093762737, + "grad_norm": 0.9402775168418884, + "learning_rate": 0.0005857113738279658, + "loss": 3.6048, + "step": 48780 + }, + { + "epoch": 3.314648729446936, + "grad_norm": 1.0506435632705688, + "learning_rate": 0.000585668908819133, + "loss": 3.5098, + "step": 48785 + }, + { + "epoch": 3.3149884495175974, + "grad_norm": 0.8823086023330688, + "learning_rate": 0.0005856264438103003, + "loss": 3.5752, + "step": 48790 + }, + { + "epoch": 3.315328169588259, + "grad_norm": 0.7824456691741943, + "learning_rate": 0.0005855839788014676, + "loss": 3.6757, + "step": 48795 + }, + { + "epoch": 3.315667889658921, + "grad_norm": 0.9050862193107605, + "learning_rate": 0.0005855415137926349, + "loss": 3.4355, + "step": 48800 + }, + { + "epoch": 3.3160076097295828, + "grad_norm": 0.9255902767181396, + "learning_rate": 0.0005854990487838022, + "loss": 3.6004, + "step": 48805 + }, + { + "epoch": 3.3163473298002444, + "grad_norm": 0.7035595178604126, + "learning_rate": 0.0005854565837749695, + "loss": 3.4515, + "step": 48810 + }, + { + "epoch": 3.3166870498709065, + "grad_norm": 1.0416179895401, + "learning_rate": 0.0005854141187661367, + "loss": 3.5305, + "step": 48815 + }, + { + "epoch": 3.317026769941568, + "grad_norm": 1.0944303274154663, + "learning_rate": 0.0005853716537573039, + "loss": 3.4974, + "step": 48820 + }, + { + "epoch": 3.3173664900122297, + "grad_norm": 0.9998109340667725, + "learning_rate": 0.0005853291887484713, + "loss": 3.6343, + "step": 48825 + }, + { + "epoch": 3.317706210082892, + "grad_norm": 0.9304737448692322, + "learning_rate": 0.0005852867237396385, + "loss": 3.3724, + "step": 48830 + }, + { + "epoch": 3.3180459301535534, + "grad_norm": 0.6992810368537903, + "learning_rate": 0.0005852442587308058, + "loss": 3.7925, + "step": 48835 + }, + { + "epoch": 3.318385650224215, + "grad_norm": 0.7170081734657288, + "learning_rate": 0.0005852017937219732, + "loss": 3.638, + "step": 48840 + }, + { + "epoch": 3.318725370294877, + "grad_norm": 0.7346291542053223, + "learning_rate": 0.0005851593287131404, + "loss": 3.7804, + "step": 48845 + }, + { + "epoch": 3.3190650903655388, + "grad_norm": 0.8415378928184509, + "learning_rate": 0.0005851168637043076, + "loss": 3.1964, + "step": 48850 + }, + { + "epoch": 3.3194048104362004, + "grad_norm": 0.8232410550117493, + "learning_rate": 0.000585074398695475, + "loss": 3.663, + "step": 48855 + }, + { + "epoch": 3.3197445305068625, + "grad_norm": 0.8720488548278809, + "learning_rate": 0.0005850319336866422, + "loss": 3.126, + "step": 48860 + }, + { + "epoch": 3.320084250577524, + "grad_norm": 1.0773766040802002, + "learning_rate": 0.0005849894686778094, + "loss": 3.5334, + "step": 48865 + }, + { + "epoch": 3.3204239706481857, + "grad_norm": 0.8816829919815063, + "learning_rate": 0.0005849470036689768, + "loss": 3.7275, + "step": 48870 + }, + { + "epoch": 3.320763690718848, + "grad_norm": 0.8054467439651489, + "learning_rate": 0.0005849045386601441, + "loss": 3.4183, + "step": 48875 + }, + { + "epoch": 3.3211034107895094, + "grad_norm": 0.9802054166793823, + "learning_rate": 0.0005848620736513113, + "loss": 3.5446, + "step": 48880 + }, + { + "epoch": 3.321443130860171, + "grad_norm": 0.9185819625854492, + "learning_rate": 0.0005848196086424786, + "loss": 3.496, + "step": 48885 + }, + { + "epoch": 3.321782850930833, + "grad_norm": 1.0449241399765015, + "learning_rate": 0.0005847771436336459, + "loss": 3.5153, + "step": 48890 + }, + { + "epoch": 3.3221225710014948, + "grad_norm": 0.9002196192741394, + "learning_rate": 0.0005847346786248132, + "loss": 3.5827, + "step": 48895 + }, + { + "epoch": 3.3224622910721564, + "grad_norm": 0.7623147368431091, + "learning_rate": 0.0005846922136159804, + "loss": 3.6484, + "step": 48900 + }, + { + "epoch": 3.3228020111428185, + "grad_norm": 0.9724432229995728, + "learning_rate": 0.0005846497486071478, + "loss": 3.4639, + "step": 48905 + }, + { + "epoch": 3.32314173121348, + "grad_norm": 0.7406185865402222, + "learning_rate": 0.0005846072835983151, + "loss": 3.8827, + "step": 48910 + }, + { + "epoch": 3.3234814512841417, + "grad_norm": 0.9950745105743408, + "learning_rate": 0.0005845648185894823, + "loss": 3.5762, + "step": 48915 + }, + { + "epoch": 3.323821171354804, + "grad_norm": 0.9344888925552368, + "learning_rate": 0.0005845223535806495, + "loss": 3.4181, + "step": 48920 + }, + { + "epoch": 3.3241608914254654, + "grad_norm": 0.8858740925788879, + "learning_rate": 0.0005844798885718169, + "loss": 3.5536, + "step": 48925 + }, + { + "epoch": 3.324500611496127, + "grad_norm": 1.044658899307251, + "learning_rate": 0.0005844374235629841, + "loss": 3.5687, + "step": 48930 + }, + { + "epoch": 3.324840331566789, + "grad_norm": 0.9449087381362915, + "learning_rate": 0.0005843949585541513, + "loss": 3.5733, + "step": 48935 + }, + { + "epoch": 3.325180051637451, + "grad_norm": 1.0380264520645142, + "learning_rate": 0.0005843524935453188, + "loss": 3.4844, + "step": 48940 + }, + { + "epoch": 3.3255197717081124, + "grad_norm": 0.7972650527954102, + "learning_rate": 0.000584310028536486, + "loss": 3.5664, + "step": 48945 + }, + { + "epoch": 3.3258594917787745, + "grad_norm": 0.9937728047370911, + "learning_rate": 0.0005842675635276532, + "loss": 3.6499, + "step": 48950 + }, + { + "epoch": 3.326199211849436, + "grad_norm": 0.8365567326545715, + "learning_rate": 0.0005842250985188206, + "loss": 3.639, + "step": 48955 + }, + { + "epoch": 3.3265389319200978, + "grad_norm": 0.8307557106018066, + "learning_rate": 0.0005841826335099878, + "loss": 3.526, + "step": 48960 + }, + { + "epoch": 3.32687865199076, + "grad_norm": 0.9083724617958069, + "learning_rate": 0.000584140168501155, + "loss": 3.7752, + "step": 48965 + }, + { + "epoch": 3.3272183720614215, + "grad_norm": 0.9049288034439087, + "learning_rate": 0.0005840977034923223, + "loss": 3.5597, + "step": 48970 + }, + { + "epoch": 3.327558092132083, + "grad_norm": 0.8801900744438171, + "learning_rate": 0.0005840552384834897, + "loss": 3.2409, + "step": 48975 + }, + { + "epoch": 3.327897812202745, + "grad_norm": 0.8086159229278564, + "learning_rate": 0.0005840127734746569, + "loss": 3.4845, + "step": 48980 + }, + { + "epoch": 3.328237532273407, + "grad_norm": 0.8561450839042664, + "learning_rate": 0.0005839703084658242, + "loss": 3.4352, + "step": 48985 + }, + { + "epoch": 3.3285772523440684, + "grad_norm": 1.0218385457992554, + "learning_rate": 0.0005839278434569915, + "loss": 3.3732, + "step": 48990 + }, + { + "epoch": 3.3289169724147305, + "grad_norm": 1.2099497318267822, + "learning_rate": 0.0005838853784481587, + "loss": 3.5525, + "step": 48995 + }, + { + "epoch": 3.329256692485392, + "grad_norm": 0.970750093460083, + "learning_rate": 0.000583842913439326, + "loss": 3.6337, + "step": 49000 + }, + { + "epoch": 3.3295964125560538, + "grad_norm": 0.9754088521003723, + "learning_rate": 0.0005838004484304932, + "loss": 3.3792, + "step": 49005 + }, + { + "epoch": 3.3299361326267154, + "grad_norm": 1.0598530769348145, + "learning_rate": 0.0005837579834216606, + "loss": 3.5458, + "step": 49010 + }, + { + "epoch": 3.3302758526973775, + "grad_norm": 1.1472526788711548, + "learning_rate": 0.0005837155184128279, + "loss": 3.3822, + "step": 49015 + }, + { + "epoch": 3.330615572768039, + "grad_norm": 1.1263474225997925, + "learning_rate": 0.0005836730534039951, + "loss": 3.4454, + "step": 49020 + }, + { + "epoch": 3.3309552928387007, + "grad_norm": 0.912329912185669, + "learning_rate": 0.0005836305883951624, + "loss": 3.732, + "step": 49025 + }, + { + "epoch": 3.331295012909363, + "grad_norm": 0.931022047996521, + "learning_rate": 0.0005835881233863297, + "loss": 3.7318, + "step": 49030 + }, + { + "epoch": 3.3316347329800244, + "grad_norm": 0.7978280782699585, + "learning_rate": 0.0005835456583774969, + "loss": 3.7725, + "step": 49035 + }, + { + "epoch": 3.331974453050686, + "grad_norm": 0.8076285123825073, + "learning_rate": 0.0005835031933686642, + "loss": 3.7428, + "step": 49040 + }, + { + "epoch": 3.332314173121348, + "grad_norm": 0.71431565284729, + "learning_rate": 0.0005834607283598316, + "loss": 3.438, + "step": 49045 + }, + { + "epoch": 3.3326538931920098, + "grad_norm": 1.2527812719345093, + "learning_rate": 0.0005834182633509988, + "loss": 3.6554, + "step": 49050 + }, + { + "epoch": 3.3329936132626714, + "grad_norm": 0.6672350168228149, + "learning_rate": 0.000583375798342166, + "loss": 3.6194, + "step": 49055 + }, + { + "epoch": 3.3333333333333335, + "grad_norm": 0.7781176567077637, + "learning_rate": 0.0005833333333333334, + "loss": 3.6281, + "step": 49060 + }, + { + "epoch": 3.333673053403995, + "grad_norm": 0.9181593656539917, + "learning_rate": 0.0005832908683245006, + "loss": 3.1371, + "step": 49065 + }, + { + "epoch": 3.3340127734746567, + "grad_norm": 1.114783763885498, + "learning_rate": 0.0005832484033156678, + "loss": 3.2219, + "step": 49070 + }, + { + "epoch": 3.334352493545319, + "grad_norm": 0.8489429354667664, + "learning_rate": 0.0005832059383068352, + "loss": 3.561, + "step": 49075 + }, + { + "epoch": 3.3346922136159804, + "grad_norm": 0.9623716473579407, + "learning_rate": 0.0005831634732980025, + "loss": 3.4225, + "step": 49080 + }, + { + "epoch": 3.335031933686642, + "grad_norm": 0.8397031426429749, + "learning_rate": 0.0005831210082891697, + "loss": 3.5534, + "step": 49085 + }, + { + "epoch": 3.335371653757304, + "grad_norm": 0.83551025390625, + "learning_rate": 0.0005830785432803371, + "loss": 3.2837, + "step": 49090 + }, + { + "epoch": 3.335711373827966, + "grad_norm": 0.8887967467308044, + "learning_rate": 0.0005830360782715043, + "loss": 3.8307, + "step": 49095 + }, + { + "epoch": 3.3360510938986274, + "grad_norm": 0.8023266792297363, + "learning_rate": 0.0005829936132626715, + "loss": 3.5649, + "step": 49100 + }, + { + "epoch": 3.3363908139692895, + "grad_norm": 0.9357014298439026, + "learning_rate": 0.0005829511482538388, + "loss": 3.3846, + "step": 49105 + }, + { + "epoch": 3.336730534039951, + "grad_norm": 0.8584401607513428, + "learning_rate": 0.0005829086832450061, + "loss": 3.7953, + "step": 49110 + }, + { + "epoch": 3.3370702541106128, + "grad_norm": 0.897729754447937, + "learning_rate": 0.0005828662182361734, + "loss": 3.4747, + "step": 49115 + }, + { + "epoch": 3.3374099741812744, + "grad_norm": 0.7875022888183594, + "learning_rate": 0.0005828237532273407, + "loss": 3.4458, + "step": 49120 + }, + { + "epoch": 3.3377496942519365, + "grad_norm": 1.0723532438278198, + "learning_rate": 0.000582781288218508, + "loss": 3.5319, + "step": 49125 + }, + { + "epoch": 3.338089414322598, + "grad_norm": 0.8995426893234253, + "learning_rate": 0.0005827388232096752, + "loss": 3.5665, + "step": 49130 + }, + { + "epoch": 3.3384291343932597, + "grad_norm": 0.7588154077529907, + "learning_rate": 0.0005826963582008425, + "loss": 3.5394, + "step": 49135 + }, + { + "epoch": 3.338768854463922, + "grad_norm": 1.076982021331787, + "learning_rate": 0.0005826538931920098, + "loss": 3.8499, + "step": 49140 + }, + { + "epoch": 3.3391085745345834, + "grad_norm": 0.9514582753181458, + "learning_rate": 0.000582611428183177, + "loss": 3.6433, + "step": 49145 + }, + { + "epoch": 3.339448294605245, + "grad_norm": 0.7473044991493225, + "learning_rate": 0.0005825689631743444, + "loss": 3.5236, + "step": 49150 + }, + { + "epoch": 3.339788014675907, + "grad_norm": 1.0321956872940063, + "learning_rate": 0.0005825264981655116, + "loss": 3.3804, + "step": 49155 + }, + { + "epoch": 3.3401277347465688, + "grad_norm": 0.9432067275047302, + "learning_rate": 0.0005824840331566789, + "loss": 3.4241, + "step": 49160 + }, + { + "epoch": 3.3404674548172304, + "grad_norm": 1.1034932136535645, + "learning_rate": 0.0005824415681478462, + "loss": 3.4387, + "step": 49165 + }, + { + "epoch": 3.3408071748878925, + "grad_norm": 0.8038483262062073, + "learning_rate": 0.0005823991031390134, + "loss": 3.5473, + "step": 49170 + }, + { + "epoch": 3.341146894958554, + "grad_norm": 0.8671546578407288, + "learning_rate": 0.0005823566381301807, + "loss": 3.5384, + "step": 49175 + }, + { + "epoch": 3.3414866150292157, + "grad_norm": 0.9980518817901611, + "learning_rate": 0.000582314173121348, + "loss": 3.3211, + "step": 49180 + }, + { + "epoch": 3.341826335099878, + "grad_norm": 0.9877550601959229, + "learning_rate": 0.0005822717081125153, + "loss": 3.154, + "step": 49185 + }, + { + "epoch": 3.3421660551705394, + "grad_norm": 0.7359760403633118, + "learning_rate": 0.0005822292431036826, + "loss": 3.3558, + "step": 49190 + }, + { + "epoch": 3.342505775241201, + "grad_norm": 0.9286291599273682, + "learning_rate": 0.0005821867780948499, + "loss": 3.3735, + "step": 49195 + }, + { + "epoch": 3.342845495311863, + "grad_norm": 1.0428755283355713, + "learning_rate": 0.0005821443130860171, + "loss": 3.5466, + "step": 49200 + }, + { + "epoch": 3.3431852153825248, + "grad_norm": 0.8222222924232483, + "learning_rate": 0.0005821018480771843, + "loss": 3.6782, + "step": 49205 + }, + { + "epoch": 3.3435249354531864, + "grad_norm": 1.1202727556228638, + "learning_rate": 0.0005820593830683517, + "loss": 3.652, + "step": 49210 + }, + { + "epoch": 3.3438646555238485, + "grad_norm": 1.102121114730835, + "learning_rate": 0.0005820169180595189, + "loss": 3.5105, + "step": 49215 + }, + { + "epoch": 3.34420437559451, + "grad_norm": 0.9617036581039429, + "learning_rate": 0.0005819744530506862, + "loss": 3.4512, + "step": 49220 + }, + { + "epoch": 3.3445440956651717, + "grad_norm": 0.963440477848053, + "learning_rate": 0.0005819319880418536, + "loss": 3.4761, + "step": 49225 + }, + { + "epoch": 3.344883815735834, + "grad_norm": 0.82782381772995, + "learning_rate": 0.0005818895230330208, + "loss": 3.6558, + "step": 49230 + }, + { + "epoch": 3.3452235358064955, + "grad_norm": 0.6698066592216492, + "learning_rate": 0.0005818470580241881, + "loss": 3.1863, + "step": 49235 + }, + { + "epoch": 3.345563255877157, + "grad_norm": 0.9416516423225403, + "learning_rate": 0.0005818045930153554, + "loss": 3.4994, + "step": 49240 + }, + { + "epoch": 3.345902975947819, + "grad_norm": 0.6994809508323669, + "learning_rate": 0.0005817621280065226, + "loss": 3.5688, + "step": 49245 + }, + { + "epoch": 3.346242696018481, + "grad_norm": 2.019428253173828, + "learning_rate": 0.0005817196629976899, + "loss": 3.602, + "step": 49250 + }, + { + "epoch": 3.3465824160891424, + "grad_norm": 1.1064766645431519, + "learning_rate": 0.0005816771979888573, + "loss": 3.6118, + "step": 49255 + }, + { + "epoch": 3.3469221361598045, + "grad_norm": 0.8234317302703857, + "learning_rate": 0.0005816347329800245, + "loss": 3.6022, + "step": 49260 + }, + { + "epoch": 3.347261856230466, + "grad_norm": 1.0805864334106445, + "learning_rate": 0.0005815922679711918, + "loss": 3.5319, + "step": 49265 + }, + { + "epoch": 3.3476015763011278, + "grad_norm": 0.9605377316474915, + "learning_rate": 0.000581549802962359, + "loss": 3.7473, + "step": 49270 + }, + { + "epoch": 3.34794129637179, + "grad_norm": 0.7837151885032654, + "learning_rate": 0.0005815073379535263, + "loss": 3.64, + "step": 49275 + }, + { + "epoch": 3.3482810164424515, + "grad_norm": 1.0982580184936523, + "learning_rate": 0.0005814648729446936, + "loss": 3.7129, + "step": 49280 + }, + { + "epoch": 3.348620736513113, + "grad_norm": 0.8103988170623779, + "learning_rate": 0.0005814224079358608, + "loss": 3.6508, + "step": 49285 + }, + { + "epoch": 3.348960456583775, + "grad_norm": 0.8725731372833252, + "learning_rate": 0.0005813799429270282, + "loss": 3.5725, + "step": 49290 + }, + { + "epoch": 3.349300176654437, + "grad_norm": 0.7639144062995911, + "learning_rate": 0.0005813374779181955, + "loss": 3.5523, + "step": 49295 + }, + { + "epoch": 3.3496398967250984, + "grad_norm": 0.856993556022644, + "learning_rate": 0.0005812950129093627, + "loss": 3.4821, + "step": 49300 + }, + { + "epoch": 3.3499796167957605, + "grad_norm": 1.027595043182373, + "learning_rate": 0.0005812525479005299, + "loss": 3.5007, + "step": 49305 + }, + { + "epoch": 3.350319336866422, + "grad_norm": 0.9101130962371826, + "learning_rate": 0.0005812100828916973, + "loss": 3.5386, + "step": 49310 + }, + { + "epoch": 3.3506590569370838, + "grad_norm": 1.6726362705230713, + "learning_rate": 0.0005811676178828645, + "loss": 3.7531, + "step": 49315 + }, + { + "epoch": 3.350998777007746, + "grad_norm": 0.8654756546020508, + "learning_rate": 0.0005811251528740317, + "loss": 3.6461, + "step": 49320 + }, + { + "epoch": 3.3513384970784075, + "grad_norm": 0.9322261810302734, + "learning_rate": 0.0005810826878651992, + "loss": 3.3982, + "step": 49325 + }, + { + "epoch": 3.351678217149069, + "grad_norm": 1.7551757097244263, + "learning_rate": 0.0005810402228563664, + "loss": 3.3868, + "step": 49330 + }, + { + "epoch": 3.352017937219731, + "grad_norm": 1.2007821798324585, + "learning_rate": 0.0005809977578475336, + "loss": 3.4701, + "step": 49335 + }, + { + "epoch": 3.352357657290393, + "grad_norm": 0.9446572065353394, + "learning_rate": 0.000580955292838701, + "loss": 3.6011, + "step": 49340 + }, + { + "epoch": 3.3526973773610544, + "grad_norm": 1.7760640382766724, + "learning_rate": 0.0005809128278298682, + "loss": 3.3534, + "step": 49345 + }, + { + "epoch": 3.353037097431716, + "grad_norm": 0.7445230484008789, + "learning_rate": 0.0005808703628210354, + "loss": 3.4277, + "step": 49350 + }, + { + "epoch": 3.353376817502378, + "grad_norm": 0.9552304148674011, + "learning_rate": 0.0005808278978122029, + "loss": 3.2389, + "step": 49355 + }, + { + "epoch": 3.35371653757304, + "grad_norm": 1.0730969905853271, + "learning_rate": 0.0005807854328033701, + "loss": 3.3811, + "step": 49360 + }, + { + "epoch": 3.3540562576437014, + "grad_norm": 0.9461817145347595, + "learning_rate": 0.0005807429677945373, + "loss": 3.4797, + "step": 49365 + }, + { + "epoch": 3.3543959777143635, + "grad_norm": 0.8003263473510742, + "learning_rate": 0.0005807005027857046, + "loss": 3.4392, + "step": 49370 + }, + { + "epoch": 3.354735697785025, + "grad_norm": 0.8905390501022339, + "learning_rate": 0.0005806580377768719, + "loss": 3.4122, + "step": 49375 + }, + { + "epoch": 3.3550754178556867, + "grad_norm": 0.9331502914428711, + "learning_rate": 0.0005806155727680391, + "loss": 3.4942, + "step": 49380 + }, + { + "epoch": 3.355415137926349, + "grad_norm": 1.1724779605865479, + "learning_rate": 0.0005805731077592064, + "loss": 3.5529, + "step": 49385 + }, + { + "epoch": 3.3557548579970105, + "grad_norm": 0.7212226390838623, + "learning_rate": 0.0005805306427503738, + "loss": 3.3232, + "step": 49390 + }, + { + "epoch": 3.356094578067672, + "grad_norm": 0.9879406690597534, + "learning_rate": 0.000580488177741541, + "loss": 3.6626, + "step": 49395 + }, + { + "epoch": 3.356434298138334, + "grad_norm": 0.9071311354637146, + "learning_rate": 0.0005804457127327083, + "loss": 3.5355, + "step": 49400 + }, + { + "epoch": 3.356774018208996, + "grad_norm": 0.7567989230155945, + "learning_rate": 0.0005804032477238755, + "loss": 3.4091, + "step": 49405 + }, + { + "epoch": 3.3571137382796574, + "grad_norm": 0.8235266208648682, + "learning_rate": 0.0005803607827150428, + "loss": 3.8166, + "step": 49410 + }, + { + "epoch": 3.3574534583503195, + "grad_norm": 0.8578166365623474, + "learning_rate": 0.0005803183177062101, + "loss": 3.3705, + "step": 49415 + }, + { + "epoch": 3.357793178420981, + "grad_norm": 0.7126883268356323, + "learning_rate": 0.0005802758526973773, + "loss": 3.5085, + "step": 49420 + }, + { + "epoch": 3.3581328984916428, + "grad_norm": 0.8250067234039307, + "learning_rate": 0.0005802333876885447, + "loss": 3.5275, + "step": 49425 + }, + { + "epoch": 3.358472618562305, + "grad_norm": 0.7594175934791565, + "learning_rate": 0.000580190922679712, + "loss": 3.4701, + "step": 49430 + }, + { + "epoch": 3.3588123386329665, + "grad_norm": 0.87702476978302, + "learning_rate": 0.0005801484576708792, + "loss": 3.1108, + "step": 49435 + }, + { + "epoch": 3.359152058703628, + "grad_norm": 1.1399182081222534, + "learning_rate": 0.0005801059926620465, + "loss": 3.1993, + "step": 49440 + }, + { + "epoch": 3.35949177877429, + "grad_norm": 1.0692555904388428, + "learning_rate": 0.0005800635276532138, + "loss": 3.2399, + "step": 49445 + }, + { + "epoch": 3.359831498844952, + "grad_norm": 0.9493743181228638, + "learning_rate": 0.000580021062644381, + "loss": 3.3503, + "step": 49450 + }, + { + "epoch": 3.3601712189156134, + "grad_norm": 1.017435908317566, + "learning_rate": 0.0005799785976355482, + "loss": 3.8118, + "step": 49455 + }, + { + "epoch": 3.360510938986275, + "grad_norm": 0.9139876961708069, + "learning_rate": 0.0005799361326267157, + "loss": 3.4476, + "step": 49460 + }, + { + "epoch": 3.360850659056937, + "grad_norm": 1.0009920597076416, + "learning_rate": 0.0005798936676178829, + "loss": 3.3259, + "step": 49465 + }, + { + "epoch": 3.3611903791275988, + "grad_norm": 0.8961812257766724, + "learning_rate": 0.0005798512026090501, + "loss": 3.5671, + "step": 49470 + }, + { + "epoch": 3.3615300991982604, + "grad_norm": 0.9386341571807861, + "learning_rate": 0.0005798087376002175, + "loss": 3.5515, + "step": 49475 + }, + { + "epoch": 3.3618698192689225, + "grad_norm": 0.8955754637718201, + "learning_rate": 0.0005797662725913847, + "loss": 3.5784, + "step": 49480 + }, + { + "epoch": 3.362209539339584, + "grad_norm": 1.1475889682769775, + "learning_rate": 0.0005797238075825519, + "loss": 3.3413, + "step": 49485 + }, + { + "epoch": 3.3625492594102457, + "grad_norm": 0.8356018662452698, + "learning_rate": 0.0005796813425737193, + "loss": 3.6213, + "step": 49490 + }, + { + "epoch": 3.362888979480908, + "grad_norm": 1.1528613567352295, + "learning_rate": 0.0005796388775648866, + "loss": 3.6929, + "step": 49495 + }, + { + "epoch": 3.3632286995515694, + "grad_norm": 0.9828994274139404, + "learning_rate": 0.0005795964125560538, + "loss": 3.5448, + "step": 49500 + }, + { + "epoch": 3.363568419622231, + "grad_norm": 0.9633510708808899, + "learning_rate": 0.0005795539475472211, + "loss": 3.3844, + "step": 49505 + }, + { + "epoch": 3.363908139692893, + "grad_norm": 0.9110558032989502, + "learning_rate": 0.0005795114825383884, + "loss": 3.4923, + "step": 49510 + }, + { + "epoch": 3.364247859763555, + "grad_norm": 1.036224603652954, + "learning_rate": 0.0005794690175295556, + "loss": 3.8053, + "step": 49515 + }, + { + "epoch": 3.3645875798342164, + "grad_norm": 0.8948330879211426, + "learning_rate": 0.0005794265525207229, + "loss": 3.3891, + "step": 49520 + }, + { + "epoch": 3.3649272999048785, + "grad_norm": 0.9554365873336792, + "learning_rate": 0.0005793840875118902, + "loss": 3.5588, + "step": 49525 + }, + { + "epoch": 3.36526701997554, + "grad_norm": 1.0419509410858154, + "learning_rate": 0.0005793416225030575, + "loss": 3.6099, + "step": 49530 + }, + { + "epoch": 3.3656067400462018, + "grad_norm": 0.8801946640014648, + "learning_rate": 0.0005792991574942248, + "loss": 3.8076, + "step": 49535 + }, + { + "epoch": 3.365946460116864, + "grad_norm": 0.8852800726890564, + "learning_rate": 0.000579256692485392, + "loss": 3.5506, + "step": 49540 + }, + { + "epoch": 3.3662861801875255, + "grad_norm": 0.9366156458854675, + "learning_rate": 0.0005792142274765593, + "loss": 3.5611, + "step": 49545 + }, + { + "epoch": 3.366625900258187, + "grad_norm": 0.768460214138031, + "learning_rate": 0.0005791717624677266, + "loss": 3.6667, + "step": 49550 + }, + { + "epoch": 3.366965620328849, + "grad_norm": 0.7904855012893677, + "learning_rate": 0.0005791292974588938, + "loss": 3.6792, + "step": 49555 + }, + { + "epoch": 3.367305340399511, + "grad_norm": 0.849022626876831, + "learning_rate": 0.0005790868324500611, + "loss": 3.2532, + "step": 49560 + }, + { + "epoch": 3.3676450604701724, + "grad_norm": 0.9996537566184998, + "learning_rate": 0.0005790443674412285, + "loss": 3.5101, + "step": 49565 + }, + { + "epoch": 3.3679847805408345, + "grad_norm": 0.9901342988014221, + "learning_rate": 0.0005790019024323957, + "loss": 3.5807, + "step": 49570 + }, + { + "epoch": 3.368324500611496, + "grad_norm": 0.760643482208252, + "learning_rate": 0.0005789594374235631, + "loss": 3.4887, + "step": 49575 + }, + { + "epoch": 3.3686642206821578, + "grad_norm": 1.0973819494247437, + "learning_rate": 0.0005789169724147303, + "loss": 3.5991, + "step": 49580 + }, + { + "epoch": 3.36900394075282, + "grad_norm": 1.0243356227874756, + "learning_rate": 0.0005788745074058975, + "loss": 3.2717, + "step": 49585 + }, + { + "epoch": 3.3693436608234815, + "grad_norm": 0.6387946605682373, + "learning_rate": 0.0005788320423970649, + "loss": 3.4676, + "step": 49590 + }, + { + "epoch": 3.369683380894143, + "grad_norm": 0.9899793863296509, + "learning_rate": 0.0005787895773882321, + "loss": 3.5605, + "step": 49595 + }, + { + "epoch": 3.370023100964805, + "grad_norm": 1.023699164390564, + "learning_rate": 0.0005787471123793994, + "loss": 3.5338, + "step": 49600 + }, + { + "epoch": 3.370362821035467, + "grad_norm": 0.8970270156860352, + "learning_rate": 0.0005787046473705667, + "loss": 3.5108, + "step": 49605 + }, + { + "epoch": 3.3707025411061284, + "grad_norm": 0.8073155879974365, + "learning_rate": 0.000578662182361734, + "loss": 3.5912, + "step": 49610 + }, + { + "epoch": 3.3710422611767905, + "grad_norm": 0.687433123588562, + "learning_rate": 0.0005786197173529012, + "loss": 3.3209, + "step": 49615 + }, + { + "epoch": 3.371381981247452, + "grad_norm": 0.8400617241859436, + "learning_rate": 0.0005785772523440685, + "loss": 3.6146, + "step": 49620 + }, + { + "epoch": 3.3717217013181138, + "grad_norm": 0.9855444431304932, + "learning_rate": 0.0005785347873352358, + "loss": 3.4652, + "step": 49625 + }, + { + "epoch": 3.372061421388776, + "grad_norm": 1.0187273025512695, + "learning_rate": 0.000578492322326403, + "loss": 3.5558, + "step": 49630 + }, + { + "epoch": 3.3724011414594375, + "grad_norm": 1.021185278892517, + "learning_rate": 0.0005784498573175704, + "loss": 3.5024, + "step": 49635 + }, + { + "epoch": 3.372740861530099, + "grad_norm": 0.7996768355369568, + "learning_rate": 0.0005784073923087377, + "loss": 3.4078, + "step": 49640 + }, + { + "epoch": 3.373080581600761, + "grad_norm": 0.8960021734237671, + "learning_rate": 0.0005783649272999049, + "loss": 3.4692, + "step": 49645 + }, + { + "epoch": 3.373420301671423, + "grad_norm": 0.7097633481025696, + "learning_rate": 0.0005783224622910722, + "loss": 3.5738, + "step": 49650 + }, + { + "epoch": 3.3737600217420844, + "grad_norm": 0.9948644638061523, + "learning_rate": 0.0005782799972822394, + "loss": 3.6191, + "step": 49655 + }, + { + "epoch": 3.3740997418127465, + "grad_norm": 1.1178803443908691, + "learning_rate": 0.0005782375322734067, + "loss": 3.3585, + "step": 49660 + }, + { + "epoch": 3.374439461883408, + "grad_norm": 1.4558680057525635, + "learning_rate": 0.000578195067264574, + "loss": 3.5182, + "step": 49665 + }, + { + "epoch": 3.37477918195407, + "grad_norm": 0.7311971187591553, + "learning_rate": 0.0005781526022557413, + "loss": 3.5622, + "step": 49670 + }, + { + "epoch": 3.375118902024732, + "grad_norm": 0.8076525926589966, + "learning_rate": 0.0005781101372469086, + "loss": 3.4742, + "step": 49675 + }, + { + "epoch": 3.3754586220953935, + "grad_norm": 1.0293940305709839, + "learning_rate": 0.0005780676722380759, + "loss": 3.3569, + "step": 49680 + }, + { + "epoch": 3.375798342166055, + "grad_norm": 0.7506266832351685, + "learning_rate": 0.0005780252072292431, + "loss": 3.5488, + "step": 49685 + }, + { + "epoch": 3.3761380622367168, + "grad_norm": 1.185490608215332, + "learning_rate": 0.0005779827422204103, + "loss": 3.4223, + "step": 49690 + }, + { + "epoch": 3.376477782307379, + "grad_norm": 0.8698786497116089, + "learning_rate": 0.0005779402772115777, + "loss": 3.7503, + "step": 49695 + }, + { + "epoch": 3.3768175023780405, + "grad_norm": 1.3655766248703003, + "learning_rate": 0.0005778978122027449, + "loss": 3.5298, + "step": 49700 + }, + { + "epoch": 3.377157222448702, + "grad_norm": 0.8765692114830017, + "learning_rate": 0.0005778553471939122, + "loss": 3.4266, + "step": 49705 + }, + { + "epoch": 3.377496942519364, + "grad_norm": 0.8900591731071472, + "learning_rate": 0.0005778128821850796, + "loss": 3.423, + "step": 49710 + }, + { + "epoch": 3.377836662590026, + "grad_norm": 0.7250720262527466, + "learning_rate": 0.0005777704171762468, + "loss": 3.4104, + "step": 49715 + }, + { + "epoch": 3.3781763826606874, + "grad_norm": 0.9090709090232849, + "learning_rate": 0.000577727952167414, + "loss": 3.6717, + "step": 49720 + }, + { + "epoch": 3.3785161027313495, + "grad_norm": 0.9523364305496216, + "learning_rate": 0.0005776854871585814, + "loss": 3.4284, + "step": 49725 + }, + { + "epoch": 3.378855822802011, + "grad_norm": 0.9779092669487, + "learning_rate": 0.0005776430221497486, + "loss": 3.5203, + "step": 49730 + }, + { + "epoch": 3.3791955428726728, + "grad_norm": 0.9906118512153625, + "learning_rate": 0.0005776005571409158, + "loss": 3.5643, + "step": 49735 + }, + { + "epoch": 3.379535262943335, + "grad_norm": 0.9301699995994568, + "learning_rate": 0.0005775580921320833, + "loss": 3.6165, + "step": 49740 + }, + { + "epoch": 3.3798749830139965, + "grad_norm": 1.045536756515503, + "learning_rate": 0.0005775156271232505, + "loss": 3.4772, + "step": 49745 + }, + { + "epoch": 3.380214703084658, + "grad_norm": 0.9242030382156372, + "learning_rate": 0.0005774731621144177, + "loss": 3.4709, + "step": 49750 + }, + { + "epoch": 3.38055442315532, + "grad_norm": 0.8237740397453308, + "learning_rate": 0.000577430697105585, + "loss": 3.485, + "step": 49755 + }, + { + "epoch": 3.380894143225982, + "grad_norm": 1.0817036628723145, + "learning_rate": 0.0005773882320967523, + "loss": 3.3025, + "step": 49760 + }, + { + "epoch": 3.3812338632966434, + "grad_norm": 1.1393084526062012, + "learning_rate": 0.0005773457670879195, + "loss": 3.5676, + "step": 49765 + }, + { + "epoch": 3.3815735833673055, + "grad_norm": 0.827745258808136, + "learning_rate": 0.0005773033020790868, + "loss": 3.366, + "step": 49770 + }, + { + "epoch": 3.381913303437967, + "grad_norm": 0.8032366037368774, + "learning_rate": 0.0005772608370702542, + "loss": 3.641, + "step": 49775 + }, + { + "epoch": 3.3822530235086288, + "grad_norm": 0.9849168062210083, + "learning_rate": 0.0005772183720614214, + "loss": 3.6279, + "step": 49780 + }, + { + "epoch": 3.382592743579291, + "grad_norm": 0.7085905075073242, + "learning_rate": 0.0005771759070525887, + "loss": 3.5161, + "step": 49785 + }, + { + "epoch": 3.3829324636499525, + "grad_norm": 0.7520939707756042, + "learning_rate": 0.000577133442043756, + "loss": 3.6268, + "step": 49790 + }, + { + "epoch": 3.383272183720614, + "grad_norm": 0.8060070872306824, + "learning_rate": 0.0005770909770349232, + "loss": 3.56, + "step": 49795 + }, + { + "epoch": 3.3836119037912757, + "grad_norm": 0.7752734422683716, + "learning_rate": 0.0005770485120260905, + "loss": 3.6275, + "step": 49800 + }, + { + "epoch": 3.383951623861938, + "grad_norm": 0.751921534538269, + "learning_rate": 0.0005770060470172577, + "loss": 3.2834, + "step": 49805 + }, + { + "epoch": 3.3842913439325994, + "grad_norm": 0.8303234577178955, + "learning_rate": 0.0005769635820084251, + "loss": 3.6246, + "step": 49810 + }, + { + "epoch": 3.384631064003261, + "grad_norm": 0.884976863861084, + "learning_rate": 0.0005769211169995924, + "loss": 3.5424, + "step": 49815 + }, + { + "epoch": 3.384970784073923, + "grad_norm": 0.8058900833129883, + "learning_rate": 0.0005768786519907596, + "loss": 3.4267, + "step": 49820 + }, + { + "epoch": 3.385310504144585, + "grad_norm": 0.8419861793518066, + "learning_rate": 0.0005768361869819269, + "loss": 3.5548, + "step": 49825 + }, + { + "epoch": 3.3856502242152464, + "grad_norm": 1.2954851388931274, + "learning_rate": 0.0005767937219730942, + "loss": 3.7868, + "step": 49830 + }, + { + "epoch": 3.3859899442859085, + "grad_norm": 1.1548715829849243, + "learning_rate": 0.0005767512569642614, + "loss": 3.8986, + "step": 49835 + }, + { + "epoch": 3.38632966435657, + "grad_norm": 0.8057295083999634, + "learning_rate": 0.0005767087919554286, + "loss": 3.3758, + "step": 49840 + }, + { + "epoch": 3.3866693844272318, + "grad_norm": 0.8722151517868042, + "learning_rate": 0.0005766663269465961, + "loss": 3.3405, + "step": 49845 + }, + { + "epoch": 3.387009104497894, + "grad_norm": 0.851099967956543, + "learning_rate": 0.0005766238619377633, + "loss": 3.7131, + "step": 49850 + }, + { + "epoch": 3.3873488245685555, + "grad_norm": 0.7858754992485046, + "learning_rate": 0.0005765813969289305, + "loss": 3.5018, + "step": 49855 + }, + { + "epoch": 3.387688544639217, + "grad_norm": 1.0244842767715454, + "learning_rate": 0.0005765389319200979, + "loss": 3.528, + "step": 49860 + }, + { + "epoch": 3.388028264709879, + "grad_norm": 0.8234319686889648, + "learning_rate": 0.0005764964669112651, + "loss": 3.5328, + "step": 49865 + }, + { + "epoch": 3.388367984780541, + "grad_norm": 0.9315900206565857, + "learning_rate": 0.0005764540019024323, + "loss": 3.6246, + "step": 49870 + }, + { + "epoch": 3.3887077048512024, + "grad_norm": 0.8407338857650757, + "learning_rate": 0.0005764115368935997, + "loss": 3.7096, + "step": 49875 + }, + { + "epoch": 3.3890474249218645, + "grad_norm": 0.8791391849517822, + "learning_rate": 0.000576369071884767, + "loss": 3.6277, + "step": 49880 + }, + { + "epoch": 3.389387144992526, + "grad_norm": 0.9804511070251465, + "learning_rate": 0.0005763266068759342, + "loss": 3.461, + "step": 49885 + }, + { + "epoch": 3.3897268650631878, + "grad_norm": 1.0360069274902344, + "learning_rate": 0.0005762841418671015, + "loss": 3.3306, + "step": 49890 + }, + { + "epoch": 3.39006658513385, + "grad_norm": 0.7966498732566833, + "learning_rate": 0.0005762416768582688, + "loss": 3.6058, + "step": 49895 + }, + { + "epoch": 3.3904063052045115, + "grad_norm": 1.0146933794021606, + "learning_rate": 0.000576199211849436, + "loss": 3.7228, + "step": 49900 + }, + { + "epoch": 3.390746025275173, + "grad_norm": 1.0984506607055664, + "learning_rate": 0.0005761567468406033, + "loss": 3.394, + "step": 49905 + }, + { + "epoch": 3.391085745345835, + "grad_norm": 0.9458656907081604, + "learning_rate": 0.0005761142818317706, + "loss": 3.456, + "step": 49910 + }, + { + "epoch": 3.391425465416497, + "grad_norm": 1.1642011404037476, + "learning_rate": 0.000576071816822938, + "loss": 3.43, + "step": 49915 + }, + { + "epoch": 3.3917651854871584, + "grad_norm": 1.1110413074493408, + "learning_rate": 0.0005760293518141052, + "loss": 3.2909, + "step": 49920 + }, + { + "epoch": 3.3921049055578205, + "grad_norm": 0.9257332682609558, + "learning_rate": 0.0005759868868052725, + "loss": 3.5936, + "step": 49925 + }, + { + "epoch": 3.392444625628482, + "grad_norm": 0.8919389247894287, + "learning_rate": 0.0005759444217964398, + "loss": 3.5385, + "step": 49930 + }, + { + "epoch": 3.3927843456991438, + "grad_norm": 0.9872719645500183, + "learning_rate": 0.000575901956787607, + "loss": 3.7157, + "step": 49935 + }, + { + "epoch": 3.393124065769806, + "grad_norm": 14.682124137878418, + "learning_rate": 0.0005758594917787742, + "loss": 3.4595, + "step": 49940 + }, + { + "epoch": 3.3934637858404675, + "grad_norm": 0.8736067414283752, + "learning_rate": 0.0005758170267699417, + "loss": 3.5281, + "step": 49945 + }, + { + "epoch": 3.393803505911129, + "grad_norm": 0.8613869547843933, + "learning_rate": 0.0005757745617611089, + "loss": 3.5113, + "step": 49950 + }, + { + "epoch": 3.394143225981791, + "grad_norm": 0.8032278418540955, + "learning_rate": 0.0005757320967522761, + "loss": 3.6965, + "step": 49955 + }, + { + "epoch": 3.394482946052453, + "grad_norm": 0.7194338440895081, + "learning_rate": 0.0005756896317434435, + "loss": 3.5288, + "step": 49960 + }, + { + "epoch": 3.3948226661231145, + "grad_norm": 0.8321855664253235, + "learning_rate": 0.0005756471667346107, + "loss": 3.5619, + "step": 49965 + }, + { + "epoch": 3.3951623861937765, + "grad_norm": 0.9656549096107483, + "learning_rate": 0.0005756047017257779, + "loss": 3.3194, + "step": 49970 + }, + { + "epoch": 3.395502106264438, + "grad_norm": 0.7215351462364197, + "learning_rate": 0.0005755622367169453, + "loss": 3.4955, + "step": 49975 + }, + { + "epoch": 3.3958418263351, + "grad_norm": 0.948928713798523, + "learning_rate": 0.0005755197717081126, + "loss": 3.6232, + "step": 49980 + }, + { + "epoch": 3.396181546405762, + "grad_norm": 0.7622394561767578, + "learning_rate": 0.0005754773066992798, + "loss": 3.4489, + "step": 49985 + }, + { + "epoch": 3.3965212664764235, + "grad_norm": 0.7495489120483398, + "learning_rate": 0.0005754348416904472, + "loss": 3.7415, + "step": 49990 + }, + { + "epoch": 3.396860986547085, + "grad_norm": 0.7207707166671753, + "learning_rate": 0.0005753923766816144, + "loss": 3.342, + "step": 49995 + }, + { + "epoch": 3.397200706617747, + "grad_norm": 0.8471470475196838, + "learning_rate": 0.0005753499116727816, + "loss": 3.6707, + "step": 50000 + }, + { + "epoch": 3.397540426688409, + "grad_norm": 1.2151710987091064, + "learning_rate": 0.0005753074466639489, + "loss": 3.6174, + "step": 50005 + }, + { + "epoch": 3.3978801467590705, + "grad_norm": 0.9039880037307739, + "learning_rate": 0.0005752649816551162, + "loss": 3.6957, + "step": 50010 + }, + { + "epoch": 3.3982198668297325, + "grad_norm": 0.9146090745925903, + "learning_rate": 0.0005752225166462835, + "loss": 3.6, + "step": 50015 + }, + { + "epoch": 3.398559586900394, + "grad_norm": 0.9287815093994141, + "learning_rate": 0.0005751800516374508, + "loss": 3.435, + "step": 50020 + }, + { + "epoch": 3.398899306971056, + "grad_norm": 1.2030836343765259, + "learning_rate": 0.0005751375866286181, + "loss": 3.7039, + "step": 50025 + }, + { + "epoch": 3.3992390270417174, + "grad_norm": 0.7052626609802246, + "learning_rate": 0.0005750951216197853, + "loss": 3.4021, + "step": 50030 + }, + { + "epoch": 3.3995787471123795, + "grad_norm": 0.8509361147880554, + "learning_rate": 0.0005750526566109526, + "loss": 3.5227, + "step": 50035 + }, + { + "epoch": 3.399918467183041, + "grad_norm": 0.9617079496383667, + "learning_rate": 0.0005750101916021198, + "loss": 3.6982, + "step": 50040 + }, + { + "epoch": 3.4002581872537028, + "grad_norm": 0.9806979894638062, + "learning_rate": 0.0005749677265932871, + "loss": 3.4404, + "step": 50045 + }, + { + "epoch": 3.400597907324365, + "grad_norm": 0.7971715331077576, + "learning_rate": 0.0005749252615844545, + "loss": 3.5532, + "step": 50050 + }, + { + "epoch": 3.4009376273950265, + "grad_norm": 0.9662729501724243, + "learning_rate": 0.0005748827965756217, + "loss": 3.5923, + "step": 50055 + }, + { + "epoch": 3.401277347465688, + "grad_norm": 0.6705602407455444, + "learning_rate": 0.000574840331566789, + "loss": 3.3271, + "step": 50060 + }, + { + "epoch": 3.40161706753635, + "grad_norm": 0.8430922031402588, + "learning_rate": 0.0005747978665579563, + "loss": 3.5735, + "step": 50065 + }, + { + "epoch": 3.401956787607012, + "grad_norm": 0.9195185899734497, + "learning_rate": 0.0005747554015491235, + "loss": 3.4749, + "step": 50070 + }, + { + "epoch": 3.4022965076776734, + "grad_norm": 1.0209014415740967, + "learning_rate": 0.0005747129365402907, + "loss": 3.5843, + "step": 50075 + }, + { + "epoch": 3.4026362277483355, + "grad_norm": 0.8268157243728638, + "learning_rate": 0.0005746704715314581, + "loss": 3.2724, + "step": 50080 + }, + { + "epoch": 3.402975947818997, + "grad_norm": 0.8612752556800842, + "learning_rate": 0.0005746280065226254, + "loss": 3.8027, + "step": 50085 + }, + { + "epoch": 3.403315667889659, + "grad_norm": 0.7740690112113953, + "learning_rate": 0.0005745855415137926, + "loss": 3.4838, + "step": 50090 + }, + { + "epoch": 3.403655387960321, + "grad_norm": 1.2786152362823486, + "learning_rate": 0.00057454307650496, + "loss": 3.735, + "step": 50095 + }, + { + "epoch": 3.4039951080309825, + "grad_norm": 0.8987647294998169, + "learning_rate": 0.0005745006114961272, + "loss": 3.5005, + "step": 50100 + }, + { + "epoch": 3.404334828101644, + "grad_norm": 0.9870299100875854, + "learning_rate": 0.0005744581464872944, + "loss": 3.588, + "step": 50105 + }, + { + "epoch": 3.404674548172306, + "grad_norm": 0.8393194079399109, + "learning_rate": 0.0005744156814784618, + "loss": 3.7215, + "step": 50110 + }, + { + "epoch": 3.405014268242968, + "grad_norm": 0.9247740507125854, + "learning_rate": 0.000574373216469629, + "loss": 3.6053, + "step": 50115 + }, + { + "epoch": 3.4053539883136295, + "grad_norm": 0.859395444393158, + "learning_rate": 0.0005743307514607963, + "loss": 3.6522, + "step": 50120 + }, + { + "epoch": 3.4056937083842915, + "grad_norm": 0.9788318276405334, + "learning_rate": 0.0005742882864519637, + "loss": 3.5815, + "step": 50125 + }, + { + "epoch": 3.406033428454953, + "grad_norm": 0.8912394642829895, + "learning_rate": 0.0005742458214431309, + "loss": 3.7071, + "step": 50130 + }, + { + "epoch": 3.406373148525615, + "grad_norm": 0.9066275358200073, + "learning_rate": 0.0005742033564342981, + "loss": 3.4305, + "step": 50135 + }, + { + "epoch": 3.4067128685962764, + "grad_norm": 0.8749808073043823, + "learning_rate": 0.0005741608914254654, + "loss": 3.595, + "step": 50140 + }, + { + "epoch": 3.4070525886669385, + "grad_norm": 0.939750611782074, + "learning_rate": 0.0005741184264166327, + "loss": 3.4813, + "step": 50145 + }, + { + "epoch": 3.4073923087376, + "grad_norm": 0.9295110702514648, + "learning_rate": 0.0005740759614077999, + "loss": 3.5136, + "step": 50150 + }, + { + "epoch": 3.4077320288082618, + "grad_norm": 0.9714298844337463, + "learning_rate": 0.0005740334963989673, + "loss": 3.7413, + "step": 50155 + }, + { + "epoch": 3.408071748878924, + "grad_norm": 0.8394303917884827, + "learning_rate": 0.0005739910313901346, + "loss": 3.4702, + "step": 50160 + }, + { + "epoch": 3.4084114689495855, + "grad_norm": 0.9525669813156128, + "learning_rate": 0.0005739485663813018, + "loss": 3.7708, + "step": 50165 + }, + { + "epoch": 3.408751189020247, + "grad_norm": 0.8759309649467468, + "learning_rate": 0.0005739061013724691, + "loss": 3.5869, + "step": 50170 + }, + { + "epoch": 3.409090909090909, + "grad_norm": 0.7641010284423828, + "learning_rate": 0.0005738636363636364, + "loss": 3.4793, + "step": 50175 + }, + { + "epoch": 3.409430629161571, + "grad_norm": 0.768287718296051, + "learning_rate": 0.0005738211713548036, + "loss": 3.4763, + "step": 50180 + }, + { + "epoch": 3.4097703492322324, + "grad_norm": 0.9428665637969971, + "learning_rate": 0.0005737787063459709, + "loss": 3.524, + "step": 50185 + }, + { + "epoch": 3.4101100693028945, + "grad_norm": 0.9918295741081238, + "learning_rate": 0.0005737362413371382, + "loss": 3.4502, + "step": 50190 + }, + { + "epoch": 3.410449789373556, + "grad_norm": 0.814432680606842, + "learning_rate": 0.0005736937763283055, + "loss": 3.3258, + "step": 50195 + }, + { + "epoch": 3.4107895094442178, + "grad_norm": 1.043674349784851, + "learning_rate": 0.0005736513113194728, + "loss": 3.7549, + "step": 50200 + }, + { + "epoch": 3.41112922951488, + "grad_norm": 1.3936353921890259, + "learning_rate": 0.00057360884631064, + "loss": 3.7159, + "step": 50205 + }, + { + "epoch": 3.4114689495855415, + "grad_norm": 0.7998273372650146, + "learning_rate": 0.0005735663813018073, + "loss": 3.2901, + "step": 50210 + }, + { + "epoch": 3.411808669656203, + "grad_norm": 0.8622553944587708, + "learning_rate": 0.0005735239162929746, + "loss": 3.6911, + "step": 50215 + }, + { + "epoch": 3.412148389726865, + "grad_norm": 0.8285670876502991, + "learning_rate": 0.0005734814512841418, + "loss": 3.5615, + "step": 50220 + }, + { + "epoch": 3.412488109797527, + "grad_norm": 1.0378657579421997, + "learning_rate": 0.0005734389862753092, + "loss": 3.439, + "step": 50225 + }, + { + "epoch": 3.4128278298681884, + "grad_norm": 1.3652055263519287, + "learning_rate": 0.0005733965212664765, + "loss": 3.4319, + "step": 50230 + }, + { + "epoch": 3.4131675499388505, + "grad_norm": 0.9098742008209229, + "learning_rate": 0.0005733540562576437, + "loss": 3.3168, + "step": 50235 + }, + { + "epoch": 3.413507270009512, + "grad_norm": 0.8177469968795776, + "learning_rate": 0.0005733115912488109, + "loss": 3.5476, + "step": 50240 + }, + { + "epoch": 3.413846990080174, + "grad_norm": 0.9842658638954163, + "learning_rate": 0.0005732691262399783, + "loss": 3.3489, + "step": 50245 + }, + { + "epoch": 3.414186710150836, + "grad_norm": 0.8692606687545776, + "learning_rate": 0.0005732266612311455, + "loss": 3.6411, + "step": 50250 + }, + { + "epoch": 3.4145264302214975, + "grad_norm": 0.7734696269035339, + "learning_rate": 0.0005731841962223128, + "loss": 3.39, + "step": 50255 + }, + { + "epoch": 3.414866150292159, + "grad_norm": 0.9279956817626953, + "learning_rate": 0.0005731417312134802, + "loss": 3.5193, + "step": 50260 + }, + { + "epoch": 3.415205870362821, + "grad_norm": 0.8650059103965759, + "learning_rate": 0.0005730992662046474, + "loss": 3.7323, + "step": 50265 + }, + { + "epoch": 3.415545590433483, + "grad_norm": 0.9717267751693726, + "learning_rate": 0.0005730568011958147, + "loss": 3.6155, + "step": 50270 + }, + { + "epoch": 3.4158853105041445, + "grad_norm": 1.0183500051498413, + "learning_rate": 0.000573014336186982, + "loss": 3.5613, + "step": 50275 + }, + { + "epoch": 3.4162250305748065, + "grad_norm": 0.7673061490058899, + "learning_rate": 0.0005729718711781492, + "loss": 3.7083, + "step": 50280 + }, + { + "epoch": 3.416564750645468, + "grad_norm": 0.8956186771392822, + "learning_rate": 0.0005729294061693165, + "loss": 3.4953, + "step": 50285 + }, + { + "epoch": 3.41690447071613, + "grad_norm": 0.8078187704086304, + "learning_rate": 0.0005728869411604837, + "loss": 3.6639, + "step": 50290 + }, + { + "epoch": 3.417244190786792, + "grad_norm": 0.8220173120498657, + "learning_rate": 0.0005728444761516511, + "loss": 3.4842, + "step": 50295 + }, + { + "epoch": 3.4175839108574535, + "grad_norm": 0.846361517906189, + "learning_rate": 0.0005728020111428184, + "loss": 3.4141, + "step": 50300 + }, + { + "epoch": 3.417923630928115, + "grad_norm": 0.7822868227958679, + "learning_rate": 0.0005727595461339856, + "loss": 3.3812, + "step": 50305 + }, + { + "epoch": 3.418263350998777, + "grad_norm": 0.8164373636245728, + "learning_rate": 0.0005727170811251529, + "loss": 3.635, + "step": 50310 + }, + { + "epoch": 3.418603071069439, + "grad_norm": 0.876943826675415, + "learning_rate": 0.0005726746161163202, + "loss": 3.3361, + "step": 50315 + }, + { + "epoch": 3.4189427911401005, + "grad_norm": 0.909516453742981, + "learning_rate": 0.0005726321511074874, + "loss": 3.604, + "step": 50320 + }, + { + "epoch": 3.4192825112107625, + "grad_norm": 0.977600634098053, + "learning_rate": 0.0005725896860986546, + "loss": 3.4293, + "step": 50325 + }, + { + "epoch": 3.419622231281424, + "grad_norm": 0.9885809421539307, + "learning_rate": 0.0005725472210898221, + "loss": 3.7721, + "step": 50330 + }, + { + "epoch": 3.419961951352086, + "grad_norm": 0.7592830657958984, + "learning_rate": 0.0005725047560809893, + "loss": 3.362, + "step": 50335 + }, + { + "epoch": 3.420301671422748, + "grad_norm": 0.9529237151145935, + "learning_rate": 0.0005724622910721565, + "loss": 3.7546, + "step": 50340 + }, + { + "epoch": 3.4206413914934095, + "grad_norm": 1.0991103649139404, + "learning_rate": 0.0005724198260633239, + "loss": 3.6324, + "step": 50345 + }, + { + "epoch": 3.420981111564071, + "grad_norm": 0.9099186658859253, + "learning_rate": 0.0005723773610544911, + "loss": 3.5146, + "step": 50350 + }, + { + "epoch": 3.421320831634733, + "grad_norm": 0.9755991101264954, + "learning_rate": 0.0005723348960456583, + "loss": 3.5894, + "step": 50355 + }, + { + "epoch": 3.421660551705395, + "grad_norm": 1.043871283531189, + "learning_rate": 0.0005722924310368257, + "loss": 3.4235, + "step": 50360 + }, + { + "epoch": 3.4220002717760565, + "grad_norm": 0.7539039850234985, + "learning_rate": 0.000572249966027993, + "loss": 3.497, + "step": 50365 + }, + { + "epoch": 3.422339991846718, + "grad_norm": 0.9017581939697266, + "learning_rate": 0.0005722075010191602, + "loss": 3.4785, + "step": 50370 + }, + { + "epoch": 3.42267971191738, + "grad_norm": 0.9821982979774475, + "learning_rate": 0.0005721650360103276, + "loss": 3.6954, + "step": 50375 + }, + { + "epoch": 3.423019431988042, + "grad_norm": 1.0090309381484985, + "learning_rate": 0.0005721225710014948, + "loss": 3.2802, + "step": 50380 + }, + { + "epoch": 3.4233591520587034, + "grad_norm": 1.425331711769104, + "learning_rate": 0.000572080105992662, + "loss": 3.7522, + "step": 50385 + }, + { + "epoch": 3.4236988721293655, + "grad_norm": 0.8659571409225464, + "learning_rate": 0.0005720376409838293, + "loss": 3.4554, + "step": 50390 + }, + { + "epoch": 3.424038592200027, + "grad_norm": 0.767653226852417, + "learning_rate": 0.0005719951759749966, + "loss": 3.4911, + "step": 50395 + }, + { + "epoch": 3.424378312270689, + "grad_norm": 0.8002367615699768, + "learning_rate": 0.0005719527109661639, + "loss": 3.5028, + "step": 50400 + }, + { + "epoch": 3.424718032341351, + "grad_norm": 0.8083398938179016, + "learning_rate": 0.0005719102459573312, + "loss": 3.3622, + "step": 50405 + }, + { + "epoch": 3.4250577524120125, + "grad_norm": 1.0708001852035522, + "learning_rate": 0.0005718677809484985, + "loss": 3.5909, + "step": 50410 + }, + { + "epoch": 3.425397472482674, + "grad_norm": 0.9617190361022949, + "learning_rate": 0.0005718253159396657, + "loss": 3.3842, + "step": 50415 + }, + { + "epoch": 3.425737192553336, + "grad_norm": 0.8474251627922058, + "learning_rate": 0.000571782850930833, + "loss": 3.4895, + "step": 50420 + }, + { + "epoch": 3.426076912623998, + "grad_norm": 0.9659502506256104, + "learning_rate": 0.0005717403859220002, + "loss": 3.6542, + "step": 50425 + }, + { + "epoch": 3.4264166326946595, + "grad_norm": 0.8797829747200012, + "learning_rate": 0.0005716979209131675, + "loss": 3.5711, + "step": 50430 + }, + { + "epoch": 3.4267563527653215, + "grad_norm": 0.7151371240615845, + "learning_rate": 0.0005716554559043349, + "loss": 3.3297, + "step": 50435 + }, + { + "epoch": 3.427096072835983, + "grad_norm": 1.003941297531128, + "learning_rate": 0.0005716129908955021, + "loss": 3.3435, + "step": 50440 + }, + { + "epoch": 3.427435792906645, + "grad_norm": 0.868050217628479, + "learning_rate": 0.0005715705258866694, + "loss": 3.5076, + "step": 50445 + }, + { + "epoch": 3.427775512977307, + "grad_norm": 0.9177842140197754, + "learning_rate": 0.0005715280608778367, + "loss": 3.3433, + "step": 50450 + }, + { + "epoch": 3.4281152330479685, + "grad_norm": 0.8501611948013306, + "learning_rate": 0.0005714855958690039, + "loss": 3.4678, + "step": 50455 + }, + { + "epoch": 3.42845495311863, + "grad_norm": 0.7744552493095398, + "learning_rate": 0.0005714431308601712, + "loss": 3.7013, + "step": 50460 + }, + { + "epoch": 3.428794673189292, + "grad_norm": 1.0888409614562988, + "learning_rate": 0.0005714006658513386, + "loss": 3.2259, + "step": 50465 + }, + { + "epoch": 3.429134393259954, + "grad_norm": 0.8661327362060547, + "learning_rate": 0.0005713582008425058, + "loss": 3.5107, + "step": 50470 + }, + { + "epoch": 3.4294741133306155, + "grad_norm": 0.8520111441612244, + "learning_rate": 0.000571315735833673, + "loss": 3.5699, + "step": 50475 + }, + { + "epoch": 3.429813833401277, + "grad_norm": 0.8589596152305603, + "learning_rate": 0.0005712732708248404, + "loss": 3.5752, + "step": 50480 + }, + { + "epoch": 3.430153553471939, + "grad_norm": 0.8942217826843262, + "learning_rate": 0.0005712308058160076, + "loss": 3.5982, + "step": 50485 + }, + { + "epoch": 3.430493273542601, + "grad_norm": 0.6301111578941345, + "learning_rate": 0.0005711883408071748, + "loss": 3.5326, + "step": 50490 + }, + { + "epoch": 3.4308329936132624, + "grad_norm": 1.0892735719680786, + "learning_rate": 0.0005711458757983422, + "loss": 3.3951, + "step": 50495 + }, + { + "epoch": 3.4311727136839245, + "grad_norm": 0.7894288301467896, + "learning_rate": 0.0005711034107895095, + "loss": 3.7086, + "step": 50500 + }, + { + "epoch": 3.431512433754586, + "grad_norm": 0.74152672290802, + "learning_rate": 0.0005710609457806767, + "loss": 3.4952, + "step": 50505 + }, + { + "epoch": 3.4318521538252478, + "grad_norm": 0.8731594681739807, + "learning_rate": 0.0005710184807718441, + "loss": 3.3273, + "step": 50510 + }, + { + "epoch": 3.43219187389591, + "grad_norm": 1.0777064561843872, + "learning_rate": 0.0005709760157630113, + "loss": 3.5356, + "step": 50515 + }, + { + "epoch": 3.4325315939665715, + "grad_norm": 1.0710819959640503, + "learning_rate": 0.0005709335507541785, + "loss": 3.5413, + "step": 50520 + }, + { + "epoch": 3.432871314037233, + "grad_norm": 0.9023421406745911, + "learning_rate": 0.0005708910857453458, + "loss": 3.1763, + "step": 50525 + }, + { + "epoch": 3.433211034107895, + "grad_norm": 0.8025294542312622, + "learning_rate": 0.0005708486207365131, + "loss": 3.5811, + "step": 50530 + }, + { + "epoch": 3.433550754178557, + "grad_norm": 1.0256314277648926, + "learning_rate": 0.0005708061557276804, + "loss": 3.286, + "step": 50535 + }, + { + "epoch": 3.4338904742492184, + "grad_norm": 0.9729945063591003, + "learning_rate": 0.0005707636907188477, + "loss": 3.5473, + "step": 50540 + }, + { + "epoch": 3.4342301943198805, + "grad_norm": 0.8616268634796143, + "learning_rate": 0.000570721225710015, + "loss": 3.4677, + "step": 50545 + }, + { + "epoch": 3.434569914390542, + "grad_norm": 0.8979142904281616, + "learning_rate": 0.0005706787607011822, + "loss": 3.6041, + "step": 50550 + }, + { + "epoch": 3.434909634461204, + "grad_norm": 0.8089253902435303, + "learning_rate": 0.0005706362956923495, + "loss": 3.455, + "step": 50555 + }, + { + "epoch": 3.435249354531866, + "grad_norm": 0.9664849042892456, + "learning_rate": 0.0005705938306835168, + "loss": 3.7483, + "step": 50560 + }, + { + "epoch": 3.4355890746025275, + "grad_norm": 0.8747547268867493, + "learning_rate": 0.000570551365674684, + "loss": 3.4744, + "step": 50565 + }, + { + "epoch": 3.435928794673189, + "grad_norm": 0.8132421970367432, + "learning_rate": 0.0005705089006658514, + "loss": 3.4626, + "step": 50570 + }, + { + "epoch": 3.436268514743851, + "grad_norm": 0.9073557257652283, + "learning_rate": 0.0005704664356570186, + "loss": 3.4309, + "step": 50575 + }, + { + "epoch": 3.436608234814513, + "grad_norm": 1.06233549118042, + "learning_rate": 0.0005704239706481859, + "loss": 3.4554, + "step": 50580 + }, + { + "epoch": 3.4369479548851745, + "grad_norm": 1.2398866415023804, + "learning_rate": 0.0005703815056393532, + "loss": 3.6278, + "step": 50585 + }, + { + "epoch": 3.4372876749558365, + "grad_norm": 1.1009074449539185, + "learning_rate": 0.0005703390406305204, + "loss": 3.5756, + "step": 50590 + }, + { + "epoch": 3.437627395026498, + "grad_norm": 0.9235627055168152, + "learning_rate": 0.0005702965756216878, + "loss": 3.6776, + "step": 50595 + }, + { + "epoch": 3.43796711509716, + "grad_norm": 0.7246569395065308, + "learning_rate": 0.000570254110612855, + "loss": 3.7002, + "step": 50600 + }, + { + "epoch": 3.438306835167822, + "grad_norm": 0.7154413461685181, + "learning_rate": 0.0005702116456040223, + "loss": 3.6034, + "step": 50605 + }, + { + "epoch": 3.4386465552384835, + "grad_norm": 0.8884794116020203, + "learning_rate": 0.0005701691805951897, + "loss": 3.5206, + "step": 50610 + }, + { + "epoch": 3.438986275309145, + "grad_norm": 0.8134838342666626, + "learning_rate": 0.0005701267155863569, + "loss": 3.2, + "step": 50615 + }, + { + "epoch": 3.439325995379807, + "grad_norm": 0.9034074544906616, + "learning_rate": 0.0005700842505775241, + "loss": 3.6551, + "step": 50620 + }, + { + "epoch": 3.439665715450469, + "grad_norm": 0.8287516832351685, + "learning_rate": 0.0005700417855686914, + "loss": 3.5064, + "step": 50625 + }, + { + "epoch": 3.4400054355211305, + "grad_norm": 0.853687584400177, + "learning_rate": 0.0005699993205598587, + "loss": 3.7684, + "step": 50630 + }, + { + "epoch": 3.4403451555917925, + "grad_norm": 0.8786543607711792, + "learning_rate": 0.0005699568555510259, + "loss": 3.4802, + "step": 50635 + }, + { + "epoch": 3.440684875662454, + "grad_norm": 0.811477780342102, + "learning_rate": 0.0005699143905421933, + "loss": 3.5606, + "step": 50640 + }, + { + "epoch": 3.441024595733116, + "grad_norm": 1.0445820093154907, + "learning_rate": 0.0005698719255333606, + "loss": 3.7642, + "step": 50645 + }, + { + "epoch": 3.441364315803778, + "grad_norm": 0.8884252309799194, + "learning_rate": 0.0005698294605245278, + "loss": 3.4542, + "step": 50650 + }, + { + "epoch": 3.4417040358744395, + "grad_norm": 1.108083963394165, + "learning_rate": 0.0005697869955156951, + "loss": 3.7055, + "step": 50655 + }, + { + "epoch": 3.442043755945101, + "grad_norm": 0.7667628526687622, + "learning_rate": 0.0005697445305068624, + "loss": 3.5844, + "step": 50660 + }, + { + "epoch": 3.442383476015763, + "grad_norm": 0.5906009078025818, + "learning_rate": 0.0005697020654980296, + "loss": 3.764, + "step": 50665 + }, + { + "epoch": 3.442723196086425, + "grad_norm": 0.8602977395057678, + "learning_rate": 0.0005696596004891969, + "loss": 3.5556, + "step": 50670 + }, + { + "epoch": 3.4430629161570865, + "grad_norm": 0.8311915993690491, + "learning_rate": 0.0005696171354803643, + "loss": 3.4764, + "step": 50675 + }, + { + "epoch": 3.4434026362277486, + "grad_norm": 0.8537005186080933, + "learning_rate": 0.0005695746704715315, + "loss": 3.5944, + "step": 50680 + }, + { + "epoch": 3.44374235629841, + "grad_norm": 0.8184798955917358, + "learning_rate": 0.0005695322054626988, + "loss": 3.6164, + "step": 50685 + }, + { + "epoch": 3.444082076369072, + "grad_norm": 0.9553748965263367, + "learning_rate": 0.000569489740453866, + "loss": 3.6228, + "step": 50690 + }, + { + "epoch": 3.444421796439734, + "grad_norm": 0.8695975542068481, + "learning_rate": 0.0005694472754450333, + "loss": 3.5587, + "step": 50695 + }, + { + "epoch": 3.4447615165103955, + "grad_norm": 0.7541364431381226, + "learning_rate": 0.0005694048104362006, + "loss": 3.4542, + "step": 50700 + }, + { + "epoch": 3.445101236581057, + "grad_norm": 0.8585219383239746, + "learning_rate": 0.0005693623454273678, + "loss": 3.4124, + "step": 50705 + }, + { + "epoch": 3.4454409566517192, + "grad_norm": 0.9003470540046692, + "learning_rate": 0.0005693198804185352, + "loss": 3.6391, + "step": 50710 + }, + { + "epoch": 3.445780676722381, + "grad_norm": 1.042048692703247, + "learning_rate": 0.0005692774154097025, + "loss": 3.7485, + "step": 50715 + }, + { + "epoch": 3.4461203967930425, + "grad_norm": 0.8912600874900818, + "learning_rate": 0.0005692349504008697, + "loss": 3.6337, + "step": 50720 + }, + { + "epoch": 3.446460116863704, + "grad_norm": 0.7305474281311035, + "learning_rate": 0.0005691924853920369, + "loss": 3.7329, + "step": 50725 + }, + { + "epoch": 3.446799836934366, + "grad_norm": 0.8587639927864075, + "learning_rate": 0.0005691500203832043, + "loss": 3.8979, + "step": 50730 + }, + { + "epoch": 3.447139557005028, + "grad_norm": 1.2681081295013428, + "learning_rate": 0.0005691075553743715, + "loss": 3.3609, + "step": 50735 + }, + { + "epoch": 3.4474792770756895, + "grad_norm": 0.7709375619888306, + "learning_rate": 0.0005690650903655387, + "loss": 3.5286, + "step": 50740 + }, + { + "epoch": 3.4478189971463515, + "grad_norm": 0.7426081299781799, + "learning_rate": 0.0005690226253567062, + "loss": 3.2639, + "step": 50745 + }, + { + "epoch": 3.448158717217013, + "grad_norm": 1.4819930791854858, + "learning_rate": 0.0005689801603478734, + "loss": 3.539, + "step": 50750 + }, + { + "epoch": 3.448498437287675, + "grad_norm": 0.8509702682495117, + "learning_rate": 0.0005689376953390406, + "loss": 3.4811, + "step": 50755 + }, + { + "epoch": 3.448838157358337, + "grad_norm": 0.8367165327072144, + "learning_rate": 0.000568895230330208, + "loss": 3.3323, + "step": 50760 + }, + { + "epoch": 3.4491778774289985, + "grad_norm": 0.9964286684989929, + "learning_rate": 0.0005688527653213752, + "loss": 3.239, + "step": 50765 + }, + { + "epoch": 3.44951759749966, + "grad_norm": 1.1147127151489258, + "learning_rate": 0.0005688103003125424, + "loss": 3.5745, + "step": 50770 + }, + { + "epoch": 3.449857317570322, + "grad_norm": 0.9712830781936646, + "learning_rate": 0.0005687678353037097, + "loss": 3.3486, + "step": 50775 + }, + { + "epoch": 3.450197037640984, + "grad_norm": 0.8550812005996704, + "learning_rate": 0.0005687253702948771, + "loss": 3.3398, + "step": 50780 + }, + { + "epoch": 3.4505367577116455, + "grad_norm": 0.8620516061782837, + "learning_rate": 0.0005686829052860443, + "loss": 3.5049, + "step": 50785 + }, + { + "epoch": 3.4508764777823075, + "grad_norm": 0.8977075219154358, + "learning_rate": 0.0005686404402772116, + "loss": 3.8658, + "step": 50790 + }, + { + "epoch": 3.451216197852969, + "grad_norm": 1.2573661804199219, + "learning_rate": 0.0005685979752683789, + "loss": 3.4595, + "step": 50795 + }, + { + "epoch": 3.451555917923631, + "grad_norm": 0.8379766941070557, + "learning_rate": 0.0005685555102595461, + "loss": 3.7034, + "step": 50800 + }, + { + "epoch": 3.451895637994293, + "grad_norm": 0.9348949193954468, + "learning_rate": 0.0005685130452507134, + "loss": 3.574, + "step": 50805 + }, + { + "epoch": 3.4522353580649545, + "grad_norm": 0.8916747570037842, + "learning_rate": 0.0005684705802418806, + "loss": 3.5446, + "step": 50810 + }, + { + "epoch": 3.452575078135616, + "grad_norm": 0.7410023808479309, + "learning_rate": 0.000568428115233048, + "loss": 3.692, + "step": 50815 + }, + { + "epoch": 3.452914798206278, + "grad_norm": 0.8096827268600464, + "learning_rate": 0.0005683856502242153, + "loss": 3.4863, + "step": 50820 + }, + { + "epoch": 3.45325451827694, + "grad_norm": 0.701348066329956, + "learning_rate": 0.0005683431852153825, + "loss": 3.5798, + "step": 50825 + }, + { + "epoch": 3.4535942383476015, + "grad_norm": 0.9151264429092407, + "learning_rate": 0.0005683007202065498, + "loss": 3.5321, + "step": 50830 + }, + { + "epoch": 3.453933958418263, + "grad_norm": 0.8722145557403564, + "learning_rate": 0.0005682582551977171, + "loss": 3.3863, + "step": 50835 + }, + { + "epoch": 3.454273678488925, + "grad_norm": 0.9006122946739197, + "learning_rate": 0.0005682157901888843, + "loss": 3.5009, + "step": 50840 + }, + { + "epoch": 3.454613398559587, + "grad_norm": 0.8197031021118164, + "learning_rate": 0.0005681733251800516, + "loss": 3.4763, + "step": 50845 + }, + { + "epoch": 3.4549531186302485, + "grad_norm": 0.8744727969169617, + "learning_rate": 0.000568130860171219, + "loss": 3.5832, + "step": 50850 + }, + { + "epoch": 3.4552928387009105, + "grad_norm": 1.117315649986267, + "learning_rate": 0.0005680883951623862, + "loss": 3.6203, + "step": 50855 + }, + { + "epoch": 3.455632558771572, + "grad_norm": 0.7963263988494873, + "learning_rate": 0.0005680459301535535, + "loss": 3.3261, + "step": 50860 + }, + { + "epoch": 3.455972278842234, + "grad_norm": 0.9327932596206665, + "learning_rate": 0.0005680034651447208, + "loss": 3.4591, + "step": 50865 + }, + { + "epoch": 3.456311998912896, + "grad_norm": 0.9840716123580933, + "learning_rate": 0.000567961000135888, + "loss": 3.557, + "step": 50870 + }, + { + "epoch": 3.4566517189835575, + "grad_norm": 1.1771697998046875, + "learning_rate": 0.0005679185351270552, + "loss": 3.5785, + "step": 50875 + }, + { + "epoch": 3.456991439054219, + "grad_norm": 0.9164397716522217, + "learning_rate": 0.0005678760701182226, + "loss": 3.5117, + "step": 50880 + }, + { + "epoch": 3.457331159124881, + "grad_norm": 0.9693806767463684, + "learning_rate": 0.0005678336051093899, + "loss": 3.4547, + "step": 50885 + }, + { + "epoch": 3.457670879195543, + "grad_norm": 0.9002485275268555, + "learning_rate": 0.0005677911401005571, + "loss": 3.6228, + "step": 50890 + }, + { + "epoch": 3.4580105992662045, + "grad_norm": 1.138275384902954, + "learning_rate": 0.0005677486750917245, + "loss": 3.297, + "step": 50895 + }, + { + "epoch": 3.4583503193368665, + "grad_norm": 0.8543419241905212, + "learning_rate": 0.0005677062100828917, + "loss": 3.6483, + "step": 50900 + }, + { + "epoch": 3.458690039407528, + "grad_norm": 0.8848923444747925, + "learning_rate": 0.0005676637450740589, + "loss": 3.6183, + "step": 50905 + }, + { + "epoch": 3.45902975947819, + "grad_norm": 0.9100379943847656, + "learning_rate": 0.0005676212800652263, + "loss": 3.521, + "step": 50910 + }, + { + "epoch": 3.459369479548852, + "grad_norm": 1.7573810815811157, + "learning_rate": 0.0005675788150563935, + "loss": 3.5461, + "step": 50915 + }, + { + "epoch": 3.4597091996195135, + "grad_norm": 1.0431642532348633, + "learning_rate": 0.0005675363500475608, + "loss": 3.4086, + "step": 50920 + }, + { + "epoch": 3.460048919690175, + "grad_norm": 0.9521357417106628, + "learning_rate": 0.0005674938850387281, + "loss": 3.3599, + "step": 50925 + }, + { + "epoch": 3.460388639760837, + "grad_norm": 0.7850194573402405, + "learning_rate": 0.0005674514200298954, + "loss": 3.2531, + "step": 50930 + }, + { + "epoch": 3.460728359831499, + "grad_norm": 1.3494582176208496, + "learning_rate": 0.0005674089550210627, + "loss": 3.6451, + "step": 50935 + }, + { + "epoch": 3.4610680799021605, + "grad_norm": 0.869604766368866, + "learning_rate": 0.0005673664900122299, + "loss": 3.4525, + "step": 50940 + }, + { + "epoch": 3.4614077999728226, + "grad_norm": 0.8547475337982178, + "learning_rate": 0.0005673240250033972, + "loss": 3.6296, + "step": 50945 + }, + { + "epoch": 3.461747520043484, + "grad_norm": 0.8615207076072693, + "learning_rate": 0.0005672815599945645, + "loss": 3.5508, + "step": 50950 + }, + { + "epoch": 3.462087240114146, + "grad_norm": 0.8000222444534302, + "learning_rate": 0.0005672390949857318, + "loss": 3.4841, + "step": 50955 + }, + { + "epoch": 3.462426960184808, + "grad_norm": 0.7710859179496765, + "learning_rate": 0.000567196629976899, + "loss": 3.7641, + "step": 50960 + }, + { + "epoch": 3.4627666802554695, + "grad_norm": 1.0300705432891846, + "learning_rate": 0.0005671541649680664, + "loss": 3.5524, + "step": 50965 + }, + { + "epoch": 3.463106400326131, + "grad_norm": 0.837216317653656, + "learning_rate": 0.0005671116999592336, + "loss": 3.4492, + "step": 50970 + }, + { + "epoch": 3.4634461203967932, + "grad_norm": 1.0632215738296509, + "learning_rate": 0.0005670692349504008, + "loss": 3.636, + "step": 50975 + }, + { + "epoch": 3.463785840467455, + "grad_norm": 0.9279181957244873, + "learning_rate": 0.0005670267699415682, + "loss": 3.5595, + "step": 50980 + }, + { + "epoch": 3.4641255605381165, + "grad_norm": 0.8463853597640991, + "learning_rate": 0.0005669843049327354, + "loss": 3.5197, + "step": 50985 + }, + { + "epoch": 3.4644652806087786, + "grad_norm": 1.0096808671951294, + "learning_rate": 0.0005669418399239027, + "loss": 3.2733, + "step": 50990 + }, + { + "epoch": 3.46480500067944, + "grad_norm": 0.8177899122238159, + "learning_rate": 0.0005668993749150701, + "loss": 3.5502, + "step": 50995 + }, + { + "epoch": 3.465144720750102, + "grad_norm": 0.6265968680381775, + "learning_rate": 0.0005668569099062373, + "loss": 3.8016, + "step": 51000 + }, + { + "epoch": 3.465484440820764, + "grad_norm": 0.7518759965896606, + "learning_rate": 0.0005668144448974045, + "loss": 3.8827, + "step": 51005 + }, + { + "epoch": 3.4658241608914255, + "grad_norm": 0.9135801196098328, + "learning_rate": 0.0005667719798885719, + "loss": 3.4915, + "step": 51010 + }, + { + "epoch": 3.466163880962087, + "grad_norm": 1.0301766395568848, + "learning_rate": 0.0005667295148797391, + "loss": 3.5549, + "step": 51015 + }, + { + "epoch": 3.4665036010327492, + "grad_norm": 1.0821129083633423, + "learning_rate": 0.0005666870498709063, + "loss": 3.4086, + "step": 51020 + }, + { + "epoch": 3.466843321103411, + "grad_norm": 1.0367754697799683, + "learning_rate": 0.0005666445848620737, + "loss": 3.5576, + "step": 51025 + }, + { + "epoch": 3.4671830411740725, + "grad_norm": 0.6967639923095703, + "learning_rate": 0.000566602119853241, + "loss": 3.5741, + "step": 51030 + }, + { + "epoch": 3.4675227612447346, + "grad_norm": 0.9032452702522278, + "learning_rate": 0.0005665596548444082, + "loss": 3.6342, + "step": 51035 + }, + { + "epoch": 3.467862481315396, + "grad_norm": 1.0133156776428223, + "learning_rate": 0.0005665171898355755, + "loss": 3.6274, + "step": 51040 + }, + { + "epoch": 3.468202201386058, + "grad_norm": 0.9498702883720398, + "learning_rate": 0.0005664747248267428, + "loss": 3.4234, + "step": 51045 + }, + { + "epoch": 3.46854192145672, + "grad_norm": 0.9075527191162109, + "learning_rate": 0.00056643225981791, + "loss": 3.6376, + "step": 51050 + }, + { + "epoch": 3.4688816415273815, + "grad_norm": 0.8078995943069458, + "learning_rate": 0.0005663897948090774, + "loss": 3.7091, + "step": 51055 + }, + { + "epoch": 3.469221361598043, + "grad_norm": 1.065435767173767, + "learning_rate": 0.0005663473298002447, + "loss": 3.4035, + "step": 51060 + }, + { + "epoch": 3.469561081668705, + "grad_norm": 0.8498144745826721, + "learning_rate": 0.0005663048647914119, + "loss": 3.3855, + "step": 51065 + }, + { + "epoch": 3.469900801739367, + "grad_norm": 0.9964304566383362, + "learning_rate": 0.0005662623997825792, + "loss": 3.4021, + "step": 51070 + }, + { + "epoch": 3.4702405218100285, + "grad_norm": 0.815377414226532, + "learning_rate": 0.0005662199347737464, + "loss": 3.3272, + "step": 51075 + }, + { + "epoch": 3.47058024188069, + "grad_norm": 1.0796416997909546, + "learning_rate": 0.0005661774697649137, + "loss": 3.5905, + "step": 51080 + }, + { + "epoch": 3.470919961951352, + "grad_norm": 0.999350368976593, + "learning_rate": 0.000566135004756081, + "loss": 3.5277, + "step": 51085 + }, + { + "epoch": 3.471259682022014, + "grad_norm": 0.9466595649719238, + "learning_rate": 0.0005660925397472483, + "loss": 3.7873, + "step": 51090 + }, + { + "epoch": 3.4715994020926755, + "grad_norm": 0.7036259174346924, + "learning_rate": 0.0005660500747384156, + "loss": 3.4771, + "step": 51095 + }, + { + "epoch": 3.4719391221633376, + "grad_norm": 0.982236385345459, + "learning_rate": 0.0005660076097295829, + "loss": 3.5734, + "step": 51100 + }, + { + "epoch": 3.472278842233999, + "grad_norm": 0.9965671300888062, + "learning_rate": 0.0005659651447207501, + "loss": 3.3784, + "step": 51105 + }, + { + "epoch": 3.472618562304661, + "grad_norm": 0.7863239645957947, + "learning_rate": 0.0005659226797119173, + "loss": 3.5013, + "step": 51110 + }, + { + "epoch": 3.472958282375323, + "grad_norm": 0.7466006278991699, + "learning_rate": 0.0005658802147030847, + "loss": 3.5878, + "step": 51115 + }, + { + "epoch": 3.4732980024459845, + "grad_norm": 0.8626487851142883, + "learning_rate": 0.0005658377496942519, + "loss": 3.6068, + "step": 51120 + }, + { + "epoch": 3.473637722516646, + "grad_norm": 1.3672266006469727, + "learning_rate": 0.0005657952846854192, + "loss": 3.5629, + "step": 51125 + }, + { + "epoch": 3.4739774425873082, + "grad_norm": 0.9647884964942932, + "learning_rate": 0.0005657528196765866, + "loss": 3.5314, + "step": 51130 + }, + { + "epoch": 3.47431716265797, + "grad_norm": 0.8028362989425659, + "learning_rate": 0.0005657103546677538, + "loss": 3.5949, + "step": 51135 + }, + { + "epoch": 3.4746568827286315, + "grad_norm": 1.1468398571014404, + "learning_rate": 0.000565667889658921, + "loss": 3.4245, + "step": 51140 + }, + { + "epoch": 3.4749966027992936, + "grad_norm": 0.8467814326286316, + "learning_rate": 0.0005656254246500884, + "loss": 3.5896, + "step": 51145 + }, + { + "epoch": 3.475336322869955, + "grad_norm": 0.6715965867042542, + "learning_rate": 0.0005655829596412556, + "loss": 3.6635, + "step": 51150 + }, + { + "epoch": 3.475676042940617, + "grad_norm": 1.0916974544525146, + "learning_rate": 0.0005655404946324228, + "loss": 3.4277, + "step": 51155 + }, + { + "epoch": 3.4760157630112785, + "grad_norm": 0.8410515785217285, + "learning_rate": 0.0005654980296235903, + "loss": 3.6335, + "step": 51160 + }, + { + "epoch": 3.4763554830819405, + "grad_norm": 0.8266076445579529, + "learning_rate": 0.0005654555646147575, + "loss": 3.3921, + "step": 51165 + }, + { + "epoch": 3.476695203152602, + "grad_norm": 0.7660947442054749, + "learning_rate": 0.0005654130996059247, + "loss": 3.3394, + "step": 51170 + }, + { + "epoch": 3.477034923223264, + "grad_norm": 0.8132218718528748, + "learning_rate": 0.000565370634597092, + "loss": 3.3233, + "step": 51175 + }, + { + "epoch": 3.477374643293926, + "grad_norm": 0.9864709377288818, + "learning_rate": 0.0005653281695882593, + "loss": 3.6753, + "step": 51180 + }, + { + "epoch": 3.4777143633645875, + "grad_norm": 0.8380324244499207, + "learning_rate": 0.0005652857045794265, + "loss": 3.7001, + "step": 51185 + }, + { + "epoch": 3.478054083435249, + "grad_norm": 0.6473556160926819, + "learning_rate": 0.0005652432395705938, + "loss": 3.6258, + "step": 51190 + }, + { + "epoch": 3.478393803505911, + "grad_norm": 0.7766239643096924, + "learning_rate": 0.0005652007745617612, + "loss": 3.5542, + "step": 51195 + }, + { + "epoch": 3.478733523576573, + "grad_norm": 0.8013241291046143, + "learning_rate": 0.0005651583095529284, + "loss": 3.3974, + "step": 51200 + }, + { + "epoch": 3.4790732436472345, + "grad_norm": 0.7030622363090515, + "learning_rate": 0.0005651158445440957, + "loss": 3.5066, + "step": 51205 + }, + { + "epoch": 3.4794129637178965, + "grad_norm": 1.1094800233840942, + "learning_rate": 0.000565073379535263, + "loss": 3.1997, + "step": 51210 + }, + { + "epoch": 3.479752683788558, + "grad_norm": 0.9074397683143616, + "learning_rate": 0.0005650309145264302, + "loss": 3.7866, + "step": 51215 + }, + { + "epoch": 3.48009240385922, + "grad_norm": 1.0409674644470215, + "learning_rate": 0.0005649884495175975, + "loss": 3.5543, + "step": 51220 + }, + { + "epoch": 3.480432123929882, + "grad_norm": 0.9041967988014221, + "learning_rate": 0.0005649459845087647, + "loss": 3.567, + "step": 51225 + }, + { + "epoch": 3.4807718440005435, + "grad_norm": 1.072012186050415, + "learning_rate": 0.0005649035194999321, + "loss": 3.5292, + "step": 51230 + }, + { + "epoch": 3.481111564071205, + "grad_norm": 0.9603410363197327, + "learning_rate": 0.0005648610544910994, + "loss": 3.6805, + "step": 51235 + }, + { + "epoch": 3.481451284141867, + "grad_norm": 0.7793058753013611, + "learning_rate": 0.0005648185894822666, + "loss": 3.4742, + "step": 51240 + }, + { + "epoch": 3.481791004212529, + "grad_norm": 0.9922502040863037, + "learning_rate": 0.0005647761244734339, + "loss": 3.3514, + "step": 51245 + }, + { + "epoch": 3.4821307242831905, + "grad_norm": 0.9909048676490784, + "learning_rate": 0.0005647336594646012, + "loss": 3.377, + "step": 51250 + }, + { + "epoch": 3.4824704443538526, + "grad_norm": 0.7987185716629028, + "learning_rate": 0.0005646911944557684, + "loss": 3.7446, + "step": 51255 + }, + { + "epoch": 3.482810164424514, + "grad_norm": 0.8964027762413025, + "learning_rate": 0.0005646487294469356, + "loss": 3.5682, + "step": 51260 + }, + { + "epoch": 3.483149884495176, + "grad_norm": 0.7689936757087708, + "learning_rate": 0.0005646062644381031, + "loss": 3.5389, + "step": 51265 + }, + { + "epoch": 3.483489604565838, + "grad_norm": 0.8254878520965576, + "learning_rate": 0.0005645637994292703, + "loss": 3.8442, + "step": 51270 + }, + { + "epoch": 3.4838293246364995, + "grad_norm": 0.8959947228431702, + "learning_rate": 0.0005645213344204376, + "loss": 3.3892, + "step": 51275 + }, + { + "epoch": 3.484169044707161, + "grad_norm": 0.7640753388404846, + "learning_rate": 0.0005644788694116049, + "loss": 3.6694, + "step": 51280 + }, + { + "epoch": 3.4845087647778232, + "grad_norm": 0.7767606377601624, + "learning_rate": 0.0005644364044027721, + "loss": 3.4289, + "step": 51285 + }, + { + "epoch": 3.484848484848485, + "grad_norm": 1.109019160270691, + "learning_rate": 0.0005643939393939394, + "loss": 3.4552, + "step": 51290 + }, + { + "epoch": 3.4851882049191465, + "grad_norm": 0.9672184586524963, + "learning_rate": 0.0005643514743851067, + "loss": 3.4566, + "step": 51295 + }, + { + "epoch": 3.4855279249898086, + "grad_norm": 0.7632936239242554, + "learning_rate": 0.000564309009376274, + "loss": 3.4585, + "step": 51300 + }, + { + "epoch": 3.48586764506047, + "grad_norm": 0.8213794827461243, + "learning_rate": 0.0005642665443674413, + "loss": 3.8212, + "step": 51305 + }, + { + "epoch": 3.486207365131132, + "grad_norm": 0.9411114454269409, + "learning_rate": 0.0005642240793586085, + "loss": 3.3897, + "step": 51310 + }, + { + "epoch": 3.486547085201794, + "grad_norm": 0.8775694370269775, + "learning_rate": 0.0005641816143497758, + "loss": 3.6353, + "step": 51315 + }, + { + "epoch": 3.4868868052724555, + "grad_norm": 0.8697211146354675, + "learning_rate": 0.0005641391493409431, + "loss": 3.1997, + "step": 51320 + }, + { + "epoch": 3.487226525343117, + "grad_norm": 0.9775707125663757, + "learning_rate": 0.0005640966843321103, + "loss": 3.6579, + "step": 51325 + }, + { + "epoch": 3.4875662454137792, + "grad_norm": 0.840651273727417, + "learning_rate": 0.0005640542193232776, + "loss": 3.6888, + "step": 51330 + }, + { + "epoch": 3.487905965484441, + "grad_norm": 0.7616641521453857, + "learning_rate": 0.000564011754314445, + "loss": 3.6693, + "step": 51335 + }, + { + "epoch": 3.4882456855551025, + "grad_norm": 0.9862614274024963, + "learning_rate": 0.0005639692893056122, + "loss": 3.5066, + "step": 51340 + }, + { + "epoch": 3.4885854056257646, + "grad_norm": 0.7202863097190857, + "learning_rate": 0.0005639268242967795, + "loss": 3.6784, + "step": 51345 + }, + { + "epoch": 3.488925125696426, + "grad_norm": 0.8098672032356262, + "learning_rate": 0.0005638843592879468, + "loss": 3.613, + "step": 51350 + }, + { + "epoch": 3.489264845767088, + "grad_norm": 0.7527655959129333, + "learning_rate": 0.000563841894279114, + "loss": 3.4663, + "step": 51355 + }, + { + "epoch": 3.48960456583775, + "grad_norm": 0.8376598358154297, + "learning_rate": 0.0005637994292702812, + "loss": 3.3015, + "step": 51360 + }, + { + "epoch": 3.4899442859084115, + "grad_norm": 0.8786569833755493, + "learning_rate": 0.0005637569642614486, + "loss": 3.6006, + "step": 51365 + }, + { + "epoch": 3.490284005979073, + "grad_norm": 0.9088473916053772, + "learning_rate": 0.0005637144992526159, + "loss": 3.5948, + "step": 51370 + }, + { + "epoch": 3.4906237260497353, + "grad_norm": 0.7834986448287964, + "learning_rate": 0.0005636720342437831, + "loss": 3.6424, + "step": 51375 + }, + { + "epoch": 3.490963446120397, + "grad_norm": 0.8700459003448486, + "learning_rate": 0.0005636295692349505, + "loss": 3.5948, + "step": 51380 + }, + { + "epoch": 3.4913031661910585, + "grad_norm": 0.8020461797714233, + "learning_rate": 0.0005635871042261177, + "loss": 3.4943, + "step": 51385 + }, + { + "epoch": 3.4916428862617206, + "grad_norm": 0.9292786717414856, + "learning_rate": 0.0005635446392172849, + "loss": 3.3214, + "step": 51390 + }, + { + "epoch": 3.491982606332382, + "grad_norm": 0.765859842300415, + "learning_rate": 0.0005635021742084523, + "loss": 3.635, + "step": 51395 + }, + { + "epoch": 3.492322326403044, + "grad_norm": 0.9569457769393921, + "learning_rate": 0.0005634597091996195, + "loss": 3.5607, + "step": 51400 + }, + { + "epoch": 3.4926620464737055, + "grad_norm": 0.9257526397705078, + "learning_rate": 0.0005634172441907868, + "loss": 3.5275, + "step": 51405 + }, + { + "epoch": 3.4930017665443676, + "grad_norm": 1.1237123012542725, + "learning_rate": 0.0005633747791819542, + "loss": 3.5096, + "step": 51410 + }, + { + "epoch": 3.493341486615029, + "grad_norm": 0.906174898147583, + "learning_rate": 0.0005633323141731214, + "loss": 3.3716, + "step": 51415 + }, + { + "epoch": 3.493681206685691, + "grad_norm": 0.9369825720787048, + "learning_rate": 0.0005632898491642886, + "loss": 3.9223, + "step": 51420 + }, + { + "epoch": 3.494020926756353, + "grad_norm": 0.9653894901275635, + "learning_rate": 0.0005632473841554559, + "loss": 3.6036, + "step": 51425 + }, + { + "epoch": 3.4943606468270145, + "grad_norm": 0.8812628388404846, + "learning_rate": 0.0005632049191466232, + "loss": 3.5882, + "step": 51430 + }, + { + "epoch": 3.494700366897676, + "grad_norm": 0.720146119594574, + "learning_rate": 0.0005631624541377904, + "loss": 3.7216, + "step": 51435 + }, + { + "epoch": 3.4950400869683382, + "grad_norm": 0.7779410481452942, + "learning_rate": 0.0005631199891289578, + "loss": 3.3995, + "step": 51440 + }, + { + "epoch": 3.495379807039, + "grad_norm": 0.9413473606109619, + "learning_rate": 0.0005630775241201251, + "loss": 3.4882, + "step": 51445 + }, + { + "epoch": 3.4957195271096615, + "grad_norm": 0.9943309426307678, + "learning_rate": 0.0005630350591112923, + "loss": 3.3377, + "step": 51450 + }, + { + "epoch": 3.4960592471803236, + "grad_norm": 1.2260239124298096, + "learning_rate": 0.0005629925941024596, + "loss": 3.5222, + "step": 51455 + }, + { + "epoch": 3.496398967250985, + "grad_norm": 0.774575412273407, + "learning_rate": 0.0005629501290936268, + "loss": 3.6552, + "step": 51460 + }, + { + "epoch": 3.496738687321647, + "grad_norm": 0.805324912071228, + "learning_rate": 0.0005629076640847941, + "loss": 3.4472, + "step": 51465 + }, + { + "epoch": 3.497078407392309, + "grad_norm": 0.8504440188407898, + "learning_rate": 0.0005628651990759614, + "loss": 3.5397, + "step": 51470 + }, + { + "epoch": 3.4974181274629705, + "grad_norm": 0.8804094791412354, + "learning_rate": 0.0005628227340671287, + "loss": 3.3276, + "step": 51475 + }, + { + "epoch": 3.497757847533632, + "grad_norm": 0.7801376581192017, + "learning_rate": 0.000562780269058296, + "loss": 3.692, + "step": 51480 + }, + { + "epoch": 3.4980975676042942, + "grad_norm": 0.9864916801452637, + "learning_rate": 0.0005627378040494633, + "loss": 3.3714, + "step": 51485 + }, + { + "epoch": 3.498437287674956, + "grad_norm": 0.8376511931419373, + "learning_rate": 0.0005626953390406305, + "loss": 3.5077, + "step": 51490 + }, + { + "epoch": 3.4987770077456175, + "grad_norm": 0.8168814182281494, + "learning_rate": 0.0005626528740317977, + "loss": 3.4695, + "step": 51495 + }, + { + "epoch": 3.499116727816279, + "grad_norm": 0.7951503396034241, + "learning_rate": 0.0005626104090229651, + "loss": 3.3997, + "step": 51500 + }, + { + "epoch": 3.499456447886941, + "grad_norm": 1.0576317310333252, + "learning_rate": 0.0005625679440141323, + "loss": 3.5215, + "step": 51505 + }, + { + "epoch": 3.499796167957603, + "grad_norm": 0.7913901209831238, + "learning_rate": 0.0005625254790052996, + "loss": 3.7335, + "step": 51510 + }, + { + "epoch": 3.5001358880282645, + "grad_norm": 0.8944619297981262, + "learning_rate": 0.000562483013996467, + "loss": 3.4656, + "step": 51515 + }, + { + "epoch": 3.5004756080989265, + "grad_norm": 0.7629538178443909, + "learning_rate": 0.0005624405489876342, + "loss": 3.2441, + "step": 51520 + }, + { + "epoch": 3.500815328169588, + "grad_norm": 0.8104501962661743, + "learning_rate": 0.0005623980839788014, + "loss": 3.3974, + "step": 51525 + }, + { + "epoch": 3.50115504824025, + "grad_norm": 0.841345489025116, + "learning_rate": 0.0005623556189699688, + "loss": 3.542, + "step": 51530 + }, + { + "epoch": 3.501494768310912, + "grad_norm": 0.7238110899925232, + "learning_rate": 0.000562313153961136, + "loss": 3.2091, + "step": 51535 + }, + { + "epoch": 3.5018344883815735, + "grad_norm": 0.8902817368507385, + "learning_rate": 0.0005622706889523032, + "loss": 3.5345, + "step": 51540 + }, + { + "epoch": 3.502174208452235, + "grad_norm": 1.0520728826522827, + "learning_rate": 0.0005622282239434707, + "loss": 3.4738, + "step": 51545 + }, + { + "epoch": 3.5025139285228972, + "grad_norm": 0.8984050154685974, + "learning_rate": 0.0005621857589346379, + "loss": 3.5549, + "step": 51550 + }, + { + "epoch": 3.502853648593559, + "grad_norm": 0.8637600541114807, + "learning_rate": 0.0005621432939258051, + "loss": 3.6015, + "step": 51555 + }, + { + "epoch": 3.5031933686642205, + "grad_norm": 0.8415752053260803, + "learning_rate": 0.0005621008289169724, + "loss": 3.3714, + "step": 51560 + }, + { + "epoch": 3.5035330887348826, + "grad_norm": 0.7472608089447021, + "learning_rate": 0.0005620583639081397, + "loss": 3.5942, + "step": 51565 + }, + { + "epoch": 3.503872808805544, + "grad_norm": 0.7850767374038696, + "learning_rate": 0.0005620158988993069, + "loss": 3.2448, + "step": 51570 + }, + { + "epoch": 3.504212528876206, + "grad_norm": 1.0866000652313232, + "learning_rate": 0.0005619734338904742, + "loss": 3.7759, + "step": 51575 + }, + { + "epoch": 3.504552248946868, + "grad_norm": 0.9364721179008484, + "learning_rate": 0.0005619309688816416, + "loss": 3.4446, + "step": 51580 + }, + { + "epoch": 3.5048919690175295, + "grad_norm": 0.7977840304374695, + "learning_rate": 0.0005618885038728088, + "loss": 3.4841, + "step": 51585 + }, + { + "epoch": 3.505231689088191, + "grad_norm": 0.769893229007721, + "learning_rate": 0.0005618460388639761, + "loss": 3.3831, + "step": 51590 + }, + { + "epoch": 3.5055714091588532, + "grad_norm": 1.0044397115707397, + "learning_rate": 0.0005618035738551434, + "loss": 3.8222, + "step": 51595 + }, + { + "epoch": 3.505911129229515, + "grad_norm": 0.9172252416610718, + "learning_rate": 0.0005617611088463106, + "loss": 3.2412, + "step": 51600 + }, + { + "epoch": 3.5062508493001765, + "grad_norm": 0.7814190983772278, + "learning_rate": 0.0005617186438374779, + "loss": 3.1614, + "step": 51605 + }, + { + "epoch": 3.5065905693708386, + "grad_norm": 0.8668371438980103, + "learning_rate": 0.0005616761788286451, + "loss": 3.3322, + "step": 51610 + }, + { + "epoch": 3.5069302894415, + "grad_norm": 0.7890636920928955, + "learning_rate": 0.0005616337138198125, + "loss": 3.2095, + "step": 51615 + }, + { + "epoch": 3.507270009512162, + "grad_norm": 0.8138740658760071, + "learning_rate": 0.0005615912488109798, + "loss": 3.7045, + "step": 51620 + }, + { + "epoch": 3.507609729582824, + "grad_norm": 1.4448821544647217, + "learning_rate": 0.000561548783802147, + "loss": 3.8606, + "step": 51625 + }, + { + "epoch": 3.5079494496534855, + "grad_norm": 1.2234591245651245, + "learning_rate": 0.0005615063187933144, + "loss": 3.5153, + "step": 51630 + }, + { + "epoch": 3.508289169724147, + "grad_norm": 0.8405647873878479, + "learning_rate": 0.0005614638537844816, + "loss": 3.4182, + "step": 51635 + }, + { + "epoch": 3.5086288897948092, + "grad_norm": 1.1307792663574219, + "learning_rate": 0.0005614213887756488, + "loss": 3.7185, + "step": 51640 + }, + { + "epoch": 3.508968609865471, + "grad_norm": 0.9317634105682373, + "learning_rate": 0.0005613789237668163, + "loss": 3.6125, + "step": 51645 + }, + { + "epoch": 3.5093083299361325, + "grad_norm": 0.9644767045974731, + "learning_rate": 0.0005613364587579835, + "loss": 3.7728, + "step": 51650 + }, + { + "epoch": 3.5096480500067946, + "grad_norm": 0.8064932823181152, + "learning_rate": 0.0005612939937491507, + "loss": 3.3894, + "step": 51655 + }, + { + "epoch": 3.509987770077456, + "grad_norm": 0.8022134900093079, + "learning_rate": 0.000561251528740318, + "loss": 3.4637, + "step": 51660 + }, + { + "epoch": 3.510327490148118, + "grad_norm": 0.6569642424583435, + "learning_rate": 0.0005612090637314853, + "loss": 3.4312, + "step": 51665 + }, + { + "epoch": 3.51066721021878, + "grad_norm": 0.7643089890480042, + "learning_rate": 0.0005611665987226525, + "loss": 3.5662, + "step": 51670 + }, + { + "epoch": 3.5110069302894416, + "grad_norm": 0.8506712913513184, + "learning_rate": 0.0005611241337138198, + "loss": 3.558, + "step": 51675 + }, + { + "epoch": 3.511346650360103, + "grad_norm": 0.9346274733543396, + "learning_rate": 0.0005610816687049872, + "loss": 3.4561, + "step": 51680 + }, + { + "epoch": 3.5116863704307653, + "grad_norm": 0.6653513312339783, + "learning_rate": 0.0005610392036961544, + "loss": 3.5411, + "step": 51685 + }, + { + "epoch": 3.512026090501427, + "grad_norm": 0.8930595517158508, + "learning_rate": 0.0005609967386873217, + "loss": 3.3545, + "step": 51690 + }, + { + "epoch": 3.5123658105720885, + "grad_norm": 1.024877667427063, + "learning_rate": 0.000560954273678489, + "loss": 3.7135, + "step": 51695 + }, + { + "epoch": 3.5127055306427506, + "grad_norm": 1.0778228044509888, + "learning_rate": 0.0005609118086696562, + "loss": 3.7303, + "step": 51700 + }, + { + "epoch": 3.5130452507134122, + "grad_norm": 0.9148048758506775, + "learning_rate": 0.0005608693436608235, + "loss": 3.3472, + "step": 51705 + }, + { + "epoch": 3.513384970784074, + "grad_norm": 0.7239642143249512, + "learning_rate": 0.0005608268786519907, + "loss": 3.618, + "step": 51710 + }, + { + "epoch": 3.513724690854736, + "grad_norm": 0.8533495664596558, + "learning_rate": 0.0005607844136431581, + "loss": 3.3132, + "step": 51715 + }, + { + "epoch": 3.5140644109253976, + "grad_norm": 0.9184256196022034, + "learning_rate": 0.0005607419486343254, + "loss": 3.581, + "step": 51720 + }, + { + "epoch": 3.514404130996059, + "grad_norm": 0.7706084847450256, + "learning_rate": 0.0005606994836254926, + "loss": 3.3683, + "step": 51725 + }, + { + "epoch": 3.5147438510667213, + "grad_norm": 0.9520463347434998, + "learning_rate": 0.0005606570186166599, + "loss": 3.2902, + "step": 51730 + }, + { + "epoch": 3.515083571137383, + "grad_norm": 0.7837512493133545, + "learning_rate": 0.0005606145536078272, + "loss": 3.4524, + "step": 51735 + }, + { + "epoch": 3.5154232912080445, + "grad_norm": 0.9515770077705383, + "learning_rate": 0.0005605720885989944, + "loss": 3.5433, + "step": 51740 + }, + { + "epoch": 3.5157630112787066, + "grad_norm": 1.1817463636398315, + "learning_rate": 0.0005605296235901616, + "loss": 3.6943, + "step": 51745 + }, + { + "epoch": 3.5161027313493682, + "grad_norm": 0.9184385538101196, + "learning_rate": 0.0005604871585813291, + "loss": 3.843, + "step": 51750 + }, + { + "epoch": 3.51644245142003, + "grad_norm": 0.8847455978393555, + "learning_rate": 0.0005604446935724963, + "loss": 3.5419, + "step": 51755 + }, + { + "epoch": 3.516782171490692, + "grad_norm": 0.8379064798355103, + "learning_rate": 0.0005604022285636635, + "loss": 3.5398, + "step": 51760 + }, + { + "epoch": 3.5171218915613536, + "grad_norm": 0.7975239157676697, + "learning_rate": 0.0005603597635548309, + "loss": 3.4863, + "step": 51765 + }, + { + "epoch": 3.517461611632015, + "grad_norm": 0.9701249599456787, + "learning_rate": 0.0005603172985459981, + "loss": 3.5517, + "step": 51770 + }, + { + "epoch": 3.517801331702677, + "grad_norm": 0.9126126766204834, + "learning_rate": 0.0005602748335371653, + "loss": 3.4602, + "step": 51775 + }, + { + "epoch": 3.518141051773339, + "grad_norm": 0.9058125615119934, + "learning_rate": 0.0005602323685283327, + "loss": 3.5894, + "step": 51780 + }, + { + "epoch": 3.5184807718440005, + "grad_norm": 1.140618920326233, + "learning_rate": 0.0005601899035195, + "loss": 3.3426, + "step": 51785 + }, + { + "epoch": 3.518820491914662, + "grad_norm": 0.9155729413032532, + "learning_rate": 0.0005601474385106672, + "loss": 3.7154, + "step": 51790 + }, + { + "epoch": 3.5191602119853242, + "grad_norm": 0.9906036853790283, + "learning_rate": 0.0005601049735018346, + "loss": 3.298, + "step": 51795 + }, + { + "epoch": 3.519499932055986, + "grad_norm": 0.9347759485244751, + "learning_rate": 0.0005600625084930018, + "loss": 3.3577, + "step": 51800 + }, + { + "epoch": 3.5198396521266475, + "grad_norm": 0.7696252465248108, + "learning_rate": 0.000560020043484169, + "loss": 3.7412, + "step": 51805 + }, + { + "epoch": 3.520179372197309, + "grad_norm": 0.7300373315811157, + "learning_rate": 0.0005599775784753363, + "loss": 3.7743, + "step": 51810 + }, + { + "epoch": 3.520519092267971, + "grad_norm": 0.9151380062103271, + "learning_rate": 0.0005599351134665036, + "loss": 3.4998, + "step": 51815 + }, + { + "epoch": 3.520858812338633, + "grad_norm": 0.919293999671936, + "learning_rate": 0.0005598926484576709, + "loss": 3.3316, + "step": 51820 + }, + { + "epoch": 3.5211985324092945, + "grad_norm": 0.813542366027832, + "learning_rate": 0.0005598501834488382, + "loss": 3.406, + "step": 51825 + }, + { + "epoch": 3.5215382524799566, + "grad_norm": 0.918489933013916, + "learning_rate": 0.0005598077184400055, + "loss": 3.5884, + "step": 51830 + }, + { + "epoch": 3.521877972550618, + "grad_norm": 0.8201988935470581, + "learning_rate": 0.0005597652534311727, + "loss": 3.7181, + "step": 51835 + }, + { + "epoch": 3.52221769262128, + "grad_norm": 1.016832709312439, + "learning_rate": 0.00055972278842234, + "loss": 3.306, + "step": 51840 + }, + { + "epoch": 3.522557412691942, + "grad_norm": 0.9122205972671509, + "learning_rate": 0.0005596803234135072, + "loss": 3.4174, + "step": 51845 + }, + { + "epoch": 3.5228971327626035, + "grad_norm": 0.9363826513290405, + "learning_rate": 0.0005596378584046745, + "loss": 3.5525, + "step": 51850 + }, + { + "epoch": 3.523236852833265, + "grad_norm": 0.806104302406311, + "learning_rate": 0.0005595953933958419, + "loss": 3.4359, + "step": 51855 + }, + { + "epoch": 3.5235765729039272, + "grad_norm": 0.9297754764556885, + "learning_rate": 0.0005595529283870091, + "loss": 3.6022, + "step": 51860 + }, + { + "epoch": 3.523916292974589, + "grad_norm": 0.7797461152076721, + "learning_rate": 0.0005595104633781764, + "loss": 3.4858, + "step": 51865 + }, + { + "epoch": 3.5242560130452505, + "grad_norm": 0.7945965528488159, + "learning_rate": 0.0005594679983693437, + "loss": 3.3539, + "step": 51870 + }, + { + "epoch": 3.5245957331159126, + "grad_norm": 0.8169232606887817, + "learning_rate": 0.0005594255333605109, + "loss": 3.6442, + "step": 51875 + }, + { + "epoch": 3.524935453186574, + "grad_norm": 0.9521108269691467, + "learning_rate": 0.0005593830683516782, + "loss": 3.2598, + "step": 51880 + }, + { + "epoch": 3.525275173257236, + "grad_norm": 0.7855775356292725, + "learning_rate": 0.0005593406033428455, + "loss": 3.2925, + "step": 51885 + }, + { + "epoch": 3.525614893327898, + "grad_norm": 0.804915189743042, + "learning_rate": 0.0005592981383340128, + "loss": 3.665, + "step": 51890 + }, + { + "epoch": 3.5259546133985595, + "grad_norm": 0.8565942049026489, + "learning_rate": 0.00055925567332518, + "loss": 3.5115, + "step": 51895 + }, + { + "epoch": 3.526294333469221, + "grad_norm": 0.9097862243652344, + "learning_rate": 0.0005592132083163474, + "loss": 3.5318, + "step": 51900 + }, + { + "epoch": 3.5266340535398832, + "grad_norm": 0.8703302145004272, + "learning_rate": 0.0005591707433075146, + "loss": 3.6007, + "step": 51905 + }, + { + "epoch": 3.526973773610545, + "grad_norm": 0.8829666376113892, + "learning_rate": 0.0005591282782986818, + "loss": 3.6236, + "step": 51910 + }, + { + "epoch": 3.5273134936812065, + "grad_norm": 1.082809329032898, + "learning_rate": 0.0005590858132898492, + "loss": 3.6727, + "step": 51915 + }, + { + "epoch": 3.5276532137518686, + "grad_norm": 0.8204389810562134, + "learning_rate": 0.0005590433482810164, + "loss": 3.2575, + "step": 51920 + }, + { + "epoch": 3.52799293382253, + "grad_norm": 0.8753173351287842, + "learning_rate": 0.0005590008832721837, + "loss": 3.5779, + "step": 51925 + }, + { + "epoch": 3.528332653893192, + "grad_norm": 0.7334783673286438, + "learning_rate": 0.0005589584182633511, + "loss": 3.7181, + "step": 51930 + }, + { + "epoch": 3.528672373963854, + "grad_norm": 0.8596217632293701, + "learning_rate": 0.0005589159532545183, + "loss": 3.4511, + "step": 51935 + }, + { + "epoch": 3.5290120940345155, + "grad_norm": 1.0429655313491821, + "learning_rate": 0.0005588734882456855, + "loss": 3.4838, + "step": 51940 + }, + { + "epoch": 3.529351814105177, + "grad_norm": 1.1335219144821167, + "learning_rate": 0.0005588310232368528, + "loss": 3.4396, + "step": 51945 + }, + { + "epoch": 3.5296915341758393, + "grad_norm": 0.9223446249961853, + "learning_rate": 0.0005587885582280201, + "loss": 3.4127, + "step": 51950 + }, + { + "epoch": 3.530031254246501, + "grad_norm": 0.8735531568527222, + "learning_rate": 0.0005587460932191873, + "loss": 3.4319, + "step": 51955 + }, + { + "epoch": 3.5303709743171625, + "grad_norm": 1.1868902444839478, + "learning_rate": 0.0005587036282103547, + "loss": 3.5537, + "step": 51960 + }, + { + "epoch": 3.5307106943878246, + "grad_norm": 0.8958865404129028, + "learning_rate": 0.000558661163201522, + "loss": 3.4669, + "step": 51965 + }, + { + "epoch": 3.531050414458486, + "grad_norm": 0.7688196897506714, + "learning_rate": 0.0005586186981926893, + "loss": 3.6802, + "step": 51970 + }, + { + "epoch": 3.531390134529148, + "grad_norm": 0.828703761100769, + "learning_rate": 0.0005585762331838565, + "loss": 3.1564, + "step": 51975 + }, + { + "epoch": 3.53172985459981, + "grad_norm": 0.8530594110488892, + "learning_rate": 0.0005585337681750238, + "loss": 3.3192, + "step": 51980 + }, + { + "epoch": 3.5320695746704716, + "grad_norm": 0.9764853119850159, + "learning_rate": 0.0005584913031661911, + "loss": 3.6971, + "step": 51985 + }, + { + "epoch": 3.532409294741133, + "grad_norm": 0.8271609544754028, + "learning_rate": 0.0005584488381573583, + "loss": 3.5326, + "step": 51990 + }, + { + "epoch": 3.5327490148117953, + "grad_norm": 1.083512783050537, + "learning_rate": 0.0005584063731485256, + "loss": 3.6448, + "step": 51995 + }, + { + "epoch": 3.533088734882457, + "grad_norm": 1.0156619548797607, + "learning_rate": 0.000558363908139693, + "loss": 3.5835, + "step": 52000 + }, + { + "epoch": 3.5334284549531185, + "grad_norm": 0.9077045321464539, + "learning_rate": 0.0005583214431308602, + "loss": 3.8323, + "step": 52005 + }, + { + "epoch": 3.5337681750237806, + "grad_norm": 0.9944930672645569, + "learning_rate": 0.0005582789781220274, + "loss": 3.2179, + "step": 52010 + }, + { + "epoch": 3.5341078950944422, + "grad_norm": 0.9028391242027283, + "learning_rate": 0.0005582365131131948, + "loss": 3.5602, + "step": 52015 + }, + { + "epoch": 3.534447615165104, + "grad_norm": 0.8399282097816467, + "learning_rate": 0.000558194048104362, + "loss": 3.4665, + "step": 52020 + }, + { + "epoch": 3.534787335235766, + "grad_norm": 0.73624587059021, + "learning_rate": 0.0005581515830955292, + "loss": 3.5786, + "step": 52025 + }, + { + "epoch": 3.5351270553064276, + "grad_norm": 0.8361068964004517, + "learning_rate": 0.0005581091180866967, + "loss": 3.7046, + "step": 52030 + }, + { + "epoch": 3.535466775377089, + "grad_norm": 1.0157029628753662, + "learning_rate": 0.0005580666530778639, + "loss": 3.6703, + "step": 52035 + }, + { + "epoch": 3.5358064954477513, + "grad_norm": 1.007638931274414, + "learning_rate": 0.0005580241880690311, + "loss": 3.5216, + "step": 52040 + }, + { + "epoch": 3.536146215518413, + "grad_norm": 0.7872365713119507, + "learning_rate": 0.0005579817230601984, + "loss": 3.6176, + "step": 52045 + }, + { + "epoch": 3.5364859355890745, + "grad_norm": 0.8825936913490295, + "learning_rate": 0.0005579392580513657, + "loss": 3.5919, + "step": 52050 + }, + { + "epoch": 3.5368256556597366, + "grad_norm": 0.9555568695068359, + "learning_rate": 0.0005578967930425329, + "loss": 3.6951, + "step": 52055 + }, + { + "epoch": 3.5371653757303982, + "grad_norm": 0.9356720447540283, + "learning_rate": 0.0005578543280337002, + "loss": 3.4531, + "step": 52060 + }, + { + "epoch": 3.53750509580106, + "grad_norm": 0.8546147346496582, + "learning_rate": 0.0005578118630248676, + "loss": 3.4239, + "step": 52065 + }, + { + "epoch": 3.537844815871722, + "grad_norm": 1.025626540184021, + "learning_rate": 0.0005577693980160348, + "loss": 3.3635, + "step": 52070 + }, + { + "epoch": 3.5381845359423836, + "grad_norm": 0.9694139361381531, + "learning_rate": 0.0005577269330072021, + "loss": 3.6169, + "step": 52075 + }, + { + "epoch": 3.538524256013045, + "grad_norm": 1.0546377897262573, + "learning_rate": 0.0005576844679983694, + "loss": 3.6338, + "step": 52080 + }, + { + "epoch": 3.5388639760837073, + "grad_norm": 1.0432416200637817, + "learning_rate": 0.0005576420029895366, + "loss": 3.3604, + "step": 52085 + }, + { + "epoch": 3.539203696154369, + "grad_norm": 1.0540459156036377, + "learning_rate": 0.0005575995379807039, + "loss": 3.3928, + "step": 52090 + }, + { + "epoch": 3.5395434162250305, + "grad_norm": 1.0021926164627075, + "learning_rate": 0.0005575570729718711, + "loss": 3.8106, + "step": 52095 + }, + { + "epoch": 3.5398831362956926, + "grad_norm": 0.8311357498168945, + "learning_rate": 0.0005575146079630385, + "loss": 3.4701, + "step": 52100 + }, + { + "epoch": 3.5402228563663543, + "grad_norm": 0.6501753330230713, + "learning_rate": 0.0005574721429542058, + "loss": 3.7566, + "step": 52105 + }, + { + "epoch": 3.540562576437016, + "grad_norm": 0.8300336003303528, + "learning_rate": 0.000557429677945373, + "loss": 3.2969, + "step": 52110 + }, + { + "epoch": 3.5409022965076775, + "grad_norm": 0.9344176650047302, + "learning_rate": 0.0005573872129365403, + "loss": 3.7353, + "step": 52115 + }, + { + "epoch": 3.5412420165783396, + "grad_norm": 0.8325251936912537, + "learning_rate": 0.0005573447479277076, + "loss": 3.4615, + "step": 52120 + }, + { + "epoch": 3.541581736649001, + "grad_norm": 0.7268164157867432, + "learning_rate": 0.0005573022829188748, + "loss": 3.3834, + "step": 52125 + }, + { + "epoch": 3.541921456719663, + "grad_norm": 0.9894169569015503, + "learning_rate": 0.000557259817910042, + "loss": 3.5761, + "step": 52130 + }, + { + "epoch": 3.542261176790325, + "grad_norm": 0.9691335558891296, + "learning_rate": 0.0005572173529012095, + "loss": 3.5849, + "step": 52135 + }, + { + "epoch": 3.5426008968609866, + "grad_norm": 0.7903117537498474, + "learning_rate": 0.0005571748878923767, + "loss": 3.4318, + "step": 52140 + }, + { + "epoch": 3.542940616931648, + "grad_norm": 0.7900323271751404, + "learning_rate": 0.0005571324228835439, + "loss": 3.588, + "step": 52145 + }, + { + "epoch": 3.54328033700231, + "grad_norm": 0.6701492667198181, + "learning_rate": 0.0005570899578747113, + "loss": 3.447, + "step": 52150 + }, + { + "epoch": 3.543620057072972, + "grad_norm": 0.9259559512138367, + "learning_rate": 0.0005570474928658785, + "loss": 3.7044, + "step": 52155 + }, + { + "epoch": 3.5439597771436335, + "grad_norm": 0.9695510864257812, + "learning_rate": 0.0005570050278570457, + "loss": 3.656, + "step": 52160 + }, + { + "epoch": 3.544299497214295, + "grad_norm": 0.7857998609542847, + "learning_rate": 0.0005569625628482131, + "loss": 3.5088, + "step": 52165 + }, + { + "epoch": 3.5446392172849572, + "grad_norm": 0.8709975481033325, + "learning_rate": 0.0005569200978393804, + "loss": 3.4101, + "step": 52170 + }, + { + "epoch": 3.544978937355619, + "grad_norm": 0.9217280149459839, + "learning_rate": 0.0005568776328305476, + "loss": 3.3223, + "step": 52175 + }, + { + "epoch": 3.5453186574262805, + "grad_norm": 0.8068895936012268, + "learning_rate": 0.000556835167821715, + "loss": 3.3833, + "step": 52180 + }, + { + "epoch": 3.5456583774969426, + "grad_norm": 0.7545412182807922, + "learning_rate": 0.0005567927028128822, + "loss": 3.0967, + "step": 52185 + }, + { + "epoch": 3.545998097567604, + "grad_norm": 0.7863417267799377, + "learning_rate": 0.0005567502378040494, + "loss": 3.5885, + "step": 52190 + }, + { + "epoch": 3.546337817638266, + "grad_norm": 0.8601597547531128, + "learning_rate": 0.0005567077727952167, + "loss": 3.2991, + "step": 52195 + }, + { + "epoch": 3.546677537708928, + "grad_norm": 1.1049659252166748, + "learning_rate": 0.000556665307786384, + "loss": 3.3829, + "step": 52200 + }, + { + "epoch": 3.5470172577795895, + "grad_norm": 1.1074718236923218, + "learning_rate": 0.0005566228427775513, + "loss": 3.4737, + "step": 52205 + }, + { + "epoch": 3.547356977850251, + "grad_norm": 0.7934736609458923, + "learning_rate": 0.0005565803777687186, + "loss": 3.5343, + "step": 52210 + }, + { + "epoch": 3.5476966979209132, + "grad_norm": 0.9247459173202515, + "learning_rate": 0.0005565379127598859, + "loss": 3.6102, + "step": 52215 + }, + { + "epoch": 3.548036417991575, + "grad_norm": 0.8647847175598145, + "learning_rate": 0.0005564954477510531, + "loss": 3.5223, + "step": 52220 + }, + { + "epoch": 3.5483761380622365, + "grad_norm": 1.0227409601211548, + "learning_rate": 0.0005564529827422204, + "loss": 3.4552, + "step": 52225 + }, + { + "epoch": 3.5487158581328986, + "grad_norm": 0.9131997227668762, + "learning_rate": 0.0005564105177333876, + "loss": 3.6005, + "step": 52230 + }, + { + "epoch": 3.54905557820356, + "grad_norm": 0.7534160614013672, + "learning_rate": 0.0005563680527245549, + "loss": 3.6412, + "step": 52235 + }, + { + "epoch": 3.549395298274222, + "grad_norm": 0.873165488243103, + "learning_rate": 0.0005563255877157223, + "loss": 3.5216, + "step": 52240 + }, + { + "epoch": 3.549735018344884, + "grad_norm": 0.9282477498054504, + "learning_rate": 0.0005562831227068895, + "loss": 3.5907, + "step": 52245 + }, + { + "epoch": 3.5500747384155455, + "grad_norm": 0.7818275094032288, + "learning_rate": 0.0005562406576980568, + "loss": 3.5158, + "step": 52250 + }, + { + "epoch": 3.550414458486207, + "grad_norm": 0.7446109056472778, + "learning_rate": 0.0005561981926892241, + "loss": 3.4593, + "step": 52255 + }, + { + "epoch": 3.5507541785568693, + "grad_norm": 1.063023328781128, + "learning_rate": 0.0005561557276803913, + "loss": 3.3804, + "step": 52260 + }, + { + "epoch": 3.551093898627531, + "grad_norm": 0.8629617094993591, + "learning_rate": 0.0005561132626715586, + "loss": 3.4822, + "step": 52265 + }, + { + "epoch": 3.5514336186981925, + "grad_norm": 0.7897460460662842, + "learning_rate": 0.000556070797662726, + "loss": 3.4463, + "step": 52270 + }, + { + "epoch": 3.5517733387688546, + "grad_norm": 0.7240408062934875, + "learning_rate": 0.0005560283326538932, + "loss": 3.7213, + "step": 52275 + }, + { + "epoch": 3.5521130588395162, + "grad_norm": 1.0587586164474487, + "learning_rate": 0.0005559858676450605, + "loss": 3.5715, + "step": 52280 + }, + { + "epoch": 3.552452778910178, + "grad_norm": 0.9337775111198425, + "learning_rate": 0.0005559434026362278, + "loss": 3.5866, + "step": 52285 + }, + { + "epoch": 3.55279249898084, + "grad_norm": 1.1152263879776, + "learning_rate": 0.000555900937627395, + "loss": 3.5777, + "step": 52290 + }, + { + "epoch": 3.5531322190515016, + "grad_norm": 1.1544393301010132, + "learning_rate": 0.0005558584726185622, + "loss": 3.8881, + "step": 52295 + }, + { + "epoch": 3.553471939122163, + "grad_norm": 0.8033807277679443, + "learning_rate": 0.0005558160076097296, + "loss": 3.3601, + "step": 52300 + }, + { + "epoch": 3.5538116591928253, + "grad_norm": 0.8602167963981628, + "learning_rate": 0.0005557735426008969, + "loss": 3.7696, + "step": 52305 + }, + { + "epoch": 3.554151379263487, + "grad_norm": 0.8778760433197021, + "learning_rate": 0.0005557310775920642, + "loss": 3.4121, + "step": 52310 + }, + { + "epoch": 3.5544910993341485, + "grad_norm": 0.734488308429718, + "learning_rate": 0.0005556886125832315, + "loss": 3.4768, + "step": 52315 + }, + { + "epoch": 3.5548308194048106, + "grad_norm": 0.9240022897720337, + "learning_rate": 0.0005556461475743987, + "loss": 3.5798, + "step": 52320 + }, + { + "epoch": 3.5551705394754722, + "grad_norm": 0.6525583267211914, + "learning_rate": 0.000555603682565566, + "loss": 3.6312, + "step": 52325 + }, + { + "epoch": 3.555510259546134, + "grad_norm": 0.9384711384773254, + "learning_rate": 0.0005555612175567333, + "loss": 3.5772, + "step": 52330 + }, + { + "epoch": 3.555849979616796, + "grad_norm": 1.2145386934280396, + "learning_rate": 0.0005555187525479005, + "loss": 3.6883, + "step": 52335 + }, + { + "epoch": 3.5561896996874576, + "grad_norm": 0.8730262517929077, + "learning_rate": 0.0005554762875390679, + "loss": 3.4113, + "step": 52340 + }, + { + "epoch": 3.556529419758119, + "grad_norm": 1.1078976392745972, + "learning_rate": 0.0005554338225302351, + "loss": 3.1874, + "step": 52345 + }, + { + "epoch": 3.5568691398287813, + "grad_norm": 0.7804155945777893, + "learning_rate": 0.0005553913575214024, + "loss": 3.4113, + "step": 52350 + }, + { + "epoch": 3.557208859899443, + "grad_norm": 1.106882095336914, + "learning_rate": 0.0005553488925125697, + "loss": 3.556, + "step": 52355 + }, + { + "epoch": 3.5575485799701045, + "grad_norm": 0.8432283997535706, + "learning_rate": 0.0005553064275037369, + "loss": 3.5888, + "step": 52360 + }, + { + "epoch": 3.5578883000407666, + "grad_norm": 0.8163414001464844, + "learning_rate": 0.0005552639624949042, + "loss": 3.4466, + "step": 52365 + }, + { + "epoch": 3.5582280201114282, + "grad_norm": 0.9727712273597717, + "learning_rate": 0.0005552214974860715, + "loss": 3.4272, + "step": 52370 + }, + { + "epoch": 3.55856774018209, + "grad_norm": 0.8184770345687866, + "learning_rate": 0.0005551790324772388, + "loss": 3.4395, + "step": 52375 + }, + { + "epoch": 3.558907460252752, + "grad_norm": 0.7420095801353455, + "learning_rate": 0.000555136567468406, + "loss": 3.531, + "step": 52380 + }, + { + "epoch": 3.5592471803234136, + "grad_norm": 0.7178859710693359, + "learning_rate": 0.0005550941024595734, + "loss": 3.3636, + "step": 52385 + }, + { + "epoch": 3.559586900394075, + "grad_norm": 1.0624867677688599, + "learning_rate": 0.0005550516374507406, + "loss": 3.3615, + "step": 52390 + }, + { + "epoch": 3.5599266204647373, + "grad_norm": 1.3558964729309082, + "learning_rate": 0.0005550091724419078, + "loss": 3.3387, + "step": 52395 + }, + { + "epoch": 3.560266340535399, + "grad_norm": 1.0010141134262085, + "learning_rate": 0.0005549667074330752, + "loss": 3.5343, + "step": 52400 + }, + { + "epoch": 3.5606060606060606, + "grad_norm": 0.847794771194458, + "learning_rate": 0.0005549242424242424, + "loss": 3.395, + "step": 52405 + }, + { + "epoch": 3.5609457806767226, + "grad_norm": 0.7556495666503906, + "learning_rate": 0.0005548817774154097, + "loss": 3.532, + "step": 52410 + }, + { + "epoch": 3.5612855007473843, + "grad_norm": 0.8140825629234314, + "learning_rate": 0.0005548393124065771, + "loss": 3.573, + "step": 52415 + }, + { + "epoch": 3.561625220818046, + "grad_norm": 0.8537191152572632, + "learning_rate": 0.0005547968473977443, + "loss": 3.6089, + "step": 52420 + }, + { + "epoch": 3.561964940888708, + "grad_norm": 0.8600668907165527, + "learning_rate": 0.0005547543823889115, + "loss": 3.6491, + "step": 52425 + }, + { + "epoch": 3.5623046609593696, + "grad_norm": 0.8466275334358215, + "learning_rate": 0.0005547119173800789, + "loss": 3.3413, + "step": 52430 + }, + { + "epoch": 3.5626443810300312, + "grad_norm": 0.8884152770042419, + "learning_rate": 0.0005546694523712461, + "loss": 3.7404, + "step": 52435 + }, + { + "epoch": 3.5629841011006933, + "grad_norm": 1.2633672952651978, + "learning_rate": 0.0005546269873624133, + "loss": 3.2408, + "step": 52440 + }, + { + "epoch": 3.563323821171355, + "grad_norm": 1.0530993938446045, + "learning_rate": 0.0005545845223535807, + "loss": 3.4943, + "step": 52445 + }, + { + "epoch": 3.5636635412420166, + "grad_norm": 1.009983777999878, + "learning_rate": 0.000554542057344748, + "loss": 3.3403, + "step": 52450 + }, + { + "epoch": 3.564003261312678, + "grad_norm": 1.1613073348999023, + "learning_rate": 0.0005544995923359152, + "loss": 3.39, + "step": 52455 + }, + { + "epoch": 3.5643429813833403, + "grad_norm": 0.8669741153717041, + "learning_rate": 0.0005544571273270825, + "loss": 3.6979, + "step": 52460 + }, + { + "epoch": 3.564682701454002, + "grad_norm": 0.8397752046585083, + "learning_rate": 0.0005544146623182498, + "loss": 3.3534, + "step": 52465 + }, + { + "epoch": 3.5650224215246635, + "grad_norm": 0.9142210483551025, + "learning_rate": 0.000554372197309417, + "loss": 3.4304, + "step": 52470 + }, + { + "epoch": 3.5653621415953256, + "grad_norm": 1.0372178554534912, + "learning_rate": 0.0005543297323005843, + "loss": 3.5681, + "step": 52475 + }, + { + "epoch": 3.5657018616659872, + "grad_norm": 0.9111691117286682, + "learning_rate": 0.0005542872672917517, + "loss": 3.5819, + "step": 52480 + }, + { + "epoch": 3.566041581736649, + "grad_norm": 0.8053564429283142, + "learning_rate": 0.0005542448022829189, + "loss": 3.6201, + "step": 52485 + }, + { + "epoch": 3.5663813018073105, + "grad_norm": 0.9566690325737, + "learning_rate": 0.0005542023372740862, + "loss": 3.5511, + "step": 52490 + }, + { + "epoch": 3.5667210218779726, + "grad_norm": 0.8295654654502869, + "learning_rate": 0.0005541598722652534, + "loss": 3.7962, + "step": 52495 + }, + { + "epoch": 3.567060741948634, + "grad_norm": 0.7827464938163757, + "learning_rate": 0.0005541174072564207, + "loss": 3.3857, + "step": 52500 + }, + { + "epoch": 3.567400462019296, + "grad_norm": 0.7262658476829529, + "learning_rate": 0.000554074942247588, + "loss": 3.3328, + "step": 52505 + }, + { + "epoch": 3.567740182089958, + "grad_norm": 0.6547884345054626, + "learning_rate": 0.0005540324772387552, + "loss": 3.3839, + "step": 52510 + }, + { + "epoch": 3.5680799021606195, + "grad_norm": 0.7867702841758728, + "learning_rate": 0.0005539900122299226, + "loss": 3.2648, + "step": 52515 + }, + { + "epoch": 3.568419622231281, + "grad_norm": 0.8483701348304749, + "learning_rate": 0.0005539475472210899, + "loss": 3.6283, + "step": 52520 + }, + { + "epoch": 3.5687593423019432, + "grad_norm": 1.0494918823242188, + "learning_rate": 0.0005539050822122571, + "loss": 3.8617, + "step": 52525 + }, + { + "epoch": 3.569099062372605, + "grad_norm": 0.8794864416122437, + "learning_rate": 0.0005538626172034243, + "loss": 3.4755, + "step": 52530 + }, + { + "epoch": 3.5694387824432665, + "grad_norm": 0.9872075319290161, + "learning_rate": 0.0005538201521945917, + "loss": 3.2696, + "step": 52535 + }, + { + "epoch": 3.5697785025139286, + "grad_norm": 0.994526743888855, + "learning_rate": 0.0005537776871857589, + "loss": 3.9037, + "step": 52540 + }, + { + "epoch": 3.57011822258459, + "grad_norm": 1.0126667022705078, + "learning_rate": 0.0005537352221769261, + "loss": 3.5087, + "step": 52545 + }, + { + "epoch": 3.570457942655252, + "grad_norm": 0.9972997307777405, + "learning_rate": 0.0005536927571680936, + "loss": 3.5819, + "step": 52550 + }, + { + "epoch": 3.570797662725914, + "grad_norm": 0.9009750485420227, + "learning_rate": 0.0005536502921592608, + "loss": 3.5678, + "step": 52555 + }, + { + "epoch": 3.5711373827965756, + "grad_norm": 0.9232698678970337, + "learning_rate": 0.000553607827150428, + "loss": 3.4753, + "step": 52560 + }, + { + "epoch": 3.571477102867237, + "grad_norm": 0.7926644086837769, + "learning_rate": 0.0005535653621415954, + "loss": 3.3699, + "step": 52565 + }, + { + "epoch": 3.5718168229378993, + "grad_norm": 0.795653760433197, + "learning_rate": 0.0005535228971327626, + "loss": 3.5941, + "step": 52570 + }, + { + "epoch": 3.572156543008561, + "grad_norm": 0.9408392310142517, + "learning_rate": 0.0005534804321239298, + "loss": 3.7129, + "step": 52575 + }, + { + "epoch": 3.5724962630792225, + "grad_norm": 0.8918939828872681, + "learning_rate": 0.0005534379671150971, + "loss": 3.5735, + "step": 52580 + }, + { + "epoch": 3.5728359831498846, + "grad_norm": 0.8696001768112183, + "learning_rate": 0.0005533955021062645, + "loss": 3.7956, + "step": 52585 + }, + { + "epoch": 3.5731757032205462, + "grad_norm": 0.862492561340332, + "learning_rate": 0.0005533530370974317, + "loss": 3.6723, + "step": 52590 + }, + { + "epoch": 3.573515423291208, + "grad_norm": 1.1803942918777466, + "learning_rate": 0.000553310572088599, + "loss": 3.4395, + "step": 52595 + }, + { + "epoch": 3.57385514336187, + "grad_norm": 0.8333913683891296, + "learning_rate": 0.0005532681070797663, + "loss": 3.5296, + "step": 52600 + }, + { + "epoch": 3.5741948634325316, + "grad_norm": 0.595969021320343, + "learning_rate": 0.0005532256420709335, + "loss": 3.4462, + "step": 52605 + }, + { + "epoch": 3.574534583503193, + "grad_norm": 1.1871641874313354, + "learning_rate": 0.0005531831770621008, + "loss": 3.7298, + "step": 52610 + }, + { + "epoch": 3.5748743035738553, + "grad_norm": 1.0818650722503662, + "learning_rate": 0.000553140712053268, + "loss": 3.6356, + "step": 52615 + }, + { + "epoch": 3.575214023644517, + "grad_norm": 0.8001614212989807, + "learning_rate": 0.0005530982470444354, + "loss": 3.5454, + "step": 52620 + }, + { + "epoch": 3.5755537437151785, + "grad_norm": 0.9451671242713928, + "learning_rate": 0.0005530557820356027, + "loss": 3.6534, + "step": 52625 + }, + { + "epoch": 3.5758934637858406, + "grad_norm": 0.9551347494125366, + "learning_rate": 0.00055301331702677, + "loss": 3.7384, + "step": 52630 + }, + { + "epoch": 3.5762331838565022, + "grad_norm": 1.0258092880249023, + "learning_rate": 0.0005529708520179372, + "loss": 3.3459, + "step": 52635 + }, + { + "epoch": 3.576572903927164, + "grad_norm": 0.9507967829704285, + "learning_rate": 0.0005529283870091045, + "loss": 3.4059, + "step": 52640 + }, + { + "epoch": 3.576912623997826, + "grad_norm": 0.803168773651123, + "learning_rate": 0.0005528859220002717, + "loss": 3.4028, + "step": 52645 + }, + { + "epoch": 3.5772523440684876, + "grad_norm": 0.9862895011901855, + "learning_rate": 0.0005528434569914391, + "loss": 3.7526, + "step": 52650 + }, + { + "epoch": 3.577592064139149, + "grad_norm": 0.8800660967826843, + "learning_rate": 0.0005528009919826064, + "loss": 3.4973, + "step": 52655 + }, + { + "epoch": 3.5779317842098113, + "grad_norm": 0.9139227271080017, + "learning_rate": 0.0005527585269737736, + "loss": 3.4352, + "step": 52660 + }, + { + "epoch": 3.578271504280473, + "grad_norm": 0.8356056809425354, + "learning_rate": 0.000552716061964941, + "loss": 3.4296, + "step": 52665 + }, + { + "epoch": 3.5786112243511345, + "grad_norm": 0.8394856452941895, + "learning_rate": 0.0005526735969561082, + "loss": 3.4842, + "step": 52670 + }, + { + "epoch": 3.5789509444217966, + "grad_norm": 0.9028449058532715, + "learning_rate": 0.0005526311319472754, + "loss": 3.5303, + "step": 52675 + }, + { + "epoch": 3.5792906644924583, + "grad_norm": 0.6592805981636047, + "learning_rate": 0.0005525886669384427, + "loss": 3.5797, + "step": 52680 + }, + { + "epoch": 3.57963038456312, + "grad_norm": 0.8321759104728699, + "learning_rate": 0.00055254620192961, + "loss": 3.6278, + "step": 52685 + }, + { + "epoch": 3.579970104633782, + "grad_norm": 0.8286914825439453, + "learning_rate": 0.0005525037369207773, + "loss": 3.6836, + "step": 52690 + }, + { + "epoch": 3.5803098247044436, + "grad_norm": 0.9556105136871338, + "learning_rate": 0.0005524612719119446, + "loss": 3.4488, + "step": 52695 + }, + { + "epoch": 3.580649544775105, + "grad_norm": 0.8201723694801331, + "learning_rate": 0.0005524188069031119, + "loss": 3.8296, + "step": 52700 + }, + { + "epoch": 3.5809892648457673, + "grad_norm": 0.8736295104026794, + "learning_rate": 0.0005523763418942791, + "loss": 3.5008, + "step": 52705 + }, + { + "epoch": 3.581328984916429, + "grad_norm": 0.7972049117088318, + "learning_rate": 0.0005523338768854464, + "loss": 3.5393, + "step": 52710 + }, + { + "epoch": 3.5816687049870906, + "grad_norm": 0.6801161766052246, + "learning_rate": 0.0005522914118766137, + "loss": 3.6074, + "step": 52715 + }, + { + "epoch": 3.5820084250577526, + "grad_norm": 0.9474793076515198, + "learning_rate": 0.0005522489468677809, + "loss": 3.6357, + "step": 52720 + }, + { + "epoch": 3.5823481451284143, + "grad_norm": 0.8200023174285889, + "learning_rate": 0.0005522064818589483, + "loss": 3.552, + "step": 52725 + }, + { + "epoch": 3.582687865199076, + "grad_norm": 0.778630256652832, + "learning_rate": 0.0005521640168501155, + "loss": 3.4114, + "step": 52730 + }, + { + "epoch": 3.583027585269738, + "grad_norm": 0.857019305229187, + "learning_rate": 0.0005521215518412828, + "loss": 3.5997, + "step": 52735 + }, + { + "epoch": 3.5833673053403996, + "grad_norm": 0.9865216612815857, + "learning_rate": 0.0005520790868324501, + "loss": 3.464, + "step": 52740 + }, + { + "epoch": 3.5837070254110612, + "grad_norm": 0.8849552273750305, + "learning_rate": 0.0005520366218236173, + "loss": 3.2026, + "step": 52745 + }, + { + "epoch": 3.5840467454817233, + "grad_norm": 0.7534281611442566, + "learning_rate": 0.0005519941568147846, + "loss": 3.4803, + "step": 52750 + }, + { + "epoch": 3.584386465552385, + "grad_norm": 0.8372395634651184, + "learning_rate": 0.000551951691805952, + "loss": 3.4433, + "step": 52755 + }, + { + "epoch": 3.5847261856230466, + "grad_norm": 1.0841163396835327, + "learning_rate": 0.0005519092267971192, + "loss": 3.7106, + "step": 52760 + }, + { + "epoch": 3.5850659056937086, + "grad_norm": 0.9874895215034485, + "learning_rate": 0.0005518667617882865, + "loss": 3.4888, + "step": 52765 + }, + { + "epoch": 3.5854056257643703, + "grad_norm": 0.7308298349380493, + "learning_rate": 0.0005518242967794538, + "loss": 3.5205, + "step": 52770 + }, + { + "epoch": 3.585745345835032, + "grad_norm": 0.9006463885307312, + "learning_rate": 0.000551781831770621, + "loss": 3.5123, + "step": 52775 + }, + { + "epoch": 3.586085065905694, + "grad_norm": 0.6696062684059143, + "learning_rate": 0.0005517393667617882, + "loss": 3.5431, + "step": 52780 + }, + { + "epoch": 3.5864247859763556, + "grad_norm": 1.0323081016540527, + "learning_rate": 0.0005516969017529556, + "loss": 3.2308, + "step": 52785 + }, + { + "epoch": 3.5867645060470172, + "grad_norm": 0.7446427345275879, + "learning_rate": 0.0005516544367441229, + "loss": 3.7679, + "step": 52790 + }, + { + "epoch": 3.587104226117679, + "grad_norm": 0.8995482921600342, + "learning_rate": 0.0005516119717352901, + "loss": 3.6302, + "step": 52795 + }, + { + "epoch": 3.587443946188341, + "grad_norm": 0.7484413981437683, + "learning_rate": 0.0005515695067264575, + "loss": 3.8534, + "step": 52800 + }, + { + "epoch": 3.5877836662590026, + "grad_norm": 0.9605672955513, + "learning_rate": 0.0005515270417176247, + "loss": 3.5409, + "step": 52805 + }, + { + "epoch": 3.588123386329664, + "grad_norm": 0.9126277565956116, + "learning_rate": 0.0005514845767087919, + "loss": 3.9055, + "step": 52810 + }, + { + "epoch": 3.5884631064003263, + "grad_norm": 0.8671048283576965, + "learning_rate": 0.0005514421116999593, + "loss": 3.6186, + "step": 52815 + }, + { + "epoch": 3.588802826470988, + "grad_norm": 1.0121009349822998, + "learning_rate": 0.0005513996466911265, + "loss": 3.6075, + "step": 52820 + }, + { + "epoch": 3.5891425465416495, + "grad_norm": 1.0537192821502686, + "learning_rate": 0.0005513571816822938, + "loss": 3.606, + "step": 52825 + }, + { + "epoch": 3.589482266612311, + "grad_norm": 0.8901971578598022, + "learning_rate": 0.0005513147166734612, + "loss": 3.7166, + "step": 52830 + }, + { + "epoch": 3.5898219866829733, + "grad_norm": 0.8190103769302368, + "learning_rate": 0.0005512722516646284, + "loss": 3.4991, + "step": 52835 + }, + { + "epoch": 3.590161706753635, + "grad_norm": 0.7430129647254944, + "learning_rate": 0.0005512297866557956, + "loss": 3.7339, + "step": 52840 + }, + { + "epoch": 3.5905014268242965, + "grad_norm": 0.865852952003479, + "learning_rate": 0.0005511873216469629, + "loss": 3.5387, + "step": 52845 + }, + { + "epoch": 3.5908411468949586, + "grad_norm": 0.9233421683311462, + "learning_rate": 0.0005511448566381302, + "loss": 3.5489, + "step": 52850 + }, + { + "epoch": 3.59118086696562, + "grad_norm": 1.1758742332458496, + "learning_rate": 0.0005511023916292974, + "loss": 3.5101, + "step": 52855 + }, + { + "epoch": 3.591520587036282, + "grad_norm": 0.759839653968811, + "learning_rate": 0.0005510599266204648, + "loss": 3.4364, + "step": 52860 + }, + { + "epoch": 3.591860307106944, + "grad_norm": 0.8615610003471375, + "learning_rate": 0.0005510174616116321, + "loss": 3.5271, + "step": 52865 + }, + { + "epoch": 3.5922000271776056, + "grad_norm": 2.8105292320251465, + "learning_rate": 0.0005509749966027993, + "loss": 3.5537, + "step": 52870 + }, + { + "epoch": 3.592539747248267, + "grad_norm": 0.8419066667556763, + "learning_rate": 0.0005509325315939666, + "loss": 3.2907, + "step": 52875 + }, + { + "epoch": 3.5928794673189293, + "grad_norm": 1.0476595163345337, + "learning_rate": 0.0005508900665851338, + "loss": 3.5564, + "step": 52880 + }, + { + "epoch": 3.593219187389591, + "grad_norm": 0.8517836332321167, + "learning_rate": 0.0005508476015763011, + "loss": 3.5576, + "step": 52885 + }, + { + "epoch": 3.5935589074602525, + "grad_norm": 0.648360550403595, + "learning_rate": 0.0005508051365674684, + "loss": 3.4788, + "step": 52890 + }, + { + "epoch": 3.5938986275309146, + "grad_norm": 0.912376880645752, + "learning_rate": 0.0005507626715586357, + "loss": 3.5519, + "step": 52895 + }, + { + "epoch": 3.5942383476015762, + "grad_norm": 0.9318661689758301, + "learning_rate": 0.000550720206549803, + "loss": 3.3773, + "step": 52900 + }, + { + "epoch": 3.594578067672238, + "grad_norm": 0.8613086342811584, + "learning_rate": 0.0005506777415409703, + "loss": 3.5423, + "step": 52905 + }, + { + "epoch": 3.5949177877429, + "grad_norm": 1.0651054382324219, + "learning_rate": 0.0005506352765321375, + "loss": 3.5235, + "step": 52910 + }, + { + "epoch": 3.5952575078135616, + "grad_norm": 0.8997407555580139, + "learning_rate": 0.0005505928115233047, + "loss": 3.7567, + "step": 52915 + }, + { + "epoch": 3.595597227884223, + "grad_norm": 0.73149174451828, + "learning_rate": 0.0005505503465144721, + "loss": 3.5582, + "step": 52920 + }, + { + "epoch": 3.5959369479548853, + "grad_norm": 0.8050642013549805, + "learning_rate": 0.0005505078815056393, + "loss": 3.1826, + "step": 52925 + }, + { + "epoch": 3.596276668025547, + "grad_norm": 0.982751727104187, + "learning_rate": 0.0005504654164968066, + "loss": 3.532, + "step": 52930 + }, + { + "epoch": 3.5966163880962085, + "grad_norm": 0.7732133865356445, + "learning_rate": 0.000550422951487974, + "loss": 3.3858, + "step": 52935 + }, + { + "epoch": 3.5969561081668706, + "grad_norm": 1.0908161401748657, + "learning_rate": 0.0005503804864791412, + "loss": 3.3649, + "step": 52940 + }, + { + "epoch": 3.5972958282375322, + "grad_norm": 0.8185614347457886, + "learning_rate": 0.0005503380214703084, + "loss": 3.3561, + "step": 52945 + }, + { + "epoch": 3.597635548308194, + "grad_norm": 0.8773089647293091, + "learning_rate": 0.0005502955564614758, + "loss": 3.5795, + "step": 52950 + }, + { + "epoch": 3.597975268378856, + "grad_norm": 1.0958915948867798, + "learning_rate": 0.000550253091452643, + "loss": 3.4933, + "step": 52955 + }, + { + "epoch": 3.5983149884495176, + "grad_norm": 0.8125223517417908, + "learning_rate": 0.0005502106264438102, + "loss": 3.4812, + "step": 52960 + }, + { + "epoch": 3.598654708520179, + "grad_norm": 0.8647769093513489, + "learning_rate": 0.0005501681614349777, + "loss": 3.5902, + "step": 52965 + }, + { + "epoch": 3.5989944285908413, + "grad_norm": 0.8423730134963989, + "learning_rate": 0.0005501256964261449, + "loss": 3.5398, + "step": 52970 + }, + { + "epoch": 3.599334148661503, + "grad_norm": 0.6446899771690369, + "learning_rate": 0.0005500832314173121, + "loss": 3.649, + "step": 52975 + }, + { + "epoch": 3.5996738687321646, + "grad_norm": 0.8032547235488892, + "learning_rate": 0.0005500407664084794, + "loss": 3.3662, + "step": 52980 + }, + { + "epoch": 3.6000135888028266, + "grad_norm": 1.1905312538146973, + "learning_rate": 0.0005499983013996467, + "loss": 3.4394, + "step": 52985 + }, + { + "epoch": 3.6003533088734883, + "grad_norm": 0.8894798755645752, + "learning_rate": 0.000549955836390814, + "loss": 3.5324, + "step": 52990 + }, + { + "epoch": 3.60069302894415, + "grad_norm": 0.9026572108268738, + "learning_rate": 0.0005499133713819812, + "loss": 3.7997, + "step": 52995 + }, + { + "epoch": 3.601032749014812, + "grad_norm": 0.7525251507759094, + "learning_rate": 0.0005498709063731486, + "loss": 3.4542, + "step": 53000 + }, + { + "epoch": 3.6013724690854736, + "grad_norm": 0.6907527446746826, + "learning_rate": 0.0005498284413643159, + "loss": 3.5605, + "step": 53005 + }, + { + "epoch": 3.6017121891561352, + "grad_norm": 0.8454573750495911, + "learning_rate": 0.0005497859763554831, + "loss": 3.4105, + "step": 53010 + }, + { + "epoch": 3.6020519092267973, + "grad_norm": 0.8242977261543274, + "learning_rate": 0.0005497435113466504, + "loss": 3.5047, + "step": 53015 + }, + { + "epoch": 3.602391629297459, + "grad_norm": 0.8369054198265076, + "learning_rate": 0.0005497010463378177, + "loss": 3.4958, + "step": 53020 + }, + { + "epoch": 3.6027313493681206, + "grad_norm": 0.9210314750671387, + "learning_rate": 0.0005496585813289849, + "loss": 3.4334, + "step": 53025 + }, + { + "epoch": 3.6030710694387826, + "grad_norm": 0.912879228591919, + "learning_rate": 0.0005496161163201521, + "loss": 3.4379, + "step": 53030 + }, + { + "epoch": 3.6034107895094443, + "grad_norm": 0.8952475190162659, + "learning_rate": 0.0005495736513113196, + "loss": 3.4311, + "step": 53035 + }, + { + "epoch": 3.603750509580106, + "grad_norm": 0.9238948822021484, + "learning_rate": 0.0005495311863024868, + "loss": 3.3197, + "step": 53040 + }, + { + "epoch": 3.604090229650768, + "grad_norm": 0.9383625388145447, + "learning_rate": 0.000549488721293654, + "loss": 3.483, + "step": 53045 + }, + { + "epoch": 3.6044299497214296, + "grad_norm": 0.9761332869529724, + "learning_rate": 0.0005494462562848214, + "loss": 3.361, + "step": 53050 + }, + { + "epoch": 3.6047696697920912, + "grad_norm": 0.6417686939239502, + "learning_rate": 0.0005494037912759886, + "loss": 3.5815, + "step": 53055 + }, + { + "epoch": 3.6051093898627533, + "grad_norm": 0.8139020800590515, + "learning_rate": 0.0005493613262671558, + "loss": 3.6731, + "step": 53060 + }, + { + "epoch": 3.605449109933415, + "grad_norm": 0.9614754915237427, + "learning_rate": 0.0005493188612583232, + "loss": 3.5456, + "step": 53065 + }, + { + "epoch": 3.6057888300040766, + "grad_norm": 0.6282964944839478, + "learning_rate": 0.0005492763962494905, + "loss": 3.5653, + "step": 53070 + }, + { + "epoch": 3.6061285500747386, + "grad_norm": 1.0063493251800537, + "learning_rate": 0.0005492339312406577, + "loss": 3.4016, + "step": 53075 + }, + { + "epoch": 3.6064682701454003, + "grad_norm": 0.8474700450897217, + "learning_rate": 0.000549191466231825, + "loss": 3.4977, + "step": 53080 + }, + { + "epoch": 3.606807990216062, + "grad_norm": 0.9115608334541321, + "learning_rate": 0.0005491490012229923, + "loss": 3.2934, + "step": 53085 + }, + { + "epoch": 3.607147710286724, + "grad_norm": 0.9855828881263733, + "learning_rate": 0.0005491065362141595, + "loss": 3.4119, + "step": 53090 + }, + { + "epoch": 3.6074874303573856, + "grad_norm": 0.7172865271568298, + "learning_rate": 0.0005490640712053268, + "loss": 3.4175, + "step": 53095 + }, + { + "epoch": 3.6078271504280472, + "grad_norm": 0.8368202447891235, + "learning_rate": 0.0005490216061964941, + "loss": 3.562, + "step": 53100 + }, + { + "epoch": 3.6081668704987093, + "grad_norm": 0.6668981909751892, + "learning_rate": 0.0005489791411876614, + "loss": 3.6515, + "step": 53105 + }, + { + "epoch": 3.608506590569371, + "grad_norm": 0.8732035160064697, + "learning_rate": 0.0005489366761788287, + "loss": 3.1435, + "step": 53110 + }, + { + "epoch": 3.6088463106400326, + "grad_norm": 0.754576563835144, + "learning_rate": 0.000548894211169996, + "loss": 3.585, + "step": 53115 + }, + { + "epoch": 3.6091860307106947, + "grad_norm": 0.877511739730835, + "learning_rate": 0.0005488517461611632, + "loss": 3.5862, + "step": 53120 + }, + { + "epoch": 3.6095257507813563, + "grad_norm": 1.0535022020339966, + "learning_rate": 0.0005488092811523305, + "loss": 3.6688, + "step": 53125 + }, + { + "epoch": 3.609865470852018, + "grad_norm": 0.9201755523681641, + "learning_rate": 0.0005487668161434977, + "loss": 3.6635, + "step": 53130 + }, + { + "epoch": 3.6102051909226796, + "grad_norm": 0.9185606837272644, + "learning_rate": 0.000548724351134665, + "loss": 3.8283, + "step": 53135 + }, + { + "epoch": 3.6105449109933416, + "grad_norm": 0.7873866558074951, + "learning_rate": 0.0005486818861258324, + "loss": 3.3596, + "step": 53140 + }, + { + "epoch": 3.6108846310640033, + "grad_norm": 0.8282515406608582, + "learning_rate": 0.0005486394211169996, + "loss": 3.4813, + "step": 53145 + }, + { + "epoch": 3.611224351134665, + "grad_norm": 0.9264218211174011, + "learning_rate": 0.0005485969561081669, + "loss": 3.5958, + "step": 53150 + }, + { + "epoch": 3.611564071205327, + "grad_norm": 0.7441644668579102, + "learning_rate": 0.0005485544910993342, + "loss": 3.5879, + "step": 53155 + }, + { + "epoch": 3.6119037912759886, + "grad_norm": 1.0126665830612183, + "learning_rate": 0.0005485120260905014, + "loss": 3.4628, + "step": 53160 + }, + { + "epoch": 3.6122435113466502, + "grad_norm": 0.9869412183761597, + "learning_rate": 0.0005484695610816686, + "loss": 3.7263, + "step": 53165 + }, + { + "epoch": 3.612583231417312, + "grad_norm": 0.7897995114326477, + "learning_rate": 0.000548427096072836, + "loss": 3.3806, + "step": 53170 + }, + { + "epoch": 3.612922951487974, + "grad_norm": 0.9668745994567871, + "learning_rate": 0.0005483846310640033, + "loss": 3.4123, + "step": 53175 + }, + { + "epoch": 3.6132626715586356, + "grad_norm": 0.9032324552536011, + "learning_rate": 0.0005483421660551705, + "loss": 3.5822, + "step": 53180 + }, + { + "epoch": 3.613602391629297, + "grad_norm": 0.9826297163963318, + "learning_rate": 0.0005482997010463379, + "loss": 3.4616, + "step": 53185 + }, + { + "epoch": 3.6139421116999593, + "grad_norm": 1.0273572206497192, + "learning_rate": 0.0005482572360375051, + "loss": 3.5533, + "step": 53190 + }, + { + "epoch": 3.614281831770621, + "grad_norm": 0.8396711349487305, + "learning_rate": 0.0005482147710286723, + "loss": 3.5326, + "step": 53195 + }, + { + "epoch": 3.6146215518412825, + "grad_norm": 1.0667747259140015, + "learning_rate": 0.0005481723060198397, + "loss": 3.5191, + "step": 53200 + }, + { + "epoch": 3.6149612719119446, + "grad_norm": 1.0240811109542847, + "learning_rate": 0.0005481298410110069, + "loss": 3.3831, + "step": 53205 + }, + { + "epoch": 3.6153009919826062, + "grad_norm": 0.8221215009689331, + "learning_rate": 0.0005480873760021742, + "loss": 3.7308, + "step": 53210 + }, + { + "epoch": 3.615640712053268, + "grad_norm": 1.0272125005722046, + "learning_rate": 0.0005480449109933416, + "loss": 3.6804, + "step": 53215 + }, + { + "epoch": 3.61598043212393, + "grad_norm": 0.8021442294120789, + "learning_rate": 0.0005480024459845088, + "loss": 3.5299, + "step": 53220 + }, + { + "epoch": 3.6163201521945916, + "grad_norm": 0.8069726824760437, + "learning_rate": 0.000547959980975676, + "loss": 3.3171, + "step": 53225 + }, + { + "epoch": 3.616659872265253, + "grad_norm": 0.7607616186141968, + "learning_rate": 0.0005479175159668433, + "loss": 3.4704, + "step": 53230 + }, + { + "epoch": 3.6169995923359153, + "grad_norm": 0.7912532091140747, + "learning_rate": 0.0005478750509580106, + "loss": 3.5246, + "step": 53235 + }, + { + "epoch": 3.617339312406577, + "grad_norm": 0.9379469156265259, + "learning_rate": 0.0005478325859491778, + "loss": 3.6859, + "step": 53240 + }, + { + "epoch": 3.6176790324772385, + "grad_norm": 1.0929460525512695, + "learning_rate": 0.0005477901209403452, + "loss": 3.3808, + "step": 53245 + }, + { + "epoch": 3.6180187525479006, + "grad_norm": 0.9988842010498047, + "learning_rate": 0.0005477476559315125, + "loss": 3.5633, + "step": 53250 + }, + { + "epoch": 3.6183584726185622, + "grad_norm": 1.0579338073730469, + "learning_rate": 0.0005477051909226797, + "loss": 3.5524, + "step": 53255 + }, + { + "epoch": 3.618698192689224, + "grad_norm": 0.875450849533081, + "learning_rate": 0.000547662725913847, + "loss": 3.4359, + "step": 53260 + }, + { + "epoch": 3.619037912759886, + "grad_norm": 0.8895996809005737, + "learning_rate": 0.0005476202609050142, + "loss": 3.3988, + "step": 53265 + }, + { + "epoch": 3.6193776328305476, + "grad_norm": 0.9649412631988525, + "learning_rate": 0.0005475777958961815, + "loss": 3.4005, + "step": 53270 + }, + { + "epoch": 3.619717352901209, + "grad_norm": 0.933057963848114, + "learning_rate": 0.0005475353308873488, + "loss": 3.487, + "step": 53275 + }, + { + "epoch": 3.6200570729718713, + "grad_norm": 0.911983072757721, + "learning_rate": 0.0005474928658785161, + "loss": 3.6928, + "step": 53280 + }, + { + "epoch": 3.620396793042533, + "grad_norm": 1.1191974878311157, + "learning_rate": 0.0005474504008696834, + "loss": 3.4711, + "step": 53285 + }, + { + "epoch": 3.6207365131131946, + "grad_norm": 0.7322220802307129, + "learning_rate": 0.0005474079358608507, + "loss": 3.4827, + "step": 53290 + }, + { + "epoch": 3.6210762331838566, + "grad_norm": 0.90591961145401, + "learning_rate": 0.0005473654708520179, + "loss": 3.4569, + "step": 53295 + }, + { + "epoch": 3.6214159532545183, + "grad_norm": 1.0407510995864868, + "learning_rate": 0.0005473230058431852, + "loss": 3.4327, + "step": 53300 + }, + { + "epoch": 3.62175567332518, + "grad_norm": 0.7971212863922119, + "learning_rate": 0.0005472805408343525, + "loss": 3.5798, + "step": 53305 + }, + { + "epoch": 3.622095393395842, + "grad_norm": 0.9153208136558533, + "learning_rate": 0.0005472380758255197, + "loss": 3.6678, + "step": 53310 + }, + { + "epoch": 3.6224351134665036, + "grad_norm": 0.789109468460083, + "learning_rate": 0.000547195610816687, + "loss": 3.4553, + "step": 53315 + }, + { + "epoch": 3.6227748335371652, + "grad_norm": 1.2554986476898193, + "learning_rate": 0.0005471531458078544, + "loss": 3.1857, + "step": 53320 + }, + { + "epoch": 3.6231145536078273, + "grad_norm": 0.778866708278656, + "learning_rate": 0.0005471106807990216, + "loss": 3.4443, + "step": 53325 + }, + { + "epoch": 3.623454273678489, + "grad_norm": 0.9483287930488586, + "learning_rate": 0.0005470682157901889, + "loss": 3.6143, + "step": 53330 + }, + { + "epoch": 3.6237939937491506, + "grad_norm": 0.7796667218208313, + "learning_rate": 0.0005470257507813562, + "loss": 3.4762, + "step": 53335 + }, + { + "epoch": 3.6241337138198126, + "grad_norm": 0.8498959541320801, + "learning_rate": 0.0005469832857725234, + "loss": 3.6811, + "step": 53340 + }, + { + "epoch": 3.6244734338904743, + "grad_norm": 0.9770117998123169, + "learning_rate": 0.0005469408207636908, + "loss": 3.2819, + "step": 53345 + }, + { + "epoch": 3.624813153961136, + "grad_norm": 0.7897945046424866, + "learning_rate": 0.0005468983557548581, + "loss": 3.3054, + "step": 53350 + }, + { + "epoch": 3.625152874031798, + "grad_norm": 0.7605648636817932, + "learning_rate": 0.0005468558907460253, + "loss": 3.5257, + "step": 53355 + }, + { + "epoch": 3.6254925941024596, + "grad_norm": 1.0064584016799927, + "learning_rate": 0.0005468134257371926, + "loss": 3.8269, + "step": 53360 + }, + { + "epoch": 3.6258323141731212, + "grad_norm": 0.6775883436203003, + "learning_rate": 0.0005467709607283598, + "loss": 3.4411, + "step": 53365 + }, + { + "epoch": 3.6261720342437833, + "grad_norm": 0.7251072525978088, + "learning_rate": 0.0005467284957195271, + "loss": 3.6757, + "step": 53370 + }, + { + "epoch": 3.626511754314445, + "grad_norm": 0.8246013522148132, + "learning_rate": 0.0005466860307106944, + "loss": 3.3034, + "step": 53375 + }, + { + "epoch": 3.6268514743851066, + "grad_norm": 0.9590049982070923, + "learning_rate": 0.0005466435657018617, + "loss": 3.6785, + "step": 53380 + }, + { + "epoch": 3.6271911944557687, + "grad_norm": 0.8394210338592529, + "learning_rate": 0.000546601100693029, + "loss": 3.661, + "step": 53385 + }, + { + "epoch": 3.6275309145264303, + "grad_norm": 1.4547656774520874, + "learning_rate": 0.0005465586356841963, + "loss": 3.406, + "step": 53390 + }, + { + "epoch": 3.627870634597092, + "grad_norm": 1.062156319618225, + "learning_rate": 0.0005465161706753635, + "loss": 3.4511, + "step": 53395 + }, + { + "epoch": 3.628210354667754, + "grad_norm": 0.8980326056480408, + "learning_rate": 0.0005464737056665308, + "loss": 3.406, + "step": 53400 + }, + { + "epoch": 3.6285500747384156, + "grad_norm": 0.9707894325256348, + "learning_rate": 0.0005464312406576981, + "loss": 3.4491, + "step": 53405 + }, + { + "epoch": 3.6288897948090773, + "grad_norm": 1.3499656915664673, + "learning_rate": 0.0005463887756488653, + "loss": 3.7684, + "step": 53410 + }, + { + "epoch": 3.6292295148797393, + "grad_norm": 0.7425840497016907, + "learning_rate": 0.0005463463106400326, + "loss": 3.4067, + "step": 53415 + }, + { + "epoch": 3.629569234950401, + "grad_norm": 0.7487313151359558, + "learning_rate": 0.0005463038456312, + "loss": 3.3259, + "step": 53420 + }, + { + "epoch": 3.6299089550210626, + "grad_norm": 0.865043580532074, + "learning_rate": 0.0005462613806223672, + "loss": 3.6391, + "step": 53425 + }, + { + "epoch": 3.6302486750917247, + "grad_norm": 1.0164997577667236, + "learning_rate": 0.0005462189156135344, + "loss": 3.4808, + "step": 53430 + }, + { + "epoch": 3.6305883951623863, + "grad_norm": 0.9972447156906128, + "learning_rate": 0.0005461764506047018, + "loss": 3.6799, + "step": 53435 + }, + { + "epoch": 3.630928115233048, + "grad_norm": 1.0803797245025635, + "learning_rate": 0.000546133985595869, + "loss": 3.5883, + "step": 53440 + }, + { + "epoch": 3.63126783530371, + "grad_norm": 0.814445972442627, + "learning_rate": 0.0005460915205870362, + "loss": 3.4415, + "step": 53445 + }, + { + "epoch": 3.6316075553743716, + "grad_norm": 0.6653794050216675, + "learning_rate": 0.0005460490555782037, + "loss": 3.5498, + "step": 53450 + }, + { + "epoch": 3.6319472754450333, + "grad_norm": 0.9789270758628845, + "learning_rate": 0.0005460065905693709, + "loss": 3.6238, + "step": 53455 + }, + { + "epoch": 3.6322869955156953, + "grad_norm": 1.101654291152954, + "learning_rate": 0.0005459641255605381, + "loss": 3.1739, + "step": 53460 + }, + { + "epoch": 3.632626715586357, + "grad_norm": 0.9288154244422913, + "learning_rate": 0.0005459216605517054, + "loss": 3.5879, + "step": 53465 + }, + { + "epoch": 3.6329664356570186, + "grad_norm": 0.6950439214706421, + "learning_rate": 0.0005458791955428727, + "loss": 3.4351, + "step": 53470 + }, + { + "epoch": 3.6333061557276802, + "grad_norm": 1.08597731590271, + "learning_rate": 0.0005458367305340399, + "loss": 3.5762, + "step": 53475 + }, + { + "epoch": 3.6336458757983423, + "grad_norm": 1.059578537940979, + "learning_rate": 0.0005457942655252072, + "loss": 3.563, + "step": 53480 + }, + { + "epoch": 3.633985595869004, + "grad_norm": 0.833982527256012, + "learning_rate": 0.0005457518005163746, + "loss": 3.5576, + "step": 53485 + }, + { + "epoch": 3.6343253159396656, + "grad_norm": 1.1361275911331177, + "learning_rate": 0.0005457093355075418, + "loss": 3.5342, + "step": 53490 + }, + { + "epoch": 3.6346650360103276, + "grad_norm": 0.881190836429596, + "learning_rate": 0.0005456668704987091, + "loss": 3.5084, + "step": 53495 + }, + { + "epoch": 3.6350047560809893, + "grad_norm": 0.6833690404891968, + "learning_rate": 0.0005456244054898764, + "loss": 3.576, + "step": 53500 + }, + { + "epoch": 3.635344476151651, + "grad_norm": 0.8589301705360413, + "learning_rate": 0.0005455819404810436, + "loss": 3.4652, + "step": 53505 + }, + { + "epoch": 3.635684196222313, + "grad_norm": 0.7770761847496033, + "learning_rate": 0.0005455394754722109, + "loss": 3.6121, + "step": 53510 + }, + { + "epoch": 3.6360239162929746, + "grad_norm": 0.7699763774871826, + "learning_rate": 0.0005454970104633781, + "loss": 3.4324, + "step": 53515 + }, + { + "epoch": 3.6363636363636362, + "grad_norm": 0.8171278834342957, + "learning_rate": 0.0005454545454545455, + "loss": 3.7349, + "step": 53520 + }, + { + "epoch": 3.636703356434298, + "grad_norm": 1.2641348838806152, + "learning_rate": 0.0005454120804457128, + "loss": 3.302, + "step": 53525 + }, + { + "epoch": 3.63704307650496, + "grad_norm": 0.8112537860870361, + "learning_rate": 0.00054536961543688, + "loss": 3.8005, + "step": 53530 + }, + { + "epoch": 3.6373827965756216, + "grad_norm": 0.7227480411529541, + "learning_rate": 0.0005453271504280473, + "loss": 3.8045, + "step": 53535 + }, + { + "epoch": 3.637722516646283, + "grad_norm": 0.9976720213890076, + "learning_rate": 0.0005452846854192146, + "loss": 3.633, + "step": 53540 + }, + { + "epoch": 3.6380622367169453, + "grad_norm": 0.7704304456710815, + "learning_rate": 0.0005452422204103818, + "loss": 3.5517, + "step": 53545 + }, + { + "epoch": 3.638401956787607, + "grad_norm": 0.8514612913131714, + "learning_rate": 0.000545199755401549, + "loss": 3.6164, + "step": 53550 + }, + { + "epoch": 3.6387416768582685, + "grad_norm": 0.7663012146949768, + "learning_rate": 0.0005451572903927165, + "loss": 3.7757, + "step": 53555 + }, + { + "epoch": 3.6390813969289306, + "grad_norm": 0.9894139170646667, + "learning_rate": 0.0005451148253838837, + "loss": 3.302, + "step": 53560 + }, + { + "epoch": 3.6394211169995923, + "grad_norm": 0.8416471481323242, + "learning_rate": 0.0005450723603750509, + "loss": 3.5129, + "step": 53565 + }, + { + "epoch": 3.639760837070254, + "grad_norm": 0.8176944851875305, + "learning_rate": 0.0005450298953662183, + "loss": 3.4485, + "step": 53570 + }, + { + "epoch": 3.640100557140916, + "grad_norm": 0.9307337999343872, + "learning_rate": 0.0005449874303573855, + "loss": 3.7807, + "step": 53575 + }, + { + "epoch": 3.6404402772115776, + "grad_norm": 0.7537496089935303, + "learning_rate": 0.0005449449653485527, + "loss": 3.6988, + "step": 53580 + }, + { + "epoch": 3.6407799972822392, + "grad_norm": 0.7648143172264099, + "learning_rate": 0.0005449025003397201, + "loss": 3.3825, + "step": 53585 + }, + { + "epoch": 3.6411197173529013, + "grad_norm": 0.8649531602859497, + "learning_rate": 0.0005448600353308874, + "loss": 3.7286, + "step": 53590 + }, + { + "epoch": 3.641459437423563, + "grad_norm": 1.1034812927246094, + "learning_rate": 0.0005448175703220546, + "loss": 3.4015, + "step": 53595 + }, + { + "epoch": 3.6417991574942246, + "grad_norm": 1.0745538473129272, + "learning_rate": 0.000544775105313222, + "loss": 3.6461, + "step": 53600 + }, + { + "epoch": 3.6421388775648866, + "grad_norm": 0.7815353870391846, + "learning_rate": 0.0005447326403043892, + "loss": 3.4524, + "step": 53605 + }, + { + "epoch": 3.6424785976355483, + "grad_norm": 1.0037766695022583, + "learning_rate": 0.0005446901752955564, + "loss": 3.4103, + "step": 53610 + }, + { + "epoch": 3.64281831770621, + "grad_norm": 0.9824424386024475, + "learning_rate": 0.0005446477102867237, + "loss": 3.5805, + "step": 53615 + }, + { + "epoch": 3.643158037776872, + "grad_norm": 0.8942891359329224, + "learning_rate": 0.000544605245277891, + "loss": 3.4682, + "step": 53620 + }, + { + "epoch": 3.6434977578475336, + "grad_norm": 0.8717586994171143, + "learning_rate": 0.0005445627802690583, + "loss": 3.6373, + "step": 53625 + }, + { + "epoch": 3.6438374779181952, + "grad_norm": 0.9676894545555115, + "learning_rate": 0.0005445203152602256, + "loss": 3.8116, + "step": 53630 + }, + { + "epoch": 3.6441771979888573, + "grad_norm": 0.8737943172454834, + "learning_rate": 0.0005444778502513929, + "loss": 3.3956, + "step": 53635 + }, + { + "epoch": 3.644516918059519, + "grad_norm": 0.850583553314209, + "learning_rate": 0.0005444353852425601, + "loss": 3.6765, + "step": 53640 + }, + { + "epoch": 3.6448566381301806, + "grad_norm": 0.8469993472099304, + "learning_rate": 0.0005443929202337274, + "loss": 3.3704, + "step": 53645 + }, + { + "epoch": 3.6451963582008426, + "grad_norm": 0.8102148175239563, + "learning_rate": 0.0005443504552248946, + "loss": 3.5487, + "step": 53650 + }, + { + "epoch": 3.6455360782715043, + "grad_norm": 0.9210461378097534, + "learning_rate": 0.0005443079902160619, + "loss": 3.4939, + "step": 53655 + }, + { + "epoch": 3.645875798342166, + "grad_norm": 0.9317465424537659, + "learning_rate": 0.0005442655252072293, + "loss": 3.5968, + "step": 53660 + }, + { + "epoch": 3.646215518412828, + "grad_norm": 1.086358666419983, + "learning_rate": 0.0005442230601983965, + "loss": 3.5471, + "step": 53665 + }, + { + "epoch": 3.6465552384834896, + "grad_norm": 0.8703083992004395, + "learning_rate": 0.0005441805951895639, + "loss": 3.6222, + "step": 53670 + }, + { + "epoch": 3.6468949585541512, + "grad_norm": 1.1880327463150024, + "learning_rate": 0.0005441381301807311, + "loss": 3.4471, + "step": 53675 + }, + { + "epoch": 3.6472346786248133, + "grad_norm": 0.7977964878082275, + "learning_rate": 0.0005440956651718983, + "loss": 3.397, + "step": 53680 + }, + { + "epoch": 3.647574398695475, + "grad_norm": 0.7841601371765137, + "learning_rate": 0.0005440532001630657, + "loss": 3.5529, + "step": 53685 + }, + { + "epoch": 3.6479141187661366, + "grad_norm": 0.8847991824150085, + "learning_rate": 0.0005440107351542329, + "loss": 3.6659, + "step": 53690 + }, + { + "epoch": 3.6482538388367987, + "grad_norm": 0.8309943079948425, + "learning_rate": 0.0005439682701454002, + "loss": 3.5767, + "step": 53695 + }, + { + "epoch": 3.6485935589074603, + "grad_norm": 0.8960649371147156, + "learning_rate": 0.0005439258051365676, + "loss": 3.5285, + "step": 53700 + }, + { + "epoch": 3.648933278978122, + "grad_norm": 1.1906756162643433, + "learning_rate": 0.0005438833401277348, + "loss": 3.5093, + "step": 53705 + }, + { + "epoch": 3.649272999048784, + "grad_norm": 0.9441413879394531, + "learning_rate": 0.000543840875118902, + "loss": 3.7124, + "step": 53710 + }, + { + "epoch": 3.6496127191194456, + "grad_norm": 0.7278782725334167, + "learning_rate": 0.0005437984101100693, + "loss": 3.7387, + "step": 53715 + }, + { + "epoch": 3.6499524391901073, + "grad_norm": 0.8797478079795837, + "learning_rate": 0.0005437559451012366, + "loss": 3.5415, + "step": 53720 + }, + { + "epoch": 3.6502921592607693, + "grad_norm": 0.7714616060256958, + "learning_rate": 0.0005437134800924038, + "loss": 3.5377, + "step": 53725 + }, + { + "epoch": 3.650631879331431, + "grad_norm": 1.0111420154571533, + "learning_rate": 0.0005436710150835712, + "loss": 3.5683, + "step": 53730 + }, + { + "epoch": 3.6509715994020926, + "grad_norm": 1.0158363580703735, + "learning_rate": 0.0005436285500747385, + "loss": 3.3675, + "step": 53735 + }, + { + "epoch": 3.6513113194727547, + "grad_norm": 0.9356029629707336, + "learning_rate": 0.0005435860850659057, + "loss": 3.4682, + "step": 53740 + }, + { + "epoch": 3.6516510395434163, + "grad_norm": 0.9091253876686096, + "learning_rate": 0.000543543620057073, + "loss": 3.3034, + "step": 53745 + }, + { + "epoch": 3.651990759614078, + "grad_norm": 0.6773715615272522, + "learning_rate": 0.0005435011550482403, + "loss": 3.4817, + "step": 53750 + }, + { + "epoch": 3.65233047968474, + "grad_norm": 0.7366059422492981, + "learning_rate": 0.0005434586900394075, + "loss": 3.6884, + "step": 53755 + }, + { + "epoch": 3.6526701997554016, + "grad_norm": 0.8861567378044128, + "learning_rate": 0.0005434162250305748, + "loss": 3.4124, + "step": 53760 + }, + { + "epoch": 3.6530099198260633, + "grad_norm": 1.0470925569534302, + "learning_rate": 0.0005433737600217421, + "loss": 3.4808, + "step": 53765 + }, + { + "epoch": 3.6533496398967253, + "grad_norm": 0.8748601078987122, + "learning_rate": 0.0005433312950129094, + "loss": 3.3202, + "step": 53770 + }, + { + "epoch": 3.653689359967387, + "grad_norm": 0.860626757144928, + "learning_rate": 0.0005432888300040767, + "loss": 3.6189, + "step": 53775 + }, + { + "epoch": 3.6540290800380486, + "grad_norm": 0.8967095017433167, + "learning_rate": 0.0005432463649952439, + "loss": 3.548, + "step": 53780 + }, + { + "epoch": 3.6543688001087107, + "grad_norm": 0.8716451525688171, + "learning_rate": 0.0005432038999864112, + "loss": 3.5785, + "step": 53785 + }, + { + "epoch": 3.6547085201793723, + "grad_norm": 0.9621819257736206, + "learning_rate": 0.0005431614349775785, + "loss": 3.4762, + "step": 53790 + }, + { + "epoch": 3.655048240250034, + "grad_norm": 0.8762222528457642, + "learning_rate": 0.0005431189699687457, + "loss": 3.5719, + "step": 53795 + }, + { + "epoch": 3.655387960320696, + "grad_norm": 0.7597521543502808, + "learning_rate": 0.000543076504959913, + "loss": 3.5219, + "step": 53800 + }, + { + "epoch": 3.6557276803913576, + "grad_norm": 1.2984267473220825, + "learning_rate": 0.0005430340399510804, + "loss": 3.6098, + "step": 53805 + }, + { + "epoch": 3.6560674004620193, + "grad_norm": 1.042135238647461, + "learning_rate": 0.0005429915749422476, + "loss": 3.3923, + "step": 53810 + }, + { + "epoch": 3.656407120532681, + "grad_norm": 0.7758339643478394, + "learning_rate": 0.0005429491099334148, + "loss": 3.5901, + "step": 53815 + }, + { + "epoch": 3.656746840603343, + "grad_norm": 1.0324559211730957, + "learning_rate": 0.0005429066449245822, + "loss": 3.6956, + "step": 53820 + }, + { + "epoch": 3.6570865606740046, + "grad_norm": 0.9685511589050293, + "learning_rate": 0.0005428641799157494, + "loss": 3.6136, + "step": 53825 + }, + { + "epoch": 3.6574262807446662, + "grad_norm": 0.827765941619873, + "learning_rate": 0.0005428217149069166, + "loss": 3.3311, + "step": 53830 + }, + { + "epoch": 3.6577660008153283, + "grad_norm": 0.8985838890075684, + "learning_rate": 0.0005427792498980841, + "loss": 3.0632, + "step": 53835 + }, + { + "epoch": 3.65810572088599, + "grad_norm": 1.1034096479415894, + "learning_rate": 0.0005427367848892513, + "loss": 3.2518, + "step": 53840 + }, + { + "epoch": 3.6584454409566516, + "grad_norm": 1.1418403387069702, + "learning_rate": 0.0005426943198804185, + "loss": 3.5205, + "step": 53845 + }, + { + "epoch": 3.6587851610273137, + "grad_norm": 1.030767560005188, + "learning_rate": 0.0005426518548715859, + "loss": 3.5257, + "step": 53850 + }, + { + "epoch": 3.6591248810979753, + "grad_norm": 1.037742018699646, + "learning_rate": 0.0005426093898627531, + "loss": 3.6175, + "step": 53855 + }, + { + "epoch": 3.659464601168637, + "grad_norm": 0.906934916973114, + "learning_rate": 0.0005425669248539203, + "loss": 3.4753, + "step": 53860 + }, + { + "epoch": 3.6598043212392986, + "grad_norm": 0.7545692920684814, + "learning_rate": 0.0005425244598450876, + "loss": 3.197, + "step": 53865 + }, + { + "epoch": 3.6601440413099606, + "grad_norm": 0.7415077090263367, + "learning_rate": 0.000542481994836255, + "loss": 3.4878, + "step": 53870 + }, + { + "epoch": 3.6604837613806223, + "grad_norm": 0.8560155034065247, + "learning_rate": 0.0005424395298274222, + "loss": 3.5776, + "step": 53875 + }, + { + "epoch": 3.660823481451284, + "grad_norm": 0.8856314420700073, + "learning_rate": 0.0005423970648185895, + "loss": 3.5032, + "step": 53880 + }, + { + "epoch": 3.661163201521946, + "grad_norm": 0.7570082545280457, + "learning_rate": 0.0005423545998097568, + "loss": 3.2862, + "step": 53885 + }, + { + "epoch": 3.6615029215926076, + "grad_norm": 0.8551943302154541, + "learning_rate": 0.000542312134800924, + "loss": 3.5629, + "step": 53890 + }, + { + "epoch": 3.6618426416632692, + "grad_norm": 0.9009014964103699, + "learning_rate": 0.0005422696697920913, + "loss": 3.7592, + "step": 53895 + }, + { + "epoch": 3.6621823617339313, + "grad_norm": 0.7558007836341858, + "learning_rate": 0.0005422272047832585, + "loss": 3.5012, + "step": 53900 + }, + { + "epoch": 3.662522081804593, + "grad_norm": 0.6824040412902832, + "learning_rate": 0.0005421847397744259, + "loss": 3.7095, + "step": 53905 + }, + { + "epoch": 3.6628618018752546, + "grad_norm": 0.8432365655899048, + "learning_rate": 0.0005421422747655932, + "loss": 3.5574, + "step": 53910 + }, + { + "epoch": 3.6632015219459166, + "grad_norm": 1.1616342067718506, + "learning_rate": 0.0005420998097567604, + "loss": 3.3939, + "step": 53915 + }, + { + "epoch": 3.6635412420165783, + "grad_norm": 0.8574005365371704, + "learning_rate": 0.0005420573447479277, + "loss": 3.6149, + "step": 53920 + }, + { + "epoch": 3.66388096208724, + "grad_norm": 0.8057312369346619, + "learning_rate": 0.000542014879739095, + "loss": 3.6741, + "step": 53925 + }, + { + "epoch": 3.664220682157902, + "grad_norm": 0.727634072303772, + "learning_rate": 0.0005419724147302622, + "loss": 3.7926, + "step": 53930 + }, + { + "epoch": 3.6645604022285636, + "grad_norm": 0.8224619626998901, + "learning_rate": 0.0005419299497214295, + "loss": 3.5678, + "step": 53935 + }, + { + "epoch": 3.6649001222992252, + "grad_norm": 0.7493871450424194, + "learning_rate": 0.0005418874847125969, + "loss": 3.6014, + "step": 53940 + }, + { + "epoch": 3.6652398423698873, + "grad_norm": 0.8806480169296265, + "learning_rate": 0.0005418450197037641, + "loss": 3.4166, + "step": 53945 + }, + { + "epoch": 3.665579562440549, + "grad_norm": 0.8685665130615234, + "learning_rate": 0.0005418025546949313, + "loss": 3.3437, + "step": 53950 + }, + { + "epoch": 3.6659192825112106, + "grad_norm": 1.0424593687057495, + "learning_rate": 0.0005417600896860987, + "loss": 3.7589, + "step": 53955 + }, + { + "epoch": 3.6662590025818727, + "grad_norm": 0.7586832642555237, + "learning_rate": 0.0005417176246772659, + "loss": 3.5474, + "step": 53960 + }, + { + "epoch": 3.6665987226525343, + "grad_norm": 0.9218701720237732, + "learning_rate": 0.0005416751596684331, + "loss": 3.7302, + "step": 53965 + }, + { + "epoch": 3.666938442723196, + "grad_norm": 1.33937668800354, + "learning_rate": 0.0005416326946596006, + "loss": 3.5739, + "step": 53970 + }, + { + "epoch": 3.667278162793858, + "grad_norm": 0.7950059771537781, + "learning_rate": 0.0005415902296507678, + "loss": 3.8805, + "step": 53975 + }, + { + "epoch": 3.6676178828645196, + "grad_norm": 0.8059768080711365, + "learning_rate": 0.000541547764641935, + "loss": 3.6719, + "step": 53980 + }, + { + "epoch": 3.6679576029351812, + "grad_norm": 1.020132064819336, + "learning_rate": 0.0005415052996331024, + "loss": 3.6698, + "step": 53985 + }, + { + "epoch": 3.6682973230058433, + "grad_norm": 0.8819435238838196, + "learning_rate": 0.0005414628346242696, + "loss": 3.2554, + "step": 53990 + }, + { + "epoch": 3.668637043076505, + "grad_norm": 0.8289917707443237, + "learning_rate": 0.0005414203696154368, + "loss": 3.5836, + "step": 53995 + }, + { + "epoch": 3.6689767631471666, + "grad_norm": 1.0032011270523071, + "learning_rate": 0.0005413779046066041, + "loss": 3.5997, + "step": 54000 + }, + { + "epoch": 3.6693164832178287, + "grad_norm": 0.8256915807723999, + "learning_rate": 0.0005413354395977715, + "loss": 3.5211, + "step": 54005 + }, + { + "epoch": 3.6696562032884903, + "grad_norm": 0.6539381742477417, + "learning_rate": 0.0005412929745889388, + "loss": 3.4966, + "step": 54010 + }, + { + "epoch": 3.669995923359152, + "grad_norm": 0.6959814429283142, + "learning_rate": 0.000541250509580106, + "loss": 3.6032, + "step": 54015 + }, + { + "epoch": 3.670335643429814, + "grad_norm": 0.7760059833526611, + "learning_rate": 0.0005412080445712733, + "loss": 3.5113, + "step": 54020 + }, + { + "epoch": 3.6706753635004756, + "grad_norm": 0.9195011258125305, + "learning_rate": 0.0005411655795624406, + "loss": 3.5846, + "step": 54025 + }, + { + "epoch": 3.6710150835711373, + "grad_norm": 0.7785404324531555, + "learning_rate": 0.0005411231145536078, + "loss": 3.605, + "step": 54030 + }, + { + "epoch": 3.6713548036417993, + "grad_norm": 0.7625938057899475, + "learning_rate": 0.000541080649544775, + "loss": 3.5765, + "step": 54035 + }, + { + "epoch": 3.671694523712461, + "grad_norm": 0.7925326228141785, + "learning_rate": 0.0005410381845359425, + "loss": 3.6898, + "step": 54040 + }, + { + "epoch": 3.6720342437831226, + "grad_norm": 0.8791647553443909, + "learning_rate": 0.0005409957195271097, + "loss": 3.6223, + "step": 54045 + }, + { + "epoch": 3.6723739638537847, + "grad_norm": 0.737472653388977, + "learning_rate": 0.000540953254518277, + "loss": 3.5384, + "step": 54050 + }, + { + "epoch": 3.6727136839244463, + "grad_norm": 1.0030609369277954, + "learning_rate": 0.0005409107895094443, + "loss": 3.6356, + "step": 54055 + }, + { + "epoch": 3.673053403995108, + "grad_norm": 0.9176384806632996, + "learning_rate": 0.0005408683245006115, + "loss": 3.716, + "step": 54060 + }, + { + "epoch": 3.67339312406577, + "grad_norm": 0.8432513475418091, + "learning_rate": 0.0005408258594917787, + "loss": 3.5127, + "step": 54065 + }, + { + "epoch": 3.6737328441364316, + "grad_norm": 1.087085247039795, + "learning_rate": 0.0005407833944829461, + "loss": 3.6224, + "step": 54070 + }, + { + "epoch": 3.6740725642070933, + "grad_norm": 0.7479074001312256, + "learning_rate": 0.0005407409294741134, + "loss": 3.809, + "step": 54075 + }, + { + "epoch": 3.6744122842777553, + "grad_norm": 0.7922296524047852, + "learning_rate": 0.0005406984644652806, + "loss": 3.6853, + "step": 54080 + }, + { + "epoch": 3.674752004348417, + "grad_norm": 1.0359100103378296, + "learning_rate": 0.000540655999456448, + "loss": 3.7038, + "step": 54085 + }, + { + "epoch": 3.6750917244190786, + "grad_norm": 0.7226911187171936, + "learning_rate": 0.0005406135344476152, + "loss": 3.56, + "step": 54090 + }, + { + "epoch": 3.6754314444897407, + "grad_norm": 0.924033522605896, + "learning_rate": 0.0005405710694387824, + "loss": 3.1501, + "step": 54095 + }, + { + "epoch": 3.6757711645604023, + "grad_norm": 0.9478504657745361, + "learning_rate": 0.0005405286044299497, + "loss": 3.6716, + "step": 54100 + }, + { + "epoch": 3.676110884631064, + "grad_norm": 0.9186589121818542, + "learning_rate": 0.000540486139421117, + "loss": 3.4695, + "step": 54105 + }, + { + "epoch": 3.676450604701726, + "grad_norm": 1.0014567375183105, + "learning_rate": 0.0005404436744122843, + "loss": 3.5712, + "step": 54110 + }, + { + "epoch": 3.6767903247723877, + "grad_norm": 0.8400470018386841, + "learning_rate": 0.0005404012094034516, + "loss": 3.4606, + "step": 54115 + }, + { + "epoch": 3.6771300448430493, + "grad_norm": 0.9921585917472839, + "learning_rate": 0.0005403587443946189, + "loss": 3.5689, + "step": 54120 + }, + { + "epoch": 3.6774697649137114, + "grad_norm": 0.9346041679382324, + "learning_rate": 0.0005403162793857861, + "loss": 3.5926, + "step": 54125 + }, + { + "epoch": 3.677809484984373, + "grad_norm": 0.7197134494781494, + "learning_rate": 0.0005402738143769534, + "loss": 3.399, + "step": 54130 + }, + { + "epoch": 3.6781492050550346, + "grad_norm": 0.8586878180503845, + "learning_rate": 0.0005402313493681207, + "loss": 3.6801, + "step": 54135 + }, + { + "epoch": 3.6784889251256967, + "grad_norm": 0.9854698181152344, + "learning_rate": 0.0005401888843592879, + "loss": 3.6226, + "step": 54140 + }, + { + "epoch": 3.6788286451963583, + "grad_norm": 0.8480338454246521, + "learning_rate": 0.0005401464193504553, + "loss": 3.6068, + "step": 54145 + }, + { + "epoch": 3.67916836526702, + "grad_norm": 0.8386926651000977, + "learning_rate": 0.0005401039543416225, + "loss": 3.5964, + "step": 54150 + }, + { + "epoch": 3.6795080853376816, + "grad_norm": 0.8429489135742188, + "learning_rate": 0.0005400614893327898, + "loss": 3.562, + "step": 54155 + }, + { + "epoch": 3.6798478054083437, + "grad_norm": 0.8664031028747559, + "learning_rate": 0.0005400190243239571, + "loss": 3.505, + "step": 54160 + }, + { + "epoch": 3.6801875254790053, + "grad_norm": 1.0885629653930664, + "learning_rate": 0.0005399765593151243, + "loss": 3.6498, + "step": 54165 + }, + { + "epoch": 3.680527245549667, + "grad_norm": 0.8554604649543762, + "learning_rate": 0.0005399340943062916, + "loss": 3.1622, + "step": 54170 + }, + { + "epoch": 3.680866965620329, + "grad_norm": 0.8289306163787842, + "learning_rate": 0.0005398916292974589, + "loss": 3.8338, + "step": 54175 + }, + { + "epoch": 3.6812066856909906, + "grad_norm": 0.8493819832801819, + "learning_rate": 0.0005398491642886262, + "loss": 3.3331, + "step": 54180 + }, + { + "epoch": 3.6815464057616523, + "grad_norm": 0.7965759038925171, + "learning_rate": 0.0005398066992797935, + "loss": 3.5322, + "step": 54185 + }, + { + "epoch": 3.6818861258323143, + "grad_norm": 0.7904520630836487, + "learning_rate": 0.0005397642342709608, + "loss": 3.5402, + "step": 54190 + }, + { + "epoch": 3.682225845902976, + "grad_norm": 0.7146794199943542, + "learning_rate": 0.000539721769262128, + "loss": 3.3843, + "step": 54195 + }, + { + "epoch": 3.6825655659736376, + "grad_norm": 0.8724516034126282, + "learning_rate": 0.0005396793042532952, + "loss": 3.5765, + "step": 54200 + }, + { + "epoch": 3.6829052860442992, + "grad_norm": 0.9540440440177917, + "learning_rate": 0.0005396368392444626, + "loss": 3.5208, + "step": 54205 + }, + { + "epoch": 3.6832450061149613, + "grad_norm": 1.0120679140090942, + "learning_rate": 0.0005395943742356298, + "loss": 3.6354, + "step": 54210 + }, + { + "epoch": 3.683584726185623, + "grad_norm": 0.8571012020111084, + "learning_rate": 0.0005395519092267971, + "loss": 3.5432, + "step": 54215 + }, + { + "epoch": 3.6839244462562846, + "grad_norm": 0.7572273015975952, + "learning_rate": 0.0005395094442179645, + "loss": 3.4351, + "step": 54220 + }, + { + "epoch": 3.6842641663269466, + "grad_norm": 1.0118818283081055, + "learning_rate": 0.0005394669792091317, + "loss": 3.4619, + "step": 54225 + }, + { + "epoch": 3.6846038863976083, + "grad_norm": 0.8593848347663879, + "learning_rate": 0.0005394245142002989, + "loss": 3.6336, + "step": 54230 + }, + { + "epoch": 3.68494360646827, + "grad_norm": 0.7839584946632385, + "learning_rate": 0.0005393820491914663, + "loss": 3.6019, + "step": 54235 + }, + { + "epoch": 3.685283326538932, + "grad_norm": 0.9891354441642761, + "learning_rate": 0.0005393395841826335, + "loss": 3.244, + "step": 54240 + }, + { + "epoch": 3.6856230466095936, + "grad_norm": 1.0445905923843384, + "learning_rate": 0.0005392971191738007, + "loss": 3.6093, + "step": 54245 + }, + { + "epoch": 3.6859627666802552, + "grad_norm": 1.0240458250045776, + "learning_rate": 0.0005392546541649682, + "loss": 3.7381, + "step": 54250 + }, + { + "epoch": 3.6863024867509173, + "grad_norm": 1.0918055772781372, + "learning_rate": 0.0005392121891561354, + "loss": 3.6994, + "step": 54255 + }, + { + "epoch": 3.686642206821579, + "grad_norm": 0.715413510799408, + "learning_rate": 0.0005391697241473026, + "loss": 3.3235, + "step": 54260 + }, + { + "epoch": 3.6869819268922406, + "grad_norm": 0.5910052061080933, + "learning_rate": 0.0005391272591384699, + "loss": 3.6812, + "step": 54265 + }, + { + "epoch": 3.6873216469629027, + "grad_norm": 0.796925961971283, + "learning_rate": 0.0005390847941296372, + "loss": 3.5641, + "step": 54270 + }, + { + "epoch": 3.6876613670335643, + "grad_norm": 0.8092347979545593, + "learning_rate": 0.0005390423291208044, + "loss": 3.373, + "step": 54275 + }, + { + "epoch": 3.688001087104226, + "grad_norm": 0.6617710590362549, + "learning_rate": 0.0005389998641119717, + "loss": 3.6916, + "step": 54280 + }, + { + "epoch": 3.688340807174888, + "grad_norm": 0.8136577010154724, + "learning_rate": 0.0005389573991031391, + "loss": 3.59, + "step": 54285 + }, + { + "epoch": 3.6886805272455496, + "grad_norm": 1.155113697052002, + "learning_rate": 0.0005389149340943063, + "loss": 3.5284, + "step": 54290 + }, + { + "epoch": 3.6890202473162113, + "grad_norm": 0.7612754702568054, + "learning_rate": 0.0005388724690854736, + "loss": 3.415, + "step": 54295 + }, + { + "epoch": 3.6893599673868733, + "grad_norm": 1.0148868560791016, + "learning_rate": 0.0005388300040766408, + "loss": 3.4938, + "step": 54300 + }, + { + "epoch": 3.689699687457535, + "grad_norm": 0.9893033504486084, + "learning_rate": 0.0005387875390678081, + "loss": 3.6539, + "step": 54305 + }, + { + "epoch": 3.6900394075281966, + "grad_norm": 0.9492884874343872, + "learning_rate": 0.0005387450740589754, + "loss": 3.2321, + "step": 54310 + }, + { + "epoch": 3.6903791275988587, + "grad_norm": 0.8600270748138428, + "learning_rate": 0.0005387026090501426, + "loss": 3.6231, + "step": 54315 + }, + { + "epoch": 3.6907188476695203, + "grad_norm": 0.9982551336288452, + "learning_rate": 0.00053866014404131, + "loss": 3.7628, + "step": 54320 + }, + { + "epoch": 3.691058567740182, + "grad_norm": 0.978286623954773, + "learning_rate": 0.0005386176790324773, + "loss": 3.4764, + "step": 54325 + }, + { + "epoch": 3.691398287810844, + "grad_norm": 0.7374218702316284, + "learning_rate": 0.0005385752140236445, + "loss": 3.3236, + "step": 54330 + }, + { + "epoch": 3.6917380078815056, + "grad_norm": 0.8285622000694275, + "learning_rate": 0.0005385327490148117, + "loss": 3.5261, + "step": 54335 + }, + { + "epoch": 3.6920777279521673, + "grad_norm": 0.8415969610214233, + "learning_rate": 0.0005384902840059791, + "loss": 3.294, + "step": 54340 + }, + { + "epoch": 3.6924174480228293, + "grad_norm": 0.8668003082275391, + "learning_rate": 0.0005384478189971463, + "loss": 3.4581, + "step": 54345 + }, + { + "epoch": 3.692757168093491, + "grad_norm": 0.7305575013160706, + "learning_rate": 0.0005384053539883136, + "loss": 3.762, + "step": 54350 + }, + { + "epoch": 3.6930968881641526, + "grad_norm": 0.7651715874671936, + "learning_rate": 0.000538362888979481, + "loss": 3.6299, + "step": 54355 + }, + { + "epoch": 3.6934366082348147, + "grad_norm": 0.8194260597229004, + "learning_rate": 0.0005383204239706482, + "loss": 3.5059, + "step": 54360 + }, + { + "epoch": 3.6937763283054763, + "grad_norm": 0.7573187351226807, + "learning_rate": 0.0005382779589618155, + "loss": 3.4695, + "step": 54365 + }, + { + "epoch": 3.694116048376138, + "grad_norm": 0.7754042148590088, + "learning_rate": 0.0005382354939529828, + "loss": 3.4687, + "step": 54370 + }, + { + "epoch": 3.6944557684468, + "grad_norm": 1.3281773328781128, + "learning_rate": 0.00053819302894415, + "loss": 3.636, + "step": 54375 + }, + { + "epoch": 3.6947954885174616, + "grad_norm": 0.8469622731208801, + "learning_rate": 0.0005381505639353173, + "loss": 3.1267, + "step": 54380 + }, + { + "epoch": 3.6951352085881233, + "grad_norm": 0.947311520576477, + "learning_rate": 0.0005381080989264845, + "loss": 3.4138, + "step": 54385 + }, + { + "epoch": 3.6954749286587854, + "grad_norm": 0.9367074370384216, + "learning_rate": 0.0005380656339176519, + "loss": 3.5586, + "step": 54390 + }, + { + "epoch": 3.695814648729447, + "grad_norm": 0.812942385673523, + "learning_rate": 0.0005380231689088192, + "loss": 3.3624, + "step": 54395 + }, + { + "epoch": 3.6961543688001086, + "grad_norm": 1.1334902048110962, + "learning_rate": 0.0005379807038999864, + "loss": 3.3966, + "step": 54400 + }, + { + "epoch": 3.6964940888707707, + "grad_norm": 0.8955584764480591, + "learning_rate": 0.0005379382388911537, + "loss": 3.7357, + "step": 54405 + }, + { + "epoch": 3.6968338089414323, + "grad_norm": 0.8426429033279419, + "learning_rate": 0.000537895773882321, + "loss": 3.6096, + "step": 54410 + }, + { + "epoch": 3.697173529012094, + "grad_norm": 0.9208871722221375, + "learning_rate": 0.0005378533088734882, + "loss": 3.507, + "step": 54415 + }, + { + "epoch": 3.697513249082756, + "grad_norm": 0.9474171996116638, + "learning_rate": 0.0005378108438646555, + "loss": 3.5504, + "step": 54420 + }, + { + "epoch": 3.6978529691534177, + "grad_norm": 0.8176616430282593, + "learning_rate": 0.0005377683788558229, + "loss": 3.4164, + "step": 54425 + }, + { + "epoch": 3.6981926892240793, + "grad_norm": 0.8975966572761536, + "learning_rate": 0.0005377259138469901, + "loss": 3.3556, + "step": 54430 + }, + { + "epoch": 3.6985324092947414, + "grad_norm": 0.9108996987342834, + "learning_rate": 0.0005376834488381574, + "loss": 3.5254, + "step": 54435 + }, + { + "epoch": 3.698872129365403, + "grad_norm": 1.0248826742172241, + "learning_rate": 0.0005376409838293247, + "loss": 3.4814, + "step": 54440 + }, + { + "epoch": 3.6992118494360646, + "grad_norm": 0.9548715353012085, + "learning_rate": 0.0005375985188204919, + "loss": 3.5446, + "step": 54445 + }, + { + "epoch": 3.6995515695067267, + "grad_norm": 0.8565791249275208, + "learning_rate": 0.0005375560538116591, + "loss": 3.4506, + "step": 54450 + }, + { + "epoch": 3.6998912895773883, + "grad_norm": 1.0968135595321655, + "learning_rate": 0.0005375135888028266, + "loss": 3.7325, + "step": 54455 + }, + { + "epoch": 3.70023100964805, + "grad_norm": 0.7906412482261658, + "learning_rate": 0.0005374711237939938, + "loss": 3.5085, + "step": 54460 + }, + { + "epoch": 3.700570729718712, + "grad_norm": 0.7648587226867676, + "learning_rate": 0.000537428658785161, + "loss": 3.4002, + "step": 54465 + }, + { + "epoch": 3.7009104497893737, + "grad_norm": 0.7430204153060913, + "learning_rate": 0.0005373861937763284, + "loss": 3.5223, + "step": 54470 + }, + { + "epoch": 3.7012501698600353, + "grad_norm": 0.7937458157539368, + "learning_rate": 0.0005373437287674956, + "loss": 3.5971, + "step": 54475 + }, + { + "epoch": 3.7015898899306974, + "grad_norm": 0.8830466270446777, + "learning_rate": 0.0005373012637586628, + "loss": 3.5855, + "step": 54480 + }, + { + "epoch": 3.701929610001359, + "grad_norm": 0.7268261313438416, + "learning_rate": 0.0005372587987498302, + "loss": 3.5045, + "step": 54485 + }, + { + "epoch": 3.7022693300720206, + "grad_norm": 0.8419830799102783, + "learning_rate": 0.0005372163337409975, + "loss": 3.5513, + "step": 54490 + }, + { + "epoch": 3.7026090501426823, + "grad_norm": 0.9575291872024536, + "learning_rate": 0.0005371738687321647, + "loss": 3.5601, + "step": 54495 + }, + { + "epoch": 3.7029487702133443, + "grad_norm": 0.7267882227897644, + "learning_rate": 0.000537131403723332, + "loss": 3.5419, + "step": 54500 + }, + { + "epoch": 3.703288490284006, + "grad_norm": 0.8549809455871582, + "learning_rate": 0.0005370889387144993, + "loss": 3.4817, + "step": 54505 + }, + { + "epoch": 3.7036282103546676, + "grad_norm": 1.1098554134368896, + "learning_rate": 0.0005370464737056665, + "loss": 3.7328, + "step": 54510 + }, + { + "epoch": 3.7039679304253297, + "grad_norm": 0.760221540927887, + "learning_rate": 0.0005370040086968338, + "loss": 3.2132, + "step": 54515 + }, + { + "epoch": 3.7043076504959913, + "grad_norm": 1.0236623287200928, + "learning_rate": 0.0005369615436880011, + "loss": 3.3995, + "step": 54520 + }, + { + "epoch": 3.704647370566653, + "grad_norm": 0.7703146934509277, + "learning_rate": 0.0005369190786791684, + "loss": 3.6967, + "step": 54525 + }, + { + "epoch": 3.704987090637315, + "grad_norm": 0.7915014028549194, + "learning_rate": 0.0005368766136703357, + "loss": 3.5892, + "step": 54530 + }, + { + "epoch": 3.7053268107079766, + "grad_norm": 0.8315061926841736, + "learning_rate": 0.000536834148661503, + "loss": 3.4938, + "step": 54535 + }, + { + "epoch": 3.7056665307786383, + "grad_norm": 0.9444687962532043, + "learning_rate": 0.0005367916836526702, + "loss": 3.5801, + "step": 54540 + }, + { + "epoch": 3.7060062508493, + "grad_norm": 0.7458621263504028, + "learning_rate": 0.0005367492186438375, + "loss": 3.6295, + "step": 54545 + }, + { + "epoch": 3.706345970919962, + "grad_norm": 0.9720023274421692, + "learning_rate": 0.0005367067536350047, + "loss": 3.5545, + "step": 54550 + }, + { + "epoch": 3.7066856909906236, + "grad_norm": 0.9546652436256409, + "learning_rate": 0.000536664288626172, + "loss": 3.4564, + "step": 54555 + }, + { + "epoch": 3.7070254110612852, + "grad_norm": 1.1242135763168335, + "learning_rate": 0.0005366218236173394, + "loss": 3.5586, + "step": 54560 + }, + { + "epoch": 3.7073651311319473, + "grad_norm": 0.8512378334999084, + "learning_rate": 0.0005365793586085066, + "loss": 3.7682, + "step": 54565 + }, + { + "epoch": 3.707704851202609, + "grad_norm": 1.0664117336273193, + "learning_rate": 0.0005365368935996739, + "loss": 3.3629, + "step": 54570 + }, + { + "epoch": 3.7080445712732706, + "grad_norm": 0.849688708782196, + "learning_rate": 0.0005364944285908412, + "loss": 3.4796, + "step": 54575 + }, + { + "epoch": 3.7083842913439327, + "grad_norm": 0.7968799471855164, + "learning_rate": 0.0005364519635820084, + "loss": 3.53, + "step": 54580 + }, + { + "epoch": 3.7087240114145943, + "grad_norm": 0.8794351816177368, + "learning_rate": 0.0005364094985731756, + "loss": 3.2573, + "step": 54585 + }, + { + "epoch": 3.709063731485256, + "grad_norm": 1.0231249332427979, + "learning_rate": 0.000536367033564343, + "loss": 3.763, + "step": 54590 + }, + { + "epoch": 3.709403451555918, + "grad_norm": 0.78081876039505, + "learning_rate": 0.0005363245685555103, + "loss": 3.545, + "step": 54595 + }, + { + "epoch": 3.7097431716265796, + "grad_norm": 0.8613443374633789, + "learning_rate": 0.0005362821035466775, + "loss": 3.4478, + "step": 54600 + }, + { + "epoch": 3.7100828916972413, + "grad_norm": 0.8601430654525757, + "learning_rate": 0.0005362396385378449, + "loss": 3.5354, + "step": 54605 + }, + { + "epoch": 3.7104226117679033, + "grad_norm": 2.1507205963134766, + "learning_rate": 0.0005361971735290121, + "loss": 3.3419, + "step": 54610 + }, + { + "epoch": 3.710762331838565, + "grad_norm": 0.8295083045959473, + "learning_rate": 0.0005361547085201793, + "loss": 3.2124, + "step": 54615 + }, + { + "epoch": 3.7111020519092266, + "grad_norm": 0.7322379946708679, + "learning_rate": 0.0005361122435113467, + "loss": 3.309, + "step": 54620 + }, + { + "epoch": 3.7114417719798887, + "grad_norm": 0.8152580857276917, + "learning_rate": 0.0005360697785025139, + "loss": 3.4282, + "step": 54625 + }, + { + "epoch": 3.7117814920505503, + "grad_norm": 1.366969347000122, + "learning_rate": 0.0005360273134936812, + "loss": 3.8529, + "step": 54630 + }, + { + "epoch": 3.712121212121212, + "grad_norm": 0.7198229432106018, + "learning_rate": 0.0005359848484848486, + "loss": 3.637, + "step": 54635 + }, + { + "epoch": 3.712460932191874, + "grad_norm": 0.9979026913642883, + "learning_rate": 0.0005359423834760158, + "loss": 3.5131, + "step": 54640 + }, + { + "epoch": 3.7128006522625356, + "grad_norm": 0.925813615322113, + "learning_rate": 0.000535899918467183, + "loss": 3.5843, + "step": 54645 + }, + { + "epoch": 3.7131403723331973, + "grad_norm": 0.8099024891853333, + "learning_rate": 0.0005358574534583503, + "loss": 3.4138, + "step": 54650 + }, + { + "epoch": 3.7134800924038593, + "grad_norm": 1.0962743759155273, + "learning_rate": 0.0005358149884495176, + "loss": 3.5525, + "step": 54655 + }, + { + "epoch": 3.713819812474521, + "grad_norm": 0.8154871463775635, + "learning_rate": 0.0005357725234406848, + "loss": 3.3077, + "step": 54660 + }, + { + "epoch": 3.7141595325451826, + "grad_norm": 0.8442808389663696, + "learning_rate": 0.0005357300584318522, + "loss": 3.3505, + "step": 54665 + }, + { + "epoch": 3.7144992526158447, + "grad_norm": 0.7866305112838745, + "learning_rate": 0.0005356875934230195, + "loss": 3.5951, + "step": 54670 + }, + { + "epoch": 3.7148389726865063, + "grad_norm": 0.741134762763977, + "learning_rate": 0.0005356451284141867, + "loss": 3.4746, + "step": 54675 + }, + { + "epoch": 3.715178692757168, + "grad_norm": 1.075347900390625, + "learning_rate": 0.000535602663405354, + "loss": 3.3624, + "step": 54680 + }, + { + "epoch": 3.71551841282783, + "grad_norm": 0.9806501865386963, + "learning_rate": 0.0005355601983965212, + "loss": 3.5982, + "step": 54685 + }, + { + "epoch": 3.7158581328984917, + "grad_norm": 1.0076103210449219, + "learning_rate": 0.0005355177333876886, + "loss": 3.6586, + "step": 54690 + }, + { + "epoch": 3.7161978529691533, + "grad_norm": 0.8461204767227173, + "learning_rate": 0.0005354752683788558, + "loss": 3.5973, + "step": 54695 + }, + { + "epoch": 3.7165375730398154, + "grad_norm": 0.935477614402771, + "learning_rate": 0.0005354328033700231, + "loss": 3.6131, + "step": 54700 + }, + { + "epoch": 3.716877293110477, + "grad_norm": 0.9814581274986267, + "learning_rate": 0.0005353903383611905, + "loss": 3.6125, + "step": 54705 + }, + { + "epoch": 3.7172170131811386, + "grad_norm": 0.7364605665206909, + "learning_rate": 0.0005353478733523577, + "loss": 3.6075, + "step": 54710 + }, + { + "epoch": 3.7175567332518007, + "grad_norm": 0.7956974506378174, + "learning_rate": 0.0005353054083435249, + "loss": 3.4073, + "step": 54715 + }, + { + "epoch": 3.7178964533224623, + "grad_norm": 0.7337539196014404, + "learning_rate": 0.0005352629433346923, + "loss": 3.4736, + "step": 54720 + }, + { + "epoch": 3.718236173393124, + "grad_norm": 0.8920583128929138, + "learning_rate": 0.0005352204783258595, + "loss": 3.4496, + "step": 54725 + }, + { + "epoch": 3.718575893463786, + "grad_norm": 0.9275780320167542, + "learning_rate": 0.0005351780133170267, + "loss": 3.5347, + "step": 54730 + }, + { + "epoch": 3.7189156135344477, + "grad_norm": 0.816299557685852, + "learning_rate": 0.0005351355483081942, + "loss": 3.2479, + "step": 54735 + }, + { + "epoch": 3.7192553336051093, + "grad_norm": 1.1205774545669556, + "learning_rate": 0.0005350930832993614, + "loss": 3.6161, + "step": 54740 + }, + { + "epoch": 3.7195950536757714, + "grad_norm": 0.8610690832138062, + "learning_rate": 0.0005350506182905286, + "loss": 3.5051, + "step": 54745 + }, + { + "epoch": 3.719934773746433, + "grad_norm": 0.8933074474334717, + "learning_rate": 0.0005350081532816959, + "loss": 3.622, + "step": 54750 + }, + { + "epoch": 3.7202744938170946, + "grad_norm": 0.6731392741203308, + "learning_rate": 0.0005349656882728632, + "loss": 3.5856, + "step": 54755 + }, + { + "epoch": 3.7206142138877567, + "grad_norm": 0.6292549967765808, + "learning_rate": 0.0005349232232640304, + "loss": 3.5979, + "step": 54760 + }, + { + "epoch": 3.7209539339584183, + "grad_norm": 0.749682605266571, + "learning_rate": 0.0005348807582551977, + "loss": 3.3863, + "step": 54765 + }, + { + "epoch": 3.72129365402908, + "grad_norm": 0.8586956262588501, + "learning_rate": 0.0005348382932463651, + "loss": 3.4992, + "step": 54770 + }, + { + "epoch": 3.721633374099742, + "grad_norm": 0.9240660667419434, + "learning_rate": 0.0005347958282375323, + "loss": 3.3668, + "step": 54775 + }, + { + "epoch": 3.7219730941704037, + "grad_norm": 0.9373857378959656, + "learning_rate": 0.0005347533632286996, + "loss": 3.6412, + "step": 54780 + }, + { + "epoch": 3.7223128142410653, + "grad_norm": 0.8075558543205261, + "learning_rate": 0.0005347108982198668, + "loss": 3.3116, + "step": 54785 + }, + { + "epoch": 3.7226525343117274, + "grad_norm": 0.8918603658676147, + "learning_rate": 0.0005346684332110341, + "loss": 3.7121, + "step": 54790 + }, + { + "epoch": 3.722992254382389, + "grad_norm": 1.1032791137695312, + "learning_rate": 0.0005346259682022014, + "loss": 3.581, + "step": 54795 + }, + { + "epoch": 3.7233319744530506, + "grad_norm": 0.7940346002578735, + "learning_rate": 0.0005345835031933686, + "loss": 3.2447, + "step": 54800 + }, + { + "epoch": 3.7236716945237127, + "grad_norm": 1.2446506023406982, + "learning_rate": 0.000534541038184536, + "loss": 3.4916, + "step": 54805 + }, + { + "epoch": 3.7240114145943743, + "grad_norm": 1.3039567470550537, + "learning_rate": 0.0005344985731757033, + "loss": 3.4576, + "step": 54810 + }, + { + "epoch": 3.724351134665036, + "grad_norm": 0.8366376161575317, + "learning_rate": 0.0005344561081668705, + "loss": 3.3213, + "step": 54815 + }, + { + "epoch": 3.724690854735698, + "grad_norm": 0.9179151058197021, + "learning_rate": 0.0005344136431580378, + "loss": 3.7871, + "step": 54820 + }, + { + "epoch": 3.7250305748063597, + "grad_norm": 0.9598586559295654, + "learning_rate": 0.0005343711781492051, + "loss": 3.348, + "step": 54825 + }, + { + "epoch": 3.7253702948770213, + "grad_norm": 1.0367140769958496, + "learning_rate": 0.0005343287131403723, + "loss": 3.3409, + "step": 54830 + }, + { + "epoch": 3.725710014947683, + "grad_norm": 0.8543326258659363, + "learning_rate": 0.0005342862481315395, + "loss": 3.4418, + "step": 54835 + }, + { + "epoch": 3.726049735018345, + "grad_norm": 0.9478472471237183, + "learning_rate": 0.000534243783122707, + "loss": 3.4504, + "step": 54840 + }, + { + "epoch": 3.7263894550890067, + "grad_norm": 0.833121657371521, + "learning_rate": 0.0005342013181138742, + "loss": 3.5395, + "step": 54845 + }, + { + "epoch": 3.7267291751596683, + "grad_norm": 0.7631881833076477, + "learning_rate": 0.0005341588531050414, + "loss": 3.2038, + "step": 54850 + }, + { + "epoch": 3.7270688952303304, + "grad_norm": 0.8639206290245056, + "learning_rate": 0.0005341163880962088, + "loss": 3.5329, + "step": 54855 + }, + { + "epoch": 3.727408615300992, + "grad_norm": 0.7909950613975525, + "learning_rate": 0.000534073923087376, + "loss": 3.6553, + "step": 54860 + }, + { + "epoch": 3.7277483353716536, + "grad_norm": 0.7943188548088074, + "learning_rate": 0.0005340314580785432, + "loss": 3.7804, + "step": 54865 + }, + { + "epoch": 3.7280880554423157, + "grad_norm": 0.9541133642196655, + "learning_rate": 0.0005339889930697106, + "loss": 3.5663, + "step": 54870 + }, + { + "epoch": 3.7284277755129773, + "grad_norm": 0.8442268967628479, + "learning_rate": 0.0005339465280608779, + "loss": 3.3964, + "step": 54875 + }, + { + "epoch": 3.728767495583639, + "grad_norm": 0.8577226400375366, + "learning_rate": 0.0005339040630520451, + "loss": 3.448, + "step": 54880 + }, + { + "epoch": 3.7291072156543006, + "grad_norm": 0.9387079477310181, + "learning_rate": 0.0005338615980432124, + "loss": 3.5225, + "step": 54885 + }, + { + "epoch": 3.7294469357249627, + "grad_norm": 1.030052900314331, + "learning_rate": 0.0005338191330343797, + "loss": 3.6208, + "step": 54890 + }, + { + "epoch": 3.7297866557956243, + "grad_norm": 1.0430415868759155, + "learning_rate": 0.0005337766680255469, + "loss": 3.671, + "step": 54895 + }, + { + "epoch": 3.730126375866286, + "grad_norm": 1.1088014841079712, + "learning_rate": 0.0005337342030167142, + "loss": 3.4393, + "step": 54900 + }, + { + "epoch": 3.730466095936948, + "grad_norm": 0.9064332246780396, + "learning_rate": 0.0005336917380078815, + "loss": 3.3686, + "step": 54905 + }, + { + "epoch": 3.7308058160076096, + "grad_norm": 0.8930426239967346, + "learning_rate": 0.0005336492729990488, + "loss": 3.5592, + "step": 54910 + }, + { + "epoch": 3.7311455360782713, + "grad_norm": 0.9819005131721497, + "learning_rate": 0.0005336068079902161, + "loss": 3.4059, + "step": 54915 + }, + { + "epoch": 3.7314852561489333, + "grad_norm": 0.8169376254081726, + "learning_rate": 0.0005335643429813834, + "loss": 3.405, + "step": 54920 + }, + { + "epoch": 3.731824976219595, + "grad_norm": 0.864361584186554, + "learning_rate": 0.0005335218779725506, + "loss": 3.5455, + "step": 54925 + }, + { + "epoch": 3.7321646962902566, + "grad_norm": 0.9856473803520203, + "learning_rate": 0.0005334794129637179, + "loss": 3.5071, + "step": 54930 + }, + { + "epoch": 3.7325044163609187, + "grad_norm": 0.8088100552558899, + "learning_rate": 0.0005334369479548851, + "loss": 3.596, + "step": 54935 + }, + { + "epoch": 3.7328441364315803, + "grad_norm": 0.8745824098587036, + "learning_rate": 0.0005333944829460524, + "loss": 3.4903, + "step": 54940 + }, + { + "epoch": 3.733183856502242, + "grad_norm": 1.0077687501907349, + "learning_rate": 0.0005333520179372198, + "loss": 3.3983, + "step": 54945 + }, + { + "epoch": 3.733523576572904, + "grad_norm": 1.019166350364685, + "learning_rate": 0.000533309552928387, + "loss": 3.3717, + "step": 54950 + }, + { + "epoch": 3.7338632966435656, + "grad_norm": 1.2827134132385254, + "learning_rate": 0.0005332670879195543, + "loss": 3.6747, + "step": 54955 + }, + { + "epoch": 3.7342030167142273, + "grad_norm": 0.8667526245117188, + "learning_rate": 0.0005332246229107216, + "loss": 3.5614, + "step": 54960 + }, + { + "epoch": 3.7345427367848893, + "grad_norm": 0.9789865612983704, + "learning_rate": 0.0005331821579018888, + "loss": 3.4407, + "step": 54965 + }, + { + "epoch": 3.734882456855551, + "grad_norm": 0.7592118978500366, + "learning_rate": 0.000533139692893056, + "loss": 3.5635, + "step": 54970 + }, + { + "epoch": 3.7352221769262126, + "grad_norm": 0.9635940194129944, + "learning_rate": 0.0005330972278842234, + "loss": 3.6552, + "step": 54975 + }, + { + "epoch": 3.7355618969968747, + "grad_norm": 0.8601370453834534, + "learning_rate": 0.0005330547628753907, + "loss": 3.5822, + "step": 54980 + }, + { + "epoch": 3.7359016170675363, + "grad_norm": 0.7420938014984131, + "learning_rate": 0.0005330122978665579, + "loss": 3.6643, + "step": 54985 + }, + { + "epoch": 3.736241337138198, + "grad_norm": 0.8298119306564331, + "learning_rate": 0.0005329698328577253, + "loss": 3.4935, + "step": 54990 + }, + { + "epoch": 3.73658105720886, + "grad_norm": 0.8837483525276184, + "learning_rate": 0.0005329273678488925, + "loss": 3.6411, + "step": 54995 + }, + { + "epoch": 3.7369207772795217, + "grad_norm": 1.6148351430892944, + "learning_rate": 0.0005328849028400597, + "loss": 3.4252, + "step": 55000 + }, + { + "epoch": 3.7372604973501833, + "grad_norm": 0.7759456038475037, + "learning_rate": 0.0005328424378312271, + "loss": 3.4859, + "step": 55005 + }, + { + "epoch": 3.7376002174208454, + "grad_norm": 0.9428431391716003, + "learning_rate": 0.0005327999728223943, + "loss": 3.5675, + "step": 55010 + }, + { + "epoch": 3.737939937491507, + "grad_norm": 0.8304556012153625, + "learning_rate": 0.0005327575078135616, + "loss": 3.7123, + "step": 55015 + }, + { + "epoch": 3.7382796575621686, + "grad_norm": 0.9272135496139526, + "learning_rate": 0.000532715042804729, + "loss": 3.7683, + "step": 55020 + }, + { + "epoch": 3.7386193776328307, + "grad_norm": 1.6169096231460571, + "learning_rate": 0.0005326725777958962, + "loss": 3.4087, + "step": 55025 + }, + { + "epoch": 3.7389590977034923, + "grad_norm": 1.0453542470932007, + "learning_rate": 0.0005326301127870635, + "loss": 3.497, + "step": 55030 + }, + { + "epoch": 3.739298817774154, + "grad_norm": 0.9136646389961243, + "learning_rate": 0.0005325876477782307, + "loss": 3.3562, + "step": 55035 + }, + { + "epoch": 3.739638537844816, + "grad_norm": 1.4324803352355957, + "learning_rate": 0.000532545182769398, + "loss": 3.5394, + "step": 55040 + }, + { + "epoch": 3.7399782579154777, + "grad_norm": 0.9308119416236877, + "learning_rate": 0.0005325027177605654, + "loss": 3.0882, + "step": 55045 + }, + { + "epoch": 3.7403179779861393, + "grad_norm": 0.9755516648292542, + "learning_rate": 0.0005324602527517326, + "loss": 3.3318, + "step": 55050 + }, + { + "epoch": 3.7406576980568014, + "grad_norm": 0.8834554553031921, + "learning_rate": 0.0005324177877428999, + "loss": 3.6192, + "step": 55055 + }, + { + "epoch": 3.740997418127463, + "grad_norm": 1.0033539533615112, + "learning_rate": 0.0005323753227340672, + "loss": 3.615, + "step": 55060 + }, + { + "epoch": 3.7413371381981246, + "grad_norm": 0.9126165509223938, + "learning_rate": 0.0005323328577252344, + "loss": 3.5484, + "step": 55065 + }, + { + "epoch": 3.7416768582687867, + "grad_norm": 0.7196186184883118, + "learning_rate": 0.0005322903927164016, + "loss": 3.6829, + "step": 55070 + }, + { + "epoch": 3.7420165783394483, + "grad_norm": 0.8776150941848755, + "learning_rate": 0.000532247927707569, + "loss": 3.8863, + "step": 55075 + }, + { + "epoch": 3.74235629841011, + "grad_norm": 0.7584242820739746, + "learning_rate": 0.0005322054626987363, + "loss": 3.3319, + "step": 55080 + }, + { + "epoch": 3.742696018480772, + "grad_norm": 0.8096224665641785, + "learning_rate": 0.0005321629976899035, + "loss": 3.3404, + "step": 55085 + }, + { + "epoch": 3.7430357385514337, + "grad_norm": 1.1541545391082764, + "learning_rate": 0.0005321205326810709, + "loss": 3.3947, + "step": 55090 + }, + { + "epoch": 3.7433754586220953, + "grad_norm": 0.8094286322593689, + "learning_rate": 0.0005320780676722381, + "loss": 3.5208, + "step": 55095 + }, + { + "epoch": 3.7437151786927574, + "grad_norm": 1.1090295314788818, + "learning_rate": 0.0005320356026634053, + "loss": 3.4305, + "step": 55100 + }, + { + "epoch": 3.744054898763419, + "grad_norm": 1.0830841064453125, + "learning_rate": 0.0005319931376545727, + "loss": 3.2967, + "step": 55105 + }, + { + "epoch": 3.7443946188340806, + "grad_norm": 0.9869200587272644, + "learning_rate": 0.0005319506726457399, + "loss": 3.4402, + "step": 55110 + }, + { + "epoch": 3.7447343389047427, + "grad_norm": 0.9945034980773926, + "learning_rate": 0.0005319082076369072, + "loss": 3.6205, + "step": 55115 + }, + { + "epoch": 3.7450740589754044, + "grad_norm": 0.7808508276939392, + "learning_rate": 0.0005318657426280746, + "loss": 3.1969, + "step": 55120 + }, + { + "epoch": 3.745413779046066, + "grad_norm": 1.0889617204666138, + "learning_rate": 0.0005318232776192418, + "loss": 3.085, + "step": 55125 + }, + { + "epoch": 3.745753499116728, + "grad_norm": 0.8158285617828369, + "learning_rate": 0.000531780812610409, + "loss": 3.5704, + "step": 55130 + }, + { + "epoch": 3.7460932191873897, + "grad_norm": 0.8806131482124329, + "learning_rate": 0.0005317383476015763, + "loss": 3.5143, + "step": 55135 + }, + { + "epoch": 3.7464329392580513, + "grad_norm": 0.685218095779419, + "learning_rate": 0.0005316958825927436, + "loss": 3.3815, + "step": 55140 + }, + { + "epoch": 3.7467726593287134, + "grad_norm": 0.9201647639274597, + "learning_rate": 0.0005316534175839108, + "loss": 3.2945, + "step": 55145 + }, + { + "epoch": 3.747112379399375, + "grad_norm": 1.0250530242919922, + "learning_rate": 0.0005316109525750782, + "loss": 3.5713, + "step": 55150 + }, + { + "epoch": 3.7474520994700367, + "grad_norm": 1.0740998983383179, + "learning_rate": 0.0005315684875662455, + "loss": 3.5787, + "step": 55155 + }, + { + "epoch": 3.7477918195406987, + "grad_norm": 0.891015887260437, + "learning_rate": 0.0005315260225574127, + "loss": 3.4705, + "step": 55160 + }, + { + "epoch": 3.7481315396113604, + "grad_norm": 0.862984299659729, + "learning_rate": 0.00053148355754858, + "loss": 3.6164, + "step": 55165 + }, + { + "epoch": 3.748471259682022, + "grad_norm": 0.8843459486961365, + "learning_rate": 0.0005314410925397473, + "loss": 3.3594, + "step": 55170 + }, + { + "epoch": 3.7488109797526836, + "grad_norm": 0.9694676995277405, + "learning_rate": 0.0005313986275309145, + "loss": 3.444, + "step": 55175 + }, + { + "epoch": 3.7491506998233457, + "grad_norm": 1.276371717453003, + "learning_rate": 0.0005313561625220818, + "loss": 3.5675, + "step": 55180 + }, + { + "epoch": 3.7494904198940073, + "grad_norm": 0.8703802824020386, + "learning_rate": 0.0005313136975132491, + "loss": 3.3623, + "step": 55185 + }, + { + "epoch": 3.749830139964669, + "grad_norm": 0.814553439617157, + "learning_rate": 0.0005312712325044164, + "loss": 3.6685, + "step": 55190 + }, + { + "epoch": 3.750169860035331, + "grad_norm": 0.9540380835533142, + "learning_rate": 0.0005312287674955837, + "loss": 3.2456, + "step": 55195 + }, + { + "epoch": 3.7505095801059927, + "grad_norm": 0.9192229509353638, + "learning_rate": 0.0005311863024867509, + "loss": 3.3748, + "step": 55200 + }, + { + "epoch": 3.7508493001766543, + "grad_norm": 0.768900990486145, + "learning_rate": 0.0005311438374779182, + "loss": 3.6243, + "step": 55205 + }, + { + "epoch": 3.7511890202473164, + "grad_norm": 1.1297773122787476, + "learning_rate": 0.0005311013724690855, + "loss": 3.5125, + "step": 55210 + }, + { + "epoch": 3.751528740317978, + "grad_norm": 0.7569725513458252, + "learning_rate": 0.0005310589074602527, + "loss": 3.476, + "step": 55215 + }, + { + "epoch": 3.7518684603886396, + "grad_norm": 1.0698707103729248, + "learning_rate": 0.00053101644245142, + "loss": 3.3171, + "step": 55220 + }, + { + "epoch": 3.7522081804593013, + "grad_norm": 0.9384840726852417, + "learning_rate": 0.0005309739774425874, + "loss": 3.4894, + "step": 55225 + }, + { + "epoch": 3.7525479005299633, + "grad_norm": 0.7603185176849365, + "learning_rate": 0.0005309315124337546, + "loss": 3.7406, + "step": 55230 + }, + { + "epoch": 3.752887620600625, + "grad_norm": 0.8559585213661194, + "learning_rate": 0.0005308890474249218, + "loss": 3.5188, + "step": 55235 + }, + { + "epoch": 3.7532273406712866, + "grad_norm": 0.769366979598999, + "learning_rate": 0.0005308465824160892, + "loss": 3.2712, + "step": 55240 + }, + { + "epoch": 3.7535670607419487, + "grad_norm": 1.2153433561325073, + "learning_rate": 0.0005308041174072564, + "loss": 3.815, + "step": 55245 + }, + { + "epoch": 3.7539067808126103, + "grad_norm": 0.8166508078575134, + "learning_rate": 0.0005307616523984236, + "loss": 3.7103, + "step": 55250 + }, + { + "epoch": 3.754246500883272, + "grad_norm": 0.6838829517364502, + "learning_rate": 0.0005307191873895911, + "loss": 3.3757, + "step": 55255 + }, + { + "epoch": 3.754586220953934, + "grad_norm": 1.6971923112869263, + "learning_rate": 0.0005306767223807583, + "loss": 3.5624, + "step": 55260 + }, + { + "epoch": 3.7549259410245956, + "grad_norm": 0.8775597214698792, + "learning_rate": 0.0005306342573719255, + "loss": 3.6253, + "step": 55265 + }, + { + "epoch": 3.7552656610952573, + "grad_norm": 0.8145177960395813, + "learning_rate": 0.0005305917923630929, + "loss": 3.5181, + "step": 55270 + }, + { + "epoch": 3.7556053811659194, + "grad_norm": 0.8092663884162903, + "learning_rate": 0.0005305493273542601, + "loss": 3.5185, + "step": 55275 + }, + { + "epoch": 3.755945101236581, + "grad_norm": 0.9875311255455017, + "learning_rate": 0.0005305068623454273, + "loss": 3.4841, + "step": 55280 + }, + { + "epoch": 3.7562848213072426, + "grad_norm": 0.9346238970756531, + "learning_rate": 0.0005304643973365946, + "loss": 3.5433, + "step": 55285 + }, + { + "epoch": 3.7566245413779047, + "grad_norm": 0.8469724059104919, + "learning_rate": 0.000530421932327762, + "loss": 3.5228, + "step": 55290 + }, + { + "epoch": 3.7569642614485663, + "grad_norm": 0.9983519911766052, + "learning_rate": 0.0005303794673189292, + "loss": 3.5077, + "step": 55295 + }, + { + "epoch": 3.757303981519228, + "grad_norm": 1.0515414476394653, + "learning_rate": 0.0005303370023100965, + "loss": 3.6276, + "step": 55300 + }, + { + "epoch": 3.75764370158989, + "grad_norm": 0.9465078115463257, + "learning_rate": 0.0005302945373012638, + "loss": 3.5428, + "step": 55305 + }, + { + "epoch": 3.7579834216605517, + "grad_norm": 1.003580093383789, + "learning_rate": 0.000530252072292431, + "loss": 3.6825, + "step": 55310 + }, + { + "epoch": 3.7583231417312133, + "grad_norm": 0.9517799615859985, + "learning_rate": 0.0005302096072835983, + "loss": 3.6813, + "step": 55315 + }, + { + "epoch": 3.7586628618018754, + "grad_norm": 0.922330915927887, + "learning_rate": 0.0005301671422747655, + "loss": 3.3574, + "step": 55320 + }, + { + "epoch": 3.759002581872537, + "grad_norm": 0.7383594512939453, + "learning_rate": 0.0005301246772659329, + "loss": 3.6515, + "step": 55325 + }, + { + "epoch": 3.7593423019431986, + "grad_norm": 1.2569410800933838, + "learning_rate": 0.0005300822122571002, + "loss": 3.5015, + "step": 55330 + }, + { + "epoch": 3.7596820220138607, + "grad_norm": 0.7399452328681946, + "learning_rate": 0.0005300397472482674, + "loss": 3.43, + "step": 55335 + }, + { + "epoch": 3.7600217420845223, + "grad_norm": 0.7153240442276001, + "learning_rate": 0.0005299972822394347, + "loss": 3.6163, + "step": 55340 + }, + { + "epoch": 3.760361462155184, + "grad_norm": 0.9509831070899963, + "learning_rate": 0.000529954817230602, + "loss": 3.6635, + "step": 55345 + }, + { + "epoch": 3.760701182225846, + "grad_norm": 0.9036169648170471, + "learning_rate": 0.0005299123522217692, + "loss": 3.28, + "step": 55350 + }, + { + "epoch": 3.7610409022965077, + "grad_norm": 0.8787235617637634, + "learning_rate": 0.0005298698872129365, + "loss": 3.7538, + "step": 55355 + }, + { + "epoch": 3.7613806223671693, + "grad_norm": 0.9020276069641113, + "learning_rate": 0.0005298274222041039, + "loss": 3.7184, + "step": 55360 + }, + { + "epoch": 3.7617203424378314, + "grad_norm": 0.8400426506996155, + "learning_rate": 0.0005297849571952711, + "loss": 3.5743, + "step": 55365 + }, + { + "epoch": 3.762060062508493, + "grad_norm": 1.2143874168395996, + "learning_rate": 0.0005297424921864385, + "loss": 3.2048, + "step": 55370 + }, + { + "epoch": 3.7623997825791546, + "grad_norm": 0.9159396290779114, + "learning_rate": 0.0005297000271776057, + "loss": 3.5762, + "step": 55375 + }, + { + "epoch": 3.7627395026498167, + "grad_norm": 0.8436747193336487, + "learning_rate": 0.0005296575621687729, + "loss": 3.5981, + "step": 55380 + }, + { + "epoch": 3.7630792227204783, + "grad_norm": 0.7627342343330383, + "learning_rate": 0.0005296150971599402, + "loss": 3.3867, + "step": 55385 + }, + { + "epoch": 3.76341894279114, + "grad_norm": 0.8818159699440002, + "learning_rate": 0.0005295726321511075, + "loss": 3.5034, + "step": 55390 + }, + { + "epoch": 3.763758662861802, + "grad_norm": 1.0074577331542969, + "learning_rate": 0.0005295301671422748, + "loss": 4.0743, + "step": 55395 + }, + { + "epoch": 3.7640983829324637, + "grad_norm": 0.8862717747688293, + "learning_rate": 0.0005294877021334421, + "loss": 3.4943, + "step": 55400 + }, + { + "epoch": 3.7644381030031253, + "grad_norm": 0.7939552068710327, + "learning_rate": 0.0005294452371246094, + "loss": 3.3237, + "step": 55405 + }, + { + "epoch": 3.7647778230737874, + "grad_norm": 0.7286713719367981, + "learning_rate": 0.0005294027721157766, + "loss": 3.5104, + "step": 55410 + }, + { + "epoch": 3.765117543144449, + "grad_norm": 0.7083745002746582, + "learning_rate": 0.0005293603071069439, + "loss": 3.4383, + "step": 55415 + }, + { + "epoch": 3.7654572632151107, + "grad_norm": 1.279093861579895, + "learning_rate": 0.0005293178420981111, + "loss": 3.6801, + "step": 55420 + }, + { + "epoch": 3.7657969832857727, + "grad_norm": 0.7574336528778076, + "learning_rate": 0.0005292753770892784, + "loss": 3.6631, + "step": 55425 + }, + { + "epoch": 3.7661367033564344, + "grad_norm": 0.9487844109535217, + "learning_rate": 0.0005292329120804458, + "loss": 3.4574, + "step": 55430 + }, + { + "epoch": 3.766476423427096, + "grad_norm": 0.9485490322113037, + "learning_rate": 0.000529190447071613, + "loss": 3.7347, + "step": 55435 + }, + { + "epoch": 3.766816143497758, + "grad_norm": 0.8379043340682983, + "learning_rate": 0.0005291479820627803, + "loss": 3.4706, + "step": 55440 + }, + { + "epoch": 3.7671558635684197, + "grad_norm": 0.9543617963790894, + "learning_rate": 0.0005291055170539476, + "loss": 3.4108, + "step": 55445 + }, + { + "epoch": 3.7674955836390813, + "grad_norm": 0.9282843470573425, + "learning_rate": 0.0005290630520451148, + "loss": 3.6889, + "step": 55450 + }, + { + "epoch": 3.7678353037097434, + "grad_norm": 0.8928327560424805, + "learning_rate": 0.000529020587036282, + "loss": 3.6793, + "step": 55455 + }, + { + "epoch": 3.768175023780405, + "grad_norm": 1.122862696647644, + "learning_rate": 0.0005289781220274494, + "loss": 3.422, + "step": 55460 + }, + { + "epoch": 3.7685147438510667, + "grad_norm": 0.8912630677223206, + "learning_rate": 0.0005289356570186167, + "loss": 3.2884, + "step": 55465 + }, + { + "epoch": 3.7688544639217287, + "grad_norm": 1.162979006767273, + "learning_rate": 0.000528893192009784, + "loss": 3.3164, + "step": 55470 + }, + { + "epoch": 3.7691941839923904, + "grad_norm": 1.2106269598007202, + "learning_rate": 0.0005288507270009513, + "loss": 3.4485, + "step": 55475 + }, + { + "epoch": 3.769533904063052, + "grad_norm": 0.845757782459259, + "learning_rate": 0.0005288082619921185, + "loss": 3.4758, + "step": 55480 + }, + { + "epoch": 3.769873624133714, + "grad_norm": 0.8576749563217163, + "learning_rate": 0.0005287657969832857, + "loss": 3.512, + "step": 55485 + }, + { + "epoch": 3.7702133442043757, + "grad_norm": 0.7643755674362183, + "learning_rate": 0.0005287233319744531, + "loss": 3.5678, + "step": 55490 + }, + { + "epoch": 3.7705530642750373, + "grad_norm": 0.8214532136917114, + "learning_rate": 0.0005286808669656203, + "loss": 3.576, + "step": 55495 + }, + { + "epoch": 3.7708927843456994, + "grad_norm": 0.8651350736618042, + "learning_rate": 0.0005286384019567876, + "loss": 3.5838, + "step": 55500 + }, + { + "epoch": 3.771232504416361, + "grad_norm": 0.9427428841590881, + "learning_rate": 0.000528595936947955, + "loss": 3.3835, + "step": 55505 + }, + { + "epoch": 3.7715722244870227, + "grad_norm": 0.887864351272583, + "learning_rate": 0.0005285534719391222, + "loss": 3.353, + "step": 55510 + }, + { + "epoch": 3.7719119445576843, + "grad_norm": 0.9157028198242188, + "learning_rate": 0.0005285110069302894, + "loss": 3.7803, + "step": 55515 + }, + { + "epoch": 3.7722516646283464, + "grad_norm": 0.6996296644210815, + "learning_rate": 0.0005284685419214567, + "loss": 3.696, + "step": 55520 + }, + { + "epoch": 3.772591384699008, + "grad_norm": 0.986216127872467, + "learning_rate": 0.000528426076912624, + "loss": 3.6178, + "step": 55525 + }, + { + "epoch": 3.7729311047696696, + "grad_norm": 0.7205371260643005, + "learning_rate": 0.0005283836119037912, + "loss": 3.2717, + "step": 55530 + }, + { + "epoch": 3.7732708248403317, + "grad_norm": 0.7644385099411011, + "learning_rate": 0.0005283411468949586, + "loss": 3.373, + "step": 55535 + }, + { + "epoch": 3.7736105449109933, + "grad_norm": 0.80385822057724, + "learning_rate": 0.0005282986818861259, + "loss": 3.4645, + "step": 55540 + }, + { + "epoch": 3.773950264981655, + "grad_norm": 0.9187750816345215, + "learning_rate": 0.0005282562168772931, + "loss": 3.6276, + "step": 55545 + }, + { + "epoch": 3.774289985052317, + "grad_norm": 0.8756825923919678, + "learning_rate": 0.0005282137518684604, + "loss": 3.6302, + "step": 55550 + }, + { + "epoch": 3.7746297051229787, + "grad_norm": 1.081735372543335, + "learning_rate": 0.0005281712868596277, + "loss": 3.5186, + "step": 55555 + }, + { + "epoch": 3.7749694251936403, + "grad_norm": 1.0715614557266235, + "learning_rate": 0.0005281288218507949, + "loss": 3.5001, + "step": 55560 + }, + { + "epoch": 3.775309145264302, + "grad_norm": 0.8683094382286072, + "learning_rate": 0.0005280863568419622, + "loss": 3.2655, + "step": 55565 + }, + { + "epoch": 3.775648865334964, + "grad_norm": 0.916874349117279, + "learning_rate": 0.0005280438918331295, + "loss": 3.4946, + "step": 55570 + }, + { + "epoch": 3.7759885854056257, + "grad_norm": 0.861598551273346, + "learning_rate": 0.0005280014268242968, + "loss": 3.8905, + "step": 55575 + }, + { + "epoch": 3.7763283054762873, + "grad_norm": 0.8323996067047119, + "learning_rate": 0.0005279589618154641, + "loss": 3.4692, + "step": 55580 + }, + { + "epoch": 3.7766680255469494, + "grad_norm": 1.052626371383667, + "learning_rate": 0.0005279164968066313, + "loss": 3.5184, + "step": 55585 + }, + { + "epoch": 3.777007745617611, + "grad_norm": 0.983048677444458, + "learning_rate": 0.0005278740317977986, + "loss": 3.6364, + "step": 55590 + }, + { + "epoch": 3.7773474656882726, + "grad_norm": 1.1843783855438232, + "learning_rate": 0.0005278315667889659, + "loss": 3.6942, + "step": 55595 + }, + { + "epoch": 3.7776871857589347, + "grad_norm": 0.8867509365081787, + "learning_rate": 0.0005277891017801331, + "loss": 3.6293, + "step": 55600 + }, + { + "epoch": 3.7780269058295963, + "grad_norm": 0.9513036012649536, + "learning_rate": 0.0005277466367713005, + "loss": 3.4245, + "step": 55605 + }, + { + "epoch": 3.778366625900258, + "grad_norm": 0.8840452432632446, + "learning_rate": 0.0005277041717624678, + "loss": 3.5631, + "step": 55610 + }, + { + "epoch": 3.77870634597092, + "grad_norm": 0.6065431833267212, + "learning_rate": 0.000527661706753635, + "loss": 3.6214, + "step": 55615 + }, + { + "epoch": 3.7790460660415817, + "grad_norm": 0.8780247569084167, + "learning_rate": 0.0005276192417448022, + "loss": 3.5385, + "step": 55620 + }, + { + "epoch": 3.7793857861122433, + "grad_norm": 0.758277177810669, + "learning_rate": 0.0005275767767359696, + "loss": 3.6123, + "step": 55625 + }, + { + "epoch": 3.7797255061829054, + "grad_norm": 0.9010626673698425, + "learning_rate": 0.0005275343117271368, + "loss": 3.7336, + "step": 55630 + }, + { + "epoch": 3.780065226253567, + "grad_norm": 0.7693651914596558, + "learning_rate": 0.000527491846718304, + "loss": 3.7174, + "step": 55635 + }, + { + "epoch": 3.7804049463242286, + "grad_norm": 0.8228757977485657, + "learning_rate": 0.0005274493817094715, + "loss": 3.7287, + "step": 55640 + }, + { + "epoch": 3.7807446663948907, + "grad_norm": 0.6740391850471497, + "learning_rate": 0.0005274069167006387, + "loss": 3.4981, + "step": 55645 + }, + { + "epoch": 3.7810843864655523, + "grad_norm": 0.8589305281639099, + "learning_rate": 0.0005273644516918059, + "loss": 3.4607, + "step": 55650 + }, + { + "epoch": 3.781424106536214, + "grad_norm": 0.7811538577079773, + "learning_rate": 0.0005273219866829733, + "loss": 3.6053, + "step": 55655 + }, + { + "epoch": 3.781763826606876, + "grad_norm": 0.9331864714622498, + "learning_rate": 0.0005272795216741405, + "loss": 3.5713, + "step": 55660 + }, + { + "epoch": 3.7821035466775377, + "grad_norm": 1.2087920904159546, + "learning_rate": 0.0005272370566653077, + "loss": 3.3601, + "step": 55665 + }, + { + "epoch": 3.7824432667481993, + "grad_norm": 0.8008260726928711, + "learning_rate": 0.0005271945916564752, + "loss": 3.705, + "step": 55670 + }, + { + "epoch": 3.7827829868188614, + "grad_norm": 0.9629670977592468, + "learning_rate": 0.0005271521266476424, + "loss": 3.4364, + "step": 55675 + }, + { + "epoch": 3.783122706889523, + "grad_norm": 0.7897980809211731, + "learning_rate": 0.0005271096616388096, + "loss": 3.5066, + "step": 55680 + }, + { + "epoch": 3.7834624269601846, + "grad_norm": 1.0639491081237793, + "learning_rate": 0.0005270671966299769, + "loss": 3.5499, + "step": 55685 + }, + { + "epoch": 3.7838021470308467, + "grad_norm": 1.1327036619186401, + "learning_rate": 0.0005270247316211442, + "loss": 3.3359, + "step": 55690 + }, + { + "epoch": 3.7841418671015083, + "grad_norm": 0.9999335408210754, + "learning_rate": 0.0005269822666123114, + "loss": 3.831, + "step": 55695 + }, + { + "epoch": 3.78448158717217, + "grad_norm": 0.8211253881454468, + "learning_rate": 0.0005269398016034787, + "loss": 3.3374, + "step": 55700 + }, + { + "epoch": 3.784821307242832, + "grad_norm": 0.9879361987113953, + "learning_rate": 0.0005268973365946461, + "loss": 3.4373, + "step": 55705 + }, + { + "epoch": 3.7851610273134937, + "grad_norm": 0.8416990637779236, + "learning_rate": 0.0005268548715858134, + "loss": 3.528, + "step": 55710 + }, + { + "epoch": 3.7855007473841553, + "grad_norm": 0.9534209966659546, + "learning_rate": 0.0005268124065769806, + "loss": 3.4609, + "step": 55715 + }, + { + "epoch": 3.7858404674548174, + "grad_norm": 0.9439724683761597, + "learning_rate": 0.0005267699415681478, + "loss": 3.5412, + "step": 55720 + }, + { + "epoch": 3.786180187525479, + "grad_norm": 0.8800684809684753, + "learning_rate": 0.0005267274765593152, + "loss": 3.4106, + "step": 55725 + }, + { + "epoch": 3.7865199075961407, + "grad_norm": 0.9825429916381836, + "learning_rate": 0.0005266850115504824, + "loss": 3.788, + "step": 55730 + }, + { + "epoch": 3.7868596276668027, + "grad_norm": 0.7447264790534973, + "learning_rate": 0.0005266425465416496, + "loss": 3.3975, + "step": 55735 + }, + { + "epoch": 3.7871993477374644, + "grad_norm": 0.882600724697113, + "learning_rate": 0.0005266000815328171, + "loss": 3.6908, + "step": 55740 + }, + { + "epoch": 3.787539067808126, + "grad_norm": 0.7813102006912231, + "learning_rate": 0.0005265576165239843, + "loss": 3.5993, + "step": 55745 + }, + { + "epoch": 3.787878787878788, + "grad_norm": 0.9392270445823669, + "learning_rate": 0.0005265151515151515, + "loss": 3.4765, + "step": 55750 + }, + { + "epoch": 3.7882185079494497, + "grad_norm": 0.8283581137657166, + "learning_rate": 0.0005264726865063189, + "loss": 3.6835, + "step": 55755 + }, + { + "epoch": 3.7885582280201113, + "grad_norm": 0.7559888958930969, + "learning_rate": 0.0005264302214974861, + "loss": 3.485, + "step": 55760 + }, + { + "epoch": 3.7888979480907734, + "grad_norm": 0.7605040669441223, + "learning_rate": 0.0005263877564886533, + "loss": 3.4826, + "step": 55765 + }, + { + "epoch": 3.789237668161435, + "grad_norm": 0.8659229278564453, + "learning_rate": 0.0005263452914798206, + "loss": 3.6602, + "step": 55770 + }, + { + "epoch": 3.7895773882320967, + "grad_norm": 0.910800039768219, + "learning_rate": 0.000526302826470988, + "loss": 3.556, + "step": 55775 + }, + { + "epoch": 3.7899171083027587, + "grad_norm": 1.0667128562927246, + "learning_rate": 0.0005262603614621552, + "loss": 3.6387, + "step": 55780 + }, + { + "epoch": 3.7902568283734204, + "grad_norm": 0.8082834482192993, + "learning_rate": 0.0005262178964533225, + "loss": 3.8618, + "step": 55785 + }, + { + "epoch": 3.790596548444082, + "grad_norm": 0.8752985596656799, + "learning_rate": 0.0005261754314444898, + "loss": 3.4569, + "step": 55790 + }, + { + "epoch": 3.790936268514744, + "grad_norm": 0.7948899269104004, + "learning_rate": 0.000526132966435657, + "loss": 3.5214, + "step": 55795 + }, + { + "epoch": 3.7912759885854057, + "grad_norm": 0.9071364998817444, + "learning_rate": 0.0005260905014268243, + "loss": 3.2972, + "step": 55800 + }, + { + "epoch": 3.7916157086560673, + "grad_norm": 1.055819034576416, + "learning_rate": 0.0005260480364179915, + "loss": 3.4413, + "step": 55805 + }, + { + "epoch": 3.7919554287267294, + "grad_norm": 0.7746694087982178, + "learning_rate": 0.0005260055714091589, + "loss": 3.4812, + "step": 55810 + }, + { + "epoch": 3.792295148797391, + "grad_norm": 0.8002044558525085, + "learning_rate": 0.0005259631064003262, + "loss": 3.4137, + "step": 55815 + }, + { + "epoch": 3.7926348688680527, + "grad_norm": 0.8296111226081848, + "learning_rate": 0.0005259206413914934, + "loss": 3.7835, + "step": 55820 + }, + { + "epoch": 3.7929745889387148, + "grad_norm": 0.9452680945396423, + "learning_rate": 0.0005258781763826607, + "loss": 3.4387, + "step": 55825 + }, + { + "epoch": 3.7933143090093764, + "grad_norm": 0.8038287162780762, + "learning_rate": 0.000525835711373828, + "loss": 3.669, + "step": 55830 + }, + { + "epoch": 3.793654029080038, + "grad_norm": 1.1663395166397095, + "learning_rate": 0.0005257932463649952, + "loss": 3.4569, + "step": 55835 + }, + { + "epoch": 3.7939937491507, + "grad_norm": 1.2264289855957031, + "learning_rate": 0.0005257507813561625, + "loss": 3.543, + "step": 55840 + }, + { + "epoch": 3.7943334692213617, + "grad_norm": 0.8041330575942993, + "learning_rate": 0.0005257083163473299, + "loss": 3.5046, + "step": 55845 + }, + { + "epoch": 3.7946731892920234, + "grad_norm": 0.918626606464386, + "learning_rate": 0.0005256658513384971, + "loss": 3.3014, + "step": 55850 + }, + { + "epoch": 3.795012909362685, + "grad_norm": 0.9897686243057251, + "learning_rate": 0.0005256233863296644, + "loss": 3.4, + "step": 55855 + }, + { + "epoch": 3.795352629433347, + "grad_norm": 0.8054947853088379, + "learning_rate": 0.0005255809213208317, + "loss": 3.5353, + "step": 55860 + }, + { + "epoch": 3.7956923495040087, + "grad_norm": 0.9137596487998962, + "learning_rate": 0.0005255384563119989, + "loss": 3.4681, + "step": 55865 + }, + { + "epoch": 3.7960320695746703, + "grad_norm": 0.9184945225715637, + "learning_rate": 0.0005254959913031661, + "loss": 3.6371, + "step": 55870 + }, + { + "epoch": 3.7963717896453324, + "grad_norm": 1.0417685508728027, + "learning_rate": 0.0005254535262943335, + "loss": 3.3319, + "step": 55875 + }, + { + "epoch": 3.796711509715994, + "grad_norm": 0.769091784954071, + "learning_rate": 0.0005254110612855008, + "loss": 3.6697, + "step": 55880 + }, + { + "epoch": 3.7970512297866557, + "grad_norm": 0.8785198926925659, + "learning_rate": 0.000525368596276668, + "loss": 3.6896, + "step": 55885 + }, + { + "epoch": 3.7973909498573177, + "grad_norm": 0.9343380928039551, + "learning_rate": 0.0005253261312678354, + "loss": 3.6572, + "step": 55890 + }, + { + "epoch": 3.7977306699279794, + "grad_norm": 0.9476320743560791, + "learning_rate": 0.0005252836662590026, + "loss": 3.3724, + "step": 55895 + }, + { + "epoch": 3.798070389998641, + "grad_norm": 1.531165599822998, + "learning_rate": 0.0005252412012501698, + "loss": 3.6377, + "step": 55900 + }, + { + "epoch": 3.7984101100693026, + "grad_norm": 1.2817524671554565, + "learning_rate": 0.0005251987362413372, + "loss": 3.7212, + "step": 55905 + }, + { + "epoch": 3.7987498301399647, + "grad_norm": 0.9495688080787659, + "learning_rate": 0.0005251562712325044, + "loss": 3.0868, + "step": 55910 + }, + { + "epoch": 3.7990895502106263, + "grad_norm": 0.716037929058075, + "learning_rate": 0.0005251138062236717, + "loss": 3.4624, + "step": 55915 + }, + { + "epoch": 3.799429270281288, + "grad_norm": 1.2376518249511719, + "learning_rate": 0.000525071341214839, + "loss": 3.6066, + "step": 55920 + }, + { + "epoch": 3.79976899035195, + "grad_norm": 0.7719739675521851, + "learning_rate": 0.0005250288762060063, + "loss": 3.6685, + "step": 55925 + }, + { + "epoch": 3.8001087104226117, + "grad_norm": 0.8896769881248474, + "learning_rate": 0.0005249864111971735, + "loss": 3.6196, + "step": 55930 + }, + { + "epoch": 3.8004484304932733, + "grad_norm": 0.694242537021637, + "learning_rate": 0.0005249439461883408, + "loss": 3.7168, + "step": 55935 + }, + { + "epoch": 3.8007881505639354, + "grad_norm": 0.826335072517395, + "learning_rate": 0.0005249014811795081, + "loss": 3.3261, + "step": 55940 + }, + { + "epoch": 3.801127870634597, + "grad_norm": 0.9762623906135559, + "learning_rate": 0.0005248590161706753, + "loss": 3.5885, + "step": 55945 + }, + { + "epoch": 3.8014675907052586, + "grad_norm": 0.9352214932441711, + "learning_rate": 0.0005248165511618427, + "loss": 3.5257, + "step": 55950 + }, + { + "epoch": 3.8018073107759207, + "grad_norm": 0.826858639717102, + "learning_rate": 0.00052477408615301, + "loss": 3.1532, + "step": 55955 + }, + { + "epoch": 3.8021470308465823, + "grad_norm": 0.7236220240592957, + "learning_rate": 0.0005247316211441772, + "loss": 3.534, + "step": 55960 + }, + { + "epoch": 3.802486750917244, + "grad_norm": 0.8774734139442444, + "learning_rate": 0.0005246891561353445, + "loss": 3.4565, + "step": 55965 + }, + { + "epoch": 3.802826470987906, + "grad_norm": 1.298330307006836, + "learning_rate": 0.0005246466911265117, + "loss": 3.6107, + "step": 55970 + }, + { + "epoch": 3.8031661910585677, + "grad_norm": 1.1160427331924438, + "learning_rate": 0.000524604226117679, + "loss": 3.5668, + "step": 55975 + }, + { + "epoch": 3.8035059111292293, + "grad_norm": 0.8108828663825989, + "learning_rate": 0.0005245617611088463, + "loss": 3.4341, + "step": 55980 + }, + { + "epoch": 3.8038456311998914, + "grad_norm": 0.838599681854248, + "learning_rate": 0.0005245192961000136, + "loss": 3.6294, + "step": 55985 + }, + { + "epoch": 3.804185351270553, + "grad_norm": 0.8860453367233276, + "learning_rate": 0.0005244768310911809, + "loss": 3.4263, + "step": 55990 + }, + { + "epoch": 3.8045250713412146, + "grad_norm": 0.9518068432807922, + "learning_rate": 0.0005244343660823482, + "loss": 3.4511, + "step": 55995 + }, + { + "epoch": 3.8048647914118767, + "grad_norm": 0.9574953317642212, + "learning_rate": 0.0005243919010735154, + "loss": 3.394, + "step": 56000 + }, + { + "epoch": 3.8052045114825384, + "grad_norm": 0.9196003079414368, + "learning_rate": 0.0005243494360646826, + "loss": 3.3477, + "step": 56005 + }, + { + "epoch": 3.8055442315532, + "grad_norm": 1.246414303779602, + "learning_rate": 0.00052430697105585, + "loss": 3.4007, + "step": 56010 + }, + { + "epoch": 3.805883951623862, + "grad_norm": 0.7405118942260742, + "learning_rate": 0.0005242645060470172, + "loss": 3.446, + "step": 56015 + }, + { + "epoch": 3.8062236716945237, + "grad_norm": 0.7583832740783691, + "learning_rate": 0.0005242220410381845, + "loss": 3.6075, + "step": 56020 + }, + { + "epoch": 3.8065633917651853, + "grad_norm": 0.9404184222221375, + "learning_rate": 0.0005241795760293519, + "loss": 3.4668, + "step": 56025 + }, + { + "epoch": 3.8069031118358474, + "grad_norm": 0.7318946123123169, + "learning_rate": 0.0005241371110205191, + "loss": 3.6657, + "step": 56030 + }, + { + "epoch": 3.807242831906509, + "grad_norm": 0.8825598359107971, + "learning_rate": 0.0005240946460116863, + "loss": 3.3975, + "step": 56035 + }, + { + "epoch": 3.8075825519771707, + "grad_norm": 0.6793971657752991, + "learning_rate": 0.0005240521810028537, + "loss": 3.5602, + "step": 56040 + }, + { + "epoch": 3.8079222720478327, + "grad_norm": 0.831713855266571, + "learning_rate": 0.0005240097159940209, + "loss": 3.4235, + "step": 56045 + }, + { + "epoch": 3.8082619921184944, + "grad_norm": 1.0683919191360474, + "learning_rate": 0.0005239672509851882, + "loss": 3.5368, + "step": 56050 + }, + { + "epoch": 3.808601712189156, + "grad_norm": 1.0225378274917603, + "learning_rate": 0.0005239247859763556, + "loss": 3.6813, + "step": 56055 + }, + { + "epoch": 3.808941432259818, + "grad_norm": 0.851654052734375, + "learning_rate": 0.0005238823209675228, + "loss": 3.3987, + "step": 56060 + }, + { + "epoch": 3.8092811523304797, + "grad_norm": 0.7730807065963745, + "learning_rate": 0.0005238398559586901, + "loss": 3.6825, + "step": 56065 + }, + { + "epoch": 3.8096208724011413, + "grad_norm": 0.9623993039131165, + "learning_rate": 0.0005237973909498573, + "loss": 3.5484, + "step": 56070 + }, + { + "epoch": 3.8099605924718034, + "grad_norm": 1.2083771228790283, + "learning_rate": 0.0005237549259410246, + "loss": 3.7746, + "step": 56075 + }, + { + "epoch": 3.810300312542465, + "grad_norm": 1.0333926677703857, + "learning_rate": 0.0005237124609321919, + "loss": 3.6467, + "step": 56080 + }, + { + "epoch": 3.8106400326131267, + "grad_norm": 0.9033041596412659, + "learning_rate": 0.0005236699959233591, + "loss": 3.5035, + "step": 56085 + }, + { + "epoch": 3.8109797526837887, + "grad_norm": 1.1437511444091797, + "learning_rate": 0.0005236275309145265, + "loss": 3.3793, + "step": 56090 + }, + { + "epoch": 3.8113194727544504, + "grad_norm": 1.888250708580017, + "learning_rate": 0.0005235850659056938, + "loss": 3.6034, + "step": 56095 + }, + { + "epoch": 3.811659192825112, + "grad_norm": 0.8686556816101074, + "learning_rate": 0.000523542600896861, + "loss": 3.6128, + "step": 56100 + }, + { + "epoch": 3.811998912895774, + "grad_norm": 1.0905582904815674, + "learning_rate": 0.0005235001358880282, + "loss": 3.6279, + "step": 56105 + }, + { + "epoch": 3.8123386329664357, + "grad_norm": 1.099255084991455, + "learning_rate": 0.0005234576708791956, + "loss": 3.7917, + "step": 56110 + }, + { + "epoch": 3.8126783530370973, + "grad_norm": 0.9166022539138794, + "learning_rate": 0.0005234152058703628, + "loss": 3.7595, + "step": 56115 + }, + { + "epoch": 3.8130180731077594, + "grad_norm": 0.8268341422080994, + "learning_rate": 0.00052337274086153, + "loss": 3.5227, + "step": 56120 + }, + { + "epoch": 3.813357793178421, + "grad_norm": 0.8269185423851013, + "learning_rate": 0.0005233302758526975, + "loss": 3.634, + "step": 56125 + }, + { + "epoch": 3.8136975132490827, + "grad_norm": 1.073610544204712, + "learning_rate": 0.0005232878108438647, + "loss": 3.494, + "step": 56130 + }, + { + "epoch": 3.8140372333197448, + "grad_norm": 0.9536941647529602, + "learning_rate": 0.0005232453458350319, + "loss": 3.3434, + "step": 56135 + }, + { + "epoch": 3.8143769533904064, + "grad_norm": 1.1666574478149414, + "learning_rate": 0.0005232028808261993, + "loss": 3.348, + "step": 56140 + }, + { + "epoch": 3.814716673461068, + "grad_norm": 1.099188208580017, + "learning_rate": 0.0005231604158173665, + "loss": 3.5287, + "step": 56145 + }, + { + "epoch": 3.81505639353173, + "grad_norm": 0.8057329654693604, + "learning_rate": 0.0005231179508085337, + "loss": 3.7105, + "step": 56150 + }, + { + "epoch": 3.8153961136023917, + "grad_norm": 1.3922618627548218, + "learning_rate": 0.0005230754857997012, + "loss": 3.6472, + "step": 56155 + }, + { + "epoch": 3.8157358336730534, + "grad_norm": 1.201080322265625, + "learning_rate": 0.0005230330207908684, + "loss": 3.3811, + "step": 56160 + }, + { + "epoch": 3.8160755537437154, + "grad_norm": 1.0848336219787598, + "learning_rate": 0.0005229905557820356, + "loss": 3.6336, + "step": 56165 + }, + { + "epoch": 3.816415273814377, + "grad_norm": 0.9715064167976379, + "learning_rate": 0.0005229480907732029, + "loss": 3.5574, + "step": 56170 + }, + { + "epoch": 3.8167549938850387, + "grad_norm": 0.7345954179763794, + "learning_rate": 0.0005229056257643702, + "loss": 3.4054, + "step": 56175 + }, + { + "epoch": 3.8170947139557008, + "grad_norm": 0.7796631455421448, + "learning_rate": 0.0005228631607555374, + "loss": 3.6312, + "step": 56180 + }, + { + "epoch": 3.8174344340263624, + "grad_norm": 0.7440600991249084, + "learning_rate": 0.0005228206957467047, + "loss": 3.6778, + "step": 56185 + }, + { + "epoch": 3.817774154097024, + "grad_norm": 0.9003073573112488, + "learning_rate": 0.0005227782307378721, + "loss": 3.6073, + "step": 56190 + }, + { + "epoch": 3.8181138741676857, + "grad_norm": 1.0145468711853027, + "learning_rate": 0.0005227357657290393, + "loss": 3.7774, + "step": 56195 + }, + { + "epoch": 3.8184535942383477, + "grad_norm": 0.8687780499458313, + "learning_rate": 0.0005226933007202066, + "loss": 3.6586, + "step": 56200 + }, + { + "epoch": 3.8187933143090094, + "grad_norm": 0.8589841723442078, + "learning_rate": 0.0005226508357113738, + "loss": 3.1222, + "step": 56205 + }, + { + "epoch": 3.819133034379671, + "grad_norm": 0.843184769153595, + "learning_rate": 0.0005226083707025411, + "loss": 3.5567, + "step": 56210 + }, + { + "epoch": 3.819472754450333, + "grad_norm": 1.095578670501709, + "learning_rate": 0.0005225659056937084, + "loss": 3.5497, + "step": 56215 + }, + { + "epoch": 3.8198124745209947, + "grad_norm": 0.9289934039115906, + "learning_rate": 0.0005225234406848756, + "loss": 3.7642, + "step": 56220 + }, + { + "epoch": 3.8201521945916563, + "grad_norm": 0.9652898907661438, + "learning_rate": 0.000522480975676043, + "loss": 3.4436, + "step": 56225 + }, + { + "epoch": 3.8204919146623184, + "grad_norm": 0.7541878819465637, + "learning_rate": 0.0005224385106672103, + "loss": 3.3053, + "step": 56230 + }, + { + "epoch": 3.82083163473298, + "grad_norm": 0.7974632978439331, + "learning_rate": 0.0005223960456583775, + "loss": 3.6799, + "step": 56235 + }, + { + "epoch": 3.8211713548036417, + "grad_norm": 0.8026483654975891, + "learning_rate": 0.0005223535806495448, + "loss": 3.6149, + "step": 56240 + }, + { + "epoch": 3.8215110748743033, + "grad_norm": 0.8584368824958801, + "learning_rate": 0.0005223111156407121, + "loss": 3.4489, + "step": 56245 + }, + { + "epoch": 3.8218507949449654, + "grad_norm": 1.3426589965820312, + "learning_rate": 0.0005222686506318793, + "loss": 3.3665, + "step": 56250 + }, + { + "epoch": 3.822190515015627, + "grad_norm": 1.054150104522705, + "learning_rate": 0.0005222261856230465, + "loss": 3.7507, + "step": 56255 + }, + { + "epoch": 3.8225302350862886, + "grad_norm": 0.9774046540260315, + "learning_rate": 0.000522183720614214, + "loss": 3.4517, + "step": 56260 + }, + { + "epoch": 3.8228699551569507, + "grad_norm": 1.6141369342803955, + "learning_rate": 0.0005221412556053812, + "loss": 3.5137, + "step": 56265 + }, + { + "epoch": 3.8232096752276123, + "grad_norm": 0.7604255080223083, + "learning_rate": 0.0005220987905965484, + "loss": 3.2613, + "step": 56270 + }, + { + "epoch": 3.823549395298274, + "grad_norm": 0.7882590293884277, + "learning_rate": 0.0005220563255877158, + "loss": 3.5131, + "step": 56275 + }, + { + "epoch": 3.823889115368936, + "grad_norm": 1.00392746925354, + "learning_rate": 0.000522013860578883, + "loss": 3.6647, + "step": 56280 + }, + { + "epoch": 3.8242288354395977, + "grad_norm": 1.240463376045227, + "learning_rate": 0.0005219713955700502, + "loss": 3.6492, + "step": 56285 + }, + { + "epoch": 3.8245685555102593, + "grad_norm": 0.9950423240661621, + "learning_rate": 0.0005219289305612176, + "loss": 3.6989, + "step": 56290 + }, + { + "epoch": 3.8249082755809214, + "grad_norm": 1.0007741451263428, + "learning_rate": 0.0005218864655523849, + "loss": 3.6167, + "step": 56295 + }, + { + "epoch": 3.825247995651583, + "grad_norm": 0.6978954672813416, + "learning_rate": 0.0005218440005435521, + "loss": 3.5706, + "step": 56300 + }, + { + "epoch": 3.8255877157222447, + "grad_norm": 0.8745890259742737, + "learning_rate": 0.0005218015355347194, + "loss": 3.526, + "step": 56305 + }, + { + "epoch": 3.8259274357929067, + "grad_norm": 4.693668842315674, + "learning_rate": 0.0005217590705258867, + "loss": 3.7068, + "step": 56310 + }, + { + "epoch": 3.8262671558635684, + "grad_norm": 0.716612696647644, + "learning_rate": 0.0005217166055170539, + "loss": 3.6121, + "step": 56315 + }, + { + "epoch": 3.82660687593423, + "grad_norm": 0.8291292190551758, + "learning_rate": 0.0005216741405082212, + "loss": 3.2219, + "step": 56320 + }, + { + "epoch": 3.826946596004892, + "grad_norm": 0.8352263569831848, + "learning_rate": 0.0005216316754993885, + "loss": 3.503, + "step": 56325 + }, + { + "epoch": 3.8272863160755537, + "grad_norm": 0.7407249808311462, + "learning_rate": 0.0005215892104905558, + "loss": 3.4592, + "step": 56330 + }, + { + "epoch": 3.8276260361462153, + "grad_norm": 0.8932452201843262, + "learning_rate": 0.0005215467454817231, + "loss": 3.5913, + "step": 56335 + }, + { + "epoch": 3.8279657562168774, + "grad_norm": 0.7984388470649719, + "learning_rate": 0.0005215042804728904, + "loss": 3.5978, + "step": 56340 + }, + { + "epoch": 3.828305476287539, + "grad_norm": 0.8221063017845154, + "learning_rate": 0.0005214618154640576, + "loss": 3.505, + "step": 56345 + }, + { + "epoch": 3.8286451963582007, + "grad_norm": 0.8480583429336548, + "learning_rate": 0.0005214193504552249, + "loss": 3.7177, + "step": 56350 + }, + { + "epoch": 3.8289849164288627, + "grad_norm": 1.1518019437789917, + "learning_rate": 0.0005213768854463921, + "loss": 3.1518, + "step": 56355 + }, + { + "epoch": 3.8293246364995244, + "grad_norm": 0.896429181098938, + "learning_rate": 0.0005213344204375594, + "loss": 3.4817, + "step": 56360 + }, + { + "epoch": 3.829664356570186, + "grad_norm": 0.9401424527168274, + "learning_rate": 0.0005212919554287268, + "loss": 3.6466, + "step": 56365 + }, + { + "epoch": 3.830004076640848, + "grad_norm": 0.7908754944801331, + "learning_rate": 0.000521249490419894, + "loss": 3.5004, + "step": 56370 + }, + { + "epoch": 3.8303437967115097, + "grad_norm": 1.1310920715332031, + "learning_rate": 0.0005212070254110613, + "loss": 3.4532, + "step": 56375 + }, + { + "epoch": 3.8306835167821713, + "grad_norm": 0.9895877242088318, + "learning_rate": 0.0005211645604022286, + "loss": 3.4025, + "step": 56380 + }, + { + "epoch": 3.8310232368528334, + "grad_norm": 0.9851517677307129, + "learning_rate": 0.0005211220953933958, + "loss": 3.4035, + "step": 56385 + }, + { + "epoch": 3.831362956923495, + "grad_norm": 1.023783564567566, + "learning_rate": 0.0005210796303845632, + "loss": 3.4235, + "step": 56390 + }, + { + "epoch": 3.8317026769941567, + "grad_norm": 1.093160629272461, + "learning_rate": 0.0005210371653757304, + "loss": 3.5946, + "step": 56395 + }, + { + "epoch": 3.8320423970648188, + "grad_norm": 0.7521139979362488, + "learning_rate": 0.0005209947003668977, + "loss": 3.5783, + "step": 56400 + }, + { + "epoch": 3.8323821171354804, + "grad_norm": 0.8869144916534424, + "learning_rate": 0.000520952235358065, + "loss": 3.6292, + "step": 56405 + }, + { + "epoch": 3.832721837206142, + "grad_norm": 0.7833171486854553, + "learning_rate": 0.0005209097703492323, + "loss": 3.6418, + "step": 56410 + }, + { + "epoch": 3.833061557276804, + "grad_norm": 0.864266574382782, + "learning_rate": 0.0005208673053403995, + "loss": 3.3513, + "step": 56415 + }, + { + "epoch": 3.8334012773474657, + "grad_norm": 0.9043874144554138, + "learning_rate": 0.0005208248403315668, + "loss": 3.3321, + "step": 56420 + }, + { + "epoch": 3.8337409974181273, + "grad_norm": 0.8984978199005127, + "learning_rate": 0.0005207823753227341, + "loss": 3.4075, + "step": 56425 + }, + { + "epoch": 3.8340807174887894, + "grad_norm": 0.9726295471191406, + "learning_rate": 0.0005207399103139013, + "loss": 3.4485, + "step": 56430 + }, + { + "epoch": 3.834420437559451, + "grad_norm": 0.948248028755188, + "learning_rate": 0.0005206974453050687, + "loss": 3.4806, + "step": 56435 + }, + { + "epoch": 3.8347601576301127, + "grad_norm": 0.9400249123573303, + "learning_rate": 0.000520654980296236, + "loss": 3.2897, + "step": 56440 + }, + { + "epoch": 3.8350998777007748, + "grad_norm": 0.8516218662261963, + "learning_rate": 0.0005206125152874032, + "loss": 3.5123, + "step": 56445 + }, + { + "epoch": 3.8354395977714364, + "grad_norm": 0.8428704142570496, + "learning_rate": 0.0005205700502785705, + "loss": 3.5174, + "step": 56450 + }, + { + "epoch": 3.835779317842098, + "grad_norm": 0.7582258582115173, + "learning_rate": 0.0005205275852697377, + "loss": 3.64, + "step": 56455 + }, + { + "epoch": 3.83611903791276, + "grad_norm": 0.7963328957557678, + "learning_rate": 0.000520485120260905, + "loss": 3.6281, + "step": 56460 + }, + { + "epoch": 3.8364587579834217, + "grad_norm": 0.8122186064720154, + "learning_rate": 0.0005204426552520723, + "loss": 3.4776, + "step": 56465 + }, + { + "epoch": 3.8367984780540834, + "grad_norm": 0.8091825842857361, + "learning_rate": 0.0005204001902432396, + "loss": 3.4745, + "step": 56470 + }, + { + "epoch": 3.8371381981247454, + "grad_norm": 0.9374104738235474, + "learning_rate": 0.0005203577252344069, + "loss": 3.6179, + "step": 56475 + }, + { + "epoch": 3.837477918195407, + "grad_norm": 1.0840269327163696, + "learning_rate": 0.0005203152602255742, + "loss": 3.5384, + "step": 56480 + }, + { + "epoch": 3.8378176382660687, + "grad_norm": 0.8359546065330505, + "learning_rate": 0.0005202727952167414, + "loss": 3.693, + "step": 56485 + }, + { + "epoch": 3.8381573583367308, + "grad_norm": 1.3815990686416626, + "learning_rate": 0.0005202303302079086, + "loss": 3.5633, + "step": 56490 + }, + { + "epoch": 3.8384970784073924, + "grad_norm": 0.7971181869506836, + "learning_rate": 0.000520187865199076, + "loss": 3.8695, + "step": 56495 + }, + { + "epoch": 3.838836798478054, + "grad_norm": 1.3229036331176758, + "learning_rate": 0.0005201454001902432, + "loss": 3.4365, + "step": 56500 + }, + { + "epoch": 3.839176518548716, + "grad_norm": 0.737005352973938, + "learning_rate": 0.0005201029351814105, + "loss": 3.5118, + "step": 56505 + }, + { + "epoch": 3.8395162386193777, + "grad_norm": 0.9250162243843079, + "learning_rate": 0.0005200604701725779, + "loss": 3.7278, + "step": 56510 + }, + { + "epoch": 3.8398559586900394, + "grad_norm": 0.8831732273101807, + "learning_rate": 0.0005200180051637451, + "loss": 3.614, + "step": 56515 + }, + { + "epoch": 3.8401956787607014, + "grad_norm": 1.194557547569275, + "learning_rate": 0.0005199755401549123, + "loss": 3.64, + "step": 56520 + }, + { + "epoch": 3.840535398831363, + "grad_norm": 0.7807213068008423, + "learning_rate": 0.0005199330751460797, + "loss": 3.3128, + "step": 56525 + }, + { + "epoch": 3.8408751189020247, + "grad_norm": 0.9032267332077026, + "learning_rate": 0.0005198906101372469, + "loss": 3.5767, + "step": 56530 + }, + { + "epoch": 3.8412148389726863, + "grad_norm": 1.3909187316894531, + "learning_rate": 0.0005198481451284141, + "loss": 3.3888, + "step": 56535 + }, + { + "epoch": 3.8415545590433484, + "grad_norm": 0.9029501676559448, + "learning_rate": 0.0005198056801195816, + "loss": 3.7472, + "step": 56540 + }, + { + "epoch": 3.84189427911401, + "grad_norm": 1.0491210222244263, + "learning_rate": 0.0005197632151107488, + "loss": 3.6895, + "step": 56545 + }, + { + "epoch": 3.8422339991846717, + "grad_norm": 0.7877796292304993, + "learning_rate": 0.000519720750101916, + "loss": 3.6207, + "step": 56550 + }, + { + "epoch": 3.8425737192553338, + "grad_norm": 0.9900450706481934, + "learning_rate": 0.0005196782850930833, + "loss": 3.711, + "step": 56555 + }, + { + "epoch": 3.8429134393259954, + "grad_norm": 0.8556744456291199, + "learning_rate": 0.0005196358200842506, + "loss": 3.6598, + "step": 56560 + }, + { + "epoch": 3.843253159396657, + "grad_norm": 0.8430207371711731, + "learning_rate": 0.0005195933550754178, + "loss": 3.4114, + "step": 56565 + }, + { + "epoch": 3.843592879467319, + "grad_norm": 0.9363009333610535, + "learning_rate": 0.0005195508900665851, + "loss": 3.4848, + "step": 56570 + }, + { + "epoch": 3.8439325995379807, + "grad_norm": 0.8533961772918701, + "learning_rate": 0.0005195084250577525, + "loss": 3.4063, + "step": 56575 + }, + { + "epoch": 3.8442723196086424, + "grad_norm": 0.6958030462265015, + "learning_rate": 0.0005194659600489197, + "loss": 3.5669, + "step": 56580 + }, + { + "epoch": 3.844612039679304, + "grad_norm": 0.9663972854614258, + "learning_rate": 0.000519423495040087, + "loss": 3.7466, + "step": 56585 + }, + { + "epoch": 3.844951759749966, + "grad_norm": 1.4495911598205566, + "learning_rate": 0.0005193810300312543, + "loss": 3.3904, + "step": 56590 + }, + { + "epoch": 3.8452914798206277, + "grad_norm": 0.7534974217414856, + "learning_rate": 0.0005193385650224215, + "loss": 3.3819, + "step": 56595 + }, + { + "epoch": 3.8456311998912893, + "grad_norm": 1.2467442750930786, + "learning_rate": 0.0005192961000135888, + "loss": 3.6057, + "step": 56600 + }, + { + "epoch": 3.8459709199619514, + "grad_norm": 0.8596462607383728, + "learning_rate": 0.000519253635004756, + "loss": 3.3968, + "step": 56605 + }, + { + "epoch": 3.846310640032613, + "grad_norm": 0.7018900513648987, + "learning_rate": 0.0005192111699959234, + "loss": 3.3342, + "step": 56610 + }, + { + "epoch": 3.8466503601032747, + "grad_norm": 1.3635196685791016, + "learning_rate": 0.0005191687049870907, + "loss": 3.7929, + "step": 56615 + }, + { + "epoch": 3.8469900801739367, + "grad_norm": 0.7594895958900452, + "learning_rate": 0.0005191262399782579, + "loss": 3.3432, + "step": 56620 + }, + { + "epoch": 3.8473298002445984, + "grad_norm": 0.751255989074707, + "learning_rate": 0.0005190837749694252, + "loss": 3.3112, + "step": 56625 + }, + { + "epoch": 3.84766952031526, + "grad_norm": 1.0504775047302246, + "learning_rate": 0.0005190413099605925, + "loss": 3.6206, + "step": 56630 + }, + { + "epoch": 3.848009240385922, + "grad_norm": 1.0043764114379883, + "learning_rate": 0.0005189988449517597, + "loss": 3.7035, + "step": 56635 + }, + { + "epoch": 3.8483489604565837, + "grad_norm": 0.7751656174659729, + "learning_rate": 0.0005189563799429269, + "loss": 3.5102, + "step": 56640 + }, + { + "epoch": 3.8486886805272453, + "grad_norm": 0.7378255724906921, + "learning_rate": 0.0005189139149340944, + "loss": 3.5598, + "step": 56645 + }, + { + "epoch": 3.8490284005979074, + "grad_norm": 0.6752565503120422, + "learning_rate": 0.0005188714499252616, + "loss": 3.4617, + "step": 56650 + }, + { + "epoch": 3.849368120668569, + "grad_norm": 0.8378898501396179, + "learning_rate": 0.0005188289849164288, + "loss": 3.4722, + "step": 56655 + }, + { + "epoch": 3.8497078407392307, + "grad_norm": 0.7867488265037537, + "learning_rate": 0.0005187865199075962, + "loss": 3.6478, + "step": 56660 + }, + { + "epoch": 3.8500475608098927, + "grad_norm": 1.0509885549545288, + "learning_rate": 0.0005187440548987634, + "loss": 3.41, + "step": 56665 + }, + { + "epoch": 3.8503872808805544, + "grad_norm": 0.7209457159042358, + "learning_rate": 0.0005187015898899306, + "loss": 3.2945, + "step": 56670 + }, + { + "epoch": 3.850727000951216, + "grad_norm": 0.8037388324737549, + "learning_rate": 0.000518659124881098, + "loss": 3.6782, + "step": 56675 + }, + { + "epoch": 3.851066721021878, + "grad_norm": 1.1898324489593506, + "learning_rate": 0.0005186166598722653, + "loss": 3.456, + "step": 56680 + }, + { + "epoch": 3.8514064410925397, + "grad_norm": 0.9519771933555603, + "learning_rate": 0.0005185741948634325, + "loss": 3.2943, + "step": 56685 + }, + { + "epoch": 3.8517461611632013, + "grad_norm": 0.8820488452911377, + "learning_rate": 0.0005185317298545999, + "loss": 3.516, + "step": 56690 + }, + { + "epoch": 3.8520858812338634, + "grad_norm": 1.0607048273086548, + "learning_rate": 0.0005184892648457671, + "loss": 3.7124, + "step": 56695 + }, + { + "epoch": 3.852425601304525, + "grad_norm": 0.9673507809638977, + "learning_rate": 0.0005184467998369343, + "loss": 3.3364, + "step": 56700 + }, + { + "epoch": 3.8527653213751867, + "grad_norm": 0.9544147253036499, + "learning_rate": 0.0005184043348281016, + "loss": 3.2988, + "step": 56705 + }, + { + "epoch": 3.8531050414458488, + "grad_norm": 1.161620020866394, + "learning_rate": 0.0005183618698192689, + "loss": 3.4615, + "step": 56710 + }, + { + "epoch": 3.8534447615165104, + "grad_norm": 0.9028362035751343, + "learning_rate": 0.0005183194048104362, + "loss": 3.4849, + "step": 56715 + }, + { + "epoch": 3.853784481587172, + "grad_norm": 0.6639185547828674, + "learning_rate": 0.0005182769398016035, + "loss": 3.4402, + "step": 56720 + }, + { + "epoch": 3.854124201657834, + "grad_norm": 0.8959195613861084, + "learning_rate": 0.0005182344747927708, + "loss": 3.5263, + "step": 56725 + }, + { + "epoch": 3.8544639217284957, + "grad_norm": 0.9006872177124023, + "learning_rate": 0.0005181920097839381, + "loss": 3.5691, + "step": 56730 + }, + { + "epoch": 3.8548036417991574, + "grad_norm": 0.7706353664398193, + "learning_rate": 0.0005181495447751053, + "loss": 3.4538, + "step": 56735 + }, + { + "epoch": 3.8551433618698194, + "grad_norm": 1.0322017669677734, + "learning_rate": 0.0005181070797662725, + "loss": 3.6076, + "step": 56740 + }, + { + "epoch": 3.855483081940481, + "grad_norm": 1.0281637907028198, + "learning_rate": 0.00051806461475744, + "loss": 3.3085, + "step": 56745 + }, + { + "epoch": 3.8558228020111427, + "grad_norm": 0.7633823156356812, + "learning_rate": 0.0005180221497486072, + "loss": 3.5971, + "step": 56750 + }, + { + "epoch": 3.8561625220818048, + "grad_norm": 0.7422765493392944, + "learning_rate": 0.0005179796847397744, + "loss": 3.5819, + "step": 56755 + }, + { + "epoch": 3.8565022421524664, + "grad_norm": 0.7503259778022766, + "learning_rate": 0.0005179372197309418, + "loss": 3.5966, + "step": 56760 + }, + { + "epoch": 3.856841962223128, + "grad_norm": 0.7888094186782837, + "learning_rate": 0.000517894754722109, + "loss": 3.3633, + "step": 56765 + }, + { + "epoch": 3.85718168229379, + "grad_norm": 1.1008520126342773, + "learning_rate": 0.0005178522897132762, + "loss": 3.7545, + "step": 56770 + }, + { + "epoch": 3.8575214023644517, + "grad_norm": 0.7092001438140869, + "learning_rate": 0.0005178098247044436, + "loss": 3.5847, + "step": 56775 + }, + { + "epoch": 3.8578611224351134, + "grad_norm": 0.8815507292747498, + "learning_rate": 0.0005177673596956109, + "loss": 3.4544, + "step": 56780 + }, + { + "epoch": 3.8582008425057754, + "grad_norm": 0.7847651839256287, + "learning_rate": 0.0005177248946867781, + "loss": 3.4706, + "step": 56785 + }, + { + "epoch": 3.858540562576437, + "grad_norm": 0.8807682394981384, + "learning_rate": 0.0005176824296779455, + "loss": 3.3279, + "step": 56790 + }, + { + "epoch": 3.8588802826470987, + "grad_norm": 0.6724677681922913, + "learning_rate": 0.0005176399646691127, + "loss": 3.6405, + "step": 56795 + }, + { + "epoch": 3.8592200027177608, + "grad_norm": 0.9270042181015015, + "learning_rate": 0.0005175974996602799, + "loss": 3.3612, + "step": 56800 + }, + { + "epoch": 3.8595597227884224, + "grad_norm": 0.9784954190254211, + "learning_rate": 0.0005175550346514472, + "loss": 3.4072, + "step": 56805 + }, + { + "epoch": 3.859899442859084, + "grad_norm": 0.834963858127594, + "learning_rate": 0.0005175125696426145, + "loss": 3.5618, + "step": 56810 + }, + { + "epoch": 3.860239162929746, + "grad_norm": 2.808793067932129, + "learning_rate": 0.0005174701046337818, + "loss": 3.403, + "step": 56815 + }, + { + "epoch": 3.8605788830004077, + "grad_norm": 0.9575893878936768, + "learning_rate": 0.0005174276396249491, + "loss": 3.6383, + "step": 56820 + }, + { + "epoch": 3.8609186030710694, + "grad_norm": 0.8541870713233948, + "learning_rate": 0.0005173851746161164, + "loss": 3.5109, + "step": 56825 + }, + { + "epoch": 3.8612583231417315, + "grad_norm": 1.168784737586975, + "learning_rate": 0.0005173427096072836, + "loss": 3.4009, + "step": 56830 + }, + { + "epoch": 3.861598043212393, + "grad_norm": 0.8494080901145935, + "learning_rate": 0.0005173002445984509, + "loss": 3.3637, + "step": 56835 + }, + { + "epoch": 3.8619377632830547, + "grad_norm": 0.9269217252731323, + "learning_rate": 0.0005172577795896181, + "loss": 3.3499, + "step": 56840 + }, + { + "epoch": 3.862277483353717, + "grad_norm": 0.9536073207855225, + "learning_rate": 0.0005172153145807854, + "loss": 3.265, + "step": 56845 + }, + { + "epoch": 3.8626172034243784, + "grad_norm": 0.8348523378372192, + "learning_rate": 0.0005171728495719528, + "loss": 3.2388, + "step": 56850 + }, + { + "epoch": 3.86295692349504, + "grad_norm": 0.9283326268196106, + "learning_rate": 0.00051713038456312, + "loss": 3.6327, + "step": 56855 + }, + { + "epoch": 3.863296643565702, + "grad_norm": 0.7020705342292786, + "learning_rate": 0.0005170879195542873, + "loss": 3.2453, + "step": 56860 + }, + { + "epoch": 3.8636363636363638, + "grad_norm": 1.2104182243347168, + "learning_rate": 0.0005170454545454546, + "loss": 3.2154, + "step": 56865 + }, + { + "epoch": 3.8639760837070254, + "grad_norm": 0.8381550908088684, + "learning_rate": 0.0005170029895366218, + "loss": 3.5047, + "step": 56870 + }, + { + "epoch": 3.864315803777687, + "grad_norm": 0.8751277327537537, + "learning_rate": 0.000516960524527789, + "loss": 3.8065, + "step": 56875 + }, + { + "epoch": 3.864655523848349, + "grad_norm": 0.9942311644554138, + "learning_rate": 0.0005169180595189564, + "loss": 3.7295, + "step": 56880 + }, + { + "epoch": 3.8649952439190107, + "grad_norm": 1.1221792697906494, + "learning_rate": 0.0005168755945101237, + "loss": 3.4266, + "step": 56885 + }, + { + "epoch": 3.8653349639896724, + "grad_norm": 0.8594583868980408, + "learning_rate": 0.000516833129501291, + "loss": 3.5325, + "step": 56890 + }, + { + "epoch": 3.8656746840603344, + "grad_norm": 0.8620944619178772, + "learning_rate": 0.0005167906644924583, + "loss": 3.5142, + "step": 56895 + }, + { + "epoch": 3.866014404130996, + "grad_norm": 0.8854902982711792, + "learning_rate": 0.0005167481994836255, + "loss": 3.5443, + "step": 56900 + }, + { + "epoch": 3.8663541242016577, + "grad_norm": 0.8330186605453491, + "learning_rate": 0.0005167057344747927, + "loss": 3.4999, + "step": 56905 + }, + { + "epoch": 3.8666938442723198, + "grad_norm": 0.971399188041687, + "learning_rate": 0.0005166632694659601, + "loss": 3.7473, + "step": 56910 + }, + { + "epoch": 3.8670335643429814, + "grad_norm": 1.0026673078536987, + "learning_rate": 0.0005166208044571273, + "loss": 3.5594, + "step": 56915 + }, + { + "epoch": 3.867373284413643, + "grad_norm": 0.8057916760444641, + "learning_rate": 0.0005165783394482946, + "loss": 3.5488, + "step": 56920 + }, + { + "epoch": 3.8677130044843047, + "grad_norm": 0.8402663469314575, + "learning_rate": 0.000516535874439462, + "loss": 3.3995, + "step": 56925 + }, + { + "epoch": 3.8680527245549667, + "grad_norm": 0.9617754220962524, + "learning_rate": 0.0005164934094306292, + "loss": 3.5485, + "step": 56930 + }, + { + "epoch": 3.8683924446256284, + "grad_norm": 0.9858745336532593, + "learning_rate": 0.0005164509444217964, + "loss": 3.3319, + "step": 56935 + }, + { + "epoch": 3.86873216469629, + "grad_norm": 1.46211576461792, + "learning_rate": 0.0005164084794129637, + "loss": 3.6278, + "step": 56940 + }, + { + "epoch": 3.869071884766952, + "grad_norm": 0.9968897700309753, + "learning_rate": 0.000516366014404131, + "loss": 3.7801, + "step": 56945 + }, + { + "epoch": 3.8694116048376137, + "grad_norm": 0.8682659268379211, + "learning_rate": 0.0005163235493952982, + "loss": 3.4278, + "step": 56950 + }, + { + "epoch": 3.8697513249082753, + "grad_norm": 0.9458279013633728, + "learning_rate": 0.0005162810843864656, + "loss": 3.499, + "step": 56955 + }, + { + "epoch": 3.8700910449789374, + "grad_norm": 1.1017022132873535, + "learning_rate": 0.0005162386193776329, + "loss": 3.6844, + "step": 56960 + }, + { + "epoch": 3.870430765049599, + "grad_norm": 0.8443915843963623, + "learning_rate": 0.0005161961543688001, + "loss": 3.6566, + "step": 56965 + }, + { + "epoch": 3.8707704851202607, + "grad_norm": 0.9440657496452332, + "learning_rate": 0.0005161536893599674, + "loss": 3.5465, + "step": 56970 + }, + { + "epoch": 3.8711102051909227, + "grad_norm": 1.0590320825576782, + "learning_rate": 0.0005161112243511347, + "loss": 3.435, + "step": 56975 + }, + { + "epoch": 3.8714499252615844, + "grad_norm": 0.7551957368850708, + "learning_rate": 0.0005160687593423019, + "loss": 3.5542, + "step": 56980 + }, + { + "epoch": 3.871789645332246, + "grad_norm": 1.1121883392333984, + "learning_rate": 0.0005160262943334692, + "loss": 3.1299, + "step": 56985 + }, + { + "epoch": 3.872129365402908, + "grad_norm": 1.3470536470413208, + "learning_rate": 0.0005159838293246365, + "loss": 3.5815, + "step": 56990 + }, + { + "epoch": 3.8724690854735697, + "grad_norm": 0.8728624582290649, + "learning_rate": 0.0005159413643158038, + "loss": 3.4315, + "step": 56995 + }, + { + "epoch": 3.8728088055442313, + "grad_norm": 0.7523467540740967, + "learning_rate": 0.0005158988993069711, + "loss": 3.4652, + "step": 57000 + }, + { + "epoch": 3.8731485256148934, + "grad_norm": 0.7703983783721924, + "learning_rate": 0.0005158564342981383, + "loss": 3.3774, + "step": 57005 + }, + { + "epoch": 3.873488245685555, + "grad_norm": 1.0352245569229126, + "learning_rate": 0.0005158139692893056, + "loss": 3.4683, + "step": 57010 + }, + { + "epoch": 3.8738279657562167, + "grad_norm": 1.3914529085159302, + "learning_rate": 0.0005157715042804729, + "loss": 3.4494, + "step": 57015 + }, + { + "epoch": 3.8741676858268788, + "grad_norm": 0.9471056461334229, + "learning_rate": 0.0005157290392716401, + "loss": 3.6172, + "step": 57020 + }, + { + "epoch": 3.8745074058975404, + "grad_norm": 0.8569105267524719, + "learning_rate": 0.0005156865742628075, + "loss": 3.4887, + "step": 57025 + }, + { + "epoch": 3.874847125968202, + "grad_norm": 0.9105156064033508, + "learning_rate": 0.0005156441092539748, + "loss": 3.3621, + "step": 57030 + }, + { + "epoch": 3.875186846038864, + "grad_norm": 1.0738773345947266, + "learning_rate": 0.000515601644245142, + "loss": 3.5846, + "step": 57035 + }, + { + "epoch": 3.8755265661095257, + "grad_norm": 0.7471365928649902, + "learning_rate": 0.0005155591792363092, + "loss": 3.365, + "step": 57040 + }, + { + "epoch": 3.8758662861801874, + "grad_norm": 0.8757900595664978, + "learning_rate": 0.0005155167142274766, + "loss": 3.3868, + "step": 57045 + }, + { + "epoch": 3.8762060062508494, + "grad_norm": 1.2710239887237549, + "learning_rate": 0.0005154742492186438, + "loss": 3.465, + "step": 57050 + }, + { + "epoch": 3.876545726321511, + "grad_norm": 1.083652138710022, + "learning_rate": 0.000515431784209811, + "loss": 3.3321, + "step": 57055 + }, + { + "epoch": 3.8768854463921727, + "grad_norm": 0.7637492418289185, + "learning_rate": 0.0005153893192009785, + "loss": 3.5403, + "step": 57060 + }, + { + "epoch": 3.8772251664628348, + "grad_norm": 0.913613498210907, + "learning_rate": 0.0005153468541921457, + "loss": 3.219, + "step": 57065 + }, + { + "epoch": 3.8775648865334964, + "grad_norm": 0.7874297499656677, + "learning_rate": 0.000515304389183313, + "loss": 3.3399, + "step": 57070 + }, + { + "epoch": 3.877904606604158, + "grad_norm": 0.9019875526428223, + "learning_rate": 0.0005152619241744803, + "loss": 3.7198, + "step": 57075 + }, + { + "epoch": 3.87824432667482, + "grad_norm": 0.9436920881271362, + "learning_rate": 0.0005152194591656475, + "loss": 3.4572, + "step": 57080 + }, + { + "epoch": 3.8785840467454817, + "grad_norm": 0.9355970621109009, + "learning_rate": 0.0005151769941568148, + "loss": 3.8913, + "step": 57085 + }, + { + "epoch": 3.8789237668161434, + "grad_norm": 1.0530781745910645, + "learning_rate": 0.000515134529147982, + "loss": 3.7054, + "step": 57090 + }, + { + "epoch": 3.8792634868868054, + "grad_norm": 1.028910517692566, + "learning_rate": 0.0005150920641391494, + "loss": 3.4881, + "step": 57095 + }, + { + "epoch": 3.879603206957467, + "grad_norm": 0.9683576822280884, + "learning_rate": 0.0005150495991303167, + "loss": 3.4541, + "step": 57100 + }, + { + "epoch": 3.8799429270281287, + "grad_norm": 0.8740150332450867, + "learning_rate": 0.0005150071341214839, + "loss": 3.5283, + "step": 57105 + }, + { + "epoch": 3.880282647098791, + "grad_norm": 0.8849396109580994, + "learning_rate": 0.0005149646691126512, + "loss": 3.4975, + "step": 57110 + }, + { + "epoch": 3.8806223671694524, + "grad_norm": 0.7530997395515442, + "learning_rate": 0.0005149222041038185, + "loss": 3.5231, + "step": 57115 + }, + { + "epoch": 3.880962087240114, + "grad_norm": 0.9675373435020447, + "learning_rate": 0.0005148797390949857, + "loss": 3.4312, + "step": 57120 + }, + { + "epoch": 3.881301807310776, + "grad_norm": 0.8525757193565369, + "learning_rate": 0.000514837274086153, + "loss": 3.4738, + "step": 57125 + }, + { + "epoch": 3.8816415273814378, + "grad_norm": 0.8331127762794495, + "learning_rate": 0.0005147948090773204, + "loss": 3.6299, + "step": 57130 + }, + { + "epoch": 3.8819812474520994, + "grad_norm": 0.8812710046768188, + "learning_rate": 0.0005147523440684876, + "loss": 3.3453, + "step": 57135 + }, + { + "epoch": 3.8823209675227615, + "grad_norm": 0.8526292443275452, + "learning_rate": 0.0005147098790596548, + "loss": 3.4998, + "step": 57140 + }, + { + "epoch": 3.882660687593423, + "grad_norm": 0.7922732830047607, + "learning_rate": 0.0005146674140508222, + "loss": 3.3101, + "step": 57145 + }, + { + "epoch": 3.8830004076640847, + "grad_norm": 0.7912677526473999, + "learning_rate": 0.0005146249490419894, + "loss": 3.4137, + "step": 57150 + }, + { + "epoch": 3.883340127734747, + "grad_norm": 2.3813555240631104, + "learning_rate": 0.0005145824840331566, + "loss": 3.3924, + "step": 57155 + }, + { + "epoch": 3.8836798478054084, + "grad_norm": 0.8601946234703064, + "learning_rate": 0.000514540019024324, + "loss": 3.8621, + "step": 57160 + }, + { + "epoch": 3.88401956787607, + "grad_norm": 0.9431272149085999, + "learning_rate": 0.0005144975540154913, + "loss": 3.6281, + "step": 57165 + }, + { + "epoch": 3.884359287946732, + "grad_norm": 0.7753309011459351, + "learning_rate": 0.0005144550890066585, + "loss": 3.5319, + "step": 57170 + }, + { + "epoch": 3.8846990080173938, + "grad_norm": 0.938681960105896, + "learning_rate": 0.0005144126239978259, + "loss": 3.3856, + "step": 57175 + }, + { + "epoch": 3.8850387280880554, + "grad_norm": 0.9258414506912231, + "learning_rate": 0.0005143701589889931, + "loss": 3.2998, + "step": 57180 + }, + { + "epoch": 3.8853784481587175, + "grad_norm": 0.8434944748878479, + "learning_rate": 0.0005143276939801603, + "loss": 3.5003, + "step": 57185 + }, + { + "epoch": 3.885718168229379, + "grad_norm": 0.7840365767478943, + "learning_rate": 0.0005142852289713276, + "loss": 3.5619, + "step": 57190 + }, + { + "epoch": 3.8860578883000407, + "grad_norm": 0.7357649207115173, + "learning_rate": 0.0005142427639624949, + "loss": 3.5696, + "step": 57195 + }, + { + "epoch": 3.886397608370703, + "grad_norm": 0.7472124695777893, + "learning_rate": 0.0005142002989536622, + "loss": 3.2473, + "step": 57200 + }, + { + "epoch": 3.8867373284413644, + "grad_norm": 0.8440210819244385, + "learning_rate": 0.0005141578339448295, + "loss": 3.5664, + "step": 57205 + }, + { + "epoch": 3.887077048512026, + "grad_norm": 0.7241885662078857, + "learning_rate": 0.0005141153689359968, + "loss": 3.6276, + "step": 57210 + }, + { + "epoch": 3.887416768582688, + "grad_norm": 1.3372856378555298, + "learning_rate": 0.000514072903927164, + "loss": 3.6902, + "step": 57215 + }, + { + "epoch": 3.8877564886533498, + "grad_norm": 0.82537841796875, + "learning_rate": 0.0005140304389183313, + "loss": 3.5656, + "step": 57220 + }, + { + "epoch": 3.8880962087240114, + "grad_norm": 0.9292195439338684, + "learning_rate": 0.0005139879739094985, + "loss": 3.4186, + "step": 57225 + }, + { + "epoch": 3.888435928794673, + "grad_norm": 1.121232032775879, + "learning_rate": 0.0005139455089006658, + "loss": 3.3884, + "step": 57230 + }, + { + "epoch": 3.888775648865335, + "grad_norm": 0.8992807865142822, + "learning_rate": 0.0005139030438918332, + "loss": 3.5825, + "step": 57235 + }, + { + "epoch": 3.8891153689359967, + "grad_norm": 0.7218926548957825, + "learning_rate": 0.0005138605788830004, + "loss": 3.361, + "step": 57240 + }, + { + "epoch": 3.8894550890066584, + "grad_norm": 0.8430278897285461, + "learning_rate": 0.0005138181138741677, + "loss": 3.418, + "step": 57245 + }, + { + "epoch": 3.8897948090773204, + "grad_norm": 0.7378146648406982, + "learning_rate": 0.000513775648865335, + "loss": 3.3837, + "step": 57250 + }, + { + "epoch": 3.890134529147982, + "grad_norm": 0.8218130469322205, + "learning_rate": 0.0005137331838565022, + "loss": 3.573, + "step": 57255 + }, + { + "epoch": 3.8904742492186437, + "grad_norm": 0.9248813390731812, + "learning_rate": 0.0005136907188476695, + "loss": 3.5349, + "step": 57260 + }, + { + "epoch": 3.8908139692893053, + "grad_norm": 1.049005389213562, + "learning_rate": 0.0005136482538388368, + "loss": 3.6043, + "step": 57265 + }, + { + "epoch": 3.8911536893599674, + "grad_norm": 0.904442310333252, + "learning_rate": 0.0005136057888300041, + "loss": 3.3792, + "step": 57270 + }, + { + "epoch": 3.891493409430629, + "grad_norm": 1.3326131105422974, + "learning_rate": 0.0005135633238211714, + "loss": 3.5085, + "step": 57275 + }, + { + "epoch": 3.8918331295012907, + "grad_norm": 0.8558046221733093, + "learning_rate": 0.0005135208588123387, + "loss": 3.3705, + "step": 57280 + }, + { + "epoch": 3.8921728495719528, + "grad_norm": 0.8421690464019775, + "learning_rate": 0.0005134783938035059, + "loss": 3.5558, + "step": 57285 + }, + { + "epoch": 3.8925125696426144, + "grad_norm": 0.8939539194107056, + "learning_rate": 0.0005134359287946731, + "loss": 3.6272, + "step": 57290 + }, + { + "epoch": 3.892852289713276, + "grad_norm": 0.8592181205749512, + "learning_rate": 0.0005133934637858405, + "loss": 3.6238, + "step": 57295 + }, + { + "epoch": 3.893192009783938, + "grad_norm": 1.068464994430542, + "learning_rate": 0.0005133509987770077, + "loss": 3.5182, + "step": 57300 + }, + { + "epoch": 3.8935317298545997, + "grad_norm": 0.8556610345840454, + "learning_rate": 0.000513308533768175, + "loss": 3.7974, + "step": 57305 + }, + { + "epoch": 3.8938714499252614, + "grad_norm": 1.110970377922058, + "learning_rate": 0.0005132660687593424, + "loss": 3.5784, + "step": 57310 + }, + { + "epoch": 3.8942111699959234, + "grad_norm": 1.2706059217453003, + "learning_rate": 0.0005132236037505096, + "loss": 3.5925, + "step": 57315 + }, + { + "epoch": 3.894550890066585, + "grad_norm": 1.3484011888504028, + "learning_rate": 0.0005131811387416768, + "loss": 3.6465, + "step": 57320 + }, + { + "epoch": 3.8948906101372467, + "grad_norm": 0.8560666441917419, + "learning_rate": 0.0005131386737328442, + "loss": 3.5376, + "step": 57325 + }, + { + "epoch": 3.8952303302079088, + "grad_norm": 0.7510702013969421, + "learning_rate": 0.0005130962087240114, + "loss": 3.5653, + "step": 57330 + }, + { + "epoch": 3.8955700502785704, + "grad_norm": 0.9162725210189819, + "learning_rate": 0.0005130537437151786, + "loss": 3.5778, + "step": 57335 + }, + { + "epoch": 3.895909770349232, + "grad_norm": 0.8164543509483337, + "learning_rate": 0.000513011278706346, + "loss": 3.3287, + "step": 57340 + }, + { + "epoch": 3.896249490419894, + "grad_norm": 1.0346342325210571, + "learning_rate": 0.0005129688136975133, + "loss": 3.5921, + "step": 57345 + }, + { + "epoch": 3.8965892104905557, + "grad_norm": 0.6747474670410156, + "learning_rate": 0.0005129263486886805, + "loss": 3.8719, + "step": 57350 + }, + { + "epoch": 3.8969289305612174, + "grad_norm": 0.7605888247489929, + "learning_rate": 0.0005128838836798478, + "loss": 3.6179, + "step": 57355 + }, + { + "epoch": 3.8972686506318794, + "grad_norm": 0.8844541907310486, + "learning_rate": 0.0005128414186710151, + "loss": 3.5693, + "step": 57360 + }, + { + "epoch": 3.897608370702541, + "grad_norm": 0.8584783673286438, + "learning_rate": 0.0005127989536621823, + "loss": 3.6423, + "step": 57365 + }, + { + "epoch": 3.8979480907732027, + "grad_norm": 0.8506049513816833, + "learning_rate": 0.0005127564886533497, + "loss": 3.6482, + "step": 57370 + }, + { + "epoch": 3.8982878108438648, + "grad_norm": 1.076507806777954, + "learning_rate": 0.000512714023644517, + "loss": 3.7152, + "step": 57375 + }, + { + "epoch": 3.8986275309145264, + "grad_norm": 1.1912646293640137, + "learning_rate": 0.0005126715586356842, + "loss": 3.7478, + "step": 57380 + }, + { + "epoch": 3.898967250985188, + "grad_norm": 0.7257078886032104, + "learning_rate": 0.0005126290936268515, + "loss": 3.4808, + "step": 57385 + }, + { + "epoch": 3.89930697105585, + "grad_norm": 0.994621217250824, + "learning_rate": 0.0005125866286180187, + "loss": 3.6285, + "step": 57390 + }, + { + "epoch": 3.8996466911265117, + "grad_norm": 1.2263237237930298, + "learning_rate": 0.000512544163609186, + "loss": 3.4403, + "step": 57395 + }, + { + "epoch": 3.8999864111971734, + "grad_norm": 1.171305775642395, + "learning_rate": 0.0005125016986003533, + "loss": 3.3856, + "step": 57400 + }, + { + "epoch": 3.9003261312678354, + "grad_norm": 0.9217913150787354, + "learning_rate": 0.0005124592335915206, + "loss": 3.3311, + "step": 57405 + }, + { + "epoch": 3.900665851338497, + "grad_norm": 0.7820585370063782, + "learning_rate": 0.000512416768582688, + "loss": 3.3996, + "step": 57410 + }, + { + "epoch": 3.9010055714091587, + "grad_norm": 0.8422624468803406, + "learning_rate": 0.0005123743035738552, + "loss": 3.3056, + "step": 57415 + }, + { + "epoch": 3.901345291479821, + "grad_norm": 0.9750136137008667, + "learning_rate": 0.0005123318385650224, + "loss": 3.4811, + "step": 57420 + }, + { + "epoch": 3.9016850115504824, + "grad_norm": 0.9572112560272217, + "learning_rate": 0.0005122893735561898, + "loss": 3.575, + "step": 57425 + }, + { + "epoch": 3.902024731621144, + "grad_norm": 0.9061694741249084, + "learning_rate": 0.000512246908547357, + "loss": 3.5866, + "step": 57430 + }, + { + "epoch": 3.902364451691806, + "grad_norm": 1.2927359342575073, + "learning_rate": 0.0005122044435385242, + "loss": 3.424, + "step": 57435 + }, + { + "epoch": 3.9027041717624678, + "grad_norm": 0.7477893233299255, + "learning_rate": 0.0005121619785296916, + "loss": 3.5954, + "step": 57440 + }, + { + "epoch": 3.9030438918331294, + "grad_norm": 1.0782461166381836, + "learning_rate": 0.0005121195135208589, + "loss": 3.2305, + "step": 57445 + }, + { + "epoch": 3.9033836119037915, + "grad_norm": 0.7344518303871155, + "learning_rate": 0.0005120770485120261, + "loss": 3.3471, + "step": 57450 + }, + { + "epoch": 3.903723331974453, + "grad_norm": 0.8119267225265503, + "learning_rate": 0.0005120345835031934, + "loss": 3.7283, + "step": 57455 + }, + { + "epoch": 3.9040630520451147, + "grad_norm": 0.7091952562332153, + "learning_rate": 0.0005119921184943607, + "loss": 3.5513, + "step": 57460 + }, + { + "epoch": 3.904402772115777, + "grad_norm": 0.7994626760482788, + "learning_rate": 0.0005119496534855279, + "loss": 3.602, + "step": 57465 + }, + { + "epoch": 3.9047424921864384, + "grad_norm": 1.053463339805603, + "learning_rate": 0.0005119071884766952, + "loss": 3.5018, + "step": 57470 + }, + { + "epoch": 3.9050822122571, + "grad_norm": 0.8438632488250732, + "learning_rate": 0.0005118647234678626, + "loss": 3.3809, + "step": 57475 + }, + { + "epoch": 3.905421932327762, + "grad_norm": 1.4400862455368042, + "learning_rate": 0.0005118222584590298, + "loss": 3.5545, + "step": 57480 + }, + { + "epoch": 3.9057616523984238, + "grad_norm": 1.162955641746521, + "learning_rate": 0.0005117797934501971, + "loss": 3.4222, + "step": 57485 + }, + { + "epoch": 3.9061013724690854, + "grad_norm": 1.1513211727142334, + "learning_rate": 0.0005117373284413643, + "loss": 3.0722, + "step": 57490 + }, + { + "epoch": 3.9064410925397475, + "grad_norm": 1.0799967050552368, + "learning_rate": 0.0005116948634325316, + "loss": 3.5244, + "step": 57495 + }, + { + "epoch": 3.906780812610409, + "grad_norm": 0.7801952958106995, + "learning_rate": 0.0005116523984236989, + "loss": 3.5727, + "step": 57500 + }, + { + "epoch": 3.9071205326810707, + "grad_norm": 0.731644868850708, + "learning_rate": 0.0005116099334148661, + "loss": 3.593, + "step": 57505 + }, + { + "epoch": 3.907460252751733, + "grad_norm": 0.7126206159591675, + "learning_rate": 0.0005115674684060335, + "loss": 3.3499, + "step": 57510 + }, + { + "epoch": 3.9077999728223944, + "grad_norm": 0.9671463966369629, + "learning_rate": 0.0005115250033972008, + "loss": 3.5789, + "step": 57515 + }, + { + "epoch": 3.908139692893056, + "grad_norm": 0.8821979761123657, + "learning_rate": 0.000511482538388368, + "loss": 3.5914, + "step": 57520 + }, + { + "epoch": 3.908479412963718, + "grad_norm": 1.115835428237915, + "learning_rate": 0.0005114400733795352, + "loss": 3.278, + "step": 57525 + }, + { + "epoch": 3.90881913303438, + "grad_norm": 1.2433019876480103, + "learning_rate": 0.0005113976083707026, + "loss": 3.4583, + "step": 57530 + }, + { + "epoch": 3.9091588531050414, + "grad_norm": 0.9704692363739014, + "learning_rate": 0.0005113551433618698, + "loss": 3.628, + "step": 57535 + }, + { + "epoch": 3.9094985731757035, + "grad_norm": 0.9200735688209534, + "learning_rate": 0.000511312678353037, + "loss": 3.4703, + "step": 57540 + }, + { + "epoch": 3.909838293246365, + "grad_norm": 0.7929883599281311, + "learning_rate": 0.0005112702133442045, + "loss": 3.5966, + "step": 57545 + }, + { + "epoch": 3.9101780133170267, + "grad_norm": 1.006364107131958, + "learning_rate": 0.0005112277483353717, + "loss": 3.5069, + "step": 57550 + }, + { + "epoch": 3.910517733387689, + "grad_norm": 0.947656512260437, + "learning_rate": 0.0005111852833265389, + "loss": 3.395, + "step": 57555 + }, + { + "epoch": 3.9108574534583505, + "grad_norm": 0.8188271522521973, + "learning_rate": 0.0005111428183177063, + "loss": 3.6131, + "step": 57560 + }, + { + "epoch": 3.911197173529012, + "grad_norm": 0.8153405785560608, + "learning_rate": 0.0005111003533088735, + "loss": 3.5333, + "step": 57565 + }, + { + "epoch": 3.9115368935996737, + "grad_norm": 0.7255359292030334, + "learning_rate": 0.0005110578883000407, + "loss": 3.4606, + "step": 57570 + }, + { + "epoch": 3.911876613670336, + "grad_norm": 0.9502444863319397, + "learning_rate": 0.000511015423291208, + "loss": 3.4896, + "step": 57575 + }, + { + "epoch": 3.9122163337409974, + "grad_norm": 0.7664276361465454, + "learning_rate": 0.0005109729582823754, + "loss": 3.555, + "step": 57580 + }, + { + "epoch": 3.912556053811659, + "grad_norm": 0.8619721531867981, + "learning_rate": 0.0005109304932735426, + "loss": 3.4382, + "step": 57585 + }, + { + "epoch": 3.912895773882321, + "grad_norm": 0.9719002842903137, + "learning_rate": 0.0005108880282647099, + "loss": 3.7332, + "step": 57590 + }, + { + "epoch": 3.9132354939529828, + "grad_norm": 1.02799654006958, + "learning_rate": 0.0005108455632558772, + "loss": 3.3307, + "step": 57595 + }, + { + "epoch": 3.9135752140236444, + "grad_norm": 1.0915124416351318, + "learning_rate": 0.0005108030982470444, + "loss": 3.5541, + "step": 57600 + }, + { + "epoch": 3.913914934094306, + "grad_norm": 0.9150426983833313, + "learning_rate": 0.0005107606332382117, + "loss": 3.6145, + "step": 57605 + }, + { + "epoch": 3.914254654164968, + "grad_norm": 0.8408659100532532, + "learning_rate": 0.000510718168229379, + "loss": 3.473, + "step": 57610 + }, + { + "epoch": 3.9145943742356297, + "grad_norm": 1.0726178884506226, + "learning_rate": 0.0005106757032205463, + "loss": 3.5011, + "step": 57615 + }, + { + "epoch": 3.9149340943062914, + "grad_norm": 0.8507474660873413, + "learning_rate": 0.0005106332382117136, + "loss": 3.5434, + "step": 57620 + }, + { + "epoch": 3.9152738143769534, + "grad_norm": 0.895517885684967, + "learning_rate": 0.0005105907732028808, + "loss": 3.7419, + "step": 57625 + }, + { + "epoch": 3.915613534447615, + "grad_norm": 1.185979962348938, + "learning_rate": 0.0005105483081940481, + "loss": 3.4073, + "step": 57630 + }, + { + "epoch": 3.9159532545182767, + "grad_norm": 0.8075200319290161, + "learning_rate": 0.0005105058431852154, + "loss": 3.6822, + "step": 57635 + }, + { + "epoch": 3.9162929745889388, + "grad_norm": 0.9030041098594666, + "learning_rate": 0.0005104633781763826, + "loss": 3.627, + "step": 57640 + }, + { + "epoch": 3.9166326946596004, + "grad_norm": 0.8038302659988403, + "learning_rate": 0.0005104209131675499, + "loss": 3.7073, + "step": 57645 + }, + { + "epoch": 3.916972414730262, + "grad_norm": 0.7335073947906494, + "learning_rate": 0.0005103784481587173, + "loss": 3.3522, + "step": 57650 + }, + { + "epoch": 3.917312134800924, + "grad_norm": 0.9355706572532654, + "learning_rate": 0.0005103359831498845, + "loss": 3.3473, + "step": 57655 + }, + { + "epoch": 3.9176518548715857, + "grad_norm": 1.360446572303772, + "learning_rate": 0.0005102935181410518, + "loss": 3.2673, + "step": 57660 + }, + { + "epoch": 3.9179915749422474, + "grad_norm": 0.6115610599517822, + "learning_rate": 0.0005102510531322191, + "loss": 3.6815, + "step": 57665 + }, + { + "epoch": 3.9183312950129094, + "grad_norm": 1.0875880718231201, + "learning_rate": 0.0005102085881233863, + "loss": 3.4687, + "step": 57670 + }, + { + "epoch": 3.918671015083571, + "grad_norm": 0.7535082101821899, + "learning_rate": 0.0005101661231145535, + "loss": 3.7056, + "step": 57675 + }, + { + "epoch": 3.9190107351542327, + "grad_norm": 0.7272796630859375, + "learning_rate": 0.0005101236581057209, + "loss": 3.5094, + "step": 57680 + }, + { + "epoch": 3.919350455224895, + "grad_norm": 0.8328291773796082, + "learning_rate": 0.0005100811930968882, + "loss": 3.5007, + "step": 57685 + }, + { + "epoch": 3.9196901752955564, + "grad_norm": 0.7181466221809387, + "learning_rate": 0.0005100387280880554, + "loss": 3.547, + "step": 57690 + }, + { + "epoch": 3.920029895366218, + "grad_norm": 0.9408383369445801, + "learning_rate": 0.0005099962630792228, + "loss": 3.8982, + "step": 57695 + }, + { + "epoch": 3.92036961543688, + "grad_norm": 0.8429552316665649, + "learning_rate": 0.00050995379807039, + "loss": 3.7638, + "step": 57700 + }, + { + "epoch": 3.9207093355075417, + "grad_norm": 0.7025629281997681, + "learning_rate": 0.0005099113330615572, + "loss": 3.6959, + "step": 57705 + }, + { + "epoch": 3.9210490555782034, + "grad_norm": 1.2279465198516846, + "learning_rate": 0.0005098688680527246, + "loss": 3.4879, + "step": 57710 + }, + { + "epoch": 3.9213887756488655, + "grad_norm": 1.05165433883667, + "learning_rate": 0.0005098264030438918, + "loss": 3.3845, + "step": 57715 + }, + { + "epoch": 3.921728495719527, + "grad_norm": 1.0895110368728638, + "learning_rate": 0.0005097839380350591, + "loss": 3.2854, + "step": 57720 + }, + { + "epoch": 3.9220682157901887, + "grad_norm": 0.7939515709877014, + "learning_rate": 0.0005097414730262264, + "loss": 3.4283, + "step": 57725 + }, + { + "epoch": 3.922407935860851, + "grad_norm": 1.3596566915512085, + "learning_rate": 0.0005096990080173937, + "loss": 3.5282, + "step": 57730 + }, + { + "epoch": 3.9227476559315124, + "grad_norm": 0.8200575113296509, + "learning_rate": 0.0005096565430085609, + "loss": 3.5375, + "step": 57735 + }, + { + "epoch": 3.923087376002174, + "grad_norm": 0.9607060551643372, + "learning_rate": 0.0005096140779997282, + "loss": 3.3503, + "step": 57740 + }, + { + "epoch": 3.923427096072836, + "grad_norm": 0.8489375710487366, + "learning_rate": 0.0005095716129908955, + "loss": 3.394, + "step": 57745 + }, + { + "epoch": 3.9237668161434978, + "grad_norm": 0.9171178936958313, + "learning_rate": 0.0005095291479820628, + "loss": 3.6719, + "step": 57750 + }, + { + "epoch": 3.9241065362141594, + "grad_norm": 1.2545490264892578, + "learning_rate": 0.0005094866829732301, + "loss": 3.5052, + "step": 57755 + }, + { + "epoch": 3.9244462562848215, + "grad_norm": 0.8707087635993958, + "learning_rate": 0.0005094442179643974, + "loss": 3.5498, + "step": 57760 + }, + { + "epoch": 3.924785976355483, + "grad_norm": 0.6219019293785095, + "learning_rate": 0.0005094017529555647, + "loss": 3.539, + "step": 57765 + }, + { + "epoch": 3.9251256964261447, + "grad_norm": 0.8875517845153809, + "learning_rate": 0.0005093592879467319, + "loss": 3.3529, + "step": 57770 + }, + { + "epoch": 3.925465416496807, + "grad_norm": 0.7593134045600891, + "learning_rate": 0.0005093168229378991, + "loss": 3.5859, + "step": 57775 + }, + { + "epoch": 3.9258051365674684, + "grad_norm": 0.8748698830604553, + "learning_rate": 0.0005092743579290665, + "loss": 3.3756, + "step": 57780 + }, + { + "epoch": 3.92614485663813, + "grad_norm": 1.0114917755126953, + "learning_rate": 0.0005092318929202337, + "loss": 3.5962, + "step": 57785 + }, + { + "epoch": 3.926484576708792, + "grad_norm": 0.7679345607757568, + "learning_rate": 0.000509189427911401, + "loss": 3.5556, + "step": 57790 + }, + { + "epoch": 3.9268242967794538, + "grad_norm": 0.7611069679260254, + "learning_rate": 0.0005091469629025684, + "loss": 3.2607, + "step": 57795 + }, + { + "epoch": 3.9271640168501154, + "grad_norm": 0.9605770707130432, + "learning_rate": 0.0005091044978937356, + "loss": 3.6002, + "step": 57800 + }, + { + "epoch": 3.9275037369207775, + "grad_norm": 0.9018978476524353, + "learning_rate": 0.0005090620328849028, + "loss": 3.6232, + "step": 57805 + }, + { + "epoch": 3.927843456991439, + "grad_norm": 1.0532951354980469, + "learning_rate": 0.0005090195678760702, + "loss": 3.4671, + "step": 57810 + }, + { + "epoch": 3.9281831770621007, + "grad_norm": 1.0418877601623535, + "learning_rate": 0.0005089771028672374, + "loss": 3.4496, + "step": 57815 + }, + { + "epoch": 3.928522897132763, + "grad_norm": 1.046834111213684, + "learning_rate": 0.0005089346378584046, + "loss": 3.3334, + "step": 57820 + }, + { + "epoch": 3.9288626172034244, + "grad_norm": 1.0155375003814697, + "learning_rate": 0.000508892172849572, + "loss": 3.6435, + "step": 57825 + }, + { + "epoch": 3.929202337274086, + "grad_norm": 0.7824139595031738, + "learning_rate": 0.0005088497078407393, + "loss": 3.5666, + "step": 57830 + }, + { + "epoch": 3.929542057344748, + "grad_norm": 0.8723594546318054, + "learning_rate": 0.0005088072428319065, + "loss": 3.2868, + "step": 57835 + }, + { + "epoch": 3.92988177741541, + "grad_norm": 0.828716516494751, + "learning_rate": 0.0005087647778230738, + "loss": 3.2986, + "step": 57840 + }, + { + "epoch": 3.9302214974860714, + "grad_norm": 0.954928994178772, + "learning_rate": 0.0005087223128142411, + "loss": 3.5946, + "step": 57845 + }, + { + "epoch": 3.9305612175567335, + "grad_norm": 0.779865026473999, + "learning_rate": 0.0005086798478054083, + "loss": 3.4754, + "step": 57850 + }, + { + "epoch": 3.930900937627395, + "grad_norm": 0.7327186465263367, + "learning_rate": 0.0005086373827965756, + "loss": 3.5298, + "step": 57855 + }, + { + "epoch": 3.9312406576980568, + "grad_norm": 0.918663501739502, + "learning_rate": 0.000508594917787743, + "loss": 3.6597, + "step": 57860 + }, + { + "epoch": 3.931580377768719, + "grad_norm": 0.9499974846839905, + "learning_rate": 0.0005085524527789102, + "loss": 3.4245, + "step": 57865 + }, + { + "epoch": 3.9319200978393805, + "grad_norm": 0.9082774519920349, + "learning_rate": 0.0005085099877700775, + "loss": 3.3617, + "step": 57870 + }, + { + "epoch": 3.932259817910042, + "grad_norm": 0.8721675872802734, + "learning_rate": 0.0005084675227612447, + "loss": 3.6073, + "step": 57875 + }, + { + "epoch": 3.932599537980704, + "grad_norm": 0.7511876821517944, + "learning_rate": 0.000508425057752412, + "loss": 3.4622, + "step": 57880 + }, + { + "epoch": 3.932939258051366, + "grad_norm": 0.9155673384666443, + "learning_rate": 0.0005083825927435793, + "loss": 3.6787, + "step": 57885 + }, + { + "epoch": 3.9332789781220274, + "grad_norm": 0.6869795918464661, + "learning_rate": 0.0005083401277347465, + "loss": 3.4943, + "step": 57890 + }, + { + "epoch": 3.9336186981926895, + "grad_norm": 0.7228953838348389, + "learning_rate": 0.0005082976627259139, + "loss": 3.4925, + "step": 57895 + }, + { + "epoch": 3.933958418263351, + "grad_norm": 0.7208672761917114, + "learning_rate": 0.0005082551977170812, + "loss": 3.3846, + "step": 57900 + }, + { + "epoch": 3.9342981383340128, + "grad_norm": 1.0521990060806274, + "learning_rate": 0.0005082127327082484, + "loss": 3.6357, + "step": 57905 + }, + { + "epoch": 3.9346378584046744, + "grad_norm": 0.849660336971283, + "learning_rate": 0.0005081702676994156, + "loss": 3.6777, + "step": 57910 + }, + { + "epoch": 3.9349775784753365, + "grad_norm": 0.778304934501648, + "learning_rate": 0.000508127802690583, + "loss": 3.6254, + "step": 57915 + }, + { + "epoch": 3.935317298545998, + "grad_norm": 1.0945446491241455, + "learning_rate": 0.0005080853376817502, + "loss": 3.2789, + "step": 57920 + }, + { + "epoch": 3.9356570186166597, + "grad_norm": 1.1343942880630493, + "learning_rate": 0.0005080428726729174, + "loss": 3.6489, + "step": 57925 + }, + { + "epoch": 3.935996738687322, + "grad_norm": 0.9594082832336426, + "learning_rate": 0.0005080004076640849, + "loss": 3.3318, + "step": 57930 + }, + { + "epoch": 3.9363364587579834, + "grad_norm": 0.8449326753616333, + "learning_rate": 0.0005079579426552521, + "loss": 3.5973, + "step": 57935 + }, + { + "epoch": 3.936676178828645, + "grad_norm": 1.0172070264816284, + "learning_rate": 0.0005079154776464193, + "loss": 3.4541, + "step": 57940 + }, + { + "epoch": 3.9370158988993067, + "grad_norm": 0.766942024230957, + "learning_rate": 0.0005078730126375867, + "loss": 3.1327, + "step": 57945 + }, + { + "epoch": 3.9373556189699688, + "grad_norm": 0.9737443327903748, + "learning_rate": 0.0005078305476287539, + "loss": 3.4705, + "step": 57950 + }, + { + "epoch": 3.9376953390406304, + "grad_norm": 0.8648078441619873, + "learning_rate": 0.0005077880826199211, + "loss": 3.5477, + "step": 57955 + }, + { + "epoch": 3.938035059111292, + "grad_norm": 0.9023715853691101, + "learning_rate": 0.0005077456176110886, + "loss": 3.2744, + "step": 57960 + }, + { + "epoch": 3.938374779181954, + "grad_norm": 0.9019187688827515, + "learning_rate": 0.0005077031526022558, + "loss": 3.6515, + "step": 57965 + }, + { + "epoch": 3.9387144992526157, + "grad_norm": 1.0867276191711426, + "learning_rate": 0.000507660687593423, + "loss": 3.7183, + "step": 57970 + }, + { + "epoch": 3.9390542193232774, + "grad_norm": 0.812721848487854, + "learning_rate": 0.0005076182225845903, + "loss": 3.489, + "step": 57975 + }, + { + "epoch": 3.9393939393939394, + "grad_norm": 0.8396128416061401, + "learning_rate": 0.0005075757575757576, + "loss": 3.2889, + "step": 57980 + }, + { + "epoch": 3.939733659464601, + "grad_norm": 0.9133723378181458, + "learning_rate": 0.0005075332925669248, + "loss": 3.5939, + "step": 57985 + }, + { + "epoch": 3.9400733795352627, + "grad_norm": 0.8736927509307861, + "learning_rate": 0.0005074908275580921, + "loss": 3.4874, + "step": 57990 + }, + { + "epoch": 3.940413099605925, + "grad_norm": 0.8803768754005432, + "learning_rate": 0.0005074483625492595, + "loss": 3.3504, + "step": 57995 + }, + { + "epoch": 3.9407528196765864, + "grad_norm": 0.9039371609687805, + "learning_rate": 0.0005074058975404267, + "loss": 3.3516, + "step": 58000 + }, + { + "epoch": 3.941092539747248, + "grad_norm": 1.075278878211975, + "learning_rate": 0.000507363432531594, + "loss": 3.2157, + "step": 58005 + }, + { + "epoch": 3.94143225981791, + "grad_norm": 0.9838477373123169, + "learning_rate": 0.0005073209675227613, + "loss": 3.5814, + "step": 58010 + }, + { + "epoch": 3.9417719798885718, + "grad_norm": 1.0882760286331177, + "learning_rate": 0.0005072785025139285, + "loss": 3.4352, + "step": 58015 + }, + { + "epoch": 3.9421116999592334, + "grad_norm": 0.798846423625946, + "learning_rate": 0.0005072360375050958, + "loss": 3.5862, + "step": 58020 + }, + { + "epoch": 3.9424514200298955, + "grad_norm": 0.9038563370704651, + "learning_rate": 0.000507193572496263, + "loss": 3.4691, + "step": 58025 + }, + { + "epoch": 3.942791140100557, + "grad_norm": 1.0672730207443237, + "learning_rate": 0.0005071511074874304, + "loss": 3.4586, + "step": 58030 + }, + { + "epoch": 3.9431308601712187, + "grad_norm": 0.7741675972938538, + "learning_rate": 0.0005071086424785977, + "loss": 3.427, + "step": 58035 + }, + { + "epoch": 3.943470580241881, + "grad_norm": 0.8939539790153503, + "learning_rate": 0.0005070661774697649, + "loss": 3.5251, + "step": 58040 + }, + { + "epoch": 3.9438103003125424, + "grad_norm": 0.8015594482421875, + "learning_rate": 0.0005070237124609322, + "loss": 3.5034, + "step": 58045 + }, + { + "epoch": 3.944150020383204, + "grad_norm": 0.8641104102134705, + "learning_rate": 0.0005069812474520995, + "loss": 3.4, + "step": 58050 + }, + { + "epoch": 3.944489740453866, + "grad_norm": 0.7646827101707458, + "learning_rate": 0.0005069387824432667, + "loss": 3.6448, + "step": 58055 + }, + { + "epoch": 3.9448294605245278, + "grad_norm": 1.0285121202468872, + "learning_rate": 0.0005068963174344339, + "loss": 3.5994, + "step": 58060 + }, + { + "epoch": 3.9451691805951894, + "grad_norm": 0.9637117385864258, + "learning_rate": 0.0005068538524256014, + "loss": 3.4345, + "step": 58065 + }, + { + "epoch": 3.9455089006658515, + "grad_norm": 0.8554577827453613, + "learning_rate": 0.0005068113874167686, + "loss": 3.5788, + "step": 58070 + }, + { + "epoch": 3.945848620736513, + "grad_norm": 0.8484289646148682, + "learning_rate": 0.0005067689224079358, + "loss": 3.6474, + "step": 58075 + }, + { + "epoch": 3.9461883408071747, + "grad_norm": 0.7393940091133118, + "learning_rate": 0.0005067264573991032, + "loss": 3.5786, + "step": 58080 + }, + { + "epoch": 3.946528060877837, + "grad_norm": 0.8794519901275635, + "learning_rate": 0.0005066839923902704, + "loss": 3.5425, + "step": 58085 + }, + { + "epoch": 3.9468677809484984, + "grad_norm": 0.8264247179031372, + "learning_rate": 0.0005066415273814377, + "loss": 3.4844, + "step": 58090 + }, + { + "epoch": 3.94720750101916, + "grad_norm": 0.830420732498169, + "learning_rate": 0.000506599062372605, + "loss": 3.4909, + "step": 58095 + }, + { + "epoch": 3.947547221089822, + "grad_norm": 1.3286415338516235, + "learning_rate": 0.0005065565973637723, + "loss": 3.7214, + "step": 58100 + }, + { + "epoch": 3.9478869411604838, + "grad_norm": 1.4399476051330566, + "learning_rate": 0.0005065141323549396, + "loss": 3.7089, + "step": 58105 + }, + { + "epoch": 3.9482266612311454, + "grad_norm": 1.0080400705337524, + "learning_rate": 0.0005064716673461069, + "loss": 3.2644, + "step": 58110 + }, + { + "epoch": 3.9485663813018075, + "grad_norm": 0.9388505220413208, + "learning_rate": 0.0005064292023372741, + "loss": 3.5571, + "step": 58115 + }, + { + "epoch": 3.948906101372469, + "grad_norm": 0.9111339449882507, + "learning_rate": 0.0005063867373284414, + "loss": 3.2708, + "step": 58120 + }, + { + "epoch": 3.9492458214431307, + "grad_norm": 0.8298900723457336, + "learning_rate": 0.0005063442723196086, + "loss": 3.339, + "step": 58125 + }, + { + "epoch": 3.949585541513793, + "grad_norm": 0.9932072758674622, + "learning_rate": 0.0005063018073107759, + "loss": 3.8228, + "step": 58130 + }, + { + "epoch": 3.9499252615844545, + "grad_norm": 0.8818564414978027, + "learning_rate": 0.0005062593423019433, + "loss": 3.5516, + "step": 58135 + }, + { + "epoch": 3.950264981655116, + "grad_norm": 1.029739499092102, + "learning_rate": 0.0005062168772931105, + "loss": 3.5174, + "step": 58140 + }, + { + "epoch": 3.950604701725778, + "grad_norm": 1.0952411890029907, + "learning_rate": 0.0005061744122842778, + "loss": 3.634, + "step": 58145 + }, + { + "epoch": 3.95094442179644, + "grad_norm": 0.8502510786056519, + "learning_rate": 0.0005061319472754451, + "loss": 3.5604, + "step": 58150 + }, + { + "epoch": 3.9512841418671014, + "grad_norm": 0.7648395299911499, + "learning_rate": 0.0005060894822666123, + "loss": 3.5383, + "step": 58155 + }, + { + "epoch": 3.9516238619377635, + "grad_norm": 0.828384518623352, + "learning_rate": 0.0005060470172577795, + "loss": 3.7654, + "step": 58160 + }, + { + "epoch": 3.951963582008425, + "grad_norm": 1.8672354221343994, + "learning_rate": 0.0005060045522489469, + "loss": 3.8266, + "step": 58165 + }, + { + "epoch": 3.9523033020790868, + "grad_norm": 0.9821152091026306, + "learning_rate": 0.0005059620872401142, + "loss": 3.4282, + "step": 58170 + }, + { + "epoch": 3.952643022149749, + "grad_norm": 0.8536328077316284, + "learning_rate": 0.0005059196222312814, + "loss": 3.4142, + "step": 58175 + }, + { + "epoch": 3.9529827422204105, + "grad_norm": 0.7489006519317627, + "learning_rate": 0.0005058771572224488, + "loss": 3.3939, + "step": 58180 + }, + { + "epoch": 3.953322462291072, + "grad_norm": 0.903169572353363, + "learning_rate": 0.000505834692213616, + "loss": 3.501, + "step": 58185 + }, + { + "epoch": 3.953662182361734, + "grad_norm": 0.7916556596755981, + "learning_rate": 0.0005057922272047832, + "loss": 3.4626, + "step": 58190 + }, + { + "epoch": 3.954001902432396, + "grad_norm": 0.8764162659645081, + "learning_rate": 0.0005057497621959506, + "loss": 3.4875, + "step": 58195 + }, + { + "epoch": 3.9543416225030574, + "grad_norm": 0.8005889654159546, + "learning_rate": 0.0005057072971871178, + "loss": 3.4745, + "step": 58200 + }, + { + "epoch": 3.9546813425737195, + "grad_norm": 1.0455639362335205, + "learning_rate": 0.0005056648321782851, + "loss": 3.6033, + "step": 58205 + }, + { + "epoch": 3.955021062644381, + "grad_norm": 0.8559653759002686, + "learning_rate": 0.0005056223671694525, + "loss": 3.2944, + "step": 58210 + }, + { + "epoch": 3.9553607827150428, + "grad_norm": 0.9199241995811462, + "learning_rate": 0.0005055799021606197, + "loss": 3.261, + "step": 58215 + }, + { + "epoch": 3.955700502785705, + "grad_norm": 0.9811674356460571, + "learning_rate": 0.0005055374371517869, + "loss": 3.7129, + "step": 58220 + }, + { + "epoch": 3.9560402228563665, + "grad_norm": 0.8099051117897034, + "learning_rate": 0.0005054949721429542, + "loss": 3.3218, + "step": 58225 + }, + { + "epoch": 3.956379942927028, + "grad_norm": 0.8604681491851807, + "learning_rate": 0.0005054525071341215, + "loss": 3.3551, + "step": 58230 + }, + { + "epoch": 3.95671966299769, + "grad_norm": 0.7500579357147217, + "learning_rate": 0.0005054100421252887, + "loss": 3.7321, + "step": 58235 + }, + { + "epoch": 3.957059383068352, + "grad_norm": 0.837807834148407, + "learning_rate": 0.0005053675771164561, + "loss": 3.6117, + "step": 58240 + }, + { + "epoch": 3.9573991031390134, + "grad_norm": 0.9512894749641418, + "learning_rate": 0.0005053251121076234, + "loss": 3.6947, + "step": 58245 + }, + { + "epoch": 3.957738823209675, + "grad_norm": 0.8467386364936829, + "learning_rate": 0.0005052826470987906, + "loss": 3.5233, + "step": 58250 + }, + { + "epoch": 3.958078543280337, + "grad_norm": 0.9090461730957031, + "learning_rate": 0.0005052401820899579, + "loss": 3.3133, + "step": 58255 + }, + { + "epoch": 3.958418263350999, + "grad_norm": 0.9531348347663879, + "learning_rate": 0.0005051977170811251, + "loss": 3.3717, + "step": 58260 + }, + { + "epoch": 3.9587579834216604, + "grad_norm": 0.9099034667015076, + "learning_rate": 0.0005051552520722924, + "loss": 3.68, + "step": 58265 + }, + { + "epoch": 3.9590977034923225, + "grad_norm": 0.9285187721252441, + "learning_rate": 0.0005051127870634597, + "loss": 3.5716, + "step": 58270 + }, + { + "epoch": 3.959437423562984, + "grad_norm": 0.934465229511261, + "learning_rate": 0.000505070322054627, + "loss": 3.2589, + "step": 58275 + }, + { + "epoch": 3.9597771436336457, + "grad_norm": 0.7939115762710571, + "learning_rate": 0.0005050278570457943, + "loss": 3.5726, + "step": 58280 + }, + { + "epoch": 3.9601168637043074, + "grad_norm": 0.9468801617622375, + "learning_rate": 0.0005049853920369616, + "loss": 3.5406, + "step": 58285 + }, + { + "epoch": 3.9604565837749695, + "grad_norm": 0.7870660424232483, + "learning_rate": 0.0005049429270281288, + "loss": 3.7892, + "step": 58290 + }, + { + "epoch": 3.960796303845631, + "grad_norm": 0.8129415512084961, + "learning_rate": 0.000504900462019296, + "loss": 3.5256, + "step": 58295 + }, + { + "epoch": 3.9611360239162927, + "grad_norm": 2.075011968612671, + "learning_rate": 0.0005048579970104634, + "loss": 3.4665, + "step": 58300 + }, + { + "epoch": 3.961475743986955, + "grad_norm": 0.9316096305847168, + "learning_rate": 0.0005048155320016306, + "loss": 3.542, + "step": 58305 + }, + { + "epoch": 3.9618154640576164, + "grad_norm": 0.8207512497901917, + "learning_rate": 0.000504773066992798, + "loss": 3.3907, + "step": 58310 + }, + { + "epoch": 3.962155184128278, + "grad_norm": 0.8904290795326233, + "learning_rate": 0.0005047306019839653, + "loss": 3.7004, + "step": 58315 + }, + { + "epoch": 3.96249490419894, + "grad_norm": 0.8550782203674316, + "learning_rate": 0.0005046881369751325, + "loss": 3.5086, + "step": 58320 + }, + { + "epoch": 3.9628346242696018, + "grad_norm": 0.7565914392471313, + "learning_rate": 0.0005046456719662997, + "loss": 3.4796, + "step": 58325 + }, + { + "epoch": 3.9631743443402634, + "grad_norm": 1.092847466468811, + "learning_rate": 0.0005046032069574671, + "loss": 3.3838, + "step": 58330 + }, + { + "epoch": 3.9635140644109255, + "grad_norm": 0.8022302389144897, + "learning_rate": 0.0005045607419486343, + "loss": 3.5696, + "step": 58335 + }, + { + "epoch": 3.963853784481587, + "grad_norm": 0.9650247693061829, + "learning_rate": 0.0005045182769398015, + "loss": 3.5475, + "step": 58340 + }, + { + "epoch": 3.9641935045522487, + "grad_norm": 0.9735673069953918, + "learning_rate": 0.000504475811930969, + "loss": 3.1982, + "step": 58345 + }, + { + "epoch": 3.964533224622911, + "grad_norm": 0.8439900875091553, + "learning_rate": 0.0005044333469221362, + "loss": 3.8076, + "step": 58350 + }, + { + "epoch": 3.9648729446935724, + "grad_norm": 0.9903268218040466, + "learning_rate": 0.0005043908819133034, + "loss": 3.2138, + "step": 58355 + }, + { + "epoch": 3.965212664764234, + "grad_norm": 0.8935691118240356, + "learning_rate": 0.0005043484169044707, + "loss": 3.402, + "step": 58360 + }, + { + "epoch": 3.965552384834896, + "grad_norm": 0.8377406597137451, + "learning_rate": 0.000504305951895638, + "loss": 3.6499, + "step": 58365 + }, + { + "epoch": 3.9658921049055578, + "grad_norm": 0.9769845604896545, + "learning_rate": 0.0005042634868868052, + "loss": 3.4396, + "step": 58370 + }, + { + "epoch": 3.9662318249762194, + "grad_norm": 0.8500931262969971, + "learning_rate": 0.0005042210218779725, + "loss": 3.4506, + "step": 58375 + }, + { + "epoch": 3.9665715450468815, + "grad_norm": 0.9201495051383972, + "learning_rate": 0.0005041785568691399, + "loss": 3.571, + "step": 58380 + }, + { + "epoch": 3.966911265117543, + "grad_norm": 0.8753264546394348, + "learning_rate": 0.0005041360918603071, + "loss": 3.6226, + "step": 58385 + }, + { + "epoch": 3.9672509851882047, + "grad_norm": 1.1276609897613525, + "learning_rate": 0.0005040936268514744, + "loss": 3.6304, + "step": 58390 + }, + { + "epoch": 3.967590705258867, + "grad_norm": 0.7851729989051819, + "learning_rate": 0.0005040511618426417, + "loss": 3.6937, + "step": 58395 + }, + { + "epoch": 3.9679304253295284, + "grad_norm": 0.803649365901947, + "learning_rate": 0.0005040086968338089, + "loss": 3.4422, + "step": 58400 + }, + { + "epoch": 3.96827014540019, + "grad_norm": 0.9867132306098938, + "learning_rate": 0.0005039662318249762, + "loss": 3.454, + "step": 58405 + }, + { + "epoch": 3.968609865470852, + "grad_norm": 0.8968766927719116, + "learning_rate": 0.0005039237668161434, + "loss": 3.4482, + "step": 58410 + }, + { + "epoch": 3.968949585541514, + "grad_norm": 0.9215635061264038, + "learning_rate": 0.0005038813018073108, + "loss": 3.265, + "step": 58415 + }, + { + "epoch": 3.9692893056121754, + "grad_norm": 0.9452293515205383, + "learning_rate": 0.0005038388367984781, + "loss": 3.4326, + "step": 58420 + }, + { + "epoch": 3.9696290256828375, + "grad_norm": 0.9856594204902649, + "learning_rate": 0.0005037963717896453, + "loss": 3.7966, + "step": 58425 + }, + { + "epoch": 3.969968745753499, + "grad_norm": 1.2272056341171265, + "learning_rate": 0.0005037539067808127, + "loss": 3.3756, + "step": 58430 + }, + { + "epoch": 3.9703084658241607, + "grad_norm": 0.759789764881134, + "learning_rate": 0.0005037114417719799, + "loss": 3.3371, + "step": 58435 + }, + { + "epoch": 3.970648185894823, + "grad_norm": 0.8865832090377808, + "learning_rate": 0.0005036689767631471, + "loss": 3.4559, + "step": 58440 + }, + { + "epoch": 3.9709879059654845, + "grad_norm": 1.0445077419281006, + "learning_rate": 0.0005036265117543146, + "loss": 3.4061, + "step": 58445 + }, + { + "epoch": 3.971327626036146, + "grad_norm": 1.873594045639038, + "learning_rate": 0.0005035840467454818, + "loss": 3.1775, + "step": 58450 + }, + { + "epoch": 3.971667346106808, + "grad_norm": 1.0005426406860352, + "learning_rate": 0.000503541581736649, + "loss": 3.5394, + "step": 58455 + }, + { + "epoch": 3.97200706617747, + "grad_norm": 2.5633442401885986, + "learning_rate": 0.0005034991167278163, + "loss": 3.4697, + "step": 58460 + }, + { + "epoch": 3.9723467862481314, + "grad_norm": 0.9882625341415405, + "learning_rate": 0.0005034566517189836, + "loss": 3.4715, + "step": 58465 + }, + { + "epoch": 3.9726865063187935, + "grad_norm": 0.9934515357017517, + "learning_rate": 0.0005034141867101508, + "loss": 3.2818, + "step": 58470 + }, + { + "epoch": 3.973026226389455, + "grad_norm": 0.9851252436637878, + "learning_rate": 0.0005033717217013181, + "loss": 3.4467, + "step": 58475 + }, + { + "epoch": 3.9733659464601168, + "grad_norm": 0.8459964990615845, + "learning_rate": 0.0005033292566924855, + "loss": 3.5555, + "step": 58480 + }, + { + "epoch": 3.973705666530779, + "grad_norm": 0.8251769542694092, + "learning_rate": 0.0005032867916836527, + "loss": 3.1989, + "step": 58485 + }, + { + "epoch": 3.9740453866014405, + "grad_norm": 0.8321053385734558, + "learning_rate": 0.00050324432667482, + "loss": 3.6527, + "step": 58490 + }, + { + "epoch": 3.974385106672102, + "grad_norm": 0.8952816724777222, + "learning_rate": 0.0005032018616659873, + "loss": 3.3618, + "step": 58495 + }, + { + "epoch": 3.974724826742764, + "grad_norm": 0.972129225730896, + "learning_rate": 0.0005031593966571545, + "loss": 3.4188, + "step": 58500 + }, + { + "epoch": 3.975064546813426, + "grad_norm": 1.0175836086273193, + "learning_rate": 0.0005031169316483218, + "loss": 3.2086, + "step": 58505 + }, + { + "epoch": 3.9754042668840874, + "grad_norm": 0.7839510440826416, + "learning_rate": 0.000503074466639489, + "loss": 3.5327, + "step": 58510 + }, + { + "epoch": 3.9757439869547495, + "grad_norm": 0.8590485453605652, + "learning_rate": 0.0005030320016306564, + "loss": 3.6787, + "step": 58515 + }, + { + "epoch": 3.976083707025411, + "grad_norm": 1.0174731016159058, + "learning_rate": 0.0005029895366218237, + "loss": 3.5586, + "step": 58520 + }, + { + "epoch": 3.9764234270960728, + "grad_norm": 0.8730679154396057, + "learning_rate": 0.0005029470716129909, + "loss": 3.6173, + "step": 58525 + }, + { + "epoch": 3.976763147166735, + "grad_norm": 1.154018759727478, + "learning_rate": 0.0005029046066041582, + "loss": 3.3978, + "step": 58530 + }, + { + "epoch": 3.9771028672373965, + "grad_norm": 1.2623858451843262, + "learning_rate": 0.0005028621415953255, + "loss": 3.5161, + "step": 58535 + }, + { + "epoch": 3.977442587308058, + "grad_norm": 0.8604353070259094, + "learning_rate": 0.0005028196765864927, + "loss": 3.5032, + "step": 58540 + }, + { + "epoch": 3.97778230737872, + "grad_norm": 1.0804920196533203, + "learning_rate": 0.00050277721157766, + "loss": 3.4831, + "step": 58545 + }, + { + "epoch": 3.978122027449382, + "grad_norm": 1.0577651262283325, + "learning_rate": 0.0005027347465688274, + "loss": 3.6307, + "step": 58550 + }, + { + "epoch": 3.9784617475200434, + "grad_norm": 0.9079492688179016, + "learning_rate": 0.0005026922815599946, + "loss": 3.7143, + "step": 58555 + }, + { + "epoch": 3.9788014675907055, + "grad_norm": 0.8260047435760498, + "learning_rate": 0.0005026498165511618, + "loss": 3.682, + "step": 58560 + }, + { + "epoch": 3.979141187661367, + "grad_norm": 0.7071512937545776, + "learning_rate": 0.0005026073515423292, + "loss": 3.5207, + "step": 58565 + }, + { + "epoch": 3.979480907732029, + "grad_norm": 0.9244656562805176, + "learning_rate": 0.0005025648865334964, + "loss": 3.542, + "step": 58570 + }, + { + "epoch": 3.979820627802691, + "grad_norm": 1.1806243658065796, + "learning_rate": 0.0005025224215246636, + "loss": 3.3665, + "step": 58575 + }, + { + "epoch": 3.9801603478733525, + "grad_norm": 0.7310066223144531, + "learning_rate": 0.000502479956515831, + "loss": 3.5364, + "step": 58580 + }, + { + "epoch": 3.980500067944014, + "grad_norm": 0.8309284448623657, + "learning_rate": 0.0005024374915069983, + "loss": 3.5225, + "step": 58585 + }, + { + "epoch": 3.9808397880146758, + "grad_norm": 0.9056680202484131, + "learning_rate": 0.0005023950264981655, + "loss": 3.468, + "step": 58590 + }, + { + "epoch": 3.981179508085338, + "grad_norm": 0.9902760982513428, + "learning_rate": 0.0005023525614893329, + "loss": 3.546, + "step": 58595 + }, + { + "epoch": 3.9815192281559995, + "grad_norm": 0.6969340443611145, + "learning_rate": 0.0005023100964805001, + "loss": 3.578, + "step": 58600 + }, + { + "epoch": 3.981858948226661, + "grad_norm": 0.660993754863739, + "learning_rate": 0.0005022676314716673, + "loss": 3.528, + "step": 58605 + }, + { + "epoch": 3.982198668297323, + "grad_norm": 0.872743546962738, + "learning_rate": 0.0005022251664628346, + "loss": 3.5203, + "step": 58610 + }, + { + "epoch": 3.982538388367985, + "grad_norm": 0.8260291814804077, + "learning_rate": 0.0005021827014540019, + "loss": 3.4115, + "step": 58615 + }, + { + "epoch": 3.9828781084386464, + "grad_norm": 0.7985127568244934, + "learning_rate": 0.0005021402364451692, + "loss": 3.6226, + "step": 58620 + }, + { + "epoch": 3.983217828509308, + "grad_norm": 0.8280341029167175, + "learning_rate": 0.0005020977714363365, + "loss": 3.4645, + "step": 58625 + }, + { + "epoch": 3.98355754857997, + "grad_norm": 0.7678314447402954, + "learning_rate": 0.0005020553064275038, + "loss": 3.5384, + "step": 58630 + }, + { + "epoch": 3.9838972686506318, + "grad_norm": 0.8187561631202698, + "learning_rate": 0.000502012841418671, + "loss": 3.2806, + "step": 58635 + }, + { + "epoch": 3.9842369887212934, + "grad_norm": 0.886323869228363, + "learning_rate": 0.0005019703764098383, + "loss": 3.6184, + "step": 58640 + }, + { + "epoch": 3.9845767087919555, + "grad_norm": 0.9094097018241882, + "learning_rate": 0.0005019279114010055, + "loss": 3.5141, + "step": 58645 + }, + { + "epoch": 3.984916428862617, + "grad_norm": 0.9438634514808655, + "learning_rate": 0.0005018854463921728, + "loss": 3.6474, + "step": 58650 + }, + { + "epoch": 3.9852561489332787, + "grad_norm": 0.7943500876426697, + "learning_rate": 0.0005018429813833402, + "loss": 3.5123, + "step": 58655 + }, + { + "epoch": 3.985595869003941, + "grad_norm": 0.8282130360603333, + "learning_rate": 0.0005018005163745074, + "loss": 3.3624, + "step": 58660 + }, + { + "epoch": 3.9859355890746024, + "grad_norm": 0.843784749507904, + "learning_rate": 0.0005017580513656747, + "loss": 3.5491, + "step": 58665 + }, + { + "epoch": 3.986275309145264, + "grad_norm": 0.8805307745933533, + "learning_rate": 0.000501715586356842, + "loss": 3.3268, + "step": 58670 + }, + { + "epoch": 3.986615029215926, + "grad_norm": 0.7629668712615967, + "learning_rate": 0.0005016731213480092, + "loss": 3.563, + "step": 58675 + }, + { + "epoch": 3.9869547492865878, + "grad_norm": 0.9756906032562256, + "learning_rate": 0.0005016306563391765, + "loss": 3.4587, + "step": 58680 + }, + { + "epoch": 3.9872944693572494, + "grad_norm": 1.0828956365585327, + "learning_rate": 0.0005015881913303438, + "loss": 3.2026, + "step": 58685 + }, + { + "epoch": 3.9876341894279115, + "grad_norm": 1.0435839891433716, + "learning_rate": 0.0005015457263215111, + "loss": 3.5432, + "step": 58690 + }, + { + "epoch": 3.987973909498573, + "grad_norm": 0.6745830178260803, + "learning_rate": 0.0005015032613126783, + "loss": 3.6717, + "step": 58695 + }, + { + "epoch": 3.9883136295692347, + "grad_norm": 0.8049938678741455, + "learning_rate": 0.0005014607963038457, + "loss": 3.3084, + "step": 58700 + }, + { + "epoch": 3.988653349639897, + "grad_norm": 0.776008665561676, + "learning_rate": 0.0005014183312950129, + "loss": 3.5405, + "step": 58705 + }, + { + "epoch": 3.9889930697105584, + "grad_norm": 0.9933795928955078, + "learning_rate": 0.0005013758662861801, + "loss": 3.4247, + "step": 58710 + }, + { + "epoch": 3.98933278978122, + "grad_norm": 0.7957560420036316, + "learning_rate": 0.0005013334012773475, + "loss": 3.6848, + "step": 58715 + }, + { + "epoch": 3.989672509851882, + "grad_norm": 1.2339847087860107, + "learning_rate": 0.0005012909362685147, + "loss": 3.7956, + "step": 58720 + }, + { + "epoch": 3.990012229922544, + "grad_norm": 1.210588812828064, + "learning_rate": 0.000501248471259682, + "loss": 3.1582, + "step": 58725 + }, + { + "epoch": 3.9903519499932054, + "grad_norm": 1.0566565990447998, + "learning_rate": 0.0005012060062508494, + "loss": 3.3669, + "step": 58730 + }, + { + "epoch": 3.9906916700638675, + "grad_norm": 1.2095428705215454, + "learning_rate": 0.0005011635412420166, + "loss": 3.3136, + "step": 58735 + }, + { + "epoch": 3.991031390134529, + "grad_norm": 0.6907721161842346, + "learning_rate": 0.0005011210762331838, + "loss": 3.4939, + "step": 58740 + }, + { + "epoch": 3.9913711102051908, + "grad_norm": 0.8217974901199341, + "learning_rate": 0.0005010786112243512, + "loss": 3.4108, + "step": 58745 + }, + { + "epoch": 3.991710830275853, + "grad_norm": 0.8160417079925537, + "learning_rate": 0.0005010361462155184, + "loss": 3.3811, + "step": 58750 + }, + { + "epoch": 3.9920505503465145, + "grad_norm": 0.8994930386543274, + "learning_rate": 0.0005009936812066856, + "loss": 3.5395, + "step": 58755 + }, + { + "epoch": 3.992390270417176, + "grad_norm": 0.9312856793403625, + "learning_rate": 0.000500951216197853, + "loss": 3.3703, + "step": 58760 + }, + { + "epoch": 3.992729990487838, + "grad_norm": 1.0386981964111328, + "learning_rate": 0.0005009087511890203, + "loss": 3.7951, + "step": 58765 + }, + { + "epoch": 3.9930697105585, + "grad_norm": 0.8984938859939575, + "learning_rate": 0.0005008662861801876, + "loss": 3.4229, + "step": 58770 + }, + { + "epoch": 3.9934094306291614, + "grad_norm": 1.0134419202804565, + "learning_rate": 0.0005008238211713548, + "loss": 3.2128, + "step": 58775 + }, + { + "epoch": 3.9937491506998235, + "grad_norm": 0.914245069026947, + "learning_rate": 0.0005007813561625221, + "loss": 3.3382, + "step": 58780 + }, + { + "epoch": 3.994088870770485, + "grad_norm": 0.9447693824768066, + "learning_rate": 0.0005007388911536894, + "loss": 3.5376, + "step": 58785 + }, + { + "epoch": 3.9944285908411468, + "grad_norm": 0.9548652768135071, + "learning_rate": 0.0005006964261448566, + "loss": 3.3727, + "step": 58790 + }, + { + "epoch": 3.994768310911809, + "grad_norm": 0.924791693687439, + "learning_rate": 0.000500653961136024, + "loss": 3.5487, + "step": 58795 + }, + { + "epoch": 3.9951080309824705, + "grad_norm": 1.0919173955917358, + "learning_rate": 0.0005006114961271913, + "loss": 3.5634, + "step": 58800 + }, + { + "epoch": 3.995447751053132, + "grad_norm": 1.2216788530349731, + "learning_rate": 0.0005005690311183585, + "loss": 3.7662, + "step": 58805 + }, + { + "epoch": 3.995787471123794, + "grad_norm": 0.9165224432945251, + "learning_rate": 0.0005005265661095257, + "loss": 3.5287, + "step": 58810 + }, + { + "epoch": 3.996127191194456, + "grad_norm": 0.9591333270072937, + "learning_rate": 0.0005004841011006931, + "loss": 3.3288, + "step": 58815 + }, + { + "epoch": 3.9964669112651174, + "grad_norm": 1.2829313278198242, + "learning_rate": 0.0005004416360918603, + "loss": 3.6745, + "step": 58820 + }, + { + "epoch": 3.9968066313357795, + "grad_norm": 0.8852560520172119, + "learning_rate": 0.0005003991710830275, + "loss": 3.5221, + "step": 58825 + }, + { + "epoch": 3.997146351406441, + "grad_norm": 0.8069028854370117, + "learning_rate": 0.000500356706074195, + "loss": 3.5751, + "step": 58830 + }, + { + "epoch": 3.9974860714771028, + "grad_norm": 0.8644586801528931, + "learning_rate": 0.0005003142410653622, + "loss": 3.2415, + "step": 58835 + }, + { + "epoch": 3.997825791547765, + "grad_norm": 1.0938270092010498, + "learning_rate": 0.0005002717760565294, + "loss": 3.4985, + "step": 58840 + }, + { + "epoch": 3.9981655116184265, + "grad_norm": 0.6521931290626526, + "learning_rate": 0.0005002293110476968, + "loss": 3.6581, + "step": 58845 + }, + { + "epoch": 3.998505231689088, + "grad_norm": 0.9686088562011719, + "learning_rate": 0.000500186846038864, + "loss": 3.3887, + "step": 58850 + }, + { + "epoch": 3.99884495175975, + "grad_norm": 0.7416316270828247, + "learning_rate": 0.0005001443810300312, + "loss": 3.5103, + "step": 58855 + }, + { + "epoch": 3.999184671830412, + "grad_norm": 0.9525768756866455, + "learning_rate": 0.0005001019160211985, + "loss": 3.688, + "step": 58860 + }, + { + "epoch": 3.9995243919010735, + "grad_norm": 1.0199936628341675, + "learning_rate": 0.0005000594510123659, + "loss": 3.7501, + "step": 58865 + }, + { + "epoch": 3.9998641119717355, + "grad_norm": 1.1737592220306396, + "learning_rate": 0.0005000169860035331, + "loss": 3.4213, + "step": 58870 + }, + { + "epoch": 4.0, + "eval_bertscore": { + "f1": 0.8385404370669559, + "precision": 0.8427524314750152, + "recall": 0.8351357977327213 + }, + "eval_bleu_4": 0.015128049595534286, + "eval_exact_match": 0.0002907258455276674, + "eval_loss": 3.432779550552368, + "eval_meteor": 0.08361186489482236, + "eval_rouge": { + "rouge1": 0.12040188242092503, + "rouge2": 0.017964918767180104, + "rougeL": 0.10407237429544142, + "rougeLsum": 0.10410244498813562 + }, + "eval_runtime": 2386.9732, + "eval_samples_per_second": 4.323, + "eval_steps_per_second": 0.54, + "step": 58872 + }, + { + "epoch": 4.000203832042397, + "grad_norm": 0.8747281432151794, + "learning_rate": 0.0004999745209947003, + "loss": 3.7358, + "step": 58875 + }, + { + "epoch": 4.000543552113059, + "grad_norm": 1.0096979141235352, + "learning_rate": 0.0004999320559858677, + "loss": 3.4673, + "step": 58880 + }, + { + "epoch": 4.000883272183721, + "grad_norm": 0.8625751733779907, + "learning_rate": 0.000499889590977035, + "loss": 3.3461, + "step": 58885 + }, + { + "epoch": 4.001222992254382, + "grad_norm": 0.7489694952964783, + "learning_rate": 0.0004998471259682022, + "loss": 3.3326, + "step": 58890 + }, + { + "epoch": 4.001562712325044, + "grad_norm": 0.9151527285575867, + "learning_rate": 0.0004998046609593694, + "loss": 3.4834, + "step": 58895 + }, + { + "epoch": 4.001902432395706, + "grad_norm": 0.9500098824501038, + "learning_rate": 0.0004997621959505368, + "loss": 3.4627, + "step": 58900 + }, + { + "epoch": 4.002242152466367, + "grad_norm": 0.8319563865661621, + "learning_rate": 0.000499719730941704, + "loss": 3.3326, + "step": 58905 + }, + { + "epoch": 4.0025818725370295, + "grad_norm": 0.6396570801734924, + "learning_rate": 0.0004996772659328713, + "loss": 3.4005, + "step": 58910 + }, + { + "epoch": 4.0029215926076915, + "grad_norm": 1.0601556301116943, + "learning_rate": 0.0004996348009240387, + "loss": 3.3145, + "step": 58915 + }, + { + "epoch": 4.003261312678353, + "grad_norm": 1.120114803314209, + "learning_rate": 0.0004995923359152059, + "loss": 3.5709, + "step": 58920 + }, + { + "epoch": 4.003601032749015, + "grad_norm": 0.8364360928535461, + "learning_rate": 0.0004995498709063731, + "loss": 3.3213, + "step": 58925 + }, + { + "epoch": 4.003940752819677, + "grad_norm": 1.0632083415985107, + "learning_rate": 0.0004995074058975405, + "loss": 3.7191, + "step": 58930 + }, + { + "epoch": 4.004280472890338, + "grad_norm": 1.0068776607513428, + "learning_rate": 0.0004994649408887077, + "loss": 3.3446, + "step": 58935 + }, + { + "epoch": 4.004620192961, + "grad_norm": 0.7390163540840149, + "learning_rate": 0.000499422475879875, + "loss": 3.5669, + "step": 58940 + }, + { + "epoch": 4.004959913031662, + "grad_norm": 0.7953309416770935, + "learning_rate": 0.0004993800108710422, + "loss": 3.4071, + "step": 58945 + }, + { + "epoch": 4.005299633102323, + "grad_norm": 0.7707774043083191, + "learning_rate": 0.0004993375458622096, + "loss": 3.4354, + "step": 58950 + }, + { + "epoch": 4.0056393531729855, + "grad_norm": 0.9267033934593201, + "learning_rate": 0.0004992950808533768, + "loss": 3.3968, + "step": 58955 + }, + { + "epoch": 4.0059790732436475, + "grad_norm": 1.0014992952346802, + "learning_rate": 0.0004992526158445441, + "loss": 3.5783, + "step": 58960 + }, + { + "epoch": 4.006318793314309, + "grad_norm": 0.8574535250663757, + "learning_rate": 0.0004992101508357114, + "loss": 3.3148, + "step": 58965 + }, + { + "epoch": 4.006658513384971, + "grad_norm": 0.8339497447013855, + "learning_rate": 0.0004991676858268787, + "loss": 3.3311, + "step": 58970 + }, + { + "epoch": 4.006998233455633, + "grad_norm": 0.8833082318305969, + "learning_rate": 0.0004991252208180459, + "loss": 3.6522, + "step": 58975 + }, + { + "epoch": 4.007337953526294, + "grad_norm": 0.866485059261322, + "learning_rate": 0.0004990827558092132, + "loss": 3.5873, + "step": 58980 + }, + { + "epoch": 4.007677673596956, + "grad_norm": 0.8681113719940186, + "learning_rate": 0.0004990402908003805, + "loss": 3.6125, + "step": 58985 + }, + { + "epoch": 4.008017393667618, + "grad_norm": 0.790141224861145, + "learning_rate": 0.0004989978257915478, + "loss": 3.4763, + "step": 58990 + }, + { + "epoch": 4.008357113738279, + "grad_norm": 0.8715270161628723, + "learning_rate": 0.000498955360782715, + "loss": 3.5145, + "step": 58995 + }, + { + "epoch": 4.0086968338089415, + "grad_norm": 0.8293136954307556, + "learning_rate": 0.0004989128957738823, + "loss": 3.6876, + "step": 59000 + }, + { + "epoch": 4.009036553879604, + "grad_norm": 0.7862138748168945, + "learning_rate": 0.0004988704307650496, + "loss": 3.4034, + "step": 59005 + }, + { + "epoch": 4.009376273950265, + "grad_norm": 0.8801670670509338, + "learning_rate": 0.0004988279657562168, + "loss": 3.426, + "step": 59010 + }, + { + "epoch": 4.009715994020927, + "grad_norm": 1.132441520690918, + "learning_rate": 0.0004987855007473842, + "loss": 3.3286, + "step": 59015 + }, + { + "epoch": 4.010055714091589, + "grad_norm": 0.6591469049453735, + "learning_rate": 0.0004987430357385515, + "loss": 3.8127, + "step": 59020 + }, + { + "epoch": 4.01039543416225, + "grad_norm": 0.9934965372085571, + "learning_rate": 0.0004987005707297187, + "loss": 3.633, + "step": 59025 + }, + { + "epoch": 4.010735154232912, + "grad_norm": 0.9317359924316406, + "learning_rate": 0.000498658105720886, + "loss": 3.4313, + "step": 59030 + }, + { + "epoch": 4.011074874303574, + "grad_norm": 0.8778597712516785, + "learning_rate": 0.0004986156407120533, + "loss": 3.3994, + "step": 59035 + }, + { + "epoch": 4.011414594374235, + "grad_norm": 1.0319303274154663, + "learning_rate": 0.0004985731757032206, + "loss": 3.291, + "step": 59040 + }, + { + "epoch": 4.0117543144448975, + "grad_norm": 0.7920326590538025, + "learning_rate": 0.0004985307106943878, + "loss": 3.5725, + "step": 59045 + }, + { + "epoch": 4.01209403451556, + "grad_norm": 0.9608009457588196, + "learning_rate": 0.0004984882456855551, + "loss": 3.471, + "step": 59050 + }, + { + "epoch": 4.012433754586221, + "grad_norm": 1.0730987787246704, + "learning_rate": 0.0004984457806767224, + "loss": 3.4814, + "step": 59055 + }, + { + "epoch": 4.012773474656883, + "grad_norm": 0.8771940469741821, + "learning_rate": 0.0004984033156678896, + "loss": 3.5399, + "step": 59060 + }, + { + "epoch": 4.013113194727545, + "grad_norm": 0.7945749163627625, + "learning_rate": 0.000498360850659057, + "loss": 3.6079, + "step": 59065 + }, + { + "epoch": 4.013452914798206, + "grad_norm": 0.9251720905303955, + "learning_rate": 0.0004983183856502243, + "loss": 3.5025, + "step": 59070 + }, + { + "epoch": 4.013792634868868, + "grad_norm": 0.879957914352417, + "learning_rate": 0.0004982759206413915, + "loss": 3.3794, + "step": 59075 + }, + { + "epoch": 4.01413235493953, + "grad_norm": 0.8275218605995178, + "learning_rate": 0.0004982334556325588, + "loss": 3.5161, + "step": 59080 + }, + { + "epoch": 4.014472075010191, + "grad_norm": 1.0662089586257935, + "learning_rate": 0.0004981909906237261, + "loss": 3.4729, + "step": 59085 + }, + { + "epoch": 4.0148117950808535, + "grad_norm": 0.924647331237793, + "learning_rate": 0.0004981485256148933, + "loss": 3.5644, + "step": 59090 + }, + { + "epoch": 4.015151515151516, + "grad_norm": 0.692802906036377, + "learning_rate": 0.0004981060606060606, + "loss": 3.5337, + "step": 59095 + }, + { + "epoch": 4.015491235222177, + "grad_norm": 0.7831913828849792, + "learning_rate": 0.0004980635955972279, + "loss": 3.6386, + "step": 59100 + }, + { + "epoch": 4.015830955292839, + "grad_norm": 0.8935628533363342, + "learning_rate": 0.0004980211305883952, + "loss": 3.374, + "step": 59105 + }, + { + "epoch": 4.016170675363501, + "grad_norm": 0.9295881390571594, + "learning_rate": 0.0004979786655795624, + "loss": 3.4827, + "step": 59110 + }, + { + "epoch": 4.016510395434162, + "grad_norm": 0.9850232601165771, + "learning_rate": 0.0004979362005707298, + "loss": 3.2893, + "step": 59115 + }, + { + "epoch": 4.016850115504824, + "grad_norm": 1.0450438261032104, + "learning_rate": 0.000497893735561897, + "loss": 3.4132, + "step": 59120 + }, + { + "epoch": 4.017189835575485, + "grad_norm": 0.9847548007965088, + "learning_rate": 0.0004978512705530643, + "loss": 3.2274, + "step": 59125 + }, + { + "epoch": 4.017529555646147, + "grad_norm": 0.6877517700195312, + "learning_rate": 0.0004978088055442316, + "loss": 3.5862, + "step": 59130 + }, + { + "epoch": 4.0178692757168095, + "grad_norm": 1.0290706157684326, + "learning_rate": 0.0004977663405353988, + "loss": 3.6977, + "step": 59135 + }, + { + "epoch": 4.018208995787471, + "grad_norm": 1.152679204940796, + "learning_rate": 0.0004977238755265661, + "loss": 3.5757, + "step": 59140 + }, + { + "epoch": 4.018548715858133, + "grad_norm": 0.7825775146484375, + "learning_rate": 0.0004976814105177334, + "loss": 3.4211, + "step": 59145 + }, + { + "epoch": 4.018888435928795, + "grad_norm": 0.7771532535552979, + "learning_rate": 0.0004976389455089007, + "loss": 3.4468, + "step": 59150 + }, + { + "epoch": 4.019228155999456, + "grad_norm": 0.6235162615776062, + "learning_rate": 0.0004975964805000679, + "loss": 3.751, + "step": 59155 + }, + { + "epoch": 4.019567876070118, + "grad_norm": 0.8276233673095703, + "learning_rate": 0.0004975540154912352, + "loss": 3.2886, + "step": 59160 + }, + { + "epoch": 4.01990759614078, + "grad_norm": 0.7455196976661682, + "learning_rate": 0.0004975115504824025, + "loss": 3.5087, + "step": 59165 + }, + { + "epoch": 4.020247316211441, + "grad_norm": 0.7343783974647522, + "learning_rate": 0.0004974690854735698, + "loss": 3.3992, + "step": 59170 + }, + { + "epoch": 4.0205870362821035, + "grad_norm": 0.8975481390953064, + "learning_rate": 0.0004974266204647371, + "loss": 3.6677, + "step": 59175 + }, + { + "epoch": 4.0209267563527655, + "grad_norm": 0.8668843507766724, + "learning_rate": 0.0004973841554559044, + "loss": 3.352, + "step": 59180 + }, + { + "epoch": 4.021266476423427, + "grad_norm": 1.0666024684906006, + "learning_rate": 0.0004973416904470716, + "loss": 3.6495, + "step": 59185 + }, + { + "epoch": 4.021606196494089, + "grad_norm": 0.83885657787323, + "learning_rate": 0.0004972992254382389, + "loss": 3.7973, + "step": 59190 + }, + { + "epoch": 4.021945916564751, + "grad_norm": 0.8596224784851074, + "learning_rate": 0.0004972567604294061, + "loss": 3.5714, + "step": 59195 + }, + { + "epoch": 4.022285636635412, + "grad_norm": 0.86994469165802, + "learning_rate": 0.0004972142954205735, + "loss": 3.4898, + "step": 59200 + }, + { + "epoch": 4.022625356706074, + "grad_norm": 0.8553299307823181, + "learning_rate": 0.0004971718304117407, + "loss": 3.4782, + "step": 59205 + }, + { + "epoch": 4.022965076776736, + "grad_norm": 0.8902169466018677, + "learning_rate": 0.000497129365402908, + "loss": 3.5854, + "step": 59210 + }, + { + "epoch": 4.023304796847397, + "grad_norm": 0.8655052781105042, + "learning_rate": 0.0004970869003940753, + "loss": 3.6774, + "step": 59215 + }, + { + "epoch": 4.0236445169180595, + "grad_norm": 0.7810933589935303, + "learning_rate": 0.0004970444353852426, + "loss": 3.4983, + "step": 59220 + }, + { + "epoch": 4.0239842369887215, + "grad_norm": 0.9515738487243652, + "learning_rate": 0.0004970019703764098, + "loss": 3.5432, + "step": 59225 + }, + { + "epoch": 4.024323957059383, + "grad_norm": 0.6742497682571411, + "learning_rate": 0.0004969595053675772, + "loss": 3.7443, + "step": 59230 + }, + { + "epoch": 4.024663677130045, + "grad_norm": 0.7862504720687866, + "learning_rate": 0.0004969170403587444, + "loss": 3.6295, + "step": 59235 + }, + { + "epoch": 4.025003397200707, + "grad_norm": 0.8000181913375854, + "learning_rate": 0.0004968745753499117, + "loss": 3.5042, + "step": 59240 + }, + { + "epoch": 4.025343117271368, + "grad_norm": 0.8202589750289917, + "learning_rate": 0.0004968321103410789, + "loss": 3.6363, + "step": 59245 + }, + { + "epoch": 4.02568283734203, + "grad_norm": 0.9175881743431091, + "learning_rate": 0.0004967896453322463, + "loss": 3.6685, + "step": 59250 + }, + { + "epoch": 4.026022557412692, + "grad_norm": 0.8016685247421265, + "learning_rate": 0.0004967471803234135, + "loss": 3.4981, + "step": 59255 + }, + { + "epoch": 4.026362277483353, + "grad_norm": 0.7623411417007446, + "learning_rate": 0.0004967047153145807, + "loss": 3.4201, + "step": 59260 + }, + { + "epoch": 4.0267019975540155, + "grad_norm": 0.8220285773277283, + "learning_rate": 0.0004966622503057481, + "loss": 3.745, + "step": 59265 + }, + { + "epoch": 4.0270417176246776, + "grad_norm": 0.9626908302307129, + "learning_rate": 0.0004966197852969154, + "loss": 3.6727, + "step": 59270 + }, + { + "epoch": 4.027381437695339, + "grad_norm": 0.963424026966095, + "learning_rate": 0.0004965773202880826, + "loss": 3.5682, + "step": 59275 + }, + { + "epoch": 4.027721157766001, + "grad_norm": 0.7740755677223206, + "learning_rate": 0.00049653485527925, + "loss": 3.3562, + "step": 59280 + }, + { + "epoch": 4.028060877836663, + "grad_norm": 0.924393355846405, + "learning_rate": 0.0004964923902704172, + "loss": 3.6037, + "step": 59285 + }, + { + "epoch": 4.028400597907324, + "grad_norm": 0.7366365790367126, + "learning_rate": 0.0004964499252615844, + "loss": 3.3526, + "step": 59290 + }, + { + "epoch": 4.028740317977986, + "grad_norm": 1.064188838005066, + "learning_rate": 0.0004964074602527517, + "loss": 3.437, + "step": 59295 + }, + { + "epoch": 4.029080038048648, + "grad_norm": 0.610778272151947, + "learning_rate": 0.0004963649952439191, + "loss": 3.5023, + "step": 59300 + }, + { + "epoch": 4.029419758119309, + "grad_norm": 0.906237781047821, + "learning_rate": 0.0004963225302350863, + "loss": 3.488, + "step": 59305 + }, + { + "epoch": 4.0297594781899715, + "grad_norm": 0.8823177218437195, + "learning_rate": 0.0004962800652262535, + "loss": 3.364, + "step": 59310 + }, + { + "epoch": 4.030099198260634, + "grad_norm": 0.8561570644378662, + "learning_rate": 0.0004962376002174209, + "loss": 3.4643, + "step": 59315 + }, + { + "epoch": 4.030438918331295, + "grad_norm": 0.7763888835906982, + "learning_rate": 0.0004961951352085881, + "loss": 3.534, + "step": 59320 + }, + { + "epoch": 4.030778638401957, + "grad_norm": 0.8408674597740173, + "learning_rate": 0.0004961526701997554, + "loss": 3.5215, + "step": 59325 + }, + { + "epoch": 4.031118358472619, + "grad_norm": 0.7309010624885559, + "learning_rate": 0.0004961102051909228, + "loss": 3.6271, + "step": 59330 + }, + { + "epoch": 4.03145807854328, + "grad_norm": 1.031607747077942, + "learning_rate": 0.00049606774018209, + "loss": 3.5652, + "step": 59335 + }, + { + "epoch": 4.031797798613942, + "grad_norm": 0.8041316866874695, + "learning_rate": 0.0004960252751732572, + "loss": 3.4452, + "step": 59340 + }, + { + "epoch": 4.032137518684604, + "grad_norm": 1.041819453239441, + "learning_rate": 0.0004959828101644245, + "loss": 3.4553, + "step": 59345 + }, + { + "epoch": 4.032477238755265, + "grad_norm": 0.8617551922798157, + "learning_rate": 0.0004959403451555918, + "loss": 3.5529, + "step": 59350 + }, + { + "epoch": 4.0328169588259275, + "grad_norm": 0.9571493864059448, + "learning_rate": 0.0004958978801467591, + "loss": 3.56, + "step": 59355 + }, + { + "epoch": 4.03315667889659, + "grad_norm": 0.964963436126709, + "learning_rate": 0.0004958554151379263, + "loss": 3.3228, + "step": 59360 + }, + { + "epoch": 4.033496398967251, + "grad_norm": 0.8677114248275757, + "learning_rate": 0.0004958129501290937, + "loss": 3.3487, + "step": 59365 + }, + { + "epoch": 4.033836119037913, + "grad_norm": 0.8580537438392639, + "learning_rate": 0.0004957704851202609, + "loss": 3.474, + "step": 59370 + }, + { + "epoch": 4.034175839108575, + "grad_norm": 0.6939281225204468, + "learning_rate": 0.0004957280201114282, + "loss": 3.397, + "step": 59375 + }, + { + "epoch": 4.034515559179236, + "grad_norm": 0.7544398307800293, + "learning_rate": 0.0004956855551025954, + "loss": 3.7162, + "step": 59380 + }, + { + "epoch": 4.034855279249898, + "grad_norm": 0.8536282777786255, + "learning_rate": 0.0004956430900937628, + "loss": 3.4894, + "step": 59385 + }, + { + "epoch": 4.03519499932056, + "grad_norm": 0.8817777037620544, + "learning_rate": 0.00049560062508493, + "loss": 3.3923, + "step": 59390 + }, + { + "epoch": 4.035534719391221, + "grad_norm": 0.9778486490249634, + "learning_rate": 0.0004955581600760973, + "loss": 3.2971, + "step": 59395 + }, + { + "epoch": 4.0358744394618835, + "grad_norm": 9.085989952087402, + "learning_rate": 0.0004955156950672646, + "loss": 3.2477, + "step": 59400 + }, + { + "epoch": 4.036214159532546, + "grad_norm": 0.7704107165336609, + "learning_rate": 0.0004954732300584319, + "loss": 3.4429, + "step": 59405 + }, + { + "epoch": 4.036553879603207, + "grad_norm": 0.968893826007843, + "learning_rate": 0.0004954307650495991, + "loss": 3.3488, + "step": 59410 + }, + { + "epoch": 4.036893599673869, + "grad_norm": 0.88393634557724, + "learning_rate": 0.0004953883000407664, + "loss": 3.3584, + "step": 59415 + }, + { + "epoch": 4.037233319744531, + "grad_norm": 1.0462424755096436, + "learning_rate": 0.0004953458350319337, + "loss": 3.3883, + "step": 59420 + }, + { + "epoch": 4.037573039815192, + "grad_norm": 0.9671331644058228, + "learning_rate": 0.000495303370023101, + "loss": 3.4942, + "step": 59425 + }, + { + "epoch": 4.037912759885854, + "grad_norm": 1.1834206581115723, + "learning_rate": 0.0004952609050142683, + "loss": 3.5535, + "step": 59430 + }, + { + "epoch": 4.038252479956516, + "grad_norm": 0.8951017260551453, + "learning_rate": 0.0004952184400054356, + "loss": 3.4215, + "step": 59435 + }, + { + "epoch": 4.0385922000271774, + "grad_norm": 0.9307129383087158, + "learning_rate": 0.0004951759749966028, + "loss": 3.4521, + "step": 59440 + }, + { + "epoch": 4.0389319200978395, + "grad_norm": 1.2310844659805298, + "learning_rate": 0.00049513350998777, + "loss": 3.5149, + "step": 59445 + }, + { + "epoch": 4.039271640168501, + "grad_norm": 0.9202895760536194, + "learning_rate": 0.0004950910449789374, + "loss": 3.3708, + "step": 59450 + }, + { + "epoch": 4.039611360239163, + "grad_norm": 0.9394950866699219, + "learning_rate": 0.0004950485799701047, + "loss": 3.7742, + "step": 59455 + }, + { + "epoch": 4.039951080309825, + "grad_norm": 0.7934741377830505, + "learning_rate": 0.0004950061149612719, + "loss": 3.5666, + "step": 59460 + }, + { + "epoch": 4.040290800380486, + "grad_norm": 0.9467673897743225, + "learning_rate": 0.0004949636499524392, + "loss": 3.4321, + "step": 59465 + }, + { + "epoch": 4.040630520451148, + "grad_norm": 0.7550963163375854, + "learning_rate": 0.0004949211849436065, + "loss": 3.6178, + "step": 59470 + }, + { + "epoch": 4.04097024052181, + "grad_norm": 1.1531764268875122, + "learning_rate": 0.0004948787199347737, + "loss": 3.5039, + "step": 59475 + }, + { + "epoch": 4.041309960592471, + "grad_norm": 0.912929117679596, + "learning_rate": 0.000494836254925941, + "loss": 3.5001, + "step": 59480 + }, + { + "epoch": 4.0416496806631335, + "grad_norm": 0.9658652544021606, + "learning_rate": 0.0004947937899171083, + "loss": 3.455, + "step": 59485 + }, + { + "epoch": 4.0419894007337955, + "grad_norm": 1.0641367435455322, + "learning_rate": 0.0004947513249082756, + "loss": 3.5756, + "step": 59490 + }, + { + "epoch": 4.042329120804457, + "grad_norm": 1.0192972421646118, + "learning_rate": 0.0004947088598994428, + "loss": 3.7225, + "step": 59495 + }, + { + "epoch": 4.042668840875119, + "grad_norm": 0.9425919651985168, + "learning_rate": 0.0004946663948906102, + "loss": 3.5573, + "step": 59500 + }, + { + "epoch": 4.043008560945781, + "grad_norm": 0.6880160570144653, + "learning_rate": 0.0004946239298817774, + "loss": 3.7185, + "step": 59505 + }, + { + "epoch": 4.043348281016442, + "grad_norm": 0.8692899346351624, + "learning_rate": 0.0004945814648729447, + "loss": 3.4777, + "step": 59510 + }, + { + "epoch": 4.043688001087104, + "grad_norm": 0.9909335374832153, + "learning_rate": 0.000494538999864112, + "loss": 3.4808, + "step": 59515 + }, + { + "epoch": 4.044027721157766, + "grad_norm": 1.2313508987426758, + "learning_rate": 0.0004944965348552792, + "loss": 3.3739, + "step": 59520 + }, + { + "epoch": 4.044367441228427, + "grad_norm": 0.7821127772331238, + "learning_rate": 0.0004944540698464465, + "loss": 3.4041, + "step": 59525 + }, + { + "epoch": 4.0447071612990895, + "grad_norm": 0.7277593612670898, + "learning_rate": 0.0004944116048376139, + "loss": 3.4754, + "step": 59530 + }, + { + "epoch": 4.0450468813697515, + "grad_norm": 0.9902859926223755, + "learning_rate": 0.0004943691398287811, + "loss": 3.6015, + "step": 59535 + }, + { + "epoch": 4.045386601440413, + "grad_norm": 0.75090491771698, + "learning_rate": 0.0004943266748199484, + "loss": 3.6644, + "step": 59540 + }, + { + "epoch": 4.045726321511075, + "grad_norm": 0.8302679061889648, + "learning_rate": 0.0004942842098111156, + "loss": 3.3744, + "step": 59545 + }, + { + "epoch": 4.046066041581737, + "grad_norm": 0.8659022450447083, + "learning_rate": 0.000494241744802283, + "loss": 3.4081, + "step": 59550 + }, + { + "epoch": 4.046405761652398, + "grad_norm": 0.8431478142738342, + "learning_rate": 0.0004941992797934502, + "loss": 3.2691, + "step": 59555 + }, + { + "epoch": 4.04674548172306, + "grad_norm": 0.8332762122154236, + "learning_rate": 0.0004941568147846175, + "loss": 3.7002, + "step": 59560 + }, + { + "epoch": 4.047085201793722, + "grad_norm": 1.0419656038284302, + "learning_rate": 0.0004941143497757848, + "loss": 3.5708, + "step": 59565 + }, + { + "epoch": 4.047424921864383, + "grad_norm": 1.036781907081604, + "learning_rate": 0.000494071884766952, + "loss": 3.3737, + "step": 59570 + }, + { + "epoch": 4.0477646419350455, + "grad_norm": 0.8540711998939514, + "learning_rate": 0.0004940294197581193, + "loss": 3.2013, + "step": 59575 + }, + { + "epoch": 4.048104362005708, + "grad_norm": 0.8660485148429871, + "learning_rate": 0.0004939869547492867, + "loss": 3.5735, + "step": 59580 + }, + { + "epoch": 4.048444082076369, + "grad_norm": 1.5885177850723267, + "learning_rate": 0.0004939444897404539, + "loss": 3.4739, + "step": 59585 + }, + { + "epoch": 4.048783802147031, + "grad_norm": 0.8557515144348145, + "learning_rate": 0.0004939020247316211, + "loss": 3.3974, + "step": 59590 + }, + { + "epoch": 4.049123522217693, + "grad_norm": 1.22386634349823, + "learning_rate": 0.0004938595597227884, + "loss": 3.4099, + "step": 59595 + }, + { + "epoch": 4.049463242288354, + "grad_norm": 0.9990693926811218, + "learning_rate": 0.0004938170947139557, + "loss": 3.6372, + "step": 59600 + }, + { + "epoch": 4.049802962359016, + "grad_norm": 1.225592851638794, + "learning_rate": 0.000493774629705123, + "loss": 3.6008, + "step": 59605 + }, + { + "epoch": 4.050142682429678, + "grad_norm": 0.8566232323646545, + "learning_rate": 0.0004937321646962903, + "loss": 3.5786, + "step": 59610 + }, + { + "epoch": 4.050482402500339, + "grad_norm": 0.8954619765281677, + "learning_rate": 0.0004936896996874576, + "loss": 3.4853, + "step": 59615 + }, + { + "epoch": 4.0508221225710015, + "grad_norm": 1.31369149684906, + "learning_rate": 0.0004936472346786248, + "loss": 3.5876, + "step": 59620 + }, + { + "epoch": 4.051161842641664, + "grad_norm": 0.9822103381156921, + "learning_rate": 0.0004936047696697921, + "loss": 3.3509, + "step": 59625 + }, + { + "epoch": 4.051501562712325, + "grad_norm": 0.8911438584327698, + "learning_rate": 0.0004935623046609593, + "loss": 3.4431, + "step": 59630 + }, + { + "epoch": 4.051841282782987, + "grad_norm": 1.1207389831542969, + "learning_rate": 0.0004935198396521267, + "loss": 3.768, + "step": 59635 + }, + { + "epoch": 4.052181002853649, + "grad_norm": 0.6248236894607544, + "learning_rate": 0.0004934773746432939, + "loss": 3.3655, + "step": 59640 + }, + { + "epoch": 4.05252072292431, + "grad_norm": 0.7271294593811035, + "learning_rate": 0.0004934349096344612, + "loss": 3.5107, + "step": 59645 + }, + { + "epoch": 4.052860442994972, + "grad_norm": 1.7869746685028076, + "learning_rate": 0.0004933924446256285, + "loss": 3.5311, + "step": 59650 + }, + { + "epoch": 4.053200163065634, + "grad_norm": 0.9492328763008118, + "learning_rate": 0.0004933499796167958, + "loss": 3.4546, + "step": 59655 + }, + { + "epoch": 4.053539883136295, + "grad_norm": 0.9061926603317261, + "learning_rate": 0.000493307514607963, + "loss": 3.3874, + "step": 59660 + }, + { + "epoch": 4.0538796032069575, + "grad_norm": 1.0815743207931519, + "learning_rate": 0.0004932650495991304, + "loss": 3.547, + "step": 59665 + }, + { + "epoch": 4.05421932327762, + "grad_norm": 0.9014431834220886, + "learning_rate": 0.0004932225845902976, + "loss": 3.6709, + "step": 59670 + }, + { + "epoch": 4.054559043348281, + "grad_norm": 3.110290765762329, + "learning_rate": 0.0004931801195814648, + "loss": 3.5499, + "step": 59675 + }, + { + "epoch": 4.054898763418943, + "grad_norm": 0.8537842631340027, + "learning_rate": 0.0004931376545726321, + "loss": 3.6232, + "step": 59680 + }, + { + "epoch": 4.055238483489605, + "grad_norm": 0.8123435974121094, + "learning_rate": 0.0004930951895637995, + "loss": 3.4592, + "step": 59685 + }, + { + "epoch": 4.055578203560266, + "grad_norm": 0.9474018812179565, + "learning_rate": 0.0004930527245549667, + "loss": 3.3542, + "step": 59690 + }, + { + "epoch": 4.055917923630928, + "grad_norm": 0.994622528553009, + "learning_rate": 0.000493010259546134, + "loss": 3.4069, + "step": 59695 + }, + { + "epoch": 4.05625764370159, + "grad_norm": 0.9391903281211853, + "learning_rate": 0.0004929677945373013, + "loss": 3.3757, + "step": 59700 + }, + { + "epoch": 4.056597363772251, + "grad_norm": 0.9274879097938538, + "learning_rate": 0.0004929253295284685, + "loss": 3.4118, + "step": 59705 + }, + { + "epoch": 4.0569370838429135, + "grad_norm": 0.8870189785957336, + "learning_rate": 0.0004928828645196358, + "loss": 3.5679, + "step": 59710 + }, + { + "epoch": 4.057276803913576, + "grad_norm": 0.8023012280464172, + "learning_rate": 0.0004928403995108032, + "loss": 3.6025, + "step": 59715 + }, + { + "epoch": 4.057616523984237, + "grad_norm": 0.8325897455215454, + "learning_rate": 0.0004927979345019704, + "loss": 3.4824, + "step": 59720 + }, + { + "epoch": 4.057956244054899, + "grad_norm": 0.7491126656532288, + "learning_rate": 0.0004927554694931376, + "loss": 3.5764, + "step": 59725 + }, + { + "epoch": 4.058295964125561, + "grad_norm": 1.0194222927093506, + "learning_rate": 0.000492713004484305, + "loss": 3.5386, + "step": 59730 + }, + { + "epoch": 4.058635684196222, + "grad_norm": 0.7877004146575928, + "learning_rate": 0.0004926705394754723, + "loss": 3.442, + "step": 59735 + }, + { + "epoch": 4.058975404266884, + "grad_norm": 1.1093658208847046, + "learning_rate": 0.0004926280744666395, + "loss": 3.5874, + "step": 59740 + }, + { + "epoch": 4.059315124337546, + "grad_norm": 0.7492717504501343, + "learning_rate": 0.0004925856094578067, + "loss": 3.5593, + "step": 59745 + }, + { + "epoch": 4.0596548444082075, + "grad_norm": 0.9673604965209961, + "learning_rate": 0.0004925431444489741, + "loss": 3.5165, + "step": 59750 + }, + { + "epoch": 4.0599945644788695, + "grad_norm": 0.8375464081764221, + "learning_rate": 0.0004925006794401413, + "loss": 3.4809, + "step": 59755 + }, + { + "epoch": 4.060334284549532, + "grad_norm": 0.9903005361557007, + "learning_rate": 0.0004924582144313086, + "loss": 3.4375, + "step": 59760 + }, + { + "epoch": 4.060674004620193, + "grad_norm": 1.018079400062561, + "learning_rate": 0.000492415749422476, + "loss": 3.3943, + "step": 59765 + }, + { + "epoch": 4.061013724690855, + "grad_norm": 1.0245575904846191, + "learning_rate": 0.0004923732844136432, + "loss": 3.4658, + "step": 59770 + }, + { + "epoch": 4.061353444761517, + "grad_norm": 1.0447276830673218, + "learning_rate": 0.0004923308194048104, + "loss": 3.5528, + "step": 59775 + }, + { + "epoch": 4.061693164832178, + "grad_norm": 0.8653343915939331, + "learning_rate": 0.0004922883543959777, + "loss": 3.4713, + "step": 59780 + }, + { + "epoch": 4.06203288490284, + "grad_norm": 1.0537867546081543, + "learning_rate": 0.000492245889387145, + "loss": 3.393, + "step": 59785 + }, + { + "epoch": 4.062372604973502, + "grad_norm": 0.9974700808525085, + "learning_rate": 0.0004922034243783123, + "loss": 3.541, + "step": 59790 + }, + { + "epoch": 4.0627123250441635, + "grad_norm": 1.0795929431915283, + "learning_rate": 0.0004921609593694795, + "loss": 3.745, + "step": 59795 + }, + { + "epoch": 4.0630520451148255, + "grad_norm": 1.1418275833129883, + "learning_rate": 0.0004921184943606469, + "loss": 3.5299, + "step": 59800 + }, + { + "epoch": 4.063391765185487, + "grad_norm": 0.7432627081871033, + "learning_rate": 0.0004920760293518141, + "loss": 3.5906, + "step": 59805 + }, + { + "epoch": 4.063731485256149, + "grad_norm": 0.8620191216468811, + "learning_rate": 0.0004920335643429814, + "loss": 3.5134, + "step": 59810 + }, + { + "epoch": 4.064071205326811, + "grad_norm": 0.9250773191452026, + "learning_rate": 0.0004919910993341487, + "loss": 3.3716, + "step": 59815 + }, + { + "epoch": 4.064410925397472, + "grad_norm": 0.7490436434745789, + "learning_rate": 0.000491948634325316, + "loss": 3.6238, + "step": 59820 + }, + { + "epoch": 4.064750645468134, + "grad_norm": 1.1570332050323486, + "learning_rate": 0.0004919061693164832, + "loss": 3.4601, + "step": 59825 + }, + { + "epoch": 4.065090365538796, + "grad_norm": 0.9343281388282776, + "learning_rate": 0.0004918637043076504, + "loss": 3.3059, + "step": 59830 + }, + { + "epoch": 4.065430085609457, + "grad_norm": 0.8995257616043091, + "learning_rate": 0.0004918212392988178, + "loss": 3.2564, + "step": 59835 + }, + { + "epoch": 4.0657698056801195, + "grad_norm": 0.8480408787727356, + "learning_rate": 0.0004917787742899851, + "loss": 3.7564, + "step": 59840 + }, + { + "epoch": 4.0661095257507816, + "grad_norm": 0.800126314163208, + "learning_rate": 0.0004917363092811523, + "loss": 3.6931, + "step": 59845 + }, + { + "epoch": 4.066449245821443, + "grad_norm": 0.9306234121322632, + "learning_rate": 0.0004916938442723196, + "loss": 3.4449, + "step": 59850 + }, + { + "epoch": 4.066788965892105, + "grad_norm": 1.2591867446899414, + "learning_rate": 0.0004916513792634869, + "loss": 3.1755, + "step": 59855 + }, + { + "epoch": 4.067128685962767, + "grad_norm": 1.1272797584533691, + "learning_rate": 0.0004916089142546541, + "loss": 3.3914, + "step": 59860 + }, + { + "epoch": 4.067468406033428, + "grad_norm": 1.0475786924362183, + "learning_rate": 0.0004915664492458215, + "loss": 3.4657, + "step": 59865 + }, + { + "epoch": 4.06780812610409, + "grad_norm": 0.8131334185600281, + "learning_rate": 0.0004915239842369888, + "loss": 3.6088, + "step": 59870 + }, + { + "epoch": 4.068147846174752, + "grad_norm": 0.8794819712638855, + "learning_rate": 0.000491481519228156, + "loss": 3.6126, + "step": 59875 + }, + { + "epoch": 4.068487566245413, + "grad_norm": 1.0365935564041138, + "learning_rate": 0.0004914390542193232, + "loss": 3.654, + "step": 59880 + }, + { + "epoch": 4.0688272863160755, + "grad_norm": 1.0568506717681885, + "learning_rate": 0.0004913965892104906, + "loss": 3.5577, + "step": 59885 + }, + { + "epoch": 4.069167006386738, + "grad_norm": 0.87224942445755, + "learning_rate": 0.0004913541242016579, + "loss": 3.4776, + "step": 59890 + }, + { + "epoch": 4.069506726457399, + "grad_norm": 1.0197595357894897, + "learning_rate": 0.0004913116591928251, + "loss": 3.495, + "step": 59895 + }, + { + "epoch": 4.069846446528061, + "grad_norm": 0.8176692128181458, + "learning_rate": 0.0004912691941839924, + "loss": 3.6596, + "step": 59900 + }, + { + "epoch": 4.070186166598723, + "grad_norm": 0.8002888560295105, + "learning_rate": 0.0004912267291751597, + "loss": 3.7552, + "step": 59905 + }, + { + "epoch": 4.070525886669384, + "grad_norm": 0.8595325946807861, + "learning_rate": 0.0004911842641663269, + "loss": 3.3244, + "step": 59910 + }, + { + "epoch": 4.070865606740046, + "grad_norm": 1.0854674577713013, + "learning_rate": 0.0004911417991574943, + "loss": 3.6076, + "step": 59915 + }, + { + "epoch": 4.071205326810708, + "grad_norm": 1.0645173788070679, + "learning_rate": 0.0004910993341486616, + "loss": 3.5576, + "step": 59920 + }, + { + "epoch": 4.071545046881369, + "grad_norm": 0.7809297442436218, + "learning_rate": 0.0004910568691398288, + "loss": 3.5359, + "step": 59925 + }, + { + "epoch": 4.0718847669520315, + "grad_norm": 0.9088901877403259, + "learning_rate": 0.000491014404130996, + "loss": 3.6544, + "step": 59930 + }, + { + "epoch": 4.072224487022694, + "grad_norm": 0.7722979784011841, + "learning_rate": 0.0004909719391221634, + "loss": 3.441, + "step": 59935 + }, + { + "epoch": 4.072564207093355, + "grad_norm": 0.8845393657684326, + "learning_rate": 0.0004909294741133306, + "loss": 3.718, + "step": 59940 + }, + { + "epoch": 4.072903927164017, + "grad_norm": 0.928385853767395, + "learning_rate": 0.0004908870091044979, + "loss": 3.8538, + "step": 59945 + }, + { + "epoch": 4.073243647234679, + "grad_norm": 0.972971498966217, + "learning_rate": 0.0004908445440956652, + "loss": 3.5335, + "step": 59950 + }, + { + "epoch": 4.07358336730534, + "grad_norm": 0.8681070804595947, + "learning_rate": 0.0004908020790868325, + "loss": 3.4464, + "step": 59955 + }, + { + "epoch": 4.073923087376002, + "grad_norm": 1.008081316947937, + "learning_rate": 0.0004907596140779997, + "loss": 3.3947, + "step": 59960 + }, + { + "epoch": 4.074262807446664, + "grad_norm": 0.9712215065956116, + "learning_rate": 0.0004907171490691671, + "loss": 3.1869, + "step": 59965 + }, + { + "epoch": 4.074602527517325, + "grad_norm": 0.7963645458221436, + "learning_rate": 0.0004906746840603343, + "loss": 3.4136, + "step": 59970 + }, + { + "epoch": 4.0749422475879875, + "grad_norm": 0.9250348210334778, + "learning_rate": 0.0004906322190515016, + "loss": 3.6041, + "step": 59975 + }, + { + "epoch": 4.07528196765865, + "grad_norm": 0.8094295263290405, + "learning_rate": 0.0004905897540426688, + "loss": 3.5135, + "step": 59980 + }, + { + "epoch": 4.075621687729311, + "grad_norm": 0.7919937372207642, + "learning_rate": 0.0004905472890338361, + "loss": 3.3237, + "step": 59985 + }, + { + "epoch": 4.075961407799973, + "grad_norm": 0.862321138381958, + "learning_rate": 0.0004905048240250034, + "loss": 3.3792, + "step": 59990 + }, + { + "epoch": 4.076301127870635, + "grad_norm": 1.0237566232681274, + "learning_rate": 0.0004904623590161707, + "loss": 3.2765, + "step": 59995 + }, + { + "epoch": 4.076640847941296, + "grad_norm": 0.9160123467445374, + "learning_rate": 0.000490419894007338, + "loss": 3.5139, + "step": 60000 + }, + { + "epoch": 4.076980568011958, + "grad_norm": 0.8233296275138855, + "learning_rate": 0.0004903774289985052, + "loss": 3.4318, + "step": 60005 + }, + { + "epoch": 4.07732028808262, + "grad_norm": 0.8600610494613647, + "learning_rate": 0.0004903349639896725, + "loss": 3.4327, + "step": 60010 + }, + { + "epoch": 4.0776600081532814, + "grad_norm": 1.3831515312194824, + "learning_rate": 0.0004902924989808397, + "loss": 3.3522, + "step": 60015 + }, + { + "epoch": 4.0779997282239435, + "grad_norm": 1.198341727256775, + "learning_rate": 0.0004902500339720071, + "loss": 3.4329, + "step": 60020 + }, + { + "epoch": 4.078339448294606, + "grad_norm": 0.8583764433860779, + "learning_rate": 0.0004902075689631744, + "loss": 3.3961, + "step": 60025 + }, + { + "epoch": 4.078679168365267, + "grad_norm": 0.7868750691413879, + "learning_rate": 0.0004901651039543416, + "loss": 3.4684, + "step": 60030 + }, + { + "epoch": 4.079018888435929, + "grad_norm": 0.8584799766540527, + "learning_rate": 0.0004901226389455089, + "loss": 3.4652, + "step": 60035 + }, + { + "epoch": 4.079358608506591, + "grad_norm": 1.0132217407226562, + "learning_rate": 0.0004900801739366762, + "loss": 3.5187, + "step": 60040 + }, + { + "epoch": 4.079698328577252, + "grad_norm": 1.039062738418579, + "learning_rate": 0.0004900377089278434, + "loss": 3.4672, + "step": 60045 + }, + { + "epoch": 4.080038048647914, + "grad_norm": 0.8012052774429321, + "learning_rate": 0.0004899952439190108, + "loss": 3.4535, + "step": 60050 + }, + { + "epoch": 4.080377768718576, + "grad_norm": 0.8432538509368896, + "learning_rate": 0.000489952778910178, + "loss": 3.5628, + "step": 60055 + }, + { + "epoch": 4.0807174887892375, + "grad_norm": 0.7210903763771057, + "learning_rate": 0.0004899103139013453, + "loss": 3.5153, + "step": 60060 + }, + { + "epoch": 4.0810572088598995, + "grad_norm": 0.8559758067131042, + "learning_rate": 0.0004898678488925125, + "loss": 3.3882, + "step": 60065 + }, + { + "epoch": 4.081396928930562, + "grad_norm": 1.042021632194519, + "learning_rate": 0.0004898253838836799, + "loss": 3.3883, + "step": 60070 + }, + { + "epoch": 4.081736649001223, + "grad_norm": 1.1754270792007446, + "learning_rate": 0.0004897829188748471, + "loss": 3.4268, + "step": 60075 + }, + { + "epoch": 4.082076369071885, + "grad_norm": 1.0656787157058716, + "learning_rate": 0.0004897404538660144, + "loss": 3.4968, + "step": 60080 + }, + { + "epoch": 4.082416089142547, + "grad_norm": 0.8314180970191956, + "learning_rate": 0.0004896979888571817, + "loss": 3.5941, + "step": 60085 + }, + { + "epoch": 4.082755809213208, + "grad_norm": 0.8209211826324463, + "learning_rate": 0.000489655523848349, + "loss": 3.6653, + "step": 60090 + }, + { + "epoch": 4.08309552928387, + "grad_norm": 1.1467338800430298, + "learning_rate": 0.0004896130588395162, + "loss": 3.933, + "step": 60095 + }, + { + "epoch": 4.083435249354532, + "grad_norm": 0.8042275905609131, + "learning_rate": 0.0004895705938306836, + "loss": 3.3902, + "step": 60100 + }, + { + "epoch": 4.0837749694251935, + "grad_norm": 0.8383256793022156, + "learning_rate": 0.0004895281288218508, + "loss": 3.2632, + "step": 60105 + }, + { + "epoch": 4.0841146894958555, + "grad_norm": 0.7507006525993347, + "learning_rate": 0.000489485663813018, + "loss": 3.5874, + "step": 60110 + }, + { + "epoch": 4.084454409566518, + "grad_norm": 0.8265739679336548, + "learning_rate": 0.0004894431988041853, + "loss": 3.6833, + "step": 60115 + }, + { + "epoch": 4.084794129637179, + "grad_norm": 0.7642436027526855, + "learning_rate": 0.0004894007337953527, + "loss": 3.4768, + "step": 60120 + }, + { + "epoch": 4.085133849707841, + "grad_norm": 0.9462882876396179, + "learning_rate": 0.0004893582687865199, + "loss": 3.6035, + "step": 60125 + }, + { + "epoch": 4.085473569778502, + "grad_norm": 0.9995160102844238, + "learning_rate": 0.0004893158037776872, + "loss": 3.4231, + "step": 60130 + }, + { + "epoch": 4.085813289849164, + "grad_norm": 1.0521057844161987, + "learning_rate": 0.0004892733387688545, + "loss": 3.7853, + "step": 60135 + }, + { + "epoch": 4.086153009919826, + "grad_norm": 1.0439428091049194, + "learning_rate": 0.0004892308737600217, + "loss": 3.3278, + "step": 60140 + }, + { + "epoch": 4.086492729990487, + "grad_norm": 0.8915185928344727, + "learning_rate": 0.000489188408751189, + "loss": 3.5125, + "step": 60145 + }, + { + "epoch": 4.0868324500611495, + "grad_norm": 0.7876999974250793, + "learning_rate": 0.0004891459437423564, + "loss": 3.5396, + "step": 60150 + }, + { + "epoch": 4.0871721701318116, + "grad_norm": 0.6847761273384094, + "learning_rate": 0.0004891034787335236, + "loss": 3.7642, + "step": 60155 + }, + { + "epoch": 4.087511890202473, + "grad_norm": 1.1301612854003906, + "learning_rate": 0.0004890610137246908, + "loss": 3.5201, + "step": 60160 + }, + { + "epoch": 4.087851610273135, + "grad_norm": 0.8413057327270508, + "learning_rate": 0.0004890185487158582, + "loss": 3.5959, + "step": 60165 + }, + { + "epoch": 4.088191330343797, + "grad_norm": 0.9388220906257629, + "learning_rate": 0.0004889760837070254, + "loss": 3.3561, + "step": 60170 + }, + { + "epoch": 4.088531050414458, + "grad_norm": 1.0634440183639526, + "learning_rate": 0.0004889336186981927, + "loss": 3.6321, + "step": 60175 + }, + { + "epoch": 4.08887077048512, + "grad_norm": 0.8490501642227173, + "learning_rate": 0.00048889115368936, + "loss": 3.7104, + "step": 60180 + }, + { + "epoch": 4.089210490555782, + "grad_norm": 0.8806116580963135, + "learning_rate": 0.0004888486886805273, + "loss": 3.2592, + "step": 60185 + }, + { + "epoch": 4.089550210626443, + "grad_norm": 0.9158693552017212, + "learning_rate": 0.0004888062236716945, + "loss": 3.5069, + "step": 60190 + }, + { + "epoch": 4.0898899306971055, + "grad_norm": 0.9916989803314209, + "learning_rate": 0.0004887637586628618, + "loss": 3.7523, + "step": 60195 + }, + { + "epoch": 4.090229650767768, + "grad_norm": 0.7990747690200806, + "learning_rate": 0.0004887212936540291, + "loss": 3.4108, + "step": 60200 + }, + { + "epoch": 4.090569370838429, + "grad_norm": 1.6441184282302856, + "learning_rate": 0.0004886788286451964, + "loss": 3.686, + "step": 60205 + }, + { + "epoch": 4.090909090909091, + "grad_norm": 0.824474036693573, + "learning_rate": 0.0004886363636363636, + "loss": 3.352, + "step": 60210 + }, + { + "epoch": 4.091248810979753, + "grad_norm": 0.8636743426322937, + "learning_rate": 0.000488593898627531, + "loss": 3.5884, + "step": 60215 + }, + { + "epoch": 4.091588531050414, + "grad_norm": 0.9154664874076843, + "learning_rate": 0.0004885514336186982, + "loss": 3.4228, + "step": 60220 + }, + { + "epoch": 4.091928251121076, + "grad_norm": 0.9216048121452332, + "learning_rate": 0.0004885089686098655, + "loss": 3.6307, + "step": 60225 + }, + { + "epoch": 4.092267971191738, + "grad_norm": 0.7878738641738892, + "learning_rate": 0.0004884665036010327, + "loss": 3.6046, + "step": 60230 + }, + { + "epoch": 4.092607691262399, + "grad_norm": 0.7026242613792419, + "learning_rate": 0.0004884240385922001, + "loss": 3.3199, + "step": 60235 + }, + { + "epoch": 4.0929474113330615, + "grad_norm": 1.1400413513183594, + "learning_rate": 0.0004883815735833673, + "loss": 3.262, + "step": 60240 + }, + { + "epoch": 4.093287131403724, + "grad_norm": 1.0244392156600952, + "learning_rate": 0.0004883391085745346, + "loss": 3.4037, + "step": 60245 + }, + { + "epoch": 4.093626851474385, + "grad_norm": 0.9020297527313232, + "learning_rate": 0.0004882966435657019, + "loss": 3.2456, + "step": 60250 + }, + { + "epoch": 4.093966571545047, + "grad_norm": 0.8943917751312256, + "learning_rate": 0.00048825417855686914, + "loss": 3.4998, + "step": 60255 + }, + { + "epoch": 4.094306291615709, + "grad_norm": 1.0855733156204224, + "learning_rate": 0.00048821171354803647, + "loss": 3.5796, + "step": 60260 + }, + { + "epoch": 4.09464601168637, + "grad_norm": 0.8314254283905029, + "learning_rate": 0.0004881692485392037, + "loss": 3.6525, + "step": 60265 + }, + { + "epoch": 4.094985731757032, + "grad_norm": 1.0861260890960693, + "learning_rate": 0.000488126783530371, + "loss": 3.3879, + "step": 60270 + }, + { + "epoch": 4.095325451827694, + "grad_norm": 0.8492047190666199, + "learning_rate": 0.00048808431852153826, + "loss": 3.3807, + "step": 60275 + }, + { + "epoch": 4.095665171898355, + "grad_norm": 0.851253092288971, + "learning_rate": 0.00048804185351270554, + "loss": 3.4912, + "step": 60280 + }, + { + "epoch": 4.0960048919690175, + "grad_norm": 0.7417199611663818, + "learning_rate": 0.0004879993885038728, + "loss": 3.3065, + "step": 60285 + }, + { + "epoch": 4.09634461203968, + "grad_norm": 0.9336373805999756, + "learning_rate": 0.0004879569234950401, + "loss": 3.439, + "step": 60290 + }, + { + "epoch": 4.096684332110341, + "grad_norm": 1.1025924682617188, + "learning_rate": 0.0004879144584862074, + "loss": 3.4208, + "step": 60295 + }, + { + "epoch": 4.097024052181003, + "grad_norm": 0.8482956886291504, + "learning_rate": 0.00048787199347737466, + "loss": 3.5746, + "step": 60300 + }, + { + "epoch": 4.097363772251665, + "grad_norm": 0.7988093495368958, + "learning_rate": 0.00048782952846854194, + "loss": 3.6002, + "step": 60305 + }, + { + "epoch": 4.097703492322326, + "grad_norm": 0.760061502456665, + "learning_rate": 0.00048778706345970916, + "loss": 3.2685, + "step": 60310 + }, + { + "epoch": 4.098043212392988, + "grad_norm": 0.8115400075912476, + "learning_rate": 0.0004877445984508765, + "loss": 3.5794, + "step": 60315 + }, + { + "epoch": 4.09838293246365, + "grad_norm": 0.8011322617530823, + "learning_rate": 0.0004877021334420438, + "loss": 3.354, + "step": 60320 + }, + { + "epoch": 4.0987226525343115, + "grad_norm": 2.6934351921081543, + "learning_rate": 0.000487659668433211, + "loss": 3.5544, + "step": 60325 + }, + { + "epoch": 4.0990623726049735, + "grad_norm": 0.8311012983322144, + "learning_rate": 0.00048761720342437834, + "loss": 3.4974, + "step": 60330 + }, + { + "epoch": 4.099402092675636, + "grad_norm": 0.9006627798080444, + "learning_rate": 0.0004875747384155456, + "loss": 3.6025, + "step": 60335 + }, + { + "epoch": 4.099741812746297, + "grad_norm": 0.9660932421684265, + "learning_rate": 0.00048753227340671284, + "loss": 3.4796, + "step": 60340 + }, + { + "epoch": 4.100081532816959, + "grad_norm": 0.825990617275238, + "learning_rate": 0.0004874898083978802, + "loss": 3.5355, + "step": 60345 + }, + { + "epoch": 4.100421252887621, + "grad_norm": 0.765449047088623, + "learning_rate": 0.00048744734338904746, + "loss": 3.6171, + "step": 60350 + }, + { + "epoch": 4.100760972958282, + "grad_norm": 0.6704328656196594, + "learning_rate": 0.0004874048783802147, + "loss": 3.544, + "step": 60355 + }, + { + "epoch": 4.101100693028944, + "grad_norm": 0.9004793167114258, + "learning_rate": 0.00048736241337138197, + "loss": 3.5444, + "step": 60360 + }, + { + "epoch": 4.101440413099606, + "grad_norm": 0.9138051271438599, + "learning_rate": 0.0004873199483625493, + "loss": 3.5035, + "step": 60365 + }, + { + "epoch": 4.1017801331702675, + "grad_norm": 0.9839326739311218, + "learning_rate": 0.0004872774833537165, + "loss": 3.4922, + "step": 60370 + }, + { + "epoch": 4.1021198532409295, + "grad_norm": 0.9757943749427795, + "learning_rate": 0.0004872350183448838, + "loss": 3.4791, + "step": 60375 + }, + { + "epoch": 4.102459573311592, + "grad_norm": 0.7705000042915344, + "learning_rate": 0.00048719255333605114, + "loss": 3.6002, + "step": 60380 + }, + { + "epoch": 4.102799293382253, + "grad_norm": 0.862695574760437, + "learning_rate": 0.00048715008832721837, + "loss": 3.4593, + "step": 60385 + }, + { + "epoch": 4.103139013452915, + "grad_norm": 0.8168323636054993, + "learning_rate": 0.00048710762331838565, + "loss": 3.413, + "step": 60390 + }, + { + "epoch": 4.103478733523577, + "grad_norm": 0.7711490988731384, + "learning_rate": 0.0004870651583095529, + "loss": 3.3285, + "step": 60395 + }, + { + "epoch": 4.103818453594238, + "grad_norm": 0.8221679329872131, + "learning_rate": 0.00048702269330072026, + "loss": 3.3133, + "step": 60400 + }, + { + "epoch": 4.1041581736649, + "grad_norm": 1.0641248226165771, + "learning_rate": 0.0004869802282918875, + "loss": 3.4725, + "step": 60405 + }, + { + "epoch": 4.104497893735562, + "grad_norm": 0.8830304145812988, + "learning_rate": 0.00048693776328305477, + "loss": 3.4793, + "step": 60410 + }, + { + "epoch": 4.1048376138062235, + "grad_norm": 0.9198535680770874, + "learning_rate": 0.0004868952982742221, + "loss": 3.4362, + "step": 60415 + }, + { + "epoch": 4.1051773338768855, + "grad_norm": 1.0214260816574097, + "learning_rate": 0.0004868528332653893, + "loss": 3.5625, + "step": 60420 + }, + { + "epoch": 4.105517053947548, + "grad_norm": 0.9782505035400391, + "learning_rate": 0.0004868103682565566, + "loss": 3.6029, + "step": 60425 + }, + { + "epoch": 4.105856774018209, + "grad_norm": 0.815485954284668, + "learning_rate": 0.0004867679032477239, + "loss": 3.4065, + "step": 60430 + }, + { + "epoch": 4.106196494088871, + "grad_norm": 0.8695691227912903, + "learning_rate": 0.00048672543823889117, + "loss": 3.1953, + "step": 60435 + }, + { + "epoch": 4.106536214159533, + "grad_norm": 0.9137519598007202, + "learning_rate": 0.00048668297323005845, + "loss": 3.5601, + "step": 60440 + }, + { + "epoch": 4.106875934230194, + "grad_norm": 0.7799518704414368, + "learning_rate": 0.0004866405082212257, + "loss": 3.5166, + "step": 60445 + }, + { + "epoch": 4.107215654300856, + "grad_norm": 0.9309407472610474, + "learning_rate": 0.000486598043212393, + "loss": 3.6408, + "step": 60450 + }, + { + "epoch": 4.107555374371518, + "grad_norm": 0.9623941779136658, + "learning_rate": 0.0004865555782035603, + "loss": 3.4567, + "step": 60455 + }, + { + "epoch": 4.1078950944421795, + "grad_norm": 0.7922086119651794, + "learning_rate": 0.00048651311319472757, + "loss": 3.4271, + "step": 60460 + }, + { + "epoch": 4.108234814512842, + "grad_norm": 1.0140390396118164, + "learning_rate": 0.0004864706481858948, + "loss": 3.3857, + "step": 60465 + }, + { + "epoch": 4.108574534583504, + "grad_norm": 0.9296822547912598, + "learning_rate": 0.0004864281831770621, + "loss": 3.4689, + "step": 60470 + }, + { + "epoch": 4.108914254654165, + "grad_norm": 1.0194514989852905, + "learning_rate": 0.0004863857181682294, + "loss": 3.6606, + "step": 60475 + }, + { + "epoch": 4.109253974724827, + "grad_norm": 0.7608421444892883, + "learning_rate": 0.00048634325315939663, + "loss": 3.5576, + "step": 60480 + }, + { + "epoch": 4.109593694795488, + "grad_norm": 0.7357890605926514, + "learning_rate": 0.00048630078815056397, + "loss": 3.4112, + "step": 60485 + }, + { + "epoch": 4.10993341486615, + "grad_norm": 0.824417233467102, + "learning_rate": 0.00048625832314173125, + "loss": 3.7389, + "step": 60490 + }, + { + "epoch": 4.110273134936812, + "grad_norm": 1.1876016855239868, + "learning_rate": 0.0004862158581328985, + "loss": 3.4617, + "step": 60495 + }, + { + "epoch": 4.110612855007473, + "grad_norm": 0.923630952835083, + "learning_rate": 0.00048617339312406575, + "loss": 3.1587, + "step": 60500 + }, + { + "epoch": 4.1109525750781355, + "grad_norm": 0.8195562362670898, + "learning_rate": 0.0004861309281152331, + "loss": 3.4922, + "step": 60505 + }, + { + "epoch": 4.111292295148798, + "grad_norm": 0.9582125544548035, + "learning_rate": 0.0004860884631064003, + "loss": 3.5156, + "step": 60510 + }, + { + "epoch": 4.111632015219459, + "grad_norm": 0.7129930257797241, + "learning_rate": 0.0004860459980975676, + "loss": 3.5218, + "step": 60515 + }, + { + "epoch": 4.111971735290121, + "grad_norm": 1.013438105583191, + "learning_rate": 0.00048600353308873493, + "loss": 3.4358, + "step": 60520 + }, + { + "epoch": 4.112311455360783, + "grad_norm": 1.8352806568145752, + "learning_rate": 0.00048596106807990215, + "loss": 3.4366, + "step": 60525 + }, + { + "epoch": 4.112651175431444, + "grad_norm": 1.054357647895813, + "learning_rate": 0.00048591860307106943, + "loss": 3.3337, + "step": 60530 + }, + { + "epoch": 4.112990895502106, + "grad_norm": 5.102238655090332, + "learning_rate": 0.0004858761380622367, + "loss": 3.2135, + "step": 60535 + }, + { + "epoch": 4.113330615572768, + "grad_norm": 0.9283230900764465, + "learning_rate": 0.000485833673053404, + "loss": 3.2923, + "step": 60540 + }, + { + "epoch": 4.113670335643429, + "grad_norm": 0.9232462644577026, + "learning_rate": 0.0004857912080445713, + "loss": 3.5495, + "step": 60545 + }, + { + "epoch": 4.1140100557140915, + "grad_norm": 0.851104199886322, + "learning_rate": 0.00048574874303573855, + "loss": 3.3081, + "step": 60550 + }, + { + "epoch": 4.114349775784754, + "grad_norm": 0.8237915635108948, + "learning_rate": 0.00048570627802690583, + "loss": 3.3159, + "step": 60555 + }, + { + "epoch": 4.114689495855415, + "grad_norm": 1.305794596672058, + "learning_rate": 0.0004856638130180731, + "loss": 3.3933, + "step": 60560 + }, + { + "epoch": 4.115029215926077, + "grad_norm": 0.8856663703918457, + "learning_rate": 0.0004856213480092404, + "loss": 3.4987, + "step": 60565 + }, + { + "epoch": 4.115368935996739, + "grad_norm": 1.0998294353485107, + "learning_rate": 0.0004855788830004077, + "loss": 3.4094, + "step": 60570 + }, + { + "epoch": 4.1157086560674, + "grad_norm": 1.2072299718856812, + "learning_rate": 0.00048553641799157495, + "loss": 3.1834, + "step": 60575 + }, + { + "epoch": 4.116048376138062, + "grad_norm": 1.1457629203796387, + "learning_rate": 0.00048549395298274223, + "loss": 3.7113, + "step": 60580 + }, + { + "epoch": 4.116388096208724, + "grad_norm": 0.7134578824043274, + "learning_rate": 0.0004854514879739095, + "loss": 3.2979, + "step": 60585 + }, + { + "epoch": 4.116727816279385, + "grad_norm": 0.8877643346786499, + "learning_rate": 0.0004854090229650768, + "loss": 3.4389, + "step": 60590 + }, + { + "epoch": 4.1170675363500475, + "grad_norm": 1.1142362356185913, + "learning_rate": 0.0004853665579562441, + "loss": 3.5986, + "step": 60595 + }, + { + "epoch": 4.11740725642071, + "grad_norm": 0.8856332898139954, + "learning_rate": 0.00048532409294741135, + "loss": 3.7596, + "step": 60600 + }, + { + "epoch": 4.117746976491371, + "grad_norm": 0.9383743405342102, + "learning_rate": 0.0004852816279385786, + "loss": 3.4659, + "step": 60605 + }, + { + "epoch": 4.118086696562033, + "grad_norm": 1.0099022388458252, + "learning_rate": 0.0004852391629297459, + "loss": 3.296, + "step": 60610 + }, + { + "epoch": 4.118426416632695, + "grad_norm": 1.203545093536377, + "learning_rate": 0.0004851966979209132, + "loss": 3.366, + "step": 60615 + }, + { + "epoch": 4.118766136703356, + "grad_norm": 0.8963096141815186, + "learning_rate": 0.0004851542329120804, + "loss": 3.5149, + "step": 60620 + }, + { + "epoch": 4.119105856774018, + "grad_norm": 0.7907694578170776, + "learning_rate": 0.00048511176790324775, + "loss": 3.5806, + "step": 60625 + }, + { + "epoch": 4.11944557684468, + "grad_norm": 0.7245348691940308, + "learning_rate": 0.00048506930289441504, + "loss": 3.2355, + "step": 60630 + }, + { + "epoch": 4.1197852969153415, + "grad_norm": 0.8068969249725342, + "learning_rate": 0.00048502683788558226, + "loss": 3.7529, + "step": 60635 + }, + { + "epoch": 4.1201250169860035, + "grad_norm": 0.8081486821174622, + "learning_rate": 0.0004849843728767496, + "loss": 3.3521, + "step": 60640 + }, + { + "epoch": 4.120464737056666, + "grad_norm": 0.8287180662155151, + "learning_rate": 0.0004849419078679169, + "loss": 3.5351, + "step": 60645 + }, + { + "epoch": 4.120804457127327, + "grad_norm": 0.8903965950012207, + "learning_rate": 0.0004848994428590841, + "loss": 3.6391, + "step": 60650 + }, + { + "epoch": 4.121144177197989, + "grad_norm": 0.8040282726287842, + "learning_rate": 0.0004848569778502514, + "loss": 3.569, + "step": 60655 + }, + { + "epoch": 4.121483897268651, + "grad_norm": 0.6821370720863342, + "learning_rate": 0.0004848145128414187, + "loss": 3.3349, + "step": 60660 + }, + { + "epoch": 4.121823617339312, + "grad_norm": 1.0014156103134155, + "learning_rate": 0.00048477204783258594, + "loss": 3.4021, + "step": 60665 + }, + { + "epoch": 4.122163337409974, + "grad_norm": 1.094258427619934, + "learning_rate": 0.0004847295828237532, + "loss": 3.3974, + "step": 60670 + }, + { + "epoch": 4.122503057480636, + "grad_norm": 1.1385911703109741, + "learning_rate": 0.00048468711781492056, + "loss": 3.3088, + "step": 60675 + }, + { + "epoch": 4.1228427775512975, + "grad_norm": 1.1799378395080566, + "learning_rate": 0.0004846446528060878, + "loss": 3.2837, + "step": 60680 + }, + { + "epoch": 4.1231824976219595, + "grad_norm": 1.2915782928466797, + "learning_rate": 0.00048460218779725506, + "loss": 3.4061, + "step": 60685 + }, + { + "epoch": 4.123522217692622, + "grad_norm": 0.7857838869094849, + "learning_rate": 0.00048455972278842234, + "loss": 3.9054, + "step": 60690 + }, + { + "epoch": 4.123861937763283, + "grad_norm": 0.9258236885070801, + "learning_rate": 0.0004845172577795896, + "loss": 3.1958, + "step": 60695 + }, + { + "epoch": 4.124201657833945, + "grad_norm": 0.9963165521621704, + "learning_rate": 0.0004844747927707569, + "loss": 3.1761, + "step": 60700 + }, + { + "epoch": 4.124541377904607, + "grad_norm": 1.0547771453857422, + "learning_rate": 0.0004844323277619242, + "loss": 3.441, + "step": 60705 + }, + { + "epoch": 4.124881097975268, + "grad_norm": 0.9425568580627441, + "learning_rate": 0.00048438986275309146, + "loss": 3.3796, + "step": 60710 + }, + { + "epoch": 4.12522081804593, + "grad_norm": 0.84581059217453, + "learning_rate": 0.00048434739774425874, + "loss": 3.462, + "step": 60715 + }, + { + "epoch": 4.125560538116592, + "grad_norm": 0.7780881524085999, + "learning_rate": 0.000484304932735426, + "loss": 3.3941, + "step": 60720 + }, + { + "epoch": 4.1259002581872535, + "grad_norm": 1.4650455713272095, + "learning_rate": 0.00048426246772659325, + "loss": 3.1523, + "step": 60725 + }, + { + "epoch": 4.1262399782579156, + "grad_norm": 1.6117115020751953, + "learning_rate": 0.0004842200027177606, + "loss": 3.4549, + "step": 60730 + }, + { + "epoch": 4.126579698328578, + "grad_norm": 0.9062334895133972, + "learning_rate": 0.00048417753770892786, + "loss": 3.4221, + "step": 60735 + }, + { + "epoch": 4.126919418399239, + "grad_norm": 0.9595715999603271, + "learning_rate": 0.00048413507270009514, + "loss": 3.3914, + "step": 60740 + }, + { + "epoch": 4.127259138469901, + "grad_norm": 1.1753897666931152, + "learning_rate": 0.0004840926076912624, + "loss": 3.1208, + "step": 60745 + }, + { + "epoch": 4.127598858540563, + "grad_norm": 0.9783079624176025, + "learning_rate": 0.0004840501426824297, + "loss": 3.7187, + "step": 60750 + }, + { + "epoch": 4.127938578611224, + "grad_norm": 0.7220286726951599, + "learning_rate": 0.000484007677673597, + "loss": 3.5384, + "step": 60755 + }, + { + "epoch": 4.128278298681886, + "grad_norm": 0.9653136730194092, + "learning_rate": 0.0004839652126647642, + "loss": 3.5373, + "step": 60760 + }, + { + "epoch": 4.128618018752548, + "grad_norm": 0.8879910111427307, + "learning_rate": 0.00048392274765593154, + "loss": 3.3493, + "step": 60765 + }, + { + "epoch": 4.1289577388232095, + "grad_norm": 1.165257453918457, + "learning_rate": 0.0004838802826470988, + "loss": 3.5048, + "step": 60770 + }, + { + "epoch": 4.129297458893872, + "grad_norm": 0.7714213132858276, + "learning_rate": 0.00048383781763826605, + "loss": 3.3744, + "step": 60775 + }, + { + "epoch": 4.129637178964534, + "grad_norm": 0.9698021411895752, + "learning_rate": 0.0004837953526294334, + "loss": 3.506, + "step": 60780 + }, + { + "epoch": 4.129976899035195, + "grad_norm": 0.6581893563270569, + "learning_rate": 0.00048375288762060066, + "loss": 3.5473, + "step": 60785 + }, + { + "epoch": 4.130316619105857, + "grad_norm": 1.1890594959259033, + "learning_rate": 0.0004837104226117679, + "loss": 3.5606, + "step": 60790 + }, + { + "epoch": 4.130656339176519, + "grad_norm": 0.8945116996765137, + "learning_rate": 0.00048366795760293517, + "loss": 3.4167, + "step": 60795 + }, + { + "epoch": 4.13099605924718, + "grad_norm": 0.9084517359733582, + "learning_rate": 0.0004836254925941025, + "loss": 3.3764, + "step": 60800 + }, + { + "epoch": 4.131335779317842, + "grad_norm": 0.8003615736961365, + "learning_rate": 0.00048358302758526973, + "loss": 3.2852, + "step": 60805 + }, + { + "epoch": 4.131675499388503, + "grad_norm": 1.0807470083236694, + "learning_rate": 0.000483540562576437, + "loss": 3.563, + "step": 60810 + }, + { + "epoch": 4.1320152194591655, + "grad_norm": 1.020410418510437, + "learning_rate": 0.00048349809756760434, + "loss": 3.3261, + "step": 60815 + }, + { + "epoch": 4.132354939529828, + "grad_norm": 0.8012077808380127, + "learning_rate": 0.00048345563255877157, + "loss": 3.5699, + "step": 60820 + }, + { + "epoch": 4.132694659600489, + "grad_norm": 0.852891743183136, + "learning_rate": 0.00048341316754993885, + "loss": 3.4496, + "step": 60825 + }, + { + "epoch": 4.133034379671151, + "grad_norm": 0.9984927773475647, + "learning_rate": 0.00048337070254110613, + "loss": 3.4363, + "step": 60830 + }, + { + "epoch": 4.133374099741813, + "grad_norm": 0.8366738557815552, + "learning_rate": 0.0004833282375322734, + "loss": 3.4055, + "step": 60835 + }, + { + "epoch": 4.133713819812474, + "grad_norm": 1.1941158771514893, + "learning_rate": 0.0004832857725234407, + "loss": 3.3566, + "step": 60840 + }, + { + "epoch": 4.134053539883136, + "grad_norm": 1.2291193008422852, + "learning_rate": 0.00048324330751460797, + "loss": 3.2872, + "step": 60845 + }, + { + "epoch": 4.134393259953798, + "grad_norm": 0.8321509957313538, + "learning_rate": 0.00048320084250577525, + "loss": 3.5308, + "step": 60850 + }, + { + "epoch": 4.134732980024459, + "grad_norm": 0.8231420516967773, + "learning_rate": 0.00048315837749694253, + "loss": 3.4741, + "step": 60855 + }, + { + "epoch": 4.1350727000951215, + "grad_norm": 1.0114644765853882, + "learning_rate": 0.0004831159124881098, + "loss": 3.5014, + "step": 60860 + }, + { + "epoch": 4.135412420165784, + "grad_norm": 0.8152646422386169, + "learning_rate": 0.00048307344747927704, + "loss": 3.5738, + "step": 60865 + }, + { + "epoch": 4.135752140236445, + "grad_norm": 0.8439964056015015, + "learning_rate": 0.00048303098247044437, + "loss": 3.4225, + "step": 60870 + }, + { + "epoch": 4.136091860307107, + "grad_norm": 0.7128999829292297, + "learning_rate": 0.00048298851746161165, + "loss": 3.457, + "step": 60875 + }, + { + "epoch": 4.136431580377769, + "grad_norm": 0.8958193063735962, + "learning_rate": 0.0004829460524527789, + "loss": 3.4807, + "step": 60880 + }, + { + "epoch": 4.13677130044843, + "grad_norm": 0.9176084399223328, + "learning_rate": 0.0004829035874439462, + "loss": 3.4527, + "step": 60885 + }, + { + "epoch": 4.137111020519092, + "grad_norm": 0.9475146532058716, + "learning_rate": 0.0004828611224351135, + "loss": 3.5877, + "step": 60890 + }, + { + "epoch": 4.137450740589754, + "grad_norm": 1.0759168863296509, + "learning_rate": 0.0004828186574262807, + "loss": 3.4587, + "step": 60895 + }, + { + "epoch": 4.1377904606604154, + "grad_norm": 0.8839792013168335, + "learning_rate": 0.000482776192417448, + "loss": 3.4785, + "step": 60900 + }, + { + "epoch": 4.1381301807310775, + "grad_norm": 0.9047330617904663, + "learning_rate": 0.00048273372740861533, + "loss": 3.4357, + "step": 60905 + }, + { + "epoch": 4.13846990080174, + "grad_norm": 0.9330630898475647, + "learning_rate": 0.0004826912623997826, + "loss": 3.5254, + "step": 60910 + }, + { + "epoch": 4.138809620872401, + "grad_norm": 0.9344311952590942, + "learning_rate": 0.00048264879739094984, + "loss": 3.6818, + "step": 60915 + }, + { + "epoch": 4.139149340943063, + "grad_norm": 1.0314078330993652, + "learning_rate": 0.00048260633238211717, + "loss": 3.39, + "step": 60920 + }, + { + "epoch": 4.139489061013725, + "grad_norm": 0.8636992573738098, + "learning_rate": 0.00048256386737328445, + "loss": 3.3487, + "step": 60925 + }, + { + "epoch": 4.139828781084386, + "grad_norm": 1.0891183614730835, + "learning_rate": 0.0004825214023644517, + "loss": 3.7929, + "step": 60930 + }, + { + "epoch": 4.140168501155048, + "grad_norm": 0.9468405246734619, + "learning_rate": 0.000482478937355619, + "loss": 3.4391, + "step": 60935 + }, + { + "epoch": 4.14050822122571, + "grad_norm": 0.7584990859031677, + "learning_rate": 0.0004824364723467863, + "loss": 3.4859, + "step": 60940 + }, + { + "epoch": 4.1408479412963715, + "grad_norm": 0.8876801133155823, + "learning_rate": 0.0004823940073379535, + "loss": 3.7726, + "step": 60945 + }, + { + "epoch": 4.1411876613670335, + "grad_norm": 0.7832413911819458, + "learning_rate": 0.0004823515423291208, + "loss": 3.6383, + "step": 60950 + }, + { + "epoch": 4.141527381437696, + "grad_norm": 0.8907681107521057, + "learning_rate": 0.00048230907732028813, + "loss": 3.1618, + "step": 60955 + }, + { + "epoch": 4.141867101508357, + "grad_norm": 0.8434287905693054, + "learning_rate": 0.00048226661231145536, + "loss": 3.3583, + "step": 60960 + }, + { + "epoch": 4.142206821579019, + "grad_norm": 0.9744192957878113, + "learning_rate": 0.00048222414730262264, + "loss": 3.429, + "step": 60965 + }, + { + "epoch": 4.142546541649681, + "grad_norm": 1.0301381349563599, + "learning_rate": 0.00048218168229378997, + "loss": 3.5584, + "step": 60970 + }, + { + "epoch": 4.142886261720342, + "grad_norm": 0.767314612865448, + "learning_rate": 0.0004821392172849572, + "loss": 3.4044, + "step": 60975 + }, + { + "epoch": 4.143225981791004, + "grad_norm": 0.7550290822982788, + "learning_rate": 0.0004820967522761245, + "loss": 3.3663, + "step": 60980 + }, + { + "epoch": 4.143565701861666, + "grad_norm": 0.9915202260017395, + "learning_rate": 0.00048205428726729176, + "loss": 3.6109, + "step": 60985 + }, + { + "epoch": 4.1439054219323275, + "grad_norm": 1.026853322982788, + "learning_rate": 0.00048201182225845904, + "loss": 3.3546, + "step": 60990 + }, + { + "epoch": 4.1442451420029895, + "grad_norm": 0.8758509159088135, + "learning_rate": 0.0004819693572496263, + "loss": 3.3082, + "step": 60995 + }, + { + "epoch": 4.144584862073652, + "grad_norm": 0.918562114238739, + "learning_rate": 0.0004819268922407936, + "loss": 3.2879, + "step": 61000 + }, + { + "epoch": 4.144924582144313, + "grad_norm": 0.858489453792572, + "learning_rate": 0.0004818844272319609, + "loss": 3.3901, + "step": 61005 + }, + { + "epoch": 4.145264302214975, + "grad_norm": 0.783415675163269, + "learning_rate": 0.00048184196222312816, + "loss": 3.5607, + "step": 61010 + }, + { + "epoch": 4.145604022285637, + "grad_norm": 0.8915145993232727, + "learning_rate": 0.00048179949721429544, + "loss": 3.5051, + "step": 61015 + }, + { + "epoch": 4.145943742356298, + "grad_norm": 0.8396680951118469, + "learning_rate": 0.00048175703220546266, + "loss": 3.4385, + "step": 61020 + }, + { + "epoch": 4.14628346242696, + "grad_norm": 0.9635047912597656, + "learning_rate": 0.00048171456719663, + "loss": 3.5492, + "step": 61025 + }, + { + "epoch": 4.146623182497622, + "grad_norm": 0.8289498090744019, + "learning_rate": 0.0004816721021877973, + "loss": 3.5478, + "step": 61030 + }, + { + "epoch": 4.1469629025682835, + "grad_norm": 0.8306374549865723, + "learning_rate": 0.0004816296371789645, + "loss": 3.4731, + "step": 61035 + }, + { + "epoch": 4.147302622638946, + "grad_norm": 0.7349468469619751, + "learning_rate": 0.00048158717217013184, + "loss": 3.5544, + "step": 61040 + }, + { + "epoch": 4.147642342709608, + "grad_norm": 0.9649933576583862, + "learning_rate": 0.0004815447071612991, + "loss": 3.7283, + "step": 61045 + }, + { + "epoch": 4.147982062780269, + "grad_norm": 1.2236928939819336, + "learning_rate": 0.00048150224215246634, + "loss": 3.4891, + "step": 61050 + }, + { + "epoch": 4.148321782850931, + "grad_norm": 0.7976474165916443, + "learning_rate": 0.0004814597771436336, + "loss": 3.5651, + "step": 61055 + }, + { + "epoch": 4.148661502921593, + "grad_norm": 0.7144026756286621, + "learning_rate": 0.00048141731213480096, + "loss": 3.4696, + "step": 61060 + }, + { + "epoch": 4.149001222992254, + "grad_norm": 0.8160803914070129, + "learning_rate": 0.0004813748471259682, + "loss": 3.4091, + "step": 61065 + }, + { + "epoch": 4.149340943062916, + "grad_norm": 0.8607046604156494, + "learning_rate": 0.00048133238211713546, + "loss": 3.5294, + "step": 61070 + }, + { + "epoch": 4.149680663133578, + "grad_norm": 0.9162683486938477, + "learning_rate": 0.0004812899171083028, + "loss": 3.4826, + "step": 61075 + }, + { + "epoch": 4.1500203832042395, + "grad_norm": 0.8390758633613586, + "learning_rate": 0.0004812474520994701, + "loss": 3.4112, + "step": 61080 + }, + { + "epoch": 4.150360103274902, + "grad_norm": 1.1057442426681519, + "learning_rate": 0.0004812049870906373, + "loss": 3.4551, + "step": 61085 + }, + { + "epoch": 4.150699823345564, + "grad_norm": 0.9776700735092163, + "learning_rate": 0.0004811625220818046, + "loss": 3.4702, + "step": 61090 + }, + { + "epoch": 4.151039543416225, + "grad_norm": 0.9008141756057739, + "learning_rate": 0.0004811200570729719, + "loss": 3.6891, + "step": 61095 + }, + { + "epoch": 4.151379263486887, + "grad_norm": 0.7931249141693115, + "learning_rate": 0.00048107759206413915, + "loss": 3.5026, + "step": 61100 + }, + { + "epoch": 4.151718983557549, + "grad_norm": 0.8162325620651245, + "learning_rate": 0.0004810351270553064, + "loss": 3.4009, + "step": 61105 + }, + { + "epoch": 4.15205870362821, + "grad_norm": 0.9525900483131409, + "learning_rate": 0.00048099266204647376, + "loss": 3.6679, + "step": 61110 + }, + { + "epoch": 4.152398423698872, + "grad_norm": 1.1448752880096436, + "learning_rate": 0.000480950197037641, + "loss": 3.5221, + "step": 61115 + }, + { + "epoch": 4.152738143769534, + "grad_norm": 0.7758772373199463, + "learning_rate": 0.00048090773202880827, + "loss": 3.4811, + "step": 61120 + }, + { + "epoch": 4.1530778638401955, + "grad_norm": 0.799312174320221, + "learning_rate": 0.00048086526701997555, + "loss": 3.5557, + "step": 61125 + }, + { + "epoch": 4.153417583910858, + "grad_norm": 0.8320074081420898, + "learning_rate": 0.0004808228020111428, + "loss": 3.374, + "step": 61130 + }, + { + "epoch": 4.15375730398152, + "grad_norm": 0.8839256167411804, + "learning_rate": 0.0004807803370023101, + "loss": 3.553, + "step": 61135 + }, + { + "epoch": 4.154097024052181, + "grad_norm": 1.1912697553634644, + "learning_rate": 0.0004807378719934774, + "loss": 3.5927, + "step": 61140 + }, + { + "epoch": 4.154436744122843, + "grad_norm": 0.8356027007102966, + "learning_rate": 0.00048069540698464467, + "loss": 3.4487, + "step": 61145 + }, + { + "epoch": 4.154776464193505, + "grad_norm": 1.07656991481781, + "learning_rate": 0.00048065294197581195, + "loss": 3.5596, + "step": 61150 + }, + { + "epoch": 4.155116184264166, + "grad_norm": 0.7973971962928772, + "learning_rate": 0.0004806104769669792, + "loss": 3.6639, + "step": 61155 + }, + { + "epoch": 4.155455904334828, + "grad_norm": 1.0467630624771118, + "learning_rate": 0.00048056801195814645, + "loss": 3.4463, + "step": 61160 + }, + { + "epoch": 4.15579562440549, + "grad_norm": 0.9411207437515259, + "learning_rate": 0.0004805255469493138, + "loss": 3.5244, + "step": 61165 + }, + { + "epoch": 4.1561353444761515, + "grad_norm": 0.8440948724746704, + "learning_rate": 0.00048048308194048107, + "loss": 3.713, + "step": 61170 + }, + { + "epoch": 4.156475064546814, + "grad_norm": 0.8739214539527893, + "learning_rate": 0.0004804406169316483, + "loss": 3.3699, + "step": 61175 + }, + { + "epoch": 4.156814784617475, + "grad_norm": 1.0458316802978516, + "learning_rate": 0.0004803981519228156, + "loss": 3.5153, + "step": 61180 + }, + { + "epoch": 4.157154504688137, + "grad_norm": 1.1161315441131592, + "learning_rate": 0.0004803556869139829, + "loss": 3.6033, + "step": 61185 + }, + { + "epoch": 4.157494224758799, + "grad_norm": 0.8622521758079529, + "learning_rate": 0.00048031322190515013, + "loss": 3.5284, + "step": 61190 + }, + { + "epoch": 4.15783394482946, + "grad_norm": 0.7702986598014832, + "learning_rate": 0.00048027075689631747, + "loss": 3.6047, + "step": 61195 + }, + { + "epoch": 4.158173664900122, + "grad_norm": 0.9935966730117798, + "learning_rate": 0.00048022829188748475, + "loss": 3.4371, + "step": 61200 + }, + { + "epoch": 4.158513384970784, + "grad_norm": 1.020129680633545, + "learning_rate": 0.00048018582687865197, + "loss": 3.2862, + "step": 61205 + }, + { + "epoch": 4.1588531050414455, + "grad_norm": 1.269012451171875, + "learning_rate": 0.00048014336186981925, + "loss": 3.8026, + "step": 61210 + }, + { + "epoch": 4.1591928251121075, + "grad_norm": 0.7302483916282654, + "learning_rate": 0.0004801008968609866, + "loss": 3.3462, + "step": 61215 + }, + { + "epoch": 4.15953254518277, + "grad_norm": 0.8843251466751099, + "learning_rate": 0.0004800584318521538, + "loss": 3.5796, + "step": 61220 + }, + { + "epoch": 4.159872265253431, + "grad_norm": 1.0964646339416504, + "learning_rate": 0.0004800159668433211, + "loss": 3.4576, + "step": 61225 + }, + { + "epoch": 4.160211985324093, + "grad_norm": 0.8815962076187134, + "learning_rate": 0.00047997350183448843, + "loss": 3.3914, + "step": 61230 + }, + { + "epoch": 4.160551705394755, + "grad_norm": 0.9033495187759399, + "learning_rate": 0.00047993103682565565, + "loss": 3.4816, + "step": 61235 + }, + { + "epoch": 4.160891425465416, + "grad_norm": 0.9238218069076538, + "learning_rate": 0.00047988857181682293, + "loss": 3.3832, + "step": 61240 + }, + { + "epoch": 4.161231145536078, + "grad_norm": 0.8261555433273315, + "learning_rate": 0.0004798461068079902, + "loss": 3.5648, + "step": 61245 + }, + { + "epoch": 4.16157086560674, + "grad_norm": 1.2196403741836548, + "learning_rate": 0.00047980364179915755, + "loss": 3.4318, + "step": 61250 + }, + { + "epoch": 4.1619105856774015, + "grad_norm": 1.1675587892532349, + "learning_rate": 0.0004797611767903248, + "loss": 3.5861, + "step": 61255 + }, + { + "epoch": 4.1622503057480635, + "grad_norm": 0.7549326419830322, + "learning_rate": 0.00047971871178149205, + "loss": 3.6748, + "step": 61260 + }, + { + "epoch": 4.162590025818726, + "grad_norm": 1.1471422910690308, + "learning_rate": 0.0004796762467726594, + "loss": 3.5278, + "step": 61265 + }, + { + "epoch": 4.162929745889387, + "grad_norm": 1.0092384815216064, + "learning_rate": 0.0004796337817638266, + "loss": 3.4407, + "step": 61270 + }, + { + "epoch": 4.163269465960049, + "grad_norm": 0.8209322690963745, + "learning_rate": 0.0004795913167549939, + "loss": 3.5374, + "step": 61275 + }, + { + "epoch": 4.163609186030711, + "grad_norm": 0.9471033215522766, + "learning_rate": 0.0004795488517461612, + "loss": 3.6338, + "step": 61280 + }, + { + "epoch": 4.163948906101372, + "grad_norm": 0.7848088145256042, + "learning_rate": 0.00047950638673732845, + "loss": 3.3755, + "step": 61285 + }, + { + "epoch": 4.164288626172034, + "grad_norm": 0.9488942623138428, + "learning_rate": 0.00047946392172849573, + "loss": 3.4213, + "step": 61290 + }, + { + "epoch": 4.164628346242696, + "grad_norm": 0.9325140714645386, + "learning_rate": 0.000479421456719663, + "loss": 3.4231, + "step": 61295 + }, + { + "epoch": 4.1649680663133575, + "grad_norm": 1.0085484981536865, + "learning_rate": 0.0004793789917108303, + "loss": 3.5084, + "step": 61300 + }, + { + "epoch": 4.1653077863840196, + "grad_norm": 0.809970498085022, + "learning_rate": 0.0004793365267019976, + "loss": 3.6421, + "step": 61305 + }, + { + "epoch": 4.165647506454682, + "grad_norm": 0.7918409705162048, + "learning_rate": 0.00047929406169316485, + "loss": 3.3706, + "step": 61310 + }, + { + "epoch": 4.165987226525343, + "grad_norm": 1.1134058237075806, + "learning_rate": 0.0004792515966843321, + "loss": 3.5831, + "step": 61315 + }, + { + "epoch": 4.166326946596005, + "grad_norm": 0.8777232766151428, + "learning_rate": 0.0004792091316754994, + "loss": 3.4488, + "step": 61320 + }, + { + "epoch": 4.166666666666667, + "grad_norm": 0.8292273879051208, + "learning_rate": 0.0004791666666666667, + "loss": 3.3655, + "step": 61325 + }, + { + "epoch": 4.167006386737328, + "grad_norm": 0.9717200994491577, + "learning_rate": 0.0004791242016578339, + "loss": 3.5134, + "step": 61330 + }, + { + "epoch": 4.16734610680799, + "grad_norm": 0.7630997896194458, + "learning_rate": 0.00047908173664900125, + "loss": 3.5761, + "step": 61335 + }, + { + "epoch": 4.167685826878652, + "grad_norm": 0.903893768787384, + "learning_rate": 0.00047903927164016853, + "loss": 3.7695, + "step": 61340 + }, + { + "epoch": 4.1680255469493135, + "grad_norm": 0.8937538266181946, + "learning_rate": 0.00047899680663133576, + "loss": 3.1996, + "step": 61345 + }, + { + "epoch": 4.168365267019976, + "grad_norm": 1.085886001586914, + "learning_rate": 0.00047895434162250304, + "loss": 3.3027, + "step": 61350 + }, + { + "epoch": 4.168704987090638, + "grad_norm": 0.880458414554596, + "learning_rate": 0.0004789118766136704, + "loss": 3.6383, + "step": 61355 + }, + { + "epoch": 4.169044707161299, + "grad_norm": 0.7507785558700562, + "learning_rate": 0.0004788694116048376, + "loss": 3.6534, + "step": 61360 + }, + { + "epoch": 4.169384427231961, + "grad_norm": 0.9068813920021057, + "learning_rate": 0.0004788269465960049, + "loss": 3.2842, + "step": 61365 + }, + { + "epoch": 4.169724147302623, + "grad_norm": 0.9297317266464233, + "learning_rate": 0.0004787844815871722, + "loss": 3.2945, + "step": 61370 + }, + { + "epoch": 4.170063867373284, + "grad_norm": 0.8585667610168457, + "learning_rate": 0.00047874201657833944, + "loss": 3.4027, + "step": 61375 + }, + { + "epoch": 4.170403587443946, + "grad_norm": 0.9907045364379883, + "learning_rate": 0.0004786995515695067, + "loss": 3.5582, + "step": 61380 + }, + { + "epoch": 4.170743307514608, + "grad_norm": 0.9121895432472229, + "learning_rate": 0.000478657086560674, + "loss": 3.5211, + "step": 61385 + }, + { + "epoch": 4.1710830275852695, + "grad_norm": 0.8772000670433044, + "learning_rate": 0.0004786146215518413, + "loss": 3.0921, + "step": 61390 + }, + { + "epoch": 4.171422747655932, + "grad_norm": 0.9559676647186279, + "learning_rate": 0.00047857215654300856, + "loss": 3.678, + "step": 61395 + }, + { + "epoch": 4.171762467726594, + "grad_norm": 1.0583782196044922, + "learning_rate": 0.00047852969153417584, + "loss": 3.4914, + "step": 61400 + }, + { + "epoch": 4.172102187797255, + "grad_norm": 0.756763756275177, + "learning_rate": 0.0004784872265253431, + "loss": 3.4113, + "step": 61405 + }, + { + "epoch": 4.172441907867917, + "grad_norm": 0.8362118601799011, + "learning_rate": 0.0004784447615165104, + "loss": 3.6134, + "step": 61410 + }, + { + "epoch": 4.172781627938579, + "grad_norm": 1.069839596748352, + "learning_rate": 0.0004784022965076777, + "loss": 3.4441, + "step": 61415 + }, + { + "epoch": 4.17312134800924, + "grad_norm": 0.9363768696784973, + "learning_rate": 0.00047835983149884496, + "loss": 3.5099, + "step": 61420 + }, + { + "epoch": 4.173461068079902, + "grad_norm": 0.9069871306419373, + "learning_rate": 0.00047831736649001224, + "loss": 3.4737, + "step": 61425 + }, + { + "epoch": 4.173800788150564, + "grad_norm": 0.8612582087516785, + "learning_rate": 0.0004782749014811795, + "loss": 3.5325, + "step": 61430 + }, + { + "epoch": 4.1741405082212255, + "grad_norm": 1.0577627420425415, + "learning_rate": 0.0004782324364723468, + "loss": 3.6395, + "step": 61435 + }, + { + "epoch": 4.174480228291888, + "grad_norm": 0.9363875985145569, + "learning_rate": 0.0004781899714635141, + "loss": 3.6046, + "step": 61440 + }, + { + "epoch": 4.17481994836255, + "grad_norm": 0.8154097199440002, + "learning_rate": 0.00047814750645468136, + "loss": 3.5131, + "step": 61445 + }, + { + "epoch": 4.175159668433211, + "grad_norm": 0.6793762445449829, + "learning_rate": 0.00047810504144584864, + "loss": 3.6875, + "step": 61450 + }, + { + "epoch": 4.175499388503873, + "grad_norm": 0.8323537707328796, + "learning_rate": 0.00047806257643701587, + "loss": 3.5121, + "step": 61455 + }, + { + "epoch": 4.175839108574535, + "grad_norm": 0.7416667342185974, + "learning_rate": 0.0004780201114281832, + "loss": 3.335, + "step": 61460 + }, + { + "epoch": 4.176178828645196, + "grad_norm": 1.047333002090454, + "learning_rate": 0.0004779776464193505, + "loss": 3.3923, + "step": 61465 + }, + { + "epoch": 4.176518548715858, + "grad_norm": 1.0007826089859009, + "learning_rate": 0.0004779351814105177, + "loss": 3.4439, + "step": 61470 + }, + { + "epoch": 4.17685826878652, + "grad_norm": 0.8062480092048645, + "learning_rate": 0.00047789271640168504, + "loss": 3.5497, + "step": 61475 + }, + { + "epoch": 4.1771979888571815, + "grad_norm": 0.7364413738250732, + "learning_rate": 0.0004778502513928523, + "loss": 3.6515, + "step": 61480 + }, + { + "epoch": 4.177537708927844, + "grad_norm": 1.056564211845398, + "learning_rate": 0.00047780778638401955, + "loss": 3.2767, + "step": 61485 + }, + { + "epoch": 4.177877428998505, + "grad_norm": 0.8792128562927246, + "learning_rate": 0.0004777653213751869, + "loss": 3.4981, + "step": 61490 + }, + { + "epoch": 4.178217149069167, + "grad_norm": 0.9576679468154907, + "learning_rate": 0.00047772285636635416, + "loss": 3.4957, + "step": 61495 + }, + { + "epoch": 4.178556869139829, + "grad_norm": 0.8643674850463867, + "learning_rate": 0.0004776803913575214, + "loss": 3.4386, + "step": 61500 + }, + { + "epoch": 4.17889658921049, + "grad_norm": 1.5490754842758179, + "learning_rate": 0.00047763792634868867, + "loss": 3.5235, + "step": 61505 + }, + { + "epoch": 4.179236309281152, + "grad_norm": 1.0392727851867676, + "learning_rate": 0.000477595461339856, + "loss": 3.4765, + "step": 61510 + }, + { + "epoch": 4.179576029351814, + "grad_norm": 1.1129608154296875, + "learning_rate": 0.00047755299633102323, + "loss": 3.2285, + "step": 61515 + }, + { + "epoch": 4.1799157494224755, + "grad_norm": 0.7832546234130859, + "learning_rate": 0.0004775105313221905, + "loss": 3.5716, + "step": 61520 + }, + { + "epoch": 4.1802554694931375, + "grad_norm": 0.9150946140289307, + "learning_rate": 0.00047746806631335784, + "loss": 3.1979, + "step": 61525 + }, + { + "epoch": 4.1805951895638, + "grad_norm": 0.64959716796875, + "learning_rate": 0.00047742560130452507, + "loss": 3.673, + "step": 61530 + }, + { + "epoch": 4.180934909634461, + "grad_norm": 0.977350652217865, + "learning_rate": 0.00047738313629569235, + "loss": 3.7289, + "step": 61535 + }, + { + "epoch": 4.181274629705123, + "grad_norm": 0.9625280499458313, + "learning_rate": 0.00047734067128685963, + "loss": 3.4824, + "step": 61540 + }, + { + "epoch": 4.181614349775785, + "grad_norm": 0.9279688000679016, + "learning_rate": 0.0004772982062780269, + "loss": 3.5479, + "step": 61545 + }, + { + "epoch": 4.181954069846446, + "grad_norm": 0.9773569107055664, + "learning_rate": 0.0004772557412691942, + "loss": 3.4389, + "step": 61550 + }, + { + "epoch": 4.182293789917108, + "grad_norm": 0.8943056464195251, + "learning_rate": 0.00047721327626036147, + "loss": 3.524, + "step": 61555 + }, + { + "epoch": 4.18263350998777, + "grad_norm": 0.8047959208488464, + "learning_rate": 0.00047717081125152875, + "loss": 3.2588, + "step": 61560 + }, + { + "epoch": 4.1829732300584315, + "grad_norm": 0.8745054006576538, + "learning_rate": 0.00047712834624269603, + "loss": 3.6413, + "step": 61565 + }, + { + "epoch": 4.1833129501290935, + "grad_norm": 0.9662274718284607, + "learning_rate": 0.0004770858812338633, + "loss": 3.4595, + "step": 61570 + }, + { + "epoch": 4.183652670199756, + "grad_norm": 0.8292756676673889, + "learning_rate": 0.00047704341622503054, + "loss": 3.6498, + "step": 61575 + }, + { + "epoch": 4.183992390270417, + "grad_norm": 0.7691963911056519, + "learning_rate": 0.00047700095121619787, + "loss": 3.6595, + "step": 61580 + }, + { + "epoch": 4.184332110341079, + "grad_norm": 0.6002717614173889, + "learning_rate": 0.00047695848620736515, + "loss": 3.5455, + "step": 61585 + }, + { + "epoch": 4.184671830411741, + "grad_norm": 0.7317185997962952, + "learning_rate": 0.00047691602119853243, + "loss": 3.5331, + "step": 61590 + }, + { + "epoch": 4.185011550482402, + "grad_norm": 0.8353995680809021, + "learning_rate": 0.0004768735561896997, + "loss": 3.5765, + "step": 61595 + }, + { + "epoch": 4.185351270553064, + "grad_norm": 0.7571285963058472, + "learning_rate": 0.000476831091180867, + "loss": 3.7165, + "step": 61600 + }, + { + "epoch": 4.185690990623726, + "grad_norm": 0.8857873678207397, + "learning_rate": 0.00047678862617203427, + "loss": 3.5826, + "step": 61605 + }, + { + "epoch": 4.1860307106943875, + "grad_norm": 1.500422716140747, + "learning_rate": 0.0004767461611632015, + "loss": 3.594, + "step": 61610 + }, + { + "epoch": 4.1863704307650496, + "grad_norm": 1.162610411643982, + "learning_rate": 0.00047670369615436883, + "loss": 3.4978, + "step": 61615 + }, + { + "epoch": 4.186710150835712, + "grad_norm": 1.0689085721969604, + "learning_rate": 0.0004766612311455361, + "loss": 3.5149, + "step": 61620 + }, + { + "epoch": 4.187049870906373, + "grad_norm": 0.910282552242279, + "learning_rate": 0.00047661876613670334, + "loss": 3.4493, + "step": 61625 + }, + { + "epoch": 4.187389590977035, + "grad_norm": 1.0557277202606201, + "learning_rate": 0.00047657630112787067, + "loss": 3.6462, + "step": 61630 + }, + { + "epoch": 4.187729311047697, + "grad_norm": 0.9505425095558167, + "learning_rate": 0.00047653383611903795, + "loss": 3.3805, + "step": 61635 + }, + { + "epoch": 4.188069031118358, + "grad_norm": 0.8598436117172241, + "learning_rate": 0.0004764913711102052, + "loss": 3.6132, + "step": 61640 + }, + { + "epoch": 4.18840875118902, + "grad_norm": 0.7603653073310852, + "learning_rate": 0.00047644890610137246, + "loss": 3.3015, + "step": 61645 + }, + { + "epoch": 4.188748471259682, + "grad_norm": 0.9686315059661865, + "learning_rate": 0.0004764064410925398, + "loss": 3.3529, + "step": 61650 + }, + { + "epoch": 4.1890881913303435, + "grad_norm": 0.904583752155304, + "learning_rate": 0.000476363976083707, + "loss": 3.4892, + "step": 61655 + }, + { + "epoch": 4.189427911401006, + "grad_norm": 1.062279462814331, + "learning_rate": 0.0004763215110748743, + "loss": 3.3009, + "step": 61660 + }, + { + "epoch": 4.189767631471668, + "grad_norm": 0.8175138235092163, + "learning_rate": 0.00047627904606604163, + "loss": 3.4636, + "step": 61665 + }, + { + "epoch": 4.190107351542329, + "grad_norm": 0.9394064545631409, + "learning_rate": 0.00047623658105720886, + "loss": 3.5016, + "step": 61670 + }, + { + "epoch": 4.190447071612991, + "grad_norm": 1.0974457263946533, + "learning_rate": 0.00047619411604837614, + "loss": 3.4715, + "step": 61675 + }, + { + "epoch": 4.190786791683653, + "grad_norm": 0.8635875582695007, + "learning_rate": 0.0004761516510395434, + "loss": 3.3515, + "step": 61680 + }, + { + "epoch": 4.191126511754314, + "grad_norm": 0.9464661478996277, + "learning_rate": 0.0004761091860307107, + "loss": 3.1918, + "step": 61685 + }, + { + "epoch": 4.191466231824976, + "grad_norm": 1.0221835374832153, + "learning_rate": 0.000476066721021878, + "loss": 3.5597, + "step": 61690 + }, + { + "epoch": 4.191805951895638, + "grad_norm": 0.932763934135437, + "learning_rate": 0.00047602425601304526, + "loss": 3.7033, + "step": 61695 + }, + { + "epoch": 4.1921456719662995, + "grad_norm": 0.8700383901596069, + "learning_rate": 0.00047598179100421254, + "loss": 3.6733, + "step": 61700 + }, + { + "epoch": 4.192485392036962, + "grad_norm": 0.8092880845069885, + "learning_rate": 0.0004759393259953798, + "loss": 3.2863, + "step": 61705 + }, + { + "epoch": 4.192825112107624, + "grad_norm": 0.9158156514167786, + "learning_rate": 0.0004758968609865471, + "loss": 3.3602, + "step": 61710 + }, + { + "epoch": 4.193164832178285, + "grad_norm": 0.9173544645309448, + "learning_rate": 0.0004758543959777143, + "loss": 3.3187, + "step": 61715 + }, + { + "epoch": 4.193504552248947, + "grad_norm": 0.8213764429092407, + "learning_rate": 0.00047581193096888166, + "loss": 3.3591, + "step": 61720 + }, + { + "epoch": 4.193844272319609, + "grad_norm": 0.8613501787185669, + "learning_rate": 0.00047576946596004894, + "loss": 3.3481, + "step": 61725 + }, + { + "epoch": 4.19418399239027, + "grad_norm": 0.8321738839149475, + "learning_rate": 0.00047572700095121616, + "loss": 3.599, + "step": 61730 + }, + { + "epoch": 4.194523712460932, + "grad_norm": 0.9807544946670532, + "learning_rate": 0.0004756845359423835, + "loss": 3.5017, + "step": 61735 + }, + { + "epoch": 4.194863432531594, + "grad_norm": 0.8135389685630798, + "learning_rate": 0.0004756420709335508, + "loss": 3.4571, + "step": 61740 + }, + { + "epoch": 4.1952031526022555, + "grad_norm": 0.9644518494606018, + "learning_rate": 0.000475599605924718, + "loss": 3.584, + "step": 61745 + }, + { + "epoch": 4.195542872672918, + "grad_norm": 0.7378988862037659, + "learning_rate": 0.0004755571409158853, + "loss": 3.6518, + "step": 61750 + }, + { + "epoch": 4.19588259274358, + "grad_norm": 0.8387075066566467, + "learning_rate": 0.0004755146759070526, + "loss": 3.5965, + "step": 61755 + }, + { + "epoch": 4.196222312814241, + "grad_norm": 0.8610532283782959, + "learning_rate": 0.0004754722108982199, + "loss": 3.4606, + "step": 61760 + }, + { + "epoch": 4.196562032884903, + "grad_norm": 0.8247411847114563, + "learning_rate": 0.0004754297458893871, + "loss": 3.3262, + "step": 61765 + }, + { + "epoch": 4.196901752955565, + "grad_norm": 0.980965793132782, + "learning_rate": 0.00047538728088055446, + "loss": 3.6096, + "step": 61770 + }, + { + "epoch": 4.197241473026226, + "grad_norm": 0.8761126399040222, + "learning_rate": 0.00047534481587172174, + "loss": 3.5151, + "step": 61775 + }, + { + "epoch": 4.197581193096888, + "grad_norm": 0.8383309245109558, + "learning_rate": 0.00047530235086288896, + "loss": 3.5543, + "step": 61780 + }, + { + "epoch": 4.19792091316755, + "grad_norm": 0.9866697788238525, + "learning_rate": 0.0004752598858540563, + "loss": 3.404, + "step": 61785 + }, + { + "epoch": 4.1982606332382115, + "grad_norm": 0.6978961229324341, + "learning_rate": 0.0004752174208452236, + "loss": 3.5538, + "step": 61790 + }, + { + "epoch": 4.198600353308874, + "grad_norm": 0.9162205457687378, + "learning_rate": 0.0004751749558363908, + "loss": 3.4356, + "step": 61795 + }, + { + "epoch": 4.198940073379536, + "grad_norm": 0.8467340469360352, + "learning_rate": 0.0004751324908275581, + "loss": 3.7083, + "step": 61800 + }, + { + "epoch": 4.199279793450197, + "grad_norm": 0.9185104370117188, + "learning_rate": 0.0004750900258187254, + "loss": 3.538, + "step": 61805 + }, + { + "epoch": 4.199619513520859, + "grad_norm": 0.9455029368400574, + "learning_rate": 0.00047504756080989265, + "loss": 3.5196, + "step": 61810 + }, + { + "epoch": 4.199959233591521, + "grad_norm": 0.8226585388183594, + "learning_rate": 0.0004750050958010599, + "loss": 3.5964, + "step": 61815 + }, + { + "epoch": 4.200298953662182, + "grad_norm": 1.0438311100006104, + "learning_rate": 0.00047496263079222726, + "loss": 3.5756, + "step": 61820 + }, + { + "epoch": 4.200638673732844, + "grad_norm": 0.9407967329025269, + "learning_rate": 0.0004749201657833945, + "loss": 3.4944, + "step": 61825 + }, + { + "epoch": 4.200978393803506, + "grad_norm": 0.8689625263214111, + "learning_rate": 0.00047487770077456177, + "loss": 3.6475, + "step": 61830 + }, + { + "epoch": 4.2013181138741675, + "grad_norm": 1.5444892644882202, + "learning_rate": 0.00047483523576572905, + "loss": 3.507, + "step": 61835 + }, + { + "epoch": 4.20165783394483, + "grad_norm": 0.8338109254837036, + "learning_rate": 0.0004747927707568963, + "loss": 3.398, + "step": 61840 + }, + { + "epoch": 4.201997554015492, + "grad_norm": 7.324791431427002, + "learning_rate": 0.0004747503057480636, + "loss": 3.4509, + "step": 61845 + }, + { + "epoch": 4.202337274086153, + "grad_norm": 1.0446778535842896, + "learning_rate": 0.0004747078407392309, + "loss": 3.3387, + "step": 61850 + }, + { + "epoch": 4.202676994156815, + "grad_norm": 1.123361349105835, + "learning_rate": 0.00047466537573039817, + "loss": 3.2169, + "step": 61855 + }, + { + "epoch": 4.203016714227476, + "grad_norm": 0.8754403591156006, + "learning_rate": 0.00047462291072156545, + "loss": 3.371, + "step": 61860 + }, + { + "epoch": 4.203356434298138, + "grad_norm": 0.8180702328681946, + "learning_rate": 0.0004745804457127327, + "loss": 3.383, + "step": 61865 + }, + { + "epoch": 4.2036961543688, + "grad_norm": 0.9273421764373779, + "learning_rate": 0.00047453798070389995, + "loss": 3.4114, + "step": 61870 + }, + { + "epoch": 4.2040358744394615, + "grad_norm": 1.1498231887817383, + "learning_rate": 0.0004744955156950673, + "loss": 3.1782, + "step": 61875 + }, + { + "epoch": 4.2043755945101235, + "grad_norm": 0.745573103427887, + "learning_rate": 0.00047445305068623457, + "loss": 3.5043, + "step": 61880 + }, + { + "epoch": 4.204715314580786, + "grad_norm": 0.7588533759117126, + "learning_rate": 0.0004744105856774018, + "loss": 3.5179, + "step": 61885 + }, + { + "epoch": 4.205055034651447, + "grad_norm": 0.9656900763511658, + "learning_rate": 0.0004743681206685691, + "loss": 3.4826, + "step": 61890 + }, + { + "epoch": 4.205394754722109, + "grad_norm": 0.964968204498291, + "learning_rate": 0.0004743256556597364, + "loss": 3.5144, + "step": 61895 + }, + { + "epoch": 4.205734474792771, + "grad_norm": 1.1871525049209595, + "learning_rate": 0.00047428319065090363, + "loss": 3.3906, + "step": 61900 + }, + { + "epoch": 4.206074194863432, + "grad_norm": 0.9983317852020264, + "learning_rate": 0.0004742407256420709, + "loss": 3.3158, + "step": 61905 + }, + { + "epoch": 4.206413914934094, + "grad_norm": 0.8124498128890991, + "learning_rate": 0.00047419826063323825, + "loss": 3.5586, + "step": 61910 + }, + { + "epoch": 4.206753635004756, + "grad_norm": 0.7715133428573608, + "learning_rate": 0.00047415579562440547, + "loss": 3.6549, + "step": 61915 + }, + { + "epoch": 4.2070933550754175, + "grad_norm": 1.1109403371810913, + "learning_rate": 0.00047411333061557275, + "loss": 3.5215, + "step": 61920 + }, + { + "epoch": 4.20743307514608, + "grad_norm": 0.7509203553199768, + "learning_rate": 0.0004740708656067401, + "loss": 3.3871, + "step": 61925 + }, + { + "epoch": 4.207772795216742, + "grad_norm": 0.8395061492919922, + "learning_rate": 0.00047402840059790737, + "loss": 3.3894, + "step": 61930 + }, + { + "epoch": 4.208112515287403, + "grad_norm": 0.6614095568656921, + "learning_rate": 0.0004739859355890746, + "loss": 3.462, + "step": 61935 + }, + { + "epoch": 4.208452235358065, + "grad_norm": 0.9735466837882996, + "learning_rate": 0.0004739434705802419, + "loss": 3.5217, + "step": 61940 + }, + { + "epoch": 4.208791955428727, + "grad_norm": 0.768747866153717, + "learning_rate": 0.0004739010055714092, + "loss": 3.3749, + "step": 61945 + }, + { + "epoch": 4.209131675499388, + "grad_norm": 0.8979070782661438, + "learning_rate": 0.00047385854056257643, + "loss": 3.667, + "step": 61950 + }, + { + "epoch": 4.20947139557005, + "grad_norm": 0.9836918115615845, + "learning_rate": 0.0004738160755537437, + "loss": 3.4834, + "step": 61955 + }, + { + "epoch": 4.209811115640712, + "grad_norm": 0.9414037466049194, + "learning_rate": 0.00047377361054491105, + "loss": 3.4967, + "step": 61960 + }, + { + "epoch": 4.2101508357113735, + "grad_norm": 1.077491283416748, + "learning_rate": 0.0004737311455360783, + "loss": 3.5865, + "step": 61965 + }, + { + "epoch": 4.210490555782036, + "grad_norm": 0.8763558864593506, + "learning_rate": 0.00047368868052724555, + "loss": 3.2732, + "step": 61970 + }, + { + "epoch": 4.210830275852698, + "grad_norm": 0.8472718596458435, + "learning_rate": 0.00047364621551841283, + "loss": 3.5139, + "step": 61975 + }, + { + "epoch": 4.211169995923359, + "grad_norm": 0.923512876033783, + "learning_rate": 0.0004736037505095801, + "loss": 3.3765, + "step": 61980 + }, + { + "epoch": 4.211509715994021, + "grad_norm": 0.9110320210456848, + "learning_rate": 0.0004735612855007474, + "loss": 3.4979, + "step": 61985 + }, + { + "epoch": 4.211849436064683, + "grad_norm": 0.9856606721878052, + "learning_rate": 0.0004735188204919147, + "loss": 3.3492, + "step": 61990 + }, + { + "epoch": 4.212189156135344, + "grad_norm": 1.143573522567749, + "learning_rate": 0.00047347635548308195, + "loss": 3.521, + "step": 61995 + }, + { + "epoch": 4.212528876206006, + "grad_norm": 0.7409008741378784, + "learning_rate": 0.00047343389047424923, + "loss": 3.5756, + "step": 62000 + }, + { + "epoch": 4.212868596276668, + "grad_norm": 1.0568146705627441, + "learning_rate": 0.0004733914254654165, + "loss": 3.5422, + "step": 62005 + }, + { + "epoch": 4.2132083163473295, + "grad_norm": 1.1648577451705933, + "learning_rate": 0.00047334896045658374, + "loss": 3.349, + "step": 62010 + }, + { + "epoch": 4.213548036417992, + "grad_norm": 0.6332630515098572, + "learning_rate": 0.0004733064954477511, + "loss": 3.5297, + "step": 62015 + }, + { + "epoch": 4.213887756488654, + "grad_norm": 0.990820586681366, + "learning_rate": 0.00047326403043891835, + "loss": 3.5881, + "step": 62020 + }, + { + "epoch": 4.214227476559315, + "grad_norm": 0.7660771012306213, + "learning_rate": 0.0004732215654300856, + "loss": 3.6545, + "step": 62025 + }, + { + "epoch": 4.214567196629977, + "grad_norm": 0.7424066066741943, + "learning_rate": 0.0004731791004212529, + "loss": 3.4747, + "step": 62030 + }, + { + "epoch": 4.214906916700639, + "grad_norm": 1.3720229864120483, + "learning_rate": 0.0004731366354124202, + "loss": 3.5914, + "step": 62035 + }, + { + "epoch": 4.2152466367713, + "grad_norm": 1.0905529260635376, + "learning_rate": 0.0004730941704035874, + "loss": 3.178, + "step": 62040 + }, + { + "epoch": 4.215586356841962, + "grad_norm": 0.8345396518707275, + "learning_rate": 0.00047305170539475475, + "loss": 3.6747, + "step": 62045 + }, + { + "epoch": 4.215926076912624, + "grad_norm": 0.7810105681419373, + "learning_rate": 0.00047300924038592203, + "loss": 3.3026, + "step": 62050 + }, + { + "epoch": 4.2162657969832855, + "grad_norm": 1.2986418008804321, + "learning_rate": 0.00047296677537708926, + "loss": 3.2723, + "step": 62055 + }, + { + "epoch": 4.216605517053948, + "grad_norm": 0.942055881023407, + "learning_rate": 0.00047292431036825654, + "loss": 3.3877, + "step": 62060 + }, + { + "epoch": 4.21694523712461, + "grad_norm": 0.7600162029266357, + "learning_rate": 0.0004728818453594239, + "loss": 3.4083, + "step": 62065 + }, + { + "epoch": 4.217284957195271, + "grad_norm": 1.151014804840088, + "learning_rate": 0.0004728393803505911, + "loss": 3.7406, + "step": 62070 + }, + { + "epoch": 4.217624677265933, + "grad_norm": 0.9654062390327454, + "learning_rate": 0.0004727969153417584, + "loss": 3.4223, + "step": 62075 + }, + { + "epoch": 4.217964397336595, + "grad_norm": 0.8574452996253967, + "learning_rate": 0.0004727544503329257, + "loss": 3.3046, + "step": 62080 + }, + { + "epoch": 4.218304117407256, + "grad_norm": 0.923534631729126, + "learning_rate": 0.00047271198532409294, + "loss": 3.5365, + "step": 62085 + }, + { + "epoch": 4.218643837477918, + "grad_norm": 1.0087449550628662, + "learning_rate": 0.0004726695203152602, + "loss": 3.521, + "step": 62090 + }, + { + "epoch": 4.21898355754858, + "grad_norm": 0.8792451620101929, + "learning_rate": 0.0004726270553064275, + "loss": 3.4545, + "step": 62095 + }, + { + "epoch": 4.2193232776192415, + "grad_norm": 0.8680570721626282, + "learning_rate": 0.00047258459029759484, + "loss": 3.5337, + "step": 62100 + }, + { + "epoch": 4.219662997689904, + "grad_norm": 0.7142009139060974, + "learning_rate": 0.00047254212528876206, + "loss": 3.6364, + "step": 62105 + }, + { + "epoch": 4.220002717760566, + "grad_norm": 0.9738248586654663, + "learning_rate": 0.00047249966027992934, + "loss": 3.1205, + "step": 62110 + }, + { + "epoch": 4.220342437831227, + "grad_norm": 0.8425074219703674, + "learning_rate": 0.0004724571952710967, + "loss": 3.4945, + "step": 62115 + }, + { + "epoch": 4.220682157901889, + "grad_norm": 0.7377592921257019, + "learning_rate": 0.0004724147302622639, + "loss": 3.5725, + "step": 62120 + }, + { + "epoch": 4.221021877972551, + "grad_norm": 0.8036885261535645, + "learning_rate": 0.0004723722652534312, + "loss": 3.4892, + "step": 62125 + }, + { + "epoch": 4.221361598043212, + "grad_norm": 0.9702445268630981, + "learning_rate": 0.00047232980024459846, + "loss": 3.5585, + "step": 62130 + }, + { + "epoch": 4.221701318113874, + "grad_norm": 0.9419636130332947, + "learning_rate": 0.00047228733523576574, + "loss": 3.3501, + "step": 62135 + }, + { + "epoch": 4.222041038184536, + "grad_norm": 0.7829832434654236, + "learning_rate": 0.000472244870226933, + "loss": 3.3922, + "step": 62140 + }, + { + "epoch": 4.2223807582551975, + "grad_norm": 0.7860432863235474, + "learning_rate": 0.0004722024052181003, + "loss": 3.5126, + "step": 62145 + }, + { + "epoch": 4.22272047832586, + "grad_norm": 0.8846808075904846, + "learning_rate": 0.0004721599402092676, + "loss": 3.6187, + "step": 62150 + }, + { + "epoch": 4.223060198396522, + "grad_norm": 0.9532198905944824, + "learning_rate": 0.00047211747520043486, + "loss": 3.6106, + "step": 62155 + }, + { + "epoch": 4.223399918467183, + "grad_norm": 0.8060060143470764, + "learning_rate": 0.00047207501019160214, + "loss": 3.3492, + "step": 62160 + }, + { + "epoch": 4.223739638537845, + "grad_norm": 0.8280306458473206, + "learning_rate": 0.00047203254518276937, + "loss": 3.3005, + "step": 62165 + }, + { + "epoch": 4.224079358608506, + "grad_norm": 1.0023374557495117, + "learning_rate": 0.0004719900801739367, + "loss": 3.2242, + "step": 62170 + }, + { + "epoch": 4.224419078679168, + "grad_norm": 0.7857472896575928, + "learning_rate": 0.000471947615165104, + "loss": 3.2931, + "step": 62175 + }, + { + "epoch": 4.22475879874983, + "grad_norm": 1.0773108005523682, + "learning_rate": 0.0004719051501562712, + "loss": 3.5876, + "step": 62180 + }, + { + "epoch": 4.2250985188204915, + "grad_norm": 0.9203125238418579, + "learning_rate": 0.00047186268514743854, + "loss": 3.3219, + "step": 62185 + }, + { + "epoch": 4.2254382388911536, + "grad_norm": 1.1758345365524292, + "learning_rate": 0.0004718202201386058, + "loss": 3.5427, + "step": 62190 + }, + { + "epoch": 4.225777958961816, + "grad_norm": 0.8343169093132019, + "learning_rate": 0.00047177775512977305, + "loss": 3.3379, + "step": 62195 + }, + { + "epoch": 4.226117679032477, + "grad_norm": 0.9148077964782715, + "learning_rate": 0.00047173529012094033, + "loss": 3.5548, + "step": 62200 + }, + { + "epoch": 4.226457399103139, + "grad_norm": 0.9804446697235107, + "learning_rate": 0.00047169282511210766, + "loss": 3.5708, + "step": 62205 + }, + { + "epoch": 4.226797119173801, + "grad_norm": 0.8714418411254883, + "learning_rate": 0.0004716503601032749, + "loss": 3.2235, + "step": 62210 + }, + { + "epoch": 4.227136839244462, + "grad_norm": 1.0064260959625244, + "learning_rate": 0.00047160789509444217, + "loss": 3.3066, + "step": 62215 + }, + { + "epoch": 4.227476559315124, + "grad_norm": 0.8281447291374207, + "learning_rate": 0.0004715654300856095, + "loss": 3.5559, + "step": 62220 + }, + { + "epoch": 4.227816279385786, + "grad_norm": 0.8623903393745422, + "learning_rate": 0.00047152296507677673, + "loss": 3.3446, + "step": 62225 + }, + { + "epoch": 4.2281559994564475, + "grad_norm": 1.0136678218841553, + "learning_rate": 0.000471480500067944, + "loss": 3.4321, + "step": 62230 + }, + { + "epoch": 4.22849571952711, + "grad_norm": 1.19075608253479, + "learning_rate": 0.0004714380350591113, + "loss": 3.4994, + "step": 62235 + }, + { + "epoch": 4.228835439597772, + "grad_norm": 1.0560334920883179, + "learning_rate": 0.00047139557005027857, + "loss": 3.5499, + "step": 62240 + }, + { + "epoch": 4.229175159668433, + "grad_norm": 0.9597827792167664, + "learning_rate": 0.00047135310504144585, + "loss": 3.5025, + "step": 62245 + }, + { + "epoch": 4.229514879739095, + "grad_norm": 0.7794488072395325, + "learning_rate": 0.00047131064003261313, + "loss": 3.3411, + "step": 62250 + }, + { + "epoch": 4.229854599809757, + "grad_norm": 0.8876743316650391, + "learning_rate": 0.0004712681750237804, + "loss": 3.4017, + "step": 62255 + }, + { + "epoch": 4.230194319880418, + "grad_norm": 0.7387467622756958, + "learning_rate": 0.0004712257100149477, + "loss": 3.7088, + "step": 62260 + }, + { + "epoch": 4.23053403995108, + "grad_norm": 0.8620684146881104, + "learning_rate": 0.00047118324500611497, + "loss": 3.4169, + "step": 62265 + }, + { + "epoch": 4.230873760021742, + "grad_norm": 1.9649810791015625, + "learning_rate": 0.00047114077999728225, + "loss": 3.3432, + "step": 62270 + }, + { + "epoch": 4.2312134800924035, + "grad_norm": 0.8399444222450256, + "learning_rate": 0.00047109831498844953, + "loss": 3.5822, + "step": 62275 + }, + { + "epoch": 4.231553200163066, + "grad_norm": 0.8617279529571533, + "learning_rate": 0.0004710558499796168, + "loss": 3.6869, + "step": 62280 + }, + { + "epoch": 4.231892920233728, + "grad_norm": 0.7887837886810303, + "learning_rate": 0.0004710133849707841, + "loss": 3.6155, + "step": 62285 + }, + { + "epoch": 4.232232640304389, + "grad_norm": 0.856587827205658, + "learning_rate": 0.00047097091996195137, + "loss": 3.5358, + "step": 62290 + }, + { + "epoch": 4.232572360375051, + "grad_norm": 0.9303193092346191, + "learning_rate": 0.00047092845495311865, + "loss": 3.4816, + "step": 62295 + }, + { + "epoch": 4.232912080445713, + "grad_norm": 0.9314841032028198, + "learning_rate": 0.00047088598994428593, + "loss": 3.47, + "step": 62300 + }, + { + "epoch": 4.233251800516374, + "grad_norm": 1.1167455911636353, + "learning_rate": 0.00047084352493545316, + "loss": 3.3843, + "step": 62305 + }, + { + "epoch": 4.233591520587036, + "grad_norm": 1.0100398063659668, + "learning_rate": 0.0004708010599266205, + "loss": 3.224, + "step": 62310 + }, + { + "epoch": 4.233931240657698, + "grad_norm": 0.895997166633606, + "learning_rate": 0.00047075859491778777, + "loss": 3.5847, + "step": 62315 + }, + { + "epoch": 4.2342709607283595, + "grad_norm": 0.837062656879425, + "learning_rate": 0.000470716129908955, + "loss": 3.1669, + "step": 62320 + }, + { + "epoch": 4.234610680799022, + "grad_norm": 1.0972082614898682, + "learning_rate": 0.00047067366490012233, + "loss": 3.2607, + "step": 62325 + }, + { + "epoch": 4.234950400869684, + "grad_norm": 1.03299880027771, + "learning_rate": 0.0004706311998912896, + "loss": 3.3383, + "step": 62330 + }, + { + "epoch": 4.235290120940345, + "grad_norm": 0.8328218460083008, + "learning_rate": 0.00047058873488245684, + "loss": 3.691, + "step": 62335 + }, + { + "epoch": 4.235629841011007, + "grad_norm": 0.7953068017959595, + "learning_rate": 0.00047054626987362417, + "loss": 3.5038, + "step": 62340 + }, + { + "epoch": 4.235969561081669, + "grad_norm": 2.4780311584472656, + "learning_rate": 0.00047050380486479145, + "loss": 3.3094, + "step": 62345 + }, + { + "epoch": 4.23630928115233, + "grad_norm": 1.0611635446548462, + "learning_rate": 0.0004704613398559587, + "loss": 3.4803, + "step": 62350 + }, + { + "epoch": 4.236649001222992, + "grad_norm": 0.8358001112937927, + "learning_rate": 0.00047041887484712596, + "loss": 3.4165, + "step": 62355 + }, + { + "epoch": 4.236988721293654, + "grad_norm": 0.8839282989501953, + "learning_rate": 0.0004703764098382933, + "loss": 3.3485, + "step": 62360 + }, + { + "epoch": 4.2373284413643155, + "grad_norm": 0.6768919229507446, + "learning_rate": 0.0004703339448294605, + "loss": 3.4817, + "step": 62365 + }, + { + "epoch": 4.237668161434978, + "grad_norm": 0.9148284196853638, + "learning_rate": 0.0004702914798206278, + "loss": 3.4989, + "step": 62370 + }, + { + "epoch": 4.23800788150564, + "grad_norm": 0.8957261443138123, + "learning_rate": 0.00047024901481179513, + "loss": 3.3334, + "step": 62375 + }, + { + "epoch": 4.238347601576301, + "grad_norm": 0.9470516443252563, + "learning_rate": 0.00047020654980296236, + "loss": 3.4824, + "step": 62380 + }, + { + "epoch": 4.238687321646963, + "grad_norm": 0.9323439002037048, + "learning_rate": 0.00047016408479412964, + "loss": 3.4644, + "step": 62385 + }, + { + "epoch": 4.239027041717625, + "grad_norm": 0.8894497752189636, + "learning_rate": 0.0004701216197852969, + "loss": 3.4867, + "step": 62390 + }, + { + "epoch": 4.239366761788286, + "grad_norm": 1.1439660787582397, + "learning_rate": 0.0004700791547764642, + "loss": 3.5749, + "step": 62395 + }, + { + "epoch": 4.239706481858948, + "grad_norm": 0.9108113050460815, + "learning_rate": 0.0004700366897676315, + "loss": 3.4178, + "step": 62400 + }, + { + "epoch": 4.24004620192961, + "grad_norm": 1.0065360069274902, + "learning_rate": 0.00046999422475879876, + "loss": 3.4966, + "step": 62405 + }, + { + "epoch": 4.2403859220002715, + "grad_norm": 0.9580250978469849, + "learning_rate": 0.00046995175974996604, + "loss": 3.6634, + "step": 62410 + }, + { + "epoch": 4.240725642070934, + "grad_norm": 0.7834501266479492, + "learning_rate": 0.0004699092947411333, + "loss": 3.6127, + "step": 62415 + }, + { + "epoch": 4.241065362141596, + "grad_norm": 0.7347567081451416, + "learning_rate": 0.0004698668297323006, + "loss": 3.6298, + "step": 62420 + }, + { + "epoch": 4.241405082212257, + "grad_norm": 1.1166601181030273, + "learning_rate": 0.0004698243647234678, + "loss": 3.5458, + "step": 62425 + }, + { + "epoch": 4.241744802282919, + "grad_norm": 0.8660541772842407, + "learning_rate": 0.00046978189971463516, + "loss": 3.2887, + "step": 62430 + }, + { + "epoch": 4.242084522353581, + "grad_norm": 0.8945692181587219, + "learning_rate": 0.00046973943470580244, + "loss": 3.3532, + "step": 62435 + }, + { + "epoch": 4.242424242424242, + "grad_norm": 0.7985753417015076, + "learning_rate": 0.0004696969696969697, + "loss": 3.2211, + "step": 62440 + }, + { + "epoch": 4.242763962494904, + "grad_norm": 0.8653902411460876, + "learning_rate": 0.000469654504688137, + "loss": 3.6749, + "step": 62445 + }, + { + "epoch": 4.243103682565566, + "grad_norm": 0.9973759651184082, + "learning_rate": 0.0004696120396793043, + "loss": 3.4082, + "step": 62450 + }, + { + "epoch": 4.2434434026362275, + "grad_norm": 0.8569865822792053, + "learning_rate": 0.00046956957467047156, + "loss": 3.617, + "step": 62455 + }, + { + "epoch": 4.24378312270689, + "grad_norm": 0.5923313498497009, + "learning_rate": 0.0004695271096616388, + "loss": 3.5664, + "step": 62460 + }, + { + "epoch": 4.244122842777552, + "grad_norm": 0.7659757733345032, + "learning_rate": 0.0004694846446528061, + "loss": 3.4558, + "step": 62465 + }, + { + "epoch": 4.244462562848213, + "grad_norm": 0.8422825336456299, + "learning_rate": 0.0004694421796439734, + "loss": 3.1768, + "step": 62470 + }, + { + "epoch": 4.244802282918875, + "grad_norm": 3.5348448753356934, + "learning_rate": 0.0004693997146351406, + "loss": 3.4113, + "step": 62475 + }, + { + "epoch": 4.245142002989537, + "grad_norm": 0.895234227180481, + "learning_rate": 0.00046935724962630796, + "loss": 3.262, + "step": 62480 + }, + { + "epoch": 4.245481723060198, + "grad_norm": 0.9319461584091187, + "learning_rate": 0.00046931478461747524, + "loss": 3.277, + "step": 62485 + }, + { + "epoch": 4.24582144313086, + "grad_norm": 1.0715136528015137, + "learning_rate": 0.00046927231960864246, + "loss": 3.3541, + "step": 62490 + }, + { + "epoch": 4.246161163201522, + "grad_norm": 0.9865347743034363, + "learning_rate": 0.00046922985459980974, + "loss": 3.3116, + "step": 62495 + }, + { + "epoch": 4.246500883272184, + "grad_norm": 1.1252214908599854, + "learning_rate": 0.0004691873895909771, + "loss": 3.5165, + "step": 62500 + }, + { + "epoch": 4.246840603342846, + "grad_norm": 0.9303372502326965, + "learning_rate": 0.0004691449245821443, + "loss": 3.4558, + "step": 62505 + }, + { + "epoch": 4.247180323413508, + "grad_norm": 0.981105387210846, + "learning_rate": 0.0004691024595733116, + "loss": 3.6547, + "step": 62510 + }, + { + "epoch": 4.247520043484169, + "grad_norm": 1.1469906568527222, + "learning_rate": 0.0004690599945644789, + "loss": 3.5213, + "step": 62515 + }, + { + "epoch": 4.247859763554831, + "grad_norm": 0.923545241355896, + "learning_rate": 0.00046901752955564615, + "loss": 3.2604, + "step": 62520 + }, + { + "epoch": 4.248199483625493, + "grad_norm": 0.8923083543777466, + "learning_rate": 0.0004689750645468134, + "loss": 3.2821, + "step": 62525 + }, + { + "epoch": 4.248539203696154, + "grad_norm": 0.7732527256011963, + "learning_rate": 0.0004689325995379807, + "loss": 3.4976, + "step": 62530 + }, + { + "epoch": 4.248878923766816, + "grad_norm": 0.8284834027290344, + "learning_rate": 0.000468890134529148, + "loss": 3.3868, + "step": 62535 + }, + { + "epoch": 4.2492186438374775, + "grad_norm": 0.9480911493301392, + "learning_rate": 0.00046884766952031527, + "loss": 3.6136, + "step": 62540 + }, + { + "epoch": 4.24955836390814, + "grad_norm": 0.9909446239471436, + "learning_rate": 0.00046880520451148255, + "loss": 3.3003, + "step": 62545 + }, + { + "epoch": 4.249898083978802, + "grad_norm": 0.9021207094192505, + "learning_rate": 0.0004687627395026498, + "loss": 3.4878, + "step": 62550 + }, + { + "epoch": 4.250237804049463, + "grad_norm": 0.749625027179718, + "learning_rate": 0.0004687202744938171, + "loss": 3.3482, + "step": 62555 + }, + { + "epoch": 4.250577524120125, + "grad_norm": 0.7491831183433533, + "learning_rate": 0.0004686778094849844, + "loss": 3.1142, + "step": 62560 + }, + { + "epoch": 4.250917244190787, + "grad_norm": 4.378658771514893, + "learning_rate": 0.0004686353444761516, + "loss": 3.5059, + "step": 62565 + }, + { + "epoch": 4.251256964261448, + "grad_norm": 0.8541637659072876, + "learning_rate": 0.00046859287946731895, + "loss": 3.5269, + "step": 62570 + }, + { + "epoch": 4.25159668433211, + "grad_norm": 0.9197339415550232, + "learning_rate": 0.0004685504144584862, + "loss": 3.5367, + "step": 62575 + }, + { + "epoch": 4.251936404402772, + "grad_norm": 0.8299925923347473, + "learning_rate": 0.00046850794944965345, + "loss": 3.2937, + "step": 62580 + }, + { + "epoch": 4.2522761244734335, + "grad_norm": 0.8622176051139832, + "learning_rate": 0.0004684654844408208, + "loss": 3.5138, + "step": 62585 + }, + { + "epoch": 4.252615844544096, + "grad_norm": 1.0262635946273804, + "learning_rate": 0.00046842301943198807, + "loss": 3.5114, + "step": 62590 + }, + { + "epoch": 4.252955564614758, + "grad_norm": 0.7360864281654358, + "learning_rate": 0.0004683805544231553, + "loss": 3.4764, + "step": 62595 + }, + { + "epoch": 4.253295284685419, + "grad_norm": 0.9645975232124329, + "learning_rate": 0.00046833808941432257, + "loss": 3.4238, + "step": 62600 + }, + { + "epoch": 4.253635004756081, + "grad_norm": 0.962471604347229, + "learning_rate": 0.0004682956244054899, + "loss": 3.1454, + "step": 62605 + }, + { + "epoch": 4.253974724826743, + "grad_norm": 0.9216309189796448, + "learning_rate": 0.0004682531593966572, + "loss": 3.4826, + "step": 62610 + }, + { + "epoch": 4.254314444897404, + "grad_norm": 0.8070670962333679, + "learning_rate": 0.0004682106943878244, + "loss": 3.4024, + "step": 62615 + }, + { + "epoch": 4.254654164968066, + "grad_norm": 0.9592640995979309, + "learning_rate": 0.00046816822937899175, + "loss": 3.5089, + "step": 62620 + }, + { + "epoch": 4.254993885038728, + "grad_norm": 0.8739386796951294, + "learning_rate": 0.000468125764370159, + "loss": 3.5299, + "step": 62625 + }, + { + "epoch": 4.2553336051093895, + "grad_norm": 0.8468846678733826, + "learning_rate": 0.00046808329936132625, + "loss": 3.7183, + "step": 62630 + }, + { + "epoch": 4.255673325180052, + "grad_norm": 1.026724100112915, + "learning_rate": 0.0004680408343524936, + "loss": 3.4499, + "step": 62635 + }, + { + "epoch": 4.256013045250714, + "grad_norm": 1.1355243921279907, + "learning_rate": 0.00046799836934366087, + "loss": 3.1432, + "step": 62640 + }, + { + "epoch": 4.256352765321375, + "grad_norm": 0.8798123598098755, + "learning_rate": 0.0004679559043348281, + "loss": 3.4388, + "step": 62645 + }, + { + "epoch": 4.256692485392037, + "grad_norm": 0.8509390354156494, + "learning_rate": 0.0004679134393259954, + "loss": 3.4567, + "step": 62650 + }, + { + "epoch": 4.257032205462699, + "grad_norm": 0.8399722576141357, + "learning_rate": 0.0004678709743171627, + "loss": 3.6202, + "step": 62655 + }, + { + "epoch": 4.25737192553336, + "grad_norm": 0.7817937731742859, + "learning_rate": 0.00046782850930832993, + "loss": 3.2841, + "step": 62660 + }, + { + "epoch": 4.257711645604022, + "grad_norm": 0.723937451839447, + "learning_rate": 0.0004677860442994972, + "loss": 3.6308, + "step": 62665 + }, + { + "epoch": 4.258051365674684, + "grad_norm": 0.8056321740150452, + "learning_rate": 0.00046774357929066455, + "loss": 3.7713, + "step": 62670 + }, + { + "epoch": 4.2583910857453455, + "grad_norm": 0.863405704498291, + "learning_rate": 0.0004677011142818318, + "loss": 3.3284, + "step": 62675 + }, + { + "epoch": 4.258730805816008, + "grad_norm": 1.2852163314819336, + "learning_rate": 0.00046765864927299905, + "loss": 3.243, + "step": 62680 + }, + { + "epoch": 4.25907052588667, + "grad_norm": 0.855522096157074, + "learning_rate": 0.00046761618426416633, + "loss": 3.3942, + "step": 62685 + }, + { + "epoch": 4.259410245957331, + "grad_norm": 1.0875468254089355, + "learning_rate": 0.0004675737192553336, + "loss": 3.3366, + "step": 62690 + }, + { + "epoch": 4.259749966027993, + "grad_norm": 0.8120050430297852, + "learning_rate": 0.0004675312542465009, + "loss": 3.3214, + "step": 62695 + }, + { + "epoch": 4.260089686098655, + "grad_norm": 0.7442243099212646, + "learning_rate": 0.0004674887892376682, + "loss": 3.531, + "step": 62700 + }, + { + "epoch": 4.260429406169316, + "grad_norm": 1.1789648532867432, + "learning_rate": 0.00046744632422883545, + "loss": 3.6601, + "step": 62705 + }, + { + "epoch": 4.260769126239978, + "grad_norm": 0.9964724183082581, + "learning_rate": 0.00046740385922000273, + "loss": 3.6971, + "step": 62710 + }, + { + "epoch": 4.26110884631064, + "grad_norm": 0.8174118995666504, + "learning_rate": 0.00046736139421117, + "loss": 3.6148, + "step": 62715 + }, + { + "epoch": 4.2614485663813015, + "grad_norm": 1.0132817029953003, + "learning_rate": 0.00046731892920233724, + "loss": 3.2955, + "step": 62720 + }, + { + "epoch": 4.261788286451964, + "grad_norm": 1.0247809886932373, + "learning_rate": 0.0004672764641935046, + "loss": 3.5491, + "step": 62725 + }, + { + "epoch": 4.262128006522626, + "grad_norm": 1.5601515769958496, + "learning_rate": 0.00046723399918467185, + "loss": 3.6784, + "step": 62730 + }, + { + "epoch": 4.262467726593287, + "grad_norm": 0.883522629737854, + "learning_rate": 0.0004671915341758391, + "loss": 3.5579, + "step": 62735 + }, + { + "epoch": 4.262807446663949, + "grad_norm": 0.8184067010879517, + "learning_rate": 0.0004671490691670064, + "loss": 3.453, + "step": 62740 + }, + { + "epoch": 4.263147166734611, + "grad_norm": 0.9404013752937317, + "learning_rate": 0.0004671066041581737, + "loss": 3.6703, + "step": 62745 + }, + { + "epoch": 4.263486886805272, + "grad_norm": 1.2359004020690918, + "learning_rate": 0.0004670641391493409, + "loss": 3.4121, + "step": 62750 + }, + { + "epoch": 4.263826606875934, + "grad_norm": 0.992039680480957, + "learning_rate": 0.0004670216741405082, + "loss": 3.3679, + "step": 62755 + }, + { + "epoch": 4.264166326946596, + "grad_norm": 0.8321199417114258, + "learning_rate": 0.00046697920913167553, + "loss": 3.6196, + "step": 62760 + }, + { + "epoch": 4.2645060470172576, + "grad_norm": 0.8964241147041321, + "learning_rate": 0.00046693674412284276, + "loss": 3.4838, + "step": 62765 + }, + { + "epoch": 4.26484576708792, + "grad_norm": 0.7762098908424377, + "learning_rate": 0.00046689427911401004, + "loss": 3.4609, + "step": 62770 + }, + { + "epoch": 4.265185487158582, + "grad_norm": 0.849297046661377, + "learning_rate": 0.0004668518141051774, + "loss": 3.5284, + "step": 62775 + }, + { + "epoch": 4.265525207229243, + "grad_norm": 1.0686362981796265, + "learning_rate": 0.00046680934909634466, + "loss": 3.2587, + "step": 62780 + }, + { + "epoch": 4.265864927299905, + "grad_norm": 1.022510290145874, + "learning_rate": 0.0004667668840875119, + "loss": 3.4365, + "step": 62785 + }, + { + "epoch": 4.266204647370567, + "grad_norm": 1.0102448463439941, + "learning_rate": 0.00046672441907867916, + "loss": 3.4856, + "step": 62790 + }, + { + "epoch": 4.266544367441228, + "grad_norm": 0.9376881718635559, + "learning_rate": 0.0004666819540698465, + "loss": 3.4935, + "step": 62795 + }, + { + "epoch": 4.26688408751189, + "grad_norm": 1.0369386672973633, + "learning_rate": 0.0004666394890610137, + "loss": 3.4773, + "step": 62800 + }, + { + "epoch": 4.267223807582552, + "grad_norm": 0.9463576078414917, + "learning_rate": 0.000466597024052181, + "loss": 3.4082, + "step": 62805 + }, + { + "epoch": 4.267563527653214, + "grad_norm": 1.0640326738357544, + "learning_rate": 0.00046655455904334834, + "loss": 3.3137, + "step": 62810 + }, + { + "epoch": 4.267903247723876, + "grad_norm": 0.9896097779273987, + "learning_rate": 0.00046651209403451556, + "loss": 3.7832, + "step": 62815 + }, + { + "epoch": 4.268242967794538, + "grad_norm": 1.0026336908340454, + "learning_rate": 0.00046646962902568284, + "loss": 3.1124, + "step": 62820 + }, + { + "epoch": 4.268582687865199, + "grad_norm": 0.9170539975166321, + "learning_rate": 0.0004664271640168501, + "loss": 3.5988, + "step": 62825 + }, + { + "epoch": 4.268922407935861, + "grad_norm": 0.8372460007667542, + "learning_rate": 0.0004663846990080174, + "loss": 3.7902, + "step": 62830 + }, + { + "epoch": 4.269262128006522, + "grad_norm": 0.8830277919769287, + "learning_rate": 0.0004663422339991847, + "loss": 3.5153, + "step": 62835 + }, + { + "epoch": 4.269601848077184, + "grad_norm": 0.7884312868118286, + "learning_rate": 0.00046629976899035196, + "loss": 3.5536, + "step": 62840 + }, + { + "epoch": 4.269941568147846, + "grad_norm": 0.9174456596374512, + "learning_rate": 0.00046625730398151924, + "loss": 3.6974, + "step": 62845 + }, + { + "epoch": 4.2702812882185075, + "grad_norm": 0.7554647326469421, + "learning_rate": 0.0004662148389726865, + "loss": 3.5009, + "step": 62850 + }, + { + "epoch": 4.27062100828917, + "grad_norm": 0.8156687021255493, + "learning_rate": 0.0004661723739638538, + "loss": 3.6231, + "step": 62855 + }, + { + "epoch": 4.270960728359832, + "grad_norm": 0.8334752321243286, + "learning_rate": 0.00046612990895502103, + "loss": 3.4368, + "step": 62860 + }, + { + "epoch": 4.271300448430493, + "grad_norm": 0.8297866582870483, + "learning_rate": 0.00046608744394618836, + "loss": 3.6961, + "step": 62865 + }, + { + "epoch": 4.271640168501155, + "grad_norm": 0.7849826216697693, + "learning_rate": 0.00046604497893735564, + "loss": 3.4026, + "step": 62870 + }, + { + "epoch": 4.271979888571817, + "grad_norm": 0.6912524700164795, + "learning_rate": 0.00046600251392852287, + "loss": 3.5435, + "step": 62875 + }, + { + "epoch": 4.272319608642478, + "grad_norm": 0.9883927702903748, + "learning_rate": 0.0004659600489196902, + "loss": 3.1717, + "step": 62880 + }, + { + "epoch": 4.27265932871314, + "grad_norm": 1.0142556428909302, + "learning_rate": 0.0004659175839108575, + "loss": 3.4933, + "step": 62885 + }, + { + "epoch": 4.272999048783802, + "grad_norm": 0.9743466973304749, + "learning_rate": 0.0004658751189020247, + "loss": 3.6622, + "step": 62890 + }, + { + "epoch": 4.2733387688544635, + "grad_norm": 1.0876703262329102, + "learning_rate": 0.000465832653893192, + "loss": 3.4563, + "step": 62895 + }, + { + "epoch": 4.273678488925126, + "grad_norm": 0.8752255439758301, + "learning_rate": 0.0004657901888843593, + "loss": 3.2035, + "step": 62900 + }, + { + "epoch": 4.274018208995788, + "grad_norm": 2.4211843013763428, + "learning_rate": 0.00046574772387552655, + "loss": 3.4968, + "step": 62905 + }, + { + "epoch": 4.274357929066449, + "grad_norm": 0.7658143043518066, + "learning_rate": 0.00046570525886669383, + "loss": 3.3747, + "step": 62910 + }, + { + "epoch": 4.274697649137111, + "grad_norm": 0.9630666971206665, + "learning_rate": 0.00046566279385786116, + "loss": 3.3537, + "step": 62915 + }, + { + "epoch": 4.275037369207773, + "grad_norm": 0.9136813879013062, + "learning_rate": 0.0004656203288490284, + "loss": 3.6859, + "step": 62920 + }, + { + "epoch": 4.275377089278434, + "grad_norm": 0.8126678466796875, + "learning_rate": 0.00046557786384019567, + "loss": 3.5053, + "step": 62925 + }, + { + "epoch": 4.275716809349096, + "grad_norm": 0.9543166160583496, + "learning_rate": 0.000465535398831363, + "loss": 3.5323, + "step": 62930 + }, + { + "epoch": 4.276056529419758, + "grad_norm": 0.8145275712013245, + "learning_rate": 0.00046549293382253023, + "loss": 3.3303, + "step": 62935 + }, + { + "epoch": 4.2763962494904195, + "grad_norm": 0.8845632076263428, + "learning_rate": 0.0004654504688136975, + "loss": 3.5354, + "step": 62940 + }, + { + "epoch": 4.276735969561082, + "grad_norm": 1.0814285278320312, + "learning_rate": 0.0004654080038048648, + "loss": 3.3671, + "step": 62945 + }, + { + "epoch": 4.277075689631744, + "grad_norm": 0.8797786831855774, + "learning_rate": 0.0004653655387960321, + "loss": 3.4992, + "step": 62950 + }, + { + "epoch": 4.277415409702405, + "grad_norm": 0.9148866534233093, + "learning_rate": 0.00046532307378719935, + "loss": 3.5765, + "step": 62955 + }, + { + "epoch": 4.277755129773067, + "grad_norm": 1.0035520792007446, + "learning_rate": 0.00046528060877836663, + "loss": 3.3923, + "step": 62960 + }, + { + "epoch": 4.278094849843729, + "grad_norm": 0.8205432295799255, + "learning_rate": 0.00046523814376953396, + "loss": 3.6367, + "step": 62965 + }, + { + "epoch": 4.27843456991439, + "grad_norm": 0.942388117313385, + "learning_rate": 0.0004651956787607012, + "loss": 3.4002, + "step": 62970 + }, + { + "epoch": 4.278774289985052, + "grad_norm": 0.8025549054145813, + "learning_rate": 0.00046515321375186847, + "loss": 3.503, + "step": 62975 + }, + { + "epoch": 4.279114010055714, + "grad_norm": 0.9052891135215759, + "learning_rate": 0.00046511074874303575, + "loss": 3.6461, + "step": 62980 + }, + { + "epoch": 4.2794537301263755, + "grad_norm": 0.7075366973876953, + "learning_rate": 0.00046506828373420303, + "loss": 3.3896, + "step": 62985 + }, + { + "epoch": 4.279793450197038, + "grad_norm": 0.7186359167098999, + "learning_rate": 0.0004650258187253703, + "loss": 3.3471, + "step": 62990 + }, + { + "epoch": 4.2801331702677, + "grad_norm": 0.8579268455505371, + "learning_rate": 0.0004649833537165376, + "loss": 3.1928, + "step": 62995 + }, + { + "epoch": 4.280472890338361, + "grad_norm": 1.119499683380127, + "learning_rate": 0.00046494088870770487, + "loss": 3.4639, + "step": 63000 + }, + { + "epoch": 4.280812610409023, + "grad_norm": 0.9440551400184631, + "learning_rate": 0.00046489842369887215, + "loss": 3.7019, + "step": 63005 + }, + { + "epoch": 4.281152330479685, + "grad_norm": 1.0752047300338745, + "learning_rate": 0.00046485595869003943, + "loss": 3.6316, + "step": 63010 + }, + { + "epoch": 4.281492050550346, + "grad_norm": 0.8332939147949219, + "learning_rate": 0.00046481349368120666, + "loss": 3.5088, + "step": 63015 + }, + { + "epoch": 4.281831770621008, + "grad_norm": 1.0204322338104248, + "learning_rate": 0.000464771028672374, + "loss": 3.6723, + "step": 63020 + }, + { + "epoch": 4.28217149069167, + "grad_norm": 0.7515679001808167, + "learning_rate": 0.00046472856366354127, + "loss": 3.5883, + "step": 63025 + }, + { + "epoch": 4.2825112107623315, + "grad_norm": 1.017456293106079, + "learning_rate": 0.0004646860986547085, + "loss": 3.5022, + "step": 63030 + }, + { + "epoch": 4.282850930832994, + "grad_norm": 0.7432921528816223, + "learning_rate": 0.00046464363364587583, + "loss": 3.6638, + "step": 63035 + }, + { + "epoch": 4.283190650903656, + "grad_norm": 0.9040016531944275, + "learning_rate": 0.0004646011686370431, + "loss": 3.5003, + "step": 63040 + }, + { + "epoch": 4.283530370974317, + "grad_norm": 0.8391094207763672, + "learning_rate": 0.00046455870362821034, + "loss": 3.5795, + "step": 63045 + }, + { + "epoch": 4.283870091044979, + "grad_norm": 0.8326451778411865, + "learning_rate": 0.0004645162386193776, + "loss": 3.4526, + "step": 63050 + }, + { + "epoch": 4.284209811115641, + "grad_norm": 0.7569173574447632, + "learning_rate": 0.00046447377361054495, + "loss": 3.6632, + "step": 63055 + }, + { + "epoch": 4.284549531186302, + "grad_norm": 0.9134597778320312, + "learning_rate": 0.0004644313086017122, + "loss": 3.6009, + "step": 63060 + }, + { + "epoch": 4.284889251256964, + "grad_norm": 0.9937013983726501, + "learning_rate": 0.00046438884359287946, + "loss": 3.667, + "step": 63065 + }, + { + "epoch": 4.285228971327626, + "grad_norm": 0.8157731294631958, + "learning_rate": 0.0004643463785840468, + "loss": 3.4999, + "step": 63070 + }, + { + "epoch": 4.2855686913982876, + "grad_norm": 1.0677434206008911, + "learning_rate": 0.000464303913575214, + "loss": 3.7102, + "step": 63075 + }, + { + "epoch": 4.28590841146895, + "grad_norm": 1.0701124668121338, + "learning_rate": 0.0004642614485663813, + "loss": 3.7098, + "step": 63080 + }, + { + "epoch": 4.286248131539612, + "grad_norm": 0.8966395258903503, + "learning_rate": 0.0004642189835575486, + "loss": 3.4293, + "step": 63085 + }, + { + "epoch": 4.286587851610273, + "grad_norm": 0.8636654019355774, + "learning_rate": 0.00046417651854871586, + "loss": 3.3805, + "step": 63090 + }, + { + "epoch": 4.286927571680935, + "grad_norm": 0.9314202070236206, + "learning_rate": 0.00046413405353988314, + "loss": 2.9326, + "step": 63095 + }, + { + "epoch": 4.287267291751597, + "grad_norm": 0.7656220197677612, + "learning_rate": 0.0004640915885310504, + "loss": 3.552, + "step": 63100 + }, + { + "epoch": 4.287607011822258, + "grad_norm": 0.8057489991188049, + "learning_rate": 0.0004640491235222177, + "loss": 3.3063, + "step": 63105 + }, + { + "epoch": 4.28794673189292, + "grad_norm": 1.1799037456512451, + "learning_rate": 0.000464006658513385, + "loss": 3.486, + "step": 63110 + }, + { + "epoch": 4.288286451963582, + "grad_norm": 0.7449032068252563, + "learning_rate": 0.00046396419350455226, + "loss": 3.4772, + "step": 63115 + }, + { + "epoch": 4.288626172034244, + "grad_norm": 1.057191252708435, + "learning_rate": 0.00046392172849571954, + "loss": 3.5552, + "step": 63120 + }, + { + "epoch": 4.288965892104906, + "grad_norm": 0.849864661693573, + "learning_rate": 0.0004638792634868868, + "loss": 3.8615, + "step": 63125 + }, + { + "epoch": 4.289305612175568, + "grad_norm": 1.073035478591919, + "learning_rate": 0.0004638367984780541, + "loss": 3.4355, + "step": 63130 + }, + { + "epoch": 4.289645332246229, + "grad_norm": 0.9139671325683594, + "learning_rate": 0.0004637943334692214, + "loss": 3.5101, + "step": 63135 + }, + { + "epoch": 4.289985052316891, + "grad_norm": 0.9718029499053955, + "learning_rate": 0.00046375186846038866, + "loss": 3.3728, + "step": 63140 + }, + { + "epoch": 4.290324772387553, + "grad_norm": 0.7044605016708374, + "learning_rate": 0.00046370940345155594, + "loss": 3.5521, + "step": 63145 + }, + { + "epoch": 4.290664492458214, + "grad_norm": 0.9600394368171692, + "learning_rate": 0.0004636669384427232, + "loss": 3.7049, + "step": 63150 + }, + { + "epoch": 4.291004212528876, + "grad_norm": 1.0087790489196777, + "learning_rate": 0.00046362447343389044, + "loss": 3.6528, + "step": 63155 + }, + { + "epoch": 4.291343932599538, + "grad_norm": 0.8499536514282227, + "learning_rate": 0.0004635820084250578, + "loss": 3.5781, + "step": 63160 + }, + { + "epoch": 4.2916836526702, + "grad_norm": 1.048994541168213, + "learning_rate": 0.00046353954341622506, + "loss": 3.2282, + "step": 63165 + }, + { + "epoch": 4.292023372740862, + "grad_norm": 1.14274001121521, + "learning_rate": 0.0004634970784073923, + "loss": 3.4732, + "step": 63170 + }, + { + "epoch": 4.292363092811524, + "grad_norm": 0.8861408829689026, + "learning_rate": 0.0004634546133985596, + "loss": 3.4481, + "step": 63175 + }, + { + "epoch": 4.292702812882185, + "grad_norm": 0.9583660364151001, + "learning_rate": 0.0004634121483897269, + "loss": 3.4129, + "step": 63180 + }, + { + "epoch": 4.293042532952847, + "grad_norm": 0.9439297318458557, + "learning_rate": 0.0004633696833808941, + "loss": 3.501, + "step": 63185 + }, + { + "epoch": 4.293382253023509, + "grad_norm": 1.1283248662948608, + "learning_rate": 0.00046332721837206146, + "loss": 3.4876, + "step": 63190 + }, + { + "epoch": 4.29372197309417, + "grad_norm": 0.702498733997345, + "learning_rate": 0.00046328475336322874, + "loss": 3.7509, + "step": 63195 + }, + { + "epoch": 4.294061693164832, + "grad_norm": 1.118389368057251, + "learning_rate": 0.00046324228835439596, + "loss": 3.285, + "step": 63200 + }, + { + "epoch": 4.294401413235494, + "grad_norm": 1.0719313621520996, + "learning_rate": 0.00046319982334556324, + "loss": 3.5133, + "step": 63205 + }, + { + "epoch": 4.294741133306156, + "grad_norm": 0.9516918063163757, + "learning_rate": 0.0004631573583367306, + "loss": 3.4966, + "step": 63210 + }, + { + "epoch": 4.295080853376818, + "grad_norm": 0.9275661110877991, + "learning_rate": 0.0004631148933278978, + "loss": 3.5664, + "step": 63215 + }, + { + "epoch": 4.29542057344748, + "grad_norm": 0.935204267501831, + "learning_rate": 0.0004630724283190651, + "loss": 3.6038, + "step": 63220 + }, + { + "epoch": 4.295760293518141, + "grad_norm": 1.0143736600875854, + "learning_rate": 0.0004630299633102324, + "loss": 3.4991, + "step": 63225 + }, + { + "epoch": 4.296100013588803, + "grad_norm": 0.8882645964622498, + "learning_rate": 0.00046298749830139965, + "loss": 3.5197, + "step": 63230 + }, + { + "epoch": 4.296439733659464, + "grad_norm": 1.0124783515930176, + "learning_rate": 0.0004629450332925669, + "loss": 3.3145, + "step": 63235 + }, + { + "epoch": 4.296779453730126, + "grad_norm": 1.1305333375930786, + "learning_rate": 0.0004629025682837342, + "loss": 3.3406, + "step": 63240 + }, + { + "epoch": 4.297119173800788, + "grad_norm": 0.7439259886741638, + "learning_rate": 0.0004628601032749015, + "loss": 3.5638, + "step": 63245 + }, + { + "epoch": 4.2974588938714495, + "grad_norm": 0.757042407989502, + "learning_rate": 0.00046281763826606877, + "loss": 3.5936, + "step": 63250 + }, + { + "epoch": 4.297798613942112, + "grad_norm": 0.9157543182373047, + "learning_rate": 0.00046277517325723605, + "loss": 3.4952, + "step": 63255 + }, + { + "epoch": 4.298138334012774, + "grad_norm": 1.0470733642578125, + "learning_rate": 0.0004627327082484033, + "loss": 3.4363, + "step": 63260 + }, + { + "epoch": 4.298478054083435, + "grad_norm": 1.0580462217330933, + "learning_rate": 0.0004626902432395706, + "loss": 3.7582, + "step": 63265 + }, + { + "epoch": 4.298817774154097, + "grad_norm": 1.00447678565979, + "learning_rate": 0.0004626477782307379, + "loss": 3.3235, + "step": 63270 + }, + { + "epoch": 4.299157494224759, + "grad_norm": 0.932273268699646, + "learning_rate": 0.0004626053132219051, + "loss": 3.6634, + "step": 63275 + }, + { + "epoch": 4.29949721429542, + "grad_norm": 0.8090956807136536, + "learning_rate": 0.00046256284821307245, + "loss": 3.5485, + "step": 63280 + }, + { + "epoch": 4.299836934366082, + "grad_norm": 1.0920292139053345, + "learning_rate": 0.0004625203832042397, + "loss": 3.514, + "step": 63285 + }, + { + "epoch": 4.300176654436744, + "grad_norm": 0.7680724859237671, + "learning_rate": 0.000462477918195407, + "loss": 3.6105, + "step": 63290 + }, + { + "epoch": 4.3005163745074055, + "grad_norm": 0.8498972654342651, + "learning_rate": 0.0004624354531865743, + "loss": 3.6357, + "step": 63295 + }, + { + "epoch": 4.300856094578068, + "grad_norm": 1.0505073070526123, + "learning_rate": 0.00046239298817774157, + "loss": 3.3726, + "step": 63300 + }, + { + "epoch": 4.30119581464873, + "grad_norm": 0.9010647535324097, + "learning_rate": 0.00046235052316890885, + "loss": 3.6079, + "step": 63305 + }, + { + "epoch": 4.301535534719391, + "grad_norm": 1.0109111070632935, + "learning_rate": 0.00046230805816007607, + "loss": 3.208, + "step": 63310 + }, + { + "epoch": 4.301875254790053, + "grad_norm": 0.9397577047348022, + "learning_rate": 0.0004622655931512434, + "loss": 3.4915, + "step": 63315 + }, + { + "epoch": 4.302214974860715, + "grad_norm": 0.8470300436019897, + "learning_rate": 0.0004622231281424107, + "loss": 3.6991, + "step": 63320 + }, + { + "epoch": 4.302554694931376, + "grad_norm": 0.7456369996070862, + "learning_rate": 0.0004621806631335779, + "loss": 3.4182, + "step": 63325 + }, + { + "epoch": 4.302894415002038, + "grad_norm": 0.9127781987190247, + "learning_rate": 0.00046213819812474525, + "loss": 3.5517, + "step": 63330 + }, + { + "epoch": 4.3032341350727, + "grad_norm": 0.9428461790084839, + "learning_rate": 0.0004620957331159125, + "loss": 3.3641, + "step": 63335 + }, + { + "epoch": 4.3035738551433615, + "grad_norm": 1.1578031778335571, + "learning_rate": 0.00046205326810707975, + "loss": 3.3254, + "step": 63340 + }, + { + "epoch": 4.303913575214024, + "grad_norm": 0.7605162262916565, + "learning_rate": 0.00046201080309824703, + "loss": 3.1745, + "step": 63345 + }, + { + "epoch": 4.304253295284686, + "grad_norm": 0.8030891418457031, + "learning_rate": 0.00046196833808941437, + "loss": 3.4805, + "step": 63350 + }, + { + "epoch": 4.304593015355347, + "grad_norm": 0.7844246625900269, + "learning_rate": 0.0004619258730805816, + "loss": 3.3293, + "step": 63355 + }, + { + "epoch": 4.304932735426009, + "grad_norm": 0.9471333026885986, + "learning_rate": 0.0004618834080717489, + "loss": 3.8534, + "step": 63360 + }, + { + "epoch": 4.305272455496671, + "grad_norm": 1.0193318128585815, + "learning_rate": 0.0004618409430629162, + "loss": 3.6415, + "step": 63365 + }, + { + "epoch": 4.305612175567332, + "grad_norm": 0.9798857569694519, + "learning_rate": 0.00046179847805408343, + "loss": 3.5176, + "step": 63370 + }, + { + "epoch": 4.305951895637994, + "grad_norm": 0.8460768461227417, + "learning_rate": 0.0004617560130452507, + "loss": 3.5642, + "step": 63375 + }, + { + "epoch": 4.306291615708656, + "grad_norm": 0.8960776329040527, + "learning_rate": 0.000461713548036418, + "loss": 3.223, + "step": 63380 + }, + { + "epoch": 4.306631335779318, + "grad_norm": 0.8703097105026245, + "learning_rate": 0.0004616710830275853, + "loss": 3.3396, + "step": 63385 + }, + { + "epoch": 4.30697105584998, + "grad_norm": 0.9503531455993652, + "learning_rate": 0.00046162861801875255, + "loss": 3.5174, + "step": 63390 + }, + { + "epoch": 4.307310775920642, + "grad_norm": 0.8498976230621338, + "learning_rate": 0.00046158615300991983, + "loss": 3.4601, + "step": 63395 + }, + { + "epoch": 4.307650495991303, + "grad_norm": 0.8161876201629639, + "learning_rate": 0.0004615436880010871, + "loss": 3.5758, + "step": 63400 + }, + { + "epoch": 4.307990216061965, + "grad_norm": 0.7444183826446533, + "learning_rate": 0.0004615012229922544, + "loss": 3.283, + "step": 63405 + }, + { + "epoch": 4.308329936132627, + "grad_norm": 0.7839939594268799, + "learning_rate": 0.0004614587579834217, + "loss": 3.514, + "step": 63410 + }, + { + "epoch": 4.308669656203288, + "grad_norm": 0.7391408085823059, + "learning_rate": 0.0004614162929745889, + "loss": 3.6163, + "step": 63415 + }, + { + "epoch": 4.30900937627395, + "grad_norm": 0.9623317718505859, + "learning_rate": 0.00046137382796575623, + "loss": 3.5905, + "step": 63420 + }, + { + "epoch": 4.309349096344612, + "grad_norm": 0.785484790802002, + "learning_rate": 0.0004613313629569235, + "loss": 3.5826, + "step": 63425 + }, + { + "epoch": 4.309688816415274, + "grad_norm": 1.0002474784851074, + "learning_rate": 0.00046128889794809074, + "loss": 3.6054, + "step": 63430 + }, + { + "epoch": 4.310028536485936, + "grad_norm": 0.732146680355072, + "learning_rate": 0.0004612464329392581, + "loss": 3.5689, + "step": 63435 + }, + { + "epoch": 4.310368256556598, + "grad_norm": 1.2116374969482422, + "learning_rate": 0.00046120396793042535, + "loss": 3.4533, + "step": 63440 + }, + { + "epoch": 4.310707976627259, + "grad_norm": 0.9260076284408569, + "learning_rate": 0.0004611615029215926, + "loss": 3.3293, + "step": 63445 + }, + { + "epoch": 4.311047696697921, + "grad_norm": 0.895195484161377, + "learning_rate": 0.00046111903791275986, + "loss": 3.5937, + "step": 63450 + }, + { + "epoch": 4.311387416768583, + "grad_norm": 0.7954394817352295, + "learning_rate": 0.0004610765729039272, + "loss": 3.7715, + "step": 63455 + }, + { + "epoch": 4.311727136839244, + "grad_norm": 1.0760308504104614, + "learning_rate": 0.0004610341078950945, + "loss": 3.6492, + "step": 63460 + }, + { + "epoch": 4.312066856909906, + "grad_norm": 0.8424020409584045, + "learning_rate": 0.0004609916428862617, + "loss": 3.3422, + "step": 63465 + }, + { + "epoch": 4.312406576980568, + "grad_norm": 0.893039345741272, + "learning_rate": 0.00046094917787742903, + "loss": 3.5619, + "step": 63470 + }, + { + "epoch": 4.31274629705123, + "grad_norm": 1.2063446044921875, + "learning_rate": 0.0004609067128685963, + "loss": 3.691, + "step": 63475 + }, + { + "epoch": 4.313086017121892, + "grad_norm": 0.9133750200271606, + "learning_rate": 0.00046086424785976354, + "loss": 3.5499, + "step": 63480 + }, + { + "epoch": 4.313425737192554, + "grad_norm": 6.863533020019531, + "learning_rate": 0.0004608217828509309, + "loss": 3.459, + "step": 63485 + }, + { + "epoch": 4.313765457263215, + "grad_norm": 0.8867606520652771, + "learning_rate": 0.00046077931784209815, + "loss": 3.561, + "step": 63490 + }, + { + "epoch": 4.314105177333877, + "grad_norm": 0.850654125213623, + "learning_rate": 0.0004607368528332654, + "loss": 3.4441, + "step": 63495 + }, + { + "epoch": 4.314444897404539, + "grad_norm": 0.7537965178489685, + "learning_rate": 0.00046069438782443266, + "loss": 3.6557, + "step": 63500 + }, + { + "epoch": 4.3147846174752, + "grad_norm": 0.8483400940895081, + "learning_rate": 0.0004606519228156, + "loss": 3.5814, + "step": 63505 + }, + { + "epoch": 4.315124337545862, + "grad_norm": 0.8274759650230408, + "learning_rate": 0.0004606094578067672, + "loss": 3.5797, + "step": 63510 + }, + { + "epoch": 4.3154640576165235, + "grad_norm": 1.2165584564208984, + "learning_rate": 0.0004605669927979345, + "loss": 3.751, + "step": 63515 + }, + { + "epoch": 4.315803777687186, + "grad_norm": 0.9291759133338928, + "learning_rate": 0.00046052452778910184, + "loss": 3.3624, + "step": 63520 + }, + { + "epoch": 4.316143497757848, + "grad_norm": 0.7642979025840759, + "learning_rate": 0.00046048206278026906, + "loss": 3.527, + "step": 63525 + }, + { + "epoch": 4.316483217828509, + "grad_norm": 0.8614895343780518, + "learning_rate": 0.00046043959777143634, + "loss": 3.264, + "step": 63530 + }, + { + "epoch": 4.316822937899171, + "grad_norm": 0.9371413588523865, + "learning_rate": 0.0004603971327626036, + "loss": 3.5584, + "step": 63535 + }, + { + "epoch": 4.317162657969833, + "grad_norm": 0.7602739334106445, + "learning_rate": 0.0004603546677537709, + "loss": 3.4627, + "step": 63540 + }, + { + "epoch": 4.317502378040494, + "grad_norm": 0.8997742533683777, + "learning_rate": 0.0004603122027449382, + "loss": 3.5373, + "step": 63545 + }, + { + "epoch": 4.317842098111156, + "grad_norm": 0.8409783840179443, + "learning_rate": 0.00046026973773610546, + "loss": 3.4716, + "step": 63550 + }, + { + "epoch": 4.318181818181818, + "grad_norm": 0.8912867307662964, + "learning_rate": 0.00046022727272727274, + "loss": 3.6941, + "step": 63555 + }, + { + "epoch": 4.3185215382524795, + "grad_norm": 0.8730382323265076, + "learning_rate": 0.00046018480771844, + "loss": 3.5182, + "step": 63560 + }, + { + "epoch": 4.318861258323142, + "grad_norm": 1.0314522981643677, + "learning_rate": 0.0004601423427096073, + "loss": 3.5165, + "step": 63565 + }, + { + "epoch": 4.319200978393804, + "grad_norm": 0.9358851909637451, + "learning_rate": 0.00046009987770077453, + "loss": 3.5095, + "step": 63570 + }, + { + "epoch": 4.319540698464465, + "grad_norm": 0.888331949710846, + "learning_rate": 0.00046005741269194186, + "loss": 3.3407, + "step": 63575 + }, + { + "epoch": 4.319880418535127, + "grad_norm": 0.7379451990127563, + "learning_rate": 0.00046001494768310914, + "loss": 3.675, + "step": 63580 + }, + { + "epoch": 4.320220138605789, + "grad_norm": 0.7439711689949036, + "learning_rate": 0.00045997248267427637, + "loss": 3.6333, + "step": 63585 + }, + { + "epoch": 4.32055985867645, + "grad_norm": 1.0048997402191162, + "learning_rate": 0.0004599300176654437, + "loss": 3.3139, + "step": 63590 + }, + { + "epoch": 4.320899578747112, + "grad_norm": 1.190597653388977, + "learning_rate": 0.000459887552656611, + "loss": 3.2786, + "step": 63595 + }, + { + "epoch": 4.321239298817774, + "grad_norm": 0.9907605051994324, + "learning_rate": 0.0004598450876477782, + "loss": 3.4597, + "step": 63600 + }, + { + "epoch": 4.3215790188884355, + "grad_norm": 0.8446999788284302, + "learning_rate": 0.0004598026226389455, + "loss": 3.5349, + "step": 63605 + }, + { + "epoch": 4.321918738959098, + "grad_norm": 1.0107758045196533, + "learning_rate": 0.0004597601576301128, + "loss": 3.4014, + "step": 63610 + }, + { + "epoch": 4.32225845902976, + "grad_norm": 0.9221784472465515, + "learning_rate": 0.00045971769262128005, + "loss": 3.5041, + "step": 63615 + }, + { + "epoch": 4.322598179100421, + "grad_norm": 0.8770763874053955, + "learning_rate": 0.00045967522761244733, + "loss": 3.4542, + "step": 63620 + }, + { + "epoch": 4.322937899171083, + "grad_norm": 0.8623422384262085, + "learning_rate": 0.00045963276260361466, + "loss": 3.4452, + "step": 63625 + }, + { + "epoch": 4.323277619241745, + "grad_norm": 1.0108615159988403, + "learning_rate": 0.00045959029759478194, + "loss": 3.6085, + "step": 63630 + }, + { + "epoch": 4.323617339312406, + "grad_norm": 0.9481383562088013, + "learning_rate": 0.00045954783258594917, + "loss": 3.2311, + "step": 63635 + }, + { + "epoch": 4.323957059383068, + "grad_norm": 0.7505761981010437, + "learning_rate": 0.00045950536757711645, + "loss": 3.5852, + "step": 63640 + }, + { + "epoch": 4.32429677945373, + "grad_norm": 0.9163798093795776, + "learning_rate": 0.0004594629025682838, + "loss": 3.5715, + "step": 63645 + }, + { + "epoch": 4.3246364995243916, + "grad_norm": 0.7255598306655884, + "learning_rate": 0.000459420437559451, + "loss": 3.5082, + "step": 63650 + }, + { + "epoch": 4.324976219595054, + "grad_norm": 0.8730499744415283, + "learning_rate": 0.0004593779725506183, + "loss": 3.3951, + "step": 63655 + }, + { + "epoch": 4.325315939665716, + "grad_norm": 0.8823094964027405, + "learning_rate": 0.0004593355075417856, + "loss": 3.2884, + "step": 63660 + }, + { + "epoch": 4.325655659736377, + "grad_norm": 1.156896710395813, + "learning_rate": 0.00045929304253295285, + "loss": 3.7539, + "step": 63665 + }, + { + "epoch": 4.325995379807039, + "grad_norm": 0.8146911859512329, + "learning_rate": 0.00045925057752412013, + "loss": 3.4958, + "step": 63670 + }, + { + "epoch": 4.326335099877701, + "grad_norm": 0.7865837216377258, + "learning_rate": 0.0004592081125152874, + "loss": 3.5582, + "step": 63675 + }, + { + "epoch": 4.326674819948362, + "grad_norm": 0.9667829871177673, + "learning_rate": 0.0004591656475064547, + "loss": 3.8062, + "step": 63680 + }, + { + "epoch": 4.327014540019024, + "grad_norm": 0.726967453956604, + "learning_rate": 0.00045912318249762197, + "loss": 3.5864, + "step": 63685 + }, + { + "epoch": 4.327354260089686, + "grad_norm": 0.8275024890899658, + "learning_rate": 0.00045908071748878925, + "loss": 3.6488, + "step": 63690 + }, + { + "epoch": 4.327693980160348, + "grad_norm": 0.9242106080055237, + "learning_rate": 0.00045903825247995653, + "loss": 3.5204, + "step": 63695 + }, + { + "epoch": 4.32803370023101, + "grad_norm": 0.8766757249832153, + "learning_rate": 0.0004589957874711238, + "loss": 3.2796, + "step": 63700 + }, + { + "epoch": 4.328373420301672, + "grad_norm": 0.8839744329452515, + "learning_rate": 0.0004589533224622911, + "loss": 3.3753, + "step": 63705 + }, + { + "epoch": 4.328713140372333, + "grad_norm": 0.7990246415138245, + "learning_rate": 0.0004589108574534583, + "loss": 3.5364, + "step": 63710 + }, + { + "epoch": 4.329052860442995, + "grad_norm": 0.8658140301704407, + "learning_rate": 0.00045886839244462565, + "loss": 3.1989, + "step": 63715 + }, + { + "epoch": 4.329392580513657, + "grad_norm": 0.7308144569396973, + "learning_rate": 0.00045882592743579293, + "loss": 3.3909, + "step": 63720 + }, + { + "epoch": 4.329732300584318, + "grad_norm": 0.7782678604125977, + "learning_rate": 0.00045878346242696016, + "loss": 3.2512, + "step": 63725 + }, + { + "epoch": 4.33007202065498, + "grad_norm": 0.874247670173645, + "learning_rate": 0.0004587409974181275, + "loss": 3.3537, + "step": 63730 + }, + { + "epoch": 4.330411740725642, + "grad_norm": 0.975520670413971, + "learning_rate": 0.00045869853240929477, + "loss": 3.1277, + "step": 63735 + }, + { + "epoch": 4.330751460796304, + "grad_norm": 0.8464484810829163, + "learning_rate": 0.000458656067400462, + "loss": 3.3297, + "step": 63740 + }, + { + "epoch": 4.331091180866966, + "grad_norm": 0.9867607355117798, + "learning_rate": 0.0004586136023916293, + "loss": 3.1022, + "step": 63745 + }, + { + "epoch": 4.331430900937628, + "grad_norm": 0.8536937832832336, + "learning_rate": 0.0004585711373827966, + "loss": 3.6501, + "step": 63750 + }, + { + "epoch": 4.331770621008289, + "grad_norm": 0.9176635146141052, + "learning_rate": 0.00045852867237396384, + "loss": 3.3217, + "step": 63755 + }, + { + "epoch": 4.332110341078951, + "grad_norm": 0.8491596579551697, + "learning_rate": 0.0004584862073651311, + "loss": 3.3535, + "step": 63760 + }, + { + "epoch": 4.332450061149613, + "grad_norm": 1.1154425144195557, + "learning_rate": 0.00045844374235629845, + "loss": 3.2336, + "step": 63765 + }, + { + "epoch": 4.332789781220274, + "grad_norm": 0.7484056949615479, + "learning_rate": 0.0004584012773474657, + "loss": 3.4865, + "step": 63770 + }, + { + "epoch": 4.333129501290936, + "grad_norm": 1.037522315979004, + "learning_rate": 0.00045835881233863296, + "loss": 3.3765, + "step": 63775 + }, + { + "epoch": 4.333469221361598, + "grad_norm": 0.8413194417953491, + "learning_rate": 0.0004583163473298003, + "loss": 3.2379, + "step": 63780 + }, + { + "epoch": 4.33380894143226, + "grad_norm": 1.2168850898742676, + "learning_rate": 0.0004582738823209675, + "loss": 3.9675, + "step": 63785 + }, + { + "epoch": 4.334148661502922, + "grad_norm": 0.746817946434021, + "learning_rate": 0.0004582314173121348, + "loss": 3.4105, + "step": 63790 + }, + { + "epoch": 4.334488381573584, + "grad_norm": 1.2545156478881836, + "learning_rate": 0.0004581889523033021, + "loss": 3.4533, + "step": 63795 + }, + { + "epoch": 4.334828101644245, + "grad_norm": 0.9837030172348022, + "learning_rate": 0.0004581464872944694, + "loss": 3.1944, + "step": 63800 + }, + { + "epoch": 4.335167821714907, + "grad_norm": 0.6560445427894592, + "learning_rate": 0.00045810402228563664, + "loss": 3.5484, + "step": 63805 + }, + { + "epoch": 4.335507541785569, + "grad_norm": 1.5255625247955322, + "learning_rate": 0.0004580615572768039, + "loss": 3.116, + "step": 63810 + }, + { + "epoch": 4.33584726185623, + "grad_norm": 0.8826807141304016, + "learning_rate": 0.00045801909226797125, + "loss": 3.7089, + "step": 63815 + }, + { + "epoch": 4.336186981926892, + "grad_norm": 0.9641222953796387, + "learning_rate": 0.0004579766272591385, + "loss": 3.5097, + "step": 63820 + }, + { + "epoch": 4.336526701997554, + "grad_norm": 0.884658932685852, + "learning_rate": 0.00045793416225030576, + "loss": 3.7325, + "step": 63825 + }, + { + "epoch": 4.336866422068216, + "grad_norm": 0.8462684750556946, + "learning_rate": 0.00045789169724147304, + "loss": 3.622, + "step": 63830 + }, + { + "epoch": 4.337206142138878, + "grad_norm": 0.8645135164260864, + "learning_rate": 0.0004578492322326403, + "loss": 3.359, + "step": 63835 + }, + { + "epoch": 4.33754586220954, + "grad_norm": 0.7918854355812073, + "learning_rate": 0.0004578067672238076, + "loss": 3.483, + "step": 63840 + }, + { + "epoch": 4.337885582280201, + "grad_norm": 0.9172031283378601, + "learning_rate": 0.0004577643022149749, + "loss": 3.5068, + "step": 63845 + }, + { + "epoch": 4.338225302350863, + "grad_norm": 1.0127224922180176, + "learning_rate": 0.00045772183720614216, + "loss": 3.0608, + "step": 63850 + }, + { + "epoch": 4.338565022421525, + "grad_norm": 0.6827316284179688, + "learning_rate": 0.00045767937219730944, + "loss": 3.4395, + "step": 63855 + }, + { + "epoch": 4.338904742492186, + "grad_norm": 0.8347532749176025, + "learning_rate": 0.0004576369071884767, + "loss": 3.4962, + "step": 63860 + }, + { + "epoch": 4.339244462562848, + "grad_norm": 1.13916015625, + "learning_rate": 0.00045759444217964394, + "loss": 3.7297, + "step": 63865 + }, + { + "epoch": 4.33958418263351, + "grad_norm": 0.9371297359466553, + "learning_rate": 0.0004575519771708113, + "loss": 3.5323, + "step": 63870 + }, + { + "epoch": 4.339923902704172, + "grad_norm": 0.7954642176628113, + "learning_rate": 0.00045750951216197856, + "loss": 3.4789, + "step": 63875 + }, + { + "epoch": 4.340263622774834, + "grad_norm": 0.84361732006073, + "learning_rate": 0.0004574670471531458, + "loss": 3.7971, + "step": 63880 + }, + { + "epoch": 4.340603342845496, + "grad_norm": 1.1230800151824951, + "learning_rate": 0.0004574245821443131, + "loss": 3.6912, + "step": 63885 + }, + { + "epoch": 4.340943062916157, + "grad_norm": 0.7453582286834717, + "learning_rate": 0.0004573821171354804, + "loss": 3.5036, + "step": 63890 + }, + { + "epoch": 4.341282782986819, + "grad_norm": 0.689608097076416, + "learning_rate": 0.0004573396521266476, + "loss": 3.3715, + "step": 63895 + }, + { + "epoch": 4.341622503057481, + "grad_norm": 0.9573660492897034, + "learning_rate": 0.0004572971871178149, + "loss": 3.3909, + "step": 63900 + }, + { + "epoch": 4.341962223128142, + "grad_norm": 0.8617348074913025, + "learning_rate": 0.00045725472210898224, + "loss": 3.3257, + "step": 63905 + }, + { + "epoch": 4.342301943198804, + "grad_norm": 1.2585164308547974, + "learning_rate": 0.00045721225710014946, + "loss": 3.3754, + "step": 63910 + }, + { + "epoch": 4.3426416632694655, + "grad_norm": 0.8391521573066711, + "learning_rate": 0.00045716979209131674, + "loss": 3.5223, + "step": 63915 + }, + { + "epoch": 4.342981383340128, + "grad_norm": 0.935937762260437, + "learning_rate": 0.0004571273270824841, + "loss": 3.5075, + "step": 63920 + }, + { + "epoch": 4.34332110341079, + "grad_norm": 0.7743483185768127, + "learning_rate": 0.0004570848620736513, + "loss": 3.8233, + "step": 63925 + }, + { + "epoch": 4.343660823481451, + "grad_norm": 0.9026889801025391, + "learning_rate": 0.0004570423970648186, + "loss": 3.72, + "step": 63930 + }, + { + "epoch": 4.344000543552113, + "grad_norm": 0.8418740034103394, + "learning_rate": 0.00045699993205598587, + "loss": 3.6296, + "step": 63935 + }, + { + "epoch": 4.344340263622775, + "grad_norm": 0.9789199829101562, + "learning_rate": 0.00045695746704715315, + "loss": 3.5162, + "step": 63940 + }, + { + "epoch": 4.344679983693436, + "grad_norm": 1.1322083473205566, + "learning_rate": 0.0004569150020383204, + "loss": 3.3734, + "step": 63945 + }, + { + "epoch": 4.345019703764098, + "grad_norm": 0.7849491834640503, + "learning_rate": 0.0004568725370294877, + "loss": 3.4661, + "step": 63950 + }, + { + "epoch": 4.34535942383476, + "grad_norm": 0.8749169111251831, + "learning_rate": 0.000456830072020655, + "loss": 3.3094, + "step": 63955 + }, + { + "epoch": 4.345699143905422, + "grad_norm": 0.8175095319747925, + "learning_rate": 0.00045678760701182227, + "loss": 3.3418, + "step": 63960 + }, + { + "epoch": 4.346038863976084, + "grad_norm": 1.2671945095062256, + "learning_rate": 0.00045674514200298955, + "loss": 3.4372, + "step": 63965 + }, + { + "epoch": 4.346378584046746, + "grad_norm": 0.8950656652450562, + "learning_rate": 0.0004567026769941568, + "loss": 3.0568, + "step": 63970 + }, + { + "epoch": 4.346718304117407, + "grad_norm": 0.8960034847259521, + "learning_rate": 0.0004566602119853241, + "loss": 3.6081, + "step": 63975 + }, + { + "epoch": 4.347058024188069, + "grad_norm": 1.1570065021514893, + "learning_rate": 0.0004566177469764914, + "loss": 3.3209, + "step": 63980 + }, + { + "epoch": 4.347397744258731, + "grad_norm": 0.7255001068115234, + "learning_rate": 0.00045657528196765867, + "loss": 3.5867, + "step": 63985 + }, + { + "epoch": 4.347737464329392, + "grad_norm": 1.0038740634918213, + "learning_rate": 0.00045653281695882595, + "loss": 3.521, + "step": 63990 + }, + { + "epoch": 4.348077184400054, + "grad_norm": 1.0589931011199951, + "learning_rate": 0.0004564903519499932, + "loss": 3.6237, + "step": 63995 + }, + { + "epoch": 4.348416904470716, + "grad_norm": 0.7493960857391357, + "learning_rate": 0.0004564478869411605, + "loss": 3.2497, + "step": 64000 + }, + { + "epoch": 4.348756624541378, + "grad_norm": 1.1200355291366577, + "learning_rate": 0.00045640542193232773, + "loss": 3.4345, + "step": 64005 + }, + { + "epoch": 4.34909634461204, + "grad_norm": 1.0469470024108887, + "learning_rate": 0.00045636295692349507, + "loss": 3.6444, + "step": 64010 + }, + { + "epoch": 4.349436064682702, + "grad_norm": 1.3059152364730835, + "learning_rate": 0.00045632049191466235, + "loss": 3.2828, + "step": 64015 + }, + { + "epoch": 4.349775784753363, + "grad_norm": 0.796098530292511, + "learning_rate": 0.00045627802690582957, + "loss": 3.6809, + "step": 64020 + }, + { + "epoch": 4.350115504824025, + "grad_norm": 1.08735990524292, + "learning_rate": 0.0004562355618969969, + "loss": 3.3172, + "step": 64025 + }, + { + "epoch": 4.350455224894687, + "grad_norm": 1.0638996362686157, + "learning_rate": 0.0004561930968881642, + "loss": 3.7254, + "step": 64030 + }, + { + "epoch": 4.350794944965348, + "grad_norm": 0.816719114780426, + "learning_rate": 0.0004561506318793314, + "loss": 3.2767, + "step": 64035 + }, + { + "epoch": 4.35113466503601, + "grad_norm": 0.975882887840271, + "learning_rate": 0.00045610816687049875, + "loss": 3.7186, + "step": 64040 + }, + { + "epoch": 4.351474385106672, + "grad_norm": 0.8562283515930176, + "learning_rate": 0.000456065701861666, + "loss": 3.4719, + "step": 64045 + }, + { + "epoch": 4.351814105177334, + "grad_norm": 0.8062695860862732, + "learning_rate": 0.00045602323685283325, + "loss": 3.1774, + "step": 64050 + }, + { + "epoch": 4.352153825247996, + "grad_norm": 0.8286985754966736, + "learning_rate": 0.00045598077184400053, + "loss": 3.303, + "step": 64055 + }, + { + "epoch": 4.352493545318658, + "grad_norm": 1.1181873083114624, + "learning_rate": 0.00045593830683516787, + "loss": 3.4947, + "step": 64060 + }, + { + "epoch": 4.352833265389319, + "grad_norm": 0.8421272039413452, + "learning_rate": 0.0004558958418263351, + "loss": 3.6738, + "step": 64065 + }, + { + "epoch": 4.353172985459981, + "grad_norm": 0.8119519948959351, + "learning_rate": 0.00045585337681750237, + "loss": 3.4687, + "step": 64070 + }, + { + "epoch": 4.353512705530643, + "grad_norm": 0.9335781335830688, + "learning_rate": 0.0004558109118086697, + "loss": 3.6479, + "step": 64075 + }, + { + "epoch": 4.353852425601304, + "grad_norm": 0.7960092425346375, + "learning_rate": 0.00045576844679983693, + "loss": 3.4977, + "step": 64080 + }, + { + "epoch": 4.354192145671966, + "grad_norm": 1.1534667015075684, + "learning_rate": 0.0004557259817910042, + "loss": 3.3004, + "step": 64085 + }, + { + "epoch": 4.354531865742628, + "grad_norm": 0.805350124835968, + "learning_rate": 0.0004556835167821715, + "loss": 3.3429, + "step": 64090 + }, + { + "epoch": 4.35487158581329, + "grad_norm": 2.7468109130859375, + "learning_rate": 0.0004556410517733388, + "loss": 3.6474, + "step": 64095 + }, + { + "epoch": 4.355211305883952, + "grad_norm": 0.8282303214073181, + "learning_rate": 0.00045559858676450605, + "loss": 3.4075, + "step": 64100 + }, + { + "epoch": 4.355551025954614, + "grad_norm": 0.8197724223136902, + "learning_rate": 0.00045555612175567333, + "loss": 3.5509, + "step": 64105 + }, + { + "epoch": 4.355890746025275, + "grad_norm": 0.8302217125892639, + "learning_rate": 0.0004555136567468406, + "loss": 3.4633, + "step": 64110 + }, + { + "epoch": 4.356230466095937, + "grad_norm": 0.7723009586334229, + "learning_rate": 0.0004554711917380079, + "loss": 3.2346, + "step": 64115 + }, + { + "epoch": 4.356570186166599, + "grad_norm": 0.9268537163734436, + "learning_rate": 0.0004554287267291752, + "loss": 3.8154, + "step": 64120 + }, + { + "epoch": 4.35690990623726, + "grad_norm": 0.9327241778373718, + "learning_rate": 0.0004553862617203424, + "loss": 3.5625, + "step": 64125 + }, + { + "epoch": 4.357249626307922, + "grad_norm": 0.8181995153427124, + "learning_rate": 0.00045534379671150973, + "loss": 3.4468, + "step": 64130 + }, + { + "epoch": 4.357589346378584, + "grad_norm": 0.884998619556427, + "learning_rate": 0.000455301331702677, + "loss": 3.4595, + "step": 64135 + }, + { + "epoch": 4.357929066449246, + "grad_norm": 0.8104830980300903, + "learning_rate": 0.0004552588666938443, + "loss": 3.4045, + "step": 64140 + }, + { + "epoch": 4.358268786519908, + "grad_norm": 1.0157188177108765, + "learning_rate": 0.0004552164016850116, + "loss": 3.6781, + "step": 64145 + }, + { + "epoch": 4.35860850659057, + "grad_norm": 0.808129608631134, + "learning_rate": 0.00045517393667617885, + "loss": 3.2705, + "step": 64150 + }, + { + "epoch": 4.358948226661231, + "grad_norm": 0.790598452091217, + "learning_rate": 0.00045513147166734613, + "loss": 3.638, + "step": 64155 + }, + { + "epoch": 4.359287946731893, + "grad_norm": 1.295979619026184, + "learning_rate": 0.00045508900665851336, + "loss": 3.4877, + "step": 64160 + }, + { + "epoch": 4.359627666802555, + "grad_norm": 0.9578388929367065, + "learning_rate": 0.0004550465416496807, + "loss": 3.3999, + "step": 64165 + }, + { + "epoch": 4.359967386873216, + "grad_norm": 1.0575147867202759, + "learning_rate": 0.000455004076640848, + "loss": 3.4249, + "step": 64170 + }, + { + "epoch": 4.360307106943878, + "grad_norm": 0.8213820457458496, + "learning_rate": 0.0004549616116320152, + "loss": 3.5628, + "step": 64175 + }, + { + "epoch": 4.36064682701454, + "grad_norm": 0.8148659467697144, + "learning_rate": 0.00045491914662318253, + "loss": 3.3695, + "step": 64180 + }, + { + "epoch": 4.360986547085202, + "grad_norm": 0.8647225499153137, + "learning_rate": 0.0004548766816143498, + "loss": 3.707, + "step": 64185 + }, + { + "epoch": 4.361326267155864, + "grad_norm": 0.8026604056358337, + "learning_rate": 0.00045483421660551704, + "loss": 3.5337, + "step": 64190 + }, + { + "epoch": 4.361665987226525, + "grad_norm": 0.9234213829040527, + "learning_rate": 0.0004547917515966843, + "loss": 3.4892, + "step": 64195 + }, + { + "epoch": 4.362005707297187, + "grad_norm": 0.9592143297195435, + "learning_rate": 0.00045474928658785165, + "loss": 3.3486, + "step": 64200 + }, + { + "epoch": 4.362345427367849, + "grad_norm": 1.0063343048095703, + "learning_rate": 0.0004547068215790189, + "loss": 3.5446, + "step": 64205 + }, + { + "epoch": 4.36268514743851, + "grad_norm": 1.1948484182357788, + "learning_rate": 0.00045466435657018616, + "loss": 3.3303, + "step": 64210 + }, + { + "epoch": 4.363024867509172, + "grad_norm": 0.8311105370521545, + "learning_rate": 0.0004546218915613535, + "loss": 3.5713, + "step": 64215 + }, + { + "epoch": 4.363364587579834, + "grad_norm": 0.8245605826377869, + "learning_rate": 0.0004545794265525207, + "loss": 3.5606, + "step": 64220 + }, + { + "epoch": 4.3637043076504956, + "grad_norm": 0.709775984287262, + "learning_rate": 0.000454536961543688, + "loss": 3.4571, + "step": 64225 + }, + { + "epoch": 4.364044027721158, + "grad_norm": 0.8208695650100708, + "learning_rate": 0.0004544944965348553, + "loss": 3.5937, + "step": 64230 + }, + { + "epoch": 4.36438374779182, + "grad_norm": 1.0970091819763184, + "learning_rate": 0.00045445203152602256, + "loss": 3.4654, + "step": 64235 + }, + { + "epoch": 4.364723467862481, + "grad_norm": 0.9484691619873047, + "learning_rate": 0.00045440956651718984, + "loss": 3.568, + "step": 64240 + }, + { + "epoch": 4.365063187933143, + "grad_norm": 0.9631114602088928, + "learning_rate": 0.0004543671015083571, + "loss": 3.6399, + "step": 64245 + }, + { + "epoch": 4.365402908003805, + "grad_norm": 1.0244508981704712, + "learning_rate": 0.0004543246364995244, + "loss": 3.4196, + "step": 64250 + }, + { + "epoch": 4.365742628074466, + "grad_norm": 0.8785644769668579, + "learning_rate": 0.0004542821714906917, + "loss": 3.2782, + "step": 64255 + }, + { + "epoch": 4.366082348145128, + "grad_norm": 1.381556510925293, + "learning_rate": 0.00045423970648185896, + "loss": 3.3896, + "step": 64260 + }, + { + "epoch": 4.36642206821579, + "grad_norm": 0.8162555694580078, + "learning_rate": 0.0004541972414730262, + "loss": 3.1273, + "step": 64265 + }, + { + "epoch": 4.366761788286452, + "grad_norm": 0.760924756526947, + "learning_rate": 0.0004541547764641935, + "loss": 3.261, + "step": 64270 + }, + { + "epoch": 4.367101508357114, + "grad_norm": 0.8858211040496826, + "learning_rate": 0.0004541123114553608, + "loss": 3.6744, + "step": 64275 + }, + { + "epoch": 4.367441228427776, + "grad_norm": 1.0548710823059082, + "learning_rate": 0.00045406984644652803, + "loss": 3.4382, + "step": 64280 + }, + { + "epoch": 4.367780948498437, + "grad_norm": 1.025705099105835, + "learning_rate": 0.00045402738143769536, + "loss": 3.5001, + "step": 64285 + }, + { + "epoch": 4.368120668569099, + "grad_norm": 1.5040837526321411, + "learning_rate": 0.00045398491642886264, + "loss": 3.403, + "step": 64290 + }, + { + "epoch": 4.368460388639761, + "grad_norm": 0.8470531105995178, + "learning_rate": 0.00045394245142002987, + "loss": 3.475, + "step": 64295 + }, + { + "epoch": 4.368800108710422, + "grad_norm": 0.9951391816139221, + "learning_rate": 0.00045389998641119715, + "loss": 3.6035, + "step": 64300 + }, + { + "epoch": 4.369139828781084, + "grad_norm": 0.9323769807815552, + "learning_rate": 0.0004538575214023645, + "loss": 3.8978, + "step": 64305 + }, + { + "epoch": 4.369479548851746, + "grad_norm": 0.9128039479255676, + "learning_rate": 0.00045381505639353176, + "loss": 3.4414, + "step": 64310 + }, + { + "epoch": 4.369819268922408, + "grad_norm": 0.7647458910942078, + "learning_rate": 0.000453772591384699, + "loss": 3.2307, + "step": 64315 + }, + { + "epoch": 4.37015898899307, + "grad_norm": 0.9004073739051819, + "learning_rate": 0.0004537301263758663, + "loss": 3.7186, + "step": 64320 + }, + { + "epoch": 4.370498709063732, + "grad_norm": 0.788355827331543, + "learning_rate": 0.0004536876613670336, + "loss": 3.3409, + "step": 64325 + }, + { + "epoch": 4.370838429134393, + "grad_norm": 0.648705244064331, + "learning_rate": 0.00045364519635820083, + "loss": 3.777, + "step": 64330 + }, + { + "epoch": 4.371178149205055, + "grad_norm": 1.0367971658706665, + "learning_rate": 0.00045360273134936816, + "loss": 3.6295, + "step": 64335 + }, + { + "epoch": 4.371517869275717, + "grad_norm": 0.9515285491943359, + "learning_rate": 0.00045356026634053544, + "loss": 3.1983, + "step": 64340 + }, + { + "epoch": 4.371857589346378, + "grad_norm": 0.8702841997146606, + "learning_rate": 0.00045351780133170267, + "loss": 3.9471, + "step": 64345 + }, + { + "epoch": 4.37219730941704, + "grad_norm": 0.8304216861724854, + "learning_rate": 0.00045347533632286995, + "loss": 3.3932, + "step": 64350 + }, + { + "epoch": 4.372537029487702, + "grad_norm": 0.8743628263473511, + "learning_rate": 0.0004534328713140373, + "loss": 3.6531, + "step": 64355 + }, + { + "epoch": 4.372876749558364, + "grad_norm": 0.9574486017227173, + "learning_rate": 0.0004533904063052045, + "loss": 3.4912, + "step": 64360 + }, + { + "epoch": 4.373216469629026, + "grad_norm": 0.936106264591217, + "learning_rate": 0.0004533479412963718, + "loss": 3.722, + "step": 64365 + }, + { + "epoch": 4.373556189699688, + "grad_norm": 1.44229257106781, + "learning_rate": 0.0004533054762875391, + "loss": 3.4364, + "step": 64370 + }, + { + "epoch": 4.373895909770349, + "grad_norm": 0.8339200615882874, + "learning_rate": 0.00045326301127870635, + "loss": 3.7202, + "step": 64375 + }, + { + "epoch": 4.374235629841011, + "grad_norm": 0.9917379021644592, + "learning_rate": 0.00045322054626987363, + "loss": 3.665, + "step": 64380 + }, + { + "epoch": 4.374575349911673, + "grad_norm": 0.8637394309043884, + "learning_rate": 0.0004531780812610409, + "loss": 3.2085, + "step": 64385 + }, + { + "epoch": 4.374915069982334, + "grad_norm": 0.9819831848144531, + "learning_rate": 0.0004531356162522082, + "loss": 3.3795, + "step": 64390 + }, + { + "epoch": 4.375254790052996, + "grad_norm": 0.9710754752159119, + "learning_rate": 0.00045309315124337547, + "loss": 3.6736, + "step": 64395 + }, + { + "epoch": 4.375594510123658, + "grad_norm": 1.006867527961731, + "learning_rate": 0.00045305068623454275, + "loss": 3.3275, + "step": 64400 + }, + { + "epoch": 4.37593423019432, + "grad_norm": 0.9713447093963623, + "learning_rate": 0.00045300822122571003, + "loss": 3.5142, + "step": 64405 + }, + { + "epoch": 4.376273950264982, + "grad_norm": 0.9223304390907288, + "learning_rate": 0.0004529657562168773, + "loss": 3.354, + "step": 64410 + }, + { + "epoch": 4.376613670335644, + "grad_norm": 0.7713569402694702, + "learning_rate": 0.0004529232912080446, + "loss": 3.4836, + "step": 64415 + }, + { + "epoch": 4.376953390406305, + "grad_norm": 0.8136366009712219, + "learning_rate": 0.0004528808261992118, + "loss": 3.6488, + "step": 64420 + }, + { + "epoch": 4.377293110476967, + "grad_norm": 1.0428146123886108, + "learning_rate": 0.00045283836119037915, + "loss": 3.503, + "step": 64425 + }, + { + "epoch": 4.377632830547629, + "grad_norm": 0.9239062666893005, + "learning_rate": 0.00045279589618154643, + "loss": 3.8233, + "step": 64430 + }, + { + "epoch": 4.37797255061829, + "grad_norm": 0.6955722570419312, + "learning_rate": 0.00045275343117271366, + "loss": 3.7353, + "step": 64435 + }, + { + "epoch": 4.378312270688952, + "grad_norm": 0.8523550033569336, + "learning_rate": 0.000452710966163881, + "loss": 3.3785, + "step": 64440 + }, + { + "epoch": 4.378651990759614, + "grad_norm": 0.8160797357559204, + "learning_rate": 0.00045266850115504827, + "loss": 3.3877, + "step": 64445 + }, + { + "epoch": 4.378991710830276, + "grad_norm": 0.9816880822181702, + "learning_rate": 0.0004526260361462155, + "loss": 3.804, + "step": 64450 + }, + { + "epoch": 4.379331430900938, + "grad_norm": 1.4800978899002075, + "learning_rate": 0.0004525835711373828, + "loss": 3.3352, + "step": 64455 + }, + { + "epoch": 4.3796711509716, + "grad_norm": 0.706000566482544, + "learning_rate": 0.0004525411061285501, + "loss": 3.3569, + "step": 64460 + }, + { + "epoch": 4.380010871042261, + "grad_norm": 0.8496415615081787, + "learning_rate": 0.00045249864111971734, + "loss": 3.5992, + "step": 64465 + }, + { + "epoch": 4.380350591112923, + "grad_norm": 0.9088304042816162, + "learning_rate": 0.0004524561761108846, + "loss": 3.3331, + "step": 64470 + }, + { + "epoch": 4.380690311183585, + "grad_norm": 0.8301309943199158, + "learning_rate": 0.00045241371110205195, + "loss": 3.4504, + "step": 64475 + }, + { + "epoch": 4.381030031254246, + "grad_norm": 1.1841665506362915, + "learning_rate": 0.00045237124609321923, + "loss": 3.529, + "step": 64480 + }, + { + "epoch": 4.381369751324908, + "grad_norm": 0.9794401526451111, + "learning_rate": 0.00045232878108438646, + "loss": 3.5589, + "step": 64485 + }, + { + "epoch": 4.38170947139557, + "grad_norm": 0.9266629219055176, + "learning_rate": 0.00045228631607555374, + "loss": 3.5896, + "step": 64490 + }, + { + "epoch": 4.382049191466232, + "grad_norm": 0.9738286137580872, + "learning_rate": 0.00045224385106672107, + "loss": 3.5562, + "step": 64495 + }, + { + "epoch": 4.382388911536894, + "grad_norm": 0.9822114706039429, + "learning_rate": 0.0004522013860578883, + "loss": 3.2382, + "step": 64500 + }, + { + "epoch": 4.382728631607556, + "grad_norm": 0.9082299470901489, + "learning_rate": 0.0004521589210490556, + "loss": 3.6357, + "step": 64505 + }, + { + "epoch": 4.383068351678217, + "grad_norm": 0.8554168939590454, + "learning_rate": 0.0004521164560402229, + "loss": 3.4156, + "step": 64510 + }, + { + "epoch": 4.383408071748879, + "grad_norm": 0.7571812868118286, + "learning_rate": 0.00045207399103139014, + "loss": 3.5565, + "step": 64515 + }, + { + "epoch": 4.383747791819541, + "grad_norm": 0.9785128831863403, + "learning_rate": 0.0004520315260225574, + "loss": 3.5728, + "step": 64520 + }, + { + "epoch": 4.384087511890202, + "grad_norm": 0.9914416670799255, + "learning_rate": 0.0004519890610137247, + "loss": 3.8714, + "step": 64525 + }, + { + "epoch": 4.384427231960864, + "grad_norm": 0.8125157952308655, + "learning_rate": 0.000451946596004892, + "loss": 3.7296, + "step": 64530 + }, + { + "epoch": 4.384766952031526, + "grad_norm": 0.8399844765663147, + "learning_rate": 0.00045190413099605926, + "loss": 3.4348, + "step": 64535 + }, + { + "epoch": 4.385106672102188, + "grad_norm": 0.9351385831832886, + "learning_rate": 0.00045186166598722654, + "loss": 3.7293, + "step": 64540 + }, + { + "epoch": 4.38544639217285, + "grad_norm": 1.010532021522522, + "learning_rate": 0.0004518192009783938, + "loss": 3.318, + "step": 64545 + }, + { + "epoch": 4.385786112243512, + "grad_norm": 1.1710469722747803, + "learning_rate": 0.0004517767359695611, + "loss": 3.6804, + "step": 64550 + }, + { + "epoch": 4.386125832314173, + "grad_norm": 0.7254777550697327, + "learning_rate": 0.0004517342709607284, + "loss": 3.4764, + "step": 64555 + }, + { + "epoch": 4.386465552384835, + "grad_norm": 0.7284232378005981, + "learning_rate": 0.0004516918059518956, + "loss": 3.442, + "step": 64560 + }, + { + "epoch": 4.386805272455497, + "grad_norm": 0.9451462030410767, + "learning_rate": 0.00045164934094306294, + "loss": 3.5935, + "step": 64565 + }, + { + "epoch": 4.387144992526158, + "grad_norm": 0.9240872263908386, + "learning_rate": 0.0004516068759342302, + "loss": 3.5202, + "step": 64570 + }, + { + "epoch": 4.38748471259682, + "grad_norm": 0.8979571461677551, + "learning_rate": 0.00045156441092539744, + "loss": 3.4783, + "step": 64575 + }, + { + "epoch": 4.3878244326674825, + "grad_norm": 0.9020591974258423, + "learning_rate": 0.0004515219459165648, + "loss": 3.6488, + "step": 64580 + }, + { + "epoch": 4.388164152738144, + "grad_norm": 0.8123366832733154, + "learning_rate": 0.00045147948090773206, + "loss": 3.5484, + "step": 64585 + }, + { + "epoch": 4.388503872808806, + "grad_norm": 0.8190121650695801, + "learning_rate": 0.0004514370158988993, + "loss": 3.4835, + "step": 64590 + }, + { + "epoch": 4.388843592879467, + "grad_norm": 0.8332828283309937, + "learning_rate": 0.00045139455089006656, + "loss": 3.628, + "step": 64595 + }, + { + "epoch": 4.389183312950129, + "grad_norm": 0.7954887747764587, + "learning_rate": 0.0004513520858812339, + "loss": 3.3835, + "step": 64600 + }, + { + "epoch": 4.389523033020791, + "grad_norm": 0.8848952651023865, + "learning_rate": 0.0004513096208724011, + "loss": 3.6436, + "step": 64605 + }, + { + "epoch": 4.389862753091452, + "grad_norm": 0.8876662850379944, + "learning_rate": 0.0004512671558635684, + "loss": 3.6554, + "step": 64610 + }, + { + "epoch": 4.390202473162114, + "grad_norm": 0.8347315788269043, + "learning_rate": 0.00045122469085473574, + "loss": 3.7396, + "step": 64615 + }, + { + "epoch": 4.390542193232776, + "grad_norm": 0.809548020362854, + "learning_rate": 0.00045118222584590296, + "loss": 3.4019, + "step": 64620 + }, + { + "epoch": 4.390881913303438, + "grad_norm": 1.0197629928588867, + "learning_rate": 0.00045113976083707024, + "loss": 3.5469, + "step": 64625 + }, + { + "epoch": 4.3912216333741, + "grad_norm": 0.8092890381813049, + "learning_rate": 0.0004510972958282376, + "loss": 3.3925, + "step": 64630 + }, + { + "epoch": 4.391561353444762, + "grad_norm": 1.1317576169967651, + "learning_rate": 0.0004510548308194048, + "loss": 3.3887, + "step": 64635 + }, + { + "epoch": 4.391901073515423, + "grad_norm": 0.8381431102752686, + "learning_rate": 0.0004510123658105721, + "loss": 3.5798, + "step": 64640 + }, + { + "epoch": 4.392240793586085, + "grad_norm": 0.8388566970825195, + "learning_rate": 0.00045096990080173936, + "loss": 3.4054, + "step": 64645 + }, + { + "epoch": 4.392580513656747, + "grad_norm": 0.9554895162582397, + "learning_rate": 0.0004509274357929067, + "loss": 3.3432, + "step": 64650 + }, + { + "epoch": 4.392920233727408, + "grad_norm": 0.898276686668396, + "learning_rate": 0.0004508849707840739, + "loss": 3.4542, + "step": 64655 + }, + { + "epoch": 4.39325995379807, + "grad_norm": 0.7132056951522827, + "learning_rate": 0.0004508425057752412, + "loss": 3.643, + "step": 64660 + }, + { + "epoch": 4.393599673868732, + "grad_norm": 1.1174218654632568, + "learning_rate": 0.00045080004076640854, + "loss": 3.3068, + "step": 64665 + }, + { + "epoch": 4.393939393939394, + "grad_norm": 0.8897207379341125, + "learning_rate": 0.00045075757575757577, + "loss": 3.4755, + "step": 64670 + }, + { + "epoch": 4.394279114010056, + "grad_norm": 0.8851433396339417, + "learning_rate": 0.00045071511074874305, + "loss": 3.2876, + "step": 64675 + }, + { + "epoch": 4.394618834080718, + "grad_norm": 0.7211104035377502, + "learning_rate": 0.0004506726457399103, + "loss": 3.687, + "step": 64680 + }, + { + "epoch": 4.394958554151379, + "grad_norm": 0.8784237504005432, + "learning_rate": 0.0004506301807310776, + "loss": 3.6228, + "step": 64685 + }, + { + "epoch": 4.395298274222041, + "grad_norm": 0.8926588296890259, + "learning_rate": 0.0004505877157222449, + "loss": 3.516, + "step": 64690 + }, + { + "epoch": 4.395637994292703, + "grad_norm": 1.0050911903381348, + "learning_rate": 0.00045054525071341217, + "loss": 3.3623, + "step": 64695 + }, + { + "epoch": 4.395977714363364, + "grad_norm": 0.9478135108947754, + "learning_rate": 0.00045050278570457945, + "loss": 3.2387, + "step": 64700 + }, + { + "epoch": 4.396317434434026, + "grad_norm": 1.6698096990585327, + "learning_rate": 0.0004504603206957467, + "loss": 3.485, + "step": 64705 + }, + { + "epoch": 4.396657154504688, + "grad_norm": 0.9323243498802185, + "learning_rate": 0.000450417855686914, + "loss": 3.3901, + "step": 64710 + }, + { + "epoch": 4.39699687457535, + "grad_norm": 0.7532806396484375, + "learning_rate": 0.00045037539067808123, + "loss": 3.4245, + "step": 64715 + }, + { + "epoch": 4.397336594646012, + "grad_norm": 0.9737496972084045, + "learning_rate": 0.00045033292566924857, + "loss": 3.2909, + "step": 64720 + }, + { + "epoch": 4.397676314716674, + "grad_norm": 0.9262939095497131, + "learning_rate": 0.00045029046066041585, + "loss": 3.4267, + "step": 64725 + }, + { + "epoch": 4.398016034787335, + "grad_norm": 0.7953515648841858, + "learning_rate": 0.00045024799565158307, + "loss": 3.3702, + "step": 64730 + }, + { + "epoch": 4.398355754857997, + "grad_norm": 0.590951681137085, + "learning_rate": 0.0004502055306427504, + "loss": 3.5927, + "step": 64735 + }, + { + "epoch": 4.398695474928659, + "grad_norm": 0.9835566878318787, + "learning_rate": 0.0004501630656339177, + "loss": 3.2923, + "step": 64740 + }, + { + "epoch": 4.39903519499932, + "grad_norm": 0.8848749399185181, + "learning_rate": 0.0004501206006250849, + "loss": 3.468, + "step": 64745 + }, + { + "epoch": 4.399374915069982, + "grad_norm": 1.1962647438049316, + "learning_rate": 0.0004500781356162522, + "loss": 3.255, + "step": 64750 + }, + { + "epoch": 4.399714635140644, + "grad_norm": 0.7615827918052673, + "learning_rate": 0.0004500356706074195, + "loss": 3.5284, + "step": 64755 + }, + { + "epoch": 4.400054355211306, + "grad_norm": 1.0787931680679321, + "learning_rate": 0.00044999320559858675, + "loss": 3.4973, + "step": 64760 + }, + { + "epoch": 4.400394075281968, + "grad_norm": 0.8010728359222412, + "learning_rate": 0.00044995074058975403, + "loss": 3.29, + "step": 64765 + }, + { + "epoch": 4.40073379535263, + "grad_norm": 0.8789026737213135, + "learning_rate": 0.00044990827558092137, + "loss": 3.3412, + "step": 64770 + }, + { + "epoch": 4.401073515423291, + "grad_norm": 0.8462004065513611, + "learning_rate": 0.0004498658105720886, + "loss": 3.4673, + "step": 64775 + }, + { + "epoch": 4.401413235493953, + "grad_norm": 0.8739003539085388, + "learning_rate": 0.00044982334556325587, + "loss": 3.3336, + "step": 64780 + }, + { + "epoch": 4.401752955564615, + "grad_norm": 0.8325495719909668, + "learning_rate": 0.00044978088055442315, + "loss": 3.2661, + "step": 64785 + }, + { + "epoch": 4.402092675635276, + "grad_norm": 1.1298160552978516, + "learning_rate": 0.00044973841554559043, + "loss": 3.5385, + "step": 64790 + }, + { + "epoch": 4.402432395705938, + "grad_norm": 0.9124968647956848, + "learning_rate": 0.0004496959505367577, + "loss": 3.4836, + "step": 64795 + }, + { + "epoch": 4.4027721157766, + "grad_norm": 0.8804593086242676, + "learning_rate": 0.000449653485527925, + "loss": 3.1837, + "step": 64800 + }, + { + "epoch": 4.403111835847262, + "grad_norm": 1.1899088621139526, + "learning_rate": 0.0004496110205190923, + "loss": 3.43, + "step": 64805 + }, + { + "epoch": 4.403451555917924, + "grad_norm": 1.0679731369018555, + "learning_rate": 0.00044956855551025955, + "loss": 3.4203, + "step": 64810 + }, + { + "epoch": 4.403791275988586, + "grad_norm": 0.970270574092865, + "learning_rate": 0.00044952609050142683, + "loss": 3.4975, + "step": 64815 + }, + { + "epoch": 4.404130996059247, + "grad_norm": 0.9771348237991333, + "learning_rate": 0.0004494836254925941, + "loss": 3.3282, + "step": 64820 + }, + { + "epoch": 4.404470716129909, + "grad_norm": 0.8059468269348145, + "learning_rate": 0.0004494411604837614, + "loss": 3.3468, + "step": 64825 + }, + { + "epoch": 4.404810436200571, + "grad_norm": 0.8164841532707214, + "learning_rate": 0.0004493986954749287, + "loss": 3.8383, + "step": 64830 + }, + { + "epoch": 4.405150156271232, + "grad_norm": 1.2418636083602905, + "learning_rate": 0.00044935623046609595, + "loss": 3.7778, + "step": 64835 + }, + { + "epoch": 4.405489876341894, + "grad_norm": 0.8264691233634949, + "learning_rate": 0.00044931376545726323, + "loss": 3.4205, + "step": 64840 + }, + { + "epoch": 4.4058295964125564, + "grad_norm": 0.8884705305099487, + "learning_rate": 0.0004492713004484305, + "loss": 3.33, + "step": 64845 + }, + { + "epoch": 4.406169316483218, + "grad_norm": 0.8003307580947876, + "learning_rate": 0.0004492288354395978, + "loss": 3.4485, + "step": 64850 + }, + { + "epoch": 4.40650903655388, + "grad_norm": 0.7426823377609253, + "learning_rate": 0.000449186370430765, + "loss": 3.4116, + "step": 64855 + }, + { + "epoch": 4.406848756624542, + "grad_norm": 1.2305539846420288, + "learning_rate": 0.00044914390542193235, + "loss": 3.2162, + "step": 64860 + }, + { + "epoch": 4.407188476695203, + "grad_norm": 0.8456860780715942, + "learning_rate": 0.00044910144041309963, + "loss": 3.626, + "step": 64865 + }, + { + "epoch": 4.407528196765865, + "grad_norm": 0.9313734769821167, + "learning_rate": 0.00044905897540426686, + "loss": 3.3284, + "step": 64870 + }, + { + "epoch": 4.407867916836526, + "grad_norm": 1.011651873588562, + "learning_rate": 0.0004490165103954342, + "loss": 3.5675, + "step": 64875 + }, + { + "epoch": 4.408207636907188, + "grad_norm": 0.9485716223716736, + "learning_rate": 0.0004489740453866015, + "loss": 3.7079, + "step": 64880 + }, + { + "epoch": 4.40854735697785, + "grad_norm": 0.8012006282806396, + "learning_rate": 0.0004489315803777687, + "loss": 3.331, + "step": 64885 + }, + { + "epoch": 4.408887077048512, + "grad_norm": 0.9415907263755798, + "learning_rate": 0.00044888911536893603, + "loss": 3.5031, + "step": 64890 + }, + { + "epoch": 4.409226797119174, + "grad_norm": 0.7125803828239441, + "learning_rate": 0.0004488466503601033, + "loss": 3.373, + "step": 64895 + }, + { + "epoch": 4.409566517189836, + "grad_norm": 0.8787747621536255, + "learning_rate": 0.00044880418535127054, + "loss": 3.403, + "step": 64900 + }, + { + "epoch": 4.409906237260497, + "grad_norm": 0.8242262005805969, + "learning_rate": 0.0004487617203424378, + "loss": 3.4112, + "step": 64905 + }, + { + "epoch": 4.410245957331159, + "grad_norm": 0.8411560654640198, + "learning_rate": 0.00044871925533360515, + "loss": 3.5182, + "step": 64910 + }, + { + "epoch": 4.410585677401821, + "grad_norm": 0.9842845797538757, + "learning_rate": 0.0004486767903247724, + "loss": 3.6295, + "step": 64915 + }, + { + "epoch": 4.410925397472482, + "grad_norm": 1.024263620376587, + "learning_rate": 0.00044863432531593966, + "loss": 3.2088, + "step": 64920 + }, + { + "epoch": 4.411265117543144, + "grad_norm": 0.79705810546875, + "learning_rate": 0.000448591860307107, + "loss": 3.2276, + "step": 64925 + }, + { + "epoch": 4.411604837613806, + "grad_norm": 0.8538332581520081, + "learning_rate": 0.0004485493952982742, + "loss": 3.3012, + "step": 64930 + }, + { + "epoch": 4.411944557684468, + "grad_norm": 0.9359583854675293, + "learning_rate": 0.0004485069302894415, + "loss": 3.6225, + "step": 64935 + }, + { + "epoch": 4.41228427775513, + "grad_norm": 0.8045932054519653, + "learning_rate": 0.0004484644652806088, + "loss": 3.516, + "step": 64940 + }, + { + "epoch": 4.412623997825792, + "grad_norm": 0.780533492565155, + "learning_rate": 0.00044842200027177606, + "loss": 3.3913, + "step": 64945 + }, + { + "epoch": 4.412963717896453, + "grad_norm": 1.0491788387298584, + "learning_rate": 0.00044837953526294334, + "loss": 3.3989, + "step": 64950 + }, + { + "epoch": 4.413303437967115, + "grad_norm": 1.1852960586547852, + "learning_rate": 0.0004483370702541106, + "loss": 3.524, + "step": 64955 + }, + { + "epoch": 4.413643158037777, + "grad_norm": 0.8996584415435791, + "learning_rate": 0.0004482946052452779, + "loss": 3.7289, + "step": 64960 + }, + { + "epoch": 4.413982878108438, + "grad_norm": 0.7434871196746826, + "learning_rate": 0.0004482521402364452, + "loss": 3.3752, + "step": 64965 + }, + { + "epoch": 4.4143225981791, + "grad_norm": 1.070533275604248, + "learning_rate": 0.00044820967522761246, + "loss": 3.6421, + "step": 64970 + }, + { + "epoch": 4.414662318249762, + "grad_norm": 0.7161136269569397, + "learning_rate": 0.0004481672102187797, + "loss": 3.3524, + "step": 64975 + }, + { + "epoch": 4.415002038320424, + "grad_norm": 0.8421730399131775, + "learning_rate": 0.000448124745209947, + "loss": 3.5382, + "step": 64980 + }, + { + "epoch": 4.415341758391086, + "grad_norm": 0.9722555875778198, + "learning_rate": 0.0004480822802011143, + "loss": 3.6208, + "step": 64985 + }, + { + "epoch": 4.415681478461748, + "grad_norm": 0.8298919200897217, + "learning_rate": 0.0004480398151922816, + "loss": 3.6129, + "step": 64990 + }, + { + "epoch": 4.416021198532409, + "grad_norm": 0.8144833445549011, + "learning_rate": 0.00044799735018344886, + "loss": 3.5713, + "step": 64995 + }, + { + "epoch": 4.416360918603071, + "grad_norm": 1.0004266500473022, + "learning_rate": 0.00044795488517461614, + "loss": 3.2115, + "step": 65000 + }, + { + "epoch": 4.416700638673733, + "grad_norm": 0.8938560485839844, + "learning_rate": 0.0004479124201657834, + "loss": 3.2683, + "step": 65005 + }, + { + "epoch": 4.417040358744394, + "grad_norm": 0.7159640789031982, + "learning_rate": 0.00044786995515695065, + "loss": 3.2967, + "step": 65010 + }, + { + "epoch": 4.417380078815056, + "grad_norm": 0.8809435367584229, + "learning_rate": 0.000447827490148118, + "loss": 3.5437, + "step": 65015 + }, + { + "epoch": 4.417719798885718, + "grad_norm": 0.8484206199645996, + "learning_rate": 0.00044778502513928526, + "loss": 3.4811, + "step": 65020 + }, + { + "epoch": 4.41805951895638, + "grad_norm": 0.7763258218765259, + "learning_rate": 0.0004477425601304525, + "loss": 3.3201, + "step": 65025 + }, + { + "epoch": 4.418399239027042, + "grad_norm": 0.8743746876716614, + "learning_rate": 0.0004477000951216198, + "loss": 3.5965, + "step": 65030 + }, + { + "epoch": 4.418738959097704, + "grad_norm": 0.9821796417236328, + "learning_rate": 0.0004476576301127871, + "loss": 3.4462, + "step": 65035 + }, + { + "epoch": 4.419078679168365, + "grad_norm": 0.7879642248153687, + "learning_rate": 0.00044761516510395433, + "loss": 3.3875, + "step": 65040 + }, + { + "epoch": 4.419418399239027, + "grad_norm": 1.009671688079834, + "learning_rate": 0.0004475727000951216, + "loss": 3.2279, + "step": 65045 + }, + { + "epoch": 4.419758119309689, + "grad_norm": 0.9736917018890381, + "learning_rate": 0.00044753023508628894, + "loss": 3.372, + "step": 65050 + }, + { + "epoch": 4.42009783938035, + "grad_norm": 0.8007050156593323, + "learning_rate": 0.00044748777007745617, + "loss": 3.6673, + "step": 65055 + }, + { + "epoch": 4.420437559451012, + "grad_norm": 0.8634523153305054, + "learning_rate": 0.00044744530506862345, + "loss": 3.6329, + "step": 65060 + }, + { + "epoch": 4.420777279521674, + "grad_norm": 1.0416736602783203, + "learning_rate": 0.0004474028400597908, + "loss": 3.5827, + "step": 65065 + }, + { + "epoch": 4.421116999592336, + "grad_norm": 0.8211742639541626, + "learning_rate": 0.000447360375050958, + "loss": 3.3254, + "step": 65070 + }, + { + "epoch": 4.421456719662998, + "grad_norm": 1.0899970531463623, + "learning_rate": 0.0004473179100421253, + "loss": 3.4016, + "step": 65075 + }, + { + "epoch": 4.42179643973366, + "grad_norm": 0.8331630825996399, + "learning_rate": 0.00044727544503329257, + "loss": 3.4202, + "step": 65080 + }, + { + "epoch": 4.422136159804321, + "grad_norm": 0.9047204256057739, + "learning_rate": 0.00044723298002445985, + "loss": 3.6312, + "step": 65085 + }, + { + "epoch": 4.422475879874983, + "grad_norm": 0.9336113929748535, + "learning_rate": 0.00044719051501562713, + "loss": 3.1886, + "step": 65090 + }, + { + "epoch": 4.422815599945645, + "grad_norm": 0.7998154759407043, + "learning_rate": 0.0004471480500067944, + "loss": 3.5472, + "step": 65095 + }, + { + "epoch": 4.423155320016306, + "grad_norm": 0.9060941338539124, + "learning_rate": 0.0004471055849979617, + "loss": 3.6218, + "step": 65100 + }, + { + "epoch": 4.423495040086968, + "grad_norm": 0.9013347029685974, + "learning_rate": 0.00044706311998912897, + "loss": 3.5244, + "step": 65105 + }, + { + "epoch": 4.42383476015763, + "grad_norm": 1.8107386827468872, + "learning_rate": 0.00044702065498029625, + "loss": 3.5768, + "step": 65110 + }, + { + "epoch": 4.424174480228292, + "grad_norm": 0.9807305335998535, + "learning_rate": 0.0004469781899714635, + "loss": 3.5115, + "step": 65115 + }, + { + "epoch": 4.424514200298954, + "grad_norm": 0.9302863478660583, + "learning_rate": 0.0004469357249626308, + "loss": 3.3412, + "step": 65120 + }, + { + "epoch": 4.424853920369616, + "grad_norm": 1.0677871704101562, + "learning_rate": 0.0004468932599537981, + "loss": 3.3396, + "step": 65125 + }, + { + "epoch": 4.425193640440277, + "grad_norm": 0.8922421932220459, + "learning_rate": 0.0004468507949449653, + "loss": 3.2335, + "step": 65130 + }, + { + "epoch": 4.425533360510939, + "grad_norm": 0.7762733697891235, + "learning_rate": 0.00044680832993613265, + "loss": 3.3928, + "step": 65135 + }, + { + "epoch": 4.425873080581601, + "grad_norm": 0.9222429394721985, + "learning_rate": 0.00044676586492729993, + "loss": 3.4203, + "step": 65140 + }, + { + "epoch": 4.426212800652262, + "grad_norm": 1.1646645069122314, + "learning_rate": 0.00044672339991846716, + "loss": 3.5337, + "step": 65145 + }, + { + "epoch": 4.426552520722924, + "grad_norm": 0.9924411773681641, + "learning_rate": 0.00044668093490963444, + "loss": 3.4629, + "step": 65150 + }, + { + "epoch": 4.4268922407935865, + "grad_norm": 0.8357965350151062, + "learning_rate": 0.00044663846990080177, + "loss": 3.5043, + "step": 65155 + }, + { + "epoch": 4.427231960864248, + "grad_norm": 0.8970503211021423, + "learning_rate": 0.00044659600489196905, + "loss": 3.8881, + "step": 65160 + }, + { + "epoch": 4.42757168093491, + "grad_norm": 0.9612279534339905, + "learning_rate": 0.0004465535398831363, + "loss": 3.8546, + "step": 65165 + }, + { + "epoch": 4.427911401005572, + "grad_norm": 0.8405584692955017, + "learning_rate": 0.0004465110748743036, + "loss": 3.626, + "step": 65170 + }, + { + "epoch": 4.428251121076233, + "grad_norm": 1.0533685684204102, + "learning_rate": 0.0004464686098654709, + "loss": 3.2728, + "step": 65175 + }, + { + "epoch": 4.428590841146895, + "grad_norm": 0.8456205725669861, + "learning_rate": 0.0004464261448566381, + "loss": 3.6083, + "step": 65180 + }, + { + "epoch": 4.428930561217557, + "grad_norm": 0.7051971554756165, + "learning_rate": 0.00044638367984780545, + "loss": 3.5593, + "step": 65185 + }, + { + "epoch": 4.429270281288218, + "grad_norm": 0.9375216364860535, + "learning_rate": 0.00044634121483897273, + "loss": 3.6292, + "step": 65190 + }, + { + "epoch": 4.42961000135888, + "grad_norm": 0.7986471056938171, + "learning_rate": 0.00044629874983013996, + "loss": 3.331, + "step": 65195 + }, + { + "epoch": 4.4299497214295425, + "grad_norm": 0.8557507991790771, + "learning_rate": 0.00044625628482130724, + "loss": 3.712, + "step": 65200 + }, + { + "epoch": 4.430289441500204, + "grad_norm": 0.9136731624603271, + "learning_rate": 0.00044621381981247457, + "loss": 3.5324, + "step": 65205 + }, + { + "epoch": 4.430629161570866, + "grad_norm": 0.8843671083450317, + "learning_rate": 0.0004461713548036418, + "loss": 3.1793, + "step": 65210 + }, + { + "epoch": 4.430968881641528, + "grad_norm": 1.2320665121078491, + "learning_rate": 0.0004461288897948091, + "loss": 3.749, + "step": 65215 + }, + { + "epoch": 4.431308601712189, + "grad_norm": 1.0043127536773682, + "learning_rate": 0.0004460864247859764, + "loss": 3.7167, + "step": 65220 + }, + { + "epoch": 4.431648321782851, + "grad_norm": 0.9097742438316345, + "learning_rate": 0.00044604395977714364, + "loss": 3.5056, + "step": 65225 + }, + { + "epoch": 4.431988041853513, + "grad_norm": 1.01972496509552, + "learning_rate": 0.0004460014947683109, + "loss": 3.5939, + "step": 65230 + }, + { + "epoch": 4.432327761924174, + "grad_norm": 1.4198657274246216, + "learning_rate": 0.0004459590297594782, + "loss": 3.7147, + "step": 65235 + }, + { + "epoch": 4.432667481994836, + "grad_norm": 1.1254534721374512, + "learning_rate": 0.0004459165647506455, + "loss": 3.4951, + "step": 65240 + }, + { + "epoch": 4.4330072020654985, + "grad_norm": 0.9748072624206543, + "learning_rate": 0.00044587409974181276, + "loss": 3.4507, + "step": 65245 + }, + { + "epoch": 4.43334692213616, + "grad_norm": 0.7694755792617798, + "learning_rate": 0.00044583163473298004, + "loss": 3.878, + "step": 65250 + }, + { + "epoch": 4.433686642206822, + "grad_norm": 1.0270973443984985, + "learning_rate": 0.0004457891697241473, + "loss": 3.4706, + "step": 65255 + }, + { + "epoch": 4.434026362277484, + "grad_norm": 0.9487940669059753, + "learning_rate": 0.0004457467047153146, + "loss": 3.3172, + "step": 65260 + }, + { + "epoch": 4.434366082348145, + "grad_norm": 1.1061460971832275, + "learning_rate": 0.0004457042397064819, + "loss": 3.6393, + "step": 65265 + }, + { + "epoch": 4.434705802418807, + "grad_norm": 0.832764208316803, + "learning_rate": 0.0004456617746976491, + "loss": 3.6784, + "step": 65270 + }, + { + "epoch": 4.435045522489468, + "grad_norm": 0.9059832692146301, + "learning_rate": 0.00044561930968881644, + "loss": 3.4462, + "step": 65275 + }, + { + "epoch": 4.43538524256013, + "grad_norm": 0.9542608261108398, + "learning_rate": 0.0004455768446799837, + "loss": 3.3771, + "step": 65280 + }, + { + "epoch": 4.435724962630792, + "grad_norm": 1.0440117120742798, + "learning_rate": 0.00044553437967115094, + "loss": 3.8525, + "step": 65285 + }, + { + "epoch": 4.436064682701454, + "grad_norm": 0.954830527305603, + "learning_rate": 0.0004454919146623183, + "loss": 3.1139, + "step": 65290 + }, + { + "epoch": 4.436404402772116, + "grad_norm": 1.2768272161483765, + "learning_rate": 0.00044544944965348556, + "loss": 3.4418, + "step": 65295 + }, + { + "epoch": 4.436744122842778, + "grad_norm": 0.9730943441390991, + "learning_rate": 0.0004454069846446528, + "loss": 3.5238, + "step": 65300 + }, + { + "epoch": 4.437083842913439, + "grad_norm": 0.9132259488105774, + "learning_rate": 0.00044536451963582006, + "loss": 3.6255, + "step": 65305 + }, + { + "epoch": 4.437423562984101, + "grad_norm": 0.9365383386611938, + "learning_rate": 0.0004453220546269874, + "loss": 3.5968, + "step": 65310 + }, + { + "epoch": 4.437763283054763, + "grad_norm": 0.9567075371742249, + "learning_rate": 0.0004452795896181546, + "loss": 3.375, + "step": 65315 + }, + { + "epoch": 4.438103003125424, + "grad_norm": 0.7754851579666138, + "learning_rate": 0.0004452371246093219, + "loss": 3.6649, + "step": 65320 + }, + { + "epoch": 4.438442723196086, + "grad_norm": 1.2116923332214355, + "learning_rate": 0.00044519465960048924, + "loss": 3.3742, + "step": 65325 + }, + { + "epoch": 4.438782443266748, + "grad_norm": 0.8112424612045288, + "learning_rate": 0.0004451521945916565, + "loss": 3.3966, + "step": 65330 + }, + { + "epoch": 4.43912216333741, + "grad_norm": 0.8631371259689331, + "learning_rate": 0.00044510972958282374, + "loss": 3.7245, + "step": 65335 + }, + { + "epoch": 4.439461883408072, + "grad_norm": 0.8542788028717041, + "learning_rate": 0.000445067264573991, + "loss": 3.3919, + "step": 65340 + }, + { + "epoch": 4.439801603478734, + "grad_norm": 0.8197934031486511, + "learning_rate": 0.00044502479956515836, + "loss": 3.4249, + "step": 65345 + }, + { + "epoch": 4.440141323549395, + "grad_norm": 0.7710437178611755, + "learning_rate": 0.0004449823345563256, + "loss": 3.6363, + "step": 65350 + }, + { + "epoch": 4.440481043620057, + "grad_norm": 0.7870102524757385, + "learning_rate": 0.00044493986954749286, + "loss": 3.3879, + "step": 65355 + }, + { + "epoch": 4.440820763690719, + "grad_norm": 0.9849935173988342, + "learning_rate": 0.0004448974045386602, + "loss": 3.4856, + "step": 65360 + }, + { + "epoch": 4.44116048376138, + "grad_norm": 0.8454204797744751, + "learning_rate": 0.0004448549395298274, + "loss": 3.6144, + "step": 65365 + }, + { + "epoch": 4.441500203832042, + "grad_norm": 0.8572724461555481, + "learning_rate": 0.0004448124745209947, + "loss": 3.3884, + "step": 65370 + }, + { + "epoch": 4.441839923902704, + "grad_norm": 1.002380609512329, + "learning_rate": 0.000444770009512162, + "loss": 3.3085, + "step": 65375 + }, + { + "epoch": 4.442179643973366, + "grad_norm": 0.9569190144538879, + "learning_rate": 0.00044472754450332927, + "loss": 3.5528, + "step": 65380 + }, + { + "epoch": 4.442519364044028, + "grad_norm": 0.901733934879303, + "learning_rate": 0.00044468507949449655, + "loss": 3.2565, + "step": 65385 + }, + { + "epoch": 4.44285908411469, + "grad_norm": 0.8372353315353394, + "learning_rate": 0.0004446426144856638, + "loss": 3.5403, + "step": 65390 + }, + { + "epoch": 4.443198804185351, + "grad_norm": 1.1170272827148438, + "learning_rate": 0.0004446001494768311, + "loss": 3.5647, + "step": 65395 + }, + { + "epoch": 4.443538524256013, + "grad_norm": 1.0626559257507324, + "learning_rate": 0.0004445576844679984, + "loss": 3.6172, + "step": 65400 + }, + { + "epoch": 4.443878244326675, + "grad_norm": 0.9299407601356506, + "learning_rate": 0.00044451521945916567, + "loss": 3.3716, + "step": 65405 + }, + { + "epoch": 4.444217964397336, + "grad_norm": 0.8780089020729065, + "learning_rate": 0.0004444727544503329, + "loss": 3.3707, + "step": 65410 + }, + { + "epoch": 4.444557684467998, + "grad_norm": 0.9045068621635437, + "learning_rate": 0.0004444302894415002, + "loss": 3.5019, + "step": 65415 + }, + { + "epoch": 4.4448974045386604, + "grad_norm": 0.7839063405990601, + "learning_rate": 0.0004443878244326675, + "loss": 3.5688, + "step": 65420 + }, + { + "epoch": 4.445237124609322, + "grad_norm": 0.9780973196029663, + "learning_rate": 0.00044434535942383473, + "loss": 3.6582, + "step": 65425 + }, + { + "epoch": 4.445576844679984, + "grad_norm": 0.9142229557037354, + "learning_rate": 0.00044430289441500207, + "loss": 3.5304, + "step": 65430 + }, + { + "epoch": 4.445916564750646, + "grad_norm": 1.0334123373031616, + "learning_rate": 0.00044426042940616935, + "loss": 3.4038, + "step": 65435 + }, + { + "epoch": 4.446256284821307, + "grad_norm": 0.7088323831558228, + "learning_rate": 0.00044421796439733657, + "loss": 3.5331, + "step": 65440 + }, + { + "epoch": 4.446596004891969, + "grad_norm": 0.938849151134491, + "learning_rate": 0.00044417549938850385, + "loss": 3.4942, + "step": 65445 + }, + { + "epoch": 4.446935724962631, + "grad_norm": 0.8618916869163513, + "learning_rate": 0.0004441330343796712, + "loss": 3.538, + "step": 65450 + }, + { + "epoch": 4.447275445033292, + "grad_norm": 1.6311894655227661, + "learning_rate": 0.0004440905693708384, + "loss": 3.6887, + "step": 65455 + }, + { + "epoch": 4.447615165103954, + "grad_norm": 0.8689915537834167, + "learning_rate": 0.0004440481043620057, + "loss": 3.2764, + "step": 65460 + }, + { + "epoch": 4.4479548851746165, + "grad_norm": 1.1043154001235962, + "learning_rate": 0.000444005639353173, + "loss": 3.6724, + "step": 65465 + }, + { + "epoch": 4.448294605245278, + "grad_norm": 0.864713191986084, + "learning_rate": 0.00044396317434434025, + "loss": 3.4313, + "step": 65470 + }, + { + "epoch": 4.44863432531594, + "grad_norm": 0.8858205676078796, + "learning_rate": 0.00044392070933550753, + "loss": 3.381, + "step": 65475 + }, + { + "epoch": 4.448974045386602, + "grad_norm": 1.025192379951477, + "learning_rate": 0.00044387824432667487, + "loss": 3.384, + "step": 65480 + }, + { + "epoch": 4.449313765457263, + "grad_norm": 0.8928725123405457, + "learning_rate": 0.0004438357793178421, + "loss": 3.2779, + "step": 65485 + }, + { + "epoch": 4.449653485527925, + "grad_norm": 0.7886186838150024, + "learning_rate": 0.00044379331430900937, + "loss": 3.5011, + "step": 65490 + }, + { + "epoch": 4.449993205598587, + "grad_norm": 0.910686731338501, + "learning_rate": 0.00044375084930017665, + "loss": 3.5708, + "step": 65495 + }, + { + "epoch": 4.450332925669248, + "grad_norm": 0.6709072589874268, + "learning_rate": 0.000443708384291344, + "loss": 3.4831, + "step": 65500 + }, + { + "epoch": 4.45067264573991, + "grad_norm": 0.8429123759269714, + "learning_rate": 0.0004436659192825112, + "loss": 3.3521, + "step": 65505 + }, + { + "epoch": 4.4510123658105725, + "grad_norm": 0.8000031113624573, + "learning_rate": 0.0004436234542736785, + "loss": 3.4365, + "step": 65510 + }, + { + "epoch": 4.451352085881234, + "grad_norm": 0.9841476082801819, + "learning_rate": 0.00044358098926484583, + "loss": 3.8182, + "step": 65515 + }, + { + "epoch": 4.451691805951896, + "grad_norm": 0.8006512522697449, + "learning_rate": 0.00044353852425601305, + "loss": 3.5815, + "step": 65520 + }, + { + "epoch": 4.452031526022558, + "grad_norm": 0.9701135158538818, + "learning_rate": 0.00044349605924718033, + "loss": 3.5722, + "step": 65525 + }, + { + "epoch": 4.452371246093219, + "grad_norm": 0.6828458309173584, + "learning_rate": 0.0004434535942383476, + "loss": 3.0316, + "step": 65530 + }, + { + "epoch": 4.452710966163881, + "grad_norm": 1.0784547328948975, + "learning_rate": 0.0004434111292295149, + "loss": 3.341, + "step": 65535 + }, + { + "epoch": 4.453050686234543, + "grad_norm": 0.9231592416763306, + "learning_rate": 0.0004433686642206822, + "loss": 3.4328, + "step": 65540 + }, + { + "epoch": 4.453390406305204, + "grad_norm": 0.9386988282203674, + "learning_rate": 0.00044332619921184945, + "loss": 3.6561, + "step": 65545 + }, + { + "epoch": 4.453730126375866, + "grad_norm": 0.8226339817047119, + "learning_rate": 0.00044328373420301673, + "loss": 3.4146, + "step": 65550 + }, + { + "epoch": 4.454069846446528, + "grad_norm": 0.7354975938796997, + "learning_rate": 0.000443241269194184, + "loss": 3.5147, + "step": 65555 + }, + { + "epoch": 4.45440956651719, + "grad_norm": 0.6774933338165283, + "learning_rate": 0.0004431988041853513, + "loss": 3.45, + "step": 65560 + }, + { + "epoch": 4.454749286587852, + "grad_norm": 0.9289747476577759, + "learning_rate": 0.0004431563391765185, + "loss": 3.386, + "step": 65565 + }, + { + "epoch": 4.455089006658513, + "grad_norm": 1.1111987829208374, + "learning_rate": 0.00044311387416768585, + "loss": 3.6929, + "step": 65570 + }, + { + "epoch": 4.455428726729175, + "grad_norm": 0.8658860921859741, + "learning_rate": 0.00044307140915885313, + "loss": 3.6657, + "step": 65575 + }, + { + "epoch": 4.455768446799837, + "grad_norm": 0.7922098636627197, + "learning_rate": 0.00044302894415002036, + "loss": 3.3009, + "step": 65580 + }, + { + "epoch": 4.456108166870498, + "grad_norm": 0.9770646095275879, + "learning_rate": 0.0004429864791411877, + "loss": 3.2895, + "step": 65585 + }, + { + "epoch": 4.45644788694116, + "grad_norm": 1.0743848085403442, + "learning_rate": 0.000442944014132355, + "loss": 3.4552, + "step": 65590 + }, + { + "epoch": 4.456787607011822, + "grad_norm": 0.8378328680992126, + "learning_rate": 0.0004429015491235222, + "loss": 3.5557, + "step": 65595 + }, + { + "epoch": 4.457127327082484, + "grad_norm": 0.9926184415817261, + "learning_rate": 0.0004428590841146895, + "loss": 3.5326, + "step": 65600 + }, + { + "epoch": 4.457467047153146, + "grad_norm": 0.8915340304374695, + "learning_rate": 0.0004428166191058568, + "loss": 3.5334, + "step": 65605 + }, + { + "epoch": 4.457806767223808, + "grad_norm": 1.261305570602417, + "learning_rate": 0.00044277415409702404, + "loss": 3.5451, + "step": 65610 + }, + { + "epoch": 4.458146487294469, + "grad_norm": 0.9417118430137634, + "learning_rate": 0.0004427316890881913, + "loss": 3.5452, + "step": 65615 + }, + { + "epoch": 4.458486207365131, + "grad_norm": 0.7899617552757263, + "learning_rate": 0.00044268922407935865, + "loss": 3.7262, + "step": 65620 + }, + { + "epoch": 4.458825927435793, + "grad_norm": 0.8470172882080078, + "learning_rate": 0.0004426467590705259, + "loss": 3.2161, + "step": 65625 + }, + { + "epoch": 4.459165647506454, + "grad_norm": 1.2755868434906006, + "learning_rate": 0.00044260429406169316, + "loss": 3.5261, + "step": 65630 + }, + { + "epoch": 4.459505367577116, + "grad_norm": 0.9039474725723267, + "learning_rate": 0.00044256182905286044, + "loss": 3.4218, + "step": 65635 + }, + { + "epoch": 4.459845087647778, + "grad_norm": 0.720127284526825, + "learning_rate": 0.0004425193640440277, + "loss": 3.7125, + "step": 65640 + }, + { + "epoch": 4.46018480771844, + "grad_norm": 1.0623605251312256, + "learning_rate": 0.000442476899035195, + "loss": 3.4404, + "step": 65645 + }, + { + "epoch": 4.460524527789102, + "grad_norm": 0.8914993405342102, + "learning_rate": 0.0004424344340263623, + "loss": 3.3855, + "step": 65650 + }, + { + "epoch": 4.460864247859764, + "grad_norm": 0.9546502828598022, + "learning_rate": 0.00044239196901752956, + "loss": 3.4651, + "step": 65655 + }, + { + "epoch": 4.461203967930425, + "grad_norm": 1.0451639890670776, + "learning_rate": 0.00044234950400869684, + "loss": 3.2596, + "step": 65660 + }, + { + "epoch": 4.461543688001087, + "grad_norm": 0.698577880859375, + "learning_rate": 0.0004423070389998641, + "loss": 3.4828, + "step": 65665 + }, + { + "epoch": 4.461883408071749, + "grad_norm": 0.9439581632614136, + "learning_rate": 0.0004422645739910314, + "loss": 3.5105, + "step": 65670 + }, + { + "epoch": 4.46222312814241, + "grad_norm": 0.9097123146057129, + "learning_rate": 0.0004422221089821987, + "loss": 3.6666, + "step": 65675 + }, + { + "epoch": 4.462562848213072, + "grad_norm": 0.6626315712928772, + "learning_rate": 0.00044217964397336596, + "loss": 3.5438, + "step": 65680 + }, + { + "epoch": 4.462902568283734, + "grad_norm": 0.9837448000907898, + "learning_rate": 0.00044213717896453324, + "loss": 3.5487, + "step": 65685 + }, + { + "epoch": 4.463242288354396, + "grad_norm": 0.9373995661735535, + "learning_rate": 0.0004420947139557005, + "loss": 3.5843, + "step": 65690 + }, + { + "epoch": 4.463582008425058, + "grad_norm": 1.002805233001709, + "learning_rate": 0.0004420522489468678, + "loss": 3.234, + "step": 65695 + }, + { + "epoch": 4.46392172849572, + "grad_norm": 0.9372333884239197, + "learning_rate": 0.0004420097839380351, + "loss": 3.4055, + "step": 65700 + }, + { + "epoch": 4.464261448566381, + "grad_norm": 0.7636845111846924, + "learning_rate": 0.0004419673189292023, + "loss": 3.4936, + "step": 65705 + }, + { + "epoch": 4.464601168637043, + "grad_norm": 0.7814468145370483, + "learning_rate": 0.00044192485392036964, + "loss": 3.4019, + "step": 65710 + }, + { + "epoch": 4.464940888707705, + "grad_norm": 0.9036640524864197, + "learning_rate": 0.0004418823889115369, + "loss": 3.4396, + "step": 65715 + }, + { + "epoch": 4.465280608778366, + "grad_norm": 1.068583369255066, + "learning_rate": 0.00044183992390270415, + "loss": 3.5664, + "step": 65720 + }, + { + "epoch": 4.465620328849028, + "grad_norm": 1.1410255432128906, + "learning_rate": 0.0004417974588938715, + "loss": 3.4444, + "step": 65725 + }, + { + "epoch": 4.4659600489196905, + "grad_norm": 0.8179012537002563, + "learning_rate": 0.00044175499388503876, + "loss": 3.3289, + "step": 65730 + }, + { + "epoch": 4.466299768990352, + "grad_norm": 0.9046792387962341, + "learning_rate": 0.000441712528876206, + "loss": 3.4473, + "step": 65735 + }, + { + "epoch": 4.466639489061014, + "grad_norm": 0.9371708035469055, + "learning_rate": 0.00044167006386737327, + "loss": 3.1384, + "step": 65740 + }, + { + "epoch": 4.466979209131676, + "grad_norm": 1.0257292985916138, + "learning_rate": 0.0004416275988585406, + "loss": 3.3667, + "step": 65745 + }, + { + "epoch": 4.467318929202337, + "grad_norm": 0.6881449222564697, + "learning_rate": 0.00044158513384970783, + "loss": 3.4212, + "step": 65750 + }, + { + "epoch": 4.467658649272999, + "grad_norm": 0.888635516166687, + "learning_rate": 0.0004415426688408751, + "loss": 3.4424, + "step": 65755 + }, + { + "epoch": 4.467998369343661, + "grad_norm": 0.7769893407821655, + "learning_rate": 0.00044150020383204244, + "loss": 3.7971, + "step": 65760 + }, + { + "epoch": 4.468338089414322, + "grad_norm": 0.8105292320251465, + "learning_rate": 0.00044145773882320967, + "loss": 3.4989, + "step": 65765 + }, + { + "epoch": 4.468677809484984, + "grad_norm": 0.9504432082176208, + "learning_rate": 0.00044141527381437695, + "loss": 3.6537, + "step": 65770 + }, + { + "epoch": 4.4690175295556465, + "grad_norm": 0.7363313436508179, + "learning_rate": 0.0004413728088055443, + "loss": 3.5313, + "step": 65775 + }, + { + "epoch": 4.469357249626308, + "grad_norm": 0.8558781743049622, + "learning_rate": 0.0004413303437967115, + "loss": 3.2675, + "step": 65780 + }, + { + "epoch": 4.46969696969697, + "grad_norm": 0.810850977897644, + "learning_rate": 0.0004412878787878788, + "loss": 3.3422, + "step": 65785 + }, + { + "epoch": 4.470036689767632, + "grad_norm": 0.812824547290802, + "learning_rate": 0.00044124541377904607, + "loss": 3.4168, + "step": 65790 + }, + { + "epoch": 4.470376409838293, + "grad_norm": 0.9039174914360046, + "learning_rate": 0.00044120294877021335, + "loss": 3.6817, + "step": 65795 + }, + { + "epoch": 4.470716129908955, + "grad_norm": 1.151842474937439, + "learning_rate": 0.00044116048376138063, + "loss": 3.4555, + "step": 65800 + }, + { + "epoch": 4.471055849979617, + "grad_norm": 1.0198984146118164, + "learning_rate": 0.0004411180187525479, + "loss": 3.5422, + "step": 65805 + }, + { + "epoch": 4.471395570050278, + "grad_norm": 0.9967764019966125, + "learning_rate": 0.0004410755537437152, + "loss": 3.6507, + "step": 65810 + }, + { + "epoch": 4.47173529012094, + "grad_norm": 0.7329361438751221, + "learning_rate": 0.00044103308873488247, + "loss": 3.3682, + "step": 65815 + }, + { + "epoch": 4.4720750101916025, + "grad_norm": 0.9602144956588745, + "learning_rate": 0.00044099062372604975, + "loss": 3.4216, + "step": 65820 + }, + { + "epoch": 4.472414730262264, + "grad_norm": 1.0068944692611694, + "learning_rate": 0.000440948158717217, + "loss": 3.3587, + "step": 65825 + }, + { + "epoch": 4.472754450332926, + "grad_norm": 0.9294483065605164, + "learning_rate": 0.0004409056937083843, + "loss": 3.4904, + "step": 65830 + }, + { + "epoch": 4.473094170403588, + "grad_norm": 0.978385865688324, + "learning_rate": 0.0004408632286995516, + "loss": 3.2934, + "step": 65835 + }, + { + "epoch": 4.473433890474249, + "grad_norm": 0.9369053244590759, + "learning_rate": 0.00044082076369071887, + "loss": 3.5831, + "step": 65840 + }, + { + "epoch": 4.473773610544911, + "grad_norm": 1.0030407905578613, + "learning_rate": 0.00044077829868188615, + "loss": 3.2767, + "step": 65845 + }, + { + "epoch": 4.474113330615573, + "grad_norm": 0.9315404891967773, + "learning_rate": 0.00044073583367305343, + "loss": 3.2655, + "step": 65850 + }, + { + "epoch": 4.474453050686234, + "grad_norm": 1.962358832359314, + "learning_rate": 0.0004406933686642207, + "loss": 3.2, + "step": 65855 + }, + { + "epoch": 4.474792770756896, + "grad_norm": 1.1417176723480225, + "learning_rate": 0.00044065090365538794, + "loss": 3.3218, + "step": 65860 + }, + { + "epoch": 4.4751324908275585, + "grad_norm": 0.8394288420677185, + "learning_rate": 0.00044060843864655527, + "loss": 3.4783, + "step": 65865 + }, + { + "epoch": 4.47547221089822, + "grad_norm": 0.7305606603622437, + "learning_rate": 0.00044056597363772255, + "loss": 3.418, + "step": 65870 + }, + { + "epoch": 4.475811930968882, + "grad_norm": 1.2269339561462402, + "learning_rate": 0.0004405235086288898, + "loss": 3.5639, + "step": 65875 + }, + { + "epoch": 4.476151651039544, + "grad_norm": 1.099797248840332, + "learning_rate": 0.0004404810436200571, + "loss": 3.4951, + "step": 65880 + }, + { + "epoch": 4.476491371110205, + "grad_norm": 1.0426537990570068, + "learning_rate": 0.0004404385786112244, + "loss": 3.3501, + "step": 65885 + }, + { + "epoch": 4.476831091180867, + "grad_norm": 5.750442028045654, + "learning_rate": 0.0004403961136023916, + "loss": 3.7314, + "step": 65890 + }, + { + "epoch": 4.477170811251529, + "grad_norm": 0.8145560026168823, + "learning_rate": 0.0004403536485935589, + "loss": 3.5004, + "step": 65895 + }, + { + "epoch": 4.47751053132219, + "grad_norm": 1.0038005113601685, + "learning_rate": 0.00044031118358472623, + "loss": 3.3481, + "step": 65900 + }, + { + "epoch": 4.477850251392852, + "grad_norm": 1.0480880737304688, + "learning_rate": 0.00044026871857589346, + "loss": 3.427, + "step": 65905 + }, + { + "epoch": 4.4781899714635145, + "grad_norm": 0.7815715074539185, + "learning_rate": 0.00044022625356706074, + "loss": 3.2897, + "step": 65910 + }, + { + "epoch": 4.478529691534176, + "grad_norm": 0.7847720384597778, + "learning_rate": 0.00044018378855822807, + "loss": 3.6785, + "step": 65915 + }, + { + "epoch": 4.478869411604838, + "grad_norm": 1.0112708806991577, + "learning_rate": 0.0004401413235493953, + "loss": 3.2978, + "step": 65920 + }, + { + "epoch": 4.4792091316755, + "grad_norm": 0.8791046142578125, + "learning_rate": 0.0004400988585405626, + "loss": 3.5416, + "step": 65925 + }, + { + "epoch": 4.479548851746161, + "grad_norm": 0.8541334271430969, + "learning_rate": 0.00044005639353172986, + "loss": 3.4144, + "step": 65930 + }, + { + "epoch": 4.479888571816823, + "grad_norm": 1.0964266061782837, + "learning_rate": 0.00044001392852289714, + "loss": 3.4693, + "step": 65935 + }, + { + "epoch": 4.480228291887485, + "grad_norm": 0.7227388620376587, + "learning_rate": 0.0004399714635140644, + "loss": 3.4547, + "step": 65940 + }, + { + "epoch": 4.480568011958146, + "grad_norm": 1.1640129089355469, + "learning_rate": 0.0004399289985052317, + "loss": 3.3635, + "step": 65945 + }, + { + "epoch": 4.480907732028808, + "grad_norm": 0.9079115390777588, + "learning_rate": 0.000439886533496399, + "loss": 3.5078, + "step": 65950 + }, + { + "epoch": 4.4812474520994705, + "grad_norm": 0.9792580008506775, + "learning_rate": 0.00043984406848756626, + "loss": 3.5263, + "step": 65955 + }, + { + "epoch": 4.481587172170132, + "grad_norm": 0.9809687733650208, + "learning_rate": 0.00043980160347873354, + "loss": 3.5562, + "step": 65960 + }, + { + "epoch": 4.481926892240794, + "grad_norm": 0.9644753932952881, + "learning_rate": 0.00043975913846990076, + "loss": 3.4382, + "step": 65965 + }, + { + "epoch": 4.482266612311455, + "grad_norm": 0.9283496737480164, + "learning_rate": 0.0004397166734610681, + "loss": 3.3248, + "step": 65970 + }, + { + "epoch": 4.482606332382117, + "grad_norm": 0.9568507671356201, + "learning_rate": 0.0004396742084522354, + "loss": 3.487, + "step": 65975 + }, + { + "epoch": 4.482946052452779, + "grad_norm": 0.8961830139160156, + "learning_rate": 0.0004396317434434026, + "loss": 3.4706, + "step": 65980 + }, + { + "epoch": 4.48328577252344, + "grad_norm": 1.1774773597717285, + "learning_rate": 0.00043958927843456994, + "loss": 3.6005, + "step": 65985 + }, + { + "epoch": 4.483625492594102, + "grad_norm": 0.7419503331184387, + "learning_rate": 0.0004395468134257372, + "loss": 3.4128, + "step": 65990 + }, + { + "epoch": 4.483965212664764, + "grad_norm": 0.8950122594833374, + "learning_rate": 0.00043950434841690444, + "loss": 3.4079, + "step": 65995 + }, + { + "epoch": 4.484304932735426, + "grad_norm": 1.3669229745864868, + "learning_rate": 0.0004394618834080717, + "loss": 3.3542, + "step": 66000 + }, + { + "epoch": 4.484644652806088, + "grad_norm": 1.194487452507019, + "learning_rate": 0.00043941941839923906, + "loss": 3.6456, + "step": 66005 + }, + { + "epoch": 4.48498437287675, + "grad_norm": 0.976218044757843, + "learning_rate": 0.00043937695339040634, + "loss": 3.3337, + "step": 66010 + }, + { + "epoch": 4.485324092947411, + "grad_norm": 0.8585435748100281, + "learning_rate": 0.00043933448838157356, + "loss": 3.5737, + "step": 66015 + }, + { + "epoch": 4.485663813018073, + "grad_norm": 0.8201165795326233, + "learning_rate": 0.0004392920233727409, + "loss": 3.4264, + "step": 66020 + }, + { + "epoch": 4.486003533088735, + "grad_norm": 0.8898891806602478, + "learning_rate": 0.0004392495583639082, + "loss": 3.3647, + "step": 66025 + }, + { + "epoch": 4.486343253159396, + "grad_norm": 1.0027565956115723, + "learning_rate": 0.0004392070933550754, + "loss": 3.403, + "step": 66030 + }, + { + "epoch": 4.486682973230058, + "grad_norm": 1.0746222734451294, + "learning_rate": 0.00043916462834624274, + "loss": 3.371, + "step": 66035 + }, + { + "epoch": 4.4870226933007205, + "grad_norm": 1.0924683809280396, + "learning_rate": 0.00043912216333741, + "loss": 3.6391, + "step": 66040 + }, + { + "epoch": 4.487362413371382, + "grad_norm": 1.2874822616577148, + "learning_rate": 0.00043907969832857724, + "loss": 3.5773, + "step": 66045 + }, + { + "epoch": 4.487702133442044, + "grad_norm": 0.8159123659133911, + "learning_rate": 0.0004390372333197445, + "loss": 3.5776, + "step": 66050 + }, + { + "epoch": 4.488041853512706, + "grad_norm": 0.8107771873474121, + "learning_rate": 0.00043899476831091186, + "loss": 3.2825, + "step": 66055 + }, + { + "epoch": 4.488381573583367, + "grad_norm": 0.8560663461685181, + "learning_rate": 0.0004389523033020791, + "loss": 3.4484, + "step": 66060 + }, + { + "epoch": 4.488721293654029, + "grad_norm": 0.7566354274749756, + "learning_rate": 0.00043890983829324636, + "loss": 3.5723, + "step": 66065 + }, + { + "epoch": 4.489061013724691, + "grad_norm": 1.0398342609405518, + "learning_rate": 0.0004388673732844137, + "loss": 3.638, + "step": 66070 + }, + { + "epoch": 4.489400733795352, + "grad_norm": 1.0090651512145996, + "learning_rate": 0.0004388249082755809, + "loss": 3.4901, + "step": 66075 + }, + { + "epoch": 4.489740453866014, + "grad_norm": 0.8337435126304626, + "learning_rate": 0.0004387824432667482, + "loss": 3.348, + "step": 66080 + }, + { + "epoch": 4.4900801739366765, + "grad_norm": 0.9805718660354614, + "learning_rate": 0.0004387399782579155, + "loss": 3.508, + "step": 66085 + }, + { + "epoch": 4.490419894007338, + "grad_norm": 0.9576119184494019, + "learning_rate": 0.00043869751324908277, + "loss": 3.5723, + "step": 66090 + }, + { + "epoch": 4.490759614078, + "grad_norm": 0.9095420837402344, + "learning_rate": 0.00043865504824025005, + "loss": 3.2335, + "step": 66095 + }, + { + "epoch": 4.491099334148662, + "grad_norm": 0.9509983658790588, + "learning_rate": 0.0004386125832314173, + "loss": 3.3261, + "step": 66100 + }, + { + "epoch": 4.491439054219323, + "grad_norm": 0.728829562664032, + "learning_rate": 0.0004385701182225846, + "loss": 3.4469, + "step": 66105 + }, + { + "epoch": 4.491778774289985, + "grad_norm": 1.111512303352356, + "learning_rate": 0.0004385276532137519, + "loss": 3.3965, + "step": 66110 + }, + { + "epoch": 4.492118494360647, + "grad_norm": 0.8686105608940125, + "learning_rate": 0.00043848518820491917, + "loss": 3.4867, + "step": 66115 + }, + { + "epoch": 4.492458214431308, + "grad_norm": 0.8222286701202393, + "learning_rate": 0.0004384427231960864, + "loss": 3.7651, + "step": 66120 + }, + { + "epoch": 4.49279793450197, + "grad_norm": 0.6755975484848022, + "learning_rate": 0.0004384002581872537, + "loss": 3.4555, + "step": 66125 + }, + { + "epoch": 4.4931376545726325, + "grad_norm": 1.147624135017395, + "learning_rate": 0.000438357793178421, + "loss": 3.4567, + "step": 66130 + }, + { + "epoch": 4.493477374643294, + "grad_norm": 0.9702037572860718, + "learning_rate": 0.00043831532816958823, + "loss": 3.4348, + "step": 66135 + }, + { + "epoch": 4.493817094713956, + "grad_norm": 0.7835829257965088, + "learning_rate": 0.00043827286316075557, + "loss": 3.5303, + "step": 66140 + }, + { + "epoch": 4.494156814784618, + "grad_norm": 0.8331335186958313, + "learning_rate": 0.00043823039815192285, + "loss": 3.4823, + "step": 66145 + }, + { + "epoch": 4.494496534855279, + "grad_norm": 0.8219634294509888, + "learning_rate": 0.00043818793314309007, + "loss": 3.5153, + "step": 66150 + }, + { + "epoch": 4.494836254925941, + "grad_norm": 0.7200407981872559, + "learning_rate": 0.00043814546813425735, + "loss": 3.4107, + "step": 66155 + }, + { + "epoch": 4.495175974996603, + "grad_norm": 0.9197948575019836, + "learning_rate": 0.0004381030031254247, + "loss": 3.6076, + "step": 66160 + }, + { + "epoch": 4.495515695067264, + "grad_norm": 0.9910964965820312, + "learning_rate": 0.0004380605381165919, + "loss": 3.4201, + "step": 66165 + }, + { + "epoch": 4.495855415137926, + "grad_norm": 1.0844299793243408, + "learning_rate": 0.0004380180731077592, + "loss": 3.3236, + "step": 66170 + }, + { + "epoch": 4.4961951352085885, + "grad_norm": 1.0974396467208862, + "learning_rate": 0.0004379756080989265, + "loss": 3.5541, + "step": 66175 + }, + { + "epoch": 4.49653485527925, + "grad_norm": 0.8741354942321777, + "learning_rate": 0.0004379331430900938, + "loss": 3.4041, + "step": 66180 + }, + { + "epoch": 4.496874575349912, + "grad_norm": 1.0638731718063354, + "learning_rate": 0.00043789067808126103, + "loss": 3.3409, + "step": 66185 + }, + { + "epoch": 4.497214295420574, + "grad_norm": 0.8131060004234314, + "learning_rate": 0.0004378482130724283, + "loss": 3.7395, + "step": 66190 + }, + { + "epoch": 4.497554015491235, + "grad_norm": 0.8135803937911987, + "learning_rate": 0.00043780574806359565, + "loss": 3.6257, + "step": 66195 + }, + { + "epoch": 4.497893735561897, + "grad_norm": 1.1259260177612305, + "learning_rate": 0.00043776328305476287, + "loss": 3.5372, + "step": 66200 + }, + { + "epoch": 4.498233455632559, + "grad_norm": 1.0854525566101074, + "learning_rate": 0.00043772081804593015, + "loss": 3.3179, + "step": 66205 + }, + { + "epoch": 4.49857317570322, + "grad_norm": 0.8077121376991272, + "learning_rate": 0.0004376783530370975, + "loss": 3.669, + "step": 66210 + }, + { + "epoch": 4.498912895773882, + "grad_norm": 0.8429264426231384, + "learning_rate": 0.0004376358880282647, + "loss": 3.3968, + "step": 66215 + }, + { + "epoch": 4.4992526158445445, + "grad_norm": 0.7988156080245972, + "learning_rate": 0.000437593423019432, + "loss": 3.4344, + "step": 66220 + }, + { + "epoch": 4.499592335915206, + "grad_norm": 1.184632658958435, + "learning_rate": 0.0004375509580105993, + "loss": 3.6033, + "step": 66225 + }, + { + "epoch": 4.499932055985868, + "grad_norm": 0.771209716796875, + "learning_rate": 0.00043750849300176655, + "loss": 3.5254, + "step": 66230 + }, + { + "epoch": 4.500271776056529, + "grad_norm": 0.8359200358390808, + "learning_rate": 0.00043746602799293383, + "loss": 3.3722, + "step": 66235 + }, + { + "epoch": 4.500611496127191, + "grad_norm": 1.0414351224899292, + "learning_rate": 0.0004374235629841011, + "loss": 3.2538, + "step": 66240 + }, + { + "epoch": 4.500951216197853, + "grad_norm": 1.0110838413238525, + "learning_rate": 0.0004373810979752684, + "loss": 3.5275, + "step": 66245 + }, + { + "epoch": 4.501290936268514, + "grad_norm": 0.9511992335319519, + "learning_rate": 0.0004373386329664357, + "loss": 3.4831, + "step": 66250 + }, + { + "epoch": 4.501630656339176, + "grad_norm": 0.8751468062400818, + "learning_rate": 0.00043729616795760295, + "loss": 3.1557, + "step": 66255 + }, + { + "epoch": 4.501970376409838, + "grad_norm": 0.8245137333869934, + "learning_rate": 0.0004372537029487702, + "loss": 3.4178, + "step": 66260 + }, + { + "epoch": 4.5023100964805, + "grad_norm": 0.807417094707489, + "learning_rate": 0.0004372112379399375, + "loss": 3.3439, + "step": 66265 + }, + { + "epoch": 4.502649816551162, + "grad_norm": 0.8834036588668823, + "learning_rate": 0.0004371687729311048, + "loss": 3.3613, + "step": 66270 + }, + { + "epoch": 4.502989536621824, + "grad_norm": 0.8848956823348999, + "learning_rate": 0.000437126307922272, + "loss": 3.4967, + "step": 66275 + }, + { + "epoch": 4.503329256692485, + "grad_norm": 1.1724224090576172, + "learning_rate": 0.00043708384291343935, + "loss": 3.4971, + "step": 66280 + }, + { + "epoch": 4.503668976763147, + "grad_norm": 0.9957519173622131, + "learning_rate": 0.00043704137790460663, + "loss": 3.5982, + "step": 66285 + }, + { + "epoch": 4.504008696833809, + "grad_norm": 0.9878401160240173, + "learning_rate": 0.00043699891289577386, + "loss": 3.6778, + "step": 66290 + }, + { + "epoch": 4.50434841690447, + "grad_norm": 0.8867027163505554, + "learning_rate": 0.00043695644788694114, + "loss": 3.4274, + "step": 66295 + }, + { + "epoch": 4.504688136975132, + "grad_norm": 0.8326494097709656, + "learning_rate": 0.0004369139828781085, + "loss": 3.5251, + "step": 66300 + }, + { + "epoch": 4.5050278570457944, + "grad_norm": 0.7408300042152405, + "learning_rate": 0.0004368715178692757, + "loss": 3.5677, + "step": 66305 + }, + { + "epoch": 4.505367577116456, + "grad_norm": 0.8756265044212341, + "learning_rate": 0.000436829052860443, + "loss": 3.2307, + "step": 66310 + }, + { + "epoch": 4.505707297187118, + "grad_norm": 0.9377771615982056, + "learning_rate": 0.0004367865878516103, + "loss": 3.4558, + "step": 66315 + }, + { + "epoch": 4.50604701725778, + "grad_norm": 0.9203988313674927, + "learning_rate": 0.00043674412284277754, + "loss": 3.5058, + "step": 66320 + }, + { + "epoch": 4.506386737328441, + "grad_norm": 0.850809633731842, + "learning_rate": 0.0004367016578339448, + "loss": 3.6603, + "step": 66325 + }, + { + "epoch": 4.506726457399103, + "grad_norm": 0.8915470242500305, + "learning_rate": 0.00043665919282511215, + "loss": 3.44, + "step": 66330 + }, + { + "epoch": 4.507066177469765, + "grad_norm": 0.7421795725822449, + "learning_rate": 0.0004366167278162794, + "loss": 3.5953, + "step": 66335 + }, + { + "epoch": 4.507405897540426, + "grad_norm": 1.2483971118927002, + "learning_rate": 0.00043657426280744666, + "loss": 3.0663, + "step": 66340 + }, + { + "epoch": 4.507745617611088, + "grad_norm": 0.9260044097900391, + "learning_rate": 0.00043653179779861394, + "loss": 3.7313, + "step": 66345 + }, + { + "epoch": 4.5080853376817505, + "grad_norm": 0.8949269652366638, + "learning_rate": 0.0004364893327897812, + "loss": 3.4812, + "step": 66350 + }, + { + "epoch": 4.508425057752412, + "grad_norm": 1.0003306865692139, + "learning_rate": 0.0004364468677809485, + "loss": 3.5219, + "step": 66355 + }, + { + "epoch": 4.508764777823074, + "grad_norm": 0.6728094816207886, + "learning_rate": 0.0004364044027721158, + "loss": 3.6028, + "step": 66360 + }, + { + "epoch": 4.509104497893736, + "grad_norm": 0.8707903027534485, + "learning_rate": 0.0004363619377632831, + "loss": 3.7663, + "step": 66365 + }, + { + "epoch": 4.509444217964397, + "grad_norm": 1.1475162506103516, + "learning_rate": 0.00043631947275445034, + "loss": 3.5086, + "step": 66370 + }, + { + "epoch": 4.509783938035059, + "grad_norm": 0.9056214094161987, + "learning_rate": 0.0004362770077456176, + "loss": 3.3424, + "step": 66375 + }, + { + "epoch": 4.510123658105721, + "grad_norm": 0.8202046751976013, + "learning_rate": 0.0004362345427367849, + "loss": 3.5289, + "step": 66380 + }, + { + "epoch": 4.510463378176382, + "grad_norm": 0.9594817161560059, + "learning_rate": 0.0004361920777279522, + "loss": 3.6661, + "step": 66385 + }, + { + "epoch": 4.510803098247044, + "grad_norm": 1.0026971101760864, + "learning_rate": 0.00043614961271911946, + "loss": 3.5113, + "step": 66390 + }, + { + "epoch": 4.5111428183177065, + "grad_norm": 0.7385213971138, + "learning_rate": 0.00043610714771028674, + "loss": 3.3318, + "step": 66395 + }, + { + "epoch": 4.511482538388368, + "grad_norm": 1.0562478303909302, + "learning_rate": 0.000436064682701454, + "loss": 3.2828, + "step": 66400 + }, + { + "epoch": 4.51182225845903, + "grad_norm": 1.2695852518081665, + "learning_rate": 0.0004360222176926213, + "loss": 3.553, + "step": 66405 + }, + { + "epoch": 4.512161978529692, + "grad_norm": 1.3713313341140747, + "learning_rate": 0.0004359797526837886, + "loss": 3.4712, + "step": 66410 + }, + { + "epoch": 4.512501698600353, + "grad_norm": 0.8170470595359802, + "learning_rate": 0.0004359372876749558, + "loss": 3.4437, + "step": 66415 + }, + { + "epoch": 4.512841418671015, + "grad_norm": 1.041155457496643, + "learning_rate": 0.00043589482266612314, + "loss": 3.4742, + "step": 66420 + }, + { + "epoch": 4.513181138741677, + "grad_norm": 0.8515216708183289, + "learning_rate": 0.0004358523576572904, + "loss": 3.3389, + "step": 66425 + }, + { + "epoch": 4.513520858812338, + "grad_norm": 0.9591215252876282, + "learning_rate": 0.00043580989264845765, + "loss": 3.302, + "step": 66430 + }, + { + "epoch": 4.513860578883, + "grad_norm": 0.8770819902420044, + "learning_rate": 0.000435767427639625, + "loss": 3.6477, + "step": 66435 + }, + { + "epoch": 4.5142002989536625, + "grad_norm": 0.9640154838562012, + "learning_rate": 0.00043572496263079226, + "loss": 3.6178, + "step": 66440 + }, + { + "epoch": 4.514540019024324, + "grad_norm": 0.9260295629501343, + "learning_rate": 0.0004356824976219595, + "loss": 3.443, + "step": 66445 + }, + { + "epoch": 4.514879739094986, + "grad_norm": 0.9172662496566772, + "learning_rate": 0.00043564003261312677, + "loss": 3.5585, + "step": 66450 + }, + { + "epoch": 4.515219459165648, + "grad_norm": 0.8893214464187622, + "learning_rate": 0.0004355975676042941, + "loss": 3.3658, + "step": 66455 + }, + { + "epoch": 4.515559179236309, + "grad_norm": 0.7086695432662964, + "learning_rate": 0.00043555510259546133, + "loss": 3.4416, + "step": 66460 + }, + { + "epoch": 4.515898899306971, + "grad_norm": 0.7693719267845154, + "learning_rate": 0.0004355126375866286, + "loss": 3.5764, + "step": 66465 + }, + { + "epoch": 4.516238619377633, + "grad_norm": 0.8776002526283264, + "learning_rate": 0.00043547017257779594, + "loss": 3.5417, + "step": 66470 + }, + { + "epoch": 4.516578339448294, + "grad_norm": 1.1325730085372925, + "learning_rate": 0.00043542770756896317, + "loss": 3.2041, + "step": 66475 + }, + { + "epoch": 4.516918059518956, + "grad_norm": 0.8800426125526428, + "learning_rate": 0.00043538524256013045, + "loss": 3.2383, + "step": 66480 + }, + { + "epoch": 4.5172577795896185, + "grad_norm": 0.7938007712364197, + "learning_rate": 0.00043534277755129773, + "loss": 3.4293, + "step": 66485 + }, + { + "epoch": 4.51759749966028, + "grad_norm": 0.8930371403694153, + "learning_rate": 0.000435300312542465, + "loss": 3.5067, + "step": 66490 + }, + { + "epoch": 4.517937219730942, + "grad_norm": 0.9529256224632263, + "learning_rate": 0.0004352578475336323, + "loss": 3.4535, + "step": 66495 + }, + { + "epoch": 4.518276939801604, + "grad_norm": 0.9461964964866638, + "learning_rate": 0.00043521538252479957, + "loss": 3.5666, + "step": 66500 + }, + { + "epoch": 4.518616659872265, + "grad_norm": 0.722118079662323, + "learning_rate": 0.00043517291751596685, + "loss": 3.3239, + "step": 66505 + }, + { + "epoch": 4.518956379942927, + "grad_norm": 1.0842740535736084, + "learning_rate": 0.00043513045250713413, + "loss": 3.5684, + "step": 66510 + }, + { + "epoch": 4.519296100013589, + "grad_norm": 0.9255064725875854, + "learning_rate": 0.0004350879874983014, + "loss": 3.4622, + "step": 66515 + }, + { + "epoch": 4.51963582008425, + "grad_norm": 0.8189423084259033, + "learning_rate": 0.00043504552248946863, + "loss": 3.5544, + "step": 66520 + }, + { + "epoch": 4.519975540154912, + "grad_norm": 0.8148610591888428, + "learning_rate": 0.00043500305748063597, + "loss": 3.5503, + "step": 66525 + }, + { + "epoch": 4.5203152602255745, + "grad_norm": 0.7528272271156311, + "learning_rate": 0.00043496059247180325, + "loss": 3.5459, + "step": 66530 + }, + { + "epoch": 4.520654980296236, + "grad_norm": 0.757683515548706, + "learning_rate": 0.00043491812746297053, + "loss": 3.3432, + "step": 66535 + }, + { + "epoch": 4.520994700366898, + "grad_norm": 0.9672046899795532, + "learning_rate": 0.0004348756624541378, + "loss": 3.6097, + "step": 66540 + }, + { + "epoch": 4.52133442043756, + "grad_norm": 0.8043573498725891, + "learning_rate": 0.0004348331974453051, + "loss": 3.4675, + "step": 66545 + }, + { + "epoch": 4.521674140508221, + "grad_norm": 0.9672068357467651, + "learning_rate": 0.00043479073243647237, + "loss": 3.5262, + "step": 66550 + }, + { + "epoch": 4.522013860578883, + "grad_norm": 0.8367587327957153, + "learning_rate": 0.0004347482674276396, + "loss": 3.4286, + "step": 66555 + }, + { + "epoch": 4.522353580649545, + "grad_norm": 0.8188043236732483, + "learning_rate": 0.00043470580241880693, + "loss": 3.3617, + "step": 66560 + }, + { + "epoch": 4.522693300720206, + "grad_norm": 0.843798041343689, + "learning_rate": 0.0004346633374099742, + "loss": 3.481, + "step": 66565 + }, + { + "epoch": 4.523033020790868, + "grad_norm": 0.7933496236801147, + "learning_rate": 0.00043462087240114144, + "loss": 3.7504, + "step": 66570 + }, + { + "epoch": 4.5233727408615305, + "grad_norm": 0.9600605368614197, + "learning_rate": 0.00043457840739230877, + "loss": 3.2795, + "step": 66575 + }, + { + "epoch": 4.523712460932192, + "grad_norm": 0.732108473777771, + "learning_rate": 0.00043453594238347605, + "loss": 3.4566, + "step": 66580 + }, + { + "epoch": 4.524052181002854, + "grad_norm": 0.9429569840431213, + "learning_rate": 0.0004344934773746433, + "loss": 3.4503, + "step": 66585 + }, + { + "epoch": 4.524391901073516, + "grad_norm": 0.7102686762809753, + "learning_rate": 0.00043445101236581056, + "loss": 3.6884, + "step": 66590 + }, + { + "epoch": 4.524731621144177, + "grad_norm": 0.6802005171775818, + "learning_rate": 0.0004344085473569779, + "loss": 3.3665, + "step": 66595 + }, + { + "epoch": 4.525071341214839, + "grad_norm": 0.925494372844696, + "learning_rate": 0.0004343660823481451, + "loss": 3.6666, + "step": 66600 + }, + { + "epoch": 4.525411061285501, + "grad_norm": 1.1452659368515015, + "learning_rate": 0.0004343236173393124, + "loss": 3.6077, + "step": 66605 + }, + { + "epoch": 4.525750781356162, + "grad_norm": 0.9466968178749084, + "learning_rate": 0.00043428115233047973, + "loss": 3.4309, + "step": 66610 + }, + { + "epoch": 4.5260905014268245, + "grad_norm": 1.0045777559280396, + "learning_rate": 0.00043423868732164696, + "loss": 3.4496, + "step": 66615 + }, + { + "epoch": 4.5264302214974865, + "grad_norm": 0.9907852411270142, + "learning_rate": 0.00043419622231281424, + "loss": 3.3222, + "step": 66620 + }, + { + "epoch": 4.526769941568148, + "grad_norm": 0.658758819103241, + "learning_rate": 0.00043415375730398157, + "loss": 3.2718, + "step": 66625 + }, + { + "epoch": 4.52710966163881, + "grad_norm": 1.2165133953094482, + "learning_rate": 0.0004341112922951488, + "loss": 3.4049, + "step": 66630 + }, + { + "epoch": 4.527449381709472, + "grad_norm": 0.8876256942749023, + "learning_rate": 0.0004340688272863161, + "loss": 3.434, + "step": 66635 + }, + { + "epoch": 4.527789101780133, + "grad_norm": 0.7738040089607239, + "learning_rate": 0.00043402636227748336, + "loss": 3.3506, + "step": 66640 + }, + { + "epoch": 4.528128821850795, + "grad_norm": 0.9086374640464783, + "learning_rate": 0.00043398389726865064, + "loss": 3.5073, + "step": 66645 + }, + { + "epoch": 4.528468541921457, + "grad_norm": 0.760643720626831, + "learning_rate": 0.0004339414322598179, + "loss": 3.4905, + "step": 66650 + }, + { + "epoch": 4.528808261992118, + "grad_norm": 0.7796370983123779, + "learning_rate": 0.0004338989672509852, + "loss": 3.5582, + "step": 66655 + }, + { + "epoch": 4.5291479820627805, + "grad_norm": 0.7682840824127197, + "learning_rate": 0.0004338565022421525, + "loss": 3.3667, + "step": 66660 + }, + { + "epoch": 4.5294877021334425, + "grad_norm": 0.9640697240829468, + "learning_rate": 0.00043381403723331976, + "loss": 3.4957, + "step": 66665 + }, + { + "epoch": 4.529827422204104, + "grad_norm": 0.9150825142860413, + "learning_rate": 0.00043377157222448704, + "loss": 3.6277, + "step": 66670 + }, + { + "epoch": 4.530167142274766, + "grad_norm": 0.9334050416946411, + "learning_rate": 0.00043372910721565426, + "loss": 3.6875, + "step": 66675 + }, + { + "epoch": 4.530506862345427, + "grad_norm": 0.9090102314949036, + "learning_rate": 0.0004336866422068216, + "loss": 3.6752, + "step": 66680 + }, + { + "epoch": 4.530846582416089, + "grad_norm": 0.8971827030181885, + "learning_rate": 0.0004336441771979889, + "loss": 3.5559, + "step": 66685 + }, + { + "epoch": 4.531186302486751, + "grad_norm": 0.8988327383995056, + "learning_rate": 0.0004336017121891561, + "loss": 3.432, + "step": 66690 + }, + { + "epoch": 4.531526022557412, + "grad_norm": 0.8552560806274414, + "learning_rate": 0.00043355924718032344, + "loss": 3.4808, + "step": 66695 + }, + { + "epoch": 4.531865742628074, + "grad_norm": 0.708320677280426, + "learning_rate": 0.0004335167821714907, + "loss": 3.4108, + "step": 66700 + }, + { + "epoch": 4.5322054626987365, + "grad_norm": 0.980912983417511, + "learning_rate": 0.000433474317162658, + "loss": 3.5147, + "step": 66705 + }, + { + "epoch": 4.532545182769398, + "grad_norm": 0.9449055790901184, + "learning_rate": 0.0004334318521538252, + "loss": 3.5526, + "step": 66710 + }, + { + "epoch": 4.53288490284006, + "grad_norm": 1.0668842792510986, + "learning_rate": 0.00043338938714499256, + "loss": 3.3522, + "step": 66715 + }, + { + "epoch": 4.533224622910722, + "grad_norm": 1.159994125366211, + "learning_rate": 0.00043334692213615984, + "loss": 3.685, + "step": 66720 + }, + { + "epoch": 4.533564342981383, + "grad_norm": 0.9329102635383606, + "learning_rate": 0.00043330445712732706, + "loss": 3.5058, + "step": 66725 + }, + { + "epoch": 4.533904063052045, + "grad_norm": 0.9496855139732361, + "learning_rate": 0.0004332619921184944, + "loss": 3.4115, + "step": 66730 + }, + { + "epoch": 4.534243783122707, + "grad_norm": 0.7862743735313416, + "learning_rate": 0.0004332195271096617, + "loss": 3.4623, + "step": 66735 + }, + { + "epoch": 4.534583503193368, + "grad_norm": 0.7804041504859924, + "learning_rate": 0.0004331770621008289, + "loss": 3.5089, + "step": 66740 + }, + { + "epoch": 4.53492322326403, + "grad_norm": 0.8277208805084229, + "learning_rate": 0.0004331345970919962, + "loss": 3.6347, + "step": 66745 + }, + { + "epoch": 4.5352629433346925, + "grad_norm": 1.4659042358398438, + "learning_rate": 0.0004330921320831635, + "loss": 3.5089, + "step": 66750 + }, + { + "epoch": 4.535602663405354, + "grad_norm": 0.9748002886772156, + "learning_rate": 0.00043304966707433074, + "loss": 3.3099, + "step": 66755 + }, + { + "epoch": 4.535942383476016, + "grad_norm": 0.7658646702766418, + "learning_rate": 0.000433007202065498, + "loss": 3.5651, + "step": 66760 + }, + { + "epoch": 4.536282103546678, + "grad_norm": 0.7752154469490051, + "learning_rate": 0.00043296473705666536, + "loss": 3.2891, + "step": 66765 + }, + { + "epoch": 4.536621823617339, + "grad_norm": 0.8406447768211365, + "learning_rate": 0.0004329222720478326, + "loss": 3.3332, + "step": 66770 + }, + { + "epoch": 4.536961543688001, + "grad_norm": 0.931098461151123, + "learning_rate": 0.00043287980703899986, + "loss": 3.5949, + "step": 66775 + }, + { + "epoch": 4.537301263758663, + "grad_norm": 0.8848597407341003, + "learning_rate": 0.00043283734203016714, + "loss": 3.2596, + "step": 66780 + }, + { + "epoch": 4.537640983829324, + "grad_norm": 1.1774585247039795, + "learning_rate": 0.0004327948770213344, + "loss": 3.6514, + "step": 66785 + }, + { + "epoch": 4.537980703899986, + "grad_norm": 1.0215868949890137, + "learning_rate": 0.0004327524120125017, + "loss": 3.5618, + "step": 66790 + }, + { + "epoch": 4.5383204239706485, + "grad_norm": 0.9636752009391785, + "learning_rate": 0.000432709947003669, + "loss": 3.6018, + "step": 66795 + }, + { + "epoch": 4.53866014404131, + "grad_norm": 0.982258141040802, + "learning_rate": 0.00043266748199483627, + "loss": 3.5931, + "step": 66800 + }, + { + "epoch": 4.538999864111972, + "grad_norm": 0.9146232604980469, + "learning_rate": 0.00043262501698600355, + "loss": 3.6283, + "step": 66805 + }, + { + "epoch": 4.539339584182634, + "grad_norm": 0.8624289035797119, + "learning_rate": 0.0004325825519771708, + "loss": 3.9439, + "step": 66810 + }, + { + "epoch": 4.539679304253295, + "grad_norm": 1.0013322830200195, + "learning_rate": 0.00043254008696833805, + "loss": 3.6557, + "step": 66815 + }, + { + "epoch": 4.540019024323957, + "grad_norm": 0.8071323037147522, + "learning_rate": 0.0004324976219595054, + "loss": 3.5187, + "step": 66820 + }, + { + "epoch": 4.540358744394619, + "grad_norm": 0.7959896326065063, + "learning_rate": 0.00043245515695067267, + "loss": 3.5998, + "step": 66825 + }, + { + "epoch": 4.54069846446528, + "grad_norm": 0.8444793224334717, + "learning_rate": 0.0004324126919418399, + "loss": 3.4634, + "step": 66830 + }, + { + "epoch": 4.541038184535942, + "grad_norm": 1.0232434272766113, + "learning_rate": 0.0004323702269330072, + "loss": 3.5174, + "step": 66835 + }, + { + "epoch": 4.5413779046066045, + "grad_norm": 0.8612986207008362, + "learning_rate": 0.0004323277619241745, + "loss": 3.4453, + "step": 66840 + }, + { + "epoch": 4.541717624677266, + "grad_norm": 1.2115777730941772, + "learning_rate": 0.00043228529691534173, + "loss": 3.3922, + "step": 66845 + }, + { + "epoch": 4.542057344747928, + "grad_norm": 0.9160543084144592, + "learning_rate": 0.000432242831906509, + "loss": 3.6223, + "step": 66850 + }, + { + "epoch": 4.54239706481859, + "grad_norm": 0.8424810767173767, + "learning_rate": 0.00043220036689767635, + "loss": 3.4271, + "step": 66855 + }, + { + "epoch": 4.542736784889251, + "grad_norm": 0.7096657156944275, + "learning_rate": 0.00043215790188884357, + "loss": 3.4862, + "step": 66860 + }, + { + "epoch": 4.543076504959913, + "grad_norm": 0.9247938990592957, + "learning_rate": 0.00043211543688001085, + "loss": 3.5417, + "step": 66865 + }, + { + "epoch": 4.543416225030575, + "grad_norm": 1.4372491836547852, + "learning_rate": 0.0004320729718711782, + "loss": 3.4718, + "step": 66870 + }, + { + "epoch": 4.543755945101236, + "grad_norm": 0.7821809649467468, + "learning_rate": 0.00043203050686234547, + "loss": 3.5486, + "step": 66875 + }, + { + "epoch": 4.5440956651718984, + "grad_norm": 0.8051093816757202, + "learning_rate": 0.0004319880418535127, + "loss": 3.631, + "step": 66880 + }, + { + "epoch": 4.54443538524256, + "grad_norm": 0.9351738095283508, + "learning_rate": 0.00043194557684468, + "loss": 3.45, + "step": 66885 + }, + { + "epoch": 4.544775105313222, + "grad_norm": 1.1868656873703003, + "learning_rate": 0.0004319031118358473, + "loss": 3.3131, + "step": 66890 + }, + { + "epoch": 4.545114825383884, + "grad_norm": 0.7916035652160645, + "learning_rate": 0.00043186064682701453, + "loss": 3.3709, + "step": 66895 + }, + { + "epoch": 4.545454545454545, + "grad_norm": 0.9459665417671204, + "learning_rate": 0.0004318181818181818, + "loss": 3.3845, + "step": 66900 + }, + { + "epoch": 4.545794265525207, + "grad_norm": 0.9013563990592957, + "learning_rate": 0.00043177571680934915, + "loss": 3.428, + "step": 66905 + }, + { + "epoch": 4.546133985595869, + "grad_norm": 0.9774910807609558, + "learning_rate": 0.00043173325180051637, + "loss": 3.6417, + "step": 66910 + }, + { + "epoch": 4.54647370566653, + "grad_norm": 1.0090702772140503, + "learning_rate": 0.00043169078679168365, + "loss": 3.4969, + "step": 66915 + }, + { + "epoch": 4.546813425737192, + "grad_norm": 0.8805792331695557, + "learning_rate": 0.000431648321782851, + "loss": 3.4828, + "step": 66920 + }, + { + "epoch": 4.5471531458078545, + "grad_norm": 0.7668918967247009, + "learning_rate": 0.0004316058567740182, + "loss": 3.651, + "step": 66925 + }, + { + "epoch": 4.547492865878516, + "grad_norm": 0.9279001951217651, + "learning_rate": 0.0004315633917651855, + "loss": 3.6542, + "step": 66930 + }, + { + "epoch": 4.547832585949178, + "grad_norm": 0.7266154289245605, + "learning_rate": 0.0004315209267563528, + "loss": 3.3583, + "step": 66935 + }, + { + "epoch": 4.54817230601984, + "grad_norm": 1.2068816423416138, + "learning_rate": 0.00043147846174752005, + "loss": 3.3139, + "step": 66940 + }, + { + "epoch": 4.548512026090501, + "grad_norm": 0.8246268033981323, + "learning_rate": 0.00043143599673868733, + "loss": 3.5793, + "step": 66945 + }, + { + "epoch": 4.548851746161163, + "grad_norm": 1.0142943859100342, + "learning_rate": 0.0004313935317298546, + "loss": 3.6196, + "step": 66950 + }, + { + "epoch": 4.549191466231825, + "grad_norm": 1.0291627645492554, + "learning_rate": 0.0004313510667210219, + "loss": 3.5292, + "step": 66955 + }, + { + "epoch": 4.549531186302486, + "grad_norm": 0.8368141651153564, + "learning_rate": 0.0004313086017121892, + "loss": 3.5346, + "step": 66960 + }, + { + "epoch": 4.549870906373148, + "grad_norm": 0.968543291091919, + "learning_rate": 0.00043126613670335645, + "loss": 3.4048, + "step": 66965 + }, + { + "epoch": 4.5502106264438105, + "grad_norm": 0.8095539808273315, + "learning_rate": 0.0004312236716945237, + "loss": 3.4625, + "step": 66970 + }, + { + "epoch": 4.550550346514472, + "grad_norm": 0.9477177262306213, + "learning_rate": 0.000431181206685691, + "loss": 3.3573, + "step": 66975 + }, + { + "epoch": 4.550890066585134, + "grad_norm": 0.8332407474517822, + "learning_rate": 0.0004311387416768583, + "loss": 3.4069, + "step": 66980 + }, + { + "epoch": 4.551229786655796, + "grad_norm": 0.8267947435379028, + "learning_rate": 0.0004310962766680255, + "loss": 3.5259, + "step": 66985 + }, + { + "epoch": 4.551569506726457, + "grad_norm": 1.01543128490448, + "learning_rate": 0.00043105381165919285, + "loss": 3.4005, + "step": 66990 + }, + { + "epoch": 4.551909226797119, + "grad_norm": 0.911358118057251, + "learning_rate": 0.00043101134665036013, + "loss": 3.525, + "step": 66995 + }, + { + "epoch": 4.552248946867781, + "grad_norm": 0.9308170676231384, + "learning_rate": 0.00043096888164152736, + "loss": 3.2679, + "step": 67000 + }, + { + "epoch": 4.552588666938442, + "grad_norm": 0.7572134137153625, + "learning_rate": 0.00043092641663269464, + "loss": 3.4205, + "step": 67005 + }, + { + "epoch": 4.552928387009104, + "grad_norm": 0.6687334775924683, + "learning_rate": 0.000430883951623862, + "loss": 3.4311, + "step": 67010 + }, + { + "epoch": 4.5532681070797665, + "grad_norm": 0.9449138641357422, + "learning_rate": 0.0004308414866150292, + "loss": 3.4864, + "step": 67015 + }, + { + "epoch": 4.553607827150428, + "grad_norm": 0.9813666939735413, + "learning_rate": 0.0004307990216061965, + "loss": 3.3891, + "step": 67020 + }, + { + "epoch": 4.55394754722109, + "grad_norm": 1.0032155513763428, + "learning_rate": 0.0004307565565973638, + "loss": 3.5482, + "step": 67025 + }, + { + "epoch": 4.554287267291752, + "grad_norm": 1.0856163501739502, + "learning_rate": 0.00043071409158853104, + "loss": 3.5185, + "step": 67030 + }, + { + "epoch": 4.554626987362413, + "grad_norm": 1.0040206909179688, + "learning_rate": 0.0004306716265796983, + "loss": 3.3138, + "step": 67035 + }, + { + "epoch": 4.554966707433075, + "grad_norm": 1.2751375436782837, + "learning_rate": 0.0004306291615708656, + "loss": 3.6216, + "step": 67040 + }, + { + "epoch": 4.555306427503737, + "grad_norm": 0.9331932067871094, + "learning_rate": 0.00043058669656203293, + "loss": 3.4206, + "step": 67045 + }, + { + "epoch": 4.555646147574398, + "grad_norm": 0.879595935344696, + "learning_rate": 0.00043054423155320016, + "loss": 3.3224, + "step": 67050 + }, + { + "epoch": 4.55598586764506, + "grad_norm": 0.9175193309783936, + "learning_rate": 0.00043050176654436744, + "loss": 3.5689, + "step": 67055 + }, + { + "epoch": 4.5563255877157225, + "grad_norm": 0.9314781427383423, + "learning_rate": 0.0004304593015355348, + "loss": 3.5726, + "step": 67060 + }, + { + "epoch": 4.556665307786384, + "grad_norm": 1.1162173748016357, + "learning_rate": 0.000430416836526702, + "loss": 3.6819, + "step": 67065 + }, + { + "epoch": 4.557005027857046, + "grad_norm": 1.0832526683807373, + "learning_rate": 0.0004303743715178693, + "loss": 3.4827, + "step": 67070 + }, + { + "epoch": 4.557344747927708, + "grad_norm": 0.9020607471466064, + "learning_rate": 0.00043033190650903656, + "loss": 3.8058, + "step": 67075 + }, + { + "epoch": 4.557684467998369, + "grad_norm": 0.854320764541626, + "learning_rate": 0.00043028944150020384, + "loss": 3.4311, + "step": 67080 + }, + { + "epoch": 4.558024188069031, + "grad_norm": 0.9069700837135315, + "learning_rate": 0.0004302469764913711, + "loss": 3.6863, + "step": 67085 + }, + { + "epoch": 4.558363908139693, + "grad_norm": 0.6865811347961426, + "learning_rate": 0.0004302045114825384, + "loss": 3.4372, + "step": 67090 + }, + { + "epoch": 4.558703628210354, + "grad_norm": 0.7132288813591003, + "learning_rate": 0.0004301620464737057, + "loss": 3.5355, + "step": 67095 + }, + { + "epoch": 4.559043348281016, + "grad_norm": 1.061149001121521, + "learning_rate": 0.00043011958146487296, + "loss": 3.2397, + "step": 67100 + }, + { + "epoch": 4.5593830683516785, + "grad_norm": 1.0076769590377808, + "learning_rate": 0.00043007711645604024, + "loss": 3.6232, + "step": 67105 + }, + { + "epoch": 4.55972278842234, + "grad_norm": 0.8603944778442383, + "learning_rate": 0.00043003465144720747, + "loss": 3.34, + "step": 67110 + }, + { + "epoch": 4.560062508493002, + "grad_norm": 0.7744854092597961, + "learning_rate": 0.0004299921864383748, + "loss": 3.4745, + "step": 67115 + }, + { + "epoch": 4.560402228563664, + "grad_norm": 1.0107568502426147, + "learning_rate": 0.0004299497214295421, + "loss": 3.6018, + "step": 67120 + }, + { + "epoch": 4.560741948634325, + "grad_norm": 0.8886268138885498, + "learning_rate": 0.0004299072564207093, + "loss": 3.2792, + "step": 67125 + }, + { + "epoch": 4.561081668704987, + "grad_norm": 0.8083389401435852, + "learning_rate": 0.00042986479141187664, + "loss": 3.4611, + "step": 67130 + }, + { + "epoch": 4.561421388775649, + "grad_norm": 1.005603551864624, + "learning_rate": 0.0004298223264030439, + "loss": 3.2947, + "step": 67135 + }, + { + "epoch": 4.56176110884631, + "grad_norm": 0.7241636514663696, + "learning_rate": 0.00042977986139421115, + "loss": 3.5465, + "step": 67140 + }, + { + "epoch": 4.562100828916972, + "grad_norm": 0.905881404876709, + "learning_rate": 0.00042973739638537843, + "loss": 3.5161, + "step": 67145 + }, + { + "epoch": 4.5624405489876345, + "grad_norm": 1.0129166841506958, + "learning_rate": 0.00042969493137654576, + "loss": 3.6072, + "step": 67150 + }, + { + "epoch": 4.562780269058296, + "grad_norm": 0.838306188583374, + "learning_rate": 0.000429652466367713, + "loss": 3.5378, + "step": 67155 + }, + { + "epoch": 4.563119989128958, + "grad_norm": 1.0791294574737549, + "learning_rate": 0.00042961000135888027, + "loss": 3.5124, + "step": 67160 + }, + { + "epoch": 4.56345970919962, + "grad_norm": 0.9212916493415833, + "learning_rate": 0.0004295675363500476, + "loss": 3.5759, + "step": 67165 + }, + { + "epoch": 4.563799429270281, + "grad_norm": 0.8432183265686035, + "learning_rate": 0.00042952507134121483, + "loss": 3.6253, + "step": 67170 + }, + { + "epoch": 4.564139149340943, + "grad_norm": 0.8497002720832825, + "learning_rate": 0.0004294826063323821, + "loss": 3.5195, + "step": 67175 + }, + { + "epoch": 4.564478869411605, + "grad_norm": 1.0626378059387207, + "learning_rate": 0.00042944014132354944, + "loss": 3.5169, + "step": 67180 + }, + { + "epoch": 4.564818589482266, + "grad_norm": 1.237396001815796, + "learning_rate": 0.00042939767631471667, + "loss": 3.1062, + "step": 67185 + }, + { + "epoch": 4.5651583095529285, + "grad_norm": 1.0051499605178833, + "learning_rate": 0.00042935521130588395, + "loss": 3.5414, + "step": 67190 + }, + { + "epoch": 4.5654980296235905, + "grad_norm": 0.8467531800270081, + "learning_rate": 0.00042931274629705123, + "loss": 3.3157, + "step": 67195 + }, + { + "epoch": 4.565837749694252, + "grad_norm": 0.9624218940734863, + "learning_rate": 0.0004292702812882185, + "loss": 3.5213, + "step": 67200 + }, + { + "epoch": 4.566177469764914, + "grad_norm": 1.1289418935775757, + "learning_rate": 0.0004292278162793858, + "loss": 3.5188, + "step": 67205 + }, + { + "epoch": 4.566517189835576, + "grad_norm": 0.9390990734100342, + "learning_rate": 0.00042918535127055307, + "loss": 3.609, + "step": 67210 + }, + { + "epoch": 4.566856909906237, + "grad_norm": 0.7995396852493286, + "learning_rate": 0.0004291428862617204, + "loss": 3.1985, + "step": 67215 + }, + { + "epoch": 4.567196629976899, + "grad_norm": 0.7994973659515381, + "learning_rate": 0.00042910042125288763, + "loss": 3.2853, + "step": 67220 + }, + { + "epoch": 4.567536350047561, + "grad_norm": 0.9507366418838501, + "learning_rate": 0.0004290579562440549, + "loss": 3.305, + "step": 67225 + }, + { + "epoch": 4.567876070118222, + "grad_norm": 0.879896879196167, + "learning_rate": 0.0004290154912352222, + "loss": 3.3514, + "step": 67230 + }, + { + "epoch": 4.5682157901888845, + "grad_norm": 0.7326902747154236, + "learning_rate": 0.00042897302622638947, + "loss": 3.7267, + "step": 67235 + }, + { + "epoch": 4.5685555102595465, + "grad_norm": 0.928202211856842, + "learning_rate": 0.00042893056121755675, + "loss": 3.4733, + "step": 67240 + }, + { + "epoch": 4.568895230330208, + "grad_norm": 0.6807061433792114, + "learning_rate": 0.00042888809620872403, + "loss": 3.5264, + "step": 67245 + }, + { + "epoch": 4.56923495040087, + "grad_norm": 1.1278740167617798, + "learning_rate": 0.0004288456311998913, + "loss": 3.7306, + "step": 67250 + }, + { + "epoch": 4.569574670471532, + "grad_norm": 0.8895540237426758, + "learning_rate": 0.0004288031661910586, + "loss": 3.4121, + "step": 67255 + }, + { + "epoch": 4.569914390542193, + "grad_norm": 1.134082555770874, + "learning_rate": 0.00042876070118222587, + "loss": 3.4065, + "step": 67260 + }, + { + "epoch": 4.570254110612855, + "grad_norm": 1.0165737867355347, + "learning_rate": 0.0004287182361733931, + "loss": 3.452, + "step": 67265 + }, + { + "epoch": 4.570593830683517, + "grad_norm": 0.8008056879043579, + "learning_rate": 0.00042867577116456043, + "loss": 3.3975, + "step": 67270 + }, + { + "epoch": 4.570933550754178, + "grad_norm": 0.8788750171661377, + "learning_rate": 0.0004286333061557277, + "loss": 3.4382, + "step": 67275 + }, + { + "epoch": 4.5712732708248405, + "grad_norm": 0.7224317193031311, + "learning_rate": 0.00042859084114689494, + "loss": 3.1885, + "step": 67280 + }, + { + "epoch": 4.5716129908955025, + "grad_norm": 0.9009097814559937, + "learning_rate": 0.00042854837613806227, + "loss": 3.4035, + "step": 67285 + }, + { + "epoch": 4.571952710966164, + "grad_norm": 1.8645020723342896, + "learning_rate": 0.00042850591112922955, + "loss": 3.6646, + "step": 67290 + }, + { + "epoch": 4.572292431036826, + "grad_norm": 0.880300760269165, + "learning_rate": 0.0004284634461203968, + "loss": 3.2987, + "step": 67295 + }, + { + "epoch": 4.572632151107488, + "grad_norm": 1.12764573097229, + "learning_rate": 0.00042842098111156406, + "loss": 3.6162, + "step": 67300 + }, + { + "epoch": 4.572971871178149, + "grad_norm": 0.8858887553215027, + "learning_rate": 0.0004283785161027314, + "loss": 3.346, + "step": 67305 + }, + { + "epoch": 4.573311591248811, + "grad_norm": 0.7338926196098328, + "learning_rate": 0.0004283360510938986, + "loss": 3.7431, + "step": 67310 + }, + { + "epoch": 4.573651311319473, + "grad_norm": 0.9242263436317444, + "learning_rate": 0.0004282935860850659, + "loss": 3.5025, + "step": 67315 + }, + { + "epoch": 4.573991031390134, + "grad_norm": 0.8383481502532959, + "learning_rate": 0.00042825112107623323, + "loss": 3.3255, + "step": 67320 + }, + { + "epoch": 4.5743307514607965, + "grad_norm": 1.196877121925354, + "learning_rate": 0.00042820865606740046, + "loss": 3.4673, + "step": 67325 + }, + { + "epoch": 4.574670471531459, + "grad_norm": 0.7028875350952148, + "learning_rate": 0.00042816619105856774, + "loss": 3.5659, + "step": 67330 + }, + { + "epoch": 4.57501019160212, + "grad_norm": 0.8862816691398621, + "learning_rate": 0.000428123726049735, + "loss": 3.5513, + "step": 67335 + }, + { + "epoch": 4.575349911672782, + "grad_norm": 0.661256730556488, + "learning_rate": 0.0004280812610409023, + "loss": 3.0223, + "step": 67340 + }, + { + "epoch": 4.575689631743444, + "grad_norm": 0.8158573508262634, + "learning_rate": 0.0004280387960320696, + "loss": 3.5244, + "step": 67345 + }, + { + "epoch": 4.576029351814105, + "grad_norm": 0.9288393259048462, + "learning_rate": 0.00042799633102323686, + "loss": 3.5275, + "step": 67350 + }, + { + "epoch": 4.576369071884767, + "grad_norm": 2.9138591289520264, + "learning_rate": 0.00042795386601440414, + "loss": 3.5622, + "step": 67355 + }, + { + "epoch": 4.576708791955428, + "grad_norm": 1.1179980039596558, + "learning_rate": 0.0004279114010055714, + "loss": 3.5335, + "step": 67360 + }, + { + "epoch": 4.57704851202609, + "grad_norm": 0.9151772856712341, + "learning_rate": 0.0004278689359967387, + "loss": 3.3553, + "step": 67365 + }, + { + "epoch": 4.5773882320967525, + "grad_norm": 1.1000335216522217, + "learning_rate": 0.0004278264709879059, + "loss": 3.2875, + "step": 67370 + }, + { + "epoch": 4.577727952167414, + "grad_norm": 0.8894139528274536, + "learning_rate": 0.00042778400597907326, + "loss": 3.3601, + "step": 67375 + }, + { + "epoch": 4.578067672238076, + "grad_norm": 0.9011496901512146, + "learning_rate": 0.00042774154097024054, + "loss": 3.6231, + "step": 67380 + }, + { + "epoch": 4.578407392308738, + "grad_norm": 0.9043906331062317, + "learning_rate": 0.0004276990759614078, + "loss": 3.6621, + "step": 67385 + }, + { + "epoch": 4.578747112379399, + "grad_norm": 0.9889025688171387, + "learning_rate": 0.0004276566109525751, + "loss": 3.4285, + "step": 67390 + }, + { + "epoch": 4.579086832450061, + "grad_norm": 0.7129039764404297, + "learning_rate": 0.0004276141459437424, + "loss": 3.3303, + "step": 67395 + }, + { + "epoch": 4.579426552520723, + "grad_norm": 1.5924601554870605, + "learning_rate": 0.00042757168093490966, + "loss": 3.4262, + "step": 67400 + }, + { + "epoch": 4.579766272591384, + "grad_norm": 0.8330124020576477, + "learning_rate": 0.0004275292159260769, + "loss": 3.6432, + "step": 67405 + }, + { + "epoch": 4.580105992662046, + "grad_norm": 0.9130191802978516, + "learning_rate": 0.0004274867509172442, + "loss": 3.6996, + "step": 67410 + }, + { + "epoch": 4.5804457127327085, + "grad_norm": 0.885295569896698, + "learning_rate": 0.0004274442859084115, + "loss": 3.4926, + "step": 67415 + }, + { + "epoch": 4.58078543280337, + "grad_norm": 0.9824835658073425, + "learning_rate": 0.0004274018208995787, + "loss": 3.3504, + "step": 67420 + }, + { + "epoch": 4.581125152874032, + "grad_norm": 0.7631675601005554, + "learning_rate": 0.00042735935589074606, + "loss": 3.388, + "step": 67425 + }, + { + "epoch": 4.581464872944694, + "grad_norm": 0.8885135650634766, + "learning_rate": 0.00042731689088191334, + "loss": 3.3984, + "step": 67430 + }, + { + "epoch": 4.581804593015355, + "grad_norm": 1.0681157112121582, + "learning_rate": 0.00042727442587308056, + "loss": 3.6506, + "step": 67435 + }, + { + "epoch": 4.582144313086017, + "grad_norm": 1.1004154682159424, + "learning_rate": 0.00042723196086424784, + "loss": 3.2761, + "step": 67440 + }, + { + "epoch": 4.582484033156679, + "grad_norm": 0.8411985039710999, + "learning_rate": 0.0004271894958554152, + "loss": 3.4207, + "step": 67445 + }, + { + "epoch": 4.58282375322734, + "grad_norm": 1.0547019243240356, + "learning_rate": 0.0004271470308465824, + "loss": 3.3656, + "step": 67450 + }, + { + "epoch": 4.583163473298002, + "grad_norm": 0.7411400079727173, + "learning_rate": 0.0004271045658377497, + "loss": 3.3009, + "step": 67455 + }, + { + "epoch": 4.5835031933686645, + "grad_norm": 1.0234463214874268, + "learning_rate": 0.000427062100828917, + "loss": 3.6862, + "step": 67460 + }, + { + "epoch": 4.583842913439326, + "grad_norm": 0.8767651319503784, + "learning_rate": 0.00042701963582008424, + "loss": 3.3956, + "step": 67465 + }, + { + "epoch": 4.584182633509988, + "grad_norm": 0.827991783618927, + "learning_rate": 0.0004269771708112515, + "loss": 3.5, + "step": 67470 + }, + { + "epoch": 4.58452235358065, + "grad_norm": 1.0578416585922241, + "learning_rate": 0.00042693470580241886, + "loss": 3.541, + "step": 67475 + }, + { + "epoch": 4.584862073651311, + "grad_norm": 1.132582187652588, + "learning_rate": 0.0004268922407935861, + "loss": 3.2494, + "step": 67480 + }, + { + "epoch": 4.585201793721973, + "grad_norm": 0.7543942332267761, + "learning_rate": 0.00042684977578475336, + "loss": 3.242, + "step": 67485 + }, + { + "epoch": 4.585541513792635, + "grad_norm": 1.258089303970337, + "learning_rate": 0.00042680731077592064, + "loss": 3.6408, + "step": 67490 + }, + { + "epoch": 4.585881233863296, + "grad_norm": 1.0118377208709717, + "learning_rate": 0.0004267648457670879, + "loss": 3.4334, + "step": 67495 + }, + { + "epoch": 4.5862209539339585, + "grad_norm": 0.8488015532493591, + "learning_rate": 0.0004267223807582552, + "loss": 3.639, + "step": 67500 + }, + { + "epoch": 4.5865606740046205, + "grad_norm": 0.8111043572425842, + "learning_rate": 0.0004266799157494225, + "loss": 3.4394, + "step": 67505 + }, + { + "epoch": 4.586900394075282, + "grad_norm": 0.8887056112289429, + "learning_rate": 0.00042663745074058976, + "loss": 3.4625, + "step": 67510 + }, + { + "epoch": 4.587240114145944, + "grad_norm": 0.7388931512832642, + "learning_rate": 0.00042659498573175705, + "loss": 3.5483, + "step": 67515 + }, + { + "epoch": 4.587579834216606, + "grad_norm": 0.9170929789543152, + "learning_rate": 0.0004265525207229243, + "loss": 3.6021, + "step": 67520 + }, + { + "epoch": 4.587919554287267, + "grad_norm": 0.8375212550163269, + "learning_rate": 0.00042651005571409155, + "loss": 3.6647, + "step": 67525 + }, + { + "epoch": 4.588259274357929, + "grad_norm": 0.9961341023445129, + "learning_rate": 0.0004264675907052589, + "loss": 3.5866, + "step": 67530 + }, + { + "epoch": 4.588598994428591, + "grad_norm": 0.861946702003479, + "learning_rate": 0.00042642512569642617, + "loss": 3.407, + "step": 67535 + }, + { + "epoch": 4.588938714499252, + "grad_norm": 1.1403319835662842, + "learning_rate": 0.0004263826606875934, + "loss": 3.5012, + "step": 67540 + }, + { + "epoch": 4.5892784345699145, + "grad_norm": 0.8945421576499939, + "learning_rate": 0.0004263401956787607, + "loss": 3.6363, + "step": 67545 + }, + { + "epoch": 4.5896181546405765, + "grad_norm": 1.1353381872177124, + "learning_rate": 0.000426297730669928, + "loss": 3.3721, + "step": 67550 + }, + { + "epoch": 4.589957874711238, + "grad_norm": 0.9434921145439148, + "learning_rate": 0.0004262552656610953, + "loss": 3.4556, + "step": 67555 + }, + { + "epoch": 4.5902975947819, + "grad_norm": 0.8050035238265991, + "learning_rate": 0.0004262128006522625, + "loss": 3.6264, + "step": 67560 + }, + { + "epoch": 4.590637314852561, + "grad_norm": 0.9937885999679565, + "learning_rate": 0.00042617033564342985, + "loss": 3.1886, + "step": 67565 + }, + { + "epoch": 4.590977034923223, + "grad_norm": 0.8570193648338318, + "learning_rate": 0.0004261278706345971, + "loss": 3.372, + "step": 67570 + }, + { + "epoch": 4.591316754993885, + "grad_norm": 0.7374115586280823, + "learning_rate": 0.00042608540562576435, + "loss": 3.5548, + "step": 67575 + }, + { + "epoch": 4.591656475064546, + "grad_norm": 0.9214220643043518, + "learning_rate": 0.0004260429406169317, + "loss": 3.4317, + "step": 67580 + }, + { + "epoch": 4.591996195135208, + "grad_norm": 0.7519167065620422, + "learning_rate": 0.00042600047560809897, + "loss": 3.5474, + "step": 67585 + }, + { + "epoch": 4.5923359152058705, + "grad_norm": 0.7849289178848267, + "learning_rate": 0.0004259580105992662, + "loss": 3.4064, + "step": 67590 + }, + { + "epoch": 4.592675635276532, + "grad_norm": 0.879451334476471, + "learning_rate": 0.00042591554559043347, + "loss": 3.2379, + "step": 67595 + }, + { + "epoch": 4.593015355347194, + "grad_norm": 0.8637433052062988, + "learning_rate": 0.0004258730805816008, + "loss": 3.3319, + "step": 67600 + }, + { + "epoch": 4.593355075417856, + "grad_norm": 0.9727707505226135, + "learning_rate": 0.00042583061557276803, + "loss": 3.5231, + "step": 67605 + }, + { + "epoch": 4.593694795488517, + "grad_norm": 0.8368589878082275, + "learning_rate": 0.0004257881505639353, + "loss": 3.4176, + "step": 67610 + }, + { + "epoch": 4.594034515559179, + "grad_norm": 1.0144131183624268, + "learning_rate": 0.00042574568555510265, + "loss": 3.428, + "step": 67615 + }, + { + "epoch": 4.594374235629841, + "grad_norm": 0.9642901420593262, + "learning_rate": 0.00042570322054626987, + "loss": 3.4172, + "step": 67620 + }, + { + "epoch": 4.594713955700502, + "grad_norm": 1.0930784940719604, + "learning_rate": 0.00042566075553743715, + "loss": 3.349, + "step": 67625 + }, + { + "epoch": 4.595053675771164, + "grad_norm": 0.7711846828460693, + "learning_rate": 0.00042561829052860443, + "loss": 3.5884, + "step": 67630 + }, + { + "epoch": 4.5953933958418265, + "grad_norm": 0.7511374950408936, + "learning_rate": 0.0004255758255197717, + "loss": 3.33, + "step": 67635 + }, + { + "epoch": 4.595733115912488, + "grad_norm": 0.968352198600769, + "learning_rate": 0.000425533360510939, + "loss": 3.1873, + "step": 67640 + }, + { + "epoch": 4.59607283598315, + "grad_norm": 1.0428951978683472, + "learning_rate": 0.00042549089550210627, + "loss": 3.7763, + "step": 67645 + }, + { + "epoch": 4.596412556053812, + "grad_norm": 0.9782671928405762, + "learning_rate": 0.00042544843049327355, + "loss": 3.5743, + "step": 67650 + }, + { + "epoch": 4.596752276124473, + "grad_norm": 0.940528929233551, + "learning_rate": 0.00042540596548444083, + "loss": 3.7496, + "step": 67655 + }, + { + "epoch": 4.597091996195135, + "grad_norm": 1.0020475387573242, + "learning_rate": 0.0004253635004756081, + "loss": 3.2844, + "step": 67660 + }, + { + "epoch": 4.597431716265797, + "grad_norm": 0.8932842016220093, + "learning_rate": 0.00042532103546677534, + "loss": 3.4531, + "step": 67665 + }, + { + "epoch": 4.597771436336458, + "grad_norm": 0.7052213549613953, + "learning_rate": 0.0004252785704579427, + "loss": 3.6569, + "step": 67670 + }, + { + "epoch": 4.59811115640712, + "grad_norm": 0.9543061256408691, + "learning_rate": 0.00042523610544910995, + "loss": 3.4588, + "step": 67675 + }, + { + "epoch": 4.5984508764777825, + "grad_norm": 0.8197562098503113, + "learning_rate": 0.0004251936404402772, + "loss": 3.3838, + "step": 67680 + }, + { + "epoch": 4.598790596548444, + "grad_norm": 1.1790635585784912, + "learning_rate": 0.0004251511754314445, + "loss": 3.3035, + "step": 67685 + }, + { + "epoch": 4.599130316619106, + "grad_norm": 0.8613609671592712, + "learning_rate": 0.0004251087104226118, + "loss": 3.4634, + "step": 67690 + }, + { + "epoch": 4.599470036689768, + "grad_norm": 0.8218375444412231, + "learning_rate": 0.000425066245413779, + "loss": 3.2852, + "step": 67695 + }, + { + "epoch": 4.599809756760429, + "grad_norm": 0.8376889228820801, + "learning_rate": 0.0004250237804049463, + "loss": 3.6469, + "step": 67700 + }, + { + "epoch": 4.600149476831091, + "grad_norm": 0.9668734073638916, + "learning_rate": 0.00042498131539611363, + "loss": 3.6957, + "step": 67705 + }, + { + "epoch": 4.600489196901753, + "grad_norm": 1.1410876512527466, + "learning_rate": 0.00042493885038728086, + "loss": 3.2336, + "step": 67710 + }, + { + "epoch": 4.600828916972414, + "grad_norm": 0.9791898131370544, + "learning_rate": 0.00042489638537844814, + "loss": 3.1929, + "step": 67715 + }, + { + "epoch": 4.601168637043076, + "grad_norm": 0.9735149145126343, + "learning_rate": 0.0004248539203696155, + "loss": 3.4115, + "step": 67720 + }, + { + "epoch": 4.6015083571137385, + "grad_norm": 0.8024521470069885, + "learning_rate": 0.00042481145536078275, + "loss": 3.4118, + "step": 67725 + }, + { + "epoch": 4.6018480771844, + "grad_norm": 0.9405158758163452, + "learning_rate": 0.00042476899035195, + "loss": 3.4588, + "step": 67730 + }, + { + "epoch": 4.602187797255062, + "grad_norm": 0.823639452457428, + "learning_rate": 0.00042472652534311726, + "loss": 3.3947, + "step": 67735 + }, + { + "epoch": 4.602527517325724, + "grad_norm": 0.9724550843238831, + "learning_rate": 0.0004246840603342846, + "loss": 3.635, + "step": 67740 + }, + { + "epoch": 4.602867237396385, + "grad_norm": 0.7851670980453491, + "learning_rate": 0.0004246415953254518, + "loss": 3.4265, + "step": 67745 + }, + { + "epoch": 4.603206957467047, + "grad_norm": 0.9942404627799988, + "learning_rate": 0.0004245991303166191, + "loss": 3.5624, + "step": 67750 + }, + { + "epoch": 4.603546677537709, + "grad_norm": 1.0582170486450195, + "learning_rate": 0.00042455666530778643, + "loss": 3.3409, + "step": 67755 + }, + { + "epoch": 4.60388639760837, + "grad_norm": 1.032801866531372, + "learning_rate": 0.00042451420029895366, + "loss": 3.5892, + "step": 67760 + }, + { + "epoch": 4.6042261176790324, + "grad_norm": 0.9317339658737183, + "learning_rate": 0.00042447173529012094, + "loss": 3.4826, + "step": 67765 + }, + { + "epoch": 4.6045658377496945, + "grad_norm": 1.2864630222320557, + "learning_rate": 0.0004244292702812883, + "loss": 3.6659, + "step": 67770 + }, + { + "epoch": 4.604905557820356, + "grad_norm": 0.9617891311645508, + "learning_rate": 0.0004243868052724555, + "loss": 3.5835, + "step": 67775 + }, + { + "epoch": 4.605245277891018, + "grad_norm": 0.868038535118103, + "learning_rate": 0.0004243443402636228, + "loss": 3.6181, + "step": 67780 + }, + { + "epoch": 4.60558499796168, + "grad_norm": 0.9905685186386108, + "learning_rate": 0.00042430187525479006, + "loss": 3.6172, + "step": 67785 + }, + { + "epoch": 4.605924718032341, + "grad_norm": 0.9184408187866211, + "learning_rate": 0.00042425941024595734, + "loss": 3.5462, + "step": 67790 + }, + { + "epoch": 4.606264438103003, + "grad_norm": 0.5900066494941711, + "learning_rate": 0.0004242169452371246, + "loss": 3.4703, + "step": 67795 + }, + { + "epoch": 4.606604158173665, + "grad_norm": 0.916496217250824, + "learning_rate": 0.0004241744802282919, + "loss": 3.3533, + "step": 67800 + }, + { + "epoch": 4.606943878244326, + "grad_norm": 0.8620443344116211, + "learning_rate": 0.0004241320152194592, + "loss": 3.464, + "step": 67805 + }, + { + "epoch": 4.6072835983149885, + "grad_norm": 0.8090815544128418, + "learning_rate": 0.00042408955021062646, + "loss": 3.2885, + "step": 67810 + }, + { + "epoch": 4.6076233183856505, + "grad_norm": 0.7371649742126465, + "learning_rate": 0.00042404708520179374, + "loss": 3.4942, + "step": 67815 + }, + { + "epoch": 4.607963038456312, + "grad_norm": 0.7098739147186279, + "learning_rate": 0.00042400462019296097, + "loss": 3.0696, + "step": 67820 + }, + { + "epoch": 4.608302758526974, + "grad_norm": 0.909998893737793, + "learning_rate": 0.0004239621551841283, + "loss": 3.6255, + "step": 67825 + }, + { + "epoch": 4.608642478597636, + "grad_norm": 0.8828195929527283, + "learning_rate": 0.0004239196901752956, + "loss": 3.8395, + "step": 67830 + }, + { + "epoch": 4.608982198668297, + "grad_norm": 0.8771922588348389, + "learning_rate": 0.0004238772251664628, + "loss": 3.4647, + "step": 67835 + }, + { + "epoch": 4.609321918738959, + "grad_norm": 0.8357383012771606, + "learning_rate": 0.00042383476015763014, + "loss": 3.354, + "step": 67840 + }, + { + "epoch": 4.609661638809621, + "grad_norm": 1.0097839832305908, + "learning_rate": 0.0004237922951487974, + "loss": 3.4152, + "step": 67845 + }, + { + "epoch": 4.610001358880282, + "grad_norm": 0.9248858094215393, + "learning_rate": 0.00042374983013996465, + "loss": 3.5807, + "step": 67850 + }, + { + "epoch": 4.6103410789509445, + "grad_norm": 0.9135956764221191, + "learning_rate": 0.00042370736513113193, + "loss": 3.4877, + "step": 67855 + }, + { + "epoch": 4.6106807990216065, + "grad_norm": 0.9012488722801208, + "learning_rate": 0.00042366490012229926, + "loss": 3.4189, + "step": 67860 + }, + { + "epoch": 4.611020519092268, + "grad_norm": 0.9534264206886292, + "learning_rate": 0.0004236224351134665, + "loss": 3.3999, + "step": 67865 + }, + { + "epoch": 4.61136023916293, + "grad_norm": 0.9386141300201416, + "learning_rate": 0.00042357997010463377, + "loss": 3.1926, + "step": 67870 + }, + { + "epoch": 4.611699959233592, + "grad_norm": 0.9828197956085205, + "learning_rate": 0.0004235375050958011, + "loss": 3.4444, + "step": 67875 + }, + { + "epoch": 4.612039679304253, + "grad_norm": 0.7708816528320312, + "learning_rate": 0.00042349504008696833, + "loss": 3.6131, + "step": 67880 + }, + { + "epoch": 4.612379399374915, + "grad_norm": 0.679082989692688, + "learning_rate": 0.0004234525750781356, + "loss": 3.4385, + "step": 67885 + }, + { + "epoch": 4.612719119445577, + "grad_norm": 0.9784545302391052, + "learning_rate": 0.0004234101100693029, + "loss": 3.6415, + "step": 67890 + }, + { + "epoch": 4.613058839516238, + "grad_norm": 1.0610295534133911, + "learning_rate": 0.0004233676450604702, + "loss": 3.2338, + "step": 67895 + }, + { + "epoch": 4.6133985595869005, + "grad_norm": 0.9436649084091187, + "learning_rate": 0.00042332518005163745, + "loss": 3.6425, + "step": 67900 + }, + { + "epoch": 4.613738279657563, + "grad_norm": 0.6794824600219727, + "learning_rate": 0.00042328271504280473, + "loss": 3.5371, + "step": 67905 + }, + { + "epoch": 4.614077999728224, + "grad_norm": 1.107001781463623, + "learning_rate": 0.00042324025003397206, + "loss": 3.2345, + "step": 67910 + }, + { + "epoch": 4.614417719798886, + "grad_norm": 0.9971826672554016, + "learning_rate": 0.0004231977850251393, + "loss": 3.4918, + "step": 67915 + }, + { + "epoch": 4.614757439869548, + "grad_norm": 0.8758735656738281, + "learning_rate": 0.00042315532001630657, + "loss": 3.4424, + "step": 67920 + }, + { + "epoch": 4.615097159940209, + "grad_norm": 0.8255982995033264, + "learning_rate": 0.00042311285500747385, + "loss": 3.7576, + "step": 67925 + }, + { + "epoch": 4.615436880010871, + "grad_norm": 0.7605299353599548, + "learning_rate": 0.00042307038999864113, + "loss": 3.3631, + "step": 67930 + }, + { + "epoch": 4.615776600081533, + "grad_norm": 0.8301802277565002, + "learning_rate": 0.0004230279249898084, + "loss": 3.4068, + "step": 67935 + }, + { + "epoch": 4.616116320152194, + "grad_norm": 0.7834235429763794, + "learning_rate": 0.0004229854599809757, + "loss": 3.4111, + "step": 67940 + }, + { + "epoch": 4.6164560402228565, + "grad_norm": 0.893020749092102, + "learning_rate": 0.00042294299497214297, + "loss": 3.6409, + "step": 67945 + }, + { + "epoch": 4.616795760293519, + "grad_norm": 0.8067091703414917, + "learning_rate": 0.00042290052996331025, + "loss": 3.5938, + "step": 67950 + }, + { + "epoch": 4.61713548036418, + "grad_norm": 0.8507208824157715, + "learning_rate": 0.00042285806495447753, + "loss": 3.6714, + "step": 67955 + }, + { + "epoch": 4.617475200434842, + "grad_norm": 1.389599323272705, + "learning_rate": 0.00042281559994564476, + "loss": 3.3246, + "step": 67960 + }, + { + "epoch": 4.617814920505504, + "grad_norm": 0.8027352690696716, + "learning_rate": 0.0004227731349368121, + "loss": 3.4091, + "step": 67965 + }, + { + "epoch": 4.618154640576165, + "grad_norm": 1.2079293727874756, + "learning_rate": 0.00042273066992797937, + "loss": 3.4204, + "step": 67970 + }, + { + "epoch": 4.618494360646827, + "grad_norm": 1.0536094903945923, + "learning_rate": 0.0004226882049191466, + "loss": 3.5493, + "step": 67975 + }, + { + "epoch": 4.618834080717489, + "grad_norm": 0.8972793817520142, + "learning_rate": 0.00042264573991031393, + "loss": 3.284, + "step": 67980 + }, + { + "epoch": 4.61917380078815, + "grad_norm": 0.879063069820404, + "learning_rate": 0.0004226032749014812, + "loss": 3.4105, + "step": 67985 + }, + { + "epoch": 4.6195135208588125, + "grad_norm": 0.8161985278129578, + "learning_rate": 0.00042256080989264844, + "loss": 3.5361, + "step": 67990 + }, + { + "epoch": 4.619853240929475, + "grad_norm": 0.9861201643943787, + "learning_rate": 0.0004225183448838157, + "loss": 3.5481, + "step": 67995 + }, + { + "epoch": 4.620192961000136, + "grad_norm": 0.964358925819397, + "learning_rate": 0.00042247587987498305, + "loss": 3.2772, + "step": 68000 + }, + { + "epoch": 4.620532681070798, + "grad_norm": 0.9929901361465454, + "learning_rate": 0.0004224334148661503, + "loss": 3.6082, + "step": 68005 + }, + { + "epoch": 4.62087240114146, + "grad_norm": 0.805381715297699, + "learning_rate": 0.00042239094985731756, + "loss": 3.4793, + "step": 68010 + }, + { + "epoch": 4.621212121212121, + "grad_norm": 0.7458207011222839, + "learning_rate": 0.0004223484848484849, + "loss": 3.3525, + "step": 68015 + }, + { + "epoch": 4.621551841282783, + "grad_norm": 0.8662701845169067, + "learning_rate": 0.0004223060198396521, + "loss": 3.3581, + "step": 68020 + }, + { + "epoch": 4.621891561353445, + "grad_norm": 0.8811438083648682, + "learning_rate": 0.0004222635548308194, + "loss": 3.3201, + "step": 68025 + }, + { + "epoch": 4.622231281424106, + "grad_norm": 3.1602609157562256, + "learning_rate": 0.00042222108982198673, + "loss": 3.4015, + "step": 68030 + }, + { + "epoch": 4.6225710014947685, + "grad_norm": 0.9522108435630798, + "learning_rate": 0.00042217862481315396, + "loss": 3.4041, + "step": 68035 + }, + { + "epoch": 4.62291072156543, + "grad_norm": 1.0106728076934814, + "learning_rate": 0.00042213615980432124, + "loss": 3.5148, + "step": 68040 + }, + { + "epoch": 4.623250441636092, + "grad_norm": 0.8571694493293762, + "learning_rate": 0.0004220936947954885, + "loss": 3.2579, + "step": 68045 + }, + { + "epoch": 4.623590161706754, + "grad_norm": 0.9172520637512207, + "learning_rate": 0.0004220512297866558, + "loss": 3.617, + "step": 68050 + }, + { + "epoch": 4.623929881777415, + "grad_norm": 0.7596650123596191, + "learning_rate": 0.0004220087647778231, + "loss": 3.2947, + "step": 68055 + }, + { + "epoch": 4.624269601848077, + "grad_norm": 1.0978385210037231, + "learning_rate": 0.00042196629976899036, + "loss": 3.6102, + "step": 68060 + }, + { + "epoch": 4.624609321918739, + "grad_norm": 0.9235647916793823, + "learning_rate": 0.0004219238347601577, + "loss": 3.7215, + "step": 68065 + }, + { + "epoch": 4.6249490419894, + "grad_norm": 1.0100572109222412, + "learning_rate": 0.0004218813697513249, + "loss": 3.1625, + "step": 68070 + }, + { + "epoch": 4.6252887620600625, + "grad_norm": 0.7820902466773987, + "learning_rate": 0.0004218389047424922, + "loss": 3.5212, + "step": 68075 + }, + { + "epoch": 4.6256284821307245, + "grad_norm": 1.044643521308899, + "learning_rate": 0.0004217964397336595, + "loss": 3.4479, + "step": 68080 + }, + { + "epoch": 4.625968202201386, + "grad_norm": 0.8095884919166565, + "learning_rate": 0.00042175397472482676, + "loss": 3.4578, + "step": 68085 + }, + { + "epoch": 4.626307922272048, + "grad_norm": 0.9433499574661255, + "learning_rate": 0.00042171150971599404, + "loss": 3.3904, + "step": 68090 + }, + { + "epoch": 4.62664764234271, + "grad_norm": 0.721687912940979, + "learning_rate": 0.0004216690447071613, + "loss": 3.5964, + "step": 68095 + }, + { + "epoch": 4.626987362413371, + "grad_norm": 1.1639763116836548, + "learning_rate": 0.0004216265796983286, + "loss": 3.6244, + "step": 68100 + }, + { + "epoch": 4.627327082484033, + "grad_norm": 0.9073703289031982, + "learning_rate": 0.0004215841146894959, + "loss": 3.3582, + "step": 68105 + }, + { + "epoch": 4.627666802554695, + "grad_norm": 0.7800865173339844, + "learning_rate": 0.00042154164968066316, + "loss": 3.3294, + "step": 68110 + }, + { + "epoch": 4.628006522625356, + "grad_norm": 1.0434372425079346, + "learning_rate": 0.0004214991846718304, + "loss": 3.6298, + "step": 68115 + }, + { + "epoch": 4.6283462426960185, + "grad_norm": 0.9065384268760681, + "learning_rate": 0.0004214567196629977, + "loss": 3.474, + "step": 68120 + }, + { + "epoch": 4.6286859627666805, + "grad_norm": 1.1309349536895752, + "learning_rate": 0.000421414254654165, + "loss": 3.6029, + "step": 68125 + }, + { + "epoch": 4.629025682837342, + "grad_norm": 0.9185808300971985, + "learning_rate": 0.0004213717896453322, + "loss": 3.6367, + "step": 68130 + }, + { + "epoch": 4.629365402908004, + "grad_norm": 0.7079995274543762, + "learning_rate": 0.00042132932463649956, + "loss": 3.4511, + "step": 68135 + }, + { + "epoch": 4.629705122978666, + "grad_norm": 0.8110623955726624, + "learning_rate": 0.00042128685962766684, + "loss": 3.4244, + "step": 68140 + }, + { + "epoch": 4.630044843049327, + "grad_norm": 0.7839182615280151, + "learning_rate": 0.00042124439461883406, + "loss": 3.4715, + "step": 68145 + }, + { + "epoch": 4.630384563119989, + "grad_norm": 1.0865899324417114, + "learning_rate": 0.00042120192961000134, + "loss": 3.4708, + "step": 68150 + }, + { + "epoch": 4.630724283190651, + "grad_norm": 0.999410092830658, + "learning_rate": 0.0004211594646011687, + "loss": 3.3514, + "step": 68155 + }, + { + "epoch": 4.631064003261312, + "grad_norm": 0.8788954019546509, + "learning_rate": 0.0004211169995923359, + "loss": 3.5602, + "step": 68160 + }, + { + "epoch": 4.6314037233319745, + "grad_norm": 0.9947049617767334, + "learning_rate": 0.0004210745345835032, + "loss": 3.4767, + "step": 68165 + }, + { + "epoch": 4.6317434434026366, + "grad_norm": 0.8769662380218506, + "learning_rate": 0.0004210320695746705, + "loss": 3.5446, + "step": 68170 + }, + { + "epoch": 4.632083163473298, + "grad_norm": 0.9333018660545349, + "learning_rate": 0.00042098960456583774, + "loss": 3.4571, + "step": 68175 + }, + { + "epoch": 4.63242288354396, + "grad_norm": 0.7598336338996887, + "learning_rate": 0.000420947139557005, + "loss": 3.4049, + "step": 68180 + }, + { + "epoch": 4.632762603614622, + "grad_norm": 1.095141887664795, + "learning_rate": 0.0004209046745481723, + "loss": 3.4888, + "step": 68185 + }, + { + "epoch": 4.633102323685283, + "grad_norm": 0.8638190627098083, + "learning_rate": 0.0004208622095393396, + "loss": 3.4684, + "step": 68190 + }, + { + "epoch": 4.633442043755945, + "grad_norm": 0.7676572799682617, + "learning_rate": 0.00042081974453050686, + "loss": 3.7041, + "step": 68195 + }, + { + "epoch": 4.633781763826607, + "grad_norm": 0.8033756613731384, + "learning_rate": 0.00042077727952167414, + "loss": 3.386, + "step": 68200 + }, + { + "epoch": 4.634121483897268, + "grad_norm": 0.8592642545700073, + "learning_rate": 0.0004207348145128414, + "loss": 3.4979, + "step": 68205 + }, + { + "epoch": 4.6344612039679305, + "grad_norm": 0.9765332341194153, + "learning_rate": 0.0004206923495040087, + "loss": 3.195, + "step": 68210 + }, + { + "epoch": 4.634800924038593, + "grad_norm": 0.8654045462608337, + "learning_rate": 0.000420649884495176, + "loss": 3.8721, + "step": 68215 + }, + { + "epoch": 4.635140644109254, + "grad_norm": 0.8636555671691895, + "learning_rate": 0.0004206074194863432, + "loss": 3.4109, + "step": 68220 + }, + { + "epoch": 4.635480364179916, + "grad_norm": 0.786760687828064, + "learning_rate": 0.00042056495447751054, + "loss": 3.6563, + "step": 68225 + }, + { + "epoch": 4.635820084250578, + "grad_norm": 0.853145956993103, + "learning_rate": 0.0004205224894686778, + "loss": 3.3037, + "step": 68230 + }, + { + "epoch": 4.636159804321239, + "grad_norm": 1.0670366287231445, + "learning_rate": 0.0004204800244598451, + "loss": 3.4527, + "step": 68235 + }, + { + "epoch": 4.636499524391901, + "grad_norm": 0.8612810373306274, + "learning_rate": 0.0004204375594510124, + "loss": 3.5309, + "step": 68240 + }, + { + "epoch": 4.636839244462563, + "grad_norm": 0.8726534247398376, + "learning_rate": 0.00042039509444217967, + "loss": 3.5563, + "step": 68245 + }, + { + "epoch": 4.637178964533224, + "grad_norm": 1.0701817274093628, + "learning_rate": 0.00042035262943334695, + "loss": 3.4879, + "step": 68250 + }, + { + "epoch": 4.6375186846038865, + "grad_norm": 0.7773467898368835, + "learning_rate": 0.00042031016442451417, + "loss": 3.4171, + "step": 68255 + }, + { + "epoch": 4.637858404674548, + "grad_norm": 0.8920998573303223, + "learning_rate": 0.0004202676994156815, + "loss": 3.5887, + "step": 68260 + }, + { + "epoch": 4.63819812474521, + "grad_norm": 0.7680952548980713, + "learning_rate": 0.0004202252344068488, + "loss": 3.3715, + "step": 68265 + }, + { + "epoch": 4.638537844815872, + "grad_norm": 0.8944039344787598, + "learning_rate": 0.000420182769398016, + "loss": 3.2417, + "step": 68270 + }, + { + "epoch": 4.638877564886533, + "grad_norm": 0.7180835008621216, + "learning_rate": 0.00042014030438918335, + "loss": 3.611, + "step": 68275 + }, + { + "epoch": 4.639217284957195, + "grad_norm": 0.9352626204490662, + "learning_rate": 0.0004200978393803506, + "loss": 3.7235, + "step": 68280 + }, + { + "epoch": 4.639557005027857, + "grad_norm": 1.074934720993042, + "learning_rate": 0.00042005537437151785, + "loss": 3.2216, + "step": 68285 + }, + { + "epoch": 4.639896725098518, + "grad_norm": 0.7853044867515564, + "learning_rate": 0.00042001290936268513, + "loss": 3.3503, + "step": 68290 + }, + { + "epoch": 4.64023644516918, + "grad_norm": 1.0257841348648071, + "learning_rate": 0.00041997044435385247, + "loss": 3.6241, + "step": 68295 + }, + { + "epoch": 4.6405761652398425, + "grad_norm": 0.9955359697341919, + "learning_rate": 0.0004199279793450197, + "loss": 3.2571, + "step": 68300 + }, + { + "epoch": 4.640915885310504, + "grad_norm": 0.8383559584617615, + "learning_rate": 0.00041988551433618697, + "loss": 3.1925, + "step": 68305 + }, + { + "epoch": 4.641255605381166, + "grad_norm": 0.9184194803237915, + "learning_rate": 0.0004198430493273543, + "loss": 3.3999, + "step": 68310 + }, + { + "epoch": 4.641595325451828, + "grad_norm": 1.046006202697754, + "learning_rate": 0.00041980058431852153, + "loss": 3.5223, + "step": 68315 + }, + { + "epoch": 4.641935045522489, + "grad_norm": 0.8355832099914551, + "learning_rate": 0.0004197581193096888, + "loss": 3.4952, + "step": 68320 + }, + { + "epoch": 4.642274765593151, + "grad_norm": 0.8756354451179504, + "learning_rate": 0.00041971565430085615, + "loss": 3.4124, + "step": 68325 + }, + { + "epoch": 4.642614485663813, + "grad_norm": 1.0545001029968262, + "learning_rate": 0.00041967318929202337, + "loss": 3.5571, + "step": 68330 + }, + { + "epoch": 4.642954205734474, + "grad_norm": 0.7567934989929199, + "learning_rate": 0.00041963072428319065, + "loss": 3.4895, + "step": 68335 + }, + { + "epoch": 4.6432939258051364, + "grad_norm": 0.9333906173706055, + "learning_rate": 0.00041958825927435793, + "loss": 3.5048, + "step": 68340 + }, + { + "epoch": 4.6436336458757985, + "grad_norm": 0.955683171749115, + "learning_rate": 0.0004195457942655252, + "loss": 3.1004, + "step": 68345 + }, + { + "epoch": 4.64397336594646, + "grad_norm": 0.9125000834465027, + "learning_rate": 0.0004195033292566925, + "loss": 3.4811, + "step": 68350 + }, + { + "epoch": 4.644313086017122, + "grad_norm": 0.8338185548782349, + "learning_rate": 0.00041946086424785977, + "loss": 3.5393, + "step": 68355 + }, + { + "epoch": 4.644652806087784, + "grad_norm": 0.8546797633171082, + "learning_rate": 0.00041941839923902705, + "loss": 3.426, + "step": 68360 + }, + { + "epoch": 4.644992526158445, + "grad_norm": 1.1002360582351685, + "learning_rate": 0.00041937593423019433, + "loss": 3.2052, + "step": 68365 + }, + { + "epoch": 4.645332246229107, + "grad_norm": 0.9460350871086121, + "learning_rate": 0.0004193334692213616, + "loss": 3.6432, + "step": 68370 + }, + { + "epoch": 4.645671966299769, + "grad_norm": 0.9006628394126892, + "learning_rate": 0.00041929100421252884, + "loss": 3.6207, + "step": 68375 + }, + { + "epoch": 4.64601168637043, + "grad_norm": 0.8346257209777832, + "learning_rate": 0.0004192485392036962, + "loss": 3.3907, + "step": 68380 + }, + { + "epoch": 4.6463514064410925, + "grad_norm": 0.7549441456794739, + "learning_rate": 0.00041920607419486345, + "loss": 3.4689, + "step": 68385 + }, + { + "epoch": 4.6466911265117545, + "grad_norm": 0.8115181922912598, + "learning_rate": 0.0004191636091860307, + "loss": 3.3721, + "step": 68390 + }, + { + "epoch": 4.647030846582416, + "grad_norm": 1.1192134618759155, + "learning_rate": 0.000419121144177198, + "loss": 3.5807, + "step": 68395 + }, + { + "epoch": 4.647370566653078, + "grad_norm": 1.872918725013733, + "learning_rate": 0.0004190786791683653, + "loss": 3.6555, + "step": 68400 + }, + { + "epoch": 4.64771028672374, + "grad_norm": 1.0189216136932373, + "learning_rate": 0.0004190362141595326, + "loss": 3.5057, + "step": 68405 + }, + { + "epoch": 4.648050006794401, + "grad_norm": 0.9271153807640076, + "learning_rate": 0.0004189937491506998, + "loss": 3.4399, + "step": 68410 + }, + { + "epoch": 4.648389726865063, + "grad_norm": 1.0508484840393066, + "learning_rate": 0.00041895128414186713, + "loss": 3.4049, + "step": 68415 + }, + { + "epoch": 4.648729446935725, + "grad_norm": 0.8426085710525513, + "learning_rate": 0.0004189088191330344, + "loss": 3.3728, + "step": 68420 + }, + { + "epoch": 4.649069167006386, + "grad_norm": 0.8193681836128235, + "learning_rate": 0.00041886635412420164, + "loss": 3.3729, + "step": 68425 + }, + { + "epoch": 4.6494088870770485, + "grad_norm": 0.9380782842636108, + "learning_rate": 0.000418823889115369, + "loss": 3.4082, + "step": 68430 + }, + { + "epoch": 4.6497486071477105, + "grad_norm": 0.9658963084220886, + "learning_rate": 0.00041878142410653625, + "loss": 3.5177, + "step": 68435 + }, + { + "epoch": 4.650088327218372, + "grad_norm": 0.7768871784210205, + "learning_rate": 0.0004187389590977035, + "loss": 3.8404, + "step": 68440 + }, + { + "epoch": 4.650428047289034, + "grad_norm": 0.8578163385391235, + "learning_rate": 0.00041869649408887076, + "loss": 3.5243, + "step": 68445 + }, + { + "epoch": 4.650767767359696, + "grad_norm": 1.1034132242202759, + "learning_rate": 0.0004186540290800381, + "loss": 3.5499, + "step": 68450 + }, + { + "epoch": 4.651107487430357, + "grad_norm": 0.7648698687553406, + "learning_rate": 0.0004186115640712053, + "loss": 3.4263, + "step": 68455 + }, + { + "epoch": 4.651447207501019, + "grad_norm": 0.7452331185340881, + "learning_rate": 0.0004185690990623726, + "loss": 3.8387, + "step": 68460 + }, + { + "epoch": 4.651786927571681, + "grad_norm": 0.7394859790802002, + "learning_rate": 0.00041852663405353993, + "loss": 3.4112, + "step": 68465 + }, + { + "epoch": 4.652126647642342, + "grad_norm": 0.8813775777816772, + "learning_rate": 0.00041848416904470716, + "loss": 3.5976, + "step": 68470 + }, + { + "epoch": 4.6524663677130045, + "grad_norm": 1.0817413330078125, + "learning_rate": 0.00041844170403587444, + "loss": 3.323, + "step": 68475 + }, + { + "epoch": 4.6528060877836666, + "grad_norm": 0.7993396520614624, + "learning_rate": 0.0004183992390270417, + "loss": 3.1931, + "step": 68480 + }, + { + "epoch": 4.653145807854328, + "grad_norm": 1.0633779764175415, + "learning_rate": 0.000418356774018209, + "loss": 3.5304, + "step": 68485 + }, + { + "epoch": 4.65348552792499, + "grad_norm": 0.8796688318252563, + "learning_rate": 0.0004183143090093763, + "loss": 3.5452, + "step": 68490 + }, + { + "epoch": 4.653825247995652, + "grad_norm": 0.9149452447891235, + "learning_rate": 0.00041827184400054356, + "loss": 3.3748, + "step": 68495 + }, + { + "epoch": 4.654164968066313, + "grad_norm": 0.8446894884109497, + "learning_rate": 0.00041822937899171084, + "loss": 3.5857, + "step": 68500 + }, + { + "epoch": 4.654504688136975, + "grad_norm": 0.8676488995552063, + "learning_rate": 0.0004181869139828781, + "loss": 3.543, + "step": 68505 + }, + { + "epoch": 4.654844408207637, + "grad_norm": 0.8865768313407898, + "learning_rate": 0.0004181444489740454, + "loss": 3.5333, + "step": 68510 + }, + { + "epoch": 4.655184128278298, + "grad_norm": 0.7793726921081543, + "learning_rate": 0.0004181019839652126, + "loss": 3.3695, + "step": 68515 + }, + { + "epoch": 4.6555238483489605, + "grad_norm": 0.8741711378097534, + "learning_rate": 0.00041805951895637996, + "loss": 3.4442, + "step": 68520 + }, + { + "epoch": 4.655863568419623, + "grad_norm": 0.9149139523506165, + "learning_rate": 0.00041801705394754724, + "loss": 3.378, + "step": 68525 + }, + { + "epoch": 4.656203288490284, + "grad_norm": 1.035473346710205, + "learning_rate": 0.00041797458893871447, + "loss": 3.4939, + "step": 68530 + }, + { + "epoch": 4.656543008560946, + "grad_norm": 0.9867655634880066, + "learning_rate": 0.0004179321239298818, + "loss": 3.5702, + "step": 68535 + }, + { + "epoch": 4.656882728631608, + "grad_norm": 0.9957058429718018, + "learning_rate": 0.0004178896589210491, + "loss": 3.5761, + "step": 68540 + }, + { + "epoch": 4.657222448702269, + "grad_norm": 0.9165830612182617, + "learning_rate": 0.0004178471939122163, + "loss": 3.54, + "step": 68545 + }, + { + "epoch": 4.657562168772931, + "grad_norm": 0.8947004675865173, + "learning_rate": 0.0004178047289033836, + "loss": 3.642, + "step": 68550 + }, + { + "epoch": 4.657901888843593, + "grad_norm": 0.8200177550315857, + "learning_rate": 0.0004177622638945509, + "loss": 3.2511, + "step": 68555 + }, + { + "epoch": 4.658241608914254, + "grad_norm": 0.9102053642272949, + "learning_rate": 0.00041771979888571815, + "loss": 3.523, + "step": 68560 + }, + { + "epoch": 4.6585813289849165, + "grad_norm": 0.8251795172691345, + "learning_rate": 0.00041767733387688543, + "loss": 3.6133, + "step": 68565 + }, + { + "epoch": 4.658921049055579, + "grad_norm": 1.107084035873413, + "learning_rate": 0.00041763486886805276, + "loss": 3.7238, + "step": 68570 + }, + { + "epoch": 4.65926076912624, + "grad_norm": 0.8720716238021851, + "learning_rate": 0.00041759240385922004, + "loss": 3.3384, + "step": 68575 + }, + { + "epoch": 4.659600489196902, + "grad_norm": 0.8229120969772339, + "learning_rate": 0.00041754993885038727, + "loss": 3.6829, + "step": 68580 + }, + { + "epoch": 4.659940209267564, + "grad_norm": 0.9354588985443115, + "learning_rate": 0.00041750747384155455, + "loss": 3.6664, + "step": 68585 + }, + { + "epoch": 4.660279929338225, + "grad_norm": 0.9079734086990356, + "learning_rate": 0.0004174650088327219, + "loss": 3.4931, + "step": 68590 + }, + { + "epoch": 4.660619649408887, + "grad_norm": 0.8546583652496338, + "learning_rate": 0.0004174225438238891, + "loss": 3.4534, + "step": 68595 + }, + { + "epoch": 4.660959369479549, + "grad_norm": 0.7542752027511597, + "learning_rate": 0.0004173800788150564, + "loss": 3.0967, + "step": 68600 + }, + { + "epoch": 4.66129908955021, + "grad_norm": 0.7804221510887146, + "learning_rate": 0.0004173376138062237, + "loss": 3.4448, + "step": 68605 + }, + { + "epoch": 4.6616388096208725, + "grad_norm": 0.9289933443069458, + "learning_rate": 0.00041729514879739095, + "loss": 3.5533, + "step": 68610 + }, + { + "epoch": 4.661978529691535, + "grad_norm": 0.9845079779624939, + "learning_rate": 0.00041725268378855823, + "loss": 3.4043, + "step": 68615 + }, + { + "epoch": 4.662318249762196, + "grad_norm": 1.1165680885314941, + "learning_rate": 0.00041721021877972556, + "loss": 3.3884, + "step": 68620 + }, + { + "epoch": 4.662657969832858, + "grad_norm": 0.7916510105133057, + "learning_rate": 0.0004171677537708928, + "loss": 3.4287, + "step": 68625 + }, + { + "epoch": 4.66299768990352, + "grad_norm": 0.8921187520027161, + "learning_rate": 0.00041712528876206007, + "loss": 3.659, + "step": 68630 + }, + { + "epoch": 4.663337409974181, + "grad_norm": 1.0192402601242065, + "learning_rate": 0.00041708282375322735, + "loss": 3.4915, + "step": 68635 + }, + { + "epoch": 4.663677130044843, + "grad_norm": 1.1442757844924927, + "learning_rate": 0.00041704035874439463, + "loss": 3.3551, + "step": 68640 + }, + { + "epoch": 4.664016850115505, + "grad_norm": 0.7416129112243652, + "learning_rate": 0.0004169978937355619, + "loss": 3.474, + "step": 68645 + }, + { + "epoch": 4.6643565701861665, + "grad_norm": 1.1482139825820923, + "learning_rate": 0.0004169554287267292, + "loss": 3.3856, + "step": 68650 + }, + { + "epoch": 4.6646962902568285, + "grad_norm": 0.7678807377815247, + "learning_rate": 0.00041691296371789647, + "loss": 3.3709, + "step": 68655 + }, + { + "epoch": 4.665036010327491, + "grad_norm": 0.7061179280281067, + "learning_rate": 0.00041687049870906375, + "loss": 3.6115, + "step": 68660 + }, + { + "epoch": 4.665375730398152, + "grad_norm": 0.8391000032424927, + "learning_rate": 0.00041682803370023103, + "loss": 3.537, + "step": 68665 + }, + { + "epoch": 4.665715450468814, + "grad_norm": 0.8942126631736755, + "learning_rate": 0.00041678556869139825, + "loss": 3.5336, + "step": 68670 + }, + { + "epoch": 4.666055170539476, + "grad_norm": 0.8027516603469849, + "learning_rate": 0.0004167431036825656, + "loss": 3.6208, + "step": 68675 + }, + { + "epoch": 4.666394890610137, + "grad_norm": 0.838283360004425, + "learning_rate": 0.00041670063867373287, + "loss": 3.2632, + "step": 68680 + }, + { + "epoch": 4.666734610680799, + "grad_norm": 1.0803709030151367, + "learning_rate": 0.0004166581736649001, + "loss": 3.6309, + "step": 68685 + }, + { + "epoch": 4.667074330751461, + "grad_norm": 1.142464518547058, + "learning_rate": 0.00041661570865606743, + "loss": 3.5949, + "step": 68690 + }, + { + "epoch": 4.6674140508221225, + "grad_norm": 0.8500204086303711, + "learning_rate": 0.0004165732436472347, + "loss": 3.3203, + "step": 68695 + }, + { + "epoch": 4.6677537708927845, + "grad_norm": 0.9047169089317322, + "learning_rate": 0.00041653077863840194, + "loss": 3.5843, + "step": 68700 + }, + { + "epoch": 4.668093490963447, + "grad_norm": 0.8354958891868591, + "learning_rate": 0.0004164883136295692, + "loss": 3.4368, + "step": 68705 + }, + { + "epoch": 4.668433211034108, + "grad_norm": 1.0198941230773926, + "learning_rate": 0.00041644584862073655, + "loss": 3.3163, + "step": 68710 + }, + { + "epoch": 4.66877293110477, + "grad_norm": 0.9637662768363953, + "learning_rate": 0.0004164033836119038, + "loss": 3.4993, + "step": 68715 + }, + { + "epoch": 4.669112651175431, + "grad_norm": 1.1615840196609497, + "learning_rate": 0.00041636091860307106, + "loss": 3.5355, + "step": 68720 + }, + { + "epoch": 4.669452371246093, + "grad_norm": 0.97249436378479, + "learning_rate": 0.0004163184535942384, + "loss": 3.6188, + "step": 68725 + }, + { + "epoch": 4.669792091316755, + "grad_norm": 1.0933294296264648, + "learning_rate": 0.0004162759885854056, + "loss": 3.5181, + "step": 68730 + }, + { + "epoch": 4.670131811387416, + "grad_norm": 1.008276343345642, + "learning_rate": 0.0004162335235765729, + "loss": 3.4522, + "step": 68735 + }, + { + "epoch": 4.6704715314580785, + "grad_norm": 1.266398549079895, + "learning_rate": 0.0004161910585677402, + "loss": 3.4191, + "step": 68740 + }, + { + "epoch": 4.6708112515287405, + "grad_norm": 0.9592137932777405, + "learning_rate": 0.0004161485935589075, + "loss": 3.306, + "step": 68745 + }, + { + "epoch": 4.671150971599402, + "grad_norm": 1.3021209239959717, + "learning_rate": 0.00041610612855007474, + "loss": 3.5541, + "step": 68750 + }, + { + "epoch": 4.671490691670064, + "grad_norm": 1.0275015830993652, + "learning_rate": 0.000416063663541242, + "loss": 3.4993, + "step": 68755 + }, + { + "epoch": 4.671830411740726, + "grad_norm": 1.0702340602874756, + "learning_rate": 0.00041602119853240935, + "loss": 3.5444, + "step": 68760 + }, + { + "epoch": 4.672170131811387, + "grad_norm": 0.7555716037750244, + "learning_rate": 0.0004159787335235766, + "loss": 3.6743, + "step": 68765 + }, + { + "epoch": 4.672509851882049, + "grad_norm": 1.057653784751892, + "learning_rate": 0.00041593626851474386, + "loss": 3.2856, + "step": 68770 + }, + { + "epoch": 4.672849571952711, + "grad_norm": 0.6843161582946777, + "learning_rate": 0.00041589380350591114, + "loss": 3.7136, + "step": 68775 + }, + { + "epoch": 4.673189292023372, + "grad_norm": 0.8282147645950317, + "learning_rate": 0.0004158513384970784, + "loss": 3.6058, + "step": 68780 + }, + { + "epoch": 4.6735290120940345, + "grad_norm": 0.7954680323600769, + "learning_rate": 0.0004158088734882457, + "loss": 3.5762, + "step": 68785 + }, + { + "epoch": 4.673868732164697, + "grad_norm": 0.7610088586807251, + "learning_rate": 0.000415766408479413, + "loss": 3.4351, + "step": 68790 + }, + { + "epoch": 4.674208452235358, + "grad_norm": 0.8840131759643555, + "learning_rate": 0.00041572394347058026, + "loss": 3.3526, + "step": 68795 + }, + { + "epoch": 4.67454817230602, + "grad_norm": 0.8599383234977722, + "learning_rate": 0.00041568147846174754, + "loss": 3.434, + "step": 68800 + }, + { + "epoch": 4.674887892376682, + "grad_norm": 0.9170916676521301, + "learning_rate": 0.0004156390134529148, + "loss": 3.2948, + "step": 68805 + }, + { + "epoch": 4.675227612447343, + "grad_norm": 0.9280619025230408, + "learning_rate": 0.00041559654844408204, + "loss": 3.6941, + "step": 68810 + }, + { + "epoch": 4.675567332518005, + "grad_norm": 0.7576864957809448, + "learning_rate": 0.0004155540834352494, + "loss": 3.491, + "step": 68815 + }, + { + "epoch": 4.675907052588667, + "grad_norm": 0.8543628454208374, + "learning_rate": 0.00041551161842641666, + "loss": 3.5155, + "step": 68820 + }, + { + "epoch": 4.676246772659328, + "grad_norm": 0.8939573764801025, + "learning_rate": 0.0004154691534175839, + "loss": 3.3863, + "step": 68825 + }, + { + "epoch": 4.6765864927299905, + "grad_norm": 0.8860607147216797, + "learning_rate": 0.0004154266884087512, + "loss": 3.3293, + "step": 68830 + }, + { + "epoch": 4.676926212800653, + "grad_norm": 1.1328390836715698, + "learning_rate": 0.0004153842233999185, + "loss": 3.5256, + "step": 68835 + }, + { + "epoch": 4.677265932871314, + "grad_norm": 0.9975154995918274, + "learning_rate": 0.0004153417583910857, + "loss": 3.2659, + "step": 68840 + }, + { + "epoch": 4.677605652941976, + "grad_norm": 0.9693921208381653, + "learning_rate": 0.000415299293382253, + "loss": 3.4272, + "step": 68845 + }, + { + "epoch": 4.677945373012638, + "grad_norm": 0.8597924709320068, + "learning_rate": 0.00041525682837342034, + "loss": 3.3507, + "step": 68850 + }, + { + "epoch": 4.678285093083299, + "grad_norm": 0.770686686038971, + "learning_rate": 0.00041521436336458756, + "loss": 3.4368, + "step": 68855 + }, + { + "epoch": 4.678624813153961, + "grad_norm": 1.0872145891189575, + "learning_rate": 0.00041517189835575484, + "loss": 3.4364, + "step": 68860 + }, + { + "epoch": 4.678964533224623, + "grad_norm": 1.0175124406814575, + "learning_rate": 0.0004151294333469222, + "loss": 3.3223, + "step": 68865 + }, + { + "epoch": 4.679304253295284, + "grad_norm": 0.9084970951080322, + "learning_rate": 0.0004150869683380894, + "loss": 3.256, + "step": 68870 + }, + { + "epoch": 4.6796439733659465, + "grad_norm": 0.9755767583847046, + "learning_rate": 0.0004150445033292567, + "loss": 3.5321, + "step": 68875 + }, + { + "epoch": 4.679983693436609, + "grad_norm": 0.8634095788002014, + "learning_rate": 0.000415002038320424, + "loss": 3.3742, + "step": 68880 + }, + { + "epoch": 4.68032341350727, + "grad_norm": 0.8646619915962219, + "learning_rate": 0.00041495957331159124, + "loss": 3.4129, + "step": 68885 + }, + { + "epoch": 4.680663133577932, + "grad_norm": 1.0422086715698242, + "learning_rate": 0.0004149171083027585, + "loss": 3.561, + "step": 68890 + }, + { + "epoch": 4.681002853648594, + "grad_norm": 0.9213076829910278, + "learning_rate": 0.0004148746432939258, + "loss": 3.4369, + "step": 68895 + }, + { + "epoch": 4.681342573719255, + "grad_norm": 1.0218091011047363, + "learning_rate": 0.0004148321782850931, + "loss": 3.4145, + "step": 68900 + }, + { + "epoch": 4.681682293789917, + "grad_norm": 1.0470527410507202, + "learning_rate": 0.00041478971327626036, + "loss": 3.5418, + "step": 68905 + }, + { + "epoch": 4.682022013860579, + "grad_norm": 0.8341613411903381, + "learning_rate": 0.00041474724826742764, + "loss": 3.3583, + "step": 68910 + }, + { + "epoch": 4.68236173393124, + "grad_norm": 1.0093225240707397, + "learning_rate": 0.000414704783258595, + "loss": 3.4729, + "step": 68915 + }, + { + "epoch": 4.6827014540019025, + "grad_norm": 0.936660647392273, + "learning_rate": 0.0004146623182497622, + "loss": 3.7194, + "step": 68920 + }, + { + "epoch": 4.683041174072565, + "grad_norm": 0.9471144080162048, + "learning_rate": 0.0004146198532409295, + "loss": 3.5766, + "step": 68925 + }, + { + "epoch": 4.683380894143226, + "grad_norm": 0.8413189649581909, + "learning_rate": 0.00041457738823209676, + "loss": 3.4148, + "step": 68930 + }, + { + "epoch": 4.683720614213888, + "grad_norm": 0.8023260235786438, + "learning_rate": 0.00041453492322326404, + "loss": 3.1295, + "step": 68935 + }, + { + "epoch": 4.684060334284549, + "grad_norm": 0.769117534160614, + "learning_rate": 0.0004144924582144313, + "loss": 3.6845, + "step": 68940 + }, + { + "epoch": 4.684400054355211, + "grad_norm": 0.7573029398918152, + "learning_rate": 0.0004144499932055986, + "loss": 3.3628, + "step": 68945 + }, + { + "epoch": 4.684739774425873, + "grad_norm": 0.9057371616363525, + "learning_rate": 0.0004144075281967659, + "loss": 3.4598, + "step": 68950 + }, + { + "epoch": 4.685079494496534, + "grad_norm": 0.9903303384780884, + "learning_rate": 0.00041436506318793317, + "loss": 3.4652, + "step": 68955 + }, + { + "epoch": 4.6854192145671965, + "grad_norm": 1.0082008838653564, + "learning_rate": 0.00041432259817910045, + "loss": 3.4539, + "step": 68960 + }, + { + "epoch": 4.6857589346378585, + "grad_norm": 0.8353513479232788, + "learning_rate": 0.00041428013317026767, + "loss": 3.4434, + "step": 68965 + }, + { + "epoch": 4.68609865470852, + "grad_norm": 0.8890240788459778, + "learning_rate": 0.000414237668161435, + "loss": 3.2423, + "step": 68970 + }, + { + "epoch": 4.686438374779182, + "grad_norm": 0.9063505530357361, + "learning_rate": 0.0004141952031526023, + "loss": 3.5485, + "step": 68975 + }, + { + "epoch": 4.686778094849844, + "grad_norm": 0.7631552815437317, + "learning_rate": 0.0004141527381437695, + "loss": 3.5129, + "step": 68980 + }, + { + "epoch": 4.687117814920505, + "grad_norm": 0.9729147553443909, + "learning_rate": 0.00041411027313493685, + "loss": 3.1509, + "step": 68985 + }, + { + "epoch": 4.687457534991167, + "grad_norm": 0.8315043449401855, + "learning_rate": 0.0004140678081261041, + "loss": 3.5632, + "step": 68990 + }, + { + "epoch": 4.687797255061829, + "grad_norm": 1.5017859935760498, + "learning_rate": 0.00041402534311727135, + "loss": 3.6263, + "step": 68995 + }, + { + "epoch": 4.68813697513249, + "grad_norm": 1.2866005897521973, + "learning_rate": 0.00041398287810843863, + "loss": 3.3679, + "step": 69000 + }, + { + "epoch": 4.6884766952031525, + "grad_norm": 0.9483434557914734, + "learning_rate": 0.00041394041309960597, + "loss": 3.684, + "step": 69005 + }, + { + "epoch": 4.6888164152738145, + "grad_norm": 0.8910866379737854, + "learning_rate": 0.0004138979480907732, + "loss": 3.6466, + "step": 69010 + }, + { + "epoch": 4.689156135344476, + "grad_norm": 0.8734315633773804, + "learning_rate": 0.00041385548308194047, + "loss": 3.7488, + "step": 69015 + }, + { + "epoch": 4.689495855415138, + "grad_norm": 1.0456229448318481, + "learning_rate": 0.0004138130180731078, + "loss": 3.541, + "step": 69020 + }, + { + "epoch": 4.6898355754858, + "grad_norm": 1.072800636291504, + "learning_rate": 0.00041377055306427503, + "loss": 3.5719, + "step": 69025 + }, + { + "epoch": 4.690175295556461, + "grad_norm": 0.8045454621315002, + "learning_rate": 0.0004137280880554423, + "loss": 3.8004, + "step": 69030 + }, + { + "epoch": 4.690515015627123, + "grad_norm": 0.7314339876174927, + "learning_rate": 0.0004136856230466096, + "loss": 3.3762, + "step": 69035 + }, + { + "epoch": 4.690854735697785, + "grad_norm": 0.8361109495162964, + "learning_rate": 0.00041364315803777687, + "loss": 3.3964, + "step": 69040 + }, + { + "epoch": 4.691194455768446, + "grad_norm": 0.9029577970504761, + "learning_rate": 0.00041360069302894415, + "loss": 3.5059, + "step": 69045 + }, + { + "epoch": 4.6915341758391085, + "grad_norm": 1.2076871395111084, + "learning_rate": 0.00041355822802011143, + "loss": 3.5094, + "step": 69050 + }, + { + "epoch": 4.6918738959097706, + "grad_norm": 0.9464370012283325, + "learning_rate": 0.0004135157630112787, + "loss": 3.6401, + "step": 69055 + }, + { + "epoch": 4.692213615980432, + "grad_norm": 1.0849424600601196, + "learning_rate": 0.000413473298002446, + "loss": 3.6149, + "step": 69060 + }, + { + "epoch": 4.692553336051094, + "grad_norm": 0.7472956776618958, + "learning_rate": 0.00041343083299361327, + "loss": 3.5002, + "step": 69065 + }, + { + "epoch": 4.692893056121756, + "grad_norm": 0.9377622008323669, + "learning_rate": 0.0004133883679847805, + "loss": 3.6878, + "step": 69070 + }, + { + "epoch": 4.693232776192417, + "grad_norm": 0.9892808198928833, + "learning_rate": 0.00041334590297594783, + "loss": 3.8907, + "step": 69075 + }, + { + "epoch": 4.693572496263079, + "grad_norm": 0.7633818984031677, + "learning_rate": 0.0004133034379671151, + "loss": 3.5556, + "step": 69080 + }, + { + "epoch": 4.693912216333741, + "grad_norm": 0.7865645885467529, + "learning_rate": 0.0004132609729582824, + "loss": 3.7338, + "step": 69085 + }, + { + "epoch": 4.694251936404402, + "grad_norm": 0.9033628702163696, + "learning_rate": 0.0004132185079494497, + "loss": 3.1633, + "step": 69090 + }, + { + "epoch": 4.6945916564750645, + "grad_norm": 0.793864369392395, + "learning_rate": 0.00041317604294061695, + "loss": 3.6217, + "step": 69095 + }, + { + "epoch": 4.694931376545727, + "grad_norm": 0.9220806360244751, + "learning_rate": 0.00041313357793178423, + "loss": 3.5634, + "step": 69100 + }, + { + "epoch": 4.695271096616388, + "grad_norm": 0.9680860638618469, + "learning_rate": 0.00041309111292295146, + "loss": 3.6296, + "step": 69105 + }, + { + "epoch": 4.69561081668705, + "grad_norm": 0.8768517971038818, + "learning_rate": 0.0004130486479141188, + "loss": 3.4488, + "step": 69110 + }, + { + "epoch": 4.695950536757712, + "grad_norm": 0.9426429271697998, + "learning_rate": 0.0004130061829052861, + "loss": 3.5465, + "step": 69115 + }, + { + "epoch": 4.696290256828373, + "grad_norm": 0.7944425344467163, + "learning_rate": 0.0004129637178964533, + "loss": 3.6723, + "step": 69120 + }, + { + "epoch": 4.696629976899035, + "grad_norm": 2.1346771717071533, + "learning_rate": 0.00041292125288762063, + "loss": 3.4964, + "step": 69125 + }, + { + "epoch": 4.696969696969697, + "grad_norm": 1.229452133178711, + "learning_rate": 0.0004128787878787879, + "loss": 3.2191, + "step": 69130 + }, + { + "epoch": 4.697309417040358, + "grad_norm": 0.7481483221054077, + "learning_rate": 0.00041283632286995514, + "loss": 3.5667, + "step": 69135 + }, + { + "epoch": 4.6976491371110205, + "grad_norm": 0.8556455373764038, + "learning_rate": 0.0004127938578611224, + "loss": 3.4788, + "step": 69140 + }, + { + "epoch": 4.697988857181683, + "grad_norm": 0.8856696486473083, + "learning_rate": 0.00041275139285228975, + "loss": 3.4867, + "step": 69145 + }, + { + "epoch": 4.698328577252344, + "grad_norm": 0.8202235102653503, + "learning_rate": 0.000412708927843457, + "loss": 3.6148, + "step": 69150 + }, + { + "epoch": 4.698668297323006, + "grad_norm": 1.042313575744629, + "learning_rate": 0.00041266646283462426, + "loss": 3.5665, + "step": 69155 + }, + { + "epoch": 4.699008017393668, + "grad_norm": 0.7703211307525635, + "learning_rate": 0.0004126239978257916, + "loss": 3.6274, + "step": 69160 + }, + { + "epoch": 4.699347737464329, + "grad_norm": 0.9743807315826416, + "learning_rate": 0.0004125815328169588, + "loss": 3.3516, + "step": 69165 + }, + { + "epoch": 4.699687457534991, + "grad_norm": 0.8901051878929138, + "learning_rate": 0.0004125390678081261, + "loss": 3.5179, + "step": 69170 + }, + { + "epoch": 4.700027177605653, + "grad_norm": 0.9366716146469116, + "learning_rate": 0.00041249660279929343, + "loss": 3.3963, + "step": 69175 + }, + { + "epoch": 4.700366897676314, + "grad_norm": 0.8757122755050659, + "learning_rate": 0.00041245413779046066, + "loss": 3.348, + "step": 69180 + }, + { + "epoch": 4.7007066177469765, + "grad_norm": 0.8713012337684631, + "learning_rate": 0.00041241167278162794, + "loss": 3.4491, + "step": 69185 + }, + { + "epoch": 4.701046337817639, + "grad_norm": 1.0504965782165527, + "learning_rate": 0.0004123692077727952, + "loss": 3.1821, + "step": 69190 + }, + { + "epoch": 4.7013860578883, + "grad_norm": 0.7504411935806274, + "learning_rate": 0.0004123267427639625, + "loss": 3.7234, + "step": 69195 + }, + { + "epoch": 4.701725777958962, + "grad_norm": 0.9374606609344482, + "learning_rate": 0.0004122842777551298, + "loss": 3.5147, + "step": 69200 + }, + { + "epoch": 4.702065498029624, + "grad_norm": 0.9495920538902283, + "learning_rate": 0.00041224181274629706, + "loss": 3.5453, + "step": 69205 + }, + { + "epoch": 4.702405218100285, + "grad_norm": 0.8455723524093628, + "learning_rate": 0.00041219934773746434, + "loss": 3.5291, + "step": 69210 + }, + { + "epoch": 4.702744938170947, + "grad_norm": 1.476053237915039, + "learning_rate": 0.0004121568827286316, + "loss": 3.5161, + "step": 69215 + }, + { + "epoch": 4.703084658241609, + "grad_norm": 1.0459927320480347, + "learning_rate": 0.0004121144177197989, + "loss": 3.5756, + "step": 69220 + }, + { + "epoch": 4.7034243783122704, + "grad_norm": 0.9662806391716003, + "learning_rate": 0.0004120719527109661, + "loss": 3.5127, + "step": 69225 + }, + { + "epoch": 4.7037640983829325, + "grad_norm": 0.782564640045166, + "learning_rate": 0.00041202948770213346, + "loss": 3.6749, + "step": 69230 + }, + { + "epoch": 4.704103818453595, + "grad_norm": 0.853685736656189, + "learning_rate": 0.00041198702269330074, + "loss": 3.4926, + "step": 69235 + }, + { + "epoch": 4.704443538524256, + "grad_norm": 0.8494096994400024, + "learning_rate": 0.00041194455768446797, + "loss": 3.5054, + "step": 69240 + }, + { + "epoch": 4.704783258594918, + "grad_norm": 0.866094172000885, + "learning_rate": 0.0004119020926756353, + "loss": 3.5804, + "step": 69245 + }, + { + "epoch": 4.70512297866558, + "grad_norm": 0.9784623980522156, + "learning_rate": 0.0004118596276668026, + "loss": 3.5316, + "step": 69250 + }, + { + "epoch": 4.705462698736241, + "grad_norm": 0.9873091578483582, + "learning_rate": 0.00041181716265796986, + "loss": 3.6019, + "step": 69255 + }, + { + "epoch": 4.705802418806903, + "grad_norm": 1.0187195539474487, + "learning_rate": 0.0004117746976491371, + "loss": 3.5431, + "step": 69260 + }, + { + "epoch": 4.706142138877565, + "grad_norm": 0.8958463668823242, + "learning_rate": 0.0004117322326403044, + "loss": 3.358, + "step": 69265 + }, + { + "epoch": 4.7064818589482265, + "grad_norm": 0.6892831325531006, + "learning_rate": 0.0004116897676314717, + "loss": 3.4991, + "step": 69270 + }, + { + "epoch": 4.7068215790188885, + "grad_norm": 0.8882495760917664, + "learning_rate": 0.00041164730262263893, + "loss": 3.3297, + "step": 69275 + }, + { + "epoch": 4.707161299089551, + "grad_norm": 0.8000712394714355, + "learning_rate": 0.00041160483761380626, + "loss": 3.3443, + "step": 69280 + }, + { + "epoch": 4.707501019160212, + "grad_norm": 0.91518634557724, + "learning_rate": 0.00041156237260497354, + "loss": 3.3768, + "step": 69285 + }, + { + "epoch": 4.707840739230874, + "grad_norm": 0.8943261504173279, + "learning_rate": 0.00041151990759614077, + "loss": 3.3386, + "step": 69290 + }, + { + "epoch": 4.708180459301536, + "grad_norm": 0.9731014966964722, + "learning_rate": 0.00041147744258730805, + "loss": 3.4249, + "step": 69295 + }, + { + "epoch": 4.708520179372197, + "grad_norm": 0.9764625430107117, + "learning_rate": 0.0004114349775784754, + "loss": 3.6803, + "step": 69300 + }, + { + "epoch": 4.708859899442859, + "grad_norm": 0.8385946154594421, + "learning_rate": 0.0004113925125696426, + "loss": 3.6429, + "step": 69305 + }, + { + "epoch": 4.709199619513521, + "grad_norm": 0.9026013612747192, + "learning_rate": 0.0004113500475608099, + "loss": 3.4236, + "step": 69310 + }, + { + "epoch": 4.7095393395841825, + "grad_norm": 0.6916206479072571, + "learning_rate": 0.0004113075825519772, + "loss": 3.5538, + "step": 69315 + }, + { + "epoch": 4.7098790596548445, + "grad_norm": 0.9716611504554749, + "learning_rate": 0.00041126511754314445, + "loss": 3.5718, + "step": 69320 + }, + { + "epoch": 4.710218779725507, + "grad_norm": 0.7719985246658325, + "learning_rate": 0.00041122265253431173, + "loss": 3.4827, + "step": 69325 + }, + { + "epoch": 4.710558499796168, + "grad_norm": 0.9900364279747009, + "learning_rate": 0.000411180187525479, + "loss": 3.4345, + "step": 69330 + }, + { + "epoch": 4.71089821986683, + "grad_norm": 0.8607082366943359, + "learning_rate": 0.0004111377225166463, + "loss": 3.3395, + "step": 69335 + }, + { + "epoch": 4.711237939937492, + "grad_norm": 0.8687829375267029, + "learning_rate": 0.00041109525750781357, + "loss": 3.8701, + "step": 69340 + }, + { + "epoch": 4.711577660008153, + "grad_norm": 0.9851251840591431, + "learning_rate": 0.00041105279249898085, + "loss": 3.5265, + "step": 69345 + }, + { + "epoch": 4.711917380078815, + "grad_norm": 0.9813066124916077, + "learning_rate": 0.00041101032749014813, + "loss": 3.2969, + "step": 69350 + }, + { + "epoch": 4.712257100149477, + "grad_norm": 0.9464268684387207, + "learning_rate": 0.0004109678624813154, + "loss": 3.3318, + "step": 69355 + }, + { + "epoch": 4.7125968202201385, + "grad_norm": 0.7016441822052002, + "learning_rate": 0.0004109253974724827, + "loss": 3.46, + "step": 69360 + }, + { + "epoch": 4.712936540290801, + "grad_norm": 0.6918023228645325, + "learning_rate": 0.0004108829324636499, + "loss": 3.6066, + "step": 69365 + }, + { + "epoch": 4.713276260361463, + "grad_norm": 0.949932873249054, + "learning_rate": 0.00041084046745481725, + "loss": 3.4233, + "step": 69370 + }, + { + "epoch": 4.713615980432124, + "grad_norm": 1.0643188953399658, + "learning_rate": 0.00041079800244598453, + "loss": 3.4068, + "step": 69375 + }, + { + "epoch": 4.713955700502786, + "grad_norm": 0.7567729353904724, + "learning_rate": 0.00041075553743715175, + "loss": 3.4116, + "step": 69380 + }, + { + "epoch": 4.714295420573448, + "grad_norm": 0.823922872543335, + "learning_rate": 0.0004107130724283191, + "loss": 3.3064, + "step": 69385 + }, + { + "epoch": 4.714635140644109, + "grad_norm": 0.8823032975196838, + "learning_rate": 0.00041067060741948637, + "loss": 3.2631, + "step": 69390 + }, + { + "epoch": 4.714974860714771, + "grad_norm": 0.8701813817024231, + "learning_rate": 0.0004106281424106536, + "loss": 3.6263, + "step": 69395 + }, + { + "epoch": 4.715314580785432, + "grad_norm": 0.9479604363441467, + "learning_rate": 0.0004105856774018209, + "loss": 3.5492, + "step": 69400 + }, + { + "epoch": 4.7156543008560945, + "grad_norm": 1.1740418672561646, + "learning_rate": 0.0004105432123929882, + "loss": 3.6143, + "step": 69405 + }, + { + "epoch": 4.715994020926757, + "grad_norm": 0.9979771971702576, + "learning_rate": 0.00041050074738415544, + "loss": 3.5649, + "step": 69410 + }, + { + "epoch": 4.716333740997418, + "grad_norm": 0.970445990562439, + "learning_rate": 0.0004104582823753227, + "loss": 3.2176, + "step": 69415 + }, + { + "epoch": 4.71667346106808, + "grad_norm": 1.3118610382080078, + "learning_rate": 0.00041041581736649005, + "loss": 3.31, + "step": 69420 + }, + { + "epoch": 4.717013181138742, + "grad_norm": 0.784374475479126, + "learning_rate": 0.00041037335235765733, + "loss": 3.6023, + "step": 69425 + }, + { + "epoch": 4.717352901209403, + "grad_norm": 0.7657243609428406, + "learning_rate": 0.00041033088734882456, + "loss": 3.48, + "step": 69430 + }, + { + "epoch": 4.717692621280065, + "grad_norm": 0.8319544196128845, + "learning_rate": 0.00041028842233999184, + "loss": 3.467, + "step": 69435 + }, + { + "epoch": 4.718032341350727, + "grad_norm": 0.7758208513259888, + "learning_rate": 0.00041024595733115917, + "loss": 3.2968, + "step": 69440 + }, + { + "epoch": 4.718372061421388, + "grad_norm": 0.9755234122276306, + "learning_rate": 0.0004102034923223264, + "loss": 3.5437, + "step": 69445 + }, + { + "epoch": 4.7187117814920505, + "grad_norm": 0.8124076128005981, + "learning_rate": 0.0004101610273134937, + "loss": 3.5811, + "step": 69450 + }, + { + "epoch": 4.719051501562713, + "grad_norm": 1.1775474548339844, + "learning_rate": 0.000410118562304661, + "loss": 3.1302, + "step": 69455 + }, + { + "epoch": 4.719391221633374, + "grad_norm": 6.396855354309082, + "learning_rate": 0.00041007609729582824, + "loss": 3.4236, + "step": 69460 + }, + { + "epoch": 4.719730941704036, + "grad_norm": 0.9627898931503296, + "learning_rate": 0.0004100336322869955, + "loss": 3.5744, + "step": 69465 + }, + { + "epoch": 4.720070661774698, + "grad_norm": 0.9507832527160645, + "learning_rate": 0.00040999116727816285, + "loss": 3.532, + "step": 69470 + }, + { + "epoch": 4.720410381845359, + "grad_norm": 1.0489801168441772, + "learning_rate": 0.0004099487022693301, + "loss": 3.3575, + "step": 69475 + }, + { + "epoch": 4.720750101916021, + "grad_norm": 0.8134302496910095, + "learning_rate": 0.00040990623726049736, + "loss": 3.5469, + "step": 69480 + }, + { + "epoch": 4.721089821986683, + "grad_norm": 1.0447328090667725, + "learning_rate": 0.00040986377225166464, + "loss": 3.2181, + "step": 69485 + }, + { + "epoch": 4.721429542057344, + "grad_norm": 0.761895477771759, + "learning_rate": 0.0004098213072428319, + "loss": 3.46, + "step": 69490 + }, + { + "epoch": 4.7217692621280065, + "grad_norm": 1.0900856256484985, + "learning_rate": 0.0004097788422339992, + "loss": 3.6524, + "step": 69495 + }, + { + "epoch": 4.722108982198669, + "grad_norm": 0.8806496262550354, + "learning_rate": 0.0004097363772251665, + "loss": 3.6766, + "step": 69500 + }, + { + "epoch": 4.72244870226933, + "grad_norm": 0.8109956383705139, + "learning_rate": 0.00040969391221633376, + "loss": 3.565, + "step": 69505 + }, + { + "epoch": 4.722788422339992, + "grad_norm": 0.9513434171676636, + "learning_rate": 0.00040965144720750104, + "loss": 3.5646, + "step": 69510 + }, + { + "epoch": 4.723128142410654, + "grad_norm": 1.059083342552185, + "learning_rate": 0.0004096089821986683, + "loss": 3.4498, + "step": 69515 + }, + { + "epoch": 4.723467862481315, + "grad_norm": 0.9925141334533691, + "learning_rate": 0.00040956651718983554, + "loss": 3.4791, + "step": 69520 + }, + { + "epoch": 4.723807582551977, + "grad_norm": 0.7216336727142334, + "learning_rate": 0.0004095240521810029, + "loss": 3.429, + "step": 69525 + }, + { + "epoch": 4.724147302622639, + "grad_norm": 1.161681890487671, + "learning_rate": 0.00040948158717217016, + "loss": 3.5496, + "step": 69530 + }, + { + "epoch": 4.7244870226933005, + "grad_norm": 0.9172937870025635, + "learning_rate": 0.0004094391221633374, + "loss": 3.2494, + "step": 69535 + }, + { + "epoch": 4.7248267427639625, + "grad_norm": 0.8286041021347046, + "learning_rate": 0.0004093966571545047, + "loss": 3.4217, + "step": 69540 + }, + { + "epoch": 4.725166462834625, + "grad_norm": 0.9892429113388062, + "learning_rate": 0.000409354192145672, + "loss": 3.6047, + "step": 69545 + }, + { + "epoch": 4.725506182905286, + "grad_norm": 0.6669461131095886, + "learning_rate": 0.0004093117271368392, + "loss": 3.9375, + "step": 69550 + }, + { + "epoch": 4.725845902975948, + "grad_norm": 0.7384819388389587, + "learning_rate": 0.0004092692621280065, + "loss": 3.2732, + "step": 69555 + }, + { + "epoch": 4.72618562304661, + "grad_norm": 0.9237121939659119, + "learning_rate": 0.00040922679711917384, + "loss": 3.1799, + "step": 69560 + }, + { + "epoch": 4.726525343117271, + "grad_norm": 1.162797451019287, + "learning_rate": 0.00040918433211034106, + "loss": 3.5168, + "step": 69565 + }, + { + "epoch": 4.726865063187933, + "grad_norm": 0.8678475618362427, + "learning_rate": 0.00040914186710150834, + "loss": 3.5217, + "step": 69570 + }, + { + "epoch": 4.727204783258595, + "grad_norm": 0.8531060814857483, + "learning_rate": 0.0004090994020926757, + "loss": 3.3051, + "step": 69575 + }, + { + "epoch": 4.7275445033292565, + "grad_norm": 0.8902525901794434, + "learning_rate": 0.0004090569370838429, + "loss": 3.5042, + "step": 69580 + }, + { + "epoch": 4.7278842233999185, + "grad_norm": 1.0029677152633667, + "learning_rate": 0.0004090144720750102, + "loss": 3.3889, + "step": 69585 + }, + { + "epoch": 4.728223943470581, + "grad_norm": 0.9958492517471313, + "learning_rate": 0.00040897200706617746, + "loss": 3.5852, + "step": 69590 + }, + { + "epoch": 4.728563663541242, + "grad_norm": 0.9596347212791443, + "learning_rate": 0.0004089295420573448, + "loss": 3.4956, + "step": 69595 + }, + { + "epoch": 4.728903383611904, + "grad_norm": 0.8545126914978027, + "learning_rate": 0.000408887077048512, + "loss": 3.4664, + "step": 69600 + }, + { + "epoch": 4.729243103682566, + "grad_norm": 0.8597524762153625, + "learning_rate": 0.0004088446120396793, + "loss": 3.4044, + "step": 69605 + }, + { + "epoch": 4.729582823753227, + "grad_norm": 0.7985163331031799, + "learning_rate": 0.00040880214703084664, + "loss": 3.6948, + "step": 69610 + }, + { + "epoch": 4.729922543823889, + "grad_norm": 1.2284164428710938, + "learning_rate": 0.00040875968202201386, + "loss": 3.2078, + "step": 69615 + }, + { + "epoch": 4.73026226389455, + "grad_norm": 0.7609068751335144, + "learning_rate": 0.00040871721701318114, + "loss": 3.4171, + "step": 69620 + }, + { + "epoch": 4.7306019839652125, + "grad_norm": 0.9071590900421143, + "learning_rate": 0.0004086747520043484, + "loss": 3.5444, + "step": 69625 + }, + { + "epoch": 4.7309417040358746, + "grad_norm": 0.8441505432128906, + "learning_rate": 0.0004086322869955157, + "loss": 3.5638, + "step": 69630 + }, + { + "epoch": 4.731281424106536, + "grad_norm": 0.9613469243049622, + "learning_rate": 0.000408589821986683, + "loss": 3.3233, + "step": 69635 + }, + { + "epoch": 4.731621144177198, + "grad_norm": 0.9090893268585205, + "learning_rate": 0.00040854735697785026, + "loss": 3.589, + "step": 69640 + }, + { + "epoch": 4.73196086424786, + "grad_norm": 0.8683876395225525, + "learning_rate": 0.00040850489196901754, + "loss": 3.4098, + "step": 69645 + }, + { + "epoch": 4.732300584318521, + "grad_norm": 0.9005845785140991, + "learning_rate": 0.0004084624269601848, + "loss": 3.3478, + "step": 69650 + }, + { + "epoch": 4.732640304389183, + "grad_norm": 0.945499062538147, + "learning_rate": 0.0004084199619513521, + "loss": 3.6038, + "step": 69655 + }, + { + "epoch": 4.732980024459845, + "grad_norm": 0.812657356262207, + "learning_rate": 0.00040837749694251933, + "loss": 3.478, + "step": 69660 + }, + { + "epoch": 4.733319744530506, + "grad_norm": 0.846960723400116, + "learning_rate": 0.00040833503193368667, + "loss": 3.1262, + "step": 69665 + }, + { + "epoch": 4.7336594646011685, + "grad_norm": 0.7913996577262878, + "learning_rate": 0.00040829256692485395, + "loss": 3.6322, + "step": 69670 + }, + { + "epoch": 4.733999184671831, + "grad_norm": 0.9658031463623047, + "learning_rate": 0.00040825010191602117, + "loss": 3.5695, + "step": 69675 + }, + { + "epoch": 4.734338904742492, + "grad_norm": 0.8523339033126831, + "learning_rate": 0.0004082076369071885, + "loss": 3.5488, + "step": 69680 + }, + { + "epoch": 4.734678624813154, + "grad_norm": 0.7953420877456665, + "learning_rate": 0.0004081651718983558, + "loss": 3.4607, + "step": 69685 + }, + { + "epoch": 4.735018344883816, + "grad_norm": 0.860140323638916, + "learning_rate": 0.000408122706889523, + "loss": 3.7551, + "step": 69690 + }, + { + "epoch": 4.735358064954477, + "grad_norm": 0.8767080903053284, + "learning_rate": 0.0004080802418806903, + "loss": 3.3991, + "step": 69695 + }, + { + "epoch": 4.735697785025139, + "grad_norm": 0.8981719613075256, + "learning_rate": 0.0004080377768718576, + "loss": 3.2561, + "step": 69700 + }, + { + "epoch": 4.736037505095801, + "grad_norm": 0.954849898815155, + "learning_rate": 0.00040799531186302485, + "loss": 3.3468, + "step": 69705 + }, + { + "epoch": 4.736377225166462, + "grad_norm": 1.090592622756958, + "learning_rate": 0.00040795284685419213, + "loss": 3.5751, + "step": 69710 + }, + { + "epoch": 4.7367169452371245, + "grad_norm": 0.9520303606987, + "learning_rate": 0.00040791038184535947, + "loss": 3.5867, + "step": 69715 + }, + { + "epoch": 4.737056665307787, + "grad_norm": 0.9403046369552612, + "learning_rate": 0.0004078679168365267, + "loss": 3.0102, + "step": 69720 + }, + { + "epoch": 4.737396385378448, + "grad_norm": 0.7786111235618591, + "learning_rate": 0.00040782545182769397, + "loss": 3.8316, + "step": 69725 + }, + { + "epoch": 4.73773610544911, + "grad_norm": 1.0550023317337036, + "learning_rate": 0.0004077829868188613, + "loss": 3.4379, + "step": 69730 + }, + { + "epoch": 4.738075825519772, + "grad_norm": 0.8314580917358398, + "learning_rate": 0.00040774052181002853, + "loss": 3.3578, + "step": 69735 + }, + { + "epoch": 4.738415545590433, + "grad_norm": 0.8088896870613098, + "learning_rate": 0.0004076980568011958, + "loss": 3.6589, + "step": 69740 + }, + { + "epoch": 4.738755265661095, + "grad_norm": 0.9897230863571167, + "learning_rate": 0.0004076555917923631, + "loss": 3.6297, + "step": 69745 + }, + { + "epoch": 4.739094985731757, + "grad_norm": 0.817943811416626, + "learning_rate": 0.00040761312678353037, + "loss": 3.6192, + "step": 69750 + }, + { + "epoch": 4.739434705802418, + "grad_norm": 0.9540905952453613, + "learning_rate": 0.00040757066177469765, + "loss": 3.6626, + "step": 69755 + }, + { + "epoch": 4.7397744258730805, + "grad_norm": 1.6650946140289307, + "learning_rate": 0.00040752819676586493, + "loss": 3.399, + "step": 69760 + }, + { + "epoch": 4.740114145943743, + "grad_norm": 0.8696034550666809, + "learning_rate": 0.00040748573175703227, + "loss": 3.6477, + "step": 69765 + }, + { + "epoch": 4.740453866014404, + "grad_norm": 1.1270350217819214, + "learning_rate": 0.0004074432667481995, + "loss": 3.4435, + "step": 69770 + }, + { + "epoch": 4.740793586085066, + "grad_norm": 0.9530674815177917, + "learning_rate": 0.00040740080173936677, + "loss": 3.3867, + "step": 69775 + }, + { + "epoch": 4.741133306155728, + "grad_norm": 1.123913288116455, + "learning_rate": 0.00040735833673053405, + "loss": 3.4357, + "step": 69780 + }, + { + "epoch": 4.741473026226389, + "grad_norm": 0.9536225199699402, + "learning_rate": 0.00040731587172170133, + "loss": 3.4923, + "step": 69785 + }, + { + "epoch": 4.741812746297051, + "grad_norm": 1.0449191331863403, + "learning_rate": 0.0004072734067128686, + "loss": 3.5092, + "step": 69790 + }, + { + "epoch": 4.742152466367713, + "grad_norm": 1.021342158317566, + "learning_rate": 0.0004072309417040359, + "loss": 3.7529, + "step": 69795 + }, + { + "epoch": 4.7424921864383744, + "grad_norm": 0.954954206943512, + "learning_rate": 0.0004071884766952032, + "loss": 3.3732, + "step": 69800 + }, + { + "epoch": 4.7428319065090365, + "grad_norm": 0.8487341403961182, + "learning_rate": 0.00040714601168637045, + "loss": 3.3324, + "step": 69805 + }, + { + "epoch": 4.743171626579699, + "grad_norm": 1.2022784948349, + "learning_rate": 0.00040710354667753773, + "loss": 3.4762, + "step": 69810 + }, + { + "epoch": 4.74351134665036, + "grad_norm": 0.9926097989082336, + "learning_rate": 0.00040706108166870496, + "loss": 2.9816, + "step": 69815 + }, + { + "epoch": 4.743851066721022, + "grad_norm": 0.9748802781105042, + "learning_rate": 0.0004070186166598723, + "loss": 3.503, + "step": 69820 + }, + { + "epoch": 4.744190786791684, + "grad_norm": 0.8919101357460022, + "learning_rate": 0.0004069761516510396, + "loss": 3.5095, + "step": 69825 + }, + { + "epoch": 4.744530506862345, + "grad_norm": 0.8129463195800781, + "learning_rate": 0.0004069336866422068, + "loss": 3.4917, + "step": 69830 + }, + { + "epoch": 4.744870226933007, + "grad_norm": 1.0738404989242554, + "learning_rate": 0.00040689122163337413, + "loss": 3.6459, + "step": 69835 + }, + { + "epoch": 4.745209947003669, + "grad_norm": 1.0326906442642212, + "learning_rate": 0.0004068487566245414, + "loss": 3.2607, + "step": 69840 + }, + { + "epoch": 4.7455496670743305, + "grad_norm": 0.9654978513717651, + "learning_rate": 0.00040680629161570864, + "loss": 3.5042, + "step": 69845 + }, + { + "epoch": 4.7458893871449925, + "grad_norm": 0.9114751219749451, + "learning_rate": 0.0004067638266068759, + "loss": 3.653, + "step": 69850 + }, + { + "epoch": 4.746229107215655, + "grad_norm": 0.970568835735321, + "learning_rate": 0.00040672136159804325, + "loss": 3.7439, + "step": 69855 + }, + { + "epoch": 4.746568827286316, + "grad_norm": 0.9871333837509155, + "learning_rate": 0.0004066788965892105, + "loss": 3.1905, + "step": 69860 + }, + { + "epoch": 4.746908547356978, + "grad_norm": 0.7907553911209106, + "learning_rate": 0.00040663643158037776, + "loss": 3.6405, + "step": 69865 + }, + { + "epoch": 4.74724826742764, + "grad_norm": 0.9636352062225342, + "learning_rate": 0.0004065939665715451, + "loss": 3.2468, + "step": 69870 + }, + { + "epoch": 4.747587987498301, + "grad_norm": 1.0626206398010254, + "learning_rate": 0.0004065515015627123, + "loss": 3.4494, + "step": 69875 + }, + { + "epoch": 4.747927707568963, + "grad_norm": 0.8582151532173157, + "learning_rate": 0.0004065090365538796, + "loss": 3.6137, + "step": 69880 + }, + { + "epoch": 4.748267427639625, + "grad_norm": 0.8358616828918457, + "learning_rate": 0.0004064665715450469, + "loss": 3.5628, + "step": 69885 + }, + { + "epoch": 4.7486071477102865, + "grad_norm": 0.869587779045105, + "learning_rate": 0.00040642410653621416, + "loss": 3.7545, + "step": 69890 + }, + { + "epoch": 4.7489468677809485, + "grad_norm": 1.04090416431427, + "learning_rate": 0.00040638164152738144, + "loss": 3.5374, + "step": 69895 + }, + { + "epoch": 4.749286587851611, + "grad_norm": 1.006638765335083, + "learning_rate": 0.0004063391765185487, + "loss": 3.4869, + "step": 69900 + }, + { + "epoch": 4.749626307922272, + "grad_norm": 0.9677208065986633, + "learning_rate": 0.000406296711509716, + "loss": 3.3301, + "step": 69905 + }, + { + "epoch": 4.749966027992934, + "grad_norm": 0.9345969557762146, + "learning_rate": 0.0004062542465008833, + "loss": 3.5362, + "step": 69910 + }, + { + "epoch": 4.750305748063596, + "grad_norm": 0.7895476222038269, + "learning_rate": 0.00040621178149205056, + "loss": 3.6504, + "step": 69915 + }, + { + "epoch": 4.750645468134257, + "grad_norm": 0.8894950151443481, + "learning_rate": 0.0004061693164832178, + "loss": 3.4556, + "step": 69920 + }, + { + "epoch": 4.750985188204919, + "grad_norm": 0.9063246846199036, + "learning_rate": 0.0004061268514743851, + "loss": 3.3695, + "step": 69925 + }, + { + "epoch": 4.751324908275581, + "grad_norm": 0.8349266648292542, + "learning_rate": 0.0004060843864655524, + "loss": 3.336, + "step": 69930 + }, + { + "epoch": 4.7516646283462425, + "grad_norm": 1.0347188711166382, + "learning_rate": 0.0004060419214567197, + "loss": 3.1784, + "step": 69935 + }, + { + "epoch": 4.7520043484169046, + "grad_norm": 0.7952750325202942, + "learning_rate": 0.00040599945644788696, + "loss": 3.5084, + "step": 69940 + }, + { + "epoch": 4.752344068487567, + "grad_norm": 1.0403187274932861, + "learning_rate": 0.00040595699143905424, + "loss": 3.7953, + "step": 69945 + }, + { + "epoch": 4.752683788558228, + "grad_norm": 0.8723675608634949, + "learning_rate": 0.0004059145264302215, + "loss": 3.2913, + "step": 69950 + }, + { + "epoch": 4.75302350862889, + "grad_norm": 0.7564780116081238, + "learning_rate": 0.00040587206142138875, + "loss": 3.4079, + "step": 69955 + }, + { + "epoch": 4.753363228699552, + "grad_norm": 0.9965308308601379, + "learning_rate": 0.0004058295964125561, + "loss": 3.5497, + "step": 69960 + }, + { + "epoch": 4.753702948770213, + "grad_norm": 0.8512517213821411, + "learning_rate": 0.00040578713140372336, + "loss": 3.6836, + "step": 69965 + }, + { + "epoch": 4.754042668840875, + "grad_norm": 0.9106812477111816, + "learning_rate": 0.0004057446663948906, + "loss": 3.5245, + "step": 69970 + }, + { + "epoch": 4.754382388911537, + "grad_norm": 1.1285918951034546, + "learning_rate": 0.0004057022013860579, + "loss": 3.1913, + "step": 69975 + }, + { + "epoch": 4.7547221089821985, + "grad_norm": 0.9773391485214233, + "learning_rate": 0.0004056597363772252, + "loss": 3.5257, + "step": 69980 + }, + { + "epoch": 4.755061829052861, + "grad_norm": 0.9700713753700256, + "learning_rate": 0.00040561727136839243, + "loss": 3.4215, + "step": 69985 + }, + { + "epoch": 4.755401549123523, + "grad_norm": 0.8964201211929321, + "learning_rate": 0.0004055748063595597, + "loss": 3.4269, + "step": 69990 + }, + { + "epoch": 4.755741269194184, + "grad_norm": 1.4229077100753784, + "learning_rate": 0.00040553234135072704, + "loss": 3.3095, + "step": 69995 + }, + { + "epoch": 4.756080989264846, + "grad_norm": 1.2170642614364624, + "learning_rate": 0.00040548987634189427, + "loss": 3.5155, + "step": 70000 + }, + { + "epoch": 4.756420709335508, + "grad_norm": 1.007277011871338, + "learning_rate": 0.00040544741133306155, + "loss": 3.8152, + "step": 70005 + }, + { + "epoch": 4.756760429406169, + "grad_norm": 0.7862555384635925, + "learning_rate": 0.0004054049463242289, + "loss": 3.5159, + "step": 70010 + }, + { + "epoch": 4.757100149476831, + "grad_norm": 0.8778907656669617, + "learning_rate": 0.0004053624813153961, + "loss": 3.4717, + "step": 70015 + }, + { + "epoch": 4.757439869547493, + "grad_norm": 0.8126395344734192, + "learning_rate": 0.0004053200163065634, + "loss": 3.4318, + "step": 70020 + }, + { + "epoch": 4.7577795896181545, + "grad_norm": 1.1871123313903809, + "learning_rate": 0.0004052775512977307, + "loss": 3.4304, + "step": 70025 + }, + { + "epoch": 4.758119309688817, + "grad_norm": 1.045343279838562, + "learning_rate": 0.00040523508628889795, + "loss": 3.5022, + "step": 70030 + }, + { + "epoch": 4.758459029759479, + "grad_norm": 0.8970746994018555, + "learning_rate": 0.00040519262128006523, + "loss": 3.6819, + "step": 70035 + }, + { + "epoch": 4.75879874983014, + "grad_norm": 0.8949733972549438, + "learning_rate": 0.0004051501562712325, + "loss": 3.5433, + "step": 70040 + }, + { + "epoch": 4.759138469900802, + "grad_norm": 0.8140722513198853, + "learning_rate": 0.0004051076912623998, + "loss": 3.6664, + "step": 70045 + }, + { + "epoch": 4.759478189971464, + "grad_norm": 0.7248872518539429, + "learning_rate": 0.00040506522625356707, + "loss": 3.075, + "step": 70050 + }, + { + "epoch": 4.759817910042125, + "grad_norm": 0.7907172441482544, + "learning_rate": 0.00040502276124473435, + "loss": 3.6868, + "step": 70055 + }, + { + "epoch": 4.760157630112787, + "grad_norm": 0.7529835104942322, + "learning_rate": 0.00040498029623590163, + "loss": 3.2642, + "step": 70060 + }, + { + "epoch": 4.760497350183449, + "grad_norm": 1.0483533143997192, + "learning_rate": 0.0004049378312270689, + "loss": 3.3425, + "step": 70065 + }, + { + "epoch": 4.7608370702541105, + "grad_norm": 1.021796464920044, + "learning_rate": 0.0004048953662182362, + "loss": 3.7086, + "step": 70070 + }, + { + "epoch": 4.761176790324773, + "grad_norm": 1.0853748321533203, + "learning_rate": 0.0004048529012094034, + "loss": 3.3763, + "step": 70075 + }, + { + "epoch": 4.761516510395434, + "grad_norm": 0.8721354603767395, + "learning_rate": 0.00040481043620057075, + "loss": 3.593, + "step": 70080 + }, + { + "epoch": 4.761856230466096, + "grad_norm": 0.8370148539543152, + "learning_rate": 0.00040476797119173803, + "loss": 3.4034, + "step": 70085 + }, + { + "epoch": 4.762195950536758, + "grad_norm": 0.7664744853973389, + "learning_rate": 0.00040472550618290525, + "loss": 3.0773, + "step": 70090 + }, + { + "epoch": 4.762535670607419, + "grad_norm": 1.2743449211120605, + "learning_rate": 0.0004046830411740726, + "loss": 3.3399, + "step": 70095 + }, + { + "epoch": 4.762875390678081, + "grad_norm": 0.8060517311096191, + "learning_rate": 0.00040464057616523987, + "loss": 3.6062, + "step": 70100 + }, + { + "epoch": 4.763215110748743, + "grad_norm": 0.8377586603164673, + "learning_rate": 0.00040459811115640715, + "loss": 3.4684, + "step": 70105 + }, + { + "epoch": 4.7635548308194045, + "grad_norm": 0.8594706654548645, + "learning_rate": 0.0004045556461475744, + "loss": 3.0675, + "step": 70110 + }, + { + "epoch": 4.7638945508900665, + "grad_norm": 0.9973412156105042, + "learning_rate": 0.0004045131811387417, + "loss": 3.3142, + "step": 70115 + }, + { + "epoch": 4.764234270960729, + "grad_norm": 1.1501307487487793, + "learning_rate": 0.000404470716129909, + "loss": 3.4585, + "step": 70120 + }, + { + "epoch": 4.76457399103139, + "grad_norm": 0.7482250332832336, + "learning_rate": 0.0004044282511210762, + "loss": 3.5328, + "step": 70125 + }, + { + "epoch": 4.764913711102052, + "grad_norm": 0.9305121898651123, + "learning_rate": 0.00040438578611224355, + "loss": 3.3082, + "step": 70130 + }, + { + "epoch": 4.765253431172714, + "grad_norm": 0.8204454779624939, + "learning_rate": 0.00040434332110341083, + "loss": 3.6905, + "step": 70135 + }, + { + "epoch": 4.765593151243375, + "grad_norm": 0.9567694664001465, + "learning_rate": 0.00040430085609457806, + "loss": 3.5189, + "step": 70140 + }, + { + "epoch": 4.765932871314037, + "grad_norm": 0.7904633283615112, + "learning_rate": 0.00040425839108574534, + "loss": 3.4518, + "step": 70145 + }, + { + "epoch": 4.766272591384699, + "grad_norm": 0.8971450328826904, + "learning_rate": 0.00040421592607691267, + "loss": 3.3259, + "step": 70150 + }, + { + "epoch": 4.7666123114553605, + "grad_norm": 0.7756615281105042, + "learning_rate": 0.0004041734610680799, + "loss": 3.3068, + "step": 70155 + }, + { + "epoch": 4.7669520315260225, + "grad_norm": 1.0709162950515747, + "learning_rate": 0.0004041309960592472, + "loss": 3.6885, + "step": 70160 + }, + { + "epoch": 4.767291751596685, + "grad_norm": 0.8551867008209229, + "learning_rate": 0.0004040885310504145, + "loss": 3.6002, + "step": 70165 + }, + { + "epoch": 4.767631471667346, + "grad_norm": 1.1767950057983398, + "learning_rate": 0.00040404606604158174, + "loss": 3.5945, + "step": 70170 + }, + { + "epoch": 4.767971191738008, + "grad_norm": 0.7802050113677979, + "learning_rate": 0.000404003601032749, + "loss": 3.423, + "step": 70175 + }, + { + "epoch": 4.76831091180867, + "grad_norm": 0.8028398156166077, + "learning_rate": 0.0004039611360239163, + "loss": 3.7693, + "step": 70180 + }, + { + "epoch": 4.768650631879331, + "grad_norm": 0.7945271730422974, + "learning_rate": 0.0004039186710150836, + "loss": 3.5705, + "step": 70185 + }, + { + "epoch": 4.768990351949993, + "grad_norm": 0.845060408115387, + "learning_rate": 0.00040387620600625086, + "loss": 3.6422, + "step": 70190 + }, + { + "epoch": 4.769330072020655, + "grad_norm": 0.7944519519805908, + "learning_rate": 0.00040383374099741814, + "loss": 3.4642, + "step": 70195 + }, + { + "epoch": 4.7696697920913165, + "grad_norm": 0.8013132810592651, + "learning_rate": 0.0004037912759885854, + "loss": 3.4119, + "step": 70200 + }, + { + "epoch": 4.7700095121619785, + "grad_norm": 0.776978611946106, + "learning_rate": 0.0004037488109797527, + "loss": 3.5294, + "step": 70205 + }, + { + "epoch": 4.770349232232641, + "grad_norm": 0.7108356952667236, + "learning_rate": 0.00040370634597092, + "loss": 3.4783, + "step": 70210 + }, + { + "epoch": 4.770688952303302, + "grad_norm": 0.8817977905273438, + "learning_rate": 0.0004036638809620872, + "loss": 3.4817, + "step": 70215 + }, + { + "epoch": 4.771028672373964, + "grad_norm": 0.8841917514801025, + "learning_rate": 0.00040362141595325454, + "loss": 3.5253, + "step": 70220 + }, + { + "epoch": 4.771368392444626, + "grad_norm": 0.8310331106185913, + "learning_rate": 0.0004035789509444218, + "loss": 3.4334, + "step": 70225 + }, + { + "epoch": 4.771708112515287, + "grad_norm": 0.8968549370765686, + "learning_rate": 0.00040353648593558904, + "loss": 3.6191, + "step": 70230 + }, + { + "epoch": 4.772047832585949, + "grad_norm": 0.9409182071685791, + "learning_rate": 0.0004034940209267564, + "loss": 3.3376, + "step": 70235 + }, + { + "epoch": 4.772387552656611, + "grad_norm": 0.8187596201896667, + "learning_rate": 0.00040345155591792366, + "loss": 3.2615, + "step": 70240 + }, + { + "epoch": 4.7727272727272725, + "grad_norm": 0.8205277919769287, + "learning_rate": 0.0004034090909090909, + "loss": 3.5039, + "step": 70245 + }, + { + "epoch": 4.773066992797935, + "grad_norm": 0.8254560828208923, + "learning_rate": 0.00040336662590025816, + "loss": 3.4936, + "step": 70250 + }, + { + "epoch": 4.773406712868597, + "grad_norm": 0.8043600916862488, + "learning_rate": 0.0004033241608914255, + "loss": 3.5621, + "step": 70255 + }, + { + "epoch": 4.773746432939258, + "grad_norm": 0.9181062579154968, + "learning_rate": 0.0004032816958825927, + "loss": 3.5296, + "step": 70260 + }, + { + "epoch": 4.77408615300992, + "grad_norm": 0.9614311456680298, + "learning_rate": 0.00040323923087376, + "loss": 3.4175, + "step": 70265 + }, + { + "epoch": 4.774425873080582, + "grad_norm": 1.0629569292068481, + "learning_rate": 0.00040319676586492734, + "loss": 3.4831, + "step": 70270 + }, + { + "epoch": 4.774765593151243, + "grad_norm": 0.7729849815368652, + "learning_rate": 0.0004031543008560946, + "loss": 3.6984, + "step": 70275 + }, + { + "epoch": 4.775105313221905, + "grad_norm": 0.9532772302627563, + "learning_rate": 0.00040311183584726184, + "loss": 3.3816, + "step": 70280 + }, + { + "epoch": 4.775445033292567, + "grad_norm": 0.7529994249343872, + "learning_rate": 0.0004030693708384291, + "loss": 3.7698, + "step": 70285 + }, + { + "epoch": 4.7757847533632285, + "grad_norm": 0.8031511306762695, + "learning_rate": 0.00040302690582959646, + "loss": 3.5054, + "step": 70290 + }, + { + "epoch": 4.776124473433891, + "grad_norm": 0.9469435214996338, + "learning_rate": 0.0004029844408207637, + "loss": 3.626, + "step": 70295 + }, + { + "epoch": 4.776464193504552, + "grad_norm": 1.2148302793502808, + "learning_rate": 0.00040294197581193096, + "loss": 3.6636, + "step": 70300 + }, + { + "epoch": 4.776803913575214, + "grad_norm": 1.1191601753234863, + "learning_rate": 0.0004028995108030983, + "loss": 3.7747, + "step": 70305 + }, + { + "epoch": 4.777143633645876, + "grad_norm": 1.1783802509307861, + "learning_rate": 0.0004028570457942655, + "loss": 3.3707, + "step": 70310 + }, + { + "epoch": 4.777483353716537, + "grad_norm": 0.8672983646392822, + "learning_rate": 0.0004028145807854328, + "loss": 3.4294, + "step": 70315 + }, + { + "epoch": 4.777823073787199, + "grad_norm": 0.7727931141853333, + "learning_rate": 0.00040277211577660014, + "loss": 3.5373, + "step": 70320 + }, + { + "epoch": 4.778162793857861, + "grad_norm": 0.8452978134155273, + "learning_rate": 0.00040272965076776736, + "loss": 3.426, + "step": 70325 + }, + { + "epoch": 4.778502513928522, + "grad_norm": 0.9340426325798035, + "learning_rate": 0.00040268718575893464, + "loss": 3.3597, + "step": 70330 + }, + { + "epoch": 4.7788422339991845, + "grad_norm": 0.8888667821884155, + "learning_rate": 0.0004026447207501019, + "loss": 3.473, + "step": 70335 + }, + { + "epoch": 4.779181954069847, + "grad_norm": 0.8960254788398743, + "learning_rate": 0.0004026022557412692, + "loss": 3.3074, + "step": 70340 + }, + { + "epoch": 4.779521674140508, + "grad_norm": 0.8607752919197083, + "learning_rate": 0.0004025597907324365, + "loss": 3.4672, + "step": 70345 + }, + { + "epoch": 4.77986139421117, + "grad_norm": 0.9392121434211731, + "learning_rate": 0.00040251732572360376, + "loss": 3.2868, + "step": 70350 + }, + { + "epoch": 4.780201114281832, + "grad_norm": 0.8517521619796753, + "learning_rate": 0.00040247486071477104, + "loss": 3.7123, + "step": 70355 + }, + { + "epoch": 4.780540834352493, + "grad_norm": 0.6652432084083557, + "learning_rate": 0.0004024323957059383, + "loss": 3.5565, + "step": 70360 + }, + { + "epoch": 4.780880554423155, + "grad_norm": 1.2006573677062988, + "learning_rate": 0.0004023899306971056, + "loss": 3.2623, + "step": 70365 + }, + { + "epoch": 4.781220274493817, + "grad_norm": 0.7220340967178345, + "learning_rate": 0.00040234746568827283, + "loss": 3.6234, + "step": 70370 + }, + { + "epoch": 4.7815599945644784, + "grad_norm": 0.881195604801178, + "learning_rate": 0.00040230500067944016, + "loss": 3.57, + "step": 70375 + }, + { + "epoch": 4.7818997146351405, + "grad_norm": 1.1372443437576294, + "learning_rate": 0.00040226253567060745, + "loss": 3.4509, + "step": 70380 + }, + { + "epoch": 4.782239434705803, + "grad_norm": 0.8250408172607422, + "learning_rate": 0.00040222007066177467, + "loss": 3.671, + "step": 70385 + }, + { + "epoch": 4.782579154776464, + "grad_norm": 0.8903779983520508, + "learning_rate": 0.000402177605652942, + "loss": 3.4131, + "step": 70390 + }, + { + "epoch": 4.782918874847126, + "grad_norm": 0.9633232355117798, + "learning_rate": 0.0004021351406441093, + "loss": 3.4399, + "step": 70395 + }, + { + "epoch": 4.783258594917788, + "grad_norm": 0.8965213298797607, + "learning_rate": 0.0004020926756352765, + "loss": 3.2851, + "step": 70400 + }, + { + "epoch": 4.783598314988449, + "grad_norm": 0.9689022898674011, + "learning_rate": 0.0004020502106264438, + "loss": 3.6069, + "step": 70405 + }, + { + "epoch": 4.783938035059111, + "grad_norm": 0.8349127173423767, + "learning_rate": 0.0004020077456176111, + "loss": 3.6365, + "step": 70410 + }, + { + "epoch": 4.784277755129773, + "grad_norm": 0.909548282623291, + "learning_rate": 0.00040196528060877835, + "loss": 3.7422, + "step": 70415 + }, + { + "epoch": 4.7846174752004345, + "grad_norm": 0.9999193549156189, + "learning_rate": 0.00040192281559994563, + "loss": 3.4312, + "step": 70420 + }, + { + "epoch": 4.7849571952710965, + "grad_norm": 0.8660966157913208, + "learning_rate": 0.00040188035059111297, + "loss": 3.2884, + "step": 70425 + }, + { + "epoch": 4.785296915341759, + "grad_norm": 0.7895473837852478, + "learning_rate": 0.0004018378855822802, + "loss": 3.6481, + "step": 70430 + }, + { + "epoch": 4.78563663541242, + "grad_norm": 0.7956475615501404, + "learning_rate": 0.00040179542057344747, + "loss": 3.4492, + "step": 70435 + }, + { + "epoch": 4.785976355483082, + "grad_norm": 1.088138461112976, + "learning_rate": 0.00040175295556461475, + "loss": 3.3027, + "step": 70440 + }, + { + "epoch": 4.786316075553744, + "grad_norm": 0.9060842394828796, + "learning_rate": 0.0004017104905557821, + "loss": 3.4391, + "step": 70445 + }, + { + "epoch": 4.786655795624405, + "grad_norm": 0.8917579054832458, + "learning_rate": 0.0004016680255469493, + "loss": 3.5534, + "step": 70450 + }, + { + "epoch": 4.786995515695067, + "grad_norm": 0.8706673979759216, + "learning_rate": 0.0004016255605381166, + "loss": 3.244, + "step": 70455 + }, + { + "epoch": 4.787335235765729, + "grad_norm": 0.8494575023651123, + "learning_rate": 0.0004015830955292839, + "loss": 3.4548, + "step": 70460 + }, + { + "epoch": 4.7876749558363905, + "grad_norm": 0.882722020149231, + "learning_rate": 0.00040154063052045115, + "loss": 3.4462, + "step": 70465 + }, + { + "epoch": 4.7880146759070525, + "grad_norm": 1.0654890537261963, + "learning_rate": 0.00040149816551161843, + "loss": 3.6834, + "step": 70470 + }, + { + "epoch": 4.788354395977715, + "grad_norm": 0.9525165557861328, + "learning_rate": 0.0004014557005027857, + "loss": 3.5394, + "step": 70475 + }, + { + "epoch": 4.788694116048376, + "grad_norm": 0.8042345643043518, + "learning_rate": 0.000401413235493953, + "loss": 3.3865, + "step": 70480 + }, + { + "epoch": 4.789033836119038, + "grad_norm": 1.1183325052261353, + "learning_rate": 0.00040137077048512027, + "loss": 3.434, + "step": 70485 + }, + { + "epoch": 4.7893735561897, + "grad_norm": 0.7519111633300781, + "learning_rate": 0.00040132830547628755, + "loss": 3.3815, + "step": 70490 + }, + { + "epoch": 4.789713276260361, + "grad_norm": 0.8972079753875732, + "learning_rate": 0.00040128584046745483, + "loss": 3.4321, + "step": 70495 + }, + { + "epoch": 4.790052996331023, + "grad_norm": 1.0112940073013306, + "learning_rate": 0.0004012433754586221, + "loss": 3.8096, + "step": 70500 + }, + { + "epoch": 4.790392716401685, + "grad_norm": 0.7080385684967041, + "learning_rate": 0.0004012009104497894, + "loss": 3.3596, + "step": 70505 + }, + { + "epoch": 4.7907324364723465, + "grad_norm": 0.8304780125617981, + "learning_rate": 0.0004011584454409566, + "loss": 3.4605, + "step": 70510 + }, + { + "epoch": 4.7910721565430086, + "grad_norm": 0.8705866932868958, + "learning_rate": 0.00040111598043212395, + "loss": 3.3854, + "step": 70515 + }, + { + "epoch": 4.791411876613671, + "grad_norm": 1.3136351108551025, + "learning_rate": 0.00040107351542329123, + "loss": 3.7872, + "step": 70520 + }, + { + "epoch": 4.791751596684332, + "grad_norm": 0.761170506477356, + "learning_rate": 0.00040103105041445846, + "loss": 3.4968, + "step": 70525 + }, + { + "epoch": 4.792091316754994, + "grad_norm": 1.1485087871551514, + "learning_rate": 0.0004009885854056258, + "loss": 3.3972, + "step": 70530 + }, + { + "epoch": 4.792431036825656, + "grad_norm": 0.9570496082305908, + "learning_rate": 0.0004009461203967931, + "loss": 3.582, + "step": 70535 + }, + { + "epoch": 4.792770756896317, + "grad_norm": 0.9269013404846191, + "learning_rate": 0.0004009036553879603, + "loss": 3.3016, + "step": 70540 + }, + { + "epoch": 4.793110476966979, + "grad_norm": 1.1033341884613037, + "learning_rate": 0.0004008611903791276, + "loss": 3.3826, + "step": 70545 + }, + { + "epoch": 4.793450197037641, + "grad_norm": 1.1749680042266846, + "learning_rate": 0.0004008187253702949, + "loss": 3.5565, + "step": 70550 + }, + { + "epoch": 4.7937899171083025, + "grad_norm": 0.8548731207847595, + "learning_rate": 0.00040077626036146214, + "loss": 3.6999, + "step": 70555 + }, + { + "epoch": 4.794129637178965, + "grad_norm": 0.7880840301513672, + "learning_rate": 0.0004007337953526294, + "loss": 3.7293, + "step": 70560 + }, + { + "epoch": 4.794469357249627, + "grad_norm": 1.099001169204712, + "learning_rate": 0.00040069133034379675, + "loss": 3.5774, + "step": 70565 + }, + { + "epoch": 4.794809077320288, + "grad_norm": 1.1060079336166382, + "learning_rate": 0.000400648865334964, + "loss": 3.7722, + "step": 70570 + }, + { + "epoch": 4.79514879739095, + "grad_norm": 1.2185269594192505, + "learning_rate": 0.00040060640032613126, + "loss": 3.3087, + "step": 70575 + }, + { + "epoch": 4.795488517461612, + "grad_norm": 0.9283305406570435, + "learning_rate": 0.00040056393531729854, + "loss": 3.403, + "step": 70580 + }, + { + "epoch": 4.795828237532273, + "grad_norm": 0.7855352759361267, + "learning_rate": 0.0004005214703084658, + "loss": 3.3949, + "step": 70585 + }, + { + "epoch": 4.796167957602935, + "grad_norm": 0.8654841184616089, + "learning_rate": 0.0004004790052996331, + "loss": 3.5531, + "step": 70590 + }, + { + "epoch": 4.796507677673597, + "grad_norm": 1.021836519241333, + "learning_rate": 0.0004004365402908004, + "loss": 3.3865, + "step": 70595 + }, + { + "epoch": 4.7968473977442585, + "grad_norm": 0.8132767081260681, + "learning_rate": 0.00040039407528196766, + "loss": 3.4587, + "step": 70600 + }, + { + "epoch": 4.797187117814921, + "grad_norm": 0.8168739080429077, + "learning_rate": 0.00040035161027313494, + "loss": 3.57, + "step": 70605 + }, + { + "epoch": 4.797526837885583, + "grad_norm": 0.7915118336677551, + "learning_rate": 0.0004003091452643022, + "loss": 3.5561, + "step": 70610 + }, + { + "epoch": 4.797866557956244, + "grad_norm": 0.8955094218254089, + "learning_rate": 0.00040026668025546955, + "loss": 3.3737, + "step": 70615 + }, + { + "epoch": 4.798206278026906, + "grad_norm": 0.7943336963653564, + "learning_rate": 0.0004002242152466368, + "loss": 3.6874, + "step": 70620 + }, + { + "epoch": 4.798545998097568, + "grad_norm": 1.0302733182907104, + "learning_rate": 0.00040018175023780406, + "loss": 3.5091, + "step": 70625 + }, + { + "epoch": 4.798885718168229, + "grad_norm": 0.8550623655319214, + "learning_rate": 0.00040013928522897134, + "loss": 3.2809, + "step": 70630 + }, + { + "epoch": 4.799225438238891, + "grad_norm": 0.8888876438140869, + "learning_rate": 0.0004000968202201386, + "loss": 3.4225, + "step": 70635 + }, + { + "epoch": 4.799565158309553, + "grad_norm": 0.9619299173355103, + "learning_rate": 0.0004000543552113059, + "loss": 3.4654, + "step": 70640 + }, + { + "epoch": 4.7999048783802145, + "grad_norm": 0.9343862533569336, + "learning_rate": 0.0004000118902024732, + "loss": 3.2595, + "step": 70645 + }, + { + "epoch": 4.800244598450877, + "grad_norm": 0.8767639398574829, + "learning_rate": 0.00039996942519364046, + "loss": 3.4761, + "step": 70650 + }, + { + "epoch": 4.800584318521539, + "grad_norm": 1.057349681854248, + "learning_rate": 0.00039992696018480774, + "loss": 3.5013, + "step": 70655 + }, + { + "epoch": 4.8009240385922, + "grad_norm": 0.8815264701843262, + "learning_rate": 0.000399884495175975, + "loss": 3.6289, + "step": 70660 + }, + { + "epoch": 4.801263758662862, + "grad_norm": 0.7199167013168335, + "learning_rate": 0.00039984203016714225, + "loss": 3.6424, + "step": 70665 + }, + { + "epoch": 4.801603478733524, + "grad_norm": 0.9723081588745117, + "learning_rate": 0.0003997995651583096, + "loss": 3.2603, + "step": 70670 + }, + { + "epoch": 4.801943198804185, + "grad_norm": 0.8908737301826477, + "learning_rate": 0.00039975710014947686, + "loss": 3.8391, + "step": 70675 + }, + { + "epoch": 4.802282918874847, + "grad_norm": 0.818849503993988, + "learning_rate": 0.0003997146351406441, + "loss": 3.497, + "step": 70680 + }, + { + "epoch": 4.802622638945509, + "grad_norm": 0.8584309816360474, + "learning_rate": 0.0003996721701318114, + "loss": 3.4156, + "step": 70685 + }, + { + "epoch": 4.8029623590161705, + "grad_norm": 1.2230474948883057, + "learning_rate": 0.0003996297051229787, + "loss": 3.5136, + "step": 70690 + }, + { + "epoch": 4.803302079086833, + "grad_norm": 1.0068387985229492, + "learning_rate": 0.00039958724011414593, + "loss": 3.6272, + "step": 70695 + }, + { + "epoch": 4.803641799157495, + "grad_norm": 0.9747718572616577, + "learning_rate": 0.0003995447751053132, + "loss": 3.3668, + "step": 70700 + }, + { + "epoch": 4.803981519228156, + "grad_norm": 1.0465805530548096, + "learning_rate": 0.00039950231009648054, + "loss": 3.439, + "step": 70705 + }, + { + "epoch": 4.804321239298818, + "grad_norm": 0.8880668878555298, + "learning_rate": 0.00039945984508764777, + "loss": 3.5246, + "step": 70710 + }, + { + "epoch": 4.80466095936948, + "grad_norm": 0.903103768825531, + "learning_rate": 0.00039941738007881505, + "loss": 3.4039, + "step": 70715 + }, + { + "epoch": 4.805000679440141, + "grad_norm": 0.6945569515228271, + "learning_rate": 0.0003993749150699824, + "loss": 3.3442, + "step": 70720 + }, + { + "epoch": 4.805340399510803, + "grad_norm": 0.8089251518249512, + "learning_rate": 0.0003993324500611496, + "loss": 3.2967, + "step": 70725 + }, + { + "epoch": 4.805680119581465, + "grad_norm": 0.8446237444877625, + "learning_rate": 0.0003992899850523169, + "loss": 3.5811, + "step": 70730 + }, + { + "epoch": 4.8060198396521265, + "grad_norm": 0.9648834466934204, + "learning_rate": 0.00039924752004348417, + "loss": 3.5093, + "step": 70735 + }, + { + "epoch": 4.806359559722789, + "grad_norm": 0.8614630103111267, + "learning_rate": 0.00039920505503465145, + "loss": 3.4752, + "step": 70740 + }, + { + "epoch": 4.806699279793451, + "grad_norm": 1.0510776042938232, + "learning_rate": 0.00039916259002581873, + "loss": 3.587, + "step": 70745 + }, + { + "epoch": 4.807038999864112, + "grad_norm": 0.958268404006958, + "learning_rate": 0.000399120125016986, + "loss": 3.6617, + "step": 70750 + }, + { + "epoch": 4.807378719934774, + "grad_norm": 0.7463881373405457, + "learning_rate": 0.0003990776600081533, + "loss": 3.523, + "step": 70755 + }, + { + "epoch": 4.807718440005435, + "grad_norm": 0.9886858463287354, + "learning_rate": 0.00039903519499932057, + "loss": 3.3387, + "step": 70760 + }, + { + "epoch": 4.808058160076097, + "grad_norm": 0.7606908082962036, + "learning_rate": 0.00039899272999048785, + "loss": 3.6394, + "step": 70765 + }, + { + "epoch": 4.808397880146759, + "grad_norm": 1.026638388633728, + "learning_rate": 0.0003989502649816551, + "loss": 3.5217, + "step": 70770 + }, + { + "epoch": 4.8087376002174205, + "grad_norm": 0.836662769317627, + "learning_rate": 0.0003989077999728224, + "loss": 3.5468, + "step": 70775 + }, + { + "epoch": 4.8090773202880825, + "grad_norm": 0.7752577662467957, + "learning_rate": 0.0003988653349639897, + "loss": 3.2342, + "step": 70780 + }, + { + "epoch": 4.809417040358745, + "grad_norm": 0.6935116052627563, + "learning_rate": 0.00039882286995515697, + "loss": 3.4789, + "step": 70785 + }, + { + "epoch": 4.809756760429406, + "grad_norm": 0.9013590216636658, + "learning_rate": 0.00039878040494632425, + "loss": 3.3288, + "step": 70790 + }, + { + "epoch": 4.810096480500068, + "grad_norm": 0.9464393854141235, + "learning_rate": 0.00039873793993749153, + "loss": 3.3385, + "step": 70795 + }, + { + "epoch": 4.81043620057073, + "grad_norm": 0.879513680934906, + "learning_rate": 0.0003986954749286588, + "loss": 3.4001, + "step": 70800 + }, + { + "epoch": 4.810775920641391, + "grad_norm": 0.8569079041481018, + "learning_rate": 0.00039865300991982603, + "loss": 3.4249, + "step": 70805 + }, + { + "epoch": 4.811115640712053, + "grad_norm": 0.841816782951355, + "learning_rate": 0.00039861054491099337, + "loss": 3.6598, + "step": 70810 + }, + { + "epoch": 4.811455360782715, + "grad_norm": 0.9392424821853638, + "learning_rate": 0.00039856807990216065, + "loss": 3.7141, + "step": 70815 + }, + { + "epoch": 4.8117950808533765, + "grad_norm": 1.0502564907073975, + "learning_rate": 0.0003985256148933279, + "loss": 3.435, + "step": 70820 + }, + { + "epoch": 4.812134800924039, + "grad_norm": 0.8750007152557373, + "learning_rate": 0.0003984831498844952, + "loss": 3.572, + "step": 70825 + }, + { + "epoch": 4.812474520994701, + "grad_norm": 0.9074890613555908, + "learning_rate": 0.0003984406848756625, + "loss": 3.5931, + "step": 70830 + }, + { + "epoch": 4.812814241065362, + "grad_norm": 1.1666885614395142, + "learning_rate": 0.0003983982198668297, + "loss": 3.5679, + "step": 70835 + }, + { + "epoch": 4.813153961136024, + "grad_norm": 1.1100159883499146, + "learning_rate": 0.000398355754857997, + "loss": 3.5664, + "step": 70840 + }, + { + "epoch": 4.813493681206686, + "grad_norm": 1.0145869255065918, + "learning_rate": 0.00039831328984916433, + "loss": 4.0115, + "step": 70845 + }, + { + "epoch": 4.813833401277347, + "grad_norm": 0.7869913578033447, + "learning_rate": 0.00039827082484033156, + "loss": 3.4886, + "step": 70850 + }, + { + "epoch": 4.814173121348009, + "grad_norm": 0.8371909856796265, + "learning_rate": 0.00039822835983149884, + "loss": 3.7582, + "step": 70855 + }, + { + "epoch": 4.814512841418671, + "grad_norm": 0.6600121259689331, + "learning_rate": 0.00039818589482266617, + "loss": 3.3642, + "step": 70860 + }, + { + "epoch": 4.8148525614893325, + "grad_norm": 0.9040317535400391, + "learning_rate": 0.0003981434298138334, + "loss": 3.4999, + "step": 70865 + }, + { + "epoch": 4.815192281559995, + "grad_norm": 0.8087959289550781, + "learning_rate": 0.0003981009648050007, + "loss": 3.3932, + "step": 70870 + }, + { + "epoch": 4.815532001630657, + "grad_norm": 0.823777437210083, + "learning_rate": 0.000398058499796168, + "loss": 3.375, + "step": 70875 + }, + { + "epoch": 4.815871721701318, + "grad_norm": 1.1131092309951782, + "learning_rate": 0.00039801603478733524, + "loss": 3.7234, + "step": 70880 + }, + { + "epoch": 4.81621144177198, + "grad_norm": 0.7496495246887207, + "learning_rate": 0.0003979735697785025, + "loss": 3.4625, + "step": 70885 + }, + { + "epoch": 4.816551161842642, + "grad_norm": 1.0396727323532104, + "learning_rate": 0.0003979311047696698, + "loss": 3.5946, + "step": 70890 + }, + { + "epoch": 4.816890881913303, + "grad_norm": 0.7298014760017395, + "learning_rate": 0.0003978886397608371, + "loss": 3.6737, + "step": 70895 + }, + { + "epoch": 4.817230601983965, + "grad_norm": 0.9425898790359497, + "learning_rate": 0.00039784617475200436, + "loss": 3.6772, + "step": 70900 + }, + { + "epoch": 4.817570322054627, + "grad_norm": 0.9819394946098328, + "learning_rate": 0.00039780370974317164, + "loss": 3.7238, + "step": 70905 + }, + { + "epoch": 4.8179100421252885, + "grad_norm": 0.9634224772453308, + "learning_rate": 0.0003977612447343389, + "loss": 3.7637, + "step": 70910 + }, + { + "epoch": 4.818249762195951, + "grad_norm": 1.0291748046875, + "learning_rate": 0.0003977187797255062, + "loss": 3.4317, + "step": 70915 + }, + { + "epoch": 4.818589482266613, + "grad_norm": 0.7748332619667053, + "learning_rate": 0.0003976763147166735, + "loss": 3.2661, + "step": 70920 + }, + { + "epoch": 4.818929202337274, + "grad_norm": 0.9932260513305664, + "learning_rate": 0.0003976338497078407, + "loss": 3.4519, + "step": 70925 + }, + { + "epoch": 4.819268922407936, + "grad_norm": 0.7910274267196655, + "learning_rate": 0.00039759138469900804, + "loss": 3.4691, + "step": 70930 + }, + { + "epoch": 4.819608642478598, + "grad_norm": 0.9375197887420654, + "learning_rate": 0.0003975489196901753, + "loss": 3.4805, + "step": 70935 + }, + { + "epoch": 4.819948362549259, + "grad_norm": 0.7566654086112976, + "learning_rate": 0.00039750645468134254, + "loss": 3.6637, + "step": 70940 + }, + { + "epoch": 4.820288082619921, + "grad_norm": 0.8344612121582031, + "learning_rate": 0.0003974639896725099, + "loss": 3.7541, + "step": 70945 + }, + { + "epoch": 4.820627802690583, + "grad_norm": 0.73778235912323, + "learning_rate": 0.00039742152466367716, + "loss": 3.4652, + "step": 70950 + }, + { + "epoch": 4.8209675227612445, + "grad_norm": 0.7157381176948547, + "learning_rate": 0.00039737905965484444, + "loss": 3.3576, + "step": 70955 + }, + { + "epoch": 4.821307242831907, + "grad_norm": 0.7827205061912537, + "learning_rate": 0.00039733659464601166, + "loss": 3.5221, + "step": 70960 + }, + { + "epoch": 4.821646962902569, + "grad_norm": 0.8035179376602173, + "learning_rate": 0.000397294129637179, + "loss": 3.6959, + "step": 70965 + }, + { + "epoch": 4.82198668297323, + "grad_norm": 0.833808183670044, + "learning_rate": 0.0003972516646283463, + "loss": 3.6502, + "step": 70970 + }, + { + "epoch": 4.822326403043892, + "grad_norm": 0.764222264289856, + "learning_rate": 0.0003972091996195135, + "loss": 3.3818, + "step": 70975 + }, + { + "epoch": 4.822666123114553, + "grad_norm": 0.7710103392601013, + "learning_rate": 0.00039716673461068084, + "loss": 3.2253, + "step": 70980 + }, + { + "epoch": 4.823005843185215, + "grad_norm": 0.9523244500160217, + "learning_rate": 0.0003971242696018481, + "loss": 3.513, + "step": 70985 + }, + { + "epoch": 4.823345563255877, + "grad_norm": 0.9166908860206604, + "learning_rate": 0.00039708180459301534, + "loss": 3.4102, + "step": 70990 + }, + { + "epoch": 4.8236852833265385, + "grad_norm": 1.000309705734253, + "learning_rate": 0.0003970393395841826, + "loss": 3.5204, + "step": 70995 + }, + { + "epoch": 4.8240250033972005, + "grad_norm": 1.0685782432556152, + "learning_rate": 0.00039699687457534996, + "loss": 3.5553, + "step": 71000 + }, + { + "epoch": 4.824364723467863, + "grad_norm": 0.7866259217262268, + "learning_rate": 0.0003969544095665172, + "loss": 3.4962, + "step": 71005 + }, + { + "epoch": 4.824704443538524, + "grad_norm": 1.2047789096832275, + "learning_rate": 0.00039691194455768446, + "loss": 3.5054, + "step": 71010 + }, + { + "epoch": 4.825044163609186, + "grad_norm": 0.8946297764778137, + "learning_rate": 0.0003968694795488518, + "loss": 2.9987, + "step": 71015 + }, + { + "epoch": 4.825383883679848, + "grad_norm": 0.7953545451164246, + "learning_rate": 0.000396827014540019, + "loss": 3.3606, + "step": 71020 + }, + { + "epoch": 4.825723603750509, + "grad_norm": 1.0809776782989502, + "learning_rate": 0.0003967845495311863, + "loss": 3.5998, + "step": 71025 + }, + { + "epoch": 4.826063323821171, + "grad_norm": 0.9757300019264221, + "learning_rate": 0.0003967420845223536, + "loss": 3.6223, + "step": 71030 + }, + { + "epoch": 4.826403043891833, + "grad_norm": 0.9193054437637329, + "learning_rate": 0.00039669961951352086, + "loss": 3.574, + "step": 71035 + }, + { + "epoch": 4.8267427639624945, + "grad_norm": 1.0651136636734009, + "learning_rate": 0.00039665715450468814, + "loss": 3.3917, + "step": 71040 + }, + { + "epoch": 4.8270824840331565, + "grad_norm": 0.9391628503799438, + "learning_rate": 0.0003966146894958554, + "loss": 3.3141, + "step": 71045 + }, + { + "epoch": 4.827422204103819, + "grad_norm": 0.8234797120094299, + "learning_rate": 0.0003965722244870227, + "loss": 3.6276, + "step": 71050 + }, + { + "epoch": 4.82776192417448, + "grad_norm": 0.9170812368392944, + "learning_rate": 0.00039652975947819, + "loss": 3.4826, + "step": 71055 + }, + { + "epoch": 4.828101644245142, + "grad_norm": 0.9189901947975159, + "learning_rate": 0.00039648729446935726, + "loss": 3.5287, + "step": 71060 + }, + { + "epoch": 4.828441364315804, + "grad_norm": 0.9926113486289978, + "learning_rate": 0.0003964448294605245, + "loss": 3.6459, + "step": 71065 + }, + { + "epoch": 4.828781084386465, + "grad_norm": 0.9156104922294617, + "learning_rate": 0.0003964023644516918, + "loss": 3.4621, + "step": 71070 + }, + { + "epoch": 4.829120804457127, + "grad_norm": 0.9316727519035339, + "learning_rate": 0.0003963598994428591, + "loss": 3.4291, + "step": 71075 + }, + { + "epoch": 4.829460524527789, + "grad_norm": 0.8657076954841614, + "learning_rate": 0.00039631743443402633, + "loss": 3.4678, + "step": 71080 + }, + { + "epoch": 4.8298002445984505, + "grad_norm": 0.802844226360321, + "learning_rate": 0.00039627496942519366, + "loss": 3.6113, + "step": 71085 + }, + { + "epoch": 4.8301399646691126, + "grad_norm": 0.7883496880531311, + "learning_rate": 0.00039623250441636094, + "loss": 3.6489, + "step": 71090 + }, + { + "epoch": 4.830479684739775, + "grad_norm": 1.0074280500411987, + "learning_rate": 0.00039619003940752817, + "loss": 3.4418, + "step": 71095 + }, + { + "epoch": 4.830819404810436, + "grad_norm": 0.822624921798706, + "learning_rate": 0.00039614757439869545, + "loss": 3.5269, + "step": 71100 + }, + { + "epoch": 4.831159124881098, + "grad_norm": 1.05500328540802, + "learning_rate": 0.0003961051093898628, + "loss": 3.449, + "step": 71105 + }, + { + "epoch": 4.83149884495176, + "grad_norm": 0.9108061790466309, + "learning_rate": 0.00039606264438103, + "loss": 3.5857, + "step": 71110 + }, + { + "epoch": 4.831838565022421, + "grad_norm": 0.928350031375885, + "learning_rate": 0.0003960201793721973, + "loss": 3.5903, + "step": 71115 + }, + { + "epoch": 4.832178285093083, + "grad_norm": 0.7900652289390564, + "learning_rate": 0.0003959777143633646, + "loss": 3.327, + "step": 71120 + }, + { + "epoch": 4.832518005163745, + "grad_norm": 0.932620644569397, + "learning_rate": 0.0003959352493545319, + "loss": 3.5904, + "step": 71125 + }, + { + "epoch": 4.8328577252344065, + "grad_norm": 0.8778575658798218, + "learning_rate": 0.00039589278434569913, + "loss": 3.3157, + "step": 71130 + }, + { + "epoch": 4.833197445305069, + "grad_norm": 0.755192756652832, + "learning_rate": 0.0003958503193368664, + "loss": 3.6799, + "step": 71135 + }, + { + "epoch": 4.833537165375731, + "grad_norm": 0.8667948842048645, + "learning_rate": 0.00039580785432803375, + "loss": 3.2194, + "step": 71140 + }, + { + "epoch": 4.833876885446392, + "grad_norm": 2.013303518295288, + "learning_rate": 0.00039576538931920097, + "loss": 3.4203, + "step": 71145 + }, + { + "epoch": 4.834216605517054, + "grad_norm": 0.7738672494888306, + "learning_rate": 0.00039572292431036825, + "loss": 3.5796, + "step": 71150 + }, + { + "epoch": 4.834556325587716, + "grad_norm": 1.3464653491973877, + "learning_rate": 0.0003956804593015356, + "loss": 3.4588, + "step": 71155 + }, + { + "epoch": 4.834896045658377, + "grad_norm": 0.9314467906951904, + "learning_rate": 0.0003956379942927028, + "loss": 3.3887, + "step": 71160 + }, + { + "epoch": 4.835235765729039, + "grad_norm": 0.963745653629303, + "learning_rate": 0.0003955955292838701, + "loss": 3.6891, + "step": 71165 + }, + { + "epoch": 4.835575485799701, + "grad_norm": 0.7535337209701538, + "learning_rate": 0.0003955530642750374, + "loss": 3.5377, + "step": 71170 + }, + { + "epoch": 4.8359152058703625, + "grad_norm": 1.0974199771881104, + "learning_rate": 0.00039551059926620465, + "loss": 3.3874, + "step": 71175 + }, + { + "epoch": 4.836254925941025, + "grad_norm": 0.6512888669967651, + "learning_rate": 0.00039546813425737193, + "loss": 3.4582, + "step": 71180 + }, + { + "epoch": 4.836594646011687, + "grad_norm": 1.0083938837051392, + "learning_rate": 0.0003954256692485392, + "loss": 3.5647, + "step": 71185 + }, + { + "epoch": 4.836934366082348, + "grad_norm": 0.8498887419700623, + "learning_rate": 0.0003953832042397065, + "loss": 3.3454, + "step": 71190 + }, + { + "epoch": 4.83727408615301, + "grad_norm": 0.9310884475708008, + "learning_rate": 0.00039534073923087377, + "loss": 3.4524, + "step": 71195 + }, + { + "epoch": 4.837613806223672, + "grad_norm": 0.9496724009513855, + "learning_rate": 0.00039529827422204105, + "loss": 3.4226, + "step": 71200 + }, + { + "epoch": 4.837953526294333, + "grad_norm": 1.0194867849349976, + "learning_rate": 0.00039525580921320833, + "loss": 3.5058, + "step": 71205 + }, + { + "epoch": 4.838293246364995, + "grad_norm": 1.0465986728668213, + "learning_rate": 0.0003952133442043756, + "loss": 3.5605, + "step": 71210 + }, + { + "epoch": 4.838632966435657, + "grad_norm": 0.9374005794525146, + "learning_rate": 0.0003951708791955429, + "loss": 3.5367, + "step": 71215 + }, + { + "epoch": 4.8389726865063185, + "grad_norm": 0.9078034162521362, + "learning_rate": 0.0003951284141867101, + "loss": 3.4267, + "step": 71220 + }, + { + "epoch": 4.839312406576981, + "grad_norm": 0.750032365322113, + "learning_rate": 0.00039508594917787745, + "loss": 3.3284, + "step": 71225 + }, + { + "epoch": 4.839652126647643, + "grad_norm": 0.771081805229187, + "learning_rate": 0.00039504348416904473, + "loss": 3.1585, + "step": 71230 + }, + { + "epoch": 4.839991846718304, + "grad_norm": 0.9651888012886047, + "learning_rate": 0.00039500101916021196, + "loss": 3.4434, + "step": 71235 + }, + { + "epoch": 4.840331566788966, + "grad_norm": 0.8739444613456726, + "learning_rate": 0.0003949585541513793, + "loss": 3.6145, + "step": 71240 + }, + { + "epoch": 4.840671286859628, + "grad_norm": 1.0616283416748047, + "learning_rate": 0.0003949160891425466, + "loss": 3.3929, + "step": 71245 + }, + { + "epoch": 4.841011006930289, + "grad_norm": 0.7762076258659363, + "learning_rate": 0.0003948736241337138, + "loss": 3.5844, + "step": 71250 + }, + { + "epoch": 4.841350727000951, + "grad_norm": 0.8426217436790466, + "learning_rate": 0.0003948311591248811, + "loss": 3.5423, + "step": 71255 + }, + { + "epoch": 4.841690447071613, + "grad_norm": 0.8437306880950928, + "learning_rate": 0.0003947886941160484, + "loss": 3.4296, + "step": 71260 + }, + { + "epoch": 4.8420301671422745, + "grad_norm": 0.9964501857757568, + "learning_rate": 0.00039474622910721564, + "loss": 3.5089, + "step": 71265 + }, + { + "epoch": 4.842369887212937, + "grad_norm": 0.7669007182121277, + "learning_rate": 0.0003947037640983829, + "loss": 3.3466, + "step": 71270 + }, + { + "epoch": 4.842709607283599, + "grad_norm": 0.8554624915122986, + "learning_rate": 0.00039466129908955025, + "loss": 3.5947, + "step": 71275 + }, + { + "epoch": 4.84304932735426, + "grad_norm": 0.8738998770713806, + "learning_rate": 0.0003946188340807175, + "loss": 3.5015, + "step": 71280 + }, + { + "epoch": 4.843389047424922, + "grad_norm": 0.9077485203742981, + "learning_rate": 0.00039457636907188476, + "loss": 3.4556, + "step": 71285 + }, + { + "epoch": 4.843728767495584, + "grad_norm": 0.9887559413909912, + "learning_rate": 0.00039453390406305204, + "loss": 3.4521, + "step": 71290 + }, + { + "epoch": 4.844068487566245, + "grad_norm": 0.8477239608764648, + "learning_rate": 0.0003944914390542194, + "loss": 3.4471, + "step": 71295 + }, + { + "epoch": 4.844408207636907, + "grad_norm": 0.7347681522369385, + "learning_rate": 0.0003944489740453866, + "loss": 3.2089, + "step": 71300 + }, + { + "epoch": 4.844747927707569, + "grad_norm": 0.6829935908317566, + "learning_rate": 0.0003944065090365539, + "loss": 3.637, + "step": 71305 + }, + { + "epoch": 4.8450876477782305, + "grad_norm": 0.9322537779808044, + "learning_rate": 0.0003943640440277212, + "loss": 3.3431, + "step": 71310 + }, + { + "epoch": 4.845427367848893, + "grad_norm": 1.0419809818267822, + "learning_rate": 0.00039432157901888844, + "loss": 3.5384, + "step": 71315 + }, + { + "epoch": 4.845767087919555, + "grad_norm": 0.8937583565711975, + "learning_rate": 0.0003942791140100557, + "loss": 3.7709, + "step": 71320 + }, + { + "epoch": 4.846106807990216, + "grad_norm": 0.9244757294654846, + "learning_rate": 0.000394236649001223, + "loss": 3.5872, + "step": 71325 + }, + { + "epoch": 4.846446528060878, + "grad_norm": 1.5295947790145874, + "learning_rate": 0.0003941941839923903, + "loss": 3.6434, + "step": 71330 + }, + { + "epoch": 4.84678624813154, + "grad_norm": 1.030714750289917, + "learning_rate": 0.00039415171898355756, + "loss": 3.6061, + "step": 71335 + }, + { + "epoch": 4.847125968202201, + "grad_norm": 1.011336088180542, + "learning_rate": 0.00039410925397472484, + "loss": 3.3618, + "step": 71340 + }, + { + "epoch": 4.847465688272863, + "grad_norm": 0.8177810311317444, + "learning_rate": 0.0003940667889658921, + "loss": 3.6246, + "step": 71345 + }, + { + "epoch": 4.847805408343525, + "grad_norm": 0.8188793659210205, + "learning_rate": 0.0003940243239570594, + "loss": 3.5037, + "step": 71350 + }, + { + "epoch": 4.8481451284141865, + "grad_norm": 1.1574666500091553, + "learning_rate": 0.0003939818589482267, + "loss": 3.8081, + "step": 71355 + }, + { + "epoch": 4.848484848484849, + "grad_norm": 0.730268657207489, + "learning_rate": 0.0003939393939393939, + "loss": 3.3167, + "step": 71360 + }, + { + "epoch": 4.848824568555511, + "grad_norm": 0.7519198060035706, + "learning_rate": 0.00039389692893056124, + "loss": 3.6142, + "step": 71365 + }, + { + "epoch": 4.849164288626172, + "grad_norm": 0.8022430539131165, + "learning_rate": 0.0003938544639217285, + "loss": 3.2648, + "step": 71370 + }, + { + "epoch": 4.849504008696834, + "grad_norm": 0.81453537940979, + "learning_rate": 0.00039381199891289575, + "loss": 3.4568, + "step": 71375 + }, + { + "epoch": 4.849843728767496, + "grad_norm": 0.9617143869400024, + "learning_rate": 0.0003937695339040631, + "loss": 3.4988, + "step": 71380 + }, + { + "epoch": 4.850183448838157, + "grad_norm": 1.1162461042404175, + "learning_rate": 0.00039372706889523036, + "loss": 3.3558, + "step": 71385 + }, + { + "epoch": 4.850523168908819, + "grad_norm": 0.7714512348175049, + "learning_rate": 0.0003936846038863976, + "loss": 3.4987, + "step": 71390 + }, + { + "epoch": 4.850862888979481, + "grad_norm": 0.9156549572944641, + "learning_rate": 0.00039364213887756487, + "loss": 3.2639, + "step": 71395 + }, + { + "epoch": 4.8512026090501426, + "grad_norm": 0.7589934468269348, + "learning_rate": 0.0003935996738687322, + "loss": 3.4715, + "step": 71400 + }, + { + "epoch": 4.851542329120805, + "grad_norm": 1.130082607269287, + "learning_rate": 0.00039355720885989943, + "loss": 3.4675, + "step": 71405 + }, + { + "epoch": 4.851882049191467, + "grad_norm": 0.96710205078125, + "learning_rate": 0.0003935147438510667, + "loss": 3.3725, + "step": 71410 + }, + { + "epoch": 4.852221769262128, + "grad_norm": 0.8954606652259827, + "learning_rate": 0.00039347227884223404, + "loss": 3.5633, + "step": 71415 + }, + { + "epoch": 4.85256148933279, + "grad_norm": 0.8415889143943787, + "learning_rate": 0.00039342981383340127, + "loss": 3.3142, + "step": 71420 + }, + { + "epoch": 4.852901209403452, + "grad_norm": 0.7687771320343018, + "learning_rate": 0.00039338734882456855, + "loss": 3.3821, + "step": 71425 + }, + { + "epoch": 4.853240929474113, + "grad_norm": 0.8235529661178589, + "learning_rate": 0.00039334488381573583, + "loss": 3.7031, + "step": 71430 + }, + { + "epoch": 4.853580649544775, + "grad_norm": 0.8341265320777893, + "learning_rate": 0.0003933024188069031, + "loss": 3.5259, + "step": 71435 + }, + { + "epoch": 4.8539203696154365, + "grad_norm": 0.8923922777175903, + "learning_rate": 0.0003932599537980704, + "loss": 3.7096, + "step": 71440 + }, + { + "epoch": 4.854260089686099, + "grad_norm": 0.9525147676467896, + "learning_rate": 0.00039321748878923767, + "loss": 3.298, + "step": 71445 + }, + { + "epoch": 4.854599809756761, + "grad_norm": 0.756129264831543, + "learning_rate": 0.00039317502378040495, + "loss": 3.4222, + "step": 71450 + }, + { + "epoch": 4.854939529827422, + "grad_norm": 0.9516236782073975, + "learning_rate": 0.00039313255877157223, + "loss": 3.4279, + "step": 71455 + }, + { + "epoch": 4.855279249898084, + "grad_norm": 0.9401823282241821, + "learning_rate": 0.0003930900937627395, + "loss": 3.4066, + "step": 71460 + }, + { + "epoch": 4.855618969968746, + "grad_norm": 0.9439297318458557, + "learning_rate": 0.00039304762875390684, + "loss": 3.3759, + "step": 71465 + }, + { + "epoch": 4.855958690039407, + "grad_norm": 0.9848126173019409, + "learning_rate": 0.00039300516374507407, + "loss": 3.4462, + "step": 71470 + }, + { + "epoch": 4.856298410110069, + "grad_norm": 1.0566965341567993, + "learning_rate": 0.00039296269873624135, + "loss": 3.69, + "step": 71475 + }, + { + "epoch": 4.856638130180731, + "grad_norm": 0.7110351920127869, + "learning_rate": 0.00039292023372740863, + "loss": 3.2803, + "step": 71480 + }, + { + "epoch": 4.8569778502513925, + "grad_norm": 0.8948861956596375, + "learning_rate": 0.0003928777687185759, + "loss": 3.7736, + "step": 71485 + }, + { + "epoch": 4.857317570322055, + "grad_norm": 0.9920291304588318, + "learning_rate": 0.0003928353037097432, + "loss": 3.6373, + "step": 71490 + }, + { + "epoch": 4.857657290392717, + "grad_norm": 0.8694120645523071, + "learning_rate": 0.00039279283870091047, + "loss": 3.3105, + "step": 71495 + }, + { + "epoch": 4.857997010463378, + "grad_norm": 0.7927242517471313, + "learning_rate": 0.00039275037369207775, + "loss": 3.7072, + "step": 71500 + }, + { + "epoch": 4.85833673053404, + "grad_norm": 1.2033480405807495, + "learning_rate": 0.00039270790868324503, + "loss": 3.2108, + "step": 71505 + }, + { + "epoch": 4.858676450604702, + "grad_norm": 0.8016459941864014, + "learning_rate": 0.0003926654436744123, + "loss": 3.2821, + "step": 71510 + }, + { + "epoch": 4.859016170675363, + "grad_norm": 1.0020811557769775, + "learning_rate": 0.00039262297866557953, + "loss": 3.3145, + "step": 71515 + }, + { + "epoch": 4.859355890746025, + "grad_norm": 0.7208513021469116, + "learning_rate": 0.00039258051365674687, + "loss": 3.3792, + "step": 71520 + }, + { + "epoch": 4.859695610816687, + "grad_norm": 0.96617591381073, + "learning_rate": 0.00039253804864791415, + "loss": 3.4933, + "step": 71525 + }, + { + "epoch": 4.8600353308873485, + "grad_norm": 0.9417550563812256, + "learning_rate": 0.0003924955836390814, + "loss": 3.4073, + "step": 71530 + }, + { + "epoch": 4.860375050958011, + "grad_norm": 1.1125959157943726, + "learning_rate": 0.0003924531186302487, + "loss": 3.4119, + "step": 71535 + }, + { + "epoch": 4.860714771028673, + "grad_norm": 0.9877198338508606, + "learning_rate": 0.000392410653621416, + "loss": 3.54, + "step": 71540 + }, + { + "epoch": 4.861054491099334, + "grad_norm": 0.8026974201202393, + "learning_rate": 0.0003923681886125832, + "loss": 3.7745, + "step": 71545 + }, + { + "epoch": 4.861394211169996, + "grad_norm": 0.7902126312255859, + "learning_rate": 0.0003923257236037505, + "loss": 3.4443, + "step": 71550 + }, + { + "epoch": 4.861733931240658, + "grad_norm": 1.0389958620071411, + "learning_rate": 0.00039228325859491783, + "loss": 3.6793, + "step": 71555 + }, + { + "epoch": 4.862073651311319, + "grad_norm": 0.9055476784706116, + "learning_rate": 0.00039224079358608506, + "loss": 3.262, + "step": 71560 + }, + { + "epoch": 4.862413371381981, + "grad_norm": 1.3160901069641113, + "learning_rate": 0.00039219832857725234, + "loss": 3.3541, + "step": 71565 + }, + { + "epoch": 4.862753091452643, + "grad_norm": 1.0848639011383057, + "learning_rate": 0.00039215586356841967, + "loss": 3.4138, + "step": 71570 + }, + { + "epoch": 4.8630928115233045, + "grad_norm": 0.8789510130882263, + "learning_rate": 0.0003921133985595869, + "loss": 3.5598, + "step": 71575 + }, + { + "epoch": 4.863432531593967, + "grad_norm": 0.9375132322311401, + "learning_rate": 0.0003920709335507542, + "loss": 3.535, + "step": 71580 + }, + { + "epoch": 4.863772251664629, + "grad_norm": 0.8606034517288208, + "learning_rate": 0.00039202846854192146, + "loss": 3.726, + "step": 71585 + }, + { + "epoch": 4.86411197173529, + "grad_norm": 0.8120411038398743, + "learning_rate": 0.00039198600353308874, + "loss": 3.2808, + "step": 71590 + }, + { + "epoch": 4.864451691805952, + "grad_norm": 0.7670430541038513, + "learning_rate": 0.000391943538524256, + "loss": 3.5733, + "step": 71595 + }, + { + "epoch": 4.864791411876614, + "grad_norm": 0.6934714913368225, + "learning_rate": 0.0003919010735154233, + "loss": 3.3441, + "step": 71600 + }, + { + "epoch": 4.865131131947275, + "grad_norm": 0.7447838187217712, + "learning_rate": 0.0003918586085065906, + "loss": 3.6007, + "step": 71605 + }, + { + "epoch": 4.865470852017937, + "grad_norm": 0.9286085367202759, + "learning_rate": 0.00039181614349775786, + "loss": 3.3004, + "step": 71610 + }, + { + "epoch": 4.865810572088599, + "grad_norm": 0.9078155755996704, + "learning_rate": 0.00039177367848892514, + "loss": 3.3515, + "step": 71615 + }, + { + "epoch": 4.8661502921592605, + "grad_norm": 1.0792973041534424, + "learning_rate": 0.00039173121348009236, + "loss": 3.4091, + "step": 71620 + }, + { + "epoch": 4.866490012229923, + "grad_norm": 0.8293854594230652, + "learning_rate": 0.0003916887484712597, + "loss": 3.6954, + "step": 71625 + }, + { + "epoch": 4.866829732300585, + "grad_norm": 0.9288709759712219, + "learning_rate": 0.000391646283462427, + "loss": 3.3769, + "step": 71630 + }, + { + "epoch": 4.867169452371246, + "grad_norm": 0.9325634241104126, + "learning_rate": 0.00039160381845359426, + "loss": 3.4428, + "step": 71635 + }, + { + "epoch": 4.867509172441908, + "grad_norm": 0.7453815937042236, + "learning_rate": 0.00039156135344476154, + "loss": 3.5358, + "step": 71640 + }, + { + "epoch": 4.86784889251257, + "grad_norm": 0.9887683391571045, + "learning_rate": 0.0003915188884359288, + "loss": 3.727, + "step": 71645 + }, + { + "epoch": 4.868188612583231, + "grad_norm": 0.709135115146637, + "learning_rate": 0.0003914764234270961, + "loss": 3.4177, + "step": 71650 + }, + { + "epoch": 4.868528332653893, + "grad_norm": 0.8720343112945557, + "learning_rate": 0.0003914339584182633, + "loss": 3.5084, + "step": 71655 + }, + { + "epoch": 4.8688680527245545, + "grad_norm": 1.0377472639083862, + "learning_rate": 0.00039139149340943066, + "loss": 3.3586, + "step": 71660 + }, + { + "epoch": 4.8692077727952165, + "grad_norm": 0.7515466213226318, + "learning_rate": 0.00039134902840059794, + "loss": 3.5479, + "step": 71665 + }, + { + "epoch": 4.869547492865879, + "grad_norm": 0.7997063994407654, + "learning_rate": 0.00039130656339176516, + "loss": 3.8232, + "step": 71670 + }, + { + "epoch": 4.86988721293654, + "grad_norm": 0.7731438279151917, + "learning_rate": 0.0003912640983829325, + "loss": 3.5762, + "step": 71675 + }, + { + "epoch": 4.870226933007202, + "grad_norm": 0.8113502860069275, + "learning_rate": 0.0003912216333740998, + "loss": 3.5236, + "step": 71680 + }, + { + "epoch": 4.870566653077864, + "grad_norm": 1.0951547622680664, + "learning_rate": 0.000391179168365267, + "loss": 3.1938, + "step": 71685 + }, + { + "epoch": 4.870906373148525, + "grad_norm": 0.8797723054885864, + "learning_rate": 0.0003911367033564343, + "loss": 3.4559, + "step": 71690 + }, + { + "epoch": 4.871246093219187, + "grad_norm": 0.9059113264083862, + "learning_rate": 0.0003910942383476016, + "loss": 3.4115, + "step": 71695 + }, + { + "epoch": 4.871585813289849, + "grad_norm": 0.8071296215057373, + "learning_rate": 0.00039105177333876884, + "loss": 3.5085, + "step": 71700 + }, + { + "epoch": 4.8719255333605105, + "grad_norm": 0.8111851811408997, + "learning_rate": 0.0003910093083299361, + "loss": 3.2816, + "step": 71705 + }, + { + "epoch": 4.872265253431173, + "grad_norm": 0.8736326098442078, + "learning_rate": 0.00039096684332110346, + "loss": 3.0608, + "step": 71710 + }, + { + "epoch": 4.872604973501835, + "grad_norm": 0.9330747723579407, + "learning_rate": 0.0003909243783122707, + "loss": 3.2157, + "step": 71715 + }, + { + "epoch": 4.872944693572496, + "grad_norm": 0.8230901956558228, + "learning_rate": 0.00039088191330343796, + "loss": 3.707, + "step": 71720 + }, + { + "epoch": 4.873284413643158, + "grad_norm": 1.0135942697525024, + "learning_rate": 0.0003908394482946053, + "loss": 3.5181, + "step": 71725 + }, + { + "epoch": 4.87362413371382, + "grad_norm": 0.7113049030303955, + "learning_rate": 0.0003907969832857725, + "loss": 3.6278, + "step": 71730 + }, + { + "epoch": 4.873963853784481, + "grad_norm": 0.9025107026100159, + "learning_rate": 0.0003907545182769398, + "loss": 3.5791, + "step": 71735 + }, + { + "epoch": 4.874303573855143, + "grad_norm": 0.7336267232894897, + "learning_rate": 0.0003907120532681071, + "loss": 3.481, + "step": 71740 + }, + { + "epoch": 4.874643293925805, + "grad_norm": 0.9404609203338623, + "learning_rate": 0.00039066958825927436, + "loss": 3.6503, + "step": 71745 + }, + { + "epoch": 4.8749830139964665, + "grad_norm": 0.7735083103179932, + "learning_rate": 0.00039062712325044164, + "loss": 3.4503, + "step": 71750 + }, + { + "epoch": 4.875322734067129, + "grad_norm": 0.8202487230300903, + "learning_rate": 0.0003905846582416089, + "loss": 3.3182, + "step": 71755 + }, + { + "epoch": 4.875662454137791, + "grad_norm": 0.7175875902175903, + "learning_rate": 0.0003905421932327762, + "loss": 3.4519, + "step": 71760 + }, + { + "epoch": 4.876002174208452, + "grad_norm": 1.2140263319015503, + "learning_rate": 0.0003904997282239435, + "loss": 3.729, + "step": 71765 + }, + { + "epoch": 4.876341894279114, + "grad_norm": 1.1485114097595215, + "learning_rate": 0.00039045726321511076, + "loss": 3.3844, + "step": 71770 + }, + { + "epoch": 4.876681614349776, + "grad_norm": 0.8044224381446838, + "learning_rate": 0.000390414798206278, + "loss": 3.592, + "step": 71775 + }, + { + "epoch": 4.877021334420437, + "grad_norm": 1.0518856048583984, + "learning_rate": 0.0003903723331974453, + "loss": 3.6643, + "step": 71780 + }, + { + "epoch": 4.877361054491099, + "grad_norm": 0.7046890258789062, + "learning_rate": 0.0003903298681886126, + "loss": 3.4309, + "step": 71785 + }, + { + "epoch": 4.877700774561761, + "grad_norm": 0.9900513887405396, + "learning_rate": 0.00039028740317977983, + "loss": 3.5369, + "step": 71790 + }, + { + "epoch": 4.8780404946324225, + "grad_norm": 0.8486366868019104, + "learning_rate": 0.00039024493817094716, + "loss": 3.2129, + "step": 71795 + }, + { + "epoch": 4.878380214703085, + "grad_norm": 0.7431617379188538, + "learning_rate": 0.00039020247316211444, + "loss": 3.5017, + "step": 71800 + }, + { + "epoch": 4.878719934773747, + "grad_norm": 1.0684456825256348, + "learning_rate": 0.0003901600081532817, + "loss": 3.3193, + "step": 71805 + }, + { + "epoch": 4.879059654844408, + "grad_norm": 1.158445119857788, + "learning_rate": 0.00039011754314444895, + "loss": 3.6273, + "step": 71810 + }, + { + "epoch": 4.87939937491507, + "grad_norm": 0.953860342502594, + "learning_rate": 0.0003900750781356163, + "loss": 3.4245, + "step": 71815 + }, + { + "epoch": 4.879739094985732, + "grad_norm": 1.0015419721603394, + "learning_rate": 0.00039003261312678357, + "loss": 3.4759, + "step": 71820 + }, + { + "epoch": 4.880078815056393, + "grad_norm": 0.7116113305091858, + "learning_rate": 0.0003899901481179508, + "loss": 3.5564, + "step": 71825 + }, + { + "epoch": 4.880418535127055, + "grad_norm": 0.909771740436554, + "learning_rate": 0.0003899476831091181, + "loss": 3.586, + "step": 71830 + }, + { + "epoch": 4.880758255197717, + "grad_norm": 0.8496811985969543, + "learning_rate": 0.0003899052181002854, + "loss": 3.6646, + "step": 71835 + }, + { + "epoch": 4.8810979752683785, + "grad_norm": 0.8220270872116089, + "learning_rate": 0.00038986275309145263, + "loss": 3.4962, + "step": 71840 + }, + { + "epoch": 4.881437695339041, + "grad_norm": 1.0029504299163818, + "learning_rate": 0.0003898202880826199, + "loss": 3.4967, + "step": 71845 + }, + { + "epoch": 4.881777415409703, + "grad_norm": 1.0897222757339478, + "learning_rate": 0.00038977782307378725, + "loss": 3.4652, + "step": 71850 + }, + { + "epoch": 4.882117135480364, + "grad_norm": 0.7791353464126587, + "learning_rate": 0.00038973535806495447, + "loss": 3.451, + "step": 71855 + }, + { + "epoch": 4.882456855551026, + "grad_norm": 0.8146374225616455, + "learning_rate": 0.00038969289305612175, + "loss": 3.6017, + "step": 71860 + }, + { + "epoch": 4.882796575621688, + "grad_norm": 0.8084748387336731, + "learning_rate": 0.0003896504280472891, + "loss": 3.7175, + "step": 71865 + }, + { + "epoch": 4.883136295692349, + "grad_norm": 0.7945273518562317, + "learning_rate": 0.0003896079630384563, + "loss": 3.6651, + "step": 71870 + }, + { + "epoch": 4.883476015763011, + "grad_norm": 1.1460814476013184, + "learning_rate": 0.0003895654980296236, + "loss": 3.4716, + "step": 71875 + }, + { + "epoch": 4.883815735833673, + "grad_norm": 0.8777429461479187, + "learning_rate": 0.00038952303302079087, + "loss": 3.6556, + "step": 71880 + }, + { + "epoch": 4.8841554559043345, + "grad_norm": 1.0097410678863525, + "learning_rate": 0.00038948056801195815, + "loss": 3.4465, + "step": 71885 + }, + { + "epoch": 4.884495175974997, + "grad_norm": 0.9611325263977051, + "learning_rate": 0.00038943810300312543, + "loss": 3.4304, + "step": 71890 + }, + { + "epoch": 4.884834896045659, + "grad_norm": 0.8928402066230774, + "learning_rate": 0.0003893956379942927, + "loss": 3.4371, + "step": 71895 + }, + { + "epoch": 4.88517461611632, + "grad_norm": 1.0716354846954346, + "learning_rate": 0.00038935317298546, + "loss": 3.3305, + "step": 71900 + }, + { + "epoch": 4.885514336186982, + "grad_norm": 0.7230565547943115, + "learning_rate": 0.00038931070797662727, + "loss": 3.5492, + "step": 71905 + }, + { + "epoch": 4.885854056257644, + "grad_norm": 1.0003303289413452, + "learning_rate": 0.00038926824296779455, + "loss": 3.3361, + "step": 71910 + }, + { + "epoch": 4.886193776328305, + "grad_norm": 0.7246459722518921, + "learning_rate": 0.0003892257779589618, + "loss": 3.3967, + "step": 71915 + }, + { + "epoch": 4.886533496398967, + "grad_norm": 0.8626651763916016, + "learning_rate": 0.0003891833129501291, + "loss": 3.4974, + "step": 71920 + }, + { + "epoch": 4.886873216469629, + "grad_norm": 0.966983437538147, + "learning_rate": 0.0003891408479412964, + "loss": 3.3502, + "step": 71925 + }, + { + "epoch": 4.8872129365402905, + "grad_norm": 0.8597303032875061, + "learning_rate": 0.0003890983829324636, + "loss": 3.5668, + "step": 71930 + }, + { + "epoch": 4.887552656610953, + "grad_norm": 0.9356186985969543, + "learning_rate": 0.00038905591792363095, + "loss": 3.4582, + "step": 71935 + }, + { + "epoch": 4.887892376681615, + "grad_norm": 0.8280949592590332, + "learning_rate": 0.00038901345291479823, + "loss": 3.7715, + "step": 71940 + }, + { + "epoch": 4.888232096752276, + "grad_norm": 0.8792530298233032, + "learning_rate": 0.00038897098790596546, + "loss": 3.5101, + "step": 71945 + }, + { + "epoch": 4.888571816822938, + "grad_norm": 0.8395430445671082, + "learning_rate": 0.00038892852289713274, + "loss": 3.5229, + "step": 71950 + }, + { + "epoch": 4.8889115368936, + "grad_norm": 0.8534421324729919, + "learning_rate": 0.0003888860578883001, + "loss": 3.7282, + "step": 71955 + }, + { + "epoch": 4.889251256964261, + "grad_norm": 0.8459038734436035, + "learning_rate": 0.0003888435928794673, + "loss": 3.1269, + "step": 71960 + }, + { + "epoch": 4.889590977034923, + "grad_norm": 0.9300627112388611, + "learning_rate": 0.0003888011278706346, + "loss": 3.3672, + "step": 71965 + }, + { + "epoch": 4.889930697105585, + "grad_norm": 0.8684754371643066, + "learning_rate": 0.0003887586628618019, + "loss": 3.3354, + "step": 71970 + }, + { + "epoch": 4.8902704171762466, + "grad_norm": 0.8709257245063782, + "learning_rate": 0.0003887161978529692, + "loss": 3.6272, + "step": 71975 + }, + { + "epoch": 4.890610137246909, + "grad_norm": 1.0498430728912354, + "learning_rate": 0.0003886737328441364, + "loss": 3.4873, + "step": 71980 + }, + { + "epoch": 4.890949857317571, + "grad_norm": 0.8447078466415405, + "learning_rate": 0.0003886312678353037, + "loss": 3.6963, + "step": 71985 + }, + { + "epoch": 4.891289577388232, + "grad_norm": 0.9700011014938354, + "learning_rate": 0.00038858880282647103, + "loss": 3.6354, + "step": 71990 + }, + { + "epoch": 4.891629297458894, + "grad_norm": 0.958055853843689, + "learning_rate": 0.00038854633781763826, + "loss": 3.7072, + "step": 71995 + }, + { + "epoch": 4.891969017529556, + "grad_norm": 0.7666856646537781, + "learning_rate": 0.00038850387280880554, + "loss": 3.533, + "step": 72000 + }, + { + "epoch": 4.892308737600217, + "grad_norm": 0.8994258046150208, + "learning_rate": 0.0003884614077999729, + "loss": 3.5281, + "step": 72005 + }, + { + "epoch": 4.892648457670879, + "grad_norm": 1.167443037033081, + "learning_rate": 0.0003884189427911401, + "loss": 3.2858, + "step": 72010 + }, + { + "epoch": 4.892988177741541, + "grad_norm": 1.1294502019882202, + "learning_rate": 0.0003883764777823074, + "loss": 3.686, + "step": 72015 + }, + { + "epoch": 4.893327897812203, + "grad_norm": 1.0230016708374023, + "learning_rate": 0.0003883340127734747, + "loss": 3.6137, + "step": 72020 + }, + { + "epoch": 4.893667617882865, + "grad_norm": 0.9003035426139832, + "learning_rate": 0.00038829154776464194, + "loss": 3.4534, + "step": 72025 + }, + { + "epoch": 4.894007337953527, + "grad_norm": 0.9287586212158203, + "learning_rate": 0.0003882490827558092, + "loss": 3.7582, + "step": 72030 + }, + { + "epoch": 4.894347058024188, + "grad_norm": 0.6925762891769409, + "learning_rate": 0.0003882066177469765, + "loss": 3.1505, + "step": 72035 + }, + { + "epoch": 4.89468677809485, + "grad_norm": 1.0912725925445557, + "learning_rate": 0.0003881641527381438, + "loss": 3.7986, + "step": 72040 + }, + { + "epoch": 4.895026498165512, + "grad_norm": 1.1536036729812622, + "learning_rate": 0.00038812168772931106, + "loss": 3.5491, + "step": 72045 + }, + { + "epoch": 4.895366218236173, + "grad_norm": 0.9433629512786865, + "learning_rate": 0.00038807922272047834, + "loss": 3.5042, + "step": 72050 + }, + { + "epoch": 4.895705938306835, + "grad_norm": 0.8793943524360657, + "learning_rate": 0.0003880367577116456, + "loss": 3.5939, + "step": 72055 + }, + { + "epoch": 4.896045658377497, + "grad_norm": 1.0590320825576782, + "learning_rate": 0.0003879942927028129, + "loss": 3.5035, + "step": 72060 + }, + { + "epoch": 4.896385378448159, + "grad_norm": 0.8599629402160645, + "learning_rate": 0.0003879518276939802, + "loss": 3.5696, + "step": 72065 + }, + { + "epoch": 4.896725098518821, + "grad_norm": 1.0057350397109985, + "learning_rate": 0.0003879093626851474, + "loss": 3.5252, + "step": 72070 + }, + { + "epoch": 4.897064818589483, + "grad_norm": 0.7453026175498962, + "learning_rate": 0.00038786689767631474, + "loss": 3.5433, + "step": 72075 + }, + { + "epoch": 4.897404538660144, + "grad_norm": 1.050764799118042, + "learning_rate": 0.000387824432667482, + "loss": 3.4845, + "step": 72080 + }, + { + "epoch": 4.897744258730806, + "grad_norm": 0.7781047224998474, + "learning_rate": 0.00038778196765864925, + "loss": 3.6845, + "step": 72085 + }, + { + "epoch": 4.898083978801468, + "grad_norm": 0.9952058792114258, + "learning_rate": 0.0003877395026498166, + "loss": 3.6216, + "step": 72090 + }, + { + "epoch": 4.898423698872129, + "grad_norm": 1.109413743019104, + "learning_rate": 0.00038769703764098386, + "loss": 3.3974, + "step": 72095 + }, + { + "epoch": 4.898763418942791, + "grad_norm": 0.8139770030975342, + "learning_rate": 0.0003876545726321511, + "loss": 3.455, + "step": 72100 + }, + { + "epoch": 4.899103139013453, + "grad_norm": 0.9477381706237793, + "learning_rate": 0.00038761210762331837, + "loss": 3.66, + "step": 72105 + }, + { + "epoch": 4.899442859084115, + "grad_norm": 1.145743489265442, + "learning_rate": 0.0003875696426144857, + "loss": 3.6234, + "step": 72110 + }, + { + "epoch": 4.899782579154777, + "grad_norm": 1.0026410818099976, + "learning_rate": 0.00038752717760565293, + "loss": 3.1466, + "step": 72115 + }, + { + "epoch": 4.900122299225439, + "grad_norm": 0.885558009147644, + "learning_rate": 0.0003874847125968202, + "loss": 3.4774, + "step": 72120 + }, + { + "epoch": 4.9004620192961, + "grad_norm": 0.7886387705802917, + "learning_rate": 0.00038744224758798754, + "loss": 3.3185, + "step": 72125 + }, + { + "epoch": 4.900801739366762, + "grad_norm": 0.9096794724464417, + "learning_rate": 0.00038739978257915477, + "loss": 3.6874, + "step": 72130 + }, + { + "epoch": 4.901141459437423, + "grad_norm": 0.7408117055892944, + "learning_rate": 0.00038735731757032205, + "loss": 3.4992, + "step": 72135 + }, + { + "epoch": 4.901481179508085, + "grad_norm": 1.1832871437072754, + "learning_rate": 0.00038731485256148933, + "loss": 3.3447, + "step": 72140 + }, + { + "epoch": 4.901820899578747, + "grad_norm": 0.925692081451416, + "learning_rate": 0.00038727238755265666, + "loss": 3.4921, + "step": 72145 + }, + { + "epoch": 4.9021606196494085, + "grad_norm": 0.7824732661247253, + "learning_rate": 0.0003872299225438239, + "loss": 3.6578, + "step": 72150 + }, + { + "epoch": 4.902500339720071, + "grad_norm": 0.9270050525665283, + "learning_rate": 0.00038718745753499117, + "loss": 3.3486, + "step": 72155 + }, + { + "epoch": 4.902840059790733, + "grad_norm": 1.3210622072219849, + "learning_rate": 0.0003871449925261585, + "loss": 3.4937, + "step": 72160 + }, + { + "epoch": 4.903179779861394, + "grad_norm": 0.738833487033844, + "learning_rate": 0.00038710252751732573, + "loss": 3.3761, + "step": 72165 + }, + { + "epoch": 4.903519499932056, + "grad_norm": 0.9803553819656372, + "learning_rate": 0.000387060062508493, + "loss": 3.4879, + "step": 72170 + }, + { + "epoch": 4.903859220002718, + "grad_norm": 0.8227695822715759, + "learning_rate": 0.0003870175974996603, + "loss": 3.5295, + "step": 72175 + }, + { + "epoch": 4.904198940073379, + "grad_norm": 0.7833067178726196, + "learning_rate": 0.00038697513249082757, + "loss": 3.5768, + "step": 72180 + }, + { + "epoch": 4.904538660144041, + "grad_norm": 0.9094823002815247, + "learning_rate": 0.00038693266748199485, + "loss": 3.6476, + "step": 72185 + }, + { + "epoch": 4.904878380214703, + "grad_norm": 0.694635808467865, + "learning_rate": 0.00038689020247316213, + "loss": 3.7602, + "step": 72190 + }, + { + "epoch": 4.9052181002853645, + "grad_norm": 1.3866500854492188, + "learning_rate": 0.0003868477374643294, + "loss": 3.4395, + "step": 72195 + }, + { + "epoch": 4.905557820356027, + "grad_norm": 0.9556769728660583, + "learning_rate": 0.0003868052724554967, + "loss": 3.5459, + "step": 72200 + }, + { + "epoch": 4.905897540426689, + "grad_norm": 0.8479629158973694, + "learning_rate": 0.00038676280744666397, + "loss": 3.4713, + "step": 72205 + }, + { + "epoch": 4.90623726049735, + "grad_norm": 0.8694672584533691, + "learning_rate": 0.0003867203424378312, + "loss": 3.5988, + "step": 72210 + }, + { + "epoch": 4.906576980568012, + "grad_norm": 0.7101976871490479, + "learning_rate": 0.00038667787742899853, + "loss": 3.4705, + "step": 72215 + }, + { + "epoch": 4.906916700638674, + "grad_norm": 1.0196577310562134, + "learning_rate": 0.0003866354124201658, + "loss": 3.3522, + "step": 72220 + }, + { + "epoch": 4.907256420709335, + "grad_norm": 0.8992969393730164, + "learning_rate": 0.00038659294741133303, + "loss": 3.4922, + "step": 72225 + }, + { + "epoch": 4.907596140779997, + "grad_norm": 1.0028029680252075, + "learning_rate": 0.00038655048240250037, + "loss": 3.2312, + "step": 72230 + }, + { + "epoch": 4.907935860850659, + "grad_norm": 0.8917765021324158, + "learning_rate": 0.00038650801739366765, + "loss": 3.31, + "step": 72235 + }, + { + "epoch": 4.9082755809213205, + "grad_norm": 0.8796781897544861, + "learning_rate": 0.0003864655523848349, + "loss": 3.5904, + "step": 72240 + }, + { + "epoch": 4.908615300991983, + "grad_norm": 0.8835317492485046, + "learning_rate": 0.00038642308737600215, + "loss": 3.6431, + "step": 72245 + }, + { + "epoch": 4.908955021062645, + "grad_norm": 0.9377812147140503, + "learning_rate": 0.0003863806223671695, + "loss": 3.3825, + "step": 72250 + }, + { + "epoch": 4.909294741133306, + "grad_norm": 1.1719474792480469, + "learning_rate": 0.0003863381573583367, + "loss": 3.4916, + "step": 72255 + }, + { + "epoch": 4.909634461203968, + "grad_norm": 0.9674480557441711, + "learning_rate": 0.000386295692349504, + "loss": 3.6128, + "step": 72260 + }, + { + "epoch": 4.90997418127463, + "grad_norm": 0.8809230923652649, + "learning_rate": 0.00038625322734067133, + "loss": 3.5105, + "step": 72265 + }, + { + "epoch": 4.910313901345291, + "grad_norm": 1.0579838752746582, + "learning_rate": 0.00038621076233183856, + "loss": 3.3242, + "step": 72270 + }, + { + "epoch": 4.910653621415953, + "grad_norm": 1.203029990196228, + "learning_rate": 0.00038616829732300584, + "loss": 3.4563, + "step": 72275 + }, + { + "epoch": 4.910993341486615, + "grad_norm": 0.7604086399078369, + "learning_rate": 0.0003861258323141731, + "loss": 3.38, + "step": 72280 + }, + { + "epoch": 4.911333061557277, + "grad_norm": 0.9430040121078491, + "learning_rate": 0.0003860833673053404, + "loss": 3.4707, + "step": 72285 + }, + { + "epoch": 4.911672781627939, + "grad_norm": 0.7291462421417236, + "learning_rate": 0.0003860409022965077, + "loss": 3.1049, + "step": 72290 + }, + { + "epoch": 4.912012501698601, + "grad_norm": 0.8246901631355286, + "learning_rate": 0.00038599843728767496, + "loss": 3.2076, + "step": 72295 + }, + { + "epoch": 4.912352221769262, + "grad_norm": 0.7850285172462463, + "learning_rate": 0.00038595597227884224, + "loss": 3.3194, + "step": 72300 + }, + { + "epoch": 4.912691941839924, + "grad_norm": 0.8531397581100464, + "learning_rate": 0.0003859135072700095, + "loss": 3.6342, + "step": 72305 + }, + { + "epoch": 4.913031661910586, + "grad_norm": 1.2020094394683838, + "learning_rate": 0.0003858710422611768, + "loss": 3.488, + "step": 72310 + }, + { + "epoch": 4.913371381981247, + "grad_norm": 1.1814930438995361, + "learning_rate": 0.00038582857725234413, + "loss": 3.4106, + "step": 72315 + }, + { + "epoch": 4.913711102051909, + "grad_norm": 0.8131977319717407, + "learning_rate": 0.00038578611224351136, + "loss": 3.4921, + "step": 72320 + }, + { + "epoch": 4.914050822122571, + "grad_norm": 0.9900912046432495, + "learning_rate": 0.00038574364723467864, + "loss": 3.5224, + "step": 72325 + }, + { + "epoch": 4.914390542193233, + "grad_norm": 0.8348836898803711, + "learning_rate": 0.0003857011822258459, + "loss": 3.2632, + "step": 72330 + }, + { + "epoch": 4.914730262263895, + "grad_norm": 1.109505534172058, + "learning_rate": 0.0003856587172170132, + "loss": 3.4626, + "step": 72335 + }, + { + "epoch": 4.915069982334556, + "grad_norm": 0.9630661010742188, + "learning_rate": 0.0003856162522081805, + "loss": 3.556, + "step": 72340 + }, + { + "epoch": 4.915409702405218, + "grad_norm": 1.0526721477508545, + "learning_rate": 0.00038557378719934776, + "loss": 3.5273, + "step": 72345 + }, + { + "epoch": 4.91574942247588, + "grad_norm": 1.1559056043624878, + "learning_rate": 0.00038553132219051504, + "loss": 3.4502, + "step": 72350 + }, + { + "epoch": 4.916089142546541, + "grad_norm": 0.7480113506317139, + "learning_rate": 0.0003854888571816823, + "loss": 3.5422, + "step": 72355 + }, + { + "epoch": 4.916428862617203, + "grad_norm": 0.7670375108718872, + "learning_rate": 0.0003854463921728496, + "loss": 3.4582, + "step": 72360 + }, + { + "epoch": 4.916768582687865, + "grad_norm": 0.9430740475654602, + "learning_rate": 0.0003854039271640168, + "loss": 3.5059, + "step": 72365 + }, + { + "epoch": 4.9171083027585265, + "grad_norm": 0.9023682475090027, + "learning_rate": 0.00038536146215518416, + "loss": 3.4228, + "step": 72370 + }, + { + "epoch": 4.917448022829189, + "grad_norm": 0.7877408862113953, + "learning_rate": 0.00038531899714635144, + "loss": 3.5894, + "step": 72375 + }, + { + "epoch": 4.917787742899851, + "grad_norm": 0.9517087340354919, + "learning_rate": 0.00038527653213751866, + "loss": 3.4393, + "step": 72380 + }, + { + "epoch": 4.918127462970512, + "grad_norm": 1.4943435192108154, + "learning_rate": 0.000385234067128686, + "loss": 3.4386, + "step": 72385 + }, + { + "epoch": 4.918467183041174, + "grad_norm": 0.8636531233787537, + "learning_rate": 0.0003851916021198533, + "loss": 3.5766, + "step": 72390 + }, + { + "epoch": 4.918806903111836, + "grad_norm": 1.0968360900878906, + "learning_rate": 0.0003851491371110205, + "loss": 3.5288, + "step": 72395 + }, + { + "epoch": 4.919146623182497, + "grad_norm": 0.7035086750984192, + "learning_rate": 0.0003851066721021878, + "loss": 3.623, + "step": 72400 + }, + { + "epoch": 4.919486343253159, + "grad_norm": 0.7638811469078064, + "learning_rate": 0.0003850642070933551, + "loss": 3.3787, + "step": 72405 + }, + { + "epoch": 4.919826063323821, + "grad_norm": 1.0079468488693237, + "learning_rate": 0.00038502174208452234, + "loss": 3.5519, + "step": 72410 + }, + { + "epoch": 4.9201657833944825, + "grad_norm": 2.7593867778778076, + "learning_rate": 0.0003849792770756896, + "loss": 3.5095, + "step": 72415 + }, + { + "epoch": 4.920505503465145, + "grad_norm": 0.7488455772399902, + "learning_rate": 0.00038493681206685696, + "loss": 3.4229, + "step": 72420 + }, + { + "epoch": 4.920845223535807, + "grad_norm": 1.2424204349517822, + "learning_rate": 0.0003848943470580242, + "loss": 3.2775, + "step": 72425 + }, + { + "epoch": 4.921184943606468, + "grad_norm": 1.3415158987045288, + "learning_rate": 0.00038485188204919146, + "loss": 3.3861, + "step": 72430 + }, + { + "epoch": 4.92152466367713, + "grad_norm": 0.7716087698936462, + "learning_rate": 0.00038480941704035874, + "loss": 3.2917, + "step": 72435 + }, + { + "epoch": 4.921864383747792, + "grad_norm": 0.8157476186752319, + "learning_rate": 0.000384766952031526, + "loss": 3.4595, + "step": 72440 + }, + { + "epoch": 4.922204103818453, + "grad_norm": 1.0020833015441895, + "learning_rate": 0.0003847244870226933, + "loss": 3.605, + "step": 72445 + }, + { + "epoch": 4.922543823889115, + "grad_norm": 1.0010167360305786, + "learning_rate": 0.0003846820220138606, + "loss": 3.2341, + "step": 72450 + }, + { + "epoch": 4.922883543959777, + "grad_norm": 0.844659149646759, + "learning_rate": 0.00038463955700502786, + "loss": 3.455, + "step": 72455 + }, + { + "epoch": 4.9232232640304385, + "grad_norm": 0.6769673228263855, + "learning_rate": 0.00038459709199619514, + "loss": 3.524, + "step": 72460 + }, + { + "epoch": 4.923562984101101, + "grad_norm": 0.8233169317245483, + "learning_rate": 0.0003845546269873624, + "loss": 3.5392, + "step": 72465 + }, + { + "epoch": 4.923902704171763, + "grad_norm": 0.6875226497650146, + "learning_rate": 0.00038451216197852965, + "loss": 3.6913, + "step": 72470 + }, + { + "epoch": 4.924242424242424, + "grad_norm": 0.8477721810340881, + "learning_rate": 0.000384469696969697, + "loss": 3.6157, + "step": 72475 + }, + { + "epoch": 4.924582144313086, + "grad_norm": 0.7096152305603027, + "learning_rate": 0.00038442723196086426, + "loss": 3.315, + "step": 72480 + }, + { + "epoch": 4.924921864383748, + "grad_norm": 1.0758352279663086, + "learning_rate": 0.00038438476695203154, + "loss": 3.4009, + "step": 72485 + }, + { + "epoch": 4.925261584454409, + "grad_norm": 0.9256690740585327, + "learning_rate": 0.0003843423019431988, + "loss": 3.3645, + "step": 72490 + }, + { + "epoch": 4.925601304525071, + "grad_norm": 1.2663222551345825, + "learning_rate": 0.0003842998369343661, + "loss": 3.3535, + "step": 72495 + }, + { + "epoch": 4.925941024595733, + "grad_norm": 1.4317413568496704, + "learning_rate": 0.0003842573719255334, + "loss": 3.4024, + "step": 72500 + }, + { + "epoch": 4.9262807446663945, + "grad_norm": 1.0781514644622803, + "learning_rate": 0.0003842149069167006, + "loss": 3.3881, + "step": 72505 + }, + { + "epoch": 4.926620464737057, + "grad_norm": 1.000072956085205, + "learning_rate": 0.00038417244190786794, + "loss": 3.6698, + "step": 72510 + }, + { + "epoch": 4.926960184807719, + "grad_norm": 0.9730064868927002, + "learning_rate": 0.0003841299768990352, + "loss": 3.7727, + "step": 72515 + }, + { + "epoch": 4.92729990487838, + "grad_norm": 0.8389480710029602, + "learning_rate": 0.00038408751189020245, + "loss": 3.5537, + "step": 72520 + }, + { + "epoch": 4.927639624949042, + "grad_norm": 0.9581026434898376, + "learning_rate": 0.0003840450468813698, + "loss": 3.4466, + "step": 72525 + }, + { + "epoch": 4.927979345019704, + "grad_norm": 0.8313193321228027, + "learning_rate": 0.00038400258187253707, + "loss": 3.2836, + "step": 72530 + }, + { + "epoch": 4.928319065090365, + "grad_norm": 0.7633522748947144, + "learning_rate": 0.0003839601168637043, + "loss": 3.3872, + "step": 72535 + }, + { + "epoch": 4.928658785161027, + "grad_norm": 0.7295286059379578, + "learning_rate": 0.00038391765185487157, + "loss": 3.7002, + "step": 72540 + }, + { + "epoch": 4.928998505231689, + "grad_norm": 0.9855348467826843, + "learning_rate": 0.0003838751868460389, + "loss": 3.3514, + "step": 72545 + }, + { + "epoch": 4.9293382253023506, + "grad_norm": 0.8892372250556946, + "learning_rate": 0.00038383272183720613, + "loss": 3.5097, + "step": 72550 + }, + { + "epoch": 4.929677945373013, + "grad_norm": 0.7191735506057739, + "learning_rate": 0.0003837902568283734, + "loss": 3.5299, + "step": 72555 + }, + { + "epoch": 4.930017665443675, + "grad_norm": 0.812643826007843, + "learning_rate": 0.00038374779181954075, + "loss": 3.4148, + "step": 72560 + }, + { + "epoch": 4.930357385514336, + "grad_norm": 1.0945944786071777, + "learning_rate": 0.00038370532681070797, + "loss": 3.5501, + "step": 72565 + }, + { + "epoch": 4.930697105584998, + "grad_norm": 0.8771920204162598, + "learning_rate": 0.00038366286180187525, + "loss": 3.4961, + "step": 72570 + }, + { + "epoch": 4.93103682565566, + "grad_norm": 1.0528018474578857, + "learning_rate": 0.00038362039679304253, + "loss": 3.7025, + "step": 72575 + }, + { + "epoch": 4.931376545726321, + "grad_norm": 0.9493814706802368, + "learning_rate": 0.0003835779317842098, + "loss": 3.5905, + "step": 72580 + }, + { + "epoch": 4.931716265796983, + "grad_norm": 0.8297226428985596, + "learning_rate": 0.0003835354667753771, + "loss": 3.5571, + "step": 72585 + }, + { + "epoch": 4.932055985867645, + "grad_norm": 0.9901010990142822, + "learning_rate": 0.00038349300176654437, + "loss": 3.4256, + "step": 72590 + }, + { + "epoch": 4.932395705938307, + "grad_norm": 0.8764402270317078, + "learning_rate": 0.00038345053675771165, + "loss": 3.638, + "step": 72595 + }, + { + "epoch": 4.932735426008969, + "grad_norm": 0.9738648533821106, + "learning_rate": 0.00038340807174887893, + "loss": 3.5583, + "step": 72600 + }, + { + "epoch": 4.933075146079631, + "grad_norm": 0.7603336572647095, + "learning_rate": 0.0003833656067400462, + "loss": 3.7526, + "step": 72605 + }, + { + "epoch": 4.933414866150292, + "grad_norm": 0.8929620385169983, + "learning_rate": 0.00038332314173121344, + "loss": 3.5815, + "step": 72610 + }, + { + "epoch": 4.933754586220954, + "grad_norm": 0.8852488398551941, + "learning_rate": 0.00038328067672238077, + "loss": 3.4233, + "step": 72615 + }, + { + "epoch": 4.934094306291616, + "grad_norm": 0.791915774345398, + "learning_rate": 0.00038323821171354805, + "loss": 3.6983, + "step": 72620 + }, + { + "epoch": 4.934434026362277, + "grad_norm": 0.9024913311004639, + "learning_rate": 0.0003831957467047153, + "loss": 3.1075, + "step": 72625 + }, + { + "epoch": 4.934773746432939, + "grad_norm": 0.9006919860839844, + "learning_rate": 0.0003831532816958826, + "loss": 3.5865, + "step": 72630 + }, + { + "epoch": 4.935113466503601, + "grad_norm": 0.7446363568305969, + "learning_rate": 0.0003831108166870499, + "loss": 3.1251, + "step": 72635 + }, + { + "epoch": 4.935453186574263, + "grad_norm": 0.8133948445320129, + "learning_rate": 0.0003830683516782171, + "loss": 3.4633, + "step": 72640 + }, + { + "epoch": 4.935792906644925, + "grad_norm": 0.9396703243255615, + "learning_rate": 0.00038302588666938445, + "loss": 3.4689, + "step": 72645 + }, + { + "epoch": 4.936132626715587, + "grad_norm": 0.9660793542861938, + "learning_rate": 0.00038298342166055173, + "loss": 3.4466, + "step": 72650 + }, + { + "epoch": 4.936472346786248, + "grad_norm": 0.7830461263656616, + "learning_rate": 0.000382940956651719, + "loss": 3.2608, + "step": 72655 + }, + { + "epoch": 4.93681206685691, + "grad_norm": 0.8772631883621216, + "learning_rate": 0.00038289849164288624, + "loss": 3.6248, + "step": 72660 + }, + { + "epoch": 4.937151786927572, + "grad_norm": 0.9882521629333496, + "learning_rate": 0.0003828560266340536, + "loss": 3.5253, + "step": 72665 + }, + { + "epoch": 4.937491506998233, + "grad_norm": 0.7614195942878723, + "learning_rate": 0.00038281356162522085, + "loss": 3.4193, + "step": 72670 + }, + { + "epoch": 4.937831227068895, + "grad_norm": 0.75351881980896, + "learning_rate": 0.0003827710966163881, + "loss": 3.4155, + "step": 72675 + }, + { + "epoch": 4.938170947139557, + "grad_norm": 0.6771587133407593, + "learning_rate": 0.0003827286316075554, + "loss": 3.3386, + "step": 72680 + }, + { + "epoch": 4.938510667210219, + "grad_norm": 0.767123281955719, + "learning_rate": 0.0003826861665987227, + "loss": 3.2334, + "step": 72685 + }, + { + "epoch": 4.938850387280881, + "grad_norm": 0.8292311429977417, + "learning_rate": 0.0003826437015898899, + "loss": 3.4624, + "step": 72690 + }, + { + "epoch": 4.939190107351543, + "grad_norm": 1.039386510848999, + "learning_rate": 0.0003826012365810572, + "loss": 3.1914, + "step": 72695 + }, + { + "epoch": 4.939529827422204, + "grad_norm": 0.9476038217544556, + "learning_rate": 0.00038255877157222453, + "loss": 3.5195, + "step": 72700 + }, + { + "epoch": 4.939869547492866, + "grad_norm": 0.8351473212242126, + "learning_rate": 0.00038251630656339176, + "loss": 3.4681, + "step": 72705 + }, + { + "epoch": 4.940209267563528, + "grad_norm": 0.9407406449317932, + "learning_rate": 0.00038247384155455904, + "loss": 3.3212, + "step": 72710 + }, + { + "epoch": 4.940548987634189, + "grad_norm": 0.7730169892311096, + "learning_rate": 0.0003824313765457264, + "loss": 3.5276, + "step": 72715 + }, + { + "epoch": 4.940888707704851, + "grad_norm": 0.8480351567268372, + "learning_rate": 0.0003823889115368936, + "loss": 3.5714, + "step": 72720 + }, + { + "epoch": 4.941228427775513, + "grad_norm": 0.7709962725639343, + "learning_rate": 0.0003823464465280609, + "loss": 3.6067, + "step": 72725 + }, + { + "epoch": 4.941568147846175, + "grad_norm": 0.9260118007659912, + "learning_rate": 0.00038230398151922816, + "loss": 3.2889, + "step": 72730 + }, + { + "epoch": 4.941907867916837, + "grad_norm": 0.8809016942977905, + "learning_rate": 0.00038226151651039544, + "loss": 3.4433, + "step": 72735 + }, + { + "epoch": 4.942247587987499, + "grad_norm": 0.9832111597061157, + "learning_rate": 0.0003822190515015627, + "loss": 3.6212, + "step": 72740 + }, + { + "epoch": 4.94258730805816, + "grad_norm": 1.0022366046905518, + "learning_rate": 0.00038217658649273, + "loss": 3.4021, + "step": 72745 + }, + { + "epoch": 4.942927028128822, + "grad_norm": 0.6964163780212402, + "learning_rate": 0.0003821341214838973, + "loss": 3.3166, + "step": 72750 + }, + { + "epoch": 4.943266748199484, + "grad_norm": 1.1478503942489624, + "learning_rate": 0.00038209165647506456, + "loss": 3.5629, + "step": 72755 + }, + { + "epoch": 4.943606468270145, + "grad_norm": 1.2431979179382324, + "learning_rate": 0.00038204919146623184, + "loss": 3.4476, + "step": 72760 + }, + { + "epoch": 4.943946188340807, + "grad_norm": 0.8297187089920044, + "learning_rate": 0.00038200672645739907, + "loss": 3.6483, + "step": 72765 + }, + { + "epoch": 4.944285908411469, + "grad_norm": 2.0198402404785156, + "learning_rate": 0.0003819642614485664, + "loss": 3.3752, + "step": 72770 + }, + { + "epoch": 4.944625628482131, + "grad_norm": 0.8844216465950012, + "learning_rate": 0.0003819217964397337, + "loss": 3.6008, + "step": 72775 + }, + { + "epoch": 4.944965348552793, + "grad_norm": 0.908076822757721, + "learning_rate": 0.0003818793314309009, + "loss": 3.7295, + "step": 72780 + }, + { + "epoch": 4.945305068623455, + "grad_norm": 0.8818939328193665, + "learning_rate": 0.00038183686642206824, + "loss": 3.2834, + "step": 72785 + }, + { + "epoch": 4.945644788694116, + "grad_norm": 1.0867094993591309, + "learning_rate": 0.0003817944014132355, + "loss": 3.6106, + "step": 72790 + }, + { + "epoch": 4.945984508764778, + "grad_norm": 0.6961829662322998, + "learning_rate": 0.00038175193640440275, + "loss": 3.4984, + "step": 72795 + }, + { + "epoch": 4.94632422883544, + "grad_norm": 0.9492461085319519, + "learning_rate": 0.00038170947139557, + "loss": 3.7323, + "step": 72800 + }, + { + "epoch": 4.946663948906101, + "grad_norm": 0.9092548489570618, + "learning_rate": 0.00038166700638673736, + "loss": 3.3685, + "step": 72805 + }, + { + "epoch": 4.947003668976763, + "grad_norm": 1.0423119068145752, + "learning_rate": 0.0003816245413779046, + "loss": 3.3472, + "step": 72810 + }, + { + "epoch": 4.9473433890474245, + "grad_norm": 0.8639079928398132, + "learning_rate": 0.00038158207636907187, + "loss": 3.6149, + "step": 72815 + }, + { + "epoch": 4.947683109118087, + "grad_norm": 0.7837599515914917, + "learning_rate": 0.0003815396113602392, + "loss": 3.1586, + "step": 72820 + }, + { + "epoch": 4.948022829188749, + "grad_norm": 0.9410359263420105, + "learning_rate": 0.0003814971463514065, + "loss": 3.3679, + "step": 72825 + }, + { + "epoch": 4.94836254925941, + "grad_norm": 1.3209228515625, + "learning_rate": 0.0003814546813425737, + "loss": 3.3612, + "step": 72830 + }, + { + "epoch": 4.948702269330072, + "grad_norm": 1.0441685914993286, + "learning_rate": 0.000381412216333741, + "loss": 3.6098, + "step": 72835 + }, + { + "epoch": 4.949041989400734, + "grad_norm": 0.8973045349121094, + "learning_rate": 0.0003813697513249083, + "loss": 3.5738, + "step": 72840 + }, + { + "epoch": 4.949381709471395, + "grad_norm": 1.0114058256149292, + "learning_rate": 0.00038132728631607555, + "loss": 3.5167, + "step": 72845 + }, + { + "epoch": 4.949721429542057, + "grad_norm": 0.891346275806427, + "learning_rate": 0.00038128482130724283, + "loss": 3.5584, + "step": 72850 + }, + { + "epoch": 4.950061149612719, + "grad_norm": 0.8078318238258362, + "learning_rate": 0.00038124235629841016, + "loss": 3.0296, + "step": 72855 + }, + { + "epoch": 4.9504008696833806, + "grad_norm": 0.8377205729484558, + "learning_rate": 0.0003811998912895774, + "loss": 3.5673, + "step": 72860 + }, + { + "epoch": 4.950740589754043, + "grad_norm": 0.931049108505249, + "learning_rate": 0.00038115742628074467, + "loss": 3.6596, + "step": 72865 + }, + { + "epoch": 4.951080309824705, + "grad_norm": 0.8170061707496643, + "learning_rate": 0.000381114961271912, + "loss": 3.6792, + "step": 72870 + }, + { + "epoch": 4.951420029895366, + "grad_norm": 0.9256570935249329, + "learning_rate": 0.00038107249626307923, + "loss": 3.5823, + "step": 72875 + }, + { + "epoch": 4.951759749966028, + "grad_norm": 0.7770896553993225, + "learning_rate": 0.0003810300312542465, + "loss": 3.446, + "step": 72880 + }, + { + "epoch": 4.95209947003669, + "grad_norm": 0.8677752017974854, + "learning_rate": 0.0003809875662454138, + "loss": 3.5394, + "step": 72885 + }, + { + "epoch": 4.952439190107351, + "grad_norm": 0.8032394051551819, + "learning_rate": 0.00038094510123658107, + "loss": 3.4439, + "step": 72890 + }, + { + "epoch": 4.952778910178013, + "grad_norm": 0.8636519908905029, + "learning_rate": 0.00038090263622774835, + "loss": 3.6256, + "step": 72895 + }, + { + "epoch": 4.953118630248675, + "grad_norm": 0.9593454003334045, + "learning_rate": 0.00038086017121891563, + "loss": 3.2868, + "step": 72900 + }, + { + "epoch": 4.953458350319337, + "grad_norm": 0.8624973297119141, + "learning_rate": 0.0003808177062100829, + "loss": 3.3436, + "step": 72905 + }, + { + "epoch": 4.953798070389999, + "grad_norm": 0.789435625076294, + "learning_rate": 0.0003807752412012502, + "loss": 3.2679, + "step": 72910 + }, + { + "epoch": 4.954137790460661, + "grad_norm": 0.7102778553962708, + "learning_rate": 0.00038073277619241747, + "loss": 3.667, + "step": 72915 + }, + { + "epoch": 4.954477510531322, + "grad_norm": 1.2156705856323242, + "learning_rate": 0.0003806903111835847, + "loss": 3.5418, + "step": 72920 + }, + { + "epoch": 4.954817230601984, + "grad_norm": 0.9636473059654236, + "learning_rate": 0.00038064784617475203, + "loss": 3.6461, + "step": 72925 + }, + { + "epoch": 4.955156950672646, + "grad_norm": 0.8010683059692383, + "learning_rate": 0.0003806053811659193, + "loss": 3.5698, + "step": 72930 + }, + { + "epoch": 4.955496670743307, + "grad_norm": 0.9754721522331238, + "learning_rate": 0.00038056291615708653, + "loss": 3.3762, + "step": 72935 + }, + { + "epoch": 4.955836390813969, + "grad_norm": 1.1058012247085571, + "learning_rate": 0.00038052045114825387, + "loss": 3.7296, + "step": 72940 + }, + { + "epoch": 4.956176110884631, + "grad_norm": 1.0266730785369873, + "learning_rate": 0.00038047798613942115, + "loss": 3.3175, + "step": 72945 + }, + { + "epoch": 4.956515830955293, + "grad_norm": 1.173547387123108, + "learning_rate": 0.0003804355211305884, + "loss": 3.5551, + "step": 72950 + }, + { + "epoch": 4.956855551025955, + "grad_norm": 0.8839654326438904, + "learning_rate": 0.00038039305612175565, + "loss": 3.3509, + "step": 72955 + }, + { + "epoch": 4.957195271096617, + "grad_norm": 0.7036253809928894, + "learning_rate": 0.000380350591112923, + "loss": 3.5851, + "step": 72960 + }, + { + "epoch": 4.957534991167278, + "grad_norm": 2.0955846309661865, + "learning_rate": 0.0003803081261040902, + "loss": 3.6668, + "step": 72965 + }, + { + "epoch": 4.95787471123794, + "grad_norm": 0.8587763905525208, + "learning_rate": 0.0003802656610952575, + "loss": 3.6165, + "step": 72970 + }, + { + "epoch": 4.958214431308602, + "grad_norm": 0.7819929122924805, + "learning_rate": 0.00038022319608642483, + "loss": 3.6493, + "step": 72975 + }, + { + "epoch": 4.958554151379263, + "grad_norm": 0.8358395099639893, + "learning_rate": 0.00038018073107759206, + "loss": 3.5317, + "step": 72980 + }, + { + "epoch": 4.958893871449925, + "grad_norm": 0.923384964466095, + "learning_rate": 0.00038013826606875934, + "loss": 3.5399, + "step": 72985 + }, + { + "epoch": 4.959233591520587, + "grad_norm": 0.9469045400619507, + "learning_rate": 0.0003800958010599266, + "loss": 3.3604, + "step": 72990 + }, + { + "epoch": 4.959573311591249, + "grad_norm": 0.9319455027580261, + "learning_rate": 0.00038005333605109395, + "loss": 3.4129, + "step": 72995 + }, + { + "epoch": 4.959913031661911, + "grad_norm": 0.9811966419219971, + "learning_rate": 0.0003800108710422612, + "loss": 3.4527, + "step": 73000 + }, + { + "epoch": 4.960252751732573, + "grad_norm": 0.7282987236976624, + "learning_rate": 0.00037996840603342846, + "loss": 3.3695, + "step": 73005 + }, + { + "epoch": 4.960592471803234, + "grad_norm": 0.8049964904785156, + "learning_rate": 0.0003799259410245958, + "loss": 3.6708, + "step": 73010 + }, + { + "epoch": 4.960932191873896, + "grad_norm": 0.7169439792633057, + "learning_rate": 0.000379883476015763, + "loss": 3.2985, + "step": 73015 + }, + { + "epoch": 4.961271911944557, + "grad_norm": 0.9029344916343689, + "learning_rate": 0.0003798410110069303, + "loss": 3.7529, + "step": 73020 + }, + { + "epoch": 4.961611632015219, + "grad_norm": 1.1080904006958008, + "learning_rate": 0.0003797985459980976, + "loss": 3.4338, + "step": 73025 + }, + { + "epoch": 4.961951352085881, + "grad_norm": 0.8612424731254578, + "learning_rate": 0.00037975608098926486, + "loss": 3.4819, + "step": 73030 + }, + { + "epoch": 4.9622910721565425, + "grad_norm": 1.0260770320892334, + "learning_rate": 0.00037971361598043214, + "loss": 3.6983, + "step": 73035 + }, + { + "epoch": 4.962630792227205, + "grad_norm": 0.7672381401062012, + "learning_rate": 0.0003796711509715994, + "loss": 3.6114, + "step": 73040 + }, + { + "epoch": 4.962970512297867, + "grad_norm": 0.881515383720398, + "learning_rate": 0.0003796286859627667, + "loss": 3.6383, + "step": 73045 + }, + { + "epoch": 4.963310232368528, + "grad_norm": 0.932043731212616, + "learning_rate": 0.000379586220953934, + "loss": 3.4375, + "step": 73050 + }, + { + "epoch": 4.96364995243919, + "grad_norm": 0.7924324870109558, + "learning_rate": 0.00037954375594510126, + "loss": 3.2162, + "step": 73055 + }, + { + "epoch": 4.963989672509852, + "grad_norm": 0.9654383659362793, + "learning_rate": 0.0003795012909362685, + "loss": 3.4231, + "step": 73060 + }, + { + "epoch": 4.964329392580513, + "grad_norm": 1.075207233428955, + "learning_rate": 0.0003794588259274358, + "loss": 3.3183, + "step": 73065 + }, + { + "epoch": 4.964669112651175, + "grad_norm": 1.3674366474151611, + "learning_rate": 0.0003794163609186031, + "loss": 3.6306, + "step": 73070 + }, + { + "epoch": 4.965008832721837, + "grad_norm": 0.8143147826194763, + "learning_rate": 0.0003793738959097703, + "loss": 3.4534, + "step": 73075 + }, + { + "epoch": 4.9653485527924985, + "grad_norm": 0.8441619873046875, + "learning_rate": 0.00037933143090093766, + "loss": 3.5078, + "step": 73080 + }, + { + "epoch": 4.965688272863161, + "grad_norm": 0.8493859767913818, + "learning_rate": 0.00037928896589210494, + "loss": 3.6795, + "step": 73085 + }, + { + "epoch": 4.966027992933823, + "grad_norm": 0.7269042134284973, + "learning_rate": 0.00037924650088327216, + "loss": 3.5263, + "step": 73090 + }, + { + "epoch": 4.966367713004484, + "grad_norm": 0.9164868593215942, + "learning_rate": 0.00037920403587443944, + "loss": 3.3496, + "step": 73095 + }, + { + "epoch": 4.966707433075146, + "grad_norm": 0.8955481648445129, + "learning_rate": 0.0003791615708656068, + "loss": 3.3098, + "step": 73100 + }, + { + "epoch": 4.967047153145808, + "grad_norm": 0.6655365228652954, + "learning_rate": 0.000379119105856774, + "loss": 3.5347, + "step": 73105 + }, + { + "epoch": 4.967386873216469, + "grad_norm": 0.787166178226471, + "learning_rate": 0.0003790766408479413, + "loss": 3.3866, + "step": 73110 + }, + { + "epoch": 4.967726593287131, + "grad_norm": 0.8208472728729248, + "learning_rate": 0.0003790341758391086, + "loss": 3.4175, + "step": 73115 + }, + { + "epoch": 4.968066313357793, + "grad_norm": 0.7683771252632141, + "learning_rate": 0.00037899171083027584, + "loss": 3.7887, + "step": 73120 + }, + { + "epoch": 4.9684060334284545, + "grad_norm": 1.0410525798797607, + "learning_rate": 0.0003789492458214431, + "loss": 3.6291, + "step": 73125 + }, + { + "epoch": 4.968745753499117, + "grad_norm": 0.9626865386962891, + "learning_rate": 0.0003789067808126104, + "loss": 3.4851, + "step": 73130 + }, + { + "epoch": 4.969085473569779, + "grad_norm": 0.7364782691001892, + "learning_rate": 0.0003788643158037777, + "loss": 3.2666, + "step": 73135 + }, + { + "epoch": 4.96942519364044, + "grad_norm": 0.8087607622146606, + "learning_rate": 0.00037882185079494496, + "loss": 3.5486, + "step": 73140 + }, + { + "epoch": 4.969764913711102, + "grad_norm": 0.8406544923782349, + "learning_rate": 0.00037877938578611224, + "loss": 3.259, + "step": 73145 + }, + { + "epoch": 4.970104633781764, + "grad_norm": 0.8677491545677185, + "learning_rate": 0.0003787369207772795, + "loss": 3.7233, + "step": 73150 + }, + { + "epoch": 4.970444353852425, + "grad_norm": 0.8326682448387146, + "learning_rate": 0.0003786944557684468, + "loss": 3.4337, + "step": 73155 + }, + { + "epoch": 4.970784073923087, + "grad_norm": 0.7446854710578918, + "learning_rate": 0.0003786519907596141, + "loss": 3.6641, + "step": 73160 + }, + { + "epoch": 4.971123793993749, + "grad_norm": 1.0132875442504883, + "learning_rate": 0.0003786095257507814, + "loss": 3.667, + "step": 73165 + }, + { + "epoch": 4.971463514064411, + "grad_norm": 0.8101994395256042, + "learning_rate": 0.00037856706074194864, + "loss": 3.4631, + "step": 73170 + }, + { + "epoch": 4.971803234135073, + "grad_norm": 1.0126091241836548, + "learning_rate": 0.0003785245957331159, + "loss": 3.6341, + "step": 73175 + }, + { + "epoch": 4.972142954205735, + "grad_norm": 0.8275250196456909, + "learning_rate": 0.0003784821307242832, + "loss": 3.3744, + "step": 73180 + }, + { + "epoch": 4.972482674276396, + "grad_norm": 0.8241539001464844, + "learning_rate": 0.0003784396657154505, + "loss": 3.3362, + "step": 73185 + }, + { + "epoch": 4.972822394347058, + "grad_norm": 0.8202075958251953, + "learning_rate": 0.00037839720070661776, + "loss": 3.629, + "step": 73190 + }, + { + "epoch": 4.97316211441772, + "grad_norm": 0.9991371631622314, + "learning_rate": 0.00037835473569778504, + "loss": 3.6351, + "step": 73195 + }, + { + "epoch": 4.973501834488381, + "grad_norm": 0.8576650619506836, + "learning_rate": 0.0003783122706889523, + "loss": 3.5665, + "step": 73200 + }, + { + "epoch": 4.973841554559043, + "grad_norm": 0.8946138620376587, + "learning_rate": 0.0003782698056801196, + "loss": 3.5967, + "step": 73205 + }, + { + "epoch": 4.974181274629705, + "grad_norm": 0.7914770245552063, + "learning_rate": 0.0003782273406712869, + "loss": 3.5534, + "step": 73210 + }, + { + "epoch": 4.974520994700367, + "grad_norm": 0.993091881275177, + "learning_rate": 0.0003781848756624541, + "loss": 3.4491, + "step": 73215 + }, + { + "epoch": 4.974860714771029, + "grad_norm": 0.8123006820678711, + "learning_rate": 0.00037814241065362144, + "loss": 3.4633, + "step": 73220 + }, + { + "epoch": 4.975200434841691, + "grad_norm": 0.9387556910514832, + "learning_rate": 0.0003780999456447887, + "loss": 3.4497, + "step": 73225 + }, + { + "epoch": 4.975540154912352, + "grad_norm": 0.9280674457550049, + "learning_rate": 0.00037805748063595595, + "loss": 3.1282, + "step": 73230 + }, + { + "epoch": 4.975879874983014, + "grad_norm": 0.7899430394172668, + "learning_rate": 0.0003780150156271233, + "loss": 3.405, + "step": 73235 + }, + { + "epoch": 4.976219595053676, + "grad_norm": 0.9494094848632812, + "learning_rate": 0.00037797255061829057, + "loss": 3.2476, + "step": 73240 + }, + { + "epoch": 4.976559315124337, + "grad_norm": 1.1741997003555298, + "learning_rate": 0.0003779300856094578, + "loss": 3.6463, + "step": 73245 + }, + { + "epoch": 4.976899035194999, + "grad_norm": 0.7904856204986572, + "learning_rate": 0.00037788762060062507, + "loss": 3.769, + "step": 73250 + }, + { + "epoch": 4.977238755265661, + "grad_norm": 0.7196319103240967, + "learning_rate": 0.0003778451555917924, + "loss": 3.5531, + "step": 73255 + }, + { + "epoch": 4.977578475336323, + "grad_norm": 0.8789392709732056, + "learning_rate": 0.00037780269058295963, + "loss": 3.5042, + "step": 73260 + }, + { + "epoch": 4.977918195406985, + "grad_norm": 0.7937629222869873, + "learning_rate": 0.0003777602255741269, + "loss": 3.6949, + "step": 73265 + }, + { + "epoch": 4.978257915477647, + "grad_norm": 0.932063102722168, + "learning_rate": 0.00037771776056529425, + "loss": 3.4947, + "step": 73270 + }, + { + "epoch": 4.978597635548308, + "grad_norm": 1.0593739748001099, + "learning_rate": 0.00037767529555646147, + "loss": 3.6267, + "step": 73275 + }, + { + "epoch": 4.97893735561897, + "grad_norm": 0.9431940913200378, + "learning_rate": 0.00037763283054762875, + "loss": 3.3194, + "step": 73280 + }, + { + "epoch": 4.979277075689632, + "grad_norm": 0.8862378597259521, + "learning_rate": 0.00037759036553879603, + "loss": 3.3446, + "step": 73285 + }, + { + "epoch": 4.979616795760293, + "grad_norm": 0.7941207885742188, + "learning_rate": 0.0003775479005299633, + "loss": 3.1871, + "step": 73290 + }, + { + "epoch": 4.979956515830955, + "grad_norm": 1.252215027809143, + "learning_rate": 0.0003775054355211306, + "loss": 3.4554, + "step": 73295 + }, + { + "epoch": 4.980296235901617, + "grad_norm": 0.8796611428260803, + "learning_rate": 0.00037746297051229787, + "loss": 3.2651, + "step": 73300 + }, + { + "epoch": 4.980635955972279, + "grad_norm": 0.8147224187850952, + "learning_rate": 0.00037742050550346515, + "loss": 3.5774, + "step": 73305 + }, + { + "epoch": 4.980975676042941, + "grad_norm": 0.7524916529655457, + "learning_rate": 0.00037737804049463243, + "loss": 3.559, + "step": 73310 + }, + { + "epoch": 4.981315396113603, + "grad_norm": 0.9657232761383057, + "learning_rate": 0.0003773355754857997, + "loss": 3.53, + "step": 73315 + }, + { + "epoch": 4.981655116184264, + "grad_norm": 1.0182019472122192, + "learning_rate": 0.00037729311047696694, + "loss": 3.4873, + "step": 73320 + }, + { + "epoch": 4.981994836254926, + "grad_norm": 0.8427720665931702, + "learning_rate": 0.00037725064546813427, + "loss": 3.4541, + "step": 73325 + }, + { + "epoch": 4.982334556325588, + "grad_norm": 0.841276228427887, + "learning_rate": 0.00037720818045930155, + "loss": 3.2288, + "step": 73330 + }, + { + "epoch": 4.982674276396249, + "grad_norm": 0.9103279113769531, + "learning_rate": 0.00037716571545046883, + "loss": 3.6237, + "step": 73335 + }, + { + "epoch": 4.983013996466911, + "grad_norm": 0.8511722683906555, + "learning_rate": 0.0003771232504416361, + "loss": 3.5972, + "step": 73340 + }, + { + "epoch": 4.983353716537573, + "grad_norm": 1.1891944408416748, + "learning_rate": 0.0003770807854328034, + "loss": 3.2608, + "step": 73345 + }, + { + "epoch": 4.983693436608235, + "grad_norm": 0.8335211873054504, + "learning_rate": 0.00037703832042397067, + "loss": 3.5282, + "step": 73350 + }, + { + "epoch": 4.984033156678897, + "grad_norm": 0.7981144189834595, + "learning_rate": 0.0003769958554151379, + "loss": 3.4086, + "step": 73355 + }, + { + "epoch": 4.984372876749559, + "grad_norm": 0.9157492518424988, + "learning_rate": 0.00037695339040630523, + "loss": 3.3645, + "step": 73360 + }, + { + "epoch": 4.98471259682022, + "grad_norm": 0.8134139180183411, + "learning_rate": 0.0003769109253974725, + "loss": 3.6794, + "step": 73365 + }, + { + "epoch": 4.985052316890882, + "grad_norm": 0.7623462677001953, + "learning_rate": 0.00037686846038863974, + "loss": 3.5264, + "step": 73370 + }, + { + "epoch": 4.985392036961544, + "grad_norm": 0.8374656438827515, + "learning_rate": 0.0003768259953798071, + "loss": 3.5954, + "step": 73375 + }, + { + "epoch": 4.985731757032205, + "grad_norm": 1.047549843788147, + "learning_rate": 0.00037678353037097435, + "loss": 3.4, + "step": 73380 + }, + { + "epoch": 4.986071477102867, + "grad_norm": 0.9485633969306946, + "learning_rate": 0.0003767410653621416, + "loss": 3.2877, + "step": 73385 + }, + { + "epoch": 4.986411197173529, + "grad_norm": 0.8473796844482422, + "learning_rate": 0.00037669860035330886, + "loss": 3.4551, + "step": 73390 + }, + { + "epoch": 4.986750917244191, + "grad_norm": 0.8003768920898438, + "learning_rate": 0.0003766561353444762, + "loss": 3.2913, + "step": 73395 + }, + { + "epoch": 4.987090637314853, + "grad_norm": 0.8610634207725525, + "learning_rate": 0.0003766136703356434, + "loss": 3.4413, + "step": 73400 + }, + { + "epoch": 4.987430357385515, + "grad_norm": 0.8350740671157837, + "learning_rate": 0.0003765712053268107, + "loss": 3.5257, + "step": 73405 + }, + { + "epoch": 4.987770077456176, + "grad_norm": 0.9380072355270386, + "learning_rate": 0.00037652874031797803, + "loss": 3.5897, + "step": 73410 + }, + { + "epoch": 4.988109797526838, + "grad_norm": 1.0014680624008179, + "learning_rate": 0.00037648627530914526, + "loss": 3.3416, + "step": 73415 + }, + { + "epoch": 4.9884495175975, + "grad_norm": 0.8285419344902039, + "learning_rate": 0.00037644381030031254, + "loss": 3.5234, + "step": 73420 + }, + { + "epoch": 4.988789237668161, + "grad_norm": 0.9550110101699829, + "learning_rate": 0.0003764013452914798, + "loss": 3.6373, + "step": 73425 + }, + { + "epoch": 4.989128957738823, + "grad_norm": 1.15923011302948, + "learning_rate": 0.0003763588802826471, + "loss": 3.5653, + "step": 73430 + }, + { + "epoch": 4.989468677809485, + "grad_norm": 0.9602484107017517, + "learning_rate": 0.0003763164152738144, + "loss": 3.6752, + "step": 73435 + }, + { + "epoch": 4.989808397880147, + "grad_norm": 0.7451989650726318, + "learning_rate": 0.00037627395026498166, + "loss": 3.548, + "step": 73440 + }, + { + "epoch": 4.990148117950809, + "grad_norm": 0.907376766204834, + "learning_rate": 0.00037623148525614894, + "loss": 3.6844, + "step": 73445 + }, + { + "epoch": 4.990487838021471, + "grad_norm": 0.7486577033996582, + "learning_rate": 0.0003761890202473162, + "loss": 3.2842, + "step": 73450 + }, + { + "epoch": 4.990827558092132, + "grad_norm": 1.1977856159210205, + "learning_rate": 0.0003761465552384835, + "loss": 3.3526, + "step": 73455 + }, + { + "epoch": 4.991167278162794, + "grad_norm": 0.8403291702270508, + "learning_rate": 0.0003761040902296507, + "loss": 3.5536, + "step": 73460 + }, + { + "epoch": 4.991506998233456, + "grad_norm": 1.088760495185852, + "learning_rate": 0.00037606162522081806, + "loss": 3.2718, + "step": 73465 + }, + { + "epoch": 4.991846718304117, + "grad_norm": 0.9433627128601074, + "learning_rate": 0.00037601916021198534, + "loss": 3.4703, + "step": 73470 + }, + { + "epoch": 4.992186438374779, + "grad_norm": 0.8565873503684998, + "learning_rate": 0.00037597669520315257, + "loss": 3.547, + "step": 73475 + }, + { + "epoch": 4.9925261584454415, + "grad_norm": 0.7623013257980347, + "learning_rate": 0.0003759342301943199, + "loss": 3.4805, + "step": 73480 + }, + { + "epoch": 4.992865878516103, + "grad_norm": 0.7753994464874268, + "learning_rate": 0.0003758917651854872, + "loss": 3.6177, + "step": 73485 + }, + { + "epoch": 4.993205598586765, + "grad_norm": 0.9224596619606018, + "learning_rate": 0.0003758493001766544, + "loss": 3.5296, + "step": 73490 + }, + { + "epoch": 4.993545318657426, + "grad_norm": 0.7430421710014343, + "learning_rate": 0.00037580683516782174, + "loss": 3.6136, + "step": 73495 + }, + { + "epoch": 4.993885038728088, + "grad_norm": 0.8180851936340332, + "learning_rate": 0.000375764370158989, + "loss": 3.3634, + "step": 73500 + }, + { + "epoch": 4.99422475879875, + "grad_norm": 0.9090211987495422, + "learning_rate": 0.0003757219051501563, + "loss": 3.3945, + "step": 73505 + }, + { + "epoch": 4.994564478869411, + "grad_norm": 0.8929524421691895, + "learning_rate": 0.0003756794401413235, + "loss": 3.6991, + "step": 73510 + }, + { + "epoch": 4.994904198940073, + "grad_norm": 0.9579527378082275, + "learning_rate": 0.00037563697513249086, + "loss": 3.5372, + "step": 73515 + }, + { + "epoch": 4.995243919010735, + "grad_norm": 1.0293318033218384, + "learning_rate": 0.00037559451012365814, + "loss": 3.4771, + "step": 73520 + }, + { + "epoch": 4.995583639081397, + "grad_norm": 0.9263602495193481, + "learning_rate": 0.00037555204511482537, + "loss": 3.3371, + "step": 73525 + }, + { + "epoch": 4.995923359152059, + "grad_norm": 0.9338143467903137, + "learning_rate": 0.0003755095801059927, + "loss": 3.4752, + "step": 73530 + }, + { + "epoch": 4.996263079222721, + "grad_norm": 0.8178796768188477, + "learning_rate": 0.00037546711509716, + "loss": 3.364, + "step": 73535 + }, + { + "epoch": 4.996602799293382, + "grad_norm": 0.8574473857879639, + "learning_rate": 0.0003754246500883272, + "loss": 3.5046, + "step": 73540 + }, + { + "epoch": 4.996942519364044, + "grad_norm": 0.8701178431510925, + "learning_rate": 0.0003753821850794945, + "loss": 3.6173, + "step": 73545 + }, + { + "epoch": 4.997282239434706, + "grad_norm": 1.0161793231964111, + "learning_rate": 0.0003753397200706618, + "loss": 3.5068, + "step": 73550 + }, + { + "epoch": 4.997621959505367, + "grad_norm": 0.870212972164154, + "learning_rate": 0.00037529725506182905, + "loss": 3.556, + "step": 73555 + }, + { + "epoch": 4.997961679576029, + "grad_norm": 0.8367553949356079, + "learning_rate": 0.00037525479005299633, + "loss": 3.6131, + "step": 73560 + }, + { + "epoch": 4.998301399646691, + "grad_norm": 1.0269298553466797, + "learning_rate": 0.00037521232504416366, + "loss": 3.2674, + "step": 73565 + }, + { + "epoch": 4.998641119717353, + "grad_norm": 0.7627138495445251, + "learning_rate": 0.0003751698600353309, + "loss": 3.4733, + "step": 73570 + }, + { + "epoch": 4.998980839788015, + "grad_norm": 0.862865149974823, + "learning_rate": 0.00037512739502649817, + "loss": 3.3547, + "step": 73575 + }, + { + "epoch": 4.999320559858677, + "grad_norm": 0.7803202867507935, + "learning_rate": 0.00037508493001766545, + "loss": 3.4797, + "step": 73580 + }, + { + "epoch": 4.999660279929338, + "grad_norm": 1.0289307832717896, + "learning_rate": 0.00037504246500883273, + "loss": 3.6836, + "step": 73585 + }, + { + "epoch": 5.0, + "grad_norm": 2.0020124912261963, + "learning_rate": 0.000375, + "loss": 3.4413, + "step": 73590 + }, + { + "epoch": 5.0, + "eval_bertscore": { + "f1": 0.8397619288343597, + "precision": 0.8431473464822108, + "recall": 0.8371518092719676 + }, + "eval_bleu_4": 0.016062640135109264, + "eval_exact_match": 0.00038763446070355656, + "eval_loss": 3.414961099624634, + "eval_meteor": 0.08555663894503344, + "eval_rouge": { + "rouge1": 0.12268302599839292, + "rouge2": 0.018551820188759085, + "rougeL": 0.10623085847093988, + "rougeLsum": 0.10624618490200036 + }, + "eval_runtime": 2350.1573, + "eval_samples_per_second": 4.391, + "eval_steps_per_second": 0.549, + "step": 73590 + }, + { + "epoch": 5.000339720070662, + "grad_norm": 0.9625300765037537, + "learning_rate": 0.0003749575349911673, + "loss": 3.4398, + "step": 73595 + }, + { + "epoch": 5.000679440141323, + "grad_norm": 0.8478984832763672, + "learning_rate": 0.00037491506998233457, + "loss": 3.3439, + "step": 73600 + }, + { + "epoch": 5.001019160211985, + "grad_norm": 0.8582783937454224, + "learning_rate": 0.00037487260497350185, + "loss": 3.598, + "step": 73605 + }, + { + "epoch": 5.001358880282647, + "grad_norm": 1.2371875047683716, + "learning_rate": 0.00037483013996466913, + "loss": 3.6129, + "step": 73610 + }, + { + "epoch": 5.001698600353309, + "grad_norm": 0.9397746920585632, + "learning_rate": 0.00037478767495583635, + "loss": 3.4562, + "step": 73615 + }, + { + "epoch": 5.002038320423971, + "grad_norm": 0.9677167534828186, + "learning_rate": 0.0003747452099470037, + "loss": 3.4077, + "step": 73620 + }, + { + "epoch": 5.002378040494633, + "grad_norm": 0.8125355243682861, + "learning_rate": 0.00037470274493817097, + "loss": 3.5458, + "step": 73625 + }, + { + "epoch": 5.002717760565294, + "grad_norm": 0.7614694237709045, + "learning_rate": 0.0003746602799293382, + "loss": 3.2286, + "step": 73630 + }, + { + "epoch": 5.003057480635956, + "grad_norm": 0.7842497825622559, + "learning_rate": 0.00037461781492050553, + "loss": 3.5837, + "step": 73635 + }, + { + "epoch": 5.003397200706618, + "grad_norm": 0.9299313426017761, + "learning_rate": 0.0003745753499116728, + "loss": 3.2729, + "step": 73640 + }, + { + "epoch": 5.003736920777279, + "grad_norm": 0.8807242512702942, + "learning_rate": 0.00037453288490284003, + "loss": 3.3807, + "step": 73645 + }, + { + "epoch": 5.004076640847941, + "grad_norm": 1.0075852870941162, + "learning_rate": 0.0003744904198940073, + "loss": 3.4866, + "step": 73650 + }, + { + "epoch": 5.004416360918603, + "grad_norm": 0.8257735967636108, + "learning_rate": 0.00037444795488517465, + "loss": 3.5268, + "step": 73655 + }, + { + "epoch": 5.004756080989265, + "grad_norm": 0.9911938309669495, + "learning_rate": 0.0003744054898763419, + "loss": 3.52, + "step": 73660 + }, + { + "epoch": 5.005095801059927, + "grad_norm": 1.0119799375534058, + "learning_rate": 0.00037436302486750915, + "loss": 3.51, + "step": 73665 + }, + { + "epoch": 5.005435521130589, + "grad_norm": 0.9239062070846558, + "learning_rate": 0.0003743205598586765, + "loss": 3.2797, + "step": 73670 + }, + { + "epoch": 5.00577524120125, + "grad_norm": 0.657340943813324, + "learning_rate": 0.0003742780948498437, + "loss": 3.5172, + "step": 73675 + }, + { + "epoch": 5.006114961271912, + "grad_norm": 0.7681381702423096, + "learning_rate": 0.000374235629841011, + "loss": 3.4016, + "step": 73680 + }, + { + "epoch": 5.006454681342574, + "grad_norm": 0.8912808299064636, + "learning_rate": 0.0003741931648321783, + "loss": 3.5468, + "step": 73685 + }, + { + "epoch": 5.006794401413235, + "grad_norm": 1.073547124862671, + "learning_rate": 0.0003741506998233456, + "loss": 3.3653, + "step": 73690 + }, + { + "epoch": 5.007134121483897, + "grad_norm": 1.0875983238220215, + "learning_rate": 0.00037410823481451284, + "loss": 3.4569, + "step": 73695 + }, + { + "epoch": 5.007473841554559, + "grad_norm": 0.8415685296058655, + "learning_rate": 0.0003740657698056801, + "loss": 3.3325, + "step": 73700 + }, + { + "epoch": 5.007813561625221, + "grad_norm": 0.8602301478385925, + "learning_rate": 0.00037402330479684745, + "loss": 3.5463, + "step": 73705 + }, + { + "epoch": 5.008153281695883, + "grad_norm": 1.1192108392715454, + "learning_rate": 0.0003739808397880147, + "loss": 3.4879, + "step": 73710 + }, + { + "epoch": 5.008493001766545, + "grad_norm": 0.9308758974075317, + "learning_rate": 0.00037393837477918196, + "loss": 3.5845, + "step": 73715 + }, + { + "epoch": 5.008832721837206, + "grad_norm": 0.8004947304725647, + "learning_rate": 0.0003738959097703493, + "loss": 3.4369, + "step": 73720 + }, + { + "epoch": 5.009172441907868, + "grad_norm": 0.7972015142440796, + "learning_rate": 0.0003738534447615165, + "loss": 3.2662, + "step": 73725 + }, + { + "epoch": 5.00951216197853, + "grad_norm": 1.042711853981018, + "learning_rate": 0.0003738109797526838, + "loss": 3.2658, + "step": 73730 + }, + { + "epoch": 5.009851882049191, + "grad_norm": 0.7924750447273254, + "learning_rate": 0.0003737685147438511, + "loss": 3.4404, + "step": 73735 + }, + { + "epoch": 5.010191602119853, + "grad_norm": 0.7492609024047852, + "learning_rate": 0.00037372604973501836, + "loss": 3.4125, + "step": 73740 + }, + { + "epoch": 5.0105313221905154, + "grad_norm": 0.7353882789611816, + "learning_rate": 0.00037368358472618564, + "loss": 3.2018, + "step": 73745 + }, + { + "epoch": 5.010871042261177, + "grad_norm": 0.8370790481567383, + "learning_rate": 0.0003736411197173529, + "loss": 3.1775, + "step": 73750 + }, + { + "epoch": 5.011210762331839, + "grad_norm": 0.9480472803115845, + "learning_rate": 0.0003735986547085202, + "loss": 3.3791, + "step": 73755 + }, + { + "epoch": 5.0115504824025, + "grad_norm": 0.864469051361084, + "learning_rate": 0.0003735561896996875, + "loss": 3.4058, + "step": 73760 + }, + { + "epoch": 5.011890202473162, + "grad_norm": 1.7910126447677612, + "learning_rate": 0.00037351372469085476, + "loss": 3.4723, + "step": 73765 + }, + { + "epoch": 5.012229922543824, + "grad_norm": 1.1197491884231567, + "learning_rate": 0.000373471259682022, + "loss": 3.1865, + "step": 73770 + }, + { + "epoch": 5.012569642614485, + "grad_norm": 0.8969390392303467, + "learning_rate": 0.0003734287946731893, + "loss": 3.3448, + "step": 73775 + }, + { + "epoch": 5.012909362685147, + "grad_norm": 0.9955067038536072, + "learning_rate": 0.0003733863296643566, + "loss": 3.4922, + "step": 73780 + }, + { + "epoch": 5.013249082755809, + "grad_norm": 0.7771676182746887, + "learning_rate": 0.0003733438646555238, + "loss": 3.5055, + "step": 73785 + }, + { + "epoch": 5.013588802826471, + "grad_norm": 0.799666702747345, + "learning_rate": 0.00037330139964669116, + "loss": 3.3613, + "step": 73790 + }, + { + "epoch": 5.013928522897133, + "grad_norm": 0.9385126829147339, + "learning_rate": 0.00037325893463785844, + "loss": 3.516, + "step": 73795 + }, + { + "epoch": 5.014268242967795, + "grad_norm": 0.9211181402206421, + "learning_rate": 0.00037321646962902566, + "loss": 3.4719, + "step": 73800 + }, + { + "epoch": 5.014607963038456, + "grad_norm": 0.9400629997253418, + "learning_rate": 0.00037317400462019294, + "loss": 3.1372, + "step": 73805 + }, + { + "epoch": 5.014947683109118, + "grad_norm": 0.9275919198989868, + "learning_rate": 0.0003731315396113603, + "loss": 3.5, + "step": 73810 + }, + { + "epoch": 5.01528740317978, + "grad_norm": 1.0918574333190918, + "learning_rate": 0.0003730890746025275, + "loss": 3.3463, + "step": 73815 + }, + { + "epoch": 5.015627123250441, + "grad_norm": 1.045360803604126, + "learning_rate": 0.0003730466095936948, + "loss": 3.509, + "step": 73820 + }, + { + "epoch": 5.015966843321103, + "grad_norm": 0.8133660554885864, + "learning_rate": 0.0003730041445848621, + "loss": 3.2984, + "step": 73825 + }, + { + "epoch": 5.016306563391765, + "grad_norm": 0.8697634339332581, + "learning_rate": 0.00037296167957602934, + "loss": 3.4741, + "step": 73830 + }, + { + "epoch": 5.016646283462427, + "grad_norm": 0.8753909468650818, + "learning_rate": 0.0003729192145671966, + "loss": 3.4926, + "step": 73835 + }, + { + "epoch": 5.016986003533089, + "grad_norm": 0.8569296002388, + "learning_rate": 0.0003728767495583639, + "loss": 3.2955, + "step": 73840 + }, + { + "epoch": 5.017325723603751, + "grad_norm": 0.8232488036155701, + "learning_rate": 0.0003728342845495312, + "loss": 3.203, + "step": 73845 + }, + { + "epoch": 5.017665443674412, + "grad_norm": 1.2155956029891968, + "learning_rate": 0.00037279181954069846, + "loss": 3.3313, + "step": 73850 + }, + { + "epoch": 5.018005163745074, + "grad_norm": 1.01833176612854, + "learning_rate": 0.00037274935453186574, + "loss": 3.4159, + "step": 73855 + }, + { + "epoch": 5.018344883815736, + "grad_norm": 0.7775917053222656, + "learning_rate": 0.0003727068895230331, + "loss": 3.3012, + "step": 73860 + }, + { + "epoch": 5.018684603886397, + "grad_norm": 0.9397168755531311, + "learning_rate": 0.0003726644245142003, + "loss": 3.3867, + "step": 73865 + }, + { + "epoch": 5.019024323957059, + "grad_norm": 0.8575489521026611, + "learning_rate": 0.0003726219595053676, + "loss": 3.4924, + "step": 73870 + }, + { + "epoch": 5.019364044027721, + "grad_norm": 0.8601073622703552, + "learning_rate": 0.00037257949449653486, + "loss": 3.5496, + "step": 73875 + }, + { + "epoch": 5.019703764098383, + "grad_norm": 0.8570936918258667, + "learning_rate": 0.00037253702948770214, + "loss": 3.3207, + "step": 73880 + }, + { + "epoch": 5.020043484169045, + "grad_norm": 1.0787328481674194, + "learning_rate": 0.0003724945644788694, + "loss": 3.3904, + "step": 73885 + }, + { + "epoch": 5.020383204239707, + "grad_norm": 1.147847294807434, + "learning_rate": 0.0003724520994700367, + "loss": 3.6, + "step": 73890 + }, + { + "epoch": 5.020722924310368, + "grad_norm": 0.9208913445472717, + "learning_rate": 0.000372409634461204, + "loss": 3.2863, + "step": 73895 + }, + { + "epoch": 5.02106264438103, + "grad_norm": 1.4002976417541504, + "learning_rate": 0.00037236716945237126, + "loss": 3.5237, + "step": 73900 + }, + { + "epoch": 5.021402364451692, + "grad_norm": 0.8220402002334595, + "learning_rate": 0.00037232470444353854, + "loss": 3.5198, + "step": 73905 + }, + { + "epoch": 5.021742084522353, + "grad_norm": 1.1295868158340454, + "learning_rate": 0.00037228223943470577, + "loss": 3.4473, + "step": 73910 + }, + { + "epoch": 5.022081804593015, + "grad_norm": 1.1026285886764526, + "learning_rate": 0.0003722397744258731, + "loss": 3.4684, + "step": 73915 + }, + { + "epoch": 5.022421524663677, + "grad_norm": 0.9584364891052246, + "learning_rate": 0.0003721973094170404, + "loss": 3.4253, + "step": 73920 + }, + { + "epoch": 5.022761244734339, + "grad_norm": 1.0277055501937866, + "learning_rate": 0.0003721548444082076, + "loss": 3.3799, + "step": 73925 + }, + { + "epoch": 5.023100964805001, + "grad_norm": 0.6846737861633301, + "learning_rate": 0.00037211237939937494, + "loss": 3.3739, + "step": 73930 + }, + { + "epoch": 5.023440684875663, + "grad_norm": 0.9156801700592041, + "learning_rate": 0.0003720699143905422, + "loss": 3.7134, + "step": 73935 + }, + { + "epoch": 5.023780404946324, + "grad_norm": 0.8014858961105347, + "learning_rate": 0.00037202744938170945, + "loss": 3.2951, + "step": 73940 + }, + { + "epoch": 5.024120125016986, + "grad_norm": 0.7329127788543701, + "learning_rate": 0.00037198498437287673, + "loss": 3.236, + "step": 73945 + }, + { + "epoch": 5.024459845087648, + "grad_norm": 0.7832569479942322, + "learning_rate": 0.00037194251936404406, + "loss": 3.21, + "step": 73950 + }, + { + "epoch": 5.024799565158309, + "grad_norm": 1.0011848211288452, + "learning_rate": 0.0003719000543552113, + "loss": 3.3147, + "step": 73955 + }, + { + "epoch": 5.025139285228971, + "grad_norm": 0.8342908620834351, + "learning_rate": 0.00037185758934637857, + "loss": 3.482, + "step": 73960 + }, + { + "epoch": 5.025479005299633, + "grad_norm": 0.8564026355743408, + "learning_rate": 0.0003718151243375459, + "loss": 3.674, + "step": 73965 + }, + { + "epoch": 5.025818725370295, + "grad_norm": 0.9077079892158508, + "learning_rate": 0.00037177265932871313, + "loss": 3.5887, + "step": 73970 + }, + { + "epoch": 5.026158445440957, + "grad_norm": 0.7185171842575073, + "learning_rate": 0.0003717301943198804, + "loss": 3.4358, + "step": 73975 + }, + { + "epoch": 5.026498165511619, + "grad_norm": 0.9403607845306396, + "learning_rate": 0.0003716877293110477, + "loss": 3.4459, + "step": 73980 + }, + { + "epoch": 5.02683788558228, + "grad_norm": 0.8399714231491089, + "learning_rate": 0.00037164526430221497, + "loss": 3.344, + "step": 73985 + }, + { + "epoch": 5.027177605652942, + "grad_norm": 0.8505078554153442, + "learning_rate": 0.00037160279929338225, + "loss": 3.3943, + "step": 73990 + }, + { + "epoch": 5.027517325723604, + "grad_norm": 0.6879952549934387, + "learning_rate": 0.00037156033428454953, + "loss": 3.3964, + "step": 73995 + }, + { + "epoch": 5.027857045794265, + "grad_norm": 0.8246248960494995, + "learning_rate": 0.0003715178692757168, + "loss": 3.2592, + "step": 74000 + }, + { + "epoch": 5.028196765864927, + "grad_norm": 0.9673071503639221, + "learning_rate": 0.0003714754042668841, + "loss": 3.6342, + "step": 74005 + }, + { + "epoch": 5.028536485935589, + "grad_norm": 1.0281087160110474, + "learning_rate": 0.00037143293925805137, + "loss": 3.4217, + "step": 74010 + }, + { + "epoch": 5.028876206006251, + "grad_norm": 1.1327872276306152, + "learning_rate": 0.0003713904742492186, + "loss": 3.4083, + "step": 74015 + }, + { + "epoch": 5.029215926076913, + "grad_norm": 1.0981696844100952, + "learning_rate": 0.00037134800924038593, + "loss": 3.2709, + "step": 74020 + }, + { + "epoch": 5.029555646147575, + "grad_norm": 0.9994499683380127, + "learning_rate": 0.0003713055442315532, + "loss": 3.6781, + "step": 74025 + }, + { + "epoch": 5.029895366218236, + "grad_norm": 0.8725029230117798, + "learning_rate": 0.0003712630792227205, + "loss": 3.4174, + "step": 74030 + }, + { + "epoch": 5.030235086288898, + "grad_norm": 0.9097709059715271, + "learning_rate": 0.00037122061421388777, + "loss": 3.4403, + "step": 74035 + }, + { + "epoch": 5.03057480635956, + "grad_norm": 0.9016314148902893, + "learning_rate": 0.00037117814920505505, + "loss": 3.3919, + "step": 74040 + }, + { + "epoch": 5.030914526430221, + "grad_norm": 1.0539602041244507, + "learning_rate": 0.00037113568419622233, + "loss": 3.2056, + "step": 74045 + }, + { + "epoch": 5.031254246500883, + "grad_norm": 1.0248746871948242, + "learning_rate": 0.0003710932191873896, + "loss": 3.7167, + "step": 74050 + }, + { + "epoch": 5.0315939665715455, + "grad_norm": 1.0809632539749146, + "learning_rate": 0.0003710507541785569, + "loss": 3.0941, + "step": 74055 + }, + { + "epoch": 5.031933686642207, + "grad_norm": 1.271547555923462, + "learning_rate": 0.00037100828916972417, + "loss": 3.168, + "step": 74060 + }, + { + "epoch": 5.032273406712869, + "grad_norm": 0.8147267699241638, + "learning_rate": 0.0003709658241608914, + "loss": 3.606, + "step": 74065 + }, + { + "epoch": 5.032613126783531, + "grad_norm": 1.0186887979507446, + "learning_rate": 0.00037092335915205873, + "loss": 3.2369, + "step": 74070 + }, + { + "epoch": 5.032952846854192, + "grad_norm": 1.0792646408081055, + "learning_rate": 0.000370880894143226, + "loss": 3.3055, + "step": 74075 + }, + { + "epoch": 5.033292566924854, + "grad_norm": 1.4125255346298218, + "learning_rate": 0.00037083842913439324, + "loss": 3.1663, + "step": 74080 + }, + { + "epoch": 5.033632286995516, + "grad_norm": 0.9496142864227295, + "learning_rate": 0.00037079596412556057, + "loss": 3.4917, + "step": 74085 + }, + { + "epoch": 5.033972007066177, + "grad_norm": 0.9657579660415649, + "learning_rate": 0.00037075349911672785, + "loss": 3.3606, + "step": 74090 + }, + { + "epoch": 5.034311727136839, + "grad_norm": 0.7909177541732788, + "learning_rate": 0.0003707110341078951, + "loss": 3.7262, + "step": 74095 + }, + { + "epoch": 5.0346514472075015, + "grad_norm": 0.8715478181838989, + "learning_rate": 0.00037066856909906236, + "loss": 3.5546, + "step": 74100 + }, + { + "epoch": 5.034991167278163, + "grad_norm": 0.9149139523506165, + "learning_rate": 0.0003706261040902297, + "loss": 3.4745, + "step": 74105 + }, + { + "epoch": 5.035330887348825, + "grad_norm": 0.9541917443275452, + "learning_rate": 0.0003705836390813969, + "loss": 3.4438, + "step": 74110 + }, + { + "epoch": 5.035670607419486, + "grad_norm": 0.804902970790863, + "learning_rate": 0.0003705411740725642, + "loss": 3.8125, + "step": 74115 + }, + { + "epoch": 5.036010327490148, + "grad_norm": 0.6748700141906738, + "learning_rate": 0.00037049870906373153, + "loss": 3.1888, + "step": 74120 + }, + { + "epoch": 5.03635004756081, + "grad_norm": 0.9283285737037659, + "learning_rate": 0.00037045624405489876, + "loss": 3.5064, + "step": 74125 + }, + { + "epoch": 5.036689767631471, + "grad_norm": 0.902118444442749, + "learning_rate": 0.00037041377904606604, + "loss": 3.6009, + "step": 74130 + }, + { + "epoch": 5.037029487702133, + "grad_norm": 1.2101463079452515, + "learning_rate": 0.0003703713140372333, + "loss": 3.317, + "step": 74135 + }, + { + "epoch": 5.037369207772795, + "grad_norm": 0.8732280731201172, + "learning_rate": 0.0003703288490284006, + "loss": 3.5872, + "step": 74140 + }, + { + "epoch": 5.037708927843457, + "grad_norm": 0.9206206202507019, + "learning_rate": 0.0003702863840195679, + "loss": 3.5645, + "step": 74145 + }, + { + "epoch": 5.038048647914119, + "grad_norm": 0.7951322197914124, + "learning_rate": 0.00037024391901073516, + "loss": 3.4903, + "step": 74150 + }, + { + "epoch": 5.038388367984781, + "grad_norm": 0.7873854041099548, + "learning_rate": 0.00037020145400190244, + "loss": 3.5114, + "step": 74155 + }, + { + "epoch": 5.038728088055442, + "grad_norm": 0.9790129661560059, + "learning_rate": 0.0003701589889930697, + "loss": 3.2833, + "step": 74160 + }, + { + "epoch": 5.039067808126104, + "grad_norm": 0.9435854554176331, + "learning_rate": 0.000370116523984237, + "loss": 3.8741, + "step": 74165 + }, + { + "epoch": 5.039407528196766, + "grad_norm": 0.8869902491569519, + "learning_rate": 0.0003700740589754042, + "loss": 3.4746, + "step": 74170 + }, + { + "epoch": 5.039747248267427, + "grad_norm": 1.0121889114379883, + "learning_rate": 0.00037003159396657156, + "loss": 3.7717, + "step": 74175 + }, + { + "epoch": 5.040086968338089, + "grad_norm": 0.8299087285995483, + "learning_rate": 0.00036998912895773884, + "loss": 3.4276, + "step": 74180 + }, + { + "epoch": 5.040426688408751, + "grad_norm": 0.7216662168502808, + "learning_rate": 0.00036994666394890607, + "loss": 3.3851, + "step": 74185 + }, + { + "epoch": 5.040766408479413, + "grad_norm": 0.7863749265670776, + "learning_rate": 0.0003699041989400734, + "loss": 3.5774, + "step": 74190 + }, + { + "epoch": 5.041106128550075, + "grad_norm": 0.8873997926712036, + "learning_rate": 0.0003698617339312407, + "loss": 3.4276, + "step": 74195 + }, + { + "epoch": 5.041445848620737, + "grad_norm": 0.9233851432800293, + "learning_rate": 0.00036981926892240796, + "loss": 3.6503, + "step": 74200 + }, + { + "epoch": 5.041785568691398, + "grad_norm": 0.7669897079467773, + "learning_rate": 0.0003697768039135752, + "loss": 3.3367, + "step": 74205 + }, + { + "epoch": 5.04212528876206, + "grad_norm": 1.0702768564224243, + "learning_rate": 0.0003697343389047425, + "loss": 3.7899, + "step": 74210 + }, + { + "epoch": 5.042465008832722, + "grad_norm": 0.8733883500099182, + "learning_rate": 0.0003696918738959098, + "loss": 3.4937, + "step": 74215 + }, + { + "epoch": 5.042804728903383, + "grad_norm": 0.8090338706970215, + "learning_rate": 0.000369649408887077, + "loss": 3.6192, + "step": 74220 + }, + { + "epoch": 5.043144448974045, + "grad_norm": 1.8956526517868042, + "learning_rate": 0.00036960694387824436, + "loss": 3.365, + "step": 74225 + }, + { + "epoch": 5.043484169044707, + "grad_norm": 1.0194029808044434, + "learning_rate": 0.00036956447886941164, + "loss": 3.6157, + "step": 74230 + }, + { + "epoch": 5.043823889115369, + "grad_norm": 0.8100787997245789, + "learning_rate": 0.00036952201386057887, + "loss": 3.2819, + "step": 74235 + }, + { + "epoch": 5.044163609186031, + "grad_norm": 1.025274395942688, + "learning_rate": 0.00036947954885174615, + "loss": 3.1214, + "step": 74240 + }, + { + "epoch": 5.044503329256693, + "grad_norm": 1.0486228466033936, + "learning_rate": 0.0003694370838429135, + "loss": 3.3469, + "step": 74245 + }, + { + "epoch": 5.044843049327354, + "grad_norm": 0.9265202879905701, + "learning_rate": 0.0003693946188340807, + "loss": 3.3911, + "step": 74250 + }, + { + "epoch": 5.045182769398016, + "grad_norm": 0.8730506300926208, + "learning_rate": 0.000369352153825248, + "loss": 3.645, + "step": 74255 + }, + { + "epoch": 5.045522489468678, + "grad_norm": 0.9687416553497314, + "learning_rate": 0.0003693096888164153, + "loss": 3.4244, + "step": 74260 + }, + { + "epoch": 5.045862209539339, + "grad_norm": 0.9287399649620056, + "learning_rate": 0.00036926722380758255, + "loss": 3.6303, + "step": 74265 + }, + { + "epoch": 5.046201929610001, + "grad_norm": 0.856453537940979, + "learning_rate": 0.00036922475879874983, + "loss": 3.4933, + "step": 74270 + }, + { + "epoch": 5.046541649680663, + "grad_norm": 1.32974374294281, + "learning_rate": 0.0003691822937899171, + "loss": 3.4778, + "step": 74275 + }, + { + "epoch": 5.046881369751325, + "grad_norm": 0.93451327085495, + "learning_rate": 0.0003691398287810844, + "loss": 3.1467, + "step": 74280 + }, + { + "epoch": 5.047221089821987, + "grad_norm": 0.8263443112373352, + "learning_rate": 0.00036909736377225167, + "loss": 3.2399, + "step": 74285 + }, + { + "epoch": 5.047560809892649, + "grad_norm": 0.7969293594360352, + "learning_rate": 0.00036905489876341895, + "loss": 3.489, + "step": 74290 + }, + { + "epoch": 5.04790052996331, + "grad_norm": 0.8431703448295593, + "learning_rate": 0.00036901243375458623, + "loss": 3.3196, + "step": 74295 + }, + { + "epoch": 5.048240250033972, + "grad_norm": 0.7845717072486877, + "learning_rate": 0.0003689699687457535, + "loss": 3.7004, + "step": 74300 + }, + { + "epoch": 5.048579970104634, + "grad_norm": 1.1778630018234253, + "learning_rate": 0.0003689275037369208, + "loss": 3.5474, + "step": 74305 + }, + { + "epoch": 5.048919690175295, + "grad_norm": 0.8924747705459595, + "learning_rate": 0.000368885038728088, + "loss": 3.8841, + "step": 74310 + }, + { + "epoch": 5.049259410245957, + "grad_norm": 0.984935462474823, + "learning_rate": 0.00036884257371925535, + "loss": 3.5027, + "step": 74315 + }, + { + "epoch": 5.0495991303166194, + "grad_norm": 0.8996263146400452, + "learning_rate": 0.00036880010871042263, + "loss": 3.6692, + "step": 74320 + }, + { + "epoch": 5.049938850387281, + "grad_norm": 0.6358514428138733, + "learning_rate": 0.00036875764370158985, + "loss": 3.4147, + "step": 74325 + }, + { + "epoch": 5.050278570457943, + "grad_norm": 0.8601886034011841, + "learning_rate": 0.0003687151786927572, + "loss": 3.1693, + "step": 74330 + }, + { + "epoch": 5.050618290528605, + "grad_norm": 0.9155727028846741, + "learning_rate": 0.00036867271368392447, + "loss": 3.5716, + "step": 74335 + }, + { + "epoch": 5.050958010599266, + "grad_norm": 0.9294515252113342, + "learning_rate": 0.0003686302486750917, + "loss": 3.2805, + "step": 74340 + }, + { + "epoch": 5.051297730669928, + "grad_norm": 0.9784578680992126, + "learning_rate": 0.00036858778366625903, + "loss": 3.1581, + "step": 74345 + }, + { + "epoch": 5.05163745074059, + "grad_norm": 0.7737109065055847, + "learning_rate": 0.0003685453186574263, + "loss": 3.4391, + "step": 74350 + }, + { + "epoch": 5.051977170811251, + "grad_norm": 0.7958144545555115, + "learning_rate": 0.00036850285364859353, + "loss": 3.5141, + "step": 74355 + }, + { + "epoch": 5.052316890881913, + "grad_norm": 1.7831401824951172, + "learning_rate": 0.0003684603886397608, + "loss": 3.3923, + "step": 74360 + }, + { + "epoch": 5.0526566109525755, + "grad_norm": 0.8024105429649353, + "learning_rate": 0.00036841792363092815, + "loss": 3.3207, + "step": 74365 + }, + { + "epoch": 5.052996331023237, + "grad_norm": 0.9502870440483093, + "learning_rate": 0.00036837545862209543, + "loss": 3.4673, + "step": 74370 + }, + { + "epoch": 5.053336051093899, + "grad_norm": 0.9811963438987732, + "learning_rate": 0.00036833299361326265, + "loss": 3.3953, + "step": 74375 + }, + { + "epoch": 5.053675771164561, + "grad_norm": 0.7844878435134888, + "learning_rate": 0.00036829052860443, + "loss": 3.1985, + "step": 74380 + }, + { + "epoch": 5.054015491235222, + "grad_norm": 1.2959247827529907, + "learning_rate": 0.00036824806359559727, + "loss": 3.302, + "step": 74385 + }, + { + "epoch": 5.054355211305884, + "grad_norm": 1.0022271871566772, + "learning_rate": 0.0003682055985867645, + "loss": 3.5885, + "step": 74390 + }, + { + "epoch": 5.054694931376546, + "grad_norm": 0.7763844132423401, + "learning_rate": 0.0003681631335779318, + "loss": 3.3302, + "step": 74395 + }, + { + "epoch": 5.055034651447207, + "grad_norm": 0.8863420486450195, + "learning_rate": 0.0003681206685690991, + "loss": 3.1322, + "step": 74400 + }, + { + "epoch": 5.055374371517869, + "grad_norm": 0.9745915532112122, + "learning_rate": 0.00036807820356026634, + "loss": 3.4016, + "step": 74405 + }, + { + "epoch": 5.0557140915885315, + "grad_norm": 0.8081738948822021, + "learning_rate": 0.0003680357385514336, + "loss": 3.3853, + "step": 74410 + }, + { + "epoch": 5.056053811659193, + "grad_norm": 0.8080282211303711, + "learning_rate": 0.00036799327354260095, + "loss": 3.462, + "step": 74415 + }, + { + "epoch": 5.056393531729855, + "grad_norm": 0.9290738701820374, + "learning_rate": 0.0003679508085337682, + "loss": 3.5414, + "step": 74420 + }, + { + "epoch": 5.056733251800517, + "grad_norm": 0.8939818143844604, + "learning_rate": 0.00036790834352493546, + "loss": 3.7895, + "step": 74425 + }, + { + "epoch": 5.057072971871178, + "grad_norm": 0.8926640152931213, + "learning_rate": 0.00036786587851610274, + "loss": 3.5479, + "step": 74430 + }, + { + "epoch": 5.05741269194184, + "grad_norm": 0.9758835434913635, + "learning_rate": 0.00036782341350727, + "loss": 3.3993, + "step": 74435 + }, + { + "epoch": 5.057752412012501, + "grad_norm": 1.341876745223999, + "learning_rate": 0.0003677809484984373, + "loss": 3.4464, + "step": 74440 + }, + { + "epoch": 5.058092132083163, + "grad_norm": 0.8545658588409424, + "learning_rate": 0.0003677384834896046, + "loss": 3.4534, + "step": 74445 + }, + { + "epoch": 5.058431852153825, + "grad_norm": 1.0669986009597778, + "learning_rate": 0.00036769601848077186, + "loss": 3.1932, + "step": 74450 + }, + { + "epoch": 5.058771572224487, + "grad_norm": 0.8410337567329407, + "learning_rate": 0.00036765355347193914, + "loss": 3.431, + "step": 74455 + }, + { + "epoch": 5.059111292295149, + "grad_norm": 1.0390547513961792, + "learning_rate": 0.0003676110884631064, + "loss": 3.4483, + "step": 74460 + }, + { + "epoch": 5.059451012365811, + "grad_norm": 1.0414466857910156, + "learning_rate": 0.00036756862345427364, + "loss": 3.4586, + "step": 74465 + }, + { + "epoch": 5.059790732436472, + "grad_norm": 0.8129729628562927, + "learning_rate": 0.000367526158445441, + "loss": 3.2754, + "step": 74470 + }, + { + "epoch": 5.060130452507134, + "grad_norm": 0.880726158618927, + "learning_rate": 0.00036748369343660826, + "loss": 3.502, + "step": 74475 + }, + { + "epoch": 5.060470172577796, + "grad_norm": 1.1044102907180786, + "learning_rate": 0.0003674412284277755, + "loss": 3.6778, + "step": 74480 + }, + { + "epoch": 5.060809892648457, + "grad_norm": 0.8495376110076904, + "learning_rate": 0.0003673987634189428, + "loss": 3.6414, + "step": 74485 + }, + { + "epoch": 5.061149612719119, + "grad_norm": 1.061542272567749, + "learning_rate": 0.0003673562984101101, + "loss": 3.3403, + "step": 74490 + }, + { + "epoch": 5.061489332789781, + "grad_norm": 1.0390080213546753, + "learning_rate": 0.0003673138334012773, + "loss": 3.4795, + "step": 74495 + }, + { + "epoch": 5.061829052860443, + "grad_norm": 1.0464240312576294, + "learning_rate": 0.0003672713683924446, + "loss": 3.6371, + "step": 74500 + }, + { + "epoch": 5.062168772931105, + "grad_norm": 0.9165144562721252, + "learning_rate": 0.00036722890338361194, + "loss": 3.2411, + "step": 74505 + }, + { + "epoch": 5.062508493001767, + "grad_norm": 0.8787360787391663, + "learning_rate": 0.00036718643837477916, + "loss": 3.4691, + "step": 74510 + }, + { + "epoch": 5.062848213072428, + "grad_norm": 0.9694712162017822, + "learning_rate": 0.00036714397336594644, + "loss": 3.4866, + "step": 74515 + }, + { + "epoch": 5.06318793314309, + "grad_norm": 1.0332369804382324, + "learning_rate": 0.0003671015083571138, + "loss": 3.5234, + "step": 74520 + }, + { + "epoch": 5.063527653213752, + "grad_norm": 0.8263710737228394, + "learning_rate": 0.000367059043348281, + "loss": 3.1496, + "step": 74525 + }, + { + "epoch": 5.063867373284413, + "grad_norm": 1.0203291177749634, + "learning_rate": 0.0003670165783394483, + "loss": 3.6649, + "step": 74530 + }, + { + "epoch": 5.064207093355075, + "grad_norm": 0.9047434329986572, + "learning_rate": 0.00036697411333061556, + "loss": 3.446, + "step": 74535 + }, + { + "epoch": 5.064546813425737, + "grad_norm": 1.1405476331710815, + "learning_rate": 0.0003669316483217829, + "loss": 3.3901, + "step": 74540 + }, + { + "epoch": 5.064886533496399, + "grad_norm": 0.9893170595169067, + "learning_rate": 0.0003668891833129501, + "loss": 3.4564, + "step": 74545 + }, + { + "epoch": 5.065226253567061, + "grad_norm": 0.8183785080909729, + "learning_rate": 0.0003668467183041174, + "loss": 3.359, + "step": 74550 + }, + { + "epoch": 5.065565973637723, + "grad_norm": 0.9511227011680603, + "learning_rate": 0.00036680425329528474, + "loss": 3.5132, + "step": 74555 + }, + { + "epoch": 5.065905693708384, + "grad_norm": 0.8187382221221924, + "learning_rate": 0.00036676178828645196, + "loss": 3.2912, + "step": 74560 + }, + { + "epoch": 5.066245413779046, + "grad_norm": 0.9733322858810425, + "learning_rate": 0.00036671932327761924, + "loss": 3.5882, + "step": 74565 + }, + { + "epoch": 5.066585133849708, + "grad_norm": 1.0464316606521606, + "learning_rate": 0.0003666768582687866, + "loss": 3.4884, + "step": 74570 + }, + { + "epoch": 5.066924853920369, + "grad_norm": 1.106223225593567, + "learning_rate": 0.0003666343932599538, + "loss": 3.6416, + "step": 74575 + }, + { + "epoch": 5.067264573991031, + "grad_norm": 0.8453660607337952, + "learning_rate": 0.0003665919282511211, + "loss": 3.177, + "step": 74580 + }, + { + "epoch": 5.067604294061693, + "grad_norm": 1.0638985633850098, + "learning_rate": 0.00036654946324228836, + "loss": 3.6729, + "step": 74585 + }, + { + "epoch": 5.067944014132355, + "grad_norm": 0.89426589012146, + "learning_rate": 0.00036650699823345564, + "loss": 3.4266, + "step": 74590 + }, + { + "epoch": 5.068283734203017, + "grad_norm": 1.0200386047363281, + "learning_rate": 0.0003664645332246229, + "loss": 3.4966, + "step": 74595 + }, + { + "epoch": 5.068623454273679, + "grad_norm": 1.131991982460022, + "learning_rate": 0.0003664220682157902, + "loss": 3.5852, + "step": 74600 + }, + { + "epoch": 5.06896317434434, + "grad_norm": 0.8159005641937256, + "learning_rate": 0.0003663796032069575, + "loss": 3.4803, + "step": 74605 + }, + { + "epoch": 5.069302894415002, + "grad_norm": 1.0808874368667603, + "learning_rate": 0.00036633713819812476, + "loss": 3.7532, + "step": 74610 + }, + { + "epoch": 5.069642614485664, + "grad_norm": 0.7710488438606262, + "learning_rate": 0.00036629467318929204, + "loss": 3.5908, + "step": 74615 + }, + { + "epoch": 5.069982334556325, + "grad_norm": 0.7490910887718201, + "learning_rate": 0.00036625220818045927, + "loss": 3.5817, + "step": 74620 + }, + { + "epoch": 5.070322054626987, + "grad_norm": 1.0362935066223145, + "learning_rate": 0.0003662097431716266, + "loss": 3.187, + "step": 74625 + }, + { + "epoch": 5.0706617746976494, + "grad_norm": 1.1559901237487793, + "learning_rate": 0.0003661672781627939, + "loss": 3.3659, + "step": 74630 + }, + { + "epoch": 5.071001494768311, + "grad_norm": 0.9697914123535156, + "learning_rate": 0.0003661248131539611, + "loss": 3.5432, + "step": 74635 + }, + { + "epoch": 5.071341214838973, + "grad_norm": 0.8743848204612732, + "learning_rate": 0.00036608234814512844, + "loss": 3.5865, + "step": 74640 + }, + { + "epoch": 5.071680934909635, + "grad_norm": 1.7949402332305908, + "learning_rate": 0.0003660398831362957, + "loss": 3.3712, + "step": 74645 + }, + { + "epoch": 5.072020654980296, + "grad_norm": 0.798200249671936, + "learning_rate": 0.00036599741812746295, + "loss": 3.3011, + "step": 74650 + }, + { + "epoch": 5.072360375050958, + "grad_norm": 0.7571073770523071, + "learning_rate": 0.00036595495311863023, + "loss": 3.4232, + "step": 74655 + }, + { + "epoch": 5.07270009512162, + "grad_norm": 1.1301369667053223, + "learning_rate": 0.00036591248810979756, + "loss": 3.4011, + "step": 74660 + }, + { + "epoch": 5.073039815192281, + "grad_norm": 0.7377216815948486, + "learning_rate": 0.0003658700231009648, + "loss": 3.6746, + "step": 74665 + }, + { + "epoch": 5.073379535262943, + "grad_norm": 0.8606254458427429, + "learning_rate": 0.00036582755809213207, + "loss": 3.9043, + "step": 74670 + }, + { + "epoch": 5.0737192553336055, + "grad_norm": 1.015848159790039, + "learning_rate": 0.0003657850930832994, + "loss": 3.2423, + "step": 74675 + }, + { + "epoch": 5.074058975404267, + "grad_norm": 0.9080487489700317, + "learning_rate": 0.00036574262807446663, + "loss": 3.344, + "step": 74680 + }, + { + "epoch": 5.074398695474929, + "grad_norm": 0.8189088106155396, + "learning_rate": 0.0003657001630656339, + "loss": 3.542, + "step": 74685 + }, + { + "epoch": 5.074738415545591, + "grad_norm": 0.8650960326194763, + "learning_rate": 0.0003656576980568012, + "loss": 3.2139, + "step": 74690 + }, + { + "epoch": 5.075078135616252, + "grad_norm": 0.9276291728019714, + "learning_rate": 0.00036561523304796847, + "loss": 3.1652, + "step": 74695 + }, + { + "epoch": 5.075417855686914, + "grad_norm": 0.7726927995681763, + "learning_rate": 0.00036557276803913575, + "loss": 3.2818, + "step": 74700 + }, + { + "epoch": 5.075757575757576, + "grad_norm": 1.0042513608932495, + "learning_rate": 0.00036553030303030303, + "loss": 3.3603, + "step": 74705 + }, + { + "epoch": 5.076097295828237, + "grad_norm": 0.9540241360664368, + "learning_rate": 0.00036548783802147037, + "loss": 3.4962, + "step": 74710 + }, + { + "epoch": 5.076437015898899, + "grad_norm": 0.9946152567863464, + "learning_rate": 0.0003654453730126376, + "loss": 3.5943, + "step": 74715 + }, + { + "epoch": 5.0767767359695615, + "grad_norm": 0.9931941628456116, + "learning_rate": 0.00036540290800380487, + "loss": 3.543, + "step": 74720 + }, + { + "epoch": 5.077116456040223, + "grad_norm": 0.8514108061790466, + "learning_rate": 0.00036536044299497215, + "loss": 3.5234, + "step": 74725 + }, + { + "epoch": 5.077456176110885, + "grad_norm": 0.9903560280799866, + "learning_rate": 0.00036531797798613943, + "loss": 3.6574, + "step": 74730 + }, + { + "epoch": 5.077795896181547, + "grad_norm": 0.8982710838317871, + "learning_rate": 0.0003652755129773067, + "loss": 3.302, + "step": 74735 + }, + { + "epoch": 5.078135616252208, + "grad_norm": 1.347965121269226, + "learning_rate": 0.000365233047968474, + "loss": 3.4945, + "step": 74740 + }, + { + "epoch": 5.07847533632287, + "grad_norm": 0.9903069734573364, + "learning_rate": 0.00036519058295964127, + "loss": 3.3341, + "step": 74745 + }, + { + "epoch": 5.078815056393532, + "grad_norm": 0.9816784262657166, + "learning_rate": 0.00036514811795080855, + "loss": 3.3578, + "step": 74750 + }, + { + "epoch": 5.079154776464193, + "grad_norm": 0.8616023063659668, + "learning_rate": 0.00036510565294197583, + "loss": 3.662, + "step": 74755 + }, + { + "epoch": 5.079494496534855, + "grad_norm": 1.0039408206939697, + "learning_rate": 0.00036506318793314306, + "loss": 3.5201, + "step": 74760 + }, + { + "epoch": 5.0798342166055175, + "grad_norm": 1.0445092916488647, + "learning_rate": 0.0003650207229243104, + "loss": 3.4898, + "step": 74765 + }, + { + "epoch": 5.080173936676179, + "grad_norm": 0.9555175304412842, + "learning_rate": 0.00036497825791547767, + "loss": 3.5357, + "step": 74770 + }, + { + "epoch": 5.080513656746841, + "grad_norm": 1.346887230873108, + "learning_rate": 0.0003649357929066449, + "loss": 3.6041, + "step": 74775 + }, + { + "epoch": 5.080853376817503, + "grad_norm": 1.1599987745285034, + "learning_rate": 0.00036489332789781223, + "loss": 3.4223, + "step": 74780 + }, + { + "epoch": 5.081193096888164, + "grad_norm": 1.493691086769104, + "learning_rate": 0.0003648508628889795, + "loss": 3.4207, + "step": 74785 + }, + { + "epoch": 5.081532816958826, + "grad_norm": 0.9721444845199585, + "learning_rate": 0.00036480839788014674, + "loss": 3.371, + "step": 74790 + }, + { + "epoch": 5.081872537029487, + "grad_norm": 0.9574981331825256, + "learning_rate": 0.000364765932871314, + "loss": 3.4279, + "step": 74795 + }, + { + "epoch": 5.082212257100149, + "grad_norm": 0.7686623334884644, + "learning_rate": 0.00036472346786248135, + "loss": 3.386, + "step": 74800 + }, + { + "epoch": 5.082551977170811, + "grad_norm": 0.7745791673660278, + "learning_rate": 0.0003646810028536486, + "loss": 3.4058, + "step": 74805 + }, + { + "epoch": 5.082891697241473, + "grad_norm": 0.983389675617218, + "learning_rate": 0.00036463853784481586, + "loss": 3.4161, + "step": 74810 + }, + { + "epoch": 5.083231417312135, + "grad_norm": 1.2050780057907104, + "learning_rate": 0.0003645960728359832, + "loss": 3.6003, + "step": 74815 + }, + { + "epoch": 5.083571137382797, + "grad_norm": 1.0247797966003418, + "learning_rate": 0.0003645536078271504, + "loss": 3.4164, + "step": 74820 + }, + { + "epoch": 5.083910857453458, + "grad_norm": 0.9998893737792969, + "learning_rate": 0.0003645111428183177, + "loss": 3.6561, + "step": 74825 + }, + { + "epoch": 5.08425057752412, + "grad_norm": 0.978882908821106, + "learning_rate": 0.000364468677809485, + "loss": 3.5452, + "step": 74830 + }, + { + "epoch": 5.084590297594782, + "grad_norm": 0.8294420838356018, + "learning_rate": 0.00036442621280065226, + "loss": 3.4461, + "step": 74835 + }, + { + "epoch": 5.084930017665443, + "grad_norm": 1.2561691999435425, + "learning_rate": 0.00036438374779181954, + "loss": 3.3592, + "step": 74840 + }, + { + "epoch": 5.085269737736105, + "grad_norm": 1.1057792901992798, + "learning_rate": 0.0003643412827829868, + "loss": 3.751, + "step": 74845 + }, + { + "epoch": 5.085609457806767, + "grad_norm": 0.7867390513420105, + "learning_rate": 0.0003642988177741541, + "loss": 3.5661, + "step": 74850 + }, + { + "epoch": 5.085949177877429, + "grad_norm": 0.9753893613815308, + "learning_rate": 0.0003642563527653214, + "loss": 3.272, + "step": 74855 + }, + { + "epoch": 5.086288897948091, + "grad_norm": 0.69923996925354, + "learning_rate": 0.00036421388775648866, + "loss": 3.4372, + "step": 74860 + }, + { + "epoch": 5.086628618018753, + "grad_norm": 0.9283983111381531, + "learning_rate": 0.0003641714227476559, + "loss": 3.4698, + "step": 74865 + }, + { + "epoch": 5.086968338089414, + "grad_norm": 0.9221273064613342, + "learning_rate": 0.0003641289577388232, + "loss": 3.2585, + "step": 74870 + }, + { + "epoch": 5.087308058160076, + "grad_norm": 1.0572346448898315, + "learning_rate": 0.0003640864927299905, + "loss": 3.4023, + "step": 74875 + }, + { + "epoch": 5.087647778230738, + "grad_norm": 0.8362987041473389, + "learning_rate": 0.0003640440277211578, + "loss": 3.6335, + "step": 74880 + }, + { + "epoch": 5.087987498301399, + "grad_norm": 0.8806021809577942, + "learning_rate": 0.00036400156271232506, + "loss": 3.318, + "step": 74885 + }, + { + "epoch": 5.088327218372061, + "grad_norm": 1.0497980117797852, + "learning_rate": 0.00036395909770349234, + "loss": 3.3717, + "step": 74890 + }, + { + "epoch": 5.088666938442723, + "grad_norm": 0.8034532070159912, + "learning_rate": 0.0003639166326946596, + "loss": 3.1862, + "step": 74895 + }, + { + "epoch": 5.089006658513385, + "grad_norm": 1.1501727104187012, + "learning_rate": 0.0003638741676858269, + "loss": 3.5848, + "step": 74900 + }, + { + "epoch": 5.089346378584047, + "grad_norm": 0.8831846117973328, + "learning_rate": 0.0003638317026769942, + "loss": 3.3995, + "step": 74905 + }, + { + "epoch": 5.089686098654709, + "grad_norm": 0.9833759069442749, + "learning_rate": 0.00036378923766816146, + "loss": 3.2713, + "step": 74910 + }, + { + "epoch": 5.09002581872537, + "grad_norm": 0.8408897519111633, + "learning_rate": 0.0003637467726593287, + "loss": 3.4248, + "step": 74915 + }, + { + "epoch": 5.090365538796032, + "grad_norm": 0.7591267228126526, + "learning_rate": 0.000363704307650496, + "loss": 3.4615, + "step": 74920 + }, + { + "epoch": 5.090705258866694, + "grad_norm": 0.8523525595664978, + "learning_rate": 0.0003636618426416633, + "loss": 3.3067, + "step": 74925 + }, + { + "epoch": 5.091044978937355, + "grad_norm": 0.8751924633979797, + "learning_rate": 0.0003636193776328305, + "loss": 3.3235, + "step": 74930 + }, + { + "epoch": 5.091384699008017, + "grad_norm": 0.8093482255935669, + "learning_rate": 0.00036357691262399786, + "loss": 3.3978, + "step": 74935 + }, + { + "epoch": 5.0917244190786795, + "grad_norm": 0.8375375866889954, + "learning_rate": 0.00036353444761516514, + "loss": 3.6556, + "step": 74940 + }, + { + "epoch": 5.092064139149341, + "grad_norm": 0.920310378074646, + "learning_rate": 0.00036349198260633237, + "loss": 3.338, + "step": 74945 + }, + { + "epoch": 5.092403859220003, + "grad_norm": 0.8462808728218079, + "learning_rate": 0.00036344951759749965, + "loss": 3.2296, + "step": 74950 + }, + { + "epoch": 5.092743579290665, + "grad_norm": 0.9120371341705322, + "learning_rate": 0.000363407052588667, + "loss": 3.5621, + "step": 74955 + }, + { + "epoch": 5.093083299361326, + "grad_norm": 0.817236065864563, + "learning_rate": 0.0003633645875798342, + "loss": 3.4538, + "step": 74960 + }, + { + "epoch": 5.093423019431988, + "grad_norm": 0.7250436544418335, + "learning_rate": 0.0003633221225710015, + "loss": 3.5455, + "step": 74965 + }, + { + "epoch": 5.09376273950265, + "grad_norm": 1.3266916275024414, + "learning_rate": 0.0003632796575621688, + "loss": 3.659, + "step": 74970 + }, + { + "epoch": 5.094102459573311, + "grad_norm": 1.0124485492706299, + "learning_rate": 0.00036323719255333605, + "loss": 3.502, + "step": 74975 + }, + { + "epoch": 5.094442179643973, + "grad_norm": 0.9973366856575012, + "learning_rate": 0.00036319472754450333, + "loss": 3.2221, + "step": 74980 + }, + { + "epoch": 5.0947818997146355, + "grad_norm": 0.9812241196632385, + "learning_rate": 0.0003631522625356706, + "loss": 3.5178, + "step": 74985 + }, + { + "epoch": 5.095121619785297, + "grad_norm": 1.0427446365356445, + "learning_rate": 0.0003631097975268379, + "loss": 3.6051, + "step": 74990 + }, + { + "epoch": 5.095461339855959, + "grad_norm": 0.7722885608673096, + "learning_rate": 0.00036306733251800517, + "loss": 3.7283, + "step": 74995 + }, + { + "epoch": 5.095801059926621, + "grad_norm": 0.9766445755958557, + "learning_rate": 0.00036302486750917245, + "loss": 3.6169, + "step": 75000 + }, + { + "epoch": 5.096140779997282, + "grad_norm": 0.8959020376205444, + "learning_rate": 0.00036298240250033973, + "loss": 3.3327, + "step": 75005 + }, + { + "epoch": 5.096480500067944, + "grad_norm": 0.886807382106781, + "learning_rate": 0.000362939937491507, + "loss": 3.6109, + "step": 75010 + }, + { + "epoch": 5.096820220138606, + "grad_norm": 0.7945123910903931, + "learning_rate": 0.0003628974724826743, + "loss": 3.7052, + "step": 75015 + }, + { + "epoch": 5.097159940209267, + "grad_norm": 1.22524893283844, + "learning_rate": 0.0003628550074738415, + "loss": 3.4699, + "step": 75020 + }, + { + "epoch": 5.097499660279929, + "grad_norm": 0.7815542221069336, + "learning_rate": 0.00036281254246500885, + "loss": 3.4951, + "step": 75025 + }, + { + "epoch": 5.0978393803505915, + "grad_norm": 0.951709508895874, + "learning_rate": 0.00036277007745617613, + "loss": 3.3323, + "step": 75030 + }, + { + "epoch": 5.098179100421253, + "grad_norm": 0.9922013282775879, + "learning_rate": 0.00036272761244734335, + "loss": 3.2862, + "step": 75035 + }, + { + "epoch": 5.098518820491915, + "grad_norm": 0.8157073855400085, + "learning_rate": 0.0003626851474385107, + "loss": 3.3791, + "step": 75040 + }, + { + "epoch": 5.098858540562577, + "grad_norm": 0.7490096688270569, + "learning_rate": 0.00036264268242967797, + "loss": 3.606, + "step": 75045 + }, + { + "epoch": 5.099198260633238, + "grad_norm": 1.0346990823745728, + "learning_rate": 0.00036260021742084525, + "loss": 3.1534, + "step": 75050 + }, + { + "epoch": 5.0995379807039, + "grad_norm": 0.9536871910095215, + "learning_rate": 0.0003625577524120125, + "loss": 3.233, + "step": 75055 + }, + { + "epoch": 5.099877700774562, + "grad_norm": 0.8778948783874512, + "learning_rate": 0.0003625152874031798, + "loss": 3.5092, + "step": 75060 + }, + { + "epoch": 5.100217420845223, + "grad_norm": 0.7315804958343506, + "learning_rate": 0.0003624728223943471, + "loss": 3.3183, + "step": 75065 + }, + { + "epoch": 5.100557140915885, + "grad_norm": 0.891664445400238, + "learning_rate": 0.0003624303573855143, + "loss": 3.5301, + "step": 75070 + }, + { + "epoch": 5.1008968609865475, + "grad_norm": 1.0686360597610474, + "learning_rate": 0.00036238789237668165, + "loss": 3.5231, + "step": 75075 + }, + { + "epoch": 5.101236581057209, + "grad_norm": 0.8331325054168701, + "learning_rate": 0.00036234542736784893, + "loss": 3.5402, + "step": 75080 + }, + { + "epoch": 5.101576301127871, + "grad_norm": 1.1884500980377197, + "learning_rate": 0.00036230296235901615, + "loss": 3.4176, + "step": 75085 + }, + { + "epoch": 5.101916021198533, + "grad_norm": 0.9484277963638306, + "learning_rate": 0.00036226049735018343, + "loss": 3.6443, + "step": 75090 + }, + { + "epoch": 5.102255741269194, + "grad_norm": 0.903917670249939, + "learning_rate": 0.00036221803234135077, + "loss": 3.3796, + "step": 75095 + }, + { + "epoch": 5.102595461339856, + "grad_norm": 0.9545102715492249, + "learning_rate": 0.000362175567332518, + "loss": 3.4833, + "step": 75100 + }, + { + "epoch": 5.102935181410518, + "grad_norm": 0.9166658520698547, + "learning_rate": 0.0003621331023236853, + "loss": 3.5251, + "step": 75105 + }, + { + "epoch": 5.103274901481179, + "grad_norm": 0.8696542382240295, + "learning_rate": 0.0003620906373148526, + "loss": 3.3384, + "step": 75110 + }, + { + "epoch": 5.103614621551841, + "grad_norm": 0.8936814069747925, + "learning_rate": 0.00036204817230601984, + "loss": 3.449, + "step": 75115 + }, + { + "epoch": 5.103954341622503, + "grad_norm": 0.7980839610099792, + "learning_rate": 0.0003620057072971871, + "loss": 3.3708, + "step": 75120 + }, + { + "epoch": 5.104294061693165, + "grad_norm": 0.9942446947097778, + "learning_rate": 0.0003619632422883544, + "loss": 3.393, + "step": 75125 + }, + { + "epoch": 5.104633781763827, + "grad_norm": 0.9312781691551208, + "learning_rate": 0.0003619207772795217, + "loss": 3.4643, + "step": 75130 + }, + { + "epoch": 5.104973501834488, + "grad_norm": 1.1152287721633911, + "learning_rate": 0.00036187831227068896, + "loss": 3.3477, + "step": 75135 + }, + { + "epoch": 5.10531322190515, + "grad_norm": 1.013667345046997, + "learning_rate": 0.00036183584726185624, + "loss": 3.1521, + "step": 75140 + }, + { + "epoch": 5.105652941975812, + "grad_norm": 0.8019955158233643, + "learning_rate": 0.0003617933822530235, + "loss": 3.2059, + "step": 75145 + }, + { + "epoch": 5.105992662046473, + "grad_norm": 0.9575315117835999, + "learning_rate": 0.0003617509172441908, + "loss": 3.544, + "step": 75150 + }, + { + "epoch": 5.106332382117135, + "grad_norm": 0.8219946622848511, + "learning_rate": 0.0003617084522353581, + "loss": 3.4609, + "step": 75155 + }, + { + "epoch": 5.106672102187797, + "grad_norm": 1.164125680923462, + "learning_rate": 0.0003616659872265253, + "loss": 3.2738, + "step": 75160 + }, + { + "epoch": 5.107011822258459, + "grad_norm": 1.005372166633606, + "learning_rate": 0.00036162352221769264, + "loss": 3.7366, + "step": 75165 + }, + { + "epoch": 5.107351542329121, + "grad_norm": 0.7238408327102661, + "learning_rate": 0.0003615810572088599, + "loss": 3.0939, + "step": 75170 + }, + { + "epoch": 5.107691262399783, + "grad_norm": 0.798172116279602, + "learning_rate": 0.00036153859220002714, + "loss": 3.5253, + "step": 75175 + }, + { + "epoch": 5.108030982470444, + "grad_norm": 0.9550918936729431, + "learning_rate": 0.0003614961271911945, + "loss": 3.2822, + "step": 75180 + }, + { + "epoch": 5.108370702541106, + "grad_norm": 0.995512068271637, + "learning_rate": 0.00036145366218236176, + "loss": 3.3756, + "step": 75185 + }, + { + "epoch": 5.108710422611768, + "grad_norm": 0.9895334243774414, + "learning_rate": 0.000361411197173529, + "loss": 3.308, + "step": 75190 + }, + { + "epoch": 5.109050142682429, + "grad_norm": 1.1365963220596313, + "learning_rate": 0.0003613687321646963, + "loss": 3.4375, + "step": 75195 + }, + { + "epoch": 5.109389862753091, + "grad_norm": 1.0738314390182495, + "learning_rate": 0.0003613262671558636, + "loss": 3.4143, + "step": 75200 + }, + { + "epoch": 5.1097295828237534, + "grad_norm": 0.7811956405639648, + "learning_rate": 0.0003612838021470308, + "loss": 3.4044, + "step": 75205 + }, + { + "epoch": 5.110069302894415, + "grad_norm": 0.804663896560669, + "learning_rate": 0.0003612413371381981, + "loss": 3.5527, + "step": 75210 + }, + { + "epoch": 5.110409022965077, + "grad_norm": 0.9579253196716309, + "learning_rate": 0.00036119887212936544, + "loss": 3.3637, + "step": 75215 + }, + { + "epoch": 5.110748743035739, + "grad_norm": 1.0096447467803955, + "learning_rate": 0.0003611564071205327, + "loss": 3.4818, + "step": 75220 + }, + { + "epoch": 5.1110884631064, + "grad_norm": 0.7714332938194275, + "learning_rate": 0.00036111394211169994, + "loss": 3.5344, + "step": 75225 + }, + { + "epoch": 5.111428183177062, + "grad_norm": 0.9728423357009888, + "learning_rate": 0.0003610714771028673, + "loss": 3.6027, + "step": 75230 + }, + { + "epoch": 5.111767903247724, + "grad_norm": 0.830195426940918, + "learning_rate": 0.00036102901209403456, + "loss": 3.4362, + "step": 75235 + }, + { + "epoch": 5.112107623318385, + "grad_norm": 0.8785569071769714, + "learning_rate": 0.0003609865470852018, + "loss": 3.39, + "step": 75240 + }, + { + "epoch": 5.112447343389047, + "grad_norm": 1.0836104154586792, + "learning_rate": 0.00036094408207636906, + "loss": 3.2883, + "step": 75245 + }, + { + "epoch": 5.1127870634597095, + "grad_norm": 1.0558947324752808, + "learning_rate": 0.0003609016170675364, + "loss": 3.6536, + "step": 75250 + }, + { + "epoch": 5.113126783530371, + "grad_norm": 0.9367012977600098, + "learning_rate": 0.0003608591520587036, + "loss": 3.1756, + "step": 75255 + }, + { + "epoch": 5.113466503601033, + "grad_norm": 0.8422859907150269, + "learning_rate": 0.0003608166870498709, + "loss": 3.3402, + "step": 75260 + }, + { + "epoch": 5.113806223671695, + "grad_norm": 0.9152809977531433, + "learning_rate": 0.00036077422204103824, + "loss": 3.5203, + "step": 75265 + }, + { + "epoch": 5.114145943742356, + "grad_norm": 0.876521646976471, + "learning_rate": 0.00036073175703220546, + "loss": 3.5446, + "step": 75270 + }, + { + "epoch": 5.114485663813018, + "grad_norm": 0.8896933794021606, + "learning_rate": 0.00036068929202337274, + "loss": 3.409, + "step": 75275 + }, + { + "epoch": 5.11482538388368, + "grad_norm": 0.7991227507591248, + "learning_rate": 0.00036064682701454, + "loss": 3.5709, + "step": 75280 + }, + { + "epoch": 5.115165103954341, + "grad_norm": 1.0065653324127197, + "learning_rate": 0.0003606043620057073, + "loss": 3.5866, + "step": 75285 + }, + { + "epoch": 5.115504824025003, + "grad_norm": 1.005728840827942, + "learning_rate": 0.0003605618969968746, + "loss": 3.3197, + "step": 75290 + }, + { + "epoch": 5.1158445440956655, + "grad_norm": 1.068459391593933, + "learning_rate": 0.00036051943198804186, + "loss": 3.505, + "step": 75295 + }, + { + "epoch": 5.116184264166327, + "grad_norm": 1.0139633417129517, + "learning_rate": 0.00036047696697920914, + "loss": 3.5115, + "step": 75300 + }, + { + "epoch": 5.116523984236989, + "grad_norm": 1.1516929864883423, + "learning_rate": 0.0003604345019703764, + "loss": 3.1883, + "step": 75305 + }, + { + "epoch": 5.116863704307651, + "grad_norm": 0.9546001553535461, + "learning_rate": 0.0003603920369615437, + "loss": 3.6388, + "step": 75310 + }, + { + "epoch": 5.117203424378312, + "grad_norm": 1.0090304613113403, + "learning_rate": 0.00036034957195271093, + "loss": 3.3868, + "step": 75315 + }, + { + "epoch": 5.117543144448974, + "grad_norm": 0.7460254430770874, + "learning_rate": 0.00036030710694387826, + "loss": 3.4881, + "step": 75320 + }, + { + "epoch": 5.117882864519636, + "grad_norm": 0.8624724745750427, + "learning_rate": 0.00036026464193504554, + "loss": 3.5399, + "step": 75325 + }, + { + "epoch": 5.118222584590297, + "grad_norm": 0.9487879872322083, + "learning_rate": 0.00036022217692621277, + "loss": 3.4304, + "step": 75330 + }, + { + "epoch": 5.118562304660959, + "grad_norm": 1.0515177249908447, + "learning_rate": 0.0003601797119173801, + "loss": 3.4221, + "step": 75335 + }, + { + "epoch": 5.1189020247316215, + "grad_norm": 0.9962283372879028, + "learning_rate": 0.0003601372469085474, + "loss": 3.477, + "step": 75340 + }, + { + "epoch": 5.119241744802283, + "grad_norm": 1.152889609336853, + "learning_rate": 0.0003600947818997146, + "loss": 3.3088, + "step": 75345 + }, + { + "epoch": 5.119581464872945, + "grad_norm": 0.8251771330833435, + "learning_rate": 0.0003600523168908819, + "loss": 3.748, + "step": 75350 + }, + { + "epoch": 5.119921184943607, + "grad_norm": 0.8877094984054565, + "learning_rate": 0.0003600098518820492, + "loss": 3.377, + "step": 75355 + }, + { + "epoch": 5.120260905014268, + "grad_norm": 0.7578794360160828, + "learning_rate": 0.00035996738687321645, + "loss": 3.2087, + "step": 75360 + }, + { + "epoch": 5.12060062508493, + "grad_norm": 0.7884721159934998, + "learning_rate": 0.00035992492186438373, + "loss": 3.5021, + "step": 75365 + }, + { + "epoch": 5.120940345155592, + "grad_norm": 1.4379349946975708, + "learning_rate": 0.00035988245685555106, + "loss": 3.5711, + "step": 75370 + }, + { + "epoch": 5.121280065226253, + "grad_norm": 1.1530681848526, + "learning_rate": 0.0003598399918467183, + "loss": 3.4801, + "step": 75375 + }, + { + "epoch": 5.121619785296915, + "grad_norm": 0.9859174489974976, + "learning_rate": 0.00035979752683788557, + "loss": 3.3444, + "step": 75380 + }, + { + "epoch": 5.1219595053675775, + "grad_norm": 0.9901753664016724, + "learning_rate": 0.00035975506182905285, + "loss": 3.1273, + "step": 75385 + }, + { + "epoch": 5.122299225438239, + "grad_norm": 0.7243141531944275, + "learning_rate": 0.0003597125968202202, + "loss": 3.1836, + "step": 75390 + }, + { + "epoch": 5.122638945508901, + "grad_norm": 0.9931227564811707, + "learning_rate": 0.0003596701318113874, + "loss": 3.3004, + "step": 75395 + }, + { + "epoch": 5.122978665579563, + "grad_norm": 0.8207472562789917, + "learning_rate": 0.0003596276668025547, + "loss": 3.503, + "step": 75400 + }, + { + "epoch": 5.123318385650224, + "grad_norm": 0.8760255575180054, + "learning_rate": 0.000359585201793722, + "loss": 3.5104, + "step": 75405 + }, + { + "epoch": 5.123658105720886, + "grad_norm": 0.9737474918365479, + "learning_rate": 0.00035954273678488925, + "loss": 3.6006, + "step": 75410 + }, + { + "epoch": 5.123997825791548, + "grad_norm": 1.2021145820617676, + "learning_rate": 0.00035950027177605653, + "loss": 3.3088, + "step": 75415 + }, + { + "epoch": 5.124337545862209, + "grad_norm": 0.815832793712616, + "learning_rate": 0.0003594578067672238, + "loss": 3.4062, + "step": 75420 + }, + { + "epoch": 5.124677265932871, + "grad_norm": 0.776910126209259, + "learning_rate": 0.0003594153417583911, + "loss": 3.2368, + "step": 75425 + }, + { + "epoch": 5.1250169860035335, + "grad_norm": 0.8547676205635071, + "learning_rate": 0.00035937287674955837, + "loss": 3.5613, + "step": 75430 + }, + { + "epoch": 5.125356706074195, + "grad_norm": 0.974787175655365, + "learning_rate": 0.00035933041174072565, + "loss": 3.4011, + "step": 75435 + }, + { + "epoch": 5.125696426144857, + "grad_norm": 1.0379118919372559, + "learning_rate": 0.00035928794673189293, + "loss": 3.5029, + "step": 75440 + }, + { + "epoch": 5.126036146215519, + "grad_norm": 0.9388253688812256, + "learning_rate": 0.0003592454817230602, + "loss": 3.6932, + "step": 75445 + }, + { + "epoch": 5.12637586628618, + "grad_norm": 0.9530711770057678, + "learning_rate": 0.0003592030167142275, + "loss": 3.4744, + "step": 75450 + }, + { + "epoch": 5.126715586356842, + "grad_norm": 0.973609983921051, + "learning_rate": 0.0003591605517053947, + "loss": 3.6192, + "step": 75455 + }, + { + "epoch": 5.127055306427504, + "grad_norm": 1.034703016281128, + "learning_rate": 0.00035911808669656205, + "loss": 3.2962, + "step": 75460 + }, + { + "epoch": 5.127395026498165, + "grad_norm": 1.0627164840698242, + "learning_rate": 0.00035907562168772933, + "loss": 3.3745, + "step": 75465 + }, + { + "epoch": 5.127734746568827, + "grad_norm": 0.9506585597991943, + "learning_rate": 0.00035903315667889656, + "loss": 3.2617, + "step": 75470 + }, + { + "epoch": 5.1280744666394895, + "grad_norm": 0.986581563949585, + "learning_rate": 0.0003589906916700639, + "loss": 2.9607, + "step": 75475 + }, + { + "epoch": 5.128414186710151, + "grad_norm": 0.9458901882171631, + "learning_rate": 0.00035894822666123117, + "loss": 3.3489, + "step": 75480 + }, + { + "epoch": 5.128753906780813, + "grad_norm": 1.1331102848052979, + "learning_rate": 0.0003589057616523984, + "loss": 3.3257, + "step": 75485 + }, + { + "epoch": 5.129093626851474, + "grad_norm": 0.7538685202598572, + "learning_rate": 0.00035886329664356573, + "loss": 3.3868, + "step": 75490 + }, + { + "epoch": 5.129433346922136, + "grad_norm": 0.8945348858833313, + "learning_rate": 0.000358820831634733, + "loss": 3.5551, + "step": 75495 + }, + { + "epoch": 5.129773066992798, + "grad_norm": 0.7266561985015869, + "learning_rate": 0.00035877836662590024, + "loss": 3.3456, + "step": 75500 + }, + { + "epoch": 5.130112787063459, + "grad_norm": 0.8124348521232605, + "learning_rate": 0.0003587359016170675, + "loss": 3.4371, + "step": 75505 + }, + { + "epoch": 5.130452507134121, + "grad_norm": 0.9724119901657104, + "learning_rate": 0.00035869343660823485, + "loss": 3.4692, + "step": 75510 + }, + { + "epoch": 5.1307922272047835, + "grad_norm": 0.9436006546020508, + "learning_rate": 0.0003586509715994021, + "loss": 3.619, + "step": 75515 + }, + { + "epoch": 5.131131947275445, + "grad_norm": 0.9247618913650513, + "learning_rate": 0.00035860850659056936, + "loss": 3.4033, + "step": 75520 + }, + { + "epoch": 5.131471667346107, + "grad_norm": 0.7660719752311707, + "learning_rate": 0.0003585660415817367, + "loss": 3.3354, + "step": 75525 + }, + { + "epoch": 5.131811387416769, + "grad_norm": 0.8121278285980225, + "learning_rate": 0.0003585235765729039, + "loss": 3.3272, + "step": 75530 + }, + { + "epoch": 5.13215110748743, + "grad_norm": 0.896870493888855, + "learning_rate": 0.0003584811115640712, + "loss": 3.3307, + "step": 75535 + }, + { + "epoch": 5.132490827558092, + "grad_norm": 0.8842107057571411, + "learning_rate": 0.0003584386465552385, + "loss": 3.4925, + "step": 75540 + }, + { + "epoch": 5.132830547628754, + "grad_norm": 0.6370797753334045, + "learning_rate": 0.00035839618154640576, + "loss": 3.8845, + "step": 75545 + }, + { + "epoch": 5.133170267699415, + "grad_norm": 0.9411664009094238, + "learning_rate": 0.00035835371653757304, + "loss": 3.249, + "step": 75550 + }, + { + "epoch": 5.133509987770077, + "grad_norm": 0.7810190320014954, + "learning_rate": 0.0003583112515287403, + "loss": 3.448, + "step": 75555 + }, + { + "epoch": 5.1338497078407395, + "grad_norm": 0.9644551277160645, + "learning_rate": 0.00035826878651990765, + "loss": 3.6004, + "step": 75560 + }, + { + "epoch": 5.134189427911401, + "grad_norm": 0.7850888967514038, + "learning_rate": 0.0003582263215110749, + "loss": 3.5523, + "step": 75565 + }, + { + "epoch": 5.134529147982063, + "grad_norm": 0.8360772132873535, + "learning_rate": 0.00035818385650224216, + "loss": 3.3816, + "step": 75570 + }, + { + "epoch": 5.134868868052725, + "grad_norm": 0.8692635893821716, + "learning_rate": 0.00035814139149340944, + "loss": 3.4182, + "step": 75575 + }, + { + "epoch": 5.135208588123386, + "grad_norm": 1.0855294466018677, + "learning_rate": 0.0003580989264845767, + "loss": 3.4477, + "step": 75580 + }, + { + "epoch": 5.135548308194048, + "grad_norm": 0.691785991191864, + "learning_rate": 0.000358056461475744, + "loss": 3.6313, + "step": 75585 + }, + { + "epoch": 5.13588802826471, + "grad_norm": 1.1764471530914307, + "learning_rate": 0.0003580139964669113, + "loss": 3.1752, + "step": 75590 + }, + { + "epoch": 5.136227748335371, + "grad_norm": 0.8215623497962952, + "learning_rate": 0.00035797153145807856, + "loss": 3.523, + "step": 75595 + }, + { + "epoch": 5.136567468406033, + "grad_norm": 0.9227715134620667, + "learning_rate": 0.00035792906644924584, + "loss": 3.4223, + "step": 75600 + }, + { + "epoch": 5.1369071884766955, + "grad_norm": 1.0373984575271606, + "learning_rate": 0.0003578866014404131, + "loss": 3.6489, + "step": 75605 + }, + { + "epoch": 5.137246908547357, + "grad_norm": 0.7685438394546509, + "learning_rate": 0.00035784413643158035, + "loss": 3.5668, + "step": 75610 + }, + { + "epoch": 5.137586628618019, + "grad_norm": 0.8674196600914001, + "learning_rate": 0.0003578016714227477, + "loss": 3.5386, + "step": 75615 + }, + { + "epoch": 5.137926348688681, + "grad_norm": 1.0068711042404175, + "learning_rate": 0.00035775920641391496, + "loss": 3.8871, + "step": 75620 + }, + { + "epoch": 5.138266068759342, + "grad_norm": 0.9329433441162109, + "learning_rate": 0.0003577167414050822, + "loss": 3.6836, + "step": 75625 + }, + { + "epoch": 5.138605788830004, + "grad_norm": 1.1959995031356812, + "learning_rate": 0.0003576742763962495, + "loss": 3.4469, + "step": 75630 + }, + { + "epoch": 5.138945508900666, + "grad_norm": 0.8033577799797058, + "learning_rate": 0.0003576318113874168, + "loss": 3.5078, + "step": 75635 + }, + { + "epoch": 5.139285228971327, + "grad_norm": 0.9147692918777466, + "learning_rate": 0.000357589346378584, + "loss": 3.2276, + "step": 75640 + }, + { + "epoch": 5.139624949041989, + "grad_norm": 0.8372762203216553, + "learning_rate": 0.0003575468813697513, + "loss": 3.3919, + "step": 75645 + }, + { + "epoch": 5.1399646691126515, + "grad_norm": 0.9196427464485168, + "learning_rate": 0.00035750441636091864, + "loss": 3.5758, + "step": 75650 + }, + { + "epoch": 5.140304389183313, + "grad_norm": 0.8652243614196777, + "learning_rate": 0.00035746195135208587, + "loss": 3.5547, + "step": 75655 + }, + { + "epoch": 5.140644109253975, + "grad_norm": 0.7755324244499207, + "learning_rate": 0.00035741948634325315, + "loss": 3.4065, + "step": 75660 + }, + { + "epoch": 5.140983829324637, + "grad_norm": 0.7884494066238403, + "learning_rate": 0.0003573770213344205, + "loss": 3.5399, + "step": 75665 + }, + { + "epoch": 5.141323549395298, + "grad_norm": 0.862424373626709, + "learning_rate": 0.0003573345563255877, + "loss": 3.4619, + "step": 75670 + }, + { + "epoch": 5.14166326946596, + "grad_norm": 0.9353483319282532, + "learning_rate": 0.000357292091316755, + "loss": 3.7566, + "step": 75675 + }, + { + "epoch": 5.142002989536622, + "grad_norm": 0.7188577055931091, + "learning_rate": 0.00035724962630792227, + "loss": 3.4008, + "step": 75680 + }, + { + "epoch": 5.142342709607283, + "grad_norm": 0.7510799765586853, + "learning_rate": 0.00035720716129908955, + "loss": 3.4323, + "step": 75685 + }, + { + "epoch": 5.142682429677945, + "grad_norm": 0.8518757224082947, + "learning_rate": 0.0003571646962902568, + "loss": 3.4956, + "step": 75690 + }, + { + "epoch": 5.1430221497486075, + "grad_norm": 0.9330177307128906, + "learning_rate": 0.0003571222312814241, + "loss": 3.4332, + "step": 75695 + }, + { + "epoch": 5.143361869819269, + "grad_norm": 0.863366961479187, + "learning_rate": 0.0003570797662725914, + "loss": 3.5285, + "step": 75700 + }, + { + "epoch": 5.143701589889931, + "grad_norm": 0.7815716862678528, + "learning_rate": 0.00035703730126375867, + "loss": 3.6571, + "step": 75705 + }, + { + "epoch": 5.144041309960593, + "grad_norm": 0.9466231465339661, + "learning_rate": 0.00035699483625492595, + "loss": 3.5089, + "step": 75710 + }, + { + "epoch": 5.144381030031254, + "grad_norm": 0.8485906720161438, + "learning_rate": 0.0003569523712460932, + "loss": 3.5112, + "step": 75715 + }, + { + "epoch": 5.144720750101916, + "grad_norm": 0.8381757140159607, + "learning_rate": 0.0003569099062372605, + "loss": 3.5854, + "step": 75720 + }, + { + "epoch": 5.145060470172578, + "grad_norm": 1.1675273180007935, + "learning_rate": 0.0003568674412284278, + "loss": 3.664, + "step": 75725 + }, + { + "epoch": 5.145400190243239, + "grad_norm": 1.169937252998352, + "learning_rate": 0.00035682497621959507, + "loss": 3.5185, + "step": 75730 + }, + { + "epoch": 5.145739910313901, + "grad_norm": 0.9915021657943726, + "learning_rate": 0.00035678251121076235, + "loss": 3.3516, + "step": 75735 + }, + { + "epoch": 5.1460796303845635, + "grad_norm": 0.7466602921485901, + "learning_rate": 0.00035674004620192963, + "loss": 3.3274, + "step": 75740 + }, + { + "epoch": 5.146419350455225, + "grad_norm": 1.2071409225463867, + "learning_rate": 0.0003566975811930969, + "loss": 3.4037, + "step": 75745 + }, + { + "epoch": 5.146759070525887, + "grad_norm": 0.9449678659439087, + "learning_rate": 0.0003566551161842642, + "loss": 3.6486, + "step": 75750 + }, + { + "epoch": 5.147098790596549, + "grad_norm": 0.8937907218933105, + "learning_rate": 0.00035661265117543147, + "loss": 3.3399, + "step": 75755 + }, + { + "epoch": 5.14743851066721, + "grad_norm": 1.1577504873275757, + "learning_rate": 0.00035657018616659875, + "loss": 3.3409, + "step": 75760 + }, + { + "epoch": 5.147778230737872, + "grad_norm": 0.8591466546058655, + "learning_rate": 0.000356527721157766, + "loss": 3.8231, + "step": 75765 + }, + { + "epoch": 5.148117950808534, + "grad_norm": 1.0236706733703613, + "learning_rate": 0.0003564852561489333, + "loss": 3.4876, + "step": 75770 + }, + { + "epoch": 5.148457670879195, + "grad_norm": 1.023123025894165, + "learning_rate": 0.0003564427911401006, + "loss": 3.4156, + "step": 75775 + }, + { + "epoch": 5.1487973909498574, + "grad_norm": 0.8643351793289185, + "learning_rate": 0.0003564003261312678, + "loss": 3.337, + "step": 75780 + }, + { + "epoch": 5.1491371110205195, + "grad_norm": 1.390157699584961, + "learning_rate": 0.00035635786112243515, + "loss": 3.3757, + "step": 75785 + }, + { + "epoch": 5.149476831091181, + "grad_norm": 0.8398975729942322, + "learning_rate": 0.00035631539611360243, + "loss": 3.4433, + "step": 75790 + }, + { + "epoch": 5.149816551161843, + "grad_norm": 0.9058846831321716, + "learning_rate": 0.00035627293110476965, + "loss": 3.3929, + "step": 75795 + }, + { + "epoch": 5.150156271232504, + "grad_norm": 0.7347093820571899, + "learning_rate": 0.00035623046609593693, + "loss": 3.4943, + "step": 75800 + }, + { + "epoch": 5.150495991303166, + "grad_norm": 0.9572609066963196, + "learning_rate": 0.00035618800108710427, + "loss": 3.2051, + "step": 75805 + }, + { + "epoch": 5.150835711373828, + "grad_norm": 0.859819233417511, + "learning_rate": 0.0003561455360782715, + "loss": 3.6448, + "step": 75810 + }, + { + "epoch": 5.151175431444489, + "grad_norm": 0.9241170883178711, + "learning_rate": 0.0003561030710694388, + "loss": 3.2694, + "step": 75815 + }, + { + "epoch": 5.151515151515151, + "grad_norm": 1.0995243787765503, + "learning_rate": 0.0003560606060606061, + "loss": 3.53, + "step": 75820 + }, + { + "epoch": 5.1518548715858135, + "grad_norm": 1.004812240600586, + "learning_rate": 0.00035601814105177333, + "loss": 3.4095, + "step": 75825 + }, + { + "epoch": 5.152194591656475, + "grad_norm": 0.8107056021690369, + "learning_rate": 0.0003559756760429406, + "loss": 3.6253, + "step": 75830 + }, + { + "epoch": 5.152534311727137, + "grad_norm": 0.842532753944397, + "learning_rate": 0.0003559332110341079, + "loss": 3.6321, + "step": 75835 + }, + { + "epoch": 5.152874031797799, + "grad_norm": 1.0183223485946655, + "learning_rate": 0.0003558907460252752, + "loss": 3.5184, + "step": 75840 + }, + { + "epoch": 5.15321375186846, + "grad_norm": 0.8119643330574036, + "learning_rate": 0.00035584828101644246, + "loss": 3.2714, + "step": 75845 + }, + { + "epoch": 5.153553471939122, + "grad_norm": 0.9546394944190979, + "learning_rate": 0.00035580581600760974, + "loss": 3.488, + "step": 75850 + }, + { + "epoch": 5.153893192009784, + "grad_norm": 0.9778090715408325, + "learning_rate": 0.000355763350998777, + "loss": 3.4579, + "step": 75855 + }, + { + "epoch": 5.154232912080445, + "grad_norm": 0.8450209498405457, + "learning_rate": 0.0003557208859899443, + "loss": 3.5419, + "step": 75860 + }, + { + "epoch": 5.154572632151107, + "grad_norm": 0.9734489321708679, + "learning_rate": 0.0003556784209811116, + "loss": 3.4745, + "step": 75865 + }, + { + "epoch": 5.1549123522217695, + "grad_norm": 0.895376443862915, + "learning_rate": 0.0003556359559722788, + "loss": 3.257, + "step": 75870 + }, + { + "epoch": 5.155252072292431, + "grad_norm": 1.0125356912612915, + "learning_rate": 0.00035559349096344614, + "loss": 3.5082, + "step": 75875 + }, + { + "epoch": 5.155591792363093, + "grad_norm": 1.5127090215682983, + "learning_rate": 0.0003555510259546134, + "loss": 3.3798, + "step": 75880 + }, + { + "epoch": 5.155931512433755, + "grad_norm": 1.0071715116500854, + "learning_rate": 0.00035550856094578064, + "loss": 3.4963, + "step": 75885 + }, + { + "epoch": 5.156271232504416, + "grad_norm": 0.9824620485305786, + "learning_rate": 0.000355466095936948, + "loss": 3.4602, + "step": 75890 + }, + { + "epoch": 5.156610952575078, + "grad_norm": 0.9819023609161377, + "learning_rate": 0.00035542363092811526, + "loss": 3.2881, + "step": 75895 + }, + { + "epoch": 5.15695067264574, + "grad_norm": 0.7317568063735962, + "learning_rate": 0.00035538116591928254, + "loss": 3.3952, + "step": 75900 + }, + { + "epoch": 5.157290392716401, + "grad_norm": 0.8449393510818481, + "learning_rate": 0.00035533870091044976, + "loss": 3.3514, + "step": 75905 + }, + { + "epoch": 5.157630112787063, + "grad_norm": 0.873768150806427, + "learning_rate": 0.0003552962359016171, + "loss": 3.4807, + "step": 75910 + }, + { + "epoch": 5.1579698328577255, + "grad_norm": 0.8254892230033875, + "learning_rate": 0.0003552537708927844, + "loss": 2.9235, + "step": 75915 + }, + { + "epoch": 5.158309552928387, + "grad_norm": 0.9388298988342285, + "learning_rate": 0.0003552113058839516, + "loss": 3.451, + "step": 75920 + }, + { + "epoch": 5.158649272999049, + "grad_norm": 0.8015468120574951, + "learning_rate": 0.00035516884087511894, + "loss": 3.5578, + "step": 75925 + }, + { + "epoch": 5.158988993069711, + "grad_norm": 0.827369749546051, + "learning_rate": 0.0003551263758662862, + "loss": 3.3029, + "step": 75930 + }, + { + "epoch": 5.159328713140372, + "grad_norm": 1.0710492134094238, + "learning_rate": 0.00035508391085745344, + "loss": 3.3515, + "step": 75935 + }, + { + "epoch": 5.159668433211034, + "grad_norm": 1.1144198179244995, + "learning_rate": 0.0003550414458486207, + "loss": 3.5147, + "step": 75940 + }, + { + "epoch": 5.160008153281696, + "grad_norm": 0.9546234607696533, + "learning_rate": 0.00035499898083978806, + "loss": 3.4145, + "step": 75945 + }, + { + "epoch": 5.160347873352357, + "grad_norm": 0.7645173668861389, + "learning_rate": 0.0003549565158309553, + "loss": 3.3884, + "step": 75950 + }, + { + "epoch": 5.160687593423019, + "grad_norm": 0.8998569250106812, + "learning_rate": 0.00035491405082212256, + "loss": 3.2952, + "step": 75955 + }, + { + "epoch": 5.1610273134936815, + "grad_norm": 1.0389199256896973, + "learning_rate": 0.0003548715858132899, + "loss": 3.5468, + "step": 75960 + }, + { + "epoch": 5.161367033564343, + "grad_norm": 1.5395762920379639, + "learning_rate": 0.0003548291208044571, + "loss": 3.6489, + "step": 75965 + }, + { + "epoch": 5.161706753635005, + "grad_norm": 0.9465786814689636, + "learning_rate": 0.0003547866557956244, + "loss": 3.3647, + "step": 75970 + }, + { + "epoch": 5.162046473705667, + "grad_norm": 0.9340637922286987, + "learning_rate": 0.0003547441907867917, + "loss": 3.3644, + "step": 75975 + }, + { + "epoch": 5.162386193776328, + "grad_norm": 0.7783132195472717, + "learning_rate": 0.00035470172577795896, + "loss": 3.6665, + "step": 75980 + }, + { + "epoch": 5.16272591384699, + "grad_norm": 0.7712247967720032, + "learning_rate": 0.00035465926076912624, + "loss": 3.2758, + "step": 75985 + }, + { + "epoch": 5.163065633917652, + "grad_norm": 0.8787266612052917, + "learning_rate": 0.0003546167957602935, + "loss": 3.4424, + "step": 75990 + }, + { + "epoch": 5.163405353988313, + "grad_norm": 0.9124729633331299, + "learning_rate": 0.0003545743307514608, + "loss": 3.5243, + "step": 75995 + }, + { + "epoch": 5.163745074058975, + "grad_norm": 0.6655544638633728, + "learning_rate": 0.0003545318657426281, + "loss": 3.67, + "step": 76000 + }, + { + "epoch": 5.1640847941296375, + "grad_norm": 0.7486337423324585, + "learning_rate": 0.00035448940073379536, + "loss": 3.5914, + "step": 76005 + }, + { + "epoch": 5.164424514200299, + "grad_norm": 1.1658858060836792, + "learning_rate": 0.0003544469357249626, + "loss": 3.3833, + "step": 76010 + }, + { + "epoch": 5.164764234270961, + "grad_norm": 0.7355794310569763, + "learning_rate": 0.0003544044707161299, + "loss": 3.5292, + "step": 76015 + }, + { + "epoch": 5.165103954341623, + "grad_norm": 1.1136115789413452, + "learning_rate": 0.0003543620057072972, + "loss": 3.6442, + "step": 76020 + }, + { + "epoch": 5.165443674412284, + "grad_norm": 1.0001121759414673, + "learning_rate": 0.00035431954069846443, + "loss": 3.2544, + "step": 76025 + }, + { + "epoch": 5.165783394482946, + "grad_norm": 1.1436100006103516, + "learning_rate": 0.00035427707568963176, + "loss": 3.2595, + "step": 76030 + }, + { + "epoch": 5.166123114553608, + "grad_norm": 1.583150029182434, + "learning_rate": 0.00035423461068079904, + "loss": 3.5627, + "step": 76035 + }, + { + "epoch": 5.166462834624269, + "grad_norm": 1.078120231628418, + "learning_rate": 0.00035419214567196627, + "loss": 3.4065, + "step": 76040 + }, + { + "epoch": 5.166802554694931, + "grad_norm": 0.90113365650177, + "learning_rate": 0.0003541496806631336, + "loss": 3.3448, + "step": 76045 + }, + { + "epoch": 5.1671422747655935, + "grad_norm": 0.9087955355644226, + "learning_rate": 0.0003541072156543009, + "loss": 3.5512, + "step": 76050 + }, + { + "epoch": 5.167481994836255, + "grad_norm": 1.0898863077163696, + "learning_rate": 0.0003540647506454681, + "loss": 3.4231, + "step": 76055 + }, + { + "epoch": 5.167821714906917, + "grad_norm": 0.9989370107650757, + "learning_rate": 0.0003540222856366354, + "loss": 3.2217, + "step": 76060 + }, + { + "epoch": 5.168161434977579, + "grad_norm": 0.7573986053466797, + "learning_rate": 0.0003539798206278027, + "loss": 3.5368, + "step": 76065 + }, + { + "epoch": 5.16850115504824, + "grad_norm": 0.8380848169326782, + "learning_rate": 0.00035393735561897, + "loss": 3.7962, + "step": 76070 + }, + { + "epoch": 5.168840875118902, + "grad_norm": 1.0448063611984253, + "learning_rate": 0.00035389489061013723, + "loss": 3.3586, + "step": 76075 + }, + { + "epoch": 5.169180595189564, + "grad_norm": 1.0887398719787598, + "learning_rate": 0.00035385242560130456, + "loss": 3.4509, + "step": 76080 + }, + { + "epoch": 5.169520315260225, + "grad_norm": 1.0384142398834229, + "learning_rate": 0.00035380996059247184, + "loss": 3.608, + "step": 76085 + }, + { + "epoch": 5.1698600353308874, + "grad_norm": 1.0547727346420288, + "learning_rate": 0.00035376749558363907, + "loss": 3.6586, + "step": 76090 + }, + { + "epoch": 5.1701997554015495, + "grad_norm": 0.9284372329711914, + "learning_rate": 0.00035372503057480635, + "loss": 3.5019, + "step": 76095 + }, + { + "epoch": 5.170539475472211, + "grad_norm": 0.981079638004303, + "learning_rate": 0.0003536825655659737, + "loss": 3.4747, + "step": 76100 + }, + { + "epoch": 5.170879195542873, + "grad_norm": 0.8659055829048157, + "learning_rate": 0.0003536401005571409, + "loss": 3.4038, + "step": 76105 + }, + { + "epoch": 5.171218915613535, + "grad_norm": 0.903406023979187, + "learning_rate": 0.0003535976355483082, + "loss": 3.3725, + "step": 76110 + }, + { + "epoch": 5.171558635684196, + "grad_norm": 1.0536607503890991, + "learning_rate": 0.0003535551705394755, + "loss": 3.6094, + "step": 76115 + }, + { + "epoch": 5.171898355754858, + "grad_norm": 0.8169236183166504, + "learning_rate": 0.00035351270553064275, + "loss": 3.6252, + "step": 76120 + }, + { + "epoch": 5.17223807582552, + "grad_norm": 0.917647659778595, + "learning_rate": 0.00035347024052181003, + "loss": 3.1956, + "step": 76125 + }, + { + "epoch": 5.172577795896181, + "grad_norm": 0.8692314624786377, + "learning_rate": 0.0003534277755129773, + "loss": 3.4552, + "step": 76130 + }, + { + "epoch": 5.1729175159668435, + "grad_norm": 1.124666452407837, + "learning_rate": 0.0003533853105041446, + "loss": 3.5335, + "step": 76135 + }, + { + "epoch": 5.1732572360375055, + "grad_norm": 0.7767394781112671, + "learning_rate": 0.00035334284549531187, + "loss": 3.2862, + "step": 76140 + }, + { + "epoch": 5.173596956108167, + "grad_norm": 0.9920660853385925, + "learning_rate": 0.00035330038048647915, + "loss": 3.2546, + "step": 76145 + }, + { + "epoch": 5.173936676178829, + "grad_norm": 0.7940046787261963, + "learning_rate": 0.00035325791547764643, + "loss": 3.3597, + "step": 76150 + }, + { + "epoch": 5.174276396249491, + "grad_norm": 0.9845365285873413, + "learning_rate": 0.0003532154504688137, + "loss": 3.4329, + "step": 76155 + }, + { + "epoch": 5.174616116320152, + "grad_norm": 0.8418992757797241, + "learning_rate": 0.000353172985459981, + "loss": 3.6057, + "step": 76160 + }, + { + "epoch": 5.174955836390814, + "grad_norm": 0.9770155549049377, + "learning_rate": 0.0003531305204511482, + "loss": 3.474, + "step": 76165 + }, + { + "epoch": 5.175295556461475, + "grad_norm": 0.8373923897743225, + "learning_rate": 0.00035308805544231555, + "loss": 3.2185, + "step": 76170 + }, + { + "epoch": 5.175635276532137, + "grad_norm": 0.9353631734848022, + "learning_rate": 0.00035304559043348283, + "loss": 3.5639, + "step": 76175 + }, + { + "epoch": 5.1759749966027995, + "grad_norm": 0.9114130735397339, + "learning_rate": 0.00035300312542465006, + "loss": 3.5495, + "step": 76180 + }, + { + "epoch": 5.176314716673461, + "grad_norm": 0.7746999263763428, + "learning_rate": 0.0003529606604158174, + "loss": 3.6279, + "step": 76185 + }, + { + "epoch": 5.176654436744123, + "grad_norm": 1.0255323648452759, + "learning_rate": 0.00035291819540698467, + "loss": 3.283, + "step": 76190 + }, + { + "epoch": 5.176994156814785, + "grad_norm": 0.9495078325271606, + "learning_rate": 0.0003528757303981519, + "loss": 3.5158, + "step": 76195 + }, + { + "epoch": 5.177333876885446, + "grad_norm": 0.9692737460136414, + "learning_rate": 0.0003528332653893192, + "loss": 3.7453, + "step": 76200 + }, + { + "epoch": 5.177673596956108, + "grad_norm": 1.1697654724121094, + "learning_rate": 0.0003527908003804865, + "loss": 3.3254, + "step": 76205 + }, + { + "epoch": 5.17801331702677, + "grad_norm": 0.97832852602005, + "learning_rate": 0.00035274833537165374, + "loss": 3.548, + "step": 76210 + }, + { + "epoch": 5.178353037097431, + "grad_norm": 1.0987861156463623, + "learning_rate": 0.000352705870362821, + "loss": 3.5084, + "step": 76215 + }, + { + "epoch": 5.178692757168093, + "grad_norm": 0.7683600187301636, + "learning_rate": 0.00035266340535398835, + "loss": 3.8778, + "step": 76220 + }, + { + "epoch": 5.1790324772387555, + "grad_norm": 0.8573612570762634, + "learning_rate": 0.0003526209403451556, + "loss": 3.7402, + "step": 76225 + }, + { + "epoch": 5.179372197309417, + "grad_norm": 0.8639705181121826, + "learning_rate": 0.00035257847533632286, + "loss": 3.3174, + "step": 76230 + }, + { + "epoch": 5.179711917380079, + "grad_norm": 0.909771203994751, + "learning_rate": 0.00035253601032749014, + "loss": 3.1867, + "step": 76235 + }, + { + "epoch": 5.180051637450741, + "grad_norm": 0.912753701210022, + "learning_rate": 0.0003524935453186575, + "loss": 3.5072, + "step": 76240 + }, + { + "epoch": 5.180391357521402, + "grad_norm": 0.7627174854278564, + "learning_rate": 0.0003524510803098247, + "loss": 3.4902, + "step": 76245 + }, + { + "epoch": 5.180731077592064, + "grad_norm": 0.8061987161636353, + "learning_rate": 0.000352408615300992, + "loss": 3.5993, + "step": 76250 + }, + { + "epoch": 5.181070797662726, + "grad_norm": 0.9024948477745056, + "learning_rate": 0.0003523661502921593, + "loss": 3.2906, + "step": 76255 + }, + { + "epoch": 5.181410517733387, + "grad_norm": 0.9783357977867126, + "learning_rate": 0.00035232368528332654, + "loss": 3.4517, + "step": 76260 + }, + { + "epoch": 5.181750237804049, + "grad_norm": 1.037613868713379, + "learning_rate": 0.0003522812202744938, + "loss": 3.5339, + "step": 76265 + }, + { + "epoch": 5.1820899578747115, + "grad_norm": 0.9578565955162048, + "learning_rate": 0.0003522387552656611, + "loss": 3.413, + "step": 76270 + }, + { + "epoch": 5.182429677945373, + "grad_norm": 0.7837210297584534, + "learning_rate": 0.0003521962902568284, + "loss": 3.5118, + "step": 76275 + }, + { + "epoch": 5.182769398016035, + "grad_norm": 0.910256028175354, + "learning_rate": 0.00035215382524799566, + "loss": 3.4385, + "step": 76280 + }, + { + "epoch": 5.183109118086697, + "grad_norm": 1.1504288911819458, + "learning_rate": 0.00035211136023916294, + "loss": 3.3793, + "step": 76285 + }, + { + "epoch": 5.183448838157358, + "grad_norm": 0.7109076976776123, + "learning_rate": 0.0003520688952303302, + "loss": 3.6799, + "step": 76290 + }, + { + "epoch": 5.18378855822802, + "grad_norm": 0.9681477546691895, + "learning_rate": 0.0003520264302214975, + "loss": 3.0246, + "step": 76295 + }, + { + "epoch": 5.184128278298682, + "grad_norm": 1.149387001991272, + "learning_rate": 0.0003519839652126648, + "loss": 3.4356, + "step": 76300 + }, + { + "epoch": 5.184467998369343, + "grad_norm": 0.9286701083183289, + "learning_rate": 0.000351941500203832, + "loss": 3.759, + "step": 76305 + }, + { + "epoch": 5.184807718440005, + "grad_norm": 0.9569326639175415, + "learning_rate": 0.00035189903519499934, + "loss": 3.4523, + "step": 76310 + }, + { + "epoch": 5.1851474385106675, + "grad_norm": 0.8626444935798645, + "learning_rate": 0.0003518565701861666, + "loss": 3.4732, + "step": 76315 + }, + { + "epoch": 5.185487158581329, + "grad_norm": 0.914947509765625, + "learning_rate": 0.00035181410517733385, + "loss": 3.4052, + "step": 76320 + }, + { + "epoch": 5.185826878651991, + "grad_norm": 1.0069433450698853, + "learning_rate": 0.0003517716401685012, + "loss": 3.3944, + "step": 76325 + }, + { + "epoch": 5.186166598722653, + "grad_norm": 1.0876588821411133, + "learning_rate": 0.00035172917515966846, + "loss": 3.4808, + "step": 76330 + }, + { + "epoch": 5.186506318793314, + "grad_norm": 0.8869509100914001, + "learning_rate": 0.0003516867101508357, + "loss": 3.4608, + "step": 76335 + }, + { + "epoch": 5.186846038863976, + "grad_norm": 1.1144438982009888, + "learning_rate": 0.000351644245142003, + "loss": 3.374, + "step": 76340 + }, + { + "epoch": 5.187185758934638, + "grad_norm": 0.8701178431510925, + "learning_rate": 0.0003516017801331703, + "loss": 3.5223, + "step": 76345 + }, + { + "epoch": 5.187525479005299, + "grad_norm": 0.9228264689445496, + "learning_rate": 0.0003515593151243375, + "loss": 3.4482, + "step": 76350 + }, + { + "epoch": 5.187865199075961, + "grad_norm": 0.9200606942176819, + "learning_rate": 0.0003515168501155048, + "loss": 3.5946, + "step": 76355 + }, + { + "epoch": 5.1882049191466235, + "grad_norm": 0.8377779722213745, + "learning_rate": 0.00035147438510667214, + "loss": 3.316, + "step": 76360 + }, + { + "epoch": 5.188544639217285, + "grad_norm": 0.7952613234519958, + "learning_rate": 0.00035143192009783937, + "loss": 3.3757, + "step": 76365 + }, + { + "epoch": 5.188884359287947, + "grad_norm": 0.7033195495605469, + "learning_rate": 0.00035138945508900665, + "loss": 3.6887, + "step": 76370 + }, + { + "epoch": 5.189224079358609, + "grad_norm": 0.8251524567604065, + "learning_rate": 0.000351346990080174, + "loss": 3.3597, + "step": 76375 + }, + { + "epoch": 5.18956379942927, + "grad_norm": 0.8204817175865173, + "learning_rate": 0.0003513045250713412, + "loss": 3.2293, + "step": 76380 + }, + { + "epoch": 5.189903519499932, + "grad_norm": 0.9454706311225891, + "learning_rate": 0.0003512620600625085, + "loss": 3.5732, + "step": 76385 + }, + { + "epoch": 5.190243239570594, + "grad_norm": 0.9697344303131104, + "learning_rate": 0.00035121959505367577, + "loss": 3.5723, + "step": 76390 + }, + { + "epoch": 5.190582959641255, + "grad_norm": 0.9776793718338013, + "learning_rate": 0.00035117713004484305, + "loss": 3.7623, + "step": 76395 + }, + { + "epoch": 5.1909226797119175, + "grad_norm": 0.8809166550636292, + "learning_rate": 0.0003511346650360103, + "loss": 3.3736, + "step": 76400 + }, + { + "epoch": 5.1912623997825795, + "grad_norm": 0.8176227807998657, + "learning_rate": 0.0003510922000271776, + "loss": 3.3903, + "step": 76405 + }, + { + "epoch": 5.191602119853241, + "grad_norm": 0.901653528213501, + "learning_rate": 0.00035104973501834494, + "loss": 3.582, + "step": 76410 + }, + { + "epoch": 5.191941839923903, + "grad_norm": 0.8055301308631897, + "learning_rate": 0.00035100727000951217, + "loss": 3.2316, + "step": 76415 + }, + { + "epoch": 5.192281559994565, + "grad_norm": 0.7280771136283875, + "learning_rate": 0.00035096480500067945, + "loss": 3.1486, + "step": 76420 + }, + { + "epoch": 5.192621280065226, + "grad_norm": 0.73358154296875, + "learning_rate": 0.00035092233999184673, + "loss": 3.7185, + "step": 76425 + }, + { + "epoch": 5.192961000135888, + "grad_norm": 0.9224925637245178, + "learning_rate": 0.000350879874983014, + "loss": 3.3484, + "step": 76430 + }, + { + "epoch": 5.19330072020655, + "grad_norm": 0.7639920711517334, + "learning_rate": 0.0003508374099741813, + "loss": 3.387, + "step": 76435 + }, + { + "epoch": 5.193640440277211, + "grad_norm": 0.8350674510002136, + "learning_rate": 0.00035079494496534857, + "loss": 3.4073, + "step": 76440 + }, + { + "epoch": 5.1939801603478735, + "grad_norm": 0.9176893830299377, + "learning_rate": 0.00035075247995651585, + "loss": 3.7313, + "step": 76445 + }, + { + "epoch": 5.1943198804185355, + "grad_norm": 0.7075564861297607, + "learning_rate": 0.00035071001494768313, + "loss": 3.6113, + "step": 76450 + }, + { + "epoch": 5.194659600489197, + "grad_norm": 0.9309481978416443, + "learning_rate": 0.0003506675499388504, + "loss": 3.5021, + "step": 76455 + }, + { + "epoch": 5.194999320559859, + "grad_norm": 0.7149673104286194, + "learning_rate": 0.00035062508493001763, + "loss": 3.6009, + "step": 76460 + }, + { + "epoch": 5.195339040630521, + "grad_norm": 0.8282140493392944, + "learning_rate": 0.00035058261992118497, + "loss": 3.5054, + "step": 76465 + }, + { + "epoch": 5.195678760701182, + "grad_norm": 0.8848045468330383, + "learning_rate": 0.00035054015491235225, + "loss": 3.4156, + "step": 76470 + }, + { + "epoch": 5.196018480771844, + "grad_norm": 0.9471142292022705, + "learning_rate": 0.0003504976899035195, + "loss": 3.5037, + "step": 76475 + }, + { + "epoch": 5.196358200842505, + "grad_norm": 0.7341698408126831, + "learning_rate": 0.0003504552248946868, + "loss": 3.539, + "step": 76480 + }, + { + "epoch": 5.196697920913167, + "grad_norm": 0.9290366768836975, + "learning_rate": 0.0003504127598858541, + "loss": 3.52, + "step": 76485 + }, + { + "epoch": 5.1970376409838295, + "grad_norm": 0.8360311388969421, + "learning_rate": 0.0003503702948770213, + "loss": 3.5625, + "step": 76490 + }, + { + "epoch": 5.197377361054491, + "grad_norm": 1.1073541641235352, + "learning_rate": 0.0003503278298681886, + "loss": 3.3813, + "step": 76495 + }, + { + "epoch": 5.197717081125153, + "grad_norm": 1.1822370290756226, + "learning_rate": 0.00035028536485935593, + "loss": 3.4282, + "step": 76500 + }, + { + "epoch": 5.198056801195815, + "grad_norm": 0.9698997139930725, + "learning_rate": 0.00035024289985052315, + "loss": 3.3957, + "step": 76505 + }, + { + "epoch": 5.198396521266476, + "grad_norm": 0.9957197904586792, + "learning_rate": 0.00035020043484169043, + "loss": 3.449, + "step": 76510 + }, + { + "epoch": 5.198736241337138, + "grad_norm": 0.7667668461799622, + "learning_rate": 0.00035015796983285777, + "loss": 3.769, + "step": 76515 + }, + { + "epoch": 5.1990759614078, + "grad_norm": 0.6687655448913574, + "learning_rate": 0.000350115504824025, + "loss": 3.3098, + "step": 76520 + }, + { + "epoch": 5.199415681478461, + "grad_norm": 1.1261318922042847, + "learning_rate": 0.0003500730398151923, + "loss": 3.458, + "step": 76525 + }, + { + "epoch": 5.199755401549123, + "grad_norm": 0.70208340883255, + "learning_rate": 0.00035003057480635955, + "loss": 3.4758, + "step": 76530 + }, + { + "epoch": 5.2000951216197855, + "grad_norm": 0.9253193736076355, + "learning_rate": 0.00034998810979752683, + "loss": 3.7037, + "step": 76535 + }, + { + "epoch": 5.200434841690447, + "grad_norm": 1.0528656244277954, + "learning_rate": 0.0003499456447886941, + "loss": 3.5488, + "step": 76540 + }, + { + "epoch": 5.200774561761109, + "grad_norm": 0.9178515672683716, + "learning_rate": 0.0003499031797798614, + "loss": 3.4896, + "step": 76545 + }, + { + "epoch": 5.201114281831771, + "grad_norm": 1.137354850769043, + "learning_rate": 0.0003498607147710287, + "loss": 3.4848, + "step": 76550 + }, + { + "epoch": 5.201454001902432, + "grad_norm": 0.8029659390449524, + "learning_rate": 0.00034981824976219596, + "loss": 3.2579, + "step": 76555 + }, + { + "epoch": 5.201793721973094, + "grad_norm": 0.9434231519699097, + "learning_rate": 0.00034977578475336324, + "loss": 3.431, + "step": 76560 + }, + { + "epoch": 5.202133442043756, + "grad_norm": 0.945254385471344, + "learning_rate": 0.00034973331974453046, + "loss": 3.465, + "step": 76565 + }, + { + "epoch": 5.202473162114417, + "grad_norm": 1.1073884963989258, + "learning_rate": 0.0003496908547356978, + "loss": 3.3489, + "step": 76570 + }, + { + "epoch": 5.202812882185079, + "grad_norm": 0.787460446357727, + "learning_rate": 0.0003496483897268651, + "loss": 3.579, + "step": 76575 + }, + { + "epoch": 5.2031526022557415, + "grad_norm": 0.9890434741973877, + "learning_rate": 0.00034960592471803236, + "loss": 3.6738, + "step": 76580 + }, + { + "epoch": 5.203492322326403, + "grad_norm": 0.7592213749885559, + "learning_rate": 0.00034956345970919964, + "loss": 3.24, + "step": 76585 + }, + { + "epoch": 5.203832042397065, + "grad_norm": 0.9535008668899536, + "learning_rate": 0.0003495209947003669, + "loss": 3.4982, + "step": 76590 + }, + { + "epoch": 5.204171762467727, + "grad_norm": 0.7897887825965881, + "learning_rate": 0.0003494785296915342, + "loss": 3.3998, + "step": 76595 + }, + { + "epoch": 5.204511482538388, + "grad_norm": 0.931279182434082, + "learning_rate": 0.0003494360646827015, + "loss": 3.5603, + "step": 76600 + }, + { + "epoch": 5.20485120260905, + "grad_norm": 0.8753998875617981, + "learning_rate": 0.00034939359967386876, + "loss": 3.2714, + "step": 76605 + }, + { + "epoch": 5.205190922679712, + "grad_norm": 0.8961510062217712, + "learning_rate": 0.00034935113466503604, + "loss": 3.3529, + "step": 76610 + }, + { + "epoch": 5.205530642750373, + "grad_norm": 0.928054690361023, + "learning_rate": 0.00034930866965620326, + "loss": 3.5166, + "step": 76615 + }, + { + "epoch": 5.205870362821035, + "grad_norm": 0.8334048986434937, + "learning_rate": 0.0003492662046473706, + "loss": 3.5691, + "step": 76620 + }, + { + "epoch": 5.2062100828916975, + "grad_norm": 1.1574712991714478, + "learning_rate": 0.0003492237396385379, + "loss": 3.6802, + "step": 76625 + }, + { + "epoch": 5.206549802962359, + "grad_norm": 0.8638137578964233, + "learning_rate": 0.0003491812746297051, + "loss": 3.4555, + "step": 76630 + }, + { + "epoch": 5.206889523033021, + "grad_norm": 0.9776730537414551, + "learning_rate": 0.00034913880962087244, + "loss": 3.642, + "step": 76635 + }, + { + "epoch": 5.207229243103683, + "grad_norm": 0.8237279653549194, + "learning_rate": 0.0003490963446120397, + "loss": 3.4523, + "step": 76640 + }, + { + "epoch": 5.207568963174344, + "grad_norm": 0.908808171749115, + "learning_rate": 0.00034905387960320694, + "loss": 3.2078, + "step": 76645 + }, + { + "epoch": 5.207908683245006, + "grad_norm": 0.9231775999069214, + "learning_rate": 0.0003490114145943742, + "loss": 3.3212, + "step": 76650 + }, + { + "epoch": 5.208248403315668, + "grad_norm": 0.8304939866065979, + "learning_rate": 0.00034896894958554156, + "loss": 3.6634, + "step": 76655 + }, + { + "epoch": 5.208588123386329, + "grad_norm": 0.8265734910964966, + "learning_rate": 0.0003489264845767088, + "loss": 3.2778, + "step": 76660 + }, + { + "epoch": 5.2089278434569914, + "grad_norm": 0.8625829219818115, + "learning_rate": 0.00034888401956787606, + "loss": 3.5005, + "step": 76665 + }, + { + "epoch": 5.2092675635276535, + "grad_norm": 0.9527860283851624, + "learning_rate": 0.0003488415545590434, + "loss": 3.2424, + "step": 76670 + }, + { + "epoch": 5.209607283598315, + "grad_norm": 0.8635834455490112, + "learning_rate": 0.0003487990895502106, + "loss": 3.4766, + "step": 76675 + }, + { + "epoch": 5.209947003668977, + "grad_norm": 0.8672298789024353, + "learning_rate": 0.0003487566245413779, + "loss": 3.4566, + "step": 76680 + }, + { + "epoch": 5.210286723739639, + "grad_norm": 0.9842963814735413, + "learning_rate": 0.0003487141595325452, + "loss": 3.3418, + "step": 76685 + }, + { + "epoch": 5.2106264438103, + "grad_norm": 0.9216187000274658, + "learning_rate": 0.00034867169452371246, + "loss": 3.2503, + "step": 76690 + }, + { + "epoch": 5.210966163880962, + "grad_norm": 0.8554538488388062, + "learning_rate": 0.00034862922951487974, + "loss": 3.4797, + "step": 76695 + }, + { + "epoch": 5.211305883951624, + "grad_norm": 0.7324966192245483, + "learning_rate": 0.000348586764506047, + "loss": 3.5288, + "step": 76700 + }, + { + "epoch": 5.211645604022285, + "grad_norm": 0.9857914447784424, + "learning_rate": 0.0003485442994972143, + "loss": 3.3858, + "step": 76705 + }, + { + "epoch": 5.2119853240929475, + "grad_norm": 0.9953253269195557, + "learning_rate": 0.0003485018344883816, + "loss": 3.4752, + "step": 76710 + }, + { + "epoch": 5.2123250441636095, + "grad_norm": 1.0670654773712158, + "learning_rate": 0.00034845936947954886, + "loss": 3.3798, + "step": 76715 + }, + { + "epoch": 5.212664764234271, + "grad_norm": 0.9089499115943909, + "learning_rate": 0.0003484169044707161, + "loss": 3.6627, + "step": 76720 + }, + { + "epoch": 5.213004484304933, + "grad_norm": 0.7362671494483948, + "learning_rate": 0.0003483744394618834, + "loss": 3.55, + "step": 76725 + }, + { + "epoch": 5.213344204375595, + "grad_norm": 1.0961148738861084, + "learning_rate": 0.0003483319744530507, + "loss": 3.4246, + "step": 76730 + }, + { + "epoch": 5.213683924446256, + "grad_norm": 0.8753455281257629, + "learning_rate": 0.00034828950944421793, + "loss": 3.5948, + "step": 76735 + }, + { + "epoch": 5.214023644516918, + "grad_norm": 1.0000290870666504, + "learning_rate": 0.00034824704443538526, + "loss": 3.4863, + "step": 76740 + }, + { + "epoch": 5.21436336458758, + "grad_norm": 0.9124593734741211, + "learning_rate": 0.00034820457942655254, + "loss": 3.4278, + "step": 76745 + }, + { + "epoch": 5.214703084658241, + "grad_norm": 1.09718656539917, + "learning_rate": 0.0003481621144177198, + "loss": 3.2415, + "step": 76750 + }, + { + "epoch": 5.2150428047289035, + "grad_norm": 1.1463552713394165, + "learning_rate": 0.00034811964940888705, + "loss": 3.2756, + "step": 76755 + }, + { + "epoch": 5.2153825247995655, + "grad_norm": 0.8576882481575012, + "learning_rate": 0.0003480771844000544, + "loss": 3.4802, + "step": 76760 + }, + { + "epoch": 5.215722244870227, + "grad_norm": 1.0838369131088257, + "learning_rate": 0.00034803471939122166, + "loss": 3.4458, + "step": 76765 + }, + { + "epoch": 5.216061964940889, + "grad_norm": 1.0962196588516235, + "learning_rate": 0.0003479922543823889, + "loss": 3.3023, + "step": 76770 + }, + { + "epoch": 5.216401685011551, + "grad_norm": 0.8995252251625061, + "learning_rate": 0.0003479497893735562, + "loss": 3.346, + "step": 76775 + }, + { + "epoch": 5.216741405082212, + "grad_norm": 0.9401818513870239, + "learning_rate": 0.0003479073243647235, + "loss": 3.3527, + "step": 76780 + }, + { + "epoch": 5.217081125152874, + "grad_norm": 1.015880823135376, + "learning_rate": 0.00034786485935589073, + "loss": 3.5075, + "step": 76785 + }, + { + "epoch": 5.217420845223536, + "grad_norm": 1.1867128610610962, + "learning_rate": 0.000347822394347058, + "loss": 3.7586, + "step": 76790 + }, + { + "epoch": 5.217760565294197, + "grad_norm": 1.0693485736846924, + "learning_rate": 0.00034777992933822534, + "loss": 3.2909, + "step": 76795 + }, + { + "epoch": 5.2181002853648595, + "grad_norm": 0.9281983375549316, + "learning_rate": 0.00034773746432939257, + "loss": 3.3984, + "step": 76800 + }, + { + "epoch": 5.2184400054355216, + "grad_norm": 0.6397440433502197, + "learning_rate": 0.00034769499932055985, + "loss": 3.6465, + "step": 76805 + }, + { + "epoch": 5.218779725506183, + "grad_norm": 0.805755615234375, + "learning_rate": 0.0003476525343117272, + "loss": 3.3819, + "step": 76810 + }, + { + "epoch": 5.219119445576845, + "grad_norm": 0.9991214275360107, + "learning_rate": 0.0003476100693028944, + "loss": 3.4524, + "step": 76815 + }, + { + "epoch": 5.219459165647507, + "grad_norm": 0.8811479210853577, + "learning_rate": 0.0003475676042940617, + "loss": 3.3575, + "step": 76820 + }, + { + "epoch": 5.219798885718168, + "grad_norm": 0.809653639793396, + "learning_rate": 0.00034752513928522897, + "loss": 3.5167, + "step": 76825 + }, + { + "epoch": 5.22013860578883, + "grad_norm": 0.8513479828834534, + "learning_rate": 0.00034748267427639625, + "loss": 3.3427, + "step": 76830 + }, + { + "epoch": 5.220478325859492, + "grad_norm": 0.847898006439209, + "learning_rate": 0.00034744020926756353, + "loss": 3.2258, + "step": 76835 + }, + { + "epoch": 5.220818045930153, + "grad_norm": 1.0589715242385864, + "learning_rate": 0.0003473977442587308, + "loss": 3.4146, + "step": 76840 + }, + { + "epoch": 5.2211577660008155, + "grad_norm": 0.682145357131958, + "learning_rate": 0.0003473552792498981, + "loss": 3.3282, + "step": 76845 + }, + { + "epoch": 5.221497486071477, + "grad_norm": 0.9379629492759705, + "learning_rate": 0.00034731281424106537, + "loss": 3.5568, + "step": 76850 + }, + { + "epoch": 5.221837206142139, + "grad_norm": 0.8612335324287415, + "learning_rate": 0.00034727034923223265, + "loss": 3.3878, + "step": 76855 + }, + { + "epoch": 5.222176926212801, + "grad_norm": 1.0778228044509888, + "learning_rate": 0.0003472278842233999, + "loss": 3.6026, + "step": 76860 + }, + { + "epoch": 5.222516646283462, + "grad_norm": 1.1476625204086304, + "learning_rate": 0.0003471854192145672, + "loss": 3.5481, + "step": 76865 + }, + { + "epoch": 5.222856366354124, + "grad_norm": 0.8820012211799622, + "learning_rate": 0.0003471429542057345, + "loss": 3.5452, + "step": 76870 + }, + { + "epoch": 5.223196086424786, + "grad_norm": 0.9331788420677185, + "learning_rate": 0.0003471004891969017, + "loss": 3.4671, + "step": 76875 + }, + { + "epoch": 5.223535806495447, + "grad_norm": 1.0273224115371704, + "learning_rate": 0.00034705802418806905, + "loss": 3.2284, + "step": 76880 + }, + { + "epoch": 5.223875526566109, + "grad_norm": 0.8899880647659302, + "learning_rate": 0.00034701555917923633, + "loss": 3.3736, + "step": 76885 + }, + { + "epoch": 5.2242152466367715, + "grad_norm": 0.7657882571220398, + "learning_rate": 0.00034697309417040356, + "loss": 3.6028, + "step": 76890 + }, + { + "epoch": 5.224554966707433, + "grad_norm": 0.9197964668273926, + "learning_rate": 0.0003469306291615709, + "loss": 3.248, + "step": 76895 + }, + { + "epoch": 5.224894686778095, + "grad_norm": 1.0294588804244995, + "learning_rate": 0.00034688816415273817, + "loss": 3.4955, + "step": 76900 + }, + { + "epoch": 5.225234406848757, + "grad_norm": 0.8837618231773376, + "learning_rate": 0.0003468456991439054, + "loss": 3.6551, + "step": 76905 + }, + { + "epoch": 5.225574126919418, + "grad_norm": 0.8975844383239746, + "learning_rate": 0.0003468032341350727, + "loss": 3.4292, + "step": 76910 + }, + { + "epoch": 5.22591384699008, + "grad_norm": 0.9473903775215149, + "learning_rate": 0.00034676076912624, + "loss": 3.4393, + "step": 76915 + }, + { + "epoch": 5.226253567060742, + "grad_norm": 1.0238534212112427, + "learning_rate": 0.0003467183041174073, + "loss": 3.5072, + "step": 76920 + }, + { + "epoch": 5.226593287131403, + "grad_norm": 0.8331368565559387, + "learning_rate": 0.0003466758391085745, + "loss": 3.5172, + "step": 76925 + }, + { + "epoch": 5.226933007202065, + "grad_norm": 1.4924378395080566, + "learning_rate": 0.00034663337409974185, + "loss": 3.6588, + "step": 76930 + }, + { + "epoch": 5.2272727272727275, + "grad_norm": 0.8971142768859863, + "learning_rate": 0.00034659090909090913, + "loss": 3.2659, + "step": 76935 + }, + { + "epoch": 5.227612447343389, + "grad_norm": 0.7911790609359741, + "learning_rate": 0.00034654844408207636, + "loss": 3.3704, + "step": 76940 + }, + { + "epoch": 5.227952167414051, + "grad_norm": 0.837675154209137, + "learning_rate": 0.00034650597907324364, + "loss": 3.5078, + "step": 76945 + }, + { + "epoch": 5.228291887484713, + "grad_norm": 0.8000457882881165, + "learning_rate": 0.00034646351406441097, + "loss": 3.4418, + "step": 76950 + }, + { + "epoch": 5.228631607555374, + "grad_norm": 0.750470757484436, + "learning_rate": 0.0003464210490555782, + "loss": 3.5679, + "step": 76955 + }, + { + "epoch": 5.228971327626036, + "grad_norm": 0.9276978373527527, + "learning_rate": 0.0003463785840467455, + "loss": 3.4533, + "step": 76960 + }, + { + "epoch": 5.229311047696698, + "grad_norm": 0.8411605358123779, + "learning_rate": 0.0003463361190379128, + "loss": 3.4244, + "step": 76965 + }, + { + "epoch": 5.229650767767359, + "grad_norm": 0.9118165969848633, + "learning_rate": 0.00034629365402908004, + "loss": 3.5258, + "step": 76970 + }, + { + "epoch": 5.2299904878380215, + "grad_norm": 1.0030111074447632, + "learning_rate": 0.0003462511890202473, + "loss": 3.6466, + "step": 76975 + }, + { + "epoch": 5.2303302079086835, + "grad_norm": 0.8474104404449463, + "learning_rate": 0.0003462087240114146, + "loss": 3.5637, + "step": 76980 + }, + { + "epoch": 5.230669927979345, + "grad_norm": 0.94627445936203, + "learning_rate": 0.0003461662590025819, + "loss": 3.372, + "step": 76985 + }, + { + "epoch": 5.231009648050007, + "grad_norm": 1.0711040496826172, + "learning_rate": 0.00034612379399374916, + "loss": 3.3669, + "step": 76990 + }, + { + "epoch": 5.231349368120669, + "grad_norm": 0.7858566045761108, + "learning_rate": 0.00034608132898491644, + "loss": 3.5099, + "step": 76995 + }, + { + "epoch": 5.23168908819133, + "grad_norm": 0.8284093141555786, + "learning_rate": 0.0003460388639760837, + "loss": 3.4334, + "step": 77000 + }, + { + "epoch": 5.232028808261992, + "grad_norm": 0.8450635671615601, + "learning_rate": 0.000345996398967251, + "loss": 3.4288, + "step": 77005 + }, + { + "epoch": 5.232368528332654, + "grad_norm": 0.9665061831474304, + "learning_rate": 0.0003459539339584183, + "loss": 3.2788, + "step": 77010 + }, + { + "epoch": 5.232708248403315, + "grad_norm": 0.8939473032951355, + "learning_rate": 0.0003459114689495855, + "loss": 3.3627, + "step": 77015 + }, + { + "epoch": 5.2330479684739775, + "grad_norm": 1.0376514196395874, + "learning_rate": 0.00034586900394075284, + "loss": 3.403, + "step": 77020 + }, + { + "epoch": 5.2333876885446395, + "grad_norm": 0.913378894329071, + "learning_rate": 0.0003458265389319201, + "loss": 3.5094, + "step": 77025 + }, + { + "epoch": 5.233727408615301, + "grad_norm": 0.8235164284706116, + "learning_rate": 0.00034578407392308735, + "loss": 3.5257, + "step": 77030 + }, + { + "epoch": 5.234067128685963, + "grad_norm": 1.1945749521255493, + "learning_rate": 0.0003457416089142547, + "loss": 3.5233, + "step": 77035 + }, + { + "epoch": 5.234406848756625, + "grad_norm": 0.8422659039497375, + "learning_rate": 0.00034569914390542196, + "loss": 3.5116, + "step": 77040 + }, + { + "epoch": 5.234746568827286, + "grad_norm": 0.9381842017173767, + "learning_rate": 0.0003456566788965892, + "loss": 3.4176, + "step": 77045 + }, + { + "epoch": 5.235086288897948, + "grad_norm": 0.8261088728904724, + "learning_rate": 0.00034561421388775647, + "loss": 3.5836, + "step": 77050 + }, + { + "epoch": 5.23542600896861, + "grad_norm": 0.8078585863113403, + "learning_rate": 0.0003455717488789238, + "loss": 3.409, + "step": 77055 + }, + { + "epoch": 5.235765729039271, + "grad_norm": 0.9065710306167603, + "learning_rate": 0.000345529283870091, + "loss": 3.3325, + "step": 77060 + }, + { + "epoch": 5.2361054491099335, + "grad_norm": 1.1779924631118774, + "learning_rate": 0.0003454868188612583, + "loss": 3.5113, + "step": 77065 + }, + { + "epoch": 5.2364451691805955, + "grad_norm": 0.871133029460907, + "learning_rate": 0.00034544435385242564, + "loss": 3.5837, + "step": 77070 + }, + { + "epoch": 5.236784889251257, + "grad_norm": 0.830227255821228, + "learning_rate": 0.00034540188884359287, + "loss": 3.4585, + "step": 77075 + }, + { + "epoch": 5.237124609321919, + "grad_norm": 0.8452599048614502, + "learning_rate": 0.00034535942383476015, + "loss": 3.4631, + "step": 77080 + }, + { + "epoch": 5.237464329392581, + "grad_norm": 1.071105718612671, + "learning_rate": 0.0003453169588259274, + "loss": 3.6018, + "step": 77085 + }, + { + "epoch": 5.237804049463242, + "grad_norm": 1.0948405265808105, + "learning_rate": 0.00034527449381709476, + "loss": 3.444, + "step": 77090 + }, + { + "epoch": 5.238143769533904, + "grad_norm": 0.9461437463760376, + "learning_rate": 0.000345232028808262, + "loss": 3.6905, + "step": 77095 + }, + { + "epoch": 5.238483489604566, + "grad_norm": 0.7600189447402954, + "learning_rate": 0.00034518956379942927, + "loss": 3.543, + "step": 77100 + }, + { + "epoch": 5.238823209675227, + "grad_norm": 0.882175087928772, + "learning_rate": 0.0003451470987905966, + "loss": 3.3845, + "step": 77105 + }, + { + "epoch": 5.2391629297458895, + "grad_norm": 1.045780897140503, + "learning_rate": 0.0003451046337817638, + "loss": 3.4564, + "step": 77110 + }, + { + "epoch": 5.239502649816552, + "grad_norm": 0.9692131876945496, + "learning_rate": 0.0003450621687729311, + "loss": 3.3799, + "step": 77115 + }, + { + "epoch": 5.239842369887213, + "grad_norm": 0.8713414669036865, + "learning_rate": 0.0003450197037640984, + "loss": 3.5173, + "step": 77120 + }, + { + "epoch": 5.240182089957875, + "grad_norm": 0.8626394867897034, + "learning_rate": 0.00034497723875526567, + "loss": 3.3149, + "step": 77125 + }, + { + "epoch": 5.240521810028537, + "grad_norm": 0.9501742124557495, + "learning_rate": 0.00034493477374643295, + "loss": 3.5311, + "step": 77130 + }, + { + "epoch": 5.240861530099198, + "grad_norm": 1.1224428415298462, + "learning_rate": 0.00034489230873760023, + "loss": 3.4602, + "step": 77135 + }, + { + "epoch": 5.24120125016986, + "grad_norm": 0.9698891043663025, + "learning_rate": 0.0003448498437287675, + "loss": 3.6282, + "step": 77140 + }, + { + "epoch": 5.241540970240522, + "grad_norm": 0.8749942779541016, + "learning_rate": 0.0003448073787199348, + "loss": 3.4227, + "step": 77145 + }, + { + "epoch": 5.241880690311183, + "grad_norm": 0.8608420491218567, + "learning_rate": 0.00034476491371110207, + "loss": 3.6252, + "step": 77150 + }, + { + "epoch": 5.2422204103818455, + "grad_norm": 0.7576118111610413, + "learning_rate": 0.0003447224487022693, + "loss": 3.3753, + "step": 77155 + }, + { + "epoch": 5.242560130452507, + "grad_norm": 0.7377164363861084, + "learning_rate": 0.00034467998369343663, + "loss": 3.3654, + "step": 77160 + }, + { + "epoch": 5.242899850523169, + "grad_norm": 0.7708435654640198, + "learning_rate": 0.0003446375186846039, + "loss": 3.5324, + "step": 77165 + }, + { + "epoch": 5.243239570593831, + "grad_norm": 0.8772801160812378, + "learning_rate": 0.00034459505367577113, + "loss": 3.5919, + "step": 77170 + }, + { + "epoch": 5.243579290664492, + "grad_norm": 1.2384616136550903, + "learning_rate": 0.00034455258866693847, + "loss": 3.3026, + "step": 77175 + }, + { + "epoch": 5.243919010735154, + "grad_norm": 1.0649309158325195, + "learning_rate": 0.00034451012365810575, + "loss": 3.5133, + "step": 77180 + }, + { + "epoch": 5.244258730805816, + "grad_norm": 0.8944038152694702, + "learning_rate": 0.000344467658649273, + "loss": 3.4353, + "step": 77185 + }, + { + "epoch": 5.244598450876477, + "grad_norm": 0.8924391865730286, + "learning_rate": 0.0003444251936404403, + "loss": 3.1274, + "step": 77190 + }, + { + "epoch": 5.244938170947139, + "grad_norm": 0.8794419765472412, + "learning_rate": 0.0003443827286316076, + "loss": 3.6789, + "step": 77195 + }, + { + "epoch": 5.2452778910178015, + "grad_norm": 0.7768568992614746, + "learning_rate": 0.0003443402636227748, + "loss": 3.4674, + "step": 77200 + }, + { + "epoch": 5.245617611088463, + "grad_norm": 0.850680947303772, + "learning_rate": 0.0003442977986139421, + "loss": 3.5749, + "step": 77205 + }, + { + "epoch": 5.245957331159125, + "grad_norm": 0.9409108757972717, + "learning_rate": 0.00034425533360510943, + "loss": 3.2891, + "step": 77210 + }, + { + "epoch": 5.246297051229787, + "grad_norm": 0.8805908560752869, + "learning_rate": 0.00034421286859627665, + "loss": 3.3258, + "step": 77215 + }, + { + "epoch": 5.246636771300448, + "grad_norm": 0.9379807710647583, + "learning_rate": 0.00034417040358744393, + "loss": 3.585, + "step": 77220 + }, + { + "epoch": 5.24697649137111, + "grad_norm": 0.8435922861099243, + "learning_rate": 0.00034412793857861127, + "loss": 3.298, + "step": 77225 + }, + { + "epoch": 5.247316211441772, + "grad_norm": 1.1961607933044434, + "learning_rate": 0.0003440854735697785, + "loss": 3.5345, + "step": 77230 + }, + { + "epoch": 5.247655931512433, + "grad_norm": 0.7640591263771057, + "learning_rate": 0.0003440430085609458, + "loss": 3.3496, + "step": 77235 + }, + { + "epoch": 5.2479956515830954, + "grad_norm": 1.0892095565795898, + "learning_rate": 0.00034400054355211305, + "loss": 3.5472, + "step": 77240 + }, + { + "epoch": 5.2483353716537575, + "grad_norm": 0.9508739709854126, + "learning_rate": 0.00034395807854328033, + "loss": 3.5762, + "step": 77245 + }, + { + "epoch": 5.248675091724419, + "grad_norm": 0.9460146427154541, + "learning_rate": 0.0003439156135344476, + "loss": 3.4755, + "step": 77250 + }, + { + "epoch": 5.249014811795081, + "grad_norm": 0.9860759973526001, + "learning_rate": 0.0003438731485256149, + "loss": 3.2882, + "step": 77255 + }, + { + "epoch": 5.249354531865743, + "grad_norm": 0.7265598177909851, + "learning_rate": 0.00034383068351678223, + "loss": 3.4948, + "step": 77260 + }, + { + "epoch": 5.249694251936404, + "grad_norm": 1.2614529132843018, + "learning_rate": 0.00034378821850794946, + "loss": 3.5251, + "step": 77265 + }, + { + "epoch": 5.250033972007066, + "grad_norm": 0.9185158610343933, + "learning_rate": 0.00034374575349911674, + "loss": 3.39, + "step": 77270 + }, + { + "epoch": 5.250373692077728, + "grad_norm": 0.9545618891716003, + "learning_rate": 0.000343703288490284, + "loss": 3.4426, + "step": 77275 + }, + { + "epoch": 5.250713412148389, + "grad_norm": 0.8243758678436279, + "learning_rate": 0.0003436608234814513, + "loss": 3.5945, + "step": 77280 + }, + { + "epoch": 5.2510531322190515, + "grad_norm": 1.0781502723693848, + "learning_rate": 0.0003436183584726186, + "loss": 3.4631, + "step": 77285 + }, + { + "epoch": 5.2513928522897135, + "grad_norm": 0.7528315782546997, + "learning_rate": 0.00034357589346378586, + "loss": 3.4565, + "step": 77290 + }, + { + "epoch": 5.251732572360375, + "grad_norm": 0.7889813184738159, + "learning_rate": 0.00034353342845495314, + "loss": 3.3936, + "step": 77295 + }, + { + "epoch": 5.252072292431037, + "grad_norm": 1.0313286781311035, + "learning_rate": 0.0003434909634461204, + "loss": 3.1429, + "step": 77300 + }, + { + "epoch": 5.252412012501699, + "grad_norm": 0.7810640931129456, + "learning_rate": 0.0003434484984372877, + "loss": 3.4239, + "step": 77305 + }, + { + "epoch": 5.25275173257236, + "grad_norm": 0.9785544276237488, + "learning_rate": 0.0003434060334284549, + "loss": 3.5322, + "step": 77310 + }, + { + "epoch": 5.253091452643022, + "grad_norm": 0.747261106967926, + "learning_rate": 0.00034336356841962226, + "loss": 3.5368, + "step": 77315 + }, + { + "epoch": 5.253431172713684, + "grad_norm": 0.8015008568763733, + "learning_rate": 0.00034332110341078954, + "loss": 3.3648, + "step": 77320 + }, + { + "epoch": 5.253770892784345, + "grad_norm": 0.8469849824905396, + "learning_rate": 0.00034327863840195676, + "loss": 3.6406, + "step": 77325 + }, + { + "epoch": 5.2541106128550075, + "grad_norm": 1.0149059295654297, + "learning_rate": 0.0003432361733931241, + "loss": 3.4733, + "step": 77330 + }, + { + "epoch": 5.2544503329256695, + "grad_norm": 0.8491706252098083, + "learning_rate": 0.0003431937083842914, + "loss": 3.4712, + "step": 77335 + }, + { + "epoch": 5.254790052996331, + "grad_norm": 0.6522251963615417, + "learning_rate": 0.0003431512433754586, + "loss": 3.5576, + "step": 77340 + }, + { + "epoch": 5.255129773066993, + "grad_norm": 0.8650041818618774, + "learning_rate": 0.0003431087783666259, + "loss": 3.617, + "step": 77345 + }, + { + "epoch": 5.255469493137655, + "grad_norm": 1.4981704950332642, + "learning_rate": 0.0003430663133577932, + "loss": 3.5542, + "step": 77350 + }, + { + "epoch": 5.255809213208316, + "grad_norm": 0.9435403347015381, + "learning_rate": 0.00034302384834896044, + "loss": 3.3994, + "step": 77355 + }, + { + "epoch": 5.256148933278978, + "grad_norm": 0.8579195141792297, + "learning_rate": 0.0003429813833401277, + "loss": 3.4151, + "step": 77360 + }, + { + "epoch": 5.25648865334964, + "grad_norm": 0.8084255456924438, + "learning_rate": 0.00034293891833129506, + "loss": 3.5837, + "step": 77365 + }, + { + "epoch": 5.256828373420301, + "grad_norm": 0.7241343855857849, + "learning_rate": 0.0003428964533224623, + "loss": 3.6721, + "step": 77370 + }, + { + "epoch": 5.2571680934909635, + "grad_norm": 0.8361693024635315, + "learning_rate": 0.00034285398831362956, + "loss": 3.4814, + "step": 77375 + }, + { + "epoch": 5.2575078135616256, + "grad_norm": 1.0515402555465698, + "learning_rate": 0.00034281152330479684, + "loss": 3.5293, + "step": 77380 + }, + { + "epoch": 5.257847533632287, + "grad_norm": 0.9794366955757141, + "learning_rate": 0.0003427690582959641, + "loss": 3.3919, + "step": 77385 + }, + { + "epoch": 5.258187253702949, + "grad_norm": 1.2499045133590698, + "learning_rate": 0.0003427265932871314, + "loss": 3.5296, + "step": 77390 + }, + { + "epoch": 5.258526973773611, + "grad_norm": 0.9121119379997253, + "learning_rate": 0.0003426841282782987, + "loss": 3.5849, + "step": 77395 + }, + { + "epoch": 5.258866693844272, + "grad_norm": 0.8153026700019836, + "learning_rate": 0.00034264166326946596, + "loss": 3.285, + "step": 77400 + }, + { + "epoch": 5.259206413914934, + "grad_norm": 0.9129648208618164, + "learning_rate": 0.00034259919826063324, + "loss": 3.527, + "step": 77405 + }, + { + "epoch": 5.259546133985596, + "grad_norm": 0.8096426129341125, + "learning_rate": 0.0003425567332518005, + "loss": 3.5453, + "step": 77410 + }, + { + "epoch": 5.259885854056257, + "grad_norm": 0.9756879806518555, + "learning_rate": 0.00034251426824296775, + "loss": 3.4736, + "step": 77415 + }, + { + "epoch": 5.2602255741269195, + "grad_norm": 1.1648398637771606, + "learning_rate": 0.0003424718032341351, + "loss": 3.711, + "step": 77420 + }, + { + "epoch": 5.260565294197582, + "grad_norm": 1.3503143787384033, + "learning_rate": 0.00034242933822530236, + "loss": 3.2874, + "step": 77425 + }, + { + "epoch": 5.260905014268243, + "grad_norm": 1.1657476425170898, + "learning_rate": 0.00034238687321646964, + "loss": 3.5511, + "step": 77430 + }, + { + "epoch": 5.261244734338905, + "grad_norm": 0.9445148706436157, + "learning_rate": 0.0003423444082076369, + "loss": 3.5345, + "step": 77435 + }, + { + "epoch": 5.261584454409567, + "grad_norm": 0.855963945388794, + "learning_rate": 0.0003423019431988042, + "loss": 3.5521, + "step": 77440 + }, + { + "epoch": 5.261924174480228, + "grad_norm": 0.9673268795013428, + "learning_rate": 0.0003422594781899715, + "loss": 3.3943, + "step": 77445 + }, + { + "epoch": 5.26226389455089, + "grad_norm": 1.0579267740249634, + "learning_rate": 0.0003422170131811387, + "loss": 3.4273, + "step": 77450 + }, + { + "epoch": 5.262603614621552, + "grad_norm": 1.1987358331680298, + "learning_rate": 0.00034217454817230604, + "loss": 3.5302, + "step": 77455 + }, + { + "epoch": 5.262943334692213, + "grad_norm": 0.9026594758033752, + "learning_rate": 0.0003421320831634733, + "loss": 3.5334, + "step": 77460 + }, + { + "epoch": 5.2632830547628755, + "grad_norm": 0.7517818808555603, + "learning_rate": 0.00034208961815464055, + "loss": 3.6289, + "step": 77465 + }, + { + "epoch": 5.263622774833538, + "grad_norm": 0.956368625164032, + "learning_rate": 0.0003420471531458079, + "loss": 3.3108, + "step": 77470 + }, + { + "epoch": 5.263962494904199, + "grad_norm": 1.05348801612854, + "learning_rate": 0.00034200468813697516, + "loss": 3.7044, + "step": 77475 + }, + { + "epoch": 5.264302214974861, + "grad_norm": 1.2108911275863647, + "learning_rate": 0.0003419622231281424, + "loss": 3.5544, + "step": 77480 + }, + { + "epoch": 5.264641935045523, + "grad_norm": 0.7418484091758728, + "learning_rate": 0.0003419197581193097, + "loss": 3.3635, + "step": 77485 + }, + { + "epoch": 5.264981655116184, + "grad_norm": 0.8016573190689087, + "learning_rate": 0.000341877293110477, + "loss": 3.3417, + "step": 77490 + }, + { + "epoch": 5.265321375186846, + "grad_norm": 1.0444839000701904, + "learning_rate": 0.00034183482810164423, + "loss": 3.398, + "step": 77495 + }, + { + "epoch": 5.265661095257508, + "grad_norm": 0.8531420826911926, + "learning_rate": 0.0003417923630928115, + "loss": 3.5893, + "step": 77500 + }, + { + "epoch": 5.266000815328169, + "grad_norm": 0.9030084609985352, + "learning_rate": 0.00034174989808397884, + "loss": 3.4608, + "step": 77505 + }, + { + "epoch": 5.2663405353988315, + "grad_norm": 0.8647962808609009, + "learning_rate": 0.00034170743307514607, + "loss": 3.4272, + "step": 77510 + }, + { + "epoch": 5.266680255469494, + "grad_norm": 0.7373958826065063, + "learning_rate": 0.00034166496806631335, + "loss": 3.5256, + "step": 77515 + }, + { + "epoch": 5.267019975540155, + "grad_norm": 1.1547341346740723, + "learning_rate": 0.0003416225030574807, + "loss": 3.5097, + "step": 77520 + }, + { + "epoch": 5.267359695610817, + "grad_norm": 0.7432810664176941, + "learning_rate": 0.0003415800380486479, + "loss": 3.4815, + "step": 77525 + }, + { + "epoch": 5.267699415681479, + "grad_norm": 0.9177041053771973, + "learning_rate": 0.0003415375730398152, + "loss": 3.3127, + "step": 77530 + }, + { + "epoch": 5.26803913575214, + "grad_norm": 1.1121187210083008, + "learning_rate": 0.00034149510803098247, + "loss": 3.431, + "step": 77535 + }, + { + "epoch": 5.268378855822802, + "grad_norm": 0.9663049578666687, + "learning_rate": 0.00034145264302214975, + "loss": 3.4525, + "step": 77540 + }, + { + "epoch": 5.268718575893463, + "grad_norm": 0.7763304710388184, + "learning_rate": 0.00034141017801331703, + "loss": 3.5386, + "step": 77545 + }, + { + "epoch": 5.2690582959641254, + "grad_norm": 0.8114166855812073, + "learning_rate": 0.0003413677130044843, + "loss": 3.4439, + "step": 77550 + }, + { + "epoch": 5.2693980160347875, + "grad_norm": 0.8475888967514038, + "learning_rate": 0.0003413252479956516, + "loss": 3.3476, + "step": 77555 + }, + { + "epoch": 5.269737736105449, + "grad_norm": 0.9251770377159119, + "learning_rate": 0.00034128278298681887, + "loss": 3.4258, + "step": 77560 + }, + { + "epoch": 5.270077456176111, + "grad_norm": 0.9561187028884888, + "learning_rate": 0.00034124031797798615, + "loss": 3.6615, + "step": 77565 + }, + { + "epoch": 5.270417176246773, + "grad_norm": 0.8605806231498718, + "learning_rate": 0.0003411978529691534, + "loss": 3.5582, + "step": 77570 + }, + { + "epoch": 5.270756896317434, + "grad_norm": 0.8247023820877075, + "learning_rate": 0.0003411553879603207, + "loss": 3.7738, + "step": 77575 + }, + { + "epoch": 5.271096616388096, + "grad_norm": 1.1633988618850708, + "learning_rate": 0.000341112922951488, + "loss": 3.4195, + "step": 77580 + }, + { + "epoch": 5.271436336458758, + "grad_norm": 0.9595105648040771, + "learning_rate": 0.0003410704579426552, + "loss": 3.3964, + "step": 77585 + }, + { + "epoch": 5.271776056529419, + "grad_norm": 0.8360251188278198, + "learning_rate": 0.00034102799293382255, + "loss": 3.2223, + "step": 77590 + }, + { + "epoch": 5.2721157766000815, + "grad_norm": 0.8594487309455872, + "learning_rate": 0.00034098552792498983, + "loss": 3.5877, + "step": 77595 + }, + { + "epoch": 5.2724554966707435, + "grad_norm": 1.0351351499557495, + "learning_rate": 0.0003409430629161571, + "loss": 3.5165, + "step": 77600 + }, + { + "epoch": 5.272795216741405, + "grad_norm": 0.8016113638877869, + "learning_rate": 0.00034090059790732434, + "loss": 3.3266, + "step": 77605 + }, + { + "epoch": 5.273134936812067, + "grad_norm": 1.0015082359313965, + "learning_rate": 0.00034085813289849167, + "loss": 3.3806, + "step": 77610 + }, + { + "epoch": 5.273474656882729, + "grad_norm": 0.9104819893836975, + "learning_rate": 0.00034081566788965895, + "loss": 3.4976, + "step": 77615 + }, + { + "epoch": 5.27381437695339, + "grad_norm": 0.9334174990653992, + "learning_rate": 0.0003407732028808262, + "loss": 3.5475, + "step": 77620 + }, + { + "epoch": 5.274154097024052, + "grad_norm": 1.066830039024353, + "learning_rate": 0.0003407307378719935, + "loss": 3.3643, + "step": 77625 + }, + { + "epoch": 5.274493817094714, + "grad_norm": 0.885061502456665, + "learning_rate": 0.0003406882728631608, + "loss": 3.5078, + "step": 77630 + }, + { + "epoch": 5.274833537165375, + "grad_norm": 0.8743835091590881, + "learning_rate": 0.000340645807854328, + "loss": 3.6292, + "step": 77635 + }, + { + "epoch": 5.2751732572360375, + "grad_norm": 0.8150795698165894, + "learning_rate": 0.0003406033428454953, + "loss": 3.5228, + "step": 77640 + }, + { + "epoch": 5.2755129773066995, + "grad_norm": 0.865960955619812, + "learning_rate": 0.00034056087783666263, + "loss": 3.3299, + "step": 77645 + }, + { + "epoch": 5.275852697377361, + "grad_norm": 1.1821253299713135, + "learning_rate": 0.00034051841282782986, + "loss": 3.3974, + "step": 77650 + }, + { + "epoch": 5.276192417448023, + "grad_norm": 0.8339661955833435, + "learning_rate": 0.00034047594781899714, + "loss": 3.4177, + "step": 77655 + }, + { + "epoch": 5.276532137518685, + "grad_norm": 0.9023543000221252, + "learning_rate": 0.00034043348281016447, + "loss": 3.5395, + "step": 77660 + }, + { + "epoch": 5.276871857589346, + "grad_norm": 0.9916113615036011, + "learning_rate": 0.0003403910178013317, + "loss": 3.3426, + "step": 77665 + }, + { + "epoch": 5.277211577660008, + "grad_norm": 0.8041478395462036, + "learning_rate": 0.000340348552792499, + "loss": 3.579, + "step": 77670 + }, + { + "epoch": 5.27755129773067, + "grad_norm": 0.9110507965087891, + "learning_rate": 0.00034030608778366626, + "loss": 3.2674, + "step": 77675 + }, + { + "epoch": 5.277891017801331, + "grad_norm": 1.0961023569107056, + "learning_rate": 0.00034026362277483354, + "loss": 3.1566, + "step": 77680 + }, + { + "epoch": 5.2782307378719935, + "grad_norm": 0.9122074246406555, + "learning_rate": 0.0003402211577660008, + "loss": 3.5073, + "step": 77685 + }, + { + "epoch": 5.278570457942656, + "grad_norm": 0.9889321327209473, + "learning_rate": 0.0003401786927571681, + "loss": 3.3913, + "step": 77690 + }, + { + "epoch": 5.278910178013317, + "grad_norm": 1.0775107145309448, + "learning_rate": 0.0003401362277483354, + "loss": 3.6616, + "step": 77695 + }, + { + "epoch": 5.279249898083979, + "grad_norm": 0.680305004119873, + "learning_rate": 0.00034009376273950266, + "loss": 3.2457, + "step": 77700 + }, + { + "epoch": 5.279589618154641, + "grad_norm": 0.8493651151657104, + "learning_rate": 0.00034005129773066994, + "loss": 3.73, + "step": 77705 + }, + { + "epoch": 5.279929338225302, + "grad_norm": 0.8255748748779297, + "learning_rate": 0.00034000883272183717, + "loss": 3.2006, + "step": 77710 + }, + { + "epoch": 5.280269058295964, + "grad_norm": 0.8718971014022827, + "learning_rate": 0.0003399663677130045, + "loss": 3.3868, + "step": 77715 + }, + { + "epoch": 5.280608778366626, + "grad_norm": 0.9038297533988953, + "learning_rate": 0.0003399239027041718, + "loss": 3.4532, + "step": 77720 + }, + { + "epoch": 5.280948498437287, + "grad_norm": 1.0276360511779785, + "learning_rate": 0.000339881437695339, + "loss": 3.5573, + "step": 77725 + }, + { + "epoch": 5.2812882185079495, + "grad_norm": 0.9996660351753235, + "learning_rate": 0.00033983897268650634, + "loss": 3.3857, + "step": 77730 + }, + { + "epoch": 5.281627938578612, + "grad_norm": 0.9069861173629761, + "learning_rate": 0.0003397965076776736, + "loss": 3.5874, + "step": 77735 + }, + { + "epoch": 5.281967658649273, + "grad_norm": 0.9187392592430115, + "learning_rate": 0.00033975404266884085, + "loss": 3.3536, + "step": 77740 + }, + { + "epoch": 5.282307378719935, + "grad_norm": 0.8645111322402954, + "learning_rate": 0.0003397115776600082, + "loss": 3.4406, + "step": 77745 + }, + { + "epoch": 5.282647098790597, + "grad_norm": 0.8234793543815613, + "learning_rate": 0.00033966911265117546, + "loss": 3.2381, + "step": 77750 + }, + { + "epoch": 5.282986818861258, + "grad_norm": 0.8133503794670105, + "learning_rate": 0.0003396266476423427, + "loss": 3.5607, + "step": 77755 + }, + { + "epoch": 5.28332653893192, + "grad_norm": 1.179540753364563, + "learning_rate": 0.00033958418263350997, + "loss": 3.5443, + "step": 77760 + }, + { + "epoch": 5.283666259002582, + "grad_norm": 1.2166885137557983, + "learning_rate": 0.0003395417176246773, + "loss": 3.045, + "step": 77765 + }, + { + "epoch": 5.284005979073243, + "grad_norm": 0.8534504175186157, + "learning_rate": 0.0003394992526158446, + "loss": 3.4228, + "step": 77770 + }, + { + "epoch": 5.2843456991439055, + "grad_norm": 0.8876750469207764, + "learning_rate": 0.0003394567876070118, + "loss": 3.5048, + "step": 77775 + }, + { + "epoch": 5.284685419214568, + "grad_norm": 0.9431858062744141, + "learning_rate": 0.00033941432259817914, + "loss": 3.2948, + "step": 77780 + }, + { + "epoch": 5.285025139285229, + "grad_norm": 0.880521833896637, + "learning_rate": 0.0003393718575893464, + "loss": 3.4363, + "step": 77785 + }, + { + "epoch": 5.285364859355891, + "grad_norm": 1.0114389657974243, + "learning_rate": 0.00033932939258051365, + "loss": 3.5478, + "step": 77790 + }, + { + "epoch": 5.285704579426553, + "grad_norm": 1.0734858512878418, + "learning_rate": 0.0003392869275716809, + "loss": 3.3463, + "step": 77795 + }, + { + "epoch": 5.286044299497214, + "grad_norm": 1.027349829673767, + "learning_rate": 0.00033924446256284826, + "loss": 3.3553, + "step": 77800 + }, + { + "epoch": 5.286384019567876, + "grad_norm": 0.8302164077758789, + "learning_rate": 0.0003392019975540155, + "loss": 3.5204, + "step": 77805 + }, + { + "epoch": 5.286723739638538, + "grad_norm": 0.8433977365493774, + "learning_rate": 0.00033915953254518277, + "loss": 3.5084, + "step": 77810 + }, + { + "epoch": 5.287063459709199, + "grad_norm": 1.0863436460494995, + "learning_rate": 0.0003391170675363501, + "loss": 3.3816, + "step": 77815 + }, + { + "epoch": 5.2874031797798615, + "grad_norm": 0.9788549542427063, + "learning_rate": 0.0003390746025275173, + "loss": 3.4923, + "step": 77820 + }, + { + "epoch": 5.287742899850523, + "grad_norm": 0.8325121402740479, + "learning_rate": 0.0003390321375186846, + "loss": 3.5172, + "step": 77825 + }, + { + "epoch": 5.288082619921185, + "grad_norm": 0.7759940028190613, + "learning_rate": 0.0003389896725098519, + "loss": 3.2122, + "step": 77830 + }, + { + "epoch": 5.288422339991847, + "grad_norm": 0.9254586696624756, + "learning_rate": 0.00033894720750101917, + "loss": 3.504, + "step": 77835 + }, + { + "epoch": 5.288762060062508, + "grad_norm": 0.800954282283783, + "learning_rate": 0.00033890474249218645, + "loss": 3.7182, + "step": 77840 + }, + { + "epoch": 5.28910178013317, + "grad_norm": 1.0384767055511475, + "learning_rate": 0.00033886227748335373, + "loss": 3.4101, + "step": 77845 + }, + { + "epoch": 5.289441500203832, + "grad_norm": 0.8862746357917786, + "learning_rate": 0.000338819812474521, + "loss": 3.3924, + "step": 77850 + }, + { + "epoch": 5.289781220274493, + "grad_norm": 1.1366990804672241, + "learning_rate": 0.0003387773474656883, + "loss": 3.5512, + "step": 77855 + }, + { + "epoch": 5.2901209403451555, + "grad_norm": 0.8754532337188721, + "learning_rate": 0.00033873488245685557, + "loss": 3.5402, + "step": 77860 + }, + { + "epoch": 5.2904606604158175, + "grad_norm": 0.9401914477348328, + "learning_rate": 0.0003386924174480228, + "loss": 3.2838, + "step": 77865 + }, + { + "epoch": 5.290800380486479, + "grad_norm": 0.6615973711013794, + "learning_rate": 0.00033864995243919013, + "loss": 3.3303, + "step": 77870 + }, + { + "epoch": 5.291140100557141, + "grad_norm": 0.9023000001907349, + "learning_rate": 0.0003386074874303574, + "loss": 3.6885, + "step": 77875 + }, + { + "epoch": 5.291479820627803, + "grad_norm": 1.2444915771484375, + "learning_rate": 0.00033856502242152463, + "loss": 3.2479, + "step": 77880 + }, + { + "epoch": 5.291819540698464, + "grad_norm": 0.8478114604949951, + "learning_rate": 0.00033852255741269197, + "loss": 3.3692, + "step": 77885 + }, + { + "epoch": 5.292159260769126, + "grad_norm": 0.8026512861251831, + "learning_rate": 0.00033848009240385925, + "loss": 3.5581, + "step": 77890 + }, + { + "epoch": 5.292498980839788, + "grad_norm": 1.0923928022384644, + "learning_rate": 0.0003384376273950265, + "loss": 3.3897, + "step": 77895 + }, + { + "epoch": 5.292838700910449, + "grad_norm": 0.9207879900932312, + "learning_rate": 0.00033839516238619375, + "loss": 3.4057, + "step": 77900 + }, + { + "epoch": 5.2931784209811115, + "grad_norm": 1.3061646223068237, + "learning_rate": 0.0003383526973773611, + "loss": 3.3043, + "step": 77905 + }, + { + "epoch": 5.2935181410517735, + "grad_norm": 1.0283002853393555, + "learning_rate": 0.0003383102323685283, + "loss": 3.4014, + "step": 77910 + }, + { + "epoch": 5.293857861122435, + "grad_norm": 0.8849559426307678, + "learning_rate": 0.0003382677673596956, + "loss": 3.4991, + "step": 77915 + }, + { + "epoch": 5.294197581193097, + "grad_norm": 1.2612664699554443, + "learning_rate": 0.00033822530235086293, + "loss": 3.3864, + "step": 77920 + }, + { + "epoch": 5.294537301263759, + "grad_norm": 0.8992102742195129, + "learning_rate": 0.00033818283734203015, + "loss": 3.2981, + "step": 77925 + }, + { + "epoch": 5.29487702133442, + "grad_norm": 0.9072105884552002, + "learning_rate": 0.00033814037233319743, + "loss": 3.4569, + "step": 77930 + }, + { + "epoch": 5.295216741405082, + "grad_norm": 0.7484358549118042, + "learning_rate": 0.0003380979073243647, + "loss": 3.6891, + "step": 77935 + }, + { + "epoch": 5.295556461475744, + "grad_norm": 1.1159650087356567, + "learning_rate": 0.00033805544231553205, + "loss": 3.3747, + "step": 77940 + }, + { + "epoch": 5.295896181546405, + "grad_norm": 1.0430089235305786, + "learning_rate": 0.0003380129773066993, + "loss": 3.5658, + "step": 77945 + }, + { + "epoch": 5.2962359016170675, + "grad_norm": 0.810356855392456, + "learning_rate": 0.00033797051229786655, + "loss": 3.6344, + "step": 77950 + }, + { + "epoch": 5.2965756216877296, + "grad_norm": 0.8487858176231384, + "learning_rate": 0.0003379280472890339, + "loss": 3.5575, + "step": 77955 + }, + { + "epoch": 5.296915341758391, + "grad_norm": 0.8651509881019592, + "learning_rate": 0.0003378855822802011, + "loss": 3.3303, + "step": 77960 + }, + { + "epoch": 5.297255061829053, + "grad_norm": 1.00436270236969, + "learning_rate": 0.0003378431172713684, + "loss": 3.44, + "step": 77965 + }, + { + "epoch": 5.297594781899715, + "grad_norm": 0.833680272102356, + "learning_rate": 0.0003378006522625357, + "loss": 3.4714, + "step": 77970 + }, + { + "epoch": 5.297934501970376, + "grad_norm": 1.0223634243011475, + "learning_rate": 0.00033775818725370295, + "loss": 3.3929, + "step": 77975 + }, + { + "epoch": 5.298274222041038, + "grad_norm": 0.8521704077720642, + "learning_rate": 0.00033771572224487024, + "loss": 3.736, + "step": 77980 + }, + { + "epoch": 5.2986139421117, + "grad_norm": 0.8357382416725159, + "learning_rate": 0.0003376732572360375, + "loss": 3.4238, + "step": 77985 + }, + { + "epoch": 5.298953662182361, + "grad_norm": 0.8170688152313232, + "learning_rate": 0.0003376307922272048, + "loss": 3.484, + "step": 77990 + }, + { + "epoch": 5.2992933822530235, + "grad_norm": 1.4321448802947998, + "learning_rate": 0.0003375883272183721, + "loss": 3.325, + "step": 77995 + }, + { + "epoch": 5.299633102323686, + "grad_norm": 0.771216094493866, + "learning_rate": 0.00033754586220953936, + "loss": 3.3302, + "step": 78000 + }, + { + "epoch": 5.299972822394347, + "grad_norm": 1.1680614948272705, + "learning_rate": 0.0003375033972007066, + "loss": 3.4355, + "step": 78005 + }, + { + "epoch": 5.300312542465009, + "grad_norm": 0.9365093111991882, + "learning_rate": 0.0003374609321918739, + "loss": 3.1969, + "step": 78010 + }, + { + "epoch": 5.300652262535671, + "grad_norm": 0.9794655442237854, + "learning_rate": 0.0003374184671830412, + "loss": 3.4951, + "step": 78015 + }, + { + "epoch": 5.300991982606332, + "grad_norm": 0.8563889861106873, + "learning_rate": 0.0003373760021742084, + "loss": 3.3351, + "step": 78020 + }, + { + "epoch": 5.301331702676994, + "grad_norm": 1.0258111953735352, + "learning_rate": 0.00033733353716537576, + "loss": 3.4315, + "step": 78025 + }, + { + "epoch": 5.301671422747656, + "grad_norm": 0.9076101183891296, + "learning_rate": 0.00033729107215654304, + "loss": 3.3646, + "step": 78030 + }, + { + "epoch": 5.302011142818317, + "grad_norm": 1.192469835281372, + "learning_rate": 0.00033724860714771026, + "loss": 3.5457, + "step": 78035 + }, + { + "epoch": 5.3023508628889795, + "grad_norm": 0.8477970957756042, + "learning_rate": 0.0003372061421388776, + "loss": 3.4808, + "step": 78040 + }, + { + "epoch": 5.302690582959642, + "grad_norm": 0.7046147584915161, + "learning_rate": 0.0003371636771300449, + "loss": 3.5354, + "step": 78045 + }, + { + "epoch": 5.303030303030303, + "grad_norm": 1.4049218893051147, + "learning_rate": 0.0003371212121212121, + "loss": 3.2365, + "step": 78050 + }, + { + "epoch": 5.303370023100965, + "grad_norm": 1.1192388534545898, + "learning_rate": 0.0003370787471123794, + "loss": 3.5744, + "step": 78055 + }, + { + "epoch": 5.303709743171627, + "grad_norm": 0.947353720664978, + "learning_rate": 0.0003370362821035467, + "loss": 3.4303, + "step": 78060 + }, + { + "epoch": 5.304049463242288, + "grad_norm": 0.9471931457519531, + "learning_rate": 0.00033699381709471394, + "loss": 3.6548, + "step": 78065 + }, + { + "epoch": 5.30438918331295, + "grad_norm": 0.9532833099365234, + "learning_rate": 0.0003369513520858812, + "loss": 3.3835, + "step": 78070 + }, + { + "epoch": 5.304728903383612, + "grad_norm": 1.1301982402801514, + "learning_rate": 0.00033690888707704856, + "loss": 3.2814, + "step": 78075 + }, + { + "epoch": 5.305068623454273, + "grad_norm": 0.9535732865333557, + "learning_rate": 0.0003368664220682158, + "loss": 3.5504, + "step": 78080 + }, + { + "epoch": 5.3054083435249355, + "grad_norm": 0.9016448259353638, + "learning_rate": 0.00033682395705938306, + "loss": 3.3218, + "step": 78085 + }, + { + "epoch": 5.305748063595598, + "grad_norm": 1.0480573177337646, + "learning_rate": 0.00033678149205055034, + "loss": 3.457, + "step": 78090 + }, + { + "epoch": 5.306087783666259, + "grad_norm": 0.8733727335929871, + "learning_rate": 0.0003367390270417176, + "loss": 3.5374, + "step": 78095 + }, + { + "epoch": 5.306427503736921, + "grad_norm": 0.954719603061676, + "learning_rate": 0.0003366965620328849, + "loss": 3.2684, + "step": 78100 + }, + { + "epoch": 5.306767223807583, + "grad_norm": 0.7705663442611694, + "learning_rate": 0.0003366540970240522, + "loss": 3.6258, + "step": 78105 + }, + { + "epoch": 5.307106943878244, + "grad_norm": 1.1256740093231201, + "learning_rate": 0.0003366116320152195, + "loss": 3.8243, + "step": 78110 + }, + { + "epoch": 5.307446663948906, + "grad_norm": 0.725265622138977, + "learning_rate": 0.00033656916700638674, + "loss": 3.6238, + "step": 78115 + }, + { + "epoch": 5.307786384019568, + "grad_norm": 0.683428168296814, + "learning_rate": 0.000336526701997554, + "loss": 3.3564, + "step": 78120 + }, + { + "epoch": 5.3081261040902294, + "grad_norm": 0.743855357170105, + "learning_rate": 0.0003364842369887213, + "loss": 3.2717, + "step": 78125 + }, + { + "epoch": 5.3084658241608915, + "grad_norm": 0.9205508232116699, + "learning_rate": 0.0003364417719798886, + "loss": 3.6135, + "step": 78130 + }, + { + "epoch": 5.308805544231554, + "grad_norm": 0.709912896156311, + "learning_rate": 0.00033639930697105586, + "loss": 3.243, + "step": 78135 + }, + { + "epoch": 5.309145264302215, + "grad_norm": 0.7392258048057556, + "learning_rate": 0.00033635684196222314, + "loss": 3.3344, + "step": 78140 + }, + { + "epoch": 5.309484984372877, + "grad_norm": 1.0169382095336914, + "learning_rate": 0.0003363143769533904, + "loss": 3.2144, + "step": 78145 + }, + { + "epoch": 5.309824704443539, + "grad_norm": 0.8694301247596741, + "learning_rate": 0.0003362719119445577, + "loss": 3.5338, + "step": 78150 + }, + { + "epoch": 5.3101644245142, + "grad_norm": 0.8137326240539551, + "learning_rate": 0.000336229446935725, + "loss": 3.5901, + "step": 78155 + }, + { + "epoch": 5.310504144584862, + "grad_norm": 0.8788939118385315, + "learning_rate": 0.0003361869819268922, + "loss": 3.5348, + "step": 78160 + }, + { + "epoch": 5.310843864655524, + "grad_norm": 0.7957729697227478, + "learning_rate": 0.00033614451691805954, + "loss": 3.8183, + "step": 78165 + }, + { + "epoch": 5.3111835847261855, + "grad_norm": 0.787786066532135, + "learning_rate": 0.0003361020519092268, + "loss": 3.4331, + "step": 78170 + }, + { + "epoch": 5.3115233047968475, + "grad_norm": 0.986140251159668, + "learning_rate": 0.00033605958690039405, + "loss": 3.65, + "step": 78175 + }, + { + "epoch": 5.31186302486751, + "grad_norm": 1.3899834156036377, + "learning_rate": 0.0003360171218915614, + "loss": 3.8465, + "step": 78180 + }, + { + "epoch": 5.312202744938171, + "grad_norm": 0.7742891907691956, + "learning_rate": 0.00033597465688272866, + "loss": 3.0846, + "step": 78185 + }, + { + "epoch": 5.312542465008833, + "grad_norm": 0.8650239706039429, + "learning_rate": 0.0003359321918738959, + "loss": 3.2248, + "step": 78190 + }, + { + "epoch": 5.312882185079495, + "grad_norm": 0.858525812625885, + "learning_rate": 0.00033588972686506317, + "loss": 3.4152, + "step": 78195 + }, + { + "epoch": 5.313221905150156, + "grad_norm": 0.8648572564125061, + "learning_rate": 0.0003358472618562305, + "loss": 3.5071, + "step": 78200 + }, + { + "epoch": 5.313561625220818, + "grad_norm": 0.9774354696273804, + "learning_rate": 0.00033580479684739773, + "loss": 3.7684, + "step": 78205 + }, + { + "epoch": 5.31390134529148, + "grad_norm": 0.8625220656394958, + "learning_rate": 0.000335762331838565, + "loss": 3.5771, + "step": 78210 + }, + { + "epoch": 5.3142410653621415, + "grad_norm": 1.0874325037002563, + "learning_rate": 0.00033571986682973234, + "loss": 3.1582, + "step": 78215 + }, + { + "epoch": 5.3145807854328035, + "grad_norm": 0.7907547354698181, + "learning_rate": 0.00033567740182089957, + "loss": 3.4979, + "step": 78220 + }, + { + "epoch": 5.314920505503465, + "grad_norm": 0.8776171803474426, + "learning_rate": 0.00033563493681206685, + "loss": 3.6613, + "step": 78225 + }, + { + "epoch": 5.315260225574127, + "grad_norm": 1.592774748802185, + "learning_rate": 0.00033559247180323413, + "loss": 3.426, + "step": 78230 + }, + { + "epoch": 5.315599945644789, + "grad_norm": 0.7880361080169678, + "learning_rate": 0.0003355500067944014, + "loss": 3.2163, + "step": 78235 + }, + { + "epoch": 5.31593966571545, + "grad_norm": 0.7951199412345886, + "learning_rate": 0.0003355075417855687, + "loss": 3.5935, + "step": 78240 + }, + { + "epoch": 5.316279385786112, + "grad_norm": 0.9114543795585632, + "learning_rate": 0.00033546507677673597, + "loss": 3.4764, + "step": 78245 + }, + { + "epoch": 5.316619105856774, + "grad_norm": 1.035859227180481, + "learning_rate": 0.00033542261176790325, + "loss": 3.3884, + "step": 78250 + }, + { + "epoch": 5.316958825927435, + "grad_norm": 1.095029354095459, + "learning_rate": 0.00033538014675907053, + "loss": 3.6163, + "step": 78255 + }, + { + "epoch": 5.3172985459980975, + "grad_norm": 0.7649465203285217, + "learning_rate": 0.0003353376817502378, + "loss": 3.5173, + "step": 78260 + }, + { + "epoch": 5.31763826606876, + "grad_norm": 0.8430017232894897, + "learning_rate": 0.00033529521674140504, + "loss": 3.2459, + "step": 78265 + }, + { + "epoch": 5.317977986139421, + "grad_norm": 0.8822588920593262, + "learning_rate": 0.00033525275173257237, + "loss": 3.5022, + "step": 78270 + }, + { + "epoch": 5.318317706210083, + "grad_norm": 0.9492565989494324, + "learning_rate": 0.00033521028672373965, + "loss": 3.4351, + "step": 78275 + }, + { + "epoch": 5.318657426280745, + "grad_norm": 0.870532751083374, + "learning_rate": 0.00033516782171490693, + "loss": 3.1444, + "step": 78280 + }, + { + "epoch": 5.318997146351406, + "grad_norm": 1.094956636428833, + "learning_rate": 0.0003351253567060742, + "loss": 3.379, + "step": 78285 + }, + { + "epoch": 5.319336866422068, + "grad_norm": 0.7777120471000671, + "learning_rate": 0.0003350828916972415, + "loss": 3.4448, + "step": 78290 + }, + { + "epoch": 5.31967658649273, + "grad_norm": 1.013400912284851, + "learning_rate": 0.00033504042668840877, + "loss": 3.495, + "step": 78295 + }, + { + "epoch": 5.320016306563391, + "grad_norm": 0.839445173740387, + "learning_rate": 0.000334997961679576, + "loss": 3.2495, + "step": 78300 + }, + { + "epoch": 5.3203560266340535, + "grad_norm": 0.8748689889907837, + "learning_rate": 0.00033495549667074333, + "loss": 3.4332, + "step": 78305 + }, + { + "epoch": 5.320695746704716, + "grad_norm": 1.0204155445098877, + "learning_rate": 0.0003349130316619106, + "loss": 3.3113, + "step": 78310 + }, + { + "epoch": 5.321035466775377, + "grad_norm": 0.9042012691497803, + "learning_rate": 0.00033487056665307784, + "loss": 3.6321, + "step": 78315 + }, + { + "epoch": 5.321375186846039, + "grad_norm": 0.9822362065315247, + "learning_rate": 0.00033482810164424517, + "loss": 3.301, + "step": 78320 + }, + { + "epoch": 5.321714906916701, + "grad_norm": 0.8440858125686646, + "learning_rate": 0.00033478563663541245, + "loss": 3.3994, + "step": 78325 + }, + { + "epoch": 5.322054626987362, + "grad_norm": 0.8402474522590637, + "learning_rate": 0.0003347431716265797, + "loss": 3.2968, + "step": 78330 + }, + { + "epoch": 5.322394347058024, + "grad_norm": 0.8311792016029358, + "learning_rate": 0.000334700706617747, + "loss": 3.4533, + "step": 78335 + }, + { + "epoch": 5.322734067128686, + "grad_norm": 0.8462395071983337, + "learning_rate": 0.0003346582416089143, + "loss": 3.5706, + "step": 78340 + }, + { + "epoch": 5.323073787199347, + "grad_norm": 0.9269742965698242, + "learning_rate": 0.0003346157766000815, + "loss": 3.5576, + "step": 78345 + }, + { + "epoch": 5.3234135072700095, + "grad_norm": 0.7809523344039917, + "learning_rate": 0.0003345733115912488, + "loss": 3.3557, + "step": 78350 + }, + { + "epoch": 5.323753227340672, + "grad_norm": 1.105766773223877, + "learning_rate": 0.00033453084658241613, + "loss": 3.5614, + "step": 78355 + }, + { + "epoch": 5.324092947411333, + "grad_norm": 0.8884000182151794, + "learning_rate": 0.00033448838157358336, + "loss": 3.5279, + "step": 78360 + }, + { + "epoch": 5.324432667481995, + "grad_norm": 0.9326366186141968, + "learning_rate": 0.00033444591656475064, + "loss": 3.3706, + "step": 78365 + }, + { + "epoch": 5.324772387552657, + "grad_norm": 0.8913698792457581, + "learning_rate": 0.00033440345155591797, + "loss": 3.4211, + "step": 78370 + }, + { + "epoch": 5.325112107623318, + "grad_norm": 0.9085662961006165, + "learning_rate": 0.0003343609865470852, + "loss": 3.3774, + "step": 78375 + }, + { + "epoch": 5.32545182769398, + "grad_norm": 0.8610137701034546, + "learning_rate": 0.0003343185215382525, + "loss": 3.6278, + "step": 78380 + }, + { + "epoch": 5.325791547764642, + "grad_norm": 0.9544657468795776, + "learning_rate": 0.00033427605652941976, + "loss": 3.6404, + "step": 78385 + }, + { + "epoch": 5.326131267835303, + "grad_norm": 0.8266671299934387, + "learning_rate": 0.00033423359152058704, + "loss": 3.3385, + "step": 78390 + }, + { + "epoch": 5.3264709879059655, + "grad_norm": 1.0041829347610474, + "learning_rate": 0.0003341911265117543, + "loss": 3.483, + "step": 78395 + }, + { + "epoch": 5.326810707976628, + "grad_norm": 1.0303107500076294, + "learning_rate": 0.0003341486615029216, + "loss": 3.3032, + "step": 78400 + }, + { + "epoch": 5.327150428047289, + "grad_norm": 0.8947954177856445, + "learning_rate": 0.0003341061964940889, + "loss": 3.4383, + "step": 78405 + }, + { + "epoch": 5.327490148117951, + "grad_norm": 0.7742104530334473, + "learning_rate": 0.00033406373148525616, + "loss": 3.4288, + "step": 78410 + }, + { + "epoch": 5.327829868188613, + "grad_norm": 0.9854509830474854, + "learning_rate": 0.00033402126647642344, + "loss": 3.4261, + "step": 78415 + }, + { + "epoch": 5.328169588259274, + "grad_norm": 1.046756625175476, + "learning_rate": 0.00033397880146759067, + "loss": 3.1523, + "step": 78420 + }, + { + "epoch": 5.328509308329936, + "grad_norm": 0.8694136738777161, + "learning_rate": 0.000333936336458758, + "loss": 3.3615, + "step": 78425 + }, + { + "epoch": 5.328849028400598, + "grad_norm": 0.7377488613128662, + "learning_rate": 0.0003338938714499253, + "loss": 3.3582, + "step": 78430 + }, + { + "epoch": 5.3291887484712595, + "grad_norm": 0.9118674993515015, + "learning_rate": 0.0003338514064410925, + "loss": 3.3059, + "step": 78435 + }, + { + "epoch": 5.3295284685419215, + "grad_norm": 1.1664576530456543, + "learning_rate": 0.00033380894143225984, + "loss": 3.617, + "step": 78440 + }, + { + "epoch": 5.329868188612584, + "grad_norm": 0.8813289403915405, + "learning_rate": 0.0003337664764234271, + "loss": 3.3742, + "step": 78445 + }, + { + "epoch": 5.330207908683245, + "grad_norm": 1.0263441801071167, + "learning_rate": 0.0003337240114145944, + "loss": 3.5435, + "step": 78450 + }, + { + "epoch": 5.330547628753907, + "grad_norm": 0.9200456738471985, + "learning_rate": 0.0003336815464057616, + "loss": 3.5949, + "step": 78455 + }, + { + "epoch": 5.330887348824569, + "grad_norm": 0.8677195906639099, + "learning_rate": 0.00033363908139692896, + "loss": 3.6441, + "step": 78460 + }, + { + "epoch": 5.33122706889523, + "grad_norm": 1.16849946975708, + "learning_rate": 0.00033359661638809624, + "loss": 3.5849, + "step": 78465 + }, + { + "epoch": 5.331566788965892, + "grad_norm": 0.9480420351028442, + "learning_rate": 0.00033355415137926347, + "loss": 3.258, + "step": 78470 + }, + { + "epoch": 5.331906509036554, + "grad_norm": 1.0793460607528687, + "learning_rate": 0.0003335116863704308, + "loss": 3.4073, + "step": 78475 + }, + { + "epoch": 5.3322462291072155, + "grad_norm": 0.8402575254440308, + "learning_rate": 0.0003334692213615981, + "loss": 3.6542, + "step": 78480 + }, + { + "epoch": 5.3325859491778775, + "grad_norm": 1.0523964166641235, + "learning_rate": 0.0003334267563527653, + "loss": 3.4659, + "step": 78485 + }, + { + "epoch": 5.33292566924854, + "grad_norm": 1.1226567029953003, + "learning_rate": 0.0003333842913439326, + "loss": 3.3645, + "step": 78490 + }, + { + "epoch": 5.333265389319201, + "grad_norm": 0.9237992167472839, + "learning_rate": 0.0003333418263350999, + "loss": 3.2472, + "step": 78495 + }, + { + "epoch": 5.333605109389863, + "grad_norm": 0.7687498927116394, + "learning_rate": 0.00033329936132626715, + "loss": 3.3268, + "step": 78500 + }, + { + "epoch": 5.333944829460524, + "grad_norm": 0.8323100209236145, + "learning_rate": 0.0003332568963174344, + "loss": 3.6368, + "step": 78505 + }, + { + "epoch": 5.334284549531186, + "grad_norm": 0.9502173066139221, + "learning_rate": 0.00033321443130860176, + "loss": 3.2524, + "step": 78510 + }, + { + "epoch": 5.334624269601848, + "grad_norm": 1.1399176120758057, + "learning_rate": 0.000333171966299769, + "loss": 3.2917, + "step": 78515 + }, + { + "epoch": 5.334963989672509, + "grad_norm": 0.9796766638755798, + "learning_rate": 0.00033312950129093627, + "loss": 3.5698, + "step": 78520 + }, + { + "epoch": 5.3353037097431715, + "grad_norm": 0.7864508032798767, + "learning_rate": 0.00033308703628210355, + "loss": 3.7325, + "step": 78525 + }, + { + "epoch": 5.3356434298138335, + "grad_norm": 0.8081654906272888, + "learning_rate": 0.0003330445712732708, + "loss": 3.5918, + "step": 78530 + }, + { + "epoch": 5.335983149884495, + "grad_norm": 0.7844996452331543, + "learning_rate": 0.0003330021062644381, + "loss": 3.5729, + "step": 78535 + }, + { + "epoch": 5.336322869955157, + "grad_norm": 0.9379047155380249, + "learning_rate": 0.0003329596412556054, + "loss": 3.1392, + "step": 78540 + }, + { + "epoch": 5.336662590025819, + "grad_norm": 0.8122665882110596, + "learning_rate": 0.00033291717624677267, + "loss": 3.3793, + "step": 78545 + }, + { + "epoch": 5.33700231009648, + "grad_norm": 0.8361498713493347, + "learning_rate": 0.00033287471123793995, + "loss": 3.2066, + "step": 78550 + }, + { + "epoch": 5.337342030167142, + "grad_norm": 0.7282031178474426, + "learning_rate": 0.00033283224622910723, + "loss": 3.3412, + "step": 78555 + }, + { + "epoch": 5.337681750237804, + "grad_norm": 0.9100168943405151, + "learning_rate": 0.00033278978122027445, + "loss": 3.7036, + "step": 78560 + }, + { + "epoch": 5.338021470308465, + "grad_norm": 1.0003844499588013, + "learning_rate": 0.0003327473162114418, + "loss": 3.2551, + "step": 78565 + }, + { + "epoch": 5.3383611903791275, + "grad_norm": 1.0355535745620728, + "learning_rate": 0.00033270485120260907, + "loss": 3.1545, + "step": 78570 + }, + { + "epoch": 5.33870091044979, + "grad_norm": 0.8881214261054993, + "learning_rate": 0.0003326623861937763, + "loss": 3.1973, + "step": 78575 + }, + { + "epoch": 5.339040630520451, + "grad_norm": 0.9087323546409607, + "learning_rate": 0.00033261992118494363, + "loss": 3.4466, + "step": 78580 + }, + { + "epoch": 5.339380350591113, + "grad_norm": 1.2324131727218628, + "learning_rate": 0.0003325774561761109, + "loss": 3.1033, + "step": 78585 + }, + { + "epoch": 5.339720070661775, + "grad_norm": 1.1290315389633179, + "learning_rate": 0.00033253499116727813, + "loss": 3.3005, + "step": 78590 + }, + { + "epoch": 5.340059790732436, + "grad_norm": 0.8952938914299011, + "learning_rate": 0.00033249252615844547, + "loss": 3.482, + "step": 78595 + }, + { + "epoch": 5.340399510803098, + "grad_norm": 1.058024525642395, + "learning_rate": 0.00033245006114961275, + "loss": 3.5366, + "step": 78600 + }, + { + "epoch": 5.34073923087376, + "grad_norm": 1.1427899599075317, + "learning_rate": 0.00033240759614078, + "loss": 3.4054, + "step": 78605 + }, + { + "epoch": 5.341078950944421, + "grad_norm": 0.9927167296409607, + "learning_rate": 0.00033236513113194725, + "loss": 3.3624, + "step": 78610 + }, + { + "epoch": 5.3414186710150835, + "grad_norm": 0.9428116083145142, + "learning_rate": 0.0003323226661231146, + "loss": 3.4031, + "step": 78615 + }, + { + "epoch": 5.341758391085746, + "grad_norm": 1.1242990493774414, + "learning_rate": 0.00033228020111428187, + "loss": 3.5939, + "step": 78620 + }, + { + "epoch": 5.342098111156407, + "grad_norm": 1.199422836303711, + "learning_rate": 0.0003322377361054491, + "loss": 3.357, + "step": 78625 + }, + { + "epoch": 5.342437831227069, + "grad_norm": 0.8705763220787048, + "learning_rate": 0.00033219527109661643, + "loss": 3.7788, + "step": 78630 + }, + { + "epoch": 5.342777551297731, + "grad_norm": 0.9926672577857971, + "learning_rate": 0.0003321528060877837, + "loss": 3.4802, + "step": 78635 + }, + { + "epoch": 5.343117271368392, + "grad_norm": 0.7579404711723328, + "learning_rate": 0.00033211034107895093, + "loss": 3.1995, + "step": 78640 + }, + { + "epoch": 5.343456991439054, + "grad_norm": 1.2075010538101196, + "learning_rate": 0.0003320678760701182, + "loss": 3.5443, + "step": 78645 + }, + { + "epoch": 5.343796711509716, + "grad_norm": 1.172472596168518, + "learning_rate": 0.00033202541106128555, + "loss": 3.2991, + "step": 78650 + }, + { + "epoch": 5.344136431580377, + "grad_norm": 0.8851138949394226, + "learning_rate": 0.0003319829460524528, + "loss": 3.3717, + "step": 78655 + }, + { + "epoch": 5.3444761516510395, + "grad_norm": 0.939117968082428, + "learning_rate": 0.00033194048104362005, + "loss": 3.6789, + "step": 78660 + }, + { + "epoch": 5.344815871721702, + "grad_norm": 1.0617202520370483, + "learning_rate": 0.0003318980160347874, + "loss": 3.4704, + "step": 78665 + }, + { + "epoch": 5.345155591792363, + "grad_norm": 0.9368365406990051, + "learning_rate": 0.0003318555510259546, + "loss": 3.5109, + "step": 78670 + }, + { + "epoch": 5.345495311863025, + "grad_norm": 0.8902553915977478, + "learning_rate": 0.0003318130860171219, + "loss": 3.3894, + "step": 78675 + }, + { + "epoch": 5.345835031933687, + "grad_norm": 1.0431864261627197, + "learning_rate": 0.0003317706210082892, + "loss": 3.4123, + "step": 78680 + }, + { + "epoch": 5.346174752004348, + "grad_norm": 0.8060005903244019, + "learning_rate": 0.00033172815599945645, + "loss": 3.4044, + "step": 78685 + }, + { + "epoch": 5.34651447207501, + "grad_norm": 0.7717661261558533, + "learning_rate": 0.00033168569099062373, + "loss": 3.5782, + "step": 78690 + }, + { + "epoch": 5.346854192145672, + "grad_norm": 0.8121126890182495, + "learning_rate": 0.000331643225981791, + "loss": 3.6172, + "step": 78695 + }, + { + "epoch": 5.3471939122163334, + "grad_norm": 0.8378850221633911, + "learning_rate": 0.0003316007609729583, + "loss": 3.5973, + "step": 78700 + }, + { + "epoch": 5.3475336322869955, + "grad_norm": 0.8326839208602905, + "learning_rate": 0.0003315582959641256, + "loss": 3.5054, + "step": 78705 + }, + { + "epoch": 5.347873352357658, + "grad_norm": 0.7965877056121826, + "learning_rate": 0.00033151583095529286, + "loss": 3.0914, + "step": 78710 + }, + { + "epoch": 5.348213072428319, + "grad_norm": 0.7931357622146606, + "learning_rate": 0.0003314733659464601, + "loss": 3.3131, + "step": 78715 + }, + { + "epoch": 5.348552792498981, + "grad_norm": 0.9721649289131165, + "learning_rate": 0.0003314309009376274, + "loss": 3.3039, + "step": 78720 + }, + { + "epoch": 5.348892512569643, + "grad_norm": 0.8588797450065613, + "learning_rate": 0.0003313884359287947, + "loss": 3.6082, + "step": 78725 + }, + { + "epoch": 5.349232232640304, + "grad_norm": 0.9456138014793396, + "learning_rate": 0.0003313459709199619, + "loss": 3.5917, + "step": 78730 + }, + { + "epoch": 5.349571952710966, + "grad_norm": 0.818286120891571, + "learning_rate": 0.00033130350591112926, + "loss": 3.407, + "step": 78735 + }, + { + "epoch": 5.349911672781628, + "grad_norm": 0.8564689755439758, + "learning_rate": 0.00033126104090229654, + "loss": 3.4993, + "step": 78740 + }, + { + "epoch": 5.3502513928522895, + "grad_norm": 1.0860605239868164, + "learning_rate": 0.00033121857589346376, + "loss": 3.3591, + "step": 78745 + }, + { + "epoch": 5.3505911129229515, + "grad_norm": 0.9525011777877808, + "learning_rate": 0.00033117611088463104, + "loss": 3.504, + "step": 78750 + }, + { + "epoch": 5.350930832993614, + "grad_norm": 0.8574268817901611, + "learning_rate": 0.0003311336458757984, + "loss": 3.5251, + "step": 78755 + }, + { + "epoch": 5.351270553064275, + "grad_norm": 0.843839168548584, + "learning_rate": 0.0003310911808669656, + "loss": 3.4859, + "step": 78760 + }, + { + "epoch": 5.351610273134937, + "grad_norm": 0.902587354183197, + "learning_rate": 0.0003310487158581329, + "loss": 3.6356, + "step": 78765 + }, + { + "epoch": 5.351949993205599, + "grad_norm": 0.8871144652366638, + "learning_rate": 0.0003310062508493002, + "loss": 3.3233, + "step": 78770 + }, + { + "epoch": 5.35228971327626, + "grad_norm": 0.8606464266777039, + "learning_rate": 0.00033096378584046744, + "loss": 3.324, + "step": 78775 + }, + { + "epoch": 5.352629433346922, + "grad_norm": 0.8877620100975037, + "learning_rate": 0.0003309213208316347, + "loss": 3.4972, + "step": 78780 + }, + { + "epoch": 5.352969153417584, + "grad_norm": 1.071241855621338, + "learning_rate": 0.000330878855822802, + "loss": 3.2461, + "step": 78785 + }, + { + "epoch": 5.3533088734882455, + "grad_norm": 0.7963054180145264, + "learning_rate": 0.00033083639081396934, + "loss": 3.5382, + "step": 78790 + }, + { + "epoch": 5.3536485935589075, + "grad_norm": 0.8161731958389282, + "learning_rate": 0.00033079392580513656, + "loss": 3.2946, + "step": 78795 + }, + { + "epoch": 5.35398831362957, + "grad_norm": 0.8352603316307068, + "learning_rate": 0.00033075146079630384, + "loss": 3.1941, + "step": 78800 + }, + { + "epoch": 5.354328033700231, + "grad_norm": 0.9309561848640442, + "learning_rate": 0.0003307089957874712, + "loss": 3.4028, + "step": 78805 + }, + { + "epoch": 5.354667753770893, + "grad_norm": 1.0791726112365723, + "learning_rate": 0.0003306665307786384, + "loss": 3.4282, + "step": 78810 + }, + { + "epoch": 5.355007473841555, + "grad_norm": 0.7636476159095764, + "learning_rate": 0.0003306240657698057, + "loss": 3.3885, + "step": 78815 + }, + { + "epoch": 5.355347193912216, + "grad_norm": 0.8618056774139404, + "learning_rate": 0.00033058160076097296, + "loss": 3.5726, + "step": 78820 + }, + { + "epoch": 5.355686913982878, + "grad_norm": 0.6847010254859924, + "learning_rate": 0.00033053913575214024, + "loss": 3.4468, + "step": 78825 + }, + { + "epoch": 5.35602663405354, + "grad_norm": 0.8719287514686584, + "learning_rate": 0.0003304966707433075, + "loss": 3.3847, + "step": 78830 + }, + { + "epoch": 5.3563663541242015, + "grad_norm": 1.0832884311676025, + "learning_rate": 0.0003304542057344748, + "loss": 3.4289, + "step": 78835 + }, + { + "epoch": 5.3567060741948636, + "grad_norm": 0.9755337834358215, + "learning_rate": 0.0003304117407256421, + "loss": 3.4823, + "step": 78840 + }, + { + "epoch": 5.357045794265526, + "grad_norm": 0.8966276049613953, + "learning_rate": 0.00033036927571680936, + "loss": 3.4288, + "step": 78845 + }, + { + "epoch": 5.357385514336187, + "grad_norm": 0.7190046906471252, + "learning_rate": 0.00033032681070797664, + "loss": 3.4137, + "step": 78850 + }, + { + "epoch": 5.357725234406849, + "grad_norm": 0.9158503413200378, + "learning_rate": 0.00033028434569914387, + "loss": 3.527, + "step": 78855 + }, + { + "epoch": 5.358064954477511, + "grad_norm": 1.1003084182739258, + "learning_rate": 0.0003302418806903112, + "loss": 3.3097, + "step": 78860 + }, + { + "epoch": 5.358404674548172, + "grad_norm": 0.8053983449935913, + "learning_rate": 0.0003301994156814785, + "loss": 3.4382, + "step": 78865 + }, + { + "epoch": 5.358744394618834, + "grad_norm": 0.738785445690155, + "learning_rate": 0.0003301569506726457, + "loss": 3.3709, + "step": 78870 + }, + { + "epoch": 5.359084114689496, + "grad_norm": 0.8625181913375854, + "learning_rate": 0.00033011448566381304, + "loss": 3.507, + "step": 78875 + }, + { + "epoch": 5.3594238347601575, + "grad_norm": 0.7818925976753235, + "learning_rate": 0.0003300720206549803, + "loss": 3.5248, + "step": 78880 + }, + { + "epoch": 5.35976355483082, + "grad_norm": 0.7681019306182861, + "learning_rate": 0.00033002955564614755, + "loss": 3.4928, + "step": 78885 + }, + { + "epoch": 5.360103274901482, + "grad_norm": 0.8309361338615417, + "learning_rate": 0.0003299870906373149, + "loss": 3.6648, + "step": 78890 + }, + { + "epoch": 5.360442994972143, + "grad_norm": 0.8129064440727234, + "learning_rate": 0.00032994462562848216, + "loss": 3.4279, + "step": 78895 + }, + { + "epoch": 5.360782715042805, + "grad_norm": 0.9137693047523499, + "learning_rate": 0.0003299021606196494, + "loss": 3.3189, + "step": 78900 + }, + { + "epoch": 5.361122435113466, + "grad_norm": 0.86076420545578, + "learning_rate": 0.00032985969561081667, + "loss": 3.4705, + "step": 78905 + }, + { + "epoch": 5.361462155184128, + "grad_norm": 0.8551079630851746, + "learning_rate": 0.000329817230601984, + "loss": 3.2848, + "step": 78910 + }, + { + "epoch": 5.36180187525479, + "grad_norm": 1.1348166465759277, + "learning_rate": 0.00032977476559315123, + "loss": 3.3761, + "step": 78915 + }, + { + "epoch": 5.362141595325451, + "grad_norm": 0.9199367761611938, + "learning_rate": 0.0003297323005843185, + "loss": 3.6059, + "step": 78920 + }, + { + "epoch": 5.3624813153961135, + "grad_norm": 1.0361356735229492, + "learning_rate": 0.00032968983557548584, + "loss": 3.4922, + "step": 78925 + }, + { + "epoch": 5.362821035466776, + "grad_norm": 0.7577531933784485, + "learning_rate": 0.00032964737056665307, + "loss": 3.4155, + "step": 78930 + }, + { + "epoch": 5.363160755537437, + "grad_norm": 0.9130941033363342, + "learning_rate": 0.00032960490555782035, + "loss": 3.5112, + "step": 78935 + }, + { + "epoch": 5.363500475608099, + "grad_norm": 1.0524455308914185, + "learning_rate": 0.00032956244054898763, + "loss": 3.4349, + "step": 78940 + }, + { + "epoch": 5.363840195678761, + "grad_norm": 1.1161905527114868, + "learning_rate": 0.0003295199755401549, + "loss": 3.4375, + "step": 78945 + }, + { + "epoch": 5.364179915749422, + "grad_norm": 0.8131364583969116, + "learning_rate": 0.0003294775105313222, + "loss": 3.5253, + "step": 78950 + }, + { + "epoch": 5.364519635820084, + "grad_norm": 1.1910043954849243, + "learning_rate": 0.00032943504552248947, + "loss": 3.5102, + "step": 78955 + }, + { + "epoch": 5.364859355890746, + "grad_norm": 0.8790878057479858, + "learning_rate": 0.0003293925805136568, + "loss": 3.2448, + "step": 78960 + }, + { + "epoch": 5.365199075961407, + "grad_norm": 0.9880867600440979, + "learning_rate": 0.00032935011550482403, + "loss": 3.3322, + "step": 78965 + }, + { + "epoch": 5.3655387960320695, + "grad_norm": 0.8204182982444763, + "learning_rate": 0.0003293076504959913, + "loss": 3.4402, + "step": 78970 + }, + { + "epoch": 5.365878516102732, + "grad_norm": 0.737518846988678, + "learning_rate": 0.0003292651854871586, + "loss": 3.7246, + "step": 78975 + }, + { + "epoch": 5.366218236173393, + "grad_norm": 0.8896304368972778, + "learning_rate": 0.00032922272047832587, + "loss": 3.3018, + "step": 78980 + }, + { + "epoch": 5.366557956244055, + "grad_norm": 0.8662015199661255, + "learning_rate": 0.00032918025546949315, + "loss": 3.388, + "step": 78985 + }, + { + "epoch": 5.366897676314717, + "grad_norm": 0.885512113571167, + "learning_rate": 0.00032913779046066043, + "loss": 3.2371, + "step": 78990 + }, + { + "epoch": 5.367237396385378, + "grad_norm": 0.8255910873413086, + "learning_rate": 0.0003290953254518277, + "loss": 3.4129, + "step": 78995 + }, + { + "epoch": 5.36757711645604, + "grad_norm": 0.9838322997093201, + "learning_rate": 0.000329052860442995, + "loss": 3.3574, + "step": 79000 + }, + { + "epoch": 5.367916836526702, + "grad_norm": 1.1750190258026123, + "learning_rate": 0.00032901039543416227, + "loss": 3.4185, + "step": 79005 + }, + { + "epoch": 5.3682565565973634, + "grad_norm": 0.8219663500785828, + "learning_rate": 0.0003289679304253295, + "loss": 3.6836, + "step": 79010 + }, + { + "epoch": 5.3685962766680255, + "grad_norm": 1.0848491191864014, + "learning_rate": 0.00032892546541649683, + "loss": 3.393, + "step": 79015 + }, + { + "epoch": 5.368935996738688, + "grad_norm": 1.1454834938049316, + "learning_rate": 0.0003288830004076641, + "loss": 3.5149, + "step": 79020 + }, + { + "epoch": 5.369275716809349, + "grad_norm": 0.9534928202629089, + "learning_rate": 0.00032884053539883134, + "loss": 3.33, + "step": 79025 + }, + { + "epoch": 5.369615436880011, + "grad_norm": 0.9042555689811707, + "learning_rate": 0.00032879807038999867, + "loss": 3.549, + "step": 79030 + }, + { + "epoch": 5.369955156950673, + "grad_norm": 0.785645067691803, + "learning_rate": 0.00032875560538116595, + "loss": 3.5386, + "step": 79035 + }, + { + "epoch": 5.370294877021334, + "grad_norm": 0.9614004492759705, + "learning_rate": 0.0003287131403723332, + "loss": 3.5598, + "step": 79040 + }, + { + "epoch": 5.370634597091996, + "grad_norm": 0.9631426334381104, + "learning_rate": 0.00032867067536350046, + "loss": 3.3806, + "step": 79045 + }, + { + "epoch": 5.370974317162658, + "grad_norm": 1.1190179586410522, + "learning_rate": 0.0003286282103546678, + "loss": 3.3421, + "step": 79050 + }, + { + "epoch": 5.3713140372333195, + "grad_norm": 1.0577495098114014, + "learning_rate": 0.000328585745345835, + "loss": 3.4473, + "step": 79055 + }, + { + "epoch": 5.3716537573039815, + "grad_norm": 0.8098303079605103, + "learning_rate": 0.0003285432803370023, + "loss": 3.4139, + "step": 79060 + }, + { + "epoch": 5.371993477374644, + "grad_norm": 0.9690068364143372, + "learning_rate": 0.00032850081532816963, + "loss": 3.5535, + "step": 79065 + }, + { + "epoch": 5.372333197445305, + "grad_norm": 0.9729496836662292, + "learning_rate": 0.00032845835031933686, + "loss": 3.3113, + "step": 79070 + }, + { + "epoch": 5.372672917515967, + "grad_norm": 0.6891639828681946, + "learning_rate": 0.00032841588531050414, + "loss": 3.3658, + "step": 79075 + }, + { + "epoch": 5.373012637586629, + "grad_norm": 0.7098758220672607, + "learning_rate": 0.0003283734203016714, + "loss": 3.3595, + "step": 79080 + }, + { + "epoch": 5.37335235765729, + "grad_norm": 0.8433969020843506, + "learning_rate": 0.0003283309552928387, + "loss": 3.3943, + "step": 79085 + }, + { + "epoch": 5.373692077727952, + "grad_norm": 1.0230627059936523, + "learning_rate": 0.000328288490284006, + "loss": 3.3344, + "step": 79090 + }, + { + "epoch": 5.374031797798614, + "grad_norm": 1.0803172588348389, + "learning_rate": 0.00032824602527517326, + "loss": 3.2611, + "step": 79095 + }, + { + "epoch": 5.3743715178692755, + "grad_norm": 0.8001290559768677, + "learning_rate": 0.00032820356026634054, + "loss": 3.2865, + "step": 79100 + }, + { + "epoch": 5.3747112379399375, + "grad_norm": 0.928577184677124, + "learning_rate": 0.0003281610952575078, + "loss": 3.4242, + "step": 79105 + }, + { + "epoch": 5.3750509580106, + "grad_norm": 0.8327693343162537, + "learning_rate": 0.0003281186302486751, + "loss": 3.3438, + "step": 79110 + }, + { + "epoch": 5.375390678081261, + "grad_norm": 0.9437814950942993, + "learning_rate": 0.0003280761652398423, + "loss": 3.2908, + "step": 79115 + }, + { + "epoch": 5.375730398151923, + "grad_norm": 1.0406646728515625, + "learning_rate": 0.00032803370023100966, + "loss": 3.4995, + "step": 79120 + }, + { + "epoch": 5.376070118222585, + "grad_norm": 0.8902111053466797, + "learning_rate": 0.00032799123522217694, + "loss": 3.364, + "step": 79125 + }, + { + "epoch": 5.376409838293246, + "grad_norm": 0.9598656296730042, + "learning_rate": 0.0003279487702133442, + "loss": 3.5307, + "step": 79130 + }, + { + "epoch": 5.376749558363908, + "grad_norm": 1.0133708715438843, + "learning_rate": 0.0003279063052045115, + "loss": 3.3971, + "step": 79135 + }, + { + "epoch": 5.37708927843457, + "grad_norm": 0.9632912874221802, + "learning_rate": 0.0003278638401956788, + "loss": 3.4648, + "step": 79140 + }, + { + "epoch": 5.3774289985052315, + "grad_norm": 1.2473727464675903, + "learning_rate": 0.00032782137518684606, + "loss": 3.6466, + "step": 79145 + }, + { + "epoch": 5.377768718575894, + "grad_norm": 0.8775381445884705, + "learning_rate": 0.0003277789101780133, + "loss": 3.4368, + "step": 79150 + }, + { + "epoch": 5.378108438646556, + "grad_norm": 0.9144920706748962, + "learning_rate": 0.0003277364451691806, + "loss": 3.3794, + "step": 79155 + }, + { + "epoch": 5.378448158717217, + "grad_norm": 0.75979083776474, + "learning_rate": 0.0003276939801603479, + "loss": 3.4923, + "step": 79160 + }, + { + "epoch": 5.378787878787879, + "grad_norm": 1.0801957845687866, + "learning_rate": 0.0003276515151515151, + "loss": 3.3828, + "step": 79165 + }, + { + "epoch": 5.379127598858541, + "grad_norm": 0.7748429775238037, + "learning_rate": 0.00032760905014268246, + "loss": 3.6666, + "step": 79170 + }, + { + "epoch": 5.379467318929202, + "grad_norm": 1.0417946577072144, + "learning_rate": 0.00032756658513384974, + "loss": 3.257, + "step": 79175 + }, + { + "epoch": 5.379807038999864, + "grad_norm": 0.9298911690711975, + "learning_rate": 0.00032752412012501697, + "loss": 3.6153, + "step": 79180 + }, + { + "epoch": 5.380146759070525, + "grad_norm": 0.8416727185249329, + "learning_rate": 0.0003274816551161843, + "loss": 3.4407, + "step": 79185 + }, + { + "epoch": 5.3804864791411875, + "grad_norm": 0.7464411854743958, + "learning_rate": 0.0003274391901073516, + "loss": 3.5058, + "step": 79190 + }, + { + "epoch": 5.38082619921185, + "grad_norm": 0.7843184471130371, + "learning_rate": 0.0003273967250985188, + "loss": 3.5201, + "step": 79195 + }, + { + "epoch": 5.381165919282511, + "grad_norm": 1.0229194164276123, + "learning_rate": 0.0003273542600896861, + "loss": 3.5614, + "step": 79200 + }, + { + "epoch": 5.381505639353173, + "grad_norm": 0.7081983685493469, + "learning_rate": 0.0003273117950808534, + "loss": 3.3381, + "step": 79205 + }, + { + "epoch": 5.381845359423835, + "grad_norm": 0.890373170375824, + "learning_rate": 0.00032726933007202065, + "loss": 3.4309, + "step": 79210 + }, + { + "epoch": 5.382185079494496, + "grad_norm": 1.044782280921936, + "learning_rate": 0.0003272268650631879, + "loss": 3.3992, + "step": 79215 + }, + { + "epoch": 5.382524799565158, + "grad_norm": 0.7827876806259155, + "learning_rate": 0.00032718440005435526, + "loss": 3.4095, + "step": 79220 + }, + { + "epoch": 5.38286451963582, + "grad_norm": 0.838117778301239, + "learning_rate": 0.0003271419350455225, + "loss": 3.3811, + "step": 79225 + }, + { + "epoch": 5.383204239706481, + "grad_norm": 1.0035732984542847, + "learning_rate": 0.00032709947003668977, + "loss": 3.631, + "step": 79230 + }, + { + "epoch": 5.3835439597771435, + "grad_norm": 1.2847179174423218, + "learning_rate": 0.00032705700502785705, + "loss": 3.6543, + "step": 79235 + }, + { + "epoch": 5.383883679847806, + "grad_norm": 0.85878986120224, + "learning_rate": 0.0003270145400190243, + "loss": 3.6099, + "step": 79240 + }, + { + "epoch": 5.384223399918467, + "grad_norm": 0.8680986762046814, + "learning_rate": 0.0003269720750101916, + "loss": 3.3341, + "step": 79245 + }, + { + "epoch": 5.384563119989129, + "grad_norm": 0.9104351997375488, + "learning_rate": 0.0003269296100013589, + "loss": 3.2486, + "step": 79250 + }, + { + "epoch": 5.384902840059791, + "grad_norm": 1.06619131565094, + "learning_rate": 0.00032688714499252617, + "loss": 3.2033, + "step": 79255 + }, + { + "epoch": 5.385242560130452, + "grad_norm": 0.8084930777549744, + "learning_rate": 0.00032684467998369345, + "loss": 3.4819, + "step": 79260 + }, + { + "epoch": 5.385582280201114, + "grad_norm": 1.017953872680664, + "learning_rate": 0.0003268022149748607, + "loss": 3.4906, + "step": 79265 + }, + { + "epoch": 5.385922000271776, + "grad_norm": 0.8811764717102051, + "learning_rate": 0.00032675974996602795, + "loss": 3.3632, + "step": 79270 + }, + { + "epoch": 5.386261720342437, + "grad_norm": 0.8606991767883301, + "learning_rate": 0.0003267172849571953, + "loss": 3.4615, + "step": 79275 + }, + { + "epoch": 5.3866014404130995, + "grad_norm": 0.7611544132232666, + "learning_rate": 0.00032667481994836257, + "loss": 3.4547, + "step": 79280 + }, + { + "epoch": 5.386941160483762, + "grad_norm": 0.7933346033096313, + "learning_rate": 0.0003266323549395298, + "loss": 3.2346, + "step": 79285 + }, + { + "epoch": 5.387280880554423, + "grad_norm": 0.7401696443557739, + "learning_rate": 0.00032658988993069713, + "loss": 3.3983, + "step": 79290 + }, + { + "epoch": 5.387620600625085, + "grad_norm": 0.7758470177650452, + "learning_rate": 0.0003265474249218644, + "loss": 3.4201, + "step": 79295 + }, + { + "epoch": 5.387960320695747, + "grad_norm": 0.8723129630088806, + "learning_rate": 0.0003265049599130317, + "loss": 3.225, + "step": 79300 + }, + { + "epoch": 5.388300040766408, + "grad_norm": 1.0921236276626587, + "learning_rate": 0.0003264624949041989, + "loss": 3.3489, + "step": 79305 + }, + { + "epoch": 5.38863976083707, + "grad_norm": 0.6928749084472656, + "learning_rate": 0.00032642002989536625, + "loss": 3.4743, + "step": 79310 + }, + { + "epoch": 5.388979480907732, + "grad_norm": 0.935758113861084, + "learning_rate": 0.00032637756488653353, + "loss": 3.3816, + "step": 79315 + }, + { + "epoch": 5.3893192009783935, + "grad_norm": 0.9053452014923096, + "learning_rate": 0.00032633509987770075, + "loss": 3.4927, + "step": 79320 + }, + { + "epoch": 5.3896589210490555, + "grad_norm": 0.9204387068748474, + "learning_rate": 0.0003262926348688681, + "loss": 3.6408, + "step": 79325 + }, + { + "epoch": 5.389998641119718, + "grad_norm": 0.8386547565460205, + "learning_rate": 0.00032625016986003537, + "loss": 3.544, + "step": 79330 + }, + { + "epoch": 5.390338361190379, + "grad_norm": 1.1313090324401855, + "learning_rate": 0.0003262077048512026, + "loss": 3.3465, + "step": 79335 + }, + { + "epoch": 5.390678081261041, + "grad_norm": 0.6973360180854797, + "learning_rate": 0.0003261652398423699, + "loss": 3.5456, + "step": 79340 + }, + { + "epoch": 5.391017801331703, + "grad_norm": 1.0095964670181274, + "learning_rate": 0.0003261227748335372, + "loss": 3.3621, + "step": 79345 + }, + { + "epoch": 5.391357521402364, + "grad_norm": 1.1098952293395996, + "learning_rate": 0.00032608030982470443, + "loss": 3.5496, + "step": 79350 + }, + { + "epoch": 5.391697241473026, + "grad_norm": 0.9603663086891174, + "learning_rate": 0.0003260378448158717, + "loss": 3.3416, + "step": 79355 + }, + { + "epoch": 5.392036961543688, + "grad_norm": 0.8548449873924255, + "learning_rate": 0.00032599537980703905, + "loss": 3.3143, + "step": 79360 + }, + { + "epoch": 5.3923766816143495, + "grad_norm": 0.987759530544281, + "learning_rate": 0.0003259529147982063, + "loss": 3.3978, + "step": 79365 + }, + { + "epoch": 5.3927164016850115, + "grad_norm": 0.9195972084999084, + "learning_rate": 0.00032591044978937355, + "loss": 3.1941, + "step": 79370 + }, + { + "epoch": 5.393056121755674, + "grad_norm": 0.8617194294929504, + "learning_rate": 0.00032586798478054083, + "loss": 3.438, + "step": 79375 + }, + { + "epoch": 5.393395841826335, + "grad_norm": 0.8858255743980408, + "learning_rate": 0.0003258255197717081, + "loss": 3.177, + "step": 79380 + }, + { + "epoch": 5.393735561896997, + "grad_norm": 0.987254798412323, + "learning_rate": 0.0003257830547628754, + "loss": 3.4465, + "step": 79385 + }, + { + "epoch": 5.394075281967659, + "grad_norm": 0.929149329662323, + "learning_rate": 0.0003257405897540427, + "loss": 3.574, + "step": 79390 + }, + { + "epoch": 5.39441500203832, + "grad_norm": 1.1347044706344604, + "learning_rate": 0.00032569812474520995, + "loss": 3.3371, + "step": 79395 + }, + { + "epoch": 5.394754722108982, + "grad_norm": 1.0154848098754883, + "learning_rate": 0.00032565565973637723, + "loss": 3.3966, + "step": 79400 + }, + { + "epoch": 5.395094442179644, + "grad_norm": 0.9472141265869141, + "learning_rate": 0.0003256131947275445, + "loss": 3.2876, + "step": 79405 + }, + { + "epoch": 5.3954341622503055, + "grad_norm": 1.385036587715149, + "learning_rate": 0.00032557072971871174, + "loss": 3.6713, + "step": 79410 + }, + { + "epoch": 5.3957738823209676, + "grad_norm": 1.0803108215332031, + "learning_rate": 0.0003255282647098791, + "loss": 3.5299, + "step": 79415 + }, + { + "epoch": 5.39611360239163, + "grad_norm": 0.6747039556503296, + "learning_rate": 0.00032548579970104636, + "loss": 3.4854, + "step": 79420 + }, + { + "epoch": 5.396453322462291, + "grad_norm": 0.8044858574867249, + "learning_rate": 0.0003254433346922136, + "loss": 3.4284, + "step": 79425 + }, + { + "epoch": 5.396793042532953, + "grad_norm": 0.927371084690094, + "learning_rate": 0.0003254008696833809, + "loss": 3.4305, + "step": 79430 + }, + { + "epoch": 5.397132762603615, + "grad_norm": 1.0387723445892334, + "learning_rate": 0.0003253584046745482, + "loss": 3.4859, + "step": 79435 + }, + { + "epoch": 5.397472482674276, + "grad_norm": 0.6959279775619507, + "learning_rate": 0.0003253159396657154, + "loss": 3.7497, + "step": 79440 + }, + { + "epoch": 5.397812202744938, + "grad_norm": 0.8004673719406128, + "learning_rate": 0.00032527347465688276, + "loss": 3.4356, + "step": 79445 + }, + { + "epoch": 5.3981519228156, + "grad_norm": 0.9559466242790222, + "learning_rate": 0.00032523100964805004, + "loss": 3.2928, + "step": 79450 + }, + { + "epoch": 5.3984916428862615, + "grad_norm": 1.3047997951507568, + "learning_rate": 0.00032518854463921726, + "loss": 3.2568, + "step": 79455 + }, + { + "epoch": 5.398831362956924, + "grad_norm": 0.8160944581031799, + "learning_rate": 0.00032514607963038454, + "loss": 3.6309, + "step": 79460 + }, + { + "epoch": 5.399171083027586, + "grad_norm": 1.0529931783676147, + "learning_rate": 0.0003251036146215519, + "loss": 3.0671, + "step": 79465 + }, + { + "epoch": 5.399510803098247, + "grad_norm": 0.8836321234703064, + "learning_rate": 0.00032506114961271916, + "loss": 3.439, + "step": 79470 + }, + { + "epoch": 5.399850523168909, + "grad_norm": 1.0307550430297852, + "learning_rate": 0.0003250186846038864, + "loss": 3.5278, + "step": 79475 + }, + { + "epoch": 5.400190243239571, + "grad_norm": 0.7754269242286682, + "learning_rate": 0.0003249762195950537, + "loss": 3.4648, + "step": 79480 + }, + { + "epoch": 5.400529963310232, + "grad_norm": 0.8441033959388733, + "learning_rate": 0.000324933754586221, + "loss": 3.1308, + "step": 79485 + }, + { + "epoch": 5.400869683380894, + "grad_norm": 1.055978775024414, + "learning_rate": 0.0003248912895773882, + "loss": 3.582, + "step": 79490 + }, + { + "epoch": 5.401209403451556, + "grad_norm": 0.8787696957588196, + "learning_rate": 0.0003248488245685555, + "loss": 3.55, + "step": 79495 + }, + { + "epoch": 5.4015491235222175, + "grad_norm": 0.9810678958892822, + "learning_rate": 0.00032480635955972284, + "loss": 3.4666, + "step": 79500 + }, + { + "epoch": 5.40188884359288, + "grad_norm": 1.0133280754089355, + "learning_rate": 0.00032476389455089006, + "loss": 3.3455, + "step": 79505 + }, + { + "epoch": 5.402228563663542, + "grad_norm": 0.8044341206550598, + "learning_rate": 0.00032472142954205734, + "loss": 3.2825, + "step": 79510 + }, + { + "epoch": 5.402568283734203, + "grad_norm": 1.0735983848571777, + "learning_rate": 0.0003246789645332247, + "loss": 3.4489, + "step": 79515 + }, + { + "epoch": 5.402908003804865, + "grad_norm": 0.9021583199501038, + "learning_rate": 0.0003246364995243919, + "loss": 3.4252, + "step": 79520 + }, + { + "epoch": 5.403247723875527, + "grad_norm": 0.9457162022590637, + "learning_rate": 0.0003245940345155592, + "loss": 3.5134, + "step": 79525 + }, + { + "epoch": 5.403587443946188, + "grad_norm": 1.0457149744033813, + "learning_rate": 0.00032455156950672646, + "loss": 3.5163, + "step": 79530 + }, + { + "epoch": 5.40392716401685, + "grad_norm": 0.9417846202850342, + "learning_rate": 0.00032450910449789374, + "loss": 3.2179, + "step": 79535 + }, + { + "epoch": 5.404266884087512, + "grad_norm": 0.9906789660453796, + "learning_rate": 0.000324466639489061, + "loss": 3.4197, + "step": 79540 + }, + { + "epoch": 5.4046066041581735, + "grad_norm": 0.8001586198806763, + "learning_rate": 0.0003244241744802283, + "loss": 3.4652, + "step": 79545 + }, + { + "epoch": 5.404946324228836, + "grad_norm": 0.8556780219078064, + "learning_rate": 0.0003243817094713956, + "loss": 3.5436, + "step": 79550 + }, + { + "epoch": 5.405286044299498, + "grad_norm": 0.8944543600082397, + "learning_rate": 0.00032433924446256286, + "loss": 3.6504, + "step": 79555 + }, + { + "epoch": 5.405625764370159, + "grad_norm": 0.8708217740058899, + "learning_rate": 0.00032429677945373014, + "loss": 3.5562, + "step": 79560 + }, + { + "epoch": 5.405965484440821, + "grad_norm": 0.9657374620437622, + "learning_rate": 0.00032425431444489737, + "loss": 3.3559, + "step": 79565 + }, + { + "epoch": 5.406305204511483, + "grad_norm": 0.8089213371276855, + "learning_rate": 0.0003242118494360647, + "loss": 3.5497, + "step": 79570 + }, + { + "epoch": 5.406644924582144, + "grad_norm": 1.0684189796447754, + "learning_rate": 0.000324169384427232, + "loss": 3.313, + "step": 79575 + }, + { + "epoch": 5.406984644652806, + "grad_norm": 1.1513299942016602, + "learning_rate": 0.0003241269194183992, + "loss": 3.5311, + "step": 79580 + }, + { + "epoch": 5.4073243647234674, + "grad_norm": 0.8903722763061523, + "learning_rate": 0.00032408445440956654, + "loss": 3.2516, + "step": 79585 + }, + { + "epoch": 5.4076640847941295, + "grad_norm": 0.8179778456687927, + "learning_rate": 0.0003240419894007338, + "loss": 3.3555, + "step": 79590 + }, + { + "epoch": 5.408003804864792, + "grad_norm": 1.0604883432388306, + "learning_rate": 0.00032399952439190105, + "loss": 3.1671, + "step": 79595 + }, + { + "epoch": 5.408343524935453, + "grad_norm": 0.7640420794487, + "learning_rate": 0.00032395705938306833, + "loss": 3.8101, + "step": 79600 + }, + { + "epoch": 5.408683245006115, + "grad_norm": 0.7050489783287048, + "learning_rate": 0.00032391459437423566, + "loss": 3.3212, + "step": 79605 + }, + { + "epoch": 5.409022965076777, + "grad_norm": 0.916081964969635, + "learning_rate": 0.0003238721293654029, + "loss": 3.4027, + "step": 79610 + }, + { + "epoch": 5.409362685147438, + "grad_norm": 0.8492266535758972, + "learning_rate": 0.00032382966435657017, + "loss": 3.1091, + "step": 79615 + }, + { + "epoch": 5.4097024052181, + "grad_norm": 0.8164465427398682, + "learning_rate": 0.0003237871993477375, + "loss": 3.2132, + "step": 79620 + }, + { + "epoch": 5.410042125288762, + "grad_norm": 0.8457860946655273, + "learning_rate": 0.00032374473433890473, + "loss": 3.6395, + "step": 79625 + }, + { + "epoch": 5.4103818453594235, + "grad_norm": 1.1181838512420654, + "learning_rate": 0.000323702269330072, + "loss": 3.3983, + "step": 79630 + }, + { + "epoch": 5.4107215654300855, + "grad_norm": 0.8906224370002747, + "learning_rate": 0.0003236598043212393, + "loss": 3.5768, + "step": 79635 + }, + { + "epoch": 5.411061285500748, + "grad_norm": 0.8479645848274231, + "learning_rate": 0.0003236173393124066, + "loss": 3.2742, + "step": 79640 + }, + { + "epoch": 5.411401005571409, + "grad_norm": 0.8457931876182556, + "learning_rate": 0.00032357487430357385, + "loss": 3.3738, + "step": 79645 + }, + { + "epoch": 5.411740725642071, + "grad_norm": 0.8787240386009216, + "learning_rate": 0.00032353240929474113, + "loss": 3.2301, + "step": 79650 + }, + { + "epoch": 5.412080445712733, + "grad_norm": 0.7399342656135559, + "learning_rate": 0.00032348994428590846, + "loss": 3.5646, + "step": 79655 + }, + { + "epoch": 5.412420165783394, + "grad_norm": 0.9568170309066772, + "learning_rate": 0.0003234474792770757, + "loss": 3.3103, + "step": 79660 + }, + { + "epoch": 5.412759885854056, + "grad_norm": 1.0208185911178589, + "learning_rate": 0.00032340501426824297, + "loss": 3.5883, + "step": 79665 + }, + { + "epoch": 5.413099605924718, + "grad_norm": 1.0199062824249268, + "learning_rate": 0.00032336254925941025, + "loss": 3.4654, + "step": 79670 + }, + { + "epoch": 5.4134393259953795, + "grad_norm": 0.8192639350891113, + "learning_rate": 0.00032332008425057753, + "loss": 3.5133, + "step": 79675 + }, + { + "epoch": 5.4137790460660415, + "grad_norm": 1.002755880355835, + "learning_rate": 0.0003232776192417448, + "loss": 3.3442, + "step": 79680 + }, + { + "epoch": 5.414118766136704, + "grad_norm": 0.8229334354400635, + "learning_rate": 0.0003232351542329121, + "loss": 3.3488, + "step": 79685 + }, + { + "epoch": 5.414458486207365, + "grad_norm": 0.9102748036384583, + "learning_rate": 0.00032319268922407937, + "loss": 3.4106, + "step": 79690 + }, + { + "epoch": 5.414798206278027, + "grad_norm": 0.8709344267845154, + "learning_rate": 0.00032315022421524665, + "loss": 3.4867, + "step": 79695 + }, + { + "epoch": 5.415137926348689, + "grad_norm": 1.1115503311157227, + "learning_rate": 0.00032310775920641393, + "loss": 3.3963, + "step": 79700 + }, + { + "epoch": 5.41547764641935, + "grad_norm": 0.6895166039466858, + "learning_rate": 0.00032306529419758116, + "loss": 3.4755, + "step": 79705 + }, + { + "epoch": 5.415817366490012, + "grad_norm": 1.021589994430542, + "learning_rate": 0.0003230228291887485, + "loss": 3.4243, + "step": 79710 + }, + { + "epoch": 5.416157086560674, + "grad_norm": 1.0303107500076294, + "learning_rate": 0.00032298036417991577, + "loss": 3.427, + "step": 79715 + }, + { + "epoch": 5.4164968066313355, + "grad_norm": 0.9345182776451111, + "learning_rate": 0.000322937899171083, + "loss": 3.433, + "step": 79720 + }, + { + "epoch": 5.416836526701998, + "grad_norm": 0.9277019500732422, + "learning_rate": 0.00032289543416225033, + "loss": 3.4604, + "step": 79725 + }, + { + "epoch": 5.41717624677266, + "grad_norm": 0.8564796447753906, + "learning_rate": 0.0003228529691534176, + "loss": 3.6353, + "step": 79730 + }, + { + "epoch": 5.417515966843321, + "grad_norm": 0.8231950998306274, + "learning_rate": 0.00032281050414458484, + "loss": 3.3832, + "step": 79735 + }, + { + "epoch": 5.417855686913983, + "grad_norm": 1.0379101037979126, + "learning_rate": 0.00032276803913575217, + "loss": 3.3708, + "step": 79740 + }, + { + "epoch": 5.418195406984645, + "grad_norm": 1.0698096752166748, + "learning_rate": 0.00032272557412691945, + "loss": 3.3811, + "step": 79745 + }, + { + "epoch": 5.418535127055306, + "grad_norm": 1.1276267766952515, + "learning_rate": 0.0003226831091180867, + "loss": 3.4829, + "step": 79750 + }, + { + "epoch": 5.418874847125968, + "grad_norm": 1.2868626117706299, + "learning_rate": 0.00032264064410925396, + "loss": 3.641, + "step": 79755 + }, + { + "epoch": 5.41921456719663, + "grad_norm": 1.095262050628662, + "learning_rate": 0.0003225981791004213, + "loss": 3.1516, + "step": 79760 + }, + { + "epoch": 5.4195542872672915, + "grad_norm": 0.9112256169319153, + "learning_rate": 0.0003225557140915885, + "loss": 3.3517, + "step": 79765 + }, + { + "epoch": 5.419894007337954, + "grad_norm": 0.9246969223022461, + "learning_rate": 0.0003225132490827558, + "loss": 3.3469, + "step": 79770 + }, + { + "epoch": 5.420233727408616, + "grad_norm": 0.829845130443573, + "learning_rate": 0.00032247078407392313, + "loss": 3.5544, + "step": 79775 + }, + { + "epoch": 5.420573447479277, + "grad_norm": 0.8716671466827393, + "learning_rate": 0.00032242831906509036, + "loss": 3.6626, + "step": 79780 + }, + { + "epoch": 5.420913167549939, + "grad_norm": 0.8169354796409607, + "learning_rate": 0.00032238585405625764, + "loss": 3.5782, + "step": 79785 + }, + { + "epoch": 5.421252887620601, + "grad_norm": 1.022579312324524, + "learning_rate": 0.0003223433890474249, + "loss": 3.4576, + "step": 79790 + }, + { + "epoch": 5.421592607691262, + "grad_norm": 0.9283916354179382, + "learning_rate": 0.0003223009240385922, + "loss": 3.2583, + "step": 79795 + }, + { + "epoch": 5.421932327761924, + "grad_norm": 0.9453358054161072, + "learning_rate": 0.0003222584590297595, + "loss": 3.2979, + "step": 79800 + }, + { + "epoch": 5.422272047832586, + "grad_norm": 0.8371495604515076, + "learning_rate": 0.00032221599402092676, + "loss": 3.2128, + "step": 79805 + }, + { + "epoch": 5.4226117679032475, + "grad_norm": 1.037532091140747, + "learning_rate": 0.0003221735290120941, + "loss": 3.3836, + "step": 79810 + }, + { + "epoch": 5.42295148797391, + "grad_norm": 1.3357410430908203, + "learning_rate": 0.0003221310640032613, + "loss": 3.6313, + "step": 79815 + }, + { + "epoch": 5.423291208044572, + "grad_norm": 0.8828228116035461, + "learning_rate": 0.0003220885989944286, + "loss": 3.4883, + "step": 79820 + }, + { + "epoch": 5.423630928115233, + "grad_norm": 0.9556611776351929, + "learning_rate": 0.0003220461339855959, + "loss": 3.5646, + "step": 79825 + }, + { + "epoch": 5.423970648185895, + "grad_norm": 0.9585297703742981, + "learning_rate": 0.00032200366897676316, + "loss": 3.5546, + "step": 79830 + }, + { + "epoch": 5.424310368256557, + "grad_norm": 0.809064507484436, + "learning_rate": 0.00032196120396793044, + "loss": 3.6639, + "step": 79835 + }, + { + "epoch": 5.424650088327218, + "grad_norm": 0.916098415851593, + "learning_rate": 0.0003219187389590977, + "loss": 3.5239, + "step": 79840 + }, + { + "epoch": 5.42498980839788, + "grad_norm": 0.8454807996749878, + "learning_rate": 0.000321876273950265, + "loss": 3.5873, + "step": 79845 + }, + { + "epoch": 5.425329528468542, + "grad_norm": 0.7013819217681885, + "learning_rate": 0.0003218338089414323, + "loss": 3.4357, + "step": 79850 + }, + { + "epoch": 5.4256692485392035, + "grad_norm": 0.8472039699554443, + "learning_rate": 0.00032179134393259956, + "loss": 3.5815, + "step": 79855 + }, + { + "epoch": 5.426008968609866, + "grad_norm": 1.1921658515930176, + "learning_rate": 0.0003217488789237668, + "loss": 3.3115, + "step": 79860 + }, + { + "epoch": 5.426348688680527, + "grad_norm": 0.9890744090080261, + "learning_rate": 0.0003217064139149341, + "loss": 3.5892, + "step": 79865 + }, + { + "epoch": 5.426688408751189, + "grad_norm": 0.8726601600646973, + "learning_rate": 0.0003216639489061014, + "loss": 3.1432, + "step": 79870 + }, + { + "epoch": 5.427028128821851, + "grad_norm": 1.2211672067642212, + "learning_rate": 0.0003216214838972686, + "loss": 3.3863, + "step": 79875 + }, + { + "epoch": 5.427367848892512, + "grad_norm": 0.6898414492607117, + "learning_rate": 0.00032157901888843596, + "loss": 3.2701, + "step": 79880 + }, + { + "epoch": 5.427707568963174, + "grad_norm": 0.7372936010360718, + "learning_rate": 0.00032153655387960324, + "loss": 3.1812, + "step": 79885 + }, + { + "epoch": 5.428047289033836, + "grad_norm": 0.896615743637085, + "learning_rate": 0.00032149408887077047, + "loss": 3.5661, + "step": 79890 + }, + { + "epoch": 5.4283870091044975, + "grad_norm": 0.754237711429596, + "learning_rate": 0.00032145162386193775, + "loss": 3.6133, + "step": 79895 + }, + { + "epoch": 5.4287267291751595, + "grad_norm": 1.1617298126220703, + "learning_rate": 0.0003214091588531051, + "loss": 3.5067, + "step": 79900 + }, + { + "epoch": 5.429066449245822, + "grad_norm": 0.8100389242172241, + "learning_rate": 0.0003213666938442723, + "loss": 3.4381, + "step": 79905 + }, + { + "epoch": 5.429406169316483, + "grad_norm": 0.9084051251411438, + "learning_rate": 0.0003213242288354396, + "loss": 3.586, + "step": 79910 + }, + { + "epoch": 5.429745889387145, + "grad_norm": 0.9597190022468567, + "learning_rate": 0.0003212817638266069, + "loss": 3.5202, + "step": 79915 + }, + { + "epoch": 5.430085609457807, + "grad_norm": 1.2846983671188354, + "learning_rate": 0.00032123929881777415, + "loss": 2.9802, + "step": 79920 + }, + { + "epoch": 5.430425329528468, + "grad_norm": 0.6910883784294128, + "learning_rate": 0.0003211968338089414, + "loss": 3.4769, + "step": 79925 + }, + { + "epoch": 5.43076504959913, + "grad_norm": 0.876245379447937, + "learning_rate": 0.0003211543688001087, + "loss": 3.2965, + "step": 79930 + }, + { + "epoch": 5.431104769669792, + "grad_norm": 0.9904869794845581, + "learning_rate": 0.000321111903791276, + "loss": 3.0782, + "step": 79935 + }, + { + "epoch": 5.4314444897404535, + "grad_norm": 0.8907330632209778, + "learning_rate": 0.00032106943878244327, + "loss": 3.4998, + "step": 79940 + }, + { + "epoch": 5.4317842098111155, + "grad_norm": 0.8442257642745972, + "learning_rate": 0.00032102697377361055, + "loss": 3.506, + "step": 79945 + }, + { + "epoch": 5.432123929881778, + "grad_norm": 0.759250819683075, + "learning_rate": 0.0003209845087647778, + "loss": 3.2729, + "step": 79950 + }, + { + "epoch": 5.432463649952439, + "grad_norm": 0.9117350578308105, + "learning_rate": 0.0003209420437559451, + "loss": 3.3955, + "step": 79955 + }, + { + "epoch": 5.432803370023101, + "grad_norm": 1.2261059284210205, + "learning_rate": 0.0003208995787471124, + "loss": 3.4951, + "step": 79960 + }, + { + "epoch": 5.433143090093763, + "grad_norm": 1.1697813272476196, + "learning_rate": 0.0003208571137382796, + "loss": 3.489, + "step": 79965 + }, + { + "epoch": 5.433482810164424, + "grad_norm": 0.8588851690292358, + "learning_rate": 0.00032081464872944695, + "loss": 3.5286, + "step": 79970 + }, + { + "epoch": 5.433822530235086, + "grad_norm": 0.8793575167655945, + "learning_rate": 0.0003207721837206142, + "loss": 3.339, + "step": 79975 + }, + { + "epoch": 5.434162250305748, + "grad_norm": 0.9356627464294434, + "learning_rate": 0.0003207297187117815, + "loss": 3.2649, + "step": 79980 + }, + { + "epoch": 5.4345019703764095, + "grad_norm": 0.9652442336082458, + "learning_rate": 0.0003206872537029488, + "loss": 3.3139, + "step": 79985 + }, + { + "epoch": 5.4348416904470715, + "grad_norm": 0.8223676681518555, + "learning_rate": 0.00032064478869411607, + "loss": 3.6272, + "step": 79990 + }, + { + "epoch": 5.435181410517734, + "grad_norm": 0.7432926893234253, + "learning_rate": 0.00032060232368528335, + "loss": 3.303, + "step": 79995 + }, + { + "epoch": 5.435521130588395, + "grad_norm": 0.8935067057609558, + "learning_rate": 0.0003205598586764506, + "loss": 3.4762, + "step": 80000 + }, + { + "epoch": 5.435860850659057, + "grad_norm": 0.8305067420005798, + "learning_rate": 0.0003205173936676179, + "loss": 3.5124, + "step": 80005 + }, + { + "epoch": 5.436200570729719, + "grad_norm": 0.9069252610206604, + "learning_rate": 0.0003204749286587852, + "loss": 3.4791, + "step": 80010 + }, + { + "epoch": 5.43654029080038, + "grad_norm": 0.8048125505447388, + "learning_rate": 0.0003204324636499524, + "loss": 3.4871, + "step": 80015 + }, + { + "epoch": 5.436880010871042, + "grad_norm": 0.9554992914199829, + "learning_rate": 0.00032038999864111975, + "loss": 3.6489, + "step": 80020 + }, + { + "epoch": 5.437219730941704, + "grad_norm": 0.8812788128852844, + "learning_rate": 0.00032034753363228703, + "loss": 3.2659, + "step": 80025 + }, + { + "epoch": 5.4375594510123655, + "grad_norm": 1.128759503364563, + "learning_rate": 0.00032030506862345425, + "loss": 3.7024, + "step": 80030 + }, + { + "epoch": 5.437899171083028, + "grad_norm": 1.1227883100509644, + "learning_rate": 0.0003202626036146216, + "loss": 3.557, + "step": 80035 + }, + { + "epoch": 5.43823889115369, + "grad_norm": 0.7424842715263367, + "learning_rate": 0.00032022013860578887, + "loss": 3.4154, + "step": 80040 + }, + { + "epoch": 5.438578611224351, + "grad_norm": 0.7760342359542847, + "learning_rate": 0.0003201776735969561, + "loss": 3.3055, + "step": 80045 + }, + { + "epoch": 5.438918331295013, + "grad_norm": 0.9621506333351135, + "learning_rate": 0.0003201352085881234, + "loss": 3.5439, + "step": 80050 + }, + { + "epoch": 5.439258051365675, + "grad_norm": 0.9321199655532837, + "learning_rate": 0.0003200927435792907, + "loss": 3.5214, + "step": 80055 + }, + { + "epoch": 5.439597771436336, + "grad_norm": 0.7454168796539307, + "learning_rate": 0.00032005027857045793, + "loss": 3.3582, + "step": 80060 + }, + { + "epoch": 5.439937491506998, + "grad_norm": 1.0646920204162598, + "learning_rate": 0.0003200078135616252, + "loss": 3.2409, + "step": 80065 + }, + { + "epoch": 5.44027721157766, + "grad_norm": 0.9061108231544495, + "learning_rate": 0.00031996534855279255, + "loss": 3.2801, + "step": 80070 + }, + { + "epoch": 5.4406169316483215, + "grad_norm": 0.8825212121009827, + "learning_rate": 0.0003199228835439598, + "loss": 3.2948, + "step": 80075 + }, + { + "epoch": 5.440956651718984, + "grad_norm": 0.9839466214179993, + "learning_rate": 0.00031988041853512705, + "loss": 3.5077, + "step": 80080 + }, + { + "epoch": 5.441296371789646, + "grad_norm": 0.7808550000190735, + "learning_rate": 0.00031983795352629433, + "loss": 3.6127, + "step": 80085 + }, + { + "epoch": 5.441636091860307, + "grad_norm": 0.8348985910415649, + "learning_rate": 0.0003197954885174616, + "loss": 3.1995, + "step": 80090 + }, + { + "epoch": 5.441975811930969, + "grad_norm": 1.035544514656067, + "learning_rate": 0.0003197530235086289, + "loss": 3.4584, + "step": 80095 + }, + { + "epoch": 5.442315532001631, + "grad_norm": 0.9968258142471313, + "learning_rate": 0.0003197105584997962, + "loss": 3.4169, + "step": 80100 + }, + { + "epoch": 5.442655252072292, + "grad_norm": 0.7317793965339661, + "learning_rate": 0.00031966809349096345, + "loss": 3.415, + "step": 80105 + }, + { + "epoch": 5.442994972142954, + "grad_norm": 1.0015166997909546, + "learning_rate": 0.00031962562848213073, + "loss": 3.4404, + "step": 80110 + }, + { + "epoch": 5.443334692213616, + "grad_norm": 0.7679945230484009, + "learning_rate": 0.000319583163473298, + "loss": 3.5983, + "step": 80115 + }, + { + "epoch": 5.4436744122842775, + "grad_norm": 0.966122031211853, + "learning_rate": 0.00031954069846446524, + "loss": 3.4565, + "step": 80120 + }, + { + "epoch": 5.44401413235494, + "grad_norm": 0.8979265689849854, + "learning_rate": 0.0003194982334556326, + "loss": 3.2175, + "step": 80125 + }, + { + "epoch": 5.444353852425602, + "grad_norm": 0.7221465110778809, + "learning_rate": 0.00031945576844679986, + "loss": 3.6694, + "step": 80130 + }, + { + "epoch": 5.444693572496263, + "grad_norm": 0.9527706503868103, + "learning_rate": 0.0003194133034379671, + "loss": 3.4806, + "step": 80135 + }, + { + "epoch": 5.445033292566925, + "grad_norm": 1.1826707124710083, + "learning_rate": 0.0003193708384291344, + "loss": 3.7939, + "step": 80140 + }, + { + "epoch": 5.445373012637587, + "grad_norm": 1.6172720193862915, + "learning_rate": 0.0003193283734203017, + "loss": 3.6342, + "step": 80145 + }, + { + "epoch": 5.445712732708248, + "grad_norm": 0.8139119744300842, + "learning_rate": 0.000319285908411469, + "loss": 3.4507, + "step": 80150 + }, + { + "epoch": 5.44605245277891, + "grad_norm": 0.8033959269523621, + "learning_rate": 0.0003192434434026362, + "loss": 3.3415, + "step": 80155 + }, + { + "epoch": 5.446392172849572, + "grad_norm": 0.8399551510810852, + "learning_rate": 0.00031920097839380354, + "loss": 3.3689, + "step": 80160 + }, + { + "epoch": 5.4467318929202335, + "grad_norm": 0.6771608591079712, + "learning_rate": 0.0003191585133849708, + "loss": 3.648, + "step": 80165 + }, + { + "epoch": 5.447071612990896, + "grad_norm": 0.9032678008079529, + "learning_rate": 0.00031911604837613804, + "loss": 3.5641, + "step": 80170 + }, + { + "epoch": 5.447411333061558, + "grad_norm": 1.138333797454834, + "learning_rate": 0.0003190735833673054, + "loss": 3.5364, + "step": 80175 + }, + { + "epoch": 5.447751053132219, + "grad_norm": 1.1148674488067627, + "learning_rate": 0.00031903111835847266, + "loss": 3.3039, + "step": 80180 + }, + { + "epoch": 5.448090773202881, + "grad_norm": 0.8793702721595764, + "learning_rate": 0.0003189886533496399, + "loss": 3.566, + "step": 80185 + }, + { + "epoch": 5.448430493273543, + "grad_norm": 0.7411450743675232, + "learning_rate": 0.00031894618834080716, + "loss": 3.5402, + "step": 80190 + }, + { + "epoch": 5.448770213344204, + "grad_norm": 0.9839465618133545, + "learning_rate": 0.0003189037233319745, + "loss": 3.4557, + "step": 80195 + }, + { + "epoch": 5.449109933414866, + "grad_norm": 0.9993599653244019, + "learning_rate": 0.0003188612583231417, + "loss": 3.4291, + "step": 80200 + }, + { + "epoch": 5.449449653485528, + "grad_norm": 0.860223114490509, + "learning_rate": 0.000318818793314309, + "loss": 3.413, + "step": 80205 + }, + { + "epoch": 5.4497893735561895, + "grad_norm": 1.0983389616012573, + "learning_rate": 0.00031877632830547634, + "loss": 3.2946, + "step": 80210 + }, + { + "epoch": 5.450129093626852, + "grad_norm": 0.9472988247871399, + "learning_rate": 0.00031873386329664356, + "loss": 3.7485, + "step": 80215 + }, + { + "epoch": 5.450468813697514, + "grad_norm": 0.8879375457763672, + "learning_rate": 0.00031869139828781084, + "loss": 3.3158, + "step": 80220 + }, + { + "epoch": 5.450808533768175, + "grad_norm": 1.0611835718154907, + "learning_rate": 0.0003186489332789781, + "loss": 3.3014, + "step": 80225 + }, + { + "epoch": 5.451148253838837, + "grad_norm": 1.0114688873291016, + "learning_rate": 0.0003186064682701454, + "loss": 3.4739, + "step": 80230 + }, + { + "epoch": 5.451487973909499, + "grad_norm": 0.8104035258293152, + "learning_rate": 0.0003185640032613127, + "loss": 3.5576, + "step": 80235 + }, + { + "epoch": 5.45182769398016, + "grad_norm": 0.7817025780677795, + "learning_rate": 0.00031852153825247996, + "loss": 3.3327, + "step": 80240 + }, + { + "epoch": 5.452167414050822, + "grad_norm": 0.8456599116325378, + "learning_rate": 0.00031847907324364724, + "loss": 3.3876, + "step": 80245 + }, + { + "epoch": 5.452507134121484, + "grad_norm": 0.9094346761703491, + "learning_rate": 0.0003184366082348145, + "loss": 3.2931, + "step": 80250 + }, + { + "epoch": 5.4528468541921455, + "grad_norm": 0.7786800861358643, + "learning_rate": 0.0003183941432259818, + "loss": 3.7141, + "step": 80255 + }, + { + "epoch": 5.453186574262808, + "grad_norm": 1.046402096748352, + "learning_rate": 0.00031835167821714903, + "loss": 3.4595, + "step": 80260 + }, + { + "epoch": 5.45352629433347, + "grad_norm": 1.0273611545562744, + "learning_rate": 0.00031830921320831636, + "loss": 3.4208, + "step": 80265 + }, + { + "epoch": 5.453866014404131, + "grad_norm": 0.9352371692657471, + "learning_rate": 0.00031826674819948364, + "loss": 3.4424, + "step": 80270 + }, + { + "epoch": 5.454205734474793, + "grad_norm": 0.9333164691925049, + "learning_rate": 0.00031822428319065087, + "loss": 3.3087, + "step": 80275 + }, + { + "epoch": 5.454545454545454, + "grad_norm": 1.1025381088256836, + "learning_rate": 0.0003181818181818182, + "loss": 3.3989, + "step": 80280 + }, + { + "epoch": 5.454885174616116, + "grad_norm": 1.0340526103973389, + "learning_rate": 0.0003181393531729855, + "loss": 3.4063, + "step": 80285 + }, + { + "epoch": 5.455224894686778, + "grad_norm": 1.2037047147750854, + "learning_rate": 0.0003180968881641527, + "loss": 3.2613, + "step": 80290 + }, + { + "epoch": 5.4555646147574395, + "grad_norm": 0.8071075081825256, + "learning_rate": 0.00031805442315532, + "loss": 3.6881, + "step": 80295 + }, + { + "epoch": 5.4559043348281016, + "grad_norm": 0.8655932545661926, + "learning_rate": 0.0003180119581464873, + "loss": 3.6935, + "step": 80300 + }, + { + "epoch": 5.456244054898764, + "grad_norm": 1.0303689241409302, + "learning_rate": 0.00031796949313765455, + "loss": 3.463, + "step": 80305 + }, + { + "epoch": 5.456583774969425, + "grad_norm": 0.8138473033905029, + "learning_rate": 0.00031792702812882183, + "loss": 3.4379, + "step": 80310 + }, + { + "epoch": 5.456923495040087, + "grad_norm": 0.9940857291221619, + "learning_rate": 0.00031788456311998916, + "loss": 3.3135, + "step": 80315 + }, + { + "epoch": 5.457263215110749, + "grad_norm": 0.9272728562355042, + "learning_rate": 0.00031784209811115644, + "loss": 3.2598, + "step": 80320 + }, + { + "epoch": 5.45760293518141, + "grad_norm": 0.9566553831100464, + "learning_rate": 0.00031779963310232367, + "loss": 3.4085, + "step": 80325 + }, + { + "epoch": 5.457942655252072, + "grad_norm": 0.8881889581680298, + "learning_rate": 0.000317757168093491, + "loss": 3.7319, + "step": 80330 + }, + { + "epoch": 5.458282375322734, + "grad_norm": 0.7367042899131775, + "learning_rate": 0.0003177147030846583, + "loss": 3.5754, + "step": 80335 + }, + { + "epoch": 5.4586220953933955, + "grad_norm": 0.9288102984428406, + "learning_rate": 0.0003176722380758255, + "loss": 3.3549, + "step": 80340 + }, + { + "epoch": 5.458961815464058, + "grad_norm": 1.0950753688812256, + "learning_rate": 0.0003176297730669928, + "loss": 3.5794, + "step": 80345 + }, + { + "epoch": 5.45930153553472, + "grad_norm": 0.8329585790634155, + "learning_rate": 0.0003175873080581601, + "loss": 3.3791, + "step": 80350 + }, + { + "epoch": 5.459641255605381, + "grad_norm": 0.9043397307395935, + "learning_rate": 0.00031754484304932735, + "loss": 3.5061, + "step": 80355 + }, + { + "epoch": 5.459980975676043, + "grad_norm": 1.1477543115615845, + "learning_rate": 0.00031750237804049463, + "loss": 3.204, + "step": 80360 + }, + { + "epoch": 5.460320695746705, + "grad_norm": 0.9725548028945923, + "learning_rate": 0.00031745991303166196, + "loss": 3.5959, + "step": 80365 + }, + { + "epoch": 5.460660415817366, + "grad_norm": 1.1579054594039917, + "learning_rate": 0.0003174174480228292, + "loss": 3.3758, + "step": 80370 + }, + { + "epoch": 5.461000135888028, + "grad_norm": 0.8023281097412109, + "learning_rate": 0.00031737498301399647, + "loss": 3.5489, + "step": 80375 + }, + { + "epoch": 5.46133985595869, + "grad_norm": 0.9975463151931763, + "learning_rate": 0.00031733251800516375, + "loss": 3.5338, + "step": 80380 + }, + { + "epoch": 5.4616795760293515, + "grad_norm": 0.7677999138832092, + "learning_rate": 0.00031729005299633103, + "loss": 3.4013, + "step": 80385 + }, + { + "epoch": 5.462019296100014, + "grad_norm": 0.9919224381446838, + "learning_rate": 0.0003172475879874983, + "loss": 3.6213, + "step": 80390 + }, + { + "epoch": 5.462359016170676, + "grad_norm": 0.8394951820373535, + "learning_rate": 0.0003172051229786656, + "loss": 3.5269, + "step": 80395 + }, + { + "epoch": 5.462698736241337, + "grad_norm": 0.8577165603637695, + "learning_rate": 0.00031716265796983287, + "loss": 3.43, + "step": 80400 + }, + { + "epoch": 5.463038456311999, + "grad_norm": 0.9205718040466309, + "learning_rate": 0.00031712019296100015, + "loss": 3.44, + "step": 80405 + }, + { + "epoch": 5.463378176382661, + "grad_norm": 0.8181172609329224, + "learning_rate": 0.00031707772795216743, + "loss": 3.5706, + "step": 80410 + }, + { + "epoch": 5.463717896453322, + "grad_norm": 0.9352506399154663, + "learning_rate": 0.00031703526294333466, + "loss": 3.644, + "step": 80415 + }, + { + "epoch": 5.464057616523984, + "grad_norm": 0.897623598575592, + "learning_rate": 0.000316992797934502, + "loss": 3.4525, + "step": 80420 + }, + { + "epoch": 5.464397336594646, + "grad_norm": 0.7987215518951416, + "learning_rate": 0.00031695033292566927, + "loss": 3.0445, + "step": 80425 + }, + { + "epoch": 5.4647370566653075, + "grad_norm": 1.0621827840805054, + "learning_rate": 0.0003169078679168365, + "loss": 3.5239, + "step": 80430 + }, + { + "epoch": 5.46507677673597, + "grad_norm": 0.7452366948127747, + "learning_rate": 0.00031686540290800383, + "loss": 3.5867, + "step": 80435 + }, + { + "epoch": 5.465416496806632, + "grad_norm": 0.8148128390312195, + "learning_rate": 0.0003168229378991711, + "loss": 3.5795, + "step": 80440 + }, + { + "epoch": 5.465756216877293, + "grad_norm": 0.9951159358024597, + "learning_rate": 0.00031678047289033834, + "loss": 3.7857, + "step": 80445 + }, + { + "epoch": 5.466095936947955, + "grad_norm": 0.9115222692489624, + "learning_rate": 0.0003167380078815056, + "loss": 3.7717, + "step": 80450 + }, + { + "epoch": 5.466435657018617, + "grad_norm": 0.7671780586242676, + "learning_rate": 0.00031669554287267295, + "loss": 3.4688, + "step": 80455 + }, + { + "epoch": 5.466775377089278, + "grad_norm": 0.8651946187019348, + "learning_rate": 0.0003166530778638402, + "loss": 3.5941, + "step": 80460 + }, + { + "epoch": 5.46711509715994, + "grad_norm": 1.0602905750274658, + "learning_rate": 0.00031661061285500746, + "loss": 3.7414, + "step": 80465 + }, + { + "epoch": 5.467454817230602, + "grad_norm": 1.0174369812011719, + "learning_rate": 0.0003165681478461748, + "loss": 3.3238, + "step": 80470 + }, + { + "epoch": 5.4677945373012635, + "grad_norm": 1.053489327430725, + "learning_rate": 0.000316525682837342, + "loss": 3.3682, + "step": 80475 + }, + { + "epoch": 5.468134257371926, + "grad_norm": 0.9659205675125122, + "learning_rate": 0.0003164832178285093, + "loss": 3.5042, + "step": 80480 + }, + { + "epoch": 5.468473977442588, + "grad_norm": 0.7990891337394714, + "learning_rate": 0.0003164407528196766, + "loss": 3.2517, + "step": 80485 + }, + { + "epoch": 5.468813697513249, + "grad_norm": 1.1702691316604614, + "learning_rate": 0.0003163982878108439, + "loss": 3.4422, + "step": 80490 + }, + { + "epoch": 5.469153417583911, + "grad_norm": 1.1447176933288574, + "learning_rate": 0.00031635582280201114, + "loss": 3.3085, + "step": 80495 + }, + { + "epoch": 5.469493137654573, + "grad_norm": 1.006929636001587, + "learning_rate": 0.0003163133577931784, + "loss": 3.2478, + "step": 80500 + }, + { + "epoch": 5.469832857725234, + "grad_norm": 0.8605318665504456, + "learning_rate": 0.00031627089278434575, + "loss": 3.5052, + "step": 80505 + }, + { + "epoch": 5.470172577795896, + "grad_norm": 1.0493879318237305, + "learning_rate": 0.000316228427775513, + "loss": 3.4422, + "step": 80510 + }, + { + "epoch": 5.470512297866558, + "grad_norm": 1.1290351152420044, + "learning_rate": 0.00031618596276668026, + "loss": 3.5213, + "step": 80515 + }, + { + "epoch": 5.4708520179372195, + "grad_norm": 1.5247526168823242, + "learning_rate": 0.00031614349775784754, + "loss": 3.2364, + "step": 80520 + }, + { + "epoch": 5.471191738007882, + "grad_norm": 1.0497881174087524, + "learning_rate": 0.0003161010327490148, + "loss": 3.3401, + "step": 80525 + }, + { + "epoch": 5.471531458078544, + "grad_norm": 0.7145054936408997, + "learning_rate": 0.0003160585677401821, + "loss": 3.4485, + "step": 80530 + }, + { + "epoch": 5.471871178149205, + "grad_norm": 0.7623755931854248, + "learning_rate": 0.0003160161027313494, + "loss": 3.2687, + "step": 80535 + }, + { + "epoch": 5.472210898219867, + "grad_norm": 1.2699828147888184, + "learning_rate": 0.00031597363772251666, + "loss": 3.4635, + "step": 80540 + }, + { + "epoch": 5.472550618290528, + "grad_norm": 1.2134913206100464, + "learning_rate": 0.00031593117271368394, + "loss": 3.5687, + "step": 80545 + }, + { + "epoch": 5.47289033836119, + "grad_norm": 0.8634642958641052, + "learning_rate": 0.0003158887077048512, + "loss": 3.4872, + "step": 80550 + }, + { + "epoch": 5.473230058431852, + "grad_norm": 0.7784424424171448, + "learning_rate": 0.00031584624269601844, + "loss": 3.4812, + "step": 80555 + }, + { + "epoch": 5.4735697785025135, + "grad_norm": 0.9676445722579956, + "learning_rate": 0.0003158037776871858, + "loss": 3.4027, + "step": 80560 + }, + { + "epoch": 5.4739094985731755, + "grad_norm": 1.1016380786895752, + "learning_rate": 0.00031576131267835306, + "loss": 3.3296, + "step": 80565 + }, + { + "epoch": 5.474249218643838, + "grad_norm": 0.7825928926467896, + "learning_rate": 0.0003157188476695203, + "loss": 3.5501, + "step": 80570 + }, + { + "epoch": 5.474588938714499, + "grad_norm": 0.8966153264045715, + "learning_rate": 0.0003156763826606876, + "loss": 3.4232, + "step": 80575 + }, + { + "epoch": 5.474928658785161, + "grad_norm": 0.9040058851242065, + "learning_rate": 0.0003156339176518549, + "loss": 3.6934, + "step": 80580 + }, + { + "epoch": 5.475268378855823, + "grad_norm": 0.9499273896217346, + "learning_rate": 0.0003155914526430221, + "loss": 3.6648, + "step": 80585 + }, + { + "epoch": 5.475608098926484, + "grad_norm": 0.8520113825798035, + "learning_rate": 0.00031554898763418946, + "loss": 3.6436, + "step": 80590 + }, + { + "epoch": 5.475947818997146, + "grad_norm": 1.1939960718154907, + "learning_rate": 0.00031550652262535674, + "loss": 3.5838, + "step": 80595 + }, + { + "epoch": 5.476287539067808, + "grad_norm": 0.7153595089912415, + "learning_rate": 0.00031546405761652397, + "loss": 3.6443, + "step": 80600 + }, + { + "epoch": 5.4766272591384695, + "grad_norm": 0.9678404331207275, + "learning_rate": 0.00031542159260769125, + "loss": 3.2944, + "step": 80605 + }, + { + "epoch": 5.476966979209132, + "grad_norm": 0.8010450601577759, + "learning_rate": 0.0003153791275988586, + "loss": 3.3811, + "step": 80610 + }, + { + "epoch": 5.477306699279794, + "grad_norm": 1.4086612462997437, + "learning_rate": 0.0003153366625900258, + "loss": 3.6514, + "step": 80615 + }, + { + "epoch": 5.477646419350455, + "grad_norm": 0.762890636920929, + "learning_rate": 0.0003152941975811931, + "loss": 3.715, + "step": 80620 + }, + { + "epoch": 5.477986139421117, + "grad_norm": 0.8555885553359985, + "learning_rate": 0.0003152517325723604, + "loss": 3.7097, + "step": 80625 + }, + { + "epoch": 5.478325859491779, + "grad_norm": 0.8632957935333252, + "learning_rate": 0.00031520926756352765, + "loss": 3.3016, + "step": 80630 + }, + { + "epoch": 5.47866557956244, + "grad_norm": 0.8068735003471375, + "learning_rate": 0.0003151668025546949, + "loss": 3.3411, + "step": 80635 + }, + { + "epoch": 5.479005299633102, + "grad_norm": 0.9254390001296997, + "learning_rate": 0.0003151243375458622, + "loss": 3.6879, + "step": 80640 + }, + { + "epoch": 5.479345019703764, + "grad_norm": 0.9061561226844788, + "learning_rate": 0.0003150818725370295, + "loss": 3.4157, + "step": 80645 + }, + { + "epoch": 5.4796847397744255, + "grad_norm": 0.7267497777938843, + "learning_rate": 0.00031503940752819677, + "loss": 3.5233, + "step": 80650 + }, + { + "epoch": 5.480024459845088, + "grad_norm": 1.007405161857605, + "learning_rate": 0.00031499694251936405, + "loss": 3.547, + "step": 80655 + }, + { + "epoch": 5.48036417991575, + "grad_norm": 0.83275306224823, + "learning_rate": 0.0003149544775105314, + "loss": 3.3576, + "step": 80660 + }, + { + "epoch": 5.480703899986411, + "grad_norm": 1.0754060745239258, + "learning_rate": 0.0003149120125016986, + "loss": 3.3933, + "step": 80665 + }, + { + "epoch": 5.481043620057073, + "grad_norm": 0.7492450475692749, + "learning_rate": 0.0003148695474928659, + "loss": 3.4898, + "step": 80670 + }, + { + "epoch": 5.481383340127735, + "grad_norm": 1.040816307067871, + "learning_rate": 0.00031482708248403317, + "loss": 3.6259, + "step": 80675 + }, + { + "epoch": 5.481723060198396, + "grad_norm": 0.8718187808990479, + "learning_rate": 0.00031478461747520045, + "loss": 3.4567, + "step": 80680 + }, + { + "epoch": 5.482062780269058, + "grad_norm": 0.962226390838623, + "learning_rate": 0.0003147421524663677, + "loss": 3.5857, + "step": 80685 + }, + { + "epoch": 5.48240250033972, + "grad_norm": 0.827300488948822, + "learning_rate": 0.000314699687457535, + "loss": 3.266, + "step": 80690 + }, + { + "epoch": 5.4827422204103815, + "grad_norm": 1.0786817073822021, + "learning_rate": 0.0003146572224487023, + "loss": 3.6099, + "step": 80695 + }, + { + "epoch": 5.483081940481044, + "grad_norm": 0.7140036821365356, + "learning_rate": 0.00031461475743986957, + "loss": 3.5164, + "step": 80700 + }, + { + "epoch": 5.483421660551706, + "grad_norm": 0.6709192395210266, + "learning_rate": 0.00031457229243103685, + "loss": 3.2747, + "step": 80705 + }, + { + "epoch": 5.483761380622367, + "grad_norm": 0.7061362266540527, + "learning_rate": 0.0003145298274222041, + "loss": 3.5543, + "step": 80710 + }, + { + "epoch": 5.484101100693029, + "grad_norm": 0.8526470065116882, + "learning_rate": 0.0003144873624133714, + "loss": 3.4602, + "step": 80715 + }, + { + "epoch": 5.484440820763691, + "grad_norm": 0.9458170533180237, + "learning_rate": 0.0003144448974045387, + "loss": 3.6625, + "step": 80720 + }, + { + "epoch": 5.484780540834352, + "grad_norm": 0.7395976781845093, + "learning_rate": 0.0003144024323957059, + "loss": 3.4018, + "step": 80725 + }, + { + "epoch": 5.485120260905014, + "grad_norm": 0.7012442350387573, + "learning_rate": 0.00031435996738687325, + "loss": 3.7142, + "step": 80730 + }, + { + "epoch": 5.485459980975676, + "grad_norm": 1.1292365789413452, + "learning_rate": 0.00031431750237804053, + "loss": 3.5992, + "step": 80735 + }, + { + "epoch": 5.4857997010463375, + "grad_norm": 0.8322694301605225, + "learning_rate": 0.00031427503736920775, + "loss": 3.3813, + "step": 80740 + }, + { + "epoch": 5.486139421117, + "grad_norm": 0.7894576191902161, + "learning_rate": 0.00031423257236037503, + "loss": 3.2208, + "step": 80745 + }, + { + "epoch": 5.486479141187662, + "grad_norm": 0.8858613967895508, + "learning_rate": 0.00031419010735154237, + "loss": 3.6104, + "step": 80750 + }, + { + "epoch": 5.486818861258323, + "grad_norm": 0.8454750776290894, + "learning_rate": 0.0003141476423427096, + "loss": 3.4152, + "step": 80755 + }, + { + "epoch": 5.487158581328985, + "grad_norm": 0.7485643029212952, + "learning_rate": 0.0003141051773338769, + "loss": 3.2746, + "step": 80760 + }, + { + "epoch": 5.487498301399647, + "grad_norm": 0.9465708136558533, + "learning_rate": 0.0003140627123250442, + "loss": 3.3082, + "step": 80765 + }, + { + "epoch": 5.487838021470308, + "grad_norm": 0.8312684893608093, + "learning_rate": 0.00031402024731621143, + "loss": 3.6827, + "step": 80770 + }, + { + "epoch": 5.48817774154097, + "grad_norm": 0.825872540473938, + "learning_rate": 0.0003139777823073787, + "loss": 3.0648, + "step": 80775 + }, + { + "epoch": 5.488517461611632, + "grad_norm": 1.0275200605392456, + "learning_rate": 0.000313935317298546, + "loss": 3.3599, + "step": 80780 + }, + { + "epoch": 5.4888571816822935, + "grad_norm": 0.9383062124252319, + "learning_rate": 0.0003138928522897133, + "loss": 3.3585, + "step": 80785 + }, + { + "epoch": 5.489196901752956, + "grad_norm": 0.8302780985832214, + "learning_rate": 0.00031385038728088055, + "loss": 3.4288, + "step": 80790 + }, + { + "epoch": 5.489536621823618, + "grad_norm": 0.7149053812026978, + "learning_rate": 0.00031380792227204783, + "loss": 3.509, + "step": 80795 + }, + { + "epoch": 5.489876341894279, + "grad_norm": 0.7897120118141174, + "learning_rate": 0.0003137654572632151, + "loss": 3.2316, + "step": 80800 + }, + { + "epoch": 5.490216061964941, + "grad_norm": 0.8637712001800537, + "learning_rate": 0.0003137229922543824, + "loss": 3.5196, + "step": 80805 + }, + { + "epoch": 5.490555782035603, + "grad_norm": 1.3040629625320435, + "learning_rate": 0.0003136805272455497, + "loss": 3.6338, + "step": 80810 + }, + { + "epoch": 5.490895502106264, + "grad_norm": 0.8398987650871277, + "learning_rate": 0.0003136380622367169, + "loss": 3.5479, + "step": 80815 + }, + { + "epoch": 5.491235222176926, + "grad_norm": 0.9866546988487244, + "learning_rate": 0.00031359559722788423, + "loss": 3.2831, + "step": 80820 + }, + { + "epoch": 5.491574942247588, + "grad_norm": 0.9165909290313721, + "learning_rate": 0.0003135531322190515, + "loss": 3.4444, + "step": 80825 + }, + { + "epoch": 5.4919146623182495, + "grad_norm": 0.9387881755828857, + "learning_rate": 0.0003135106672102188, + "loss": 3.4661, + "step": 80830 + }, + { + "epoch": 5.492254382388912, + "grad_norm": 0.8058584332466125, + "learning_rate": 0.0003134682022013861, + "loss": 3.4278, + "step": 80835 + }, + { + "epoch": 5.492594102459574, + "grad_norm": 0.7975155115127563, + "learning_rate": 0.00031342573719255336, + "loss": 3.6128, + "step": 80840 + }, + { + "epoch": 5.492933822530235, + "grad_norm": 1.0221580266952515, + "learning_rate": 0.00031338327218372064, + "loss": 3.177, + "step": 80845 + }, + { + "epoch": 5.493273542600897, + "grad_norm": 0.8448030352592468, + "learning_rate": 0.00031334080717488786, + "loss": 3.3803, + "step": 80850 + }, + { + "epoch": 5.493613262671559, + "grad_norm": 0.9093098044395447, + "learning_rate": 0.0003132983421660552, + "loss": 3.3047, + "step": 80855 + }, + { + "epoch": 5.49395298274222, + "grad_norm": 0.7240036129951477, + "learning_rate": 0.0003132558771572225, + "loss": 3.7104, + "step": 80860 + }, + { + "epoch": 5.494292702812882, + "grad_norm": 0.8219504356384277, + "learning_rate": 0.0003132134121483897, + "loss": 3.3929, + "step": 80865 + }, + { + "epoch": 5.494632422883544, + "grad_norm": 0.91567462682724, + "learning_rate": 0.00031317094713955704, + "loss": 3.7203, + "step": 80870 + }, + { + "epoch": 5.4949721429542056, + "grad_norm": 0.9021643996238708, + "learning_rate": 0.0003131284821307243, + "loss": 3.5359, + "step": 80875 + }, + { + "epoch": 5.495311863024868, + "grad_norm": 0.9679797291755676, + "learning_rate": 0.00031308601712189154, + "loss": 3.6355, + "step": 80880 + }, + { + "epoch": 5.49565158309553, + "grad_norm": 0.8407672047615051, + "learning_rate": 0.0003130435521130589, + "loss": 3.2346, + "step": 80885 + }, + { + "epoch": 5.495991303166191, + "grad_norm": 0.8310339450836182, + "learning_rate": 0.00031300108710422616, + "loss": 3.4323, + "step": 80890 + }, + { + "epoch": 5.496331023236853, + "grad_norm": 0.8400173783302307, + "learning_rate": 0.0003129586220953934, + "loss": 3.4867, + "step": 80895 + }, + { + "epoch": 5.496670743307515, + "grad_norm": 0.8919368386268616, + "learning_rate": 0.00031291615708656066, + "loss": 3.1451, + "step": 80900 + }, + { + "epoch": 5.497010463378176, + "grad_norm": 0.9530535936355591, + "learning_rate": 0.000312873692077728, + "loss": 3.7148, + "step": 80905 + }, + { + "epoch": 5.497350183448838, + "grad_norm": 1.0884747505187988, + "learning_rate": 0.0003128312270688952, + "loss": 3.3274, + "step": 80910 + }, + { + "epoch": 5.4976899035195, + "grad_norm": 0.8325161933898926, + "learning_rate": 0.0003127887620600625, + "loss": 3.8002, + "step": 80915 + }, + { + "epoch": 5.498029623590162, + "grad_norm": 0.7367621064186096, + "learning_rate": 0.00031274629705122984, + "loss": 3.4238, + "step": 80920 + }, + { + "epoch": 5.498369343660824, + "grad_norm": 1.0020225048065186, + "learning_rate": 0.00031270383204239706, + "loss": 3.464, + "step": 80925 + }, + { + "epoch": 5.498709063731486, + "grad_norm": 1.4268810749053955, + "learning_rate": 0.00031266136703356434, + "loss": 3.7458, + "step": 80930 + }, + { + "epoch": 5.499048783802147, + "grad_norm": 0.941967248916626, + "learning_rate": 0.0003126189020247316, + "loss": 3.544, + "step": 80935 + }, + { + "epoch": 5.499388503872809, + "grad_norm": 0.6870529055595398, + "learning_rate": 0.0003125764370158989, + "loss": 3.3328, + "step": 80940 + }, + { + "epoch": 5.499728223943471, + "grad_norm": 0.7225164175033569, + "learning_rate": 0.0003125339720070662, + "loss": 3.5093, + "step": 80945 + }, + { + "epoch": 5.500067944014132, + "grad_norm": 0.8607456088066101, + "learning_rate": 0.00031249150699823346, + "loss": 3.4201, + "step": 80950 + }, + { + "epoch": 5.500407664084794, + "grad_norm": 0.8955379128456116, + "learning_rate": 0.00031244904198940074, + "loss": 3.501, + "step": 80955 + }, + { + "epoch": 5.500747384155456, + "grad_norm": 1.0551795959472656, + "learning_rate": 0.000312406576980568, + "loss": 3.5283, + "step": 80960 + }, + { + "epoch": 5.501087104226118, + "grad_norm": 0.8208785057067871, + "learning_rate": 0.0003123641119717353, + "loss": 3.1868, + "step": 80965 + }, + { + "epoch": 5.50142682429678, + "grad_norm": 0.8330332636833191, + "learning_rate": 0.00031232164696290253, + "loss": 3.5737, + "step": 80970 + }, + { + "epoch": 5.501766544367442, + "grad_norm": 0.8054198026657104, + "learning_rate": 0.00031227918195406986, + "loss": 3.3362, + "step": 80975 + }, + { + "epoch": 5.502106264438103, + "grad_norm": 0.8531404137611389, + "learning_rate": 0.00031223671694523714, + "loss": 3.6338, + "step": 80980 + }, + { + "epoch": 5.502445984508765, + "grad_norm": 0.8784241676330566, + "learning_rate": 0.00031219425193640437, + "loss": 3.4301, + "step": 80985 + }, + { + "epoch": 5.502785704579426, + "grad_norm": 1.2337952852249146, + "learning_rate": 0.0003121517869275717, + "loss": 3.2355, + "step": 80990 + }, + { + "epoch": 5.503125424650088, + "grad_norm": 0.9148406982421875, + "learning_rate": 0.000312109321918739, + "loss": 3.0834, + "step": 80995 + }, + { + "epoch": 5.50346514472075, + "grad_norm": 0.8355448842048645, + "learning_rate": 0.0003120668569099062, + "loss": 3.5353, + "step": 81000 + }, + { + "epoch": 5.5038048647914115, + "grad_norm": 0.8001610040664673, + "learning_rate": 0.0003120243919010735, + "loss": 3.5185, + "step": 81005 + }, + { + "epoch": 5.504144584862074, + "grad_norm": 0.7142758965492249, + "learning_rate": 0.0003119819268922408, + "loss": 3.2596, + "step": 81010 + }, + { + "epoch": 5.504484304932736, + "grad_norm": 1.0178089141845703, + "learning_rate": 0.0003119394618834081, + "loss": 3.3429, + "step": 81015 + }, + { + "epoch": 5.504824025003397, + "grad_norm": 0.7752333879470825, + "learning_rate": 0.00031189699687457533, + "loss": 3.5237, + "step": 81020 + }, + { + "epoch": 5.505163745074059, + "grad_norm": 0.9457226991653442, + "learning_rate": 0.00031185453186574266, + "loss": 3.3177, + "step": 81025 + }, + { + "epoch": 5.505503465144721, + "grad_norm": 0.7255674004554749, + "learning_rate": 0.00031181206685690994, + "loss": 3.3541, + "step": 81030 + }, + { + "epoch": 5.505843185215382, + "grad_norm": 0.8334184885025024, + "learning_rate": 0.00031176960184807717, + "loss": 3.4337, + "step": 81035 + }, + { + "epoch": 5.506182905286044, + "grad_norm": 0.9134833812713623, + "learning_rate": 0.00031172713683924445, + "loss": 3.4832, + "step": 81040 + }, + { + "epoch": 5.506522625356706, + "grad_norm": 1.888919472694397, + "learning_rate": 0.0003116846718304118, + "loss": 3.5251, + "step": 81045 + }, + { + "epoch": 5.5068623454273675, + "grad_norm": 0.7706668972969055, + "learning_rate": 0.000311642206821579, + "loss": 3.4754, + "step": 81050 + }, + { + "epoch": 5.50720206549803, + "grad_norm": 0.8986238837242126, + "learning_rate": 0.0003115997418127463, + "loss": 3.5841, + "step": 81055 + }, + { + "epoch": 5.507541785568692, + "grad_norm": 1.1827826499938965, + "learning_rate": 0.0003115572768039136, + "loss": 3.3321, + "step": 81060 + }, + { + "epoch": 5.507881505639353, + "grad_norm": 0.8335099816322327, + "learning_rate": 0.00031151481179508085, + "loss": 3.4912, + "step": 81065 + }, + { + "epoch": 5.508221225710015, + "grad_norm": 0.7566081285476685, + "learning_rate": 0.00031147234678624813, + "loss": 3.5137, + "step": 81070 + }, + { + "epoch": 5.508560945780677, + "grad_norm": 1.066643476486206, + "learning_rate": 0.0003114298817774154, + "loss": 3.4599, + "step": 81075 + }, + { + "epoch": 5.508900665851338, + "grad_norm": 0.9735884666442871, + "learning_rate": 0.0003113874167685827, + "loss": 3.1983, + "step": 81080 + }, + { + "epoch": 5.509240385922, + "grad_norm": 0.9810009598731995, + "learning_rate": 0.00031134495175974997, + "loss": 3.419, + "step": 81085 + }, + { + "epoch": 5.509580105992662, + "grad_norm": 0.8970182538032532, + "learning_rate": 0.00031130248675091725, + "loss": 3.3858, + "step": 81090 + }, + { + "epoch": 5.5099198260633235, + "grad_norm": 1.007788896560669, + "learning_rate": 0.00031126002174208453, + "loss": 3.5593, + "step": 81095 + }, + { + "epoch": 5.510259546133986, + "grad_norm": 0.8802146911621094, + "learning_rate": 0.0003112175567332518, + "loss": 3.5622, + "step": 81100 + }, + { + "epoch": 5.510599266204648, + "grad_norm": 1.0850130319595337, + "learning_rate": 0.0003111750917244191, + "loss": 3.5951, + "step": 81105 + }, + { + "epoch": 5.510938986275309, + "grad_norm": 0.9907547831535339, + "learning_rate": 0.0003111326267155863, + "loss": 3.5906, + "step": 81110 + }, + { + "epoch": 5.511278706345971, + "grad_norm": 0.7389823794364929, + "learning_rate": 0.00031109016170675365, + "loss": 3.3717, + "step": 81115 + }, + { + "epoch": 5.511618426416633, + "grad_norm": 0.6855267286300659, + "learning_rate": 0.00031104769669792093, + "loss": 3.4491, + "step": 81120 + }, + { + "epoch": 5.511958146487294, + "grad_norm": 0.8731095194816589, + "learning_rate": 0.00031100523168908816, + "loss": 3.2034, + "step": 81125 + }, + { + "epoch": 5.512297866557956, + "grad_norm": 1.0679022073745728, + "learning_rate": 0.0003109627666802555, + "loss": 3.5769, + "step": 81130 + }, + { + "epoch": 5.512637586628618, + "grad_norm": 0.9910141229629517, + "learning_rate": 0.00031092030167142277, + "loss": 3.4737, + "step": 81135 + }, + { + "epoch": 5.5129773066992795, + "grad_norm": 0.8359428644180298, + "learning_rate": 0.00031087783666259, + "loss": 3.3912, + "step": 81140 + }, + { + "epoch": 5.513317026769942, + "grad_norm": 1.4860799312591553, + "learning_rate": 0.0003108353716537573, + "loss": 3.5919, + "step": 81145 + }, + { + "epoch": 5.513656746840604, + "grad_norm": 0.8569153547286987, + "learning_rate": 0.0003107929066449246, + "loss": 3.5638, + "step": 81150 + }, + { + "epoch": 5.513996466911265, + "grad_norm": 0.6971043348312378, + "learning_rate": 0.00031075044163609184, + "loss": 3.363, + "step": 81155 + }, + { + "epoch": 5.514336186981927, + "grad_norm": 1.0334975719451904, + "learning_rate": 0.0003107079766272591, + "loss": 3.7928, + "step": 81160 + }, + { + "epoch": 5.514675907052589, + "grad_norm": 0.9720942974090576, + "learning_rate": 0.00031066551161842645, + "loss": 3.652, + "step": 81165 + }, + { + "epoch": 5.51501562712325, + "grad_norm": 1.0383622646331787, + "learning_rate": 0.0003106230466095937, + "loss": 3.5169, + "step": 81170 + }, + { + "epoch": 5.515355347193912, + "grad_norm": 1.071237564086914, + "learning_rate": 0.00031058058160076096, + "loss": 3.3401, + "step": 81175 + }, + { + "epoch": 5.515695067264574, + "grad_norm": 0.8561421632766724, + "learning_rate": 0.0003105381165919283, + "loss": 3.528, + "step": 81180 + }, + { + "epoch": 5.516034787335236, + "grad_norm": 1.0647273063659668, + "learning_rate": 0.00031049565158309557, + "loss": 3.5147, + "step": 81185 + }, + { + "epoch": 5.516374507405898, + "grad_norm": 0.8341363072395325, + "learning_rate": 0.0003104531865742628, + "loss": 3.5733, + "step": 81190 + }, + { + "epoch": 5.516714227476559, + "grad_norm": 0.9674589037895203, + "learning_rate": 0.0003104107215654301, + "loss": 3.4614, + "step": 81195 + }, + { + "epoch": 5.517053947547221, + "grad_norm": 0.7738350033760071, + "learning_rate": 0.0003103682565565974, + "loss": 3.5352, + "step": 81200 + }, + { + "epoch": 5.517393667617883, + "grad_norm": 0.9460663199424744, + "learning_rate": 0.00031032579154776464, + "loss": 3.3693, + "step": 81205 + }, + { + "epoch": 5.517733387688544, + "grad_norm": 0.9426915049552917, + "learning_rate": 0.0003102833265389319, + "loss": 3.6383, + "step": 81210 + }, + { + "epoch": 5.518073107759206, + "grad_norm": 0.7740843296051025, + "learning_rate": 0.00031024086153009925, + "loss": 3.5264, + "step": 81215 + }, + { + "epoch": 5.518412827829868, + "grad_norm": 0.763244092464447, + "learning_rate": 0.0003101983965212665, + "loss": 3.5959, + "step": 81220 + }, + { + "epoch": 5.5187525479005295, + "grad_norm": 0.8581451177597046, + "learning_rate": 0.00031015593151243376, + "loss": 3.7106, + "step": 81225 + }, + { + "epoch": 5.519092267971192, + "grad_norm": 0.9373214840888977, + "learning_rate": 0.00031011346650360104, + "loss": 3.3907, + "step": 81230 + }, + { + "epoch": 5.519431988041854, + "grad_norm": 0.784247636795044, + "learning_rate": 0.0003100710014947683, + "loss": 3.6591, + "step": 81235 + }, + { + "epoch": 5.519771708112515, + "grad_norm": 0.7847091555595398, + "learning_rate": 0.0003100285364859356, + "loss": 3.7057, + "step": 81240 + }, + { + "epoch": 5.520111428183177, + "grad_norm": 0.7519651055335999, + "learning_rate": 0.0003099860714771029, + "loss": 3.736, + "step": 81245 + }, + { + "epoch": 5.520451148253839, + "grad_norm": 0.8794171810150146, + "learning_rate": 0.00030994360646827016, + "loss": 3.7029, + "step": 81250 + }, + { + "epoch": 5.5207908683245, + "grad_norm": 0.8762409687042236, + "learning_rate": 0.00030990114145943744, + "loss": 3.4056, + "step": 81255 + }, + { + "epoch": 5.521130588395162, + "grad_norm": 1.152618169784546, + "learning_rate": 0.0003098586764506047, + "loss": 3.7391, + "step": 81260 + }, + { + "epoch": 5.521470308465824, + "grad_norm": 1.0222599506378174, + "learning_rate": 0.00030981621144177194, + "loss": 3.383, + "step": 81265 + }, + { + "epoch": 5.5218100285364855, + "grad_norm": 0.9656743407249451, + "learning_rate": 0.0003097737464329393, + "loss": 3.3963, + "step": 81270 + }, + { + "epoch": 5.522149748607148, + "grad_norm": 0.9820765852928162, + "learning_rate": 0.00030973128142410656, + "loss": 3.3651, + "step": 81275 + }, + { + "epoch": 5.52248946867781, + "grad_norm": 0.9623928666114807, + "learning_rate": 0.0003096888164152738, + "loss": 3.6043, + "step": 81280 + }, + { + "epoch": 5.522829188748471, + "grad_norm": 0.8027721047401428, + "learning_rate": 0.0003096463514064411, + "loss": 3.6281, + "step": 81285 + }, + { + "epoch": 5.523168908819133, + "grad_norm": 1.0896185636520386, + "learning_rate": 0.0003096038863976084, + "loss": 3.704, + "step": 81290 + }, + { + "epoch": 5.523508628889795, + "grad_norm": 1.062348484992981, + "learning_rate": 0.0003095614213887756, + "loss": 3.4156, + "step": 81295 + }, + { + "epoch": 5.523848348960456, + "grad_norm": 0.8389953970909119, + "learning_rate": 0.0003095189563799429, + "loss": 3.4855, + "step": 81300 + }, + { + "epoch": 5.524188069031118, + "grad_norm": 0.7962811589241028, + "learning_rate": 0.00030947649137111024, + "loss": 3.5014, + "step": 81305 + }, + { + "epoch": 5.52452778910178, + "grad_norm": 1.152745008468628, + "learning_rate": 0.00030943402636227747, + "loss": 3.6012, + "step": 81310 + }, + { + "epoch": 5.5248675091724415, + "grad_norm": 0.880626380443573, + "learning_rate": 0.00030939156135344475, + "loss": 3.3709, + "step": 81315 + }, + { + "epoch": 5.525207229243104, + "grad_norm": 1.0217185020446777, + "learning_rate": 0.0003093490963446121, + "loss": 3.2734, + "step": 81320 + }, + { + "epoch": 5.525546949313766, + "grad_norm": 0.7716484665870667, + "learning_rate": 0.0003093066313357793, + "loss": 3.3719, + "step": 81325 + }, + { + "epoch": 5.525886669384427, + "grad_norm": 0.7668516039848328, + "learning_rate": 0.0003092641663269466, + "loss": 3.4088, + "step": 81330 + }, + { + "epoch": 5.526226389455089, + "grad_norm": 0.8431546688079834, + "learning_rate": 0.00030922170131811387, + "loss": 3.7341, + "step": 81335 + }, + { + "epoch": 5.526566109525751, + "grad_norm": 0.7565270066261292, + "learning_rate": 0.00030917923630928115, + "loss": 3.5849, + "step": 81340 + }, + { + "epoch": 5.526905829596412, + "grad_norm": 1.013279914855957, + "learning_rate": 0.0003091367713004484, + "loss": 3.4597, + "step": 81345 + }, + { + "epoch": 5.527245549667074, + "grad_norm": 0.8725472092628479, + "learning_rate": 0.0003090943062916157, + "loss": 3.6743, + "step": 81350 + }, + { + "epoch": 5.527585269737736, + "grad_norm": 0.9197516441345215, + "learning_rate": 0.00030905184128278304, + "loss": 3.5573, + "step": 81355 + }, + { + "epoch": 5.5279249898083975, + "grad_norm": 0.8649563789367676, + "learning_rate": 0.00030900937627395027, + "loss": 3.2833, + "step": 81360 + }, + { + "epoch": 5.52826470987906, + "grad_norm": 0.9522911906242371, + "learning_rate": 0.00030896691126511755, + "loss": 3.4647, + "step": 81365 + }, + { + "epoch": 5.528604429949722, + "grad_norm": 0.971309244632721, + "learning_rate": 0.0003089244462562848, + "loss": 3.098, + "step": 81370 + }, + { + "epoch": 5.528944150020383, + "grad_norm": 0.8210400342941284, + "learning_rate": 0.0003088819812474521, + "loss": 3.4412, + "step": 81375 + }, + { + "epoch": 5.529283870091045, + "grad_norm": 0.922982931137085, + "learning_rate": 0.0003088395162386194, + "loss": 3.4404, + "step": 81380 + }, + { + "epoch": 5.529623590161707, + "grad_norm": 0.864836573600769, + "learning_rate": 0.00030879705122978667, + "loss": 3.6085, + "step": 81385 + }, + { + "epoch": 5.529963310232368, + "grad_norm": 0.7084240913391113, + "learning_rate": 0.00030875458622095395, + "loss": 3.46, + "step": 81390 + }, + { + "epoch": 5.53030303030303, + "grad_norm": 0.9100731611251831, + "learning_rate": 0.0003087121212121212, + "loss": 3.3204, + "step": 81395 + }, + { + "epoch": 5.530642750373692, + "grad_norm": 1.1598457098007202, + "learning_rate": 0.0003086696562032885, + "loss": 3.3535, + "step": 81400 + }, + { + "epoch": 5.5309824704443535, + "grad_norm": 0.759475827217102, + "learning_rate": 0.00030862719119445573, + "loss": 3.5396, + "step": 81405 + }, + { + "epoch": 5.531322190515016, + "grad_norm": 0.9689304828643799, + "learning_rate": 0.00030858472618562307, + "loss": 3.4808, + "step": 81410 + }, + { + "epoch": 5.531661910585678, + "grad_norm": 0.9662784934043884, + "learning_rate": 0.00030854226117679035, + "loss": 3.638, + "step": 81415 + }, + { + "epoch": 5.532001630656339, + "grad_norm": 0.7969033718109131, + "learning_rate": 0.0003084997961679576, + "loss": 3.6212, + "step": 81420 + }, + { + "epoch": 5.532341350727001, + "grad_norm": 0.8373355865478516, + "learning_rate": 0.0003084573311591249, + "loss": 3.2992, + "step": 81425 + }, + { + "epoch": 5.532681070797663, + "grad_norm": 0.75406414270401, + "learning_rate": 0.0003084148661502922, + "loss": 3.3365, + "step": 81430 + }, + { + "epoch": 5.533020790868324, + "grad_norm": 0.8164325952529907, + "learning_rate": 0.0003083724011414594, + "loss": 3.5745, + "step": 81435 + }, + { + "epoch": 5.533360510938986, + "grad_norm": 0.8856513500213623, + "learning_rate": 0.00030832993613262675, + "loss": 3.5332, + "step": 81440 + }, + { + "epoch": 5.533700231009648, + "grad_norm": 0.92999666929245, + "learning_rate": 0.00030828747112379403, + "loss": 3.5749, + "step": 81445 + }, + { + "epoch": 5.5340399510803095, + "grad_norm": 1.1994752883911133, + "learning_rate": 0.00030824500611496125, + "loss": 3.2454, + "step": 81450 + }, + { + "epoch": 5.534379671150972, + "grad_norm": 0.7879796624183655, + "learning_rate": 0.00030820254110612853, + "loss": 3.4838, + "step": 81455 + }, + { + "epoch": 5.534719391221634, + "grad_norm": 0.9186739921569824, + "learning_rate": 0.00030816007609729587, + "loss": 3.4186, + "step": 81460 + }, + { + "epoch": 5.535059111292295, + "grad_norm": 0.9426395297050476, + "learning_rate": 0.0003081176110884631, + "loss": 3.5337, + "step": 81465 + }, + { + "epoch": 5.535398831362957, + "grad_norm": 0.9048250317573547, + "learning_rate": 0.0003080751460796304, + "loss": 3.3633, + "step": 81470 + }, + { + "epoch": 5.535738551433619, + "grad_norm": 0.8763076663017273, + "learning_rate": 0.0003080326810707977, + "loss": 3.6328, + "step": 81475 + }, + { + "epoch": 5.53607827150428, + "grad_norm": 0.9969365000724792, + "learning_rate": 0.00030799021606196493, + "loss": 3.4891, + "step": 81480 + }, + { + "epoch": 5.536417991574942, + "grad_norm": 1.0034382343292236, + "learning_rate": 0.0003079477510531322, + "loss": 3.4807, + "step": 81485 + }, + { + "epoch": 5.536757711645604, + "grad_norm": 0.7597475051879883, + "learning_rate": 0.0003079052860442995, + "loss": 3.4156, + "step": 81490 + }, + { + "epoch": 5.537097431716266, + "grad_norm": 0.9130927324295044, + "learning_rate": 0.0003078628210354668, + "loss": 3.6135, + "step": 81495 + }, + { + "epoch": 5.537437151786928, + "grad_norm": 0.8239119052886963, + "learning_rate": 0.00030782035602663405, + "loss": 3.3812, + "step": 81500 + }, + { + "epoch": 5.53777687185759, + "grad_norm": 0.8413383364677429, + "learning_rate": 0.00030777789101780133, + "loss": 3.8013, + "step": 81505 + }, + { + "epoch": 5.538116591928251, + "grad_norm": 1.042863130569458, + "learning_rate": 0.0003077354260089686, + "loss": 3.2555, + "step": 81510 + }, + { + "epoch": 5.538456311998913, + "grad_norm": 1.0980210304260254, + "learning_rate": 0.0003076929610001359, + "loss": 3.4268, + "step": 81515 + }, + { + "epoch": 5.538796032069575, + "grad_norm": 0.9523441195487976, + "learning_rate": 0.0003076504959913032, + "loss": 3.4618, + "step": 81520 + }, + { + "epoch": 5.539135752140236, + "grad_norm": 0.7750239968299866, + "learning_rate": 0.00030760803098247045, + "loss": 3.6177, + "step": 81525 + }, + { + "epoch": 5.539475472210898, + "grad_norm": 0.8557862043380737, + "learning_rate": 0.00030756556597363773, + "loss": 3.6136, + "step": 81530 + }, + { + "epoch": 5.53981519228156, + "grad_norm": 0.9203059673309326, + "learning_rate": 0.000307523100964805, + "loss": 3.611, + "step": 81535 + }, + { + "epoch": 5.540154912352222, + "grad_norm": 0.9897331595420837, + "learning_rate": 0.0003074806359559723, + "loss": 3.5575, + "step": 81540 + }, + { + "epoch": 5.540494632422884, + "grad_norm": 0.725673496723175, + "learning_rate": 0.0003074381709471396, + "loss": 3.503, + "step": 81545 + }, + { + "epoch": 5.540834352493546, + "grad_norm": 0.845152735710144, + "learning_rate": 0.00030739570593830685, + "loss": 3.5146, + "step": 81550 + }, + { + "epoch": 5.541174072564207, + "grad_norm": 0.92283034324646, + "learning_rate": 0.00030735324092947414, + "loss": 3.5243, + "step": 81555 + }, + { + "epoch": 5.541513792634869, + "grad_norm": 0.7134180068969727, + "learning_rate": 0.00030731077592064136, + "loss": 3.6435, + "step": 81560 + }, + { + "epoch": 5.541853512705531, + "grad_norm": 0.815041720867157, + "learning_rate": 0.0003072683109118087, + "loss": 3.232, + "step": 81565 + }, + { + "epoch": 5.542193232776192, + "grad_norm": 0.95195072889328, + "learning_rate": 0.000307225845902976, + "loss": 3.4998, + "step": 81570 + }, + { + "epoch": 5.542532952846854, + "grad_norm": 1.0015026330947876, + "learning_rate": 0.0003071833808941432, + "loss": 3.2192, + "step": 81575 + }, + { + "epoch": 5.542872672917516, + "grad_norm": 0.7975209355354309, + "learning_rate": 0.00030714091588531054, + "loss": 3.5386, + "step": 81580 + }, + { + "epoch": 5.543212392988178, + "grad_norm": 1.0146900415420532, + "learning_rate": 0.0003070984508764778, + "loss": 3.5969, + "step": 81585 + }, + { + "epoch": 5.54355211305884, + "grad_norm": 1.1870993375778198, + "learning_rate": 0.00030705598586764504, + "loss": 3.4088, + "step": 81590 + }, + { + "epoch": 5.543891833129502, + "grad_norm": 0.7946213483810425, + "learning_rate": 0.0003070135208588123, + "loss": 3.3111, + "step": 81595 + }, + { + "epoch": 5.544231553200163, + "grad_norm": 0.741422712802887, + "learning_rate": 0.00030697105584997966, + "loss": 3.4492, + "step": 81600 + }, + { + "epoch": 5.544571273270825, + "grad_norm": 0.8405625820159912, + "learning_rate": 0.0003069285908411469, + "loss": 3.5581, + "step": 81605 + }, + { + "epoch": 5.544910993341487, + "grad_norm": 0.791593611240387, + "learning_rate": 0.00030688612583231416, + "loss": 3.5571, + "step": 81610 + }, + { + "epoch": 5.545250713412148, + "grad_norm": 0.9306030869483948, + "learning_rate": 0.0003068436608234815, + "loss": 3.2705, + "step": 81615 + }, + { + "epoch": 5.54559043348281, + "grad_norm": 1.1521319150924683, + "learning_rate": 0.0003068011958146487, + "loss": 3.246, + "step": 81620 + }, + { + "epoch": 5.545930153553472, + "grad_norm": 0.9561100602149963, + "learning_rate": 0.000306758730805816, + "loss": 3.5496, + "step": 81625 + }, + { + "epoch": 5.546269873624134, + "grad_norm": 1.022749900817871, + "learning_rate": 0.0003067162657969833, + "loss": 3.4377, + "step": 81630 + }, + { + "epoch": 5.546609593694796, + "grad_norm": 0.7863710522651672, + "learning_rate": 0.00030667380078815056, + "loss": 3.256, + "step": 81635 + }, + { + "epoch": 5.546949313765458, + "grad_norm": 1.0092450380325317, + "learning_rate": 0.00030663133577931784, + "loss": 3.0453, + "step": 81640 + }, + { + "epoch": 5.547289033836119, + "grad_norm": 0.7912927865982056, + "learning_rate": 0.0003065888707704851, + "loss": 3.3555, + "step": 81645 + }, + { + "epoch": 5.547628753906781, + "grad_norm": 0.6961928606033325, + "learning_rate": 0.0003065464057616524, + "loss": 3.734, + "step": 81650 + }, + { + "epoch": 5.547968473977443, + "grad_norm": 0.9818047881126404, + "learning_rate": 0.0003065039407528197, + "loss": 3.1864, + "step": 81655 + }, + { + "epoch": 5.548308194048104, + "grad_norm": 0.6745341420173645, + "learning_rate": 0.00030646147574398696, + "loss": 3.5434, + "step": 81660 + }, + { + "epoch": 5.548647914118766, + "grad_norm": 0.8936575055122375, + "learning_rate": 0.0003064190107351542, + "loss": 3.5217, + "step": 81665 + }, + { + "epoch": 5.5489876341894275, + "grad_norm": 1.0344130992889404, + "learning_rate": 0.0003063765457263215, + "loss": 3.5695, + "step": 81670 + }, + { + "epoch": 5.54932735426009, + "grad_norm": 0.7889003753662109, + "learning_rate": 0.0003063340807174888, + "loss": 3.1696, + "step": 81675 + }, + { + "epoch": 5.549667074330752, + "grad_norm": 1.233446717262268, + "learning_rate": 0.00030629161570865603, + "loss": 3.4801, + "step": 81680 + }, + { + "epoch": 5.550006794401413, + "grad_norm": 0.7653019428253174, + "learning_rate": 0.00030624915069982336, + "loss": 3.4087, + "step": 81685 + }, + { + "epoch": 5.550346514472075, + "grad_norm": 0.867329478263855, + "learning_rate": 0.00030620668569099064, + "loss": 3.3949, + "step": 81690 + }, + { + "epoch": 5.550686234542737, + "grad_norm": 0.950331449508667, + "learning_rate": 0.0003061642206821579, + "loss": 3.4376, + "step": 81695 + }, + { + "epoch": 5.551025954613398, + "grad_norm": 1.0427284240722656, + "learning_rate": 0.00030612175567332515, + "loss": 3.5348, + "step": 81700 + }, + { + "epoch": 5.55136567468406, + "grad_norm": 0.9118134379386902, + "learning_rate": 0.0003060792906644925, + "loss": 3.4512, + "step": 81705 + }, + { + "epoch": 5.551705394754722, + "grad_norm": 0.9592965841293335, + "learning_rate": 0.00030603682565565976, + "loss": 3.157, + "step": 81710 + }, + { + "epoch": 5.5520451148253835, + "grad_norm": 0.9162136316299438, + "learning_rate": 0.000305994360646827, + "loss": 3.4959, + "step": 81715 + }, + { + "epoch": 5.552384834896046, + "grad_norm": 1.0603339672088623, + "learning_rate": 0.0003059518956379943, + "loss": 3.4895, + "step": 81720 + }, + { + "epoch": 5.552724554966708, + "grad_norm": 1.4533050060272217, + "learning_rate": 0.0003059094306291616, + "loss": 3.6457, + "step": 81725 + }, + { + "epoch": 5.553064275037369, + "grad_norm": 0.9384618997573853, + "learning_rate": 0.00030586696562032883, + "loss": 3.33, + "step": 81730 + }, + { + "epoch": 5.553403995108031, + "grad_norm": 2.9775397777557373, + "learning_rate": 0.00030582450061149616, + "loss": 3.5713, + "step": 81735 + }, + { + "epoch": 5.553743715178693, + "grad_norm": 0.8903524875640869, + "learning_rate": 0.00030578203560266344, + "loss": 3.7137, + "step": 81740 + }, + { + "epoch": 5.554083435249354, + "grad_norm": 0.730891764163971, + "learning_rate": 0.00030573957059383067, + "loss": 3.5549, + "step": 81745 + }, + { + "epoch": 5.554423155320016, + "grad_norm": 1.0943650007247925, + "learning_rate": 0.00030569710558499795, + "loss": 3.2535, + "step": 81750 + }, + { + "epoch": 5.554762875390678, + "grad_norm": 0.7178515195846558, + "learning_rate": 0.0003056546405761653, + "loss": 3.3708, + "step": 81755 + }, + { + "epoch": 5.5551025954613396, + "grad_norm": 0.8439033031463623, + "learning_rate": 0.0003056121755673325, + "loss": 3.4483, + "step": 81760 + }, + { + "epoch": 5.555442315532002, + "grad_norm": 1.1084353923797607, + "learning_rate": 0.0003055697105584998, + "loss": 3.3225, + "step": 81765 + }, + { + "epoch": 5.555782035602664, + "grad_norm": 0.864029049873352, + "learning_rate": 0.0003055272455496671, + "loss": 3.3363, + "step": 81770 + }, + { + "epoch": 5.556121755673325, + "grad_norm": 1.131706714630127, + "learning_rate": 0.00030548478054083435, + "loss": 3.5936, + "step": 81775 + }, + { + "epoch": 5.556461475743987, + "grad_norm": 0.9713962078094482, + "learning_rate": 0.00030544231553200163, + "loss": 3.4988, + "step": 81780 + }, + { + "epoch": 5.556801195814649, + "grad_norm": 0.815556526184082, + "learning_rate": 0.0003053998505231689, + "loss": 3.704, + "step": 81785 + }, + { + "epoch": 5.55714091588531, + "grad_norm": 0.9397998452186584, + "learning_rate": 0.0003053573855143362, + "loss": 3.3946, + "step": 81790 + }, + { + "epoch": 5.557480635955972, + "grad_norm": 0.7259123921394348, + "learning_rate": 0.00030531492050550347, + "loss": 3.2194, + "step": 81795 + }, + { + "epoch": 5.557820356026634, + "grad_norm": 0.9877852201461792, + "learning_rate": 0.00030527245549667075, + "loss": 3.4022, + "step": 81800 + }, + { + "epoch": 5.558160076097296, + "grad_norm": 0.8011809587478638, + "learning_rate": 0.00030522999048783803, + "loss": 3.5469, + "step": 81805 + }, + { + "epoch": 5.558499796167958, + "grad_norm": 1.0196787118911743, + "learning_rate": 0.0003051875254790053, + "loss": 3.3883, + "step": 81810 + }, + { + "epoch": 5.55883951623862, + "grad_norm": 0.8918800950050354, + "learning_rate": 0.0003051450604701726, + "loss": 3.5412, + "step": 81815 + }, + { + "epoch": 5.559179236309281, + "grad_norm": 0.9680745005607605, + "learning_rate": 0.0003051025954613398, + "loss": 3.5482, + "step": 81820 + }, + { + "epoch": 5.559518956379943, + "grad_norm": 1.0865769386291504, + "learning_rate": 0.00030506013045250715, + "loss": 3.4213, + "step": 81825 + }, + { + "epoch": 5.559858676450605, + "grad_norm": 0.8647723197937012, + "learning_rate": 0.00030501766544367443, + "loss": 3.4375, + "step": 81830 + }, + { + "epoch": 5.560198396521266, + "grad_norm": 0.9392686486244202, + "learning_rate": 0.00030497520043484166, + "loss": 3.5077, + "step": 81835 + }, + { + "epoch": 5.560538116591928, + "grad_norm": 0.91786789894104, + "learning_rate": 0.000304932735426009, + "loss": 3.5334, + "step": 81840 + }, + { + "epoch": 5.56087783666259, + "grad_norm": 0.9021332859992981, + "learning_rate": 0.00030489027041717627, + "loss": 3.527, + "step": 81845 + }, + { + "epoch": 5.561217556733252, + "grad_norm": 0.8272998929023743, + "learning_rate": 0.0003048478054083435, + "loss": 3.4024, + "step": 81850 + }, + { + "epoch": 5.561557276803914, + "grad_norm": 0.9198755621910095, + "learning_rate": 0.0003048053403995108, + "loss": 3.657, + "step": 81855 + }, + { + "epoch": 5.561896996874576, + "grad_norm": 0.832099437713623, + "learning_rate": 0.0003047628753906781, + "loss": 3.4243, + "step": 81860 + }, + { + "epoch": 5.562236716945237, + "grad_norm": 0.887787401676178, + "learning_rate": 0.0003047204103818454, + "loss": 3.043, + "step": 81865 + }, + { + "epoch": 5.562576437015899, + "grad_norm": 0.9349852800369263, + "learning_rate": 0.0003046779453730126, + "loss": 3.541, + "step": 81870 + }, + { + "epoch": 5.56291615708656, + "grad_norm": 0.7738796472549438, + "learning_rate": 0.00030463548036417995, + "loss": 3.2429, + "step": 81875 + }, + { + "epoch": 5.563255877157222, + "grad_norm": 0.9025421142578125, + "learning_rate": 0.00030459301535534723, + "loss": 3.1446, + "step": 81880 + }, + { + "epoch": 5.563595597227884, + "grad_norm": 0.9583965539932251, + "learning_rate": 0.00030455055034651446, + "loss": 3.2242, + "step": 81885 + }, + { + "epoch": 5.5639353172985455, + "grad_norm": 0.8384267091751099, + "learning_rate": 0.00030450808533768174, + "loss": 3.4183, + "step": 81890 + }, + { + "epoch": 5.564275037369208, + "grad_norm": 0.885161817073822, + "learning_rate": 0.00030446562032884907, + "loss": 3.2714, + "step": 81895 + }, + { + "epoch": 5.56461475743987, + "grad_norm": 0.8946053981781006, + "learning_rate": 0.0003044231553200163, + "loss": 3.3755, + "step": 81900 + }, + { + "epoch": 5.564954477510531, + "grad_norm": 0.9393993616104126, + "learning_rate": 0.0003043806903111836, + "loss": 3.504, + "step": 81905 + }, + { + "epoch": 5.565294197581193, + "grad_norm": 0.7906018495559692, + "learning_rate": 0.0003043382253023509, + "loss": 3.5454, + "step": 81910 + }, + { + "epoch": 5.565633917651855, + "grad_norm": 0.8550724387168884, + "learning_rate": 0.00030429576029351814, + "loss": 3.4956, + "step": 81915 + }, + { + "epoch": 5.565973637722516, + "grad_norm": 1.1295901536941528, + "learning_rate": 0.0003042532952846854, + "loss": 3.0116, + "step": 81920 + }, + { + "epoch": 5.566313357793178, + "grad_norm": 0.987320065498352, + "learning_rate": 0.0003042108302758527, + "loss": 3.4867, + "step": 81925 + }, + { + "epoch": 5.56665307786384, + "grad_norm": 0.8098143339157104, + "learning_rate": 0.00030416836526702, + "loss": 3.2854, + "step": 81930 + }, + { + "epoch": 5.5669927979345015, + "grad_norm": 1.0826642513275146, + "learning_rate": 0.00030412590025818726, + "loss": 3.4618, + "step": 81935 + }, + { + "epoch": 5.567332518005164, + "grad_norm": 0.7229319214820862, + "learning_rate": 0.00030408343524935454, + "loss": 3.7403, + "step": 81940 + }, + { + "epoch": 5.567672238075826, + "grad_norm": 1.0227519273757935, + "learning_rate": 0.0003040409702405218, + "loss": 3.3575, + "step": 81945 + }, + { + "epoch": 5.568011958146487, + "grad_norm": 0.928403913974762, + "learning_rate": 0.0003039985052316891, + "loss": 3.6202, + "step": 81950 + }, + { + "epoch": 5.568351678217149, + "grad_norm": 1.0340694189071655, + "learning_rate": 0.0003039560402228564, + "loss": 3.2517, + "step": 81955 + }, + { + "epoch": 5.568691398287811, + "grad_norm": 0.906521201133728, + "learning_rate": 0.0003039135752140236, + "loss": 3.4598, + "step": 81960 + }, + { + "epoch": 5.569031118358472, + "grad_norm": 0.8268351554870605, + "learning_rate": 0.00030387111020519094, + "loss": 3.433, + "step": 81965 + }, + { + "epoch": 5.569370838429134, + "grad_norm": 0.8038462400436401, + "learning_rate": 0.0003038286451963582, + "loss": 3.5709, + "step": 81970 + }, + { + "epoch": 5.569710558499796, + "grad_norm": 1.0649347305297852, + "learning_rate": 0.00030378618018752544, + "loss": 3.5176, + "step": 81975 + }, + { + "epoch": 5.5700502785704575, + "grad_norm": 0.9329836368560791, + "learning_rate": 0.0003037437151786928, + "loss": 3.4801, + "step": 81980 + }, + { + "epoch": 5.57038999864112, + "grad_norm": 0.8207629323005676, + "learning_rate": 0.00030370125016986006, + "loss": 3.5338, + "step": 81985 + }, + { + "epoch": 5.570729718711782, + "grad_norm": 0.9361269474029541, + "learning_rate": 0.0003036587851610273, + "loss": 3.2225, + "step": 81990 + }, + { + "epoch": 5.571069438782443, + "grad_norm": 0.8225507140159607, + "learning_rate": 0.00030361632015219456, + "loss": 3.4942, + "step": 81995 + }, + { + "epoch": 5.571409158853105, + "grad_norm": 0.9554885029792786, + "learning_rate": 0.0003035738551433619, + "loss": 3.4684, + "step": 82000 + }, + { + "epoch": 5.571748878923767, + "grad_norm": 0.8908816576004028, + "learning_rate": 0.0003035313901345291, + "loss": 3.3062, + "step": 82005 + }, + { + "epoch": 5.572088598994428, + "grad_norm": 1.008134365081787, + "learning_rate": 0.0003034889251256964, + "loss": 3.4236, + "step": 82010 + }, + { + "epoch": 5.57242831906509, + "grad_norm": 1.0869617462158203, + "learning_rate": 0.00030344646011686374, + "loss": 3.5957, + "step": 82015 + }, + { + "epoch": 5.572768039135752, + "grad_norm": 0.9642946124076843, + "learning_rate": 0.00030340399510803097, + "loss": 3.3498, + "step": 82020 + }, + { + "epoch": 5.5731077592064135, + "grad_norm": 0.719622015953064, + "learning_rate": 0.00030336153009919825, + "loss": 3.6525, + "step": 82025 + }, + { + "epoch": 5.573447479277076, + "grad_norm": 0.7496836185455322, + "learning_rate": 0.0003033190650903656, + "loss": 3.6613, + "step": 82030 + }, + { + "epoch": 5.573787199347738, + "grad_norm": 0.9266257286071777, + "learning_rate": 0.00030327660008153286, + "loss": 3.8335, + "step": 82035 + }, + { + "epoch": 5.574126919418399, + "grad_norm": 0.861636757850647, + "learning_rate": 0.0003032341350727001, + "loss": 3.6411, + "step": 82040 + }, + { + "epoch": 5.574466639489061, + "grad_norm": 0.8968741297721863, + "learning_rate": 0.00030319167006386737, + "loss": 3.4372, + "step": 82045 + }, + { + "epoch": 5.574806359559723, + "grad_norm": 0.8568574786186218, + "learning_rate": 0.0003031492050550347, + "loss": 3.5561, + "step": 82050 + }, + { + "epoch": 5.575146079630384, + "grad_norm": 0.817346453666687, + "learning_rate": 0.0003031067400462019, + "loss": 3.3599, + "step": 82055 + }, + { + "epoch": 5.575485799701046, + "grad_norm": 0.9702475070953369, + "learning_rate": 0.0003030642750373692, + "loss": 3.429, + "step": 82060 + }, + { + "epoch": 5.575825519771708, + "grad_norm": 0.8469544053077698, + "learning_rate": 0.00030302181002853654, + "loss": 3.2703, + "step": 82065 + }, + { + "epoch": 5.57616523984237, + "grad_norm": 0.7062801718711853, + "learning_rate": 0.00030297934501970377, + "loss": 3.4251, + "step": 82070 + }, + { + "epoch": 5.576504959913032, + "grad_norm": 0.8736145496368408, + "learning_rate": 0.00030293688001087105, + "loss": 3.3132, + "step": 82075 + }, + { + "epoch": 5.576844679983694, + "grad_norm": 0.8844289183616638, + "learning_rate": 0.0003028944150020383, + "loss": 3.5241, + "step": 82080 + }, + { + "epoch": 5.577184400054355, + "grad_norm": 1.1152395009994507, + "learning_rate": 0.0003028519499932056, + "loss": 3.6191, + "step": 82085 + }, + { + "epoch": 5.577524120125017, + "grad_norm": 0.8079606890678406, + "learning_rate": 0.0003028094849843729, + "loss": 3.6467, + "step": 82090 + }, + { + "epoch": 5.577863840195679, + "grad_norm": 0.951145589351654, + "learning_rate": 0.00030276701997554017, + "loss": 3.3736, + "step": 82095 + }, + { + "epoch": 5.57820356026634, + "grad_norm": 1.0148407220840454, + "learning_rate": 0.00030272455496670745, + "loss": 3.6894, + "step": 82100 + }, + { + "epoch": 5.578543280337002, + "grad_norm": 0.8269964456558228, + "learning_rate": 0.0003026820899578747, + "loss": 3.4921, + "step": 82105 + }, + { + "epoch": 5.578883000407664, + "grad_norm": 0.8297431468963623, + "learning_rate": 0.000302639624949042, + "loss": 3.4525, + "step": 82110 + }, + { + "epoch": 5.579222720478326, + "grad_norm": 1.009473204612732, + "learning_rate": 0.00030259715994020923, + "loss": 3.1861, + "step": 82115 + }, + { + "epoch": 5.579562440548988, + "grad_norm": 0.9926433563232422, + "learning_rate": 0.00030255469493137657, + "loss": 3.5976, + "step": 82120 + }, + { + "epoch": 5.57990216061965, + "grad_norm": 0.9074004888534546, + "learning_rate": 0.00030251222992254385, + "loss": 3.4014, + "step": 82125 + }, + { + "epoch": 5.580241880690311, + "grad_norm": 1.0215978622436523, + "learning_rate": 0.00030246976491371107, + "loss": 3.3353, + "step": 82130 + }, + { + "epoch": 5.580581600760973, + "grad_norm": 1.1065857410430908, + "learning_rate": 0.0003024272999048784, + "loss": 3.4733, + "step": 82135 + }, + { + "epoch": 5.580921320831635, + "grad_norm": 0.9878951907157898, + "learning_rate": 0.0003023848348960457, + "loss": 3.4551, + "step": 82140 + }, + { + "epoch": 5.581261040902296, + "grad_norm": 0.8208097219467163, + "learning_rate": 0.0003023423698872129, + "loss": 3.7223, + "step": 82145 + }, + { + "epoch": 5.581600760972958, + "grad_norm": 1.0541378259658813, + "learning_rate": 0.0003022999048783802, + "loss": 3.6887, + "step": 82150 + }, + { + "epoch": 5.58194048104362, + "grad_norm": 0.8775576949119568, + "learning_rate": 0.00030225743986954753, + "loss": 3.5467, + "step": 82155 + }, + { + "epoch": 5.582280201114282, + "grad_norm": 0.9530203938484192, + "learning_rate": 0.00030221497486071475, + "loss": 3.7089, + "step": 82160 + }, + { + "epoch": 5.582619921184944, + "grad_norm": 0.822160005569458, + "learning_rate": 0.00030217250985188203, + "loss": 3.4028, + "step": 82165 + }, + { + "epoch": 5.582959641255606, + "grad_norm": 1.0128660202026367, + "learning_rate": 0.00030213004484304937, + "loss": 3.5106, + "step": 82170 + }, + { + "epoch": 5.583299361326267, + "grad_norm": 0.8754316568374634, + "learning_rate": 0.0003020875798342166, + "loss": 3.6489, + "step": 82175 + }, + { + "epoch": 5.583639081396929, + "grad_norm": 0.8674492835998535, + "learning_rate": 0.0003020451148253839, + "loss": 3.1899, + "step": 82180 + }, + { + "epoch": 5.583978801467591, + "grad_norm": 0.9375045299530029, + "learning_rate": 0.00030200264981655115, + "loss": 3.3223, + "step": 82185 + }, + { + "epoch": 5.584318521538252, + "grad_norm": 0.963342547416687, + "learning_rate": 0.00030196018480771843, + "loss": 3.4384, + "step": 82190 + }, + { + "epoch": 5.584658241608914, + "grad_norm": 0.9302758574485779, + "learning_rate": 0.0003019177197988857, + "loss": 3.2992, + "step": 82195 + }, + { + "epoch": 5.584997961679576, + "grad_norm": 0.7429613471031189, + "learning_rate": 0.000301875254790053, + "loss": 3.7038, + "step": 82200 + }, + { + "epoch": 5.585337681750238, + "grad_norm": 0.8120657205581665, + "learning_rate": 0.00030183278978122033, + "loss": 3.4356, + "step": 82205 + }, + { + "epoch": 5.5856774018209, + "grad_norm": 0.8632986545562744, + "learning_rate": 0.00030179032477238755, + "loss": 3.1139, + "step": 82210 + }, + { + "epoch": 5.586017121891562, + "grad_norm": 0.9031504392623901, + "learning_rate": 0.00030174785976355483, + "loss": 3.5911, + "step": 82215 + }, + { + "epoch": 5.586356841962223, + "grad_norm": 0.8817985653877258, + "learning_rate": 0.0003017053947547221, + "loss": 3.3429, + "step": 82220 + }, + { + "epoch": 5.586696562032885, + "grad_norm": 1.2318918704986572, + "learning_rate": 0.0003016629297458894, + "loss": 3.5567, + "step": 82225 + }, + { + "epoch": 5.587036282103547, + "grad_norm": 0.9268613457679749, + "learning_rate": 0.0003016204647370567, + "loss": 3.4792, + "step": 82230 + }, + { + "epoch": 5.587376002174208, + "grad_norm": 0.9057188034057617, + "learning_rate": 0.00030157799972822395, + "loss": 3.6509, + "step": 82235 + }, + { + "epoch": 5.58771572224487, + "grad_norm": 1.0919981002807617, + "learning_rate": 0.00030153553471939123, + "loss": 3.2178, + "step": 82240 + }, + { + "epoch": 5.588055442315532, + "grad_norm": 0.8234882950782776, + "learning_rate": 0.0003014930697105585, + "loss": 3.5765, + "step": 82245 + }, + { + "epoch": 5.588395162386194, + "grad_norm": 0.8547834157943726, + "learning_rate": 0.0003014506047017258, + "loss": 3.434, + "step": 82250 + }, + { + "epoch": 5.588734882456856, + "grad_norm": 0.7923328876495361, + "learning_rate": 0.000301408139692893, + "loss": 3.6263, + "step": 82255 + }, + { + "epoch": 5.589074602527518, + "grad_norm": 0.7645610570907593, + "learning_rate": 0.00030136567468406035, + "loss": 3.7071, + "step": 82260 + }, + { + "epoch": 5.589414322598179, + "grad_norm": 0.8536590933799744, + "learning_rate": 0.00030132320967522763, + "loss": 3.3274, + "step": 82265 + }, + { + "epoch": 5.589754042668841, + "grad_norm": 1.1325973272323608, + "learning_rate": 0.00030128074466639486, + "loss": 3.5502, + "step": 82270 + }, + { + "epoch": 5.590093762739503, + "grad_norm": 0.9416400790214539, + "learning_rate": 0.0003012382796575622, + "loss": 3.3037, + "step": 82275 + }, + { + "epoch": 5.590433482810164, + "grad_norm": 1.1977800130844116, + "learning_rate": 0.0003011958146487295, + "loss": 3.3514, + "step": 82280 + }, + { + "epoch": 5.590773202880826, + "grad_norm": 0.8603570461273193, + "learning_rate": 0.0003011533496398967, + "loss": 3.1126, + "step": 82285 + }, + { + "epoch": 5.591112922951488, + "grad_norm": 1.0111265182495117, + "learning_rate": 0.000301110884631064, + "loss": 3.4435, + "step": 82290 + }, + { + "epoch": 5.59145264302215, + "grad_norm": 1.0420277118682861, + "learning_rate": 0.0003010684196222313, + "loss": 3.2959, + "step": 82295 + }, + { + "epoch": 5.591792363092812, + "grad_norm": 0.8204990029335022, + "learning_rate": 0.00030102595461339854, + "loss": 3.5823, + "step": 82300 + }, + { + "epoch": 5.592132083163474, + "grad_norm": 0.8032492995262146, + "learning_rate": 0.0003009834896045658, + "loss": 3.4152, + "step": 82305 + }, + { + "epoch": 5.592471803234135, + "grad_norm": 0.8204959630966187, + "learning_rate": 0.00030094102459573316, + "loss": 3.4062, + "step": 82310 + }, + { + "epoch": 5.592811523304797, + "grad_norm": 1.252389669418335, + "learning_rate": 0.0003008985595869004, + "loss": 3.5131, + "step": 82315 + }, + { + "epoch": 5.593151243375459, + "grad_norm": 0.7937635183334351, + "learning_rate": 0.00030085609457806766, + "loss": 3.373, + "step": 82320 + }, + { + "epoch": 5.59349096344612, + "grad_norm": 0.972148597240448, + "learning_rate": 0.000300813629569235, + "loss": 3.3431, + "step": 82325 + }, + { + "epoch": 5.593830683516782, + "grad_norm": 1.1939748525619507, + "learning_rate": 0.0003007711645604022, + "loss": 3.6258, + "step": 82330 + }, + { + "epoch": 5.594170403587444, + "grad_norm": 0.9143811464309692, + "learning_rate": 0.0003007286995515695, + "loss": 3.484, + "step": 82335 + }, + { + "epoch": 5.594510123658106, + "grad_norm": 1.2436470985412598, + "learning_rate": 0.0003006862345427368, + "loss": 3.646, + "step": 82340 + }, + { + "epoch": 5.594849843728768, + "grad_norm": 1.177282691001892, + "learning_rate": 0.00030064376953390406, + "loss": 3.4994, + "step": 82345 + }, + { + "epoch": 5.595189563799429, + "grad_norm": 0.8559488654136658, + "learning_rate": 0.00030060130452507134, + "loss": 3.1118, + "step": 82350 + }, + { + "epoch": 5.595529283870091, + "grad_norm": 0.9672789573669434, + "learning_rate": 0.0003005588395162386, + "loss": 3.7992, + "step": 82355 + }, + { + "epoch": 5.595869003940753, + "grad_norm": 0.804573655128479, + "learning_rate": 0.0003005163745074059, + "loss": 3.6284, + "step": 82360 + }, + { + "epoch": 5.596208724011414, + "grad_norm": 0.8855844140052795, + "learning_rate": 0.0003004739094985732, + "loss": 3.7174, + "step": 82365 + }, + { + "epoch": 5.596548444082076, + "grad_norm": 1.0248981714248657, + "learning_rate": 0.00030043144448974046, + "loss": 3.5449, + "step": 82370 + }, + { + "epoch": 5.596888164152738, + "grad_norm": 0.8718444108963013, + "learning_rate": 0.00030038897948090774, + "loss": 3.3219, + "step": 82375 + }, + { + "epoch": 5.5972278842234, + "grad_norm": 1.1217143535614014, + "learning_rate": 0.000300346514472075, + "loss": 3.5923, + "step": 82380 + }, + { + "epoch": 5.597567604294062, + "grad_norm": 1.0151934623718262, + "learning_rate": 0.0003003040494632423, + "loss": 3.2789, + "step": 82385 + }, + { + "epoch": 5.597907324364724, + "grad_norm": 0.8510819673538208, + "learning_rate": 0.0003002615844544096, + "loss": 3.4468, + "step": 82390 + }, + { + "epoch": 5.598247044435385, + "grad_norm": 0.717692494392395, + "learning_rate": 0.00030021911944557686, + "loss": 3.4964, + "step": 82395 + }, + { + "epoch": 5.598586764506047, + "grad_norm": 0.8688861131668091, + "learning_rate": 0.00030017665443674414, + "loss": 3.4789, + "step": 82400 + }, + { + "epoch": 5.598926484576709, + "grad_norm": 0.8709871172904968, + "learning_rate": 0.0003001341894279114, + "loss": 3.3047, + "step": 82405 + }, + { + "epoch": 5.59926620464737, + "grad_norm": 0.7539446949958801, + "learning_rate": 0.00030009172441907865, + "loss": 3.3955, + "step": 82410 + }, + { + "epoch": 5.599605924718032, + "grad_norm": 1.0892256498336792, + "learning_rate": 0.000300049259410246, + "loss": 3.6333, + "step": 82415 + }, + { + "epoch": 5.599945644788694, + "grad_norm": 0.9068273901939392, + "learning_rate": 0.00030000679440141326, + "loss": 3.5849, + "step": 82420 + }, + { + "epoch": 5.600285364859356, + "grad_norm": 0.8445501923561096, + "learning_rate": 0.0002999643293925805, + "loss": 3.4866, + "step": 82425 + }, + { + "epoch": 5.600625084930018, + "grad_norm": 0.8555974364280701, + "learning_rate": 0.0002999218643837478, + "loss": 3.4909, + "step": 82430 + }, + { + "epoch": 5.60096480500068, + "grad_norm": 0.8991062641143799, + "learning_rate": 0.0002998793993749151, + "loss": 3.4082, + "step": 82435 + }, + { + "epoch": 5.601304525071341, + "grad_norm": 0.8401778936386108, + "learning_rate": 0.00029983693436608233, + "loss": 3.3235, + "step": 82440 + }, + { + "epoch": 5.601644245142003, + "grad_norm": 0.7574227452278137, + "learning_rate": 0.0002997944693572496, + "loss": 3.62, + "step": 82445 + }, + { + "epoch": 5.601983965212665, + "grad_norm": 0.9731252789497375, + "learning_rate": 0.00029975200434841694, + "loss": 3.4487, + "step": 82450 + }, + { + "epoch": 5.602323685283326, + "grad_norm": 0.9891376495361328, + "learning_rate": 0.00029970953933958417, + "loss": 3.6082, + "step": 82455 + }, + { + "epoch": 5.602663405353988, + "grad_norm": 0.9286349415779114, + "learning_rate": 0.00029966707433075145, + "loss": 3.3799, + "step": 82460 + }, + { + "epoch": 5.60300312542465, + "grad_norm": 1.0710163116455078, + "learning_rate": 0.0002996246093219188, + "loss": 3.4492, + "step": 82465 + }, + { + "epoch": 5.603342845495312, + "grad_norm": 1.0180164575576782, + "learning_rate": 0.000299582144313086, + "loss": 3.6341, + "step": 82470 + }, + { + "epoch": 5.603682565565974, + "grad_norm": 0.9899030327796936, + "learning_rate": 0.0002995396793042533, + "loss": 3.5178, + "step": 82475 + }, + { + "epoch": 5.604022285636636, + "grad_norm": 0.8733294606208801, + "learning_rate": 0.00029949721429542057, + "loss": 3.4119, + "step": 82480 + }, + { + "epoch": 5.604362005707297, + "grad_norm": 0.8700166940689087, + "learning_rate": 0.00029945474928658785, + "loss": 3.3, + "step": 82485 + }, + { + "epoch": 5.604701725777959, + "grad_norm": 0.8279207944869995, + "learning_rate": 0.00029941228427775513, + "loss": 3.2474, + "step": 82490 + }, + { + "epoch": 5.605041445848621, + "grad_norm": 1.1284507513046265, + "learning_rate": 0.0002993698192689224, + "loss": 3.3856, + "step": 82495 + }, + { + "epoch": 5.605381165919282, + "grad_norm": 0.8666946291923523, + "learning_rate": 0.0002993273542600897, + "loss": 3.4129, + "step": 82500 + }, + { + "epoch": 5.605720885989944, + "grad_norm": 1.0889859199523926, + "learning_rate": 0.00029928488925125697, + "loss": 3.542, + "step": 82505 + }, + { + "epoch": 5.606060606060606, + "grad_norm": 0.8910405039787292, + "learning_rate": 0.00029924242424242425, + "loss": 3.5168, + "step": 82510 + }, + { + "epoch": 5.606400326131268, + "grad_norm": 0.6801512837409973, + "learning_rate": 0.0002991999592335915, + "loss": 3.5961, + "step": 82515 + }, + { + "epoch": 5.60674004620193, + "grad_norm": 0.8186582922935486, + "learning_rate": 0.0002991574942247588, + "loss": 3.5429, + "step": 82520 + }, + { + "epoch": 5.607079766272592, + "grad_norm": 0.7269889712333679, + "learning_rate": 0.0002991150292159261, + "loss": 3.3532, + "step": 82525 + }, + { + "epoch": 5.607419486343253, + "grad_norm": 1.0843061208724976, + "learning_rate": 0.0002990725642070933, + "loss": 3.6661, + "step": 82530 + }, + { + "epoch": 5.607759206413915, + "grad_norm": 0.6639176607131958, + "learning_rate": 0.00029903009919826065, + "loss": 3.4398, + "step": 82535 + }, + { + "epoch": 5.608098926484577, + "grad_norm": 1.049216389656067, + "learning_rate": 0.00029898763418942793, + "loss": 3.4465, + "step": 82540 + }, + { + "epoch": 5.608438646555238, + "grad_norm": 0.7577747106552124, + "learning_rate": 0.0002989451691805952, + "loss": 3.5817, + "step": 82545 + }, + { + "epoch": 5.6087783666259, + "grad_norm": 1.0697002410888672, + "learning_rate": 0.00029890270417176244, + "loss": 3.656, + "step": 82550 + }, + { + "epoch": 5.6091180866965615, + "grad_norm": 1.2251524925231934, + "learning_rate": 0.00029886023916292977, + "loss": 3.162, + "step": 82555 + }, + { + "epoch": 5.609457806767224, + "grad_norm": 0.6835396885871887, + "learning_rate": 0.00029881777415409705, + "loss": 3.376, + "step": 82560 + }, + { + "epoch": 5.609797526837886, + "grad_norm": 0.7857998013496399, + "learning_rate": 0.0002987753091452643, + "loss": 3.4854, + "step": 82565 + }, + { + "epoch": 5.610137246908547, + "grad_norm": 0.8086017370223999, + "learning_rate": 0.0002987328441364316, + "loss": 3.5351, + "step": 82570 + }, + { + "epoch": 5.610476966979209, + "grad_norm": 1.0776985883712769, + "learning_rate": 0.0002986903791275989, + "loss": 3.232, + "step": 82575 + }, + { + "epoch": 5.610816687049871, + "grad_norm": 0.8785591125488281, + "learning_rate": 0.0002986479141187661, + "loss": 3.4459, + "step": 82580 + }, + { + "epoch": 5.611156407120532, + "grad_norm": 0.907768964767456, + "learning_rate": 0.00029860544910993345, + "loss": 3.6141, + "step": 82585 + }, + { + "epoch": 5.611496127191194, + "grad_norm": 1.0870754718780518, + "learning_rate": 0.00029856298410110073, + "loss": 3.4037, + "step": 82590 + }, + { + "epoch": 5.611835847261856, + "grad_norm": 0.8357994556427002, + "learning_rate": 0.00029852051909226796, + "loss": 3.4175, + "step": 82595 + }, + { + "epoch": 5.6121755673325175, + "grad_norm": 0.8477683663368225, + "learning_rate": 0.00029847805408343524, + "loss": 3.2592, + "step": 82600 + }, + { + "epoch": 5.61251528740318, + "grad_norm": 0.8902766108512878, + "learning_rate": 0.00029843558907460257, + "loss": 3.3841, + "step": 82605 + }, + { + "epoch": 5.612855007473842, + "grad_norm": 0.9501713514328003, + "learning_rate": 0.0002983931240657698, + "loss": 3.3027, + "step": 82610 + }, + { + "epoch": 5.613194727544503, + "grad_norm": 1.094098448753357, + "learning_rate": 0.0002983506590569371, + "loss": 3.3747, + "step": 82615 + }, + { + "epoch": 5.613534447615165, + "grad_norm": 1.1341629028320312, + "learning_rate": 0.0002983081940481044, + "loss": 3.4693, + "step": 82620 + }, + { + "epoch": 5.613874167685827, + "grad_norm": 1.0982837677001953, + "learning_rate": 0.00029826572903927164, + "loss": 3.6718, + "step": 82625 + }, + { + "epoch": 5.614213887756488, + "grad_norm": 1.18268883228302, + "learning_rate": 0.0002982232640304389, + "loss": 3.5742, + "step": 82630 + }, + { + "epoch": 5.61455360782715, + "grad_norm": 0.7063042521476746, + "learning_rate": 0.0002981807990216062, + "loss": 3.317, + "step": 82635 + }, + { + "epoch": 5.614893327897812, + "grad_norm": 0.8306232690811157, + "learning_rate": 0.0002981383340127735, + "loss": 3.2071, + "step": 82640 + }, + { + "epoch": 5.615233047968474, + "grad_norm": 0.9124130010604858, + "learning_rate": 0.00029809586900394076, + "loss": 3.4146, + "step": 82645 + }, + { + "epoch": 5.615572768039136, + "grad_norm": 0.8041988015174866, + "learning_rate": 0.00029805340399510804, + "loss": 3.4698, + "step": 82650 + }, + { + "epoch": 5.615912488109798, + "grad_norm": 0.9369439482688904, + "learning_rate": 0.0002980109389862753, + "loss": 3.43, + "step": 82655 + }, + { + "epoch": 5.616252208180459, + "grad_norm": 0.7106771469116211, + "learning_rate": 0.0002979684739774426, + "loss": 3.5662, + "step": 82660 + }, + { + "epoch": 5.616591928251121, + "grad_norm": 0.9854356050491333, + "learning_rate": 0.0002979260089686099, + "loss": 3.3666, + "step": 82665 + }, + { + "epoch": 5.616931648321783, + "grad_norm": 1.1525728702545166, + "learning_rate": 0.0002978835439597771, + "loss": 3.1969, + "step": 82670 + }, + { + "epoch": 5.617271368392444, + "grad_norm": 0.8339700698852539, + "learning_rate": 0.00029784107895094444, + "loss": 3.4877, + "step": 82675 + }, + { + "epoch": 5.617611088463106, + "grad_norm": 0.9361932277679443, + "learning_rate": 0.0002977986139421117, + "loss": 3.4459, + "step": 82680 + }, + { + "epoch": 5.617950808533768, + "grad_norm": 1.0960323810577393, + "learning_rate": 0.00029775614893327894, + "loss": 3.5247, + "step": 82685 + }, + { + "epoch": 5.61829052860443, + "grad_norm": 0.8406267762184143, + "learning_rate": 0.0002977136839244463, + "loss": 3.3203, + "step": 82690 + }, + { + "epoch": 5.618630248675092, + "grad_norm": 0.7366962432861328, + "learning_rate": 0.00029767121891561356, + "loss": 3.4736, + "step": 82695 + }, + { + "epoch": 5.618969968745754, + "grad_norm": 1.05165433883667, + "learning_rate": 0.0002976287539067808, + "loss": 3.7444, + "step": 82700 + }, + { + "epoch": 5.619309688816415, + "grad_norm": 1.192793369293213, + "learning_rate": 0.00029758628889794806, + "loss": 3.5398, + "step": 82705 + }, + { + "epoch": 5.619649408887077, + "grad_norm": 0.9434531927108765, + "learning_rate": 0.0002975438238891154, + "loss": 3.5305, + "step": 82710 + }, + { + "epoch": 5.619989128957739, + "grad_norm": 0.7988432049751282, + "learning_rate": 0.0002975013588802827, + "loss": 3.3358, + "step": 82715 + }, + { + "epoch": 5.6203288490284, + "grad_norm": 0.8350309729576111, + "learning_rate": 0.0002974588938714499, + "loss": 3.1934, + "step": 82720 + }, + { + "epoch": 5.620668569099062, + "grad_norm": 0.783007800579071, + "learning_rate": 0.00029741642886261724, + "loss": 3.6222, + "step": 82725 + }, + { + "epoch": 5.621008289169724, + "grad_norm": 1.0800529718399048, + "learning_rate": 0.0002973739638537845, + "loss": 3.2762, + "step": 82730 + }, + { + "epoch": 5.621348009240386, + "grad_norm": 0.7524433135986328, + "learning_rate": 0.00029733149884495175, + "loss": 3.62, + "step": 82735 + }, + { + "epoch": 5.621687729311048, + "grad_norm": 0.7950888276100159, + "learning_rate": 0.000297289033836119, + "loss": 3.5863, + "step": 82740 + }, + { + "epoch": 5.62202744938171, + "grad_norm": 0.914823591709137, + "learning_rate": 0.00029724656882728636, + "loss": 3.3415, + "step": 82745 + }, + { + "epoch": 5.622367169452371, + "grad_norm": 0.7424940466880798, + "learning_rate": 0.0002972041038184536, + "loss": 3.4915, + "step": 82750 + }, + { + "epoch": 5.622706889523033, + "grad_norm": 0.9059765338897705, + "learning_rate": 0.00029716163880962087, + "loss": 3.5569, + "step": 82755 + }, + { + "epoch": 5.623046609593695, + "grad_norm": 0.9582024812698364, + "learning_rate": 0.0002971191738007882, + "loss": 3.5173, + "step": 82760 + }, + { + "epoch": 5.623386329664356, + "grad_norm": 0.879798173904419, + "learning_rate": 0.0002970767087919554, + "loss": 3.4334, + "step": 82765 + }, + { + "epoch": 5.623726049735018, + "grad_norm": 1.0461952686309814, + "learning_rate": 0.0002970342437831227, + "loss": 3.4366, + "step": 82770 + }, + { + "epoch": 5.62406576980568, + "grad_norm": 1.1164079904556274, + "learning_rate": 0.00029699177877429, + "loss": 3.3137, + "step": 82775 + }, + { + "epoch": 5.624405489876342, + "grad_norm": 1.247663140296936, + "learning_rate": 0.00029694931376545727, + "loss": 3.5439, + "step": 82780 + }, + { + "epoch": 5.624745209947004, + "grad_norm": 1.0324838161468506, + "learning_rate": 0.00029690684875662455, + "loss": 3.8394, + "step": 82785 + }, + { + "epoch": 5.625084930017666, + "grad_norm": 0.7930219173431396, + "learning_rate": 0.0002968643837477918, + "loss": 3.4222, + "step": 82790 + }, + { + "epoch": 5.625424650088327, + "grad_norm": 0.9668696522712708, + "learning_rate": 0.0002968219187389591, + "loss": 3.4342, + "step": 82795 + }, + { + "epoch": 5.625764370158989, + "grad_norm": 1.05758798122406, + "learning_rate": 0.0002967794537301264, + "loss": 3.6813, + "step": 82800 + }, + { + "epoch": 5.626104090229651, + "grad_norm": 0.9212996959686279, + "learning_rate": 0.00029673698872129367, + "loss": 3.5015, + "step": 82805 + }, + { + "epoch": 5.626443810300312, + "grad_norm": 1.0778874158859253, + "learning_rate": 0.0002966945237124609, + "loss": 3.3348, + "step": 82810 + }, + { + "epoch": 5.626783530370974, + "grad_norm": 1.132836103439331, + "learning_rate": 0.0002966520587036282, + "loss": 3.4537, + "step": 82815 + }, + { + "epoch": 5.627123250441636, + "grad_norm": 1.0223679542541504, + "learning_rate": 0.0002966095936947955, + "loss": 3.2584, + "step": 82820 + }, + { + "epoch": 5.627462970512298, + "grad_norm": 0.9438114166259766, + "learning_rate": 0.00029656712868596273, + "loss": 3.3805, + "step": 82825 + }, + { + "epoch": 5.62780269058296, + "grad_norm": 0.9413001537322998, + "learning_rate": 0.00029652466367713007, + "loss": 3.499, + "step": 82830 + }, + { + "epoch": 5.628142410653622, + "grad_norm": 0.8503909707069397, + "learning_rate": 0.00029648219866829735, + "loss": 3.2795, + "step": 82835 + }, + { + "epoch": 5.628482130724283, + "grad_norm": 0.7844239473342896, + "learning_rate": 0.00029643973365946457, + "loss": 3.5971, + "step": 82840 + }, + { + "epoch": 5.628821850794945, + "grad_norm": 0.8048662543296814, + "learning_rate": 0.00029639726865063185, + "loss": 3.5358, + "step": 82845 + }, + { + "epoch": 5.629161570865607, + "grad_norm": 0.7842504978179932, + "learning_rate": 0.0002963548036417992, + "loss": 3.3627, + "step": 82850 + }, + { + "epoch": 5.629501290936268, + "grad_norm": 1.1938397884368896, + "learning_rate": 0.0002963123386329664, + "loss": 3.8508, + "step": 82855 + }, + { + "epoch": 5.62984101100693, + "grad_norm": 1.0448285341262817, + "learning_rate": 0.0002962698736241337, + "loss": 3.4857, + "step": 82860 + }, + { + "epoch": 5.630180731077592, + "grad_norm": 0.67790287733078, + "learning_rate": 0.00029622740861530103, + "loss": 3.2893, + "step": 82865 + }, + { + "epoch": 5.630520451148254, + "grad_norm": 0.7852063775062561, + "learning_rate": 0.00029618494360646825, + "loss": 3.4405, + "step": 82870 + }, + { + "epoch": 5.630860171218916, + "grad_norm": 1.1292860507965088, + "learning_rate": 0.00029614247859763553, + "loss": 3.3046, + "step": 82875 + }, + { + "epoch": 5.631199891289578, + "grad_norm": 1.324384093284607, + "learning_rate": 0.00029610001358880287, + "loss": 3.6211, + "step": 82880 + }, + { + "epoch": 5.631539611360239, + "grad_norm": 0.8519487380981445, + "learning_rate": 0.00029605754857997015, + "loss": 3.3598, + "step": 82885 + }, + { + "epoch": 5.631879331430901, + "grad_norm": 1.4642263650894165, + "learning_rate": 0.0002960150835711374, + "loss": 3.566, + "step": 82890 + }, + { + "epoch": 5.632219051501563, + "grad_norm": 1.4540491104125977, + "learning_rate": 0.00029597261856230465, + "loss": 3.3431, + "step": 82895 + }, + { + "epoch": 5.632558771572224, + "grad_norm": 1.0076764822006226, + "learning_rate": 0.000295930153553472, + "loss": 3.4225, + "step": 82900 + }, + { + "epoch": 5.632898491642886, + "grad_norm": 0.6310834884643555, + "learning_rate": 0.0002958876885446392, + "loss": 3.6629, + "step": 82905 + }, + { + "epoch": 5.633238211713548, + "grad_norm": 0.9057756662368774, + "learning_rate": 0.0002958452235358065, + "loss": 3.5102, + "step": 82910 + }, + { + "epoch": 5.63357793178421, + "grad_norm": 0.8441143035888672, + "learning_rate": 0.00029580275852697383, + "loss": 3.6229, + "step": 82915 + }, + { + "epoch": 5.633917651854872, + "grad_norm": 0.921527624130249, + "learning_rate": 0.00029576029351814105, + "loss": 3.4045, + "step": 82920 + }, + { + "epoch": 5.634257371925534, + "grad_norm": 0.8931593894958496, + "learning_rate": 0.00029571782850930833, + "loss": 3.2953, + "step": 82925 + }, + { + "epoch": 5.634597091996195, + "grad_norm": 1.0581187009811401, + "learning_rate": 0.0002956753635004756, + "loss": 3.4309, + "step": 82930 + }, + { + "epoch": 5.634936812066857, + "grad_norm": 1.0409610271453857, + "learning_rate": 0.0002956328984916429, + "loss": 3.4529, + "step": 82935 + }, + { + "epoch": 5.635276532137519, + "grad_norm": 0.7524537444114685, + "learning_rate": 0.0002955904334828102, + "loss": 2.9636, + "step": 82940 + }, + { + "epoch": 5.63561625220818, + "grad_norm": 1.0245060920715332, + "learning_rate": 0.00029554796847397745, + "loss": 3.4309, + "step": 82945 + }, + { + "epoch": 5.635955972278842, + "grad_norm": 0.9190667271614075, + "learning_rate": 0.00029550550346514473, + "loss": 3.465, + "step": 82950 + }, + { + "epoch": 5.6362956923495044, + "grad_norm": 0.7650712728500366, + "learning_rate": 0.000295463038456312, + "loss": 3.4938, + "step": 82955 + }, + { + "epoch": 5.636635412420166, + "grad_norm": 1.1156895160675049, + "learning_rate": 0.0002954205734474793, + "loss": 3.7372, + "step": 82960 + }, + { + "epoch": 5.636975132490828, + "grad_norm": 0.7459725737571716, + "learning_rate": 0.0002953781084386465, + "loss": 3.2657, + "step": 82965 + }, + { + "epoch": 5.63731485256149, + "grad_norm": 0.7078388929367065, + "learning_rate": 0.00029533564342981385, + "loss": 3.4057, + "step": 82970 + }, + { + "epoch": 5.637654572632151, + "grad_norm": 0.9115403890609741, + "learning_rate": 0.00029529317842098113, + "loss": 3.8266, + "step": 82975 + }, + { + "epoch": 5.637994292702813, + "grad_norm": 0.9840308427810669, + "learning_rate": 0.00029525071341214836, + "loss": 3.3176, + "step": 82980 + }, + { + "epoch": 5.638334012773475, + "grad_norm": 1.2488564252853394, + "learning_rate": 0.0002952082484033157, + "loss": 3.4897, + "step": 82985 + }, + { + "epoch": 5.638673732844136, + "grad_norm": 0.9639122486114502, + "learning_rate": 0.000295165783394483, + "loss": 3.5619, + "step": 82990 + }, + { + "epoch": 5.639013452914798, + "grad_norm": 0.7075731158256531, + "learning_rate": 0.0002951233183856502, + "loss": 3.2407, + "step": 82995 + }, + { + "epoch": 5.6393531729854605, + "grad_norm": 0.7933985590934753, + "learning_rate": 0.0002950808533768175, + "loss": 3.5547, + "step": 83000 + }, + { + "epoch": 5.639692893056122, + "grad_norm": 0.8658381700515747, + "learning_rate": 0.0002950383883679848, + "loss": 3.3827, + "step": 83005 + }, + { + "epoch": 5.640032613126784, + "grad_norm": 0.905755341053009, + "learning_rate": 0.00029499592335915204, + "loss": 3.6374, + "step": 83010 + }, + { + "epoch": 5.640372333197446, + "grad_norm": 1.0243123769760132, + "learning_rate": 0.0002949534583503193, + "loss": 3.5549, + "step": 83015 + }, + { + "epoch": 5.640712053268107, + "grad_norm": 1.1589465141296387, + "learning_rate": 0.00029491099334148666, + "loss": 3.2596, + "step": 83020 + }, + { + "epoch": 5.641051773338769, + "grad_norm": 0.9430339932441711, + "learning_rate": 0.0002948685283326539, + "loss": 3.6477, + "step": 83025 + }, + { + "epoch": 5.64139149340943, + "grad_norm": 0.7901031970977783, + "learning_rate": 0.00029482606332382116, + "loss": 3.444, + "step": 83030 + }, + { + "epoch": 5.641731213480092, + "grad_norm": 0.9946638345718384, + "learning_rate": 0.00029478359831498844, + "loss": 3.4469, + "step": 83035 + }, + { + "epoch": 5.642070933550754, + "grad_norm": 0.8680121302604675, + "learning_rate": 0.0002947411333061557, + "loss": 3.6005, + "step": 83040 + }, + { + "epoch": 5.642410653621416, + "grad_norm": 1.1426748037338257, + "learning_rate": 0.000294698668297323, + "loss": 3.3126, + "step": 83045 + }, + { + "epoch": 5.642750373692078, + "grad_norm": 0.8053727149963379, + "learning_rate": 0.0002946562032884903, + "loss": 3.5187, + "step": 83050 + }, + { + "epoch": 5.64309009376274, + "grad_norm": 0.9972378015518188, + "learning_rate": 0.0002946137382796576, + "loss": 3.5392, + "step": 83055 + }, + { + "epoch": 5.643429813833401, + "grad_norm": 0.9491370320320129, + "learning_rate": 0.00029457127327082484, + "loss": 3.5256, + "step": 83060 + }, + { + "epoch": 5.643769533904063, + "grad_norm": 0.7507970929145813, + "learning_rate": 0.0002945288082619921, + "loss": 3.5311, + "step": 83065 + }, + { + "epoch": 5.644109253974725, + "grad_norm": 0.7628971338272095, + "learning_rate": 0.0002944863432531594, + "loss": 3.3957, + "step": 83070 + }, + { + "epoch": 5.644448974045386, + "grad_norm": 0.7985711693763733, + "learning_rate": 0.0002944438782443267, + "loss": 3.3527, + "step": 83075 + }, + { + "epoch": 5.644788694116048, + "grad_norm": 0.8339905738830566, + "learning_rate": 0.00029440141323549396, + "loss": 3.2678, + "step": 83080 + }, + { + "epoch": 5.64512841418671, + "grad_norm": 1.2876183986663818, + "learning_rate": 0.00029435894822666124, + "loss": 3.3757, + "step": 83085 + }, + { + "epoch": 5.645468134257372, + "grad_norm": 0.9001734256744385, + "learning_rate": 0.0002943164832178285, + "loss": 3.4509, + "step": 83090 + }, + { + "epoch": 5.645807854328034, + "grad_norm": 0.8219545483589172, + "learning_rate": 0.0002942740182089958, + "loss": 3.426, + "step": 83095 + }, + { + "epoch": 5.646147574398696, + "grad_norm": 0.9660037755966187, + "learning_rate": 0.0002942315532001631, + "loss": 3.5295, + "step": 83100 + }, + { + "epoch": 5.646487294469357, + "grad_norm": 1.0905600786209106, + "learning_rate": 0.0002941890881913303, + "loss": 3.3748, + "step": 83105 + }, + { + "epoch": 5.646827014540019, + "grad_norm": 0.7156327366828918, + "learning_rate": 0.00029414662318249764, + "loss": 3.3843, + "step": 83110 + }, + { + "epoch": 5.647166734610681, + "grad_norm": 0.968073308467865, + "learning_rate": 0.0002941041581736649, + "loss": 3.7197, + "step": 83115 + }, + { + "epoch": 5.647506454681342, + "grad_norm": 0.8887528777122498, + "learning_rate": 0.00029406169316483215, + "loss": 3.4534, + "step": 83120 + }, + { + "epoch": 5.647846174752004, + "grad_norm": 1.0617400407791138, + "learning_rate": 0.0002940192281559995, + "loss": 3.383, + "step": 83125 + }, + { + "epoch": 5.648185894822666, + "grad_norm": 0.9726810455322266, + "learning_rate": 0.00029397676314716676, + "loss": 3.5561, + "step": 83130 + }, + { + "epoch": 5.648525614893328, + "grad_norm": 0.8281834125518799, + "learning_rate": 0.000293934298138334, + "loss": 3.4291, + "step": 83135 + }, + { + "epoch": 5.64886533496399, + "grad_norm": 0.8042598366737366, + "learning_rate": 0.00029389183312950127, + "loss": 3.4293, + "step": 83140 + }, + { + "epoch": 5.649205055034652, + "grad_norm": 0.8832095265388489, + "learning_rate": 0.0002938493681206686, + "loss": 3.1667, + "step": 83145 + }, + { + "epoch": 5.649544775105313, + "grad_norm": 0.7883518934249878, + "learning_rate": 0.00029380690311183583, + "loss": 3.5576, + "step": 83150 + }, + { + "epoch": 5.649884495175975, + "grad_norm": 0.9501177668571472, + "learning_rate": 0.0002937644381030031, + "loss": 3.4591, + "step": 83155 + }, + { + "epoch": 5.650224215246637, + "grad_norm": 1.1314674615859985, + "learning_rate": 0.00029372197309417044, + "loss": 3.7386, + "step": 83160 + }, + { + "epoch": 5.650563935317298, + "grad_norm": 0.9495134949684143, + "learning_rate": 0.00029367950808533767, + "loss": 3.5485, + "step": 83165 + }, + { + "epoch": 5.65090365538796, + "grad_norm": 0.868606448173523, + "learning_rate": 0.00029363704307650495, + "loss": 3.4141, + "step": 83170 + }, + { + "epoch": 5.651243375458622, + "grad_norm": 1.0000779628753662, + "learning_rate": 0.0002935945780676723, + "loss": 3.5225, + "step": 83175 + }, + { + "epoch": 5.651583095529284, + "grad_norm": 1.0089242458343506, + "learning_rate": 0.0002935521130588395, + "loss": 3.2065, + "step": 83180 + }, + { + "epoch": 5.651922815599946, + "grad_norm": 1.2553049325942993, + "learning_rate": 0.0002935096480500068, + "loss": 3.4016, + "step": 83185 + }, + { + "epoch": 5.652262535670608, + "grad_norm": 1.1665568351745605, + "learning_rate": 0.00029346718304117407, + "loss": 3.3261, + "step": 83190 + }, + { + "epoch": 5.652602255741269, + "grad_norm": 0.8038183450698853, + "learning_rate": 0.00029342471803234135, + "loss": 3.3971, + "step": 83195 + }, + { + "epoch": 5.652941975811931, + "grad_norm": 0.7910287380218506, + "learning_rate": 0.00029338225302350863, + "loss": 3.4475, + "step": 83200 + }, + { + "epoch": 5.653281695882593, + "grad_norm": 0.7997781038284302, + "learning_rate": 0.0002933397880146759, + "loss": 3.3672, + "step": 83205 + }, + { + "epoch": 5.653621415953254, + "grad_norm": 0.9859561920166016, + "learning_rate": 0.0002932973230058432, + "loss": 3.6528, + "step": 83210 + }, + { + "epoch": 5.653961136023916, + "grad_norm": 0.9638369083404541, + "learning_rate": 0.00029325485799701047, + "loss": 3.4317, + "step": 83215 + }, + { + "epoch": 5.654300856094578, + "grad_norm": 0.7725545167922974, + "learning_rate": 0.00029321239298817775, + "loss": 3.5945, + "step": 83220 + }, + { + "epoch": 5.65464057616524, + "grad_norm": 0.8084056377410889, + "learning_rate": 0.00029316992797934503, + "loss": 3.4803, + "step": 83225 + }, + { + "epoch": 5.654980296235902, + "grad_norm": 0.8588467240333557, + "learning_rate": 0.0002931274629705123, + "loss": 3.2939, + "step": 83230 + }, + { + "epoch": 5.655320016306564, + "grad_norm": 0.9415102005004883, + "learning_rate": 0.0002930849979616796, + "loss": 3.395, + "step": 83235 + }, + { + "epoch": 5.655659736377225, + "grad_norm": 1.0770758390426636, + "learning_rate": 0.00029304253295284687, + "loss": 3.4179, + "step": 83240 + }, + { + "epoch": 5.655999456447887, + "grad_norm": 0.8784584403038025, + "learning_rate": 0.00029300006794401415, + "loss": 3.3974, + "step": 83245 + }, + { + "epoch": 5.656339176518548, + "grad_norm": 0.7465420365333557, + "learning_rate": 0.00029295760293518143, + "loss": 3.1769, + "step": 83250 + }, + { + "epoch": 5.65667889658921, + "grad_norm": 1.0063602924346924, + "learning_rate": 0.0002929151379263487, + "loss": 3.7082, + "step": 83255 + }, + { + "epoch": 5.657018616659872, + "grad_norm": 0.9924637675285339, + "learning_rate": 0.00029287267291751594, + "loss": 3.2542, + "step": 83260 + }, + { + "epoch": 5.657358336730534, + "grad_norm": 1.0415393114089966, + "learning_rate": 0.00029283020790868327, + "loss": 3.625, + "step": 83265 + }, + { + "epoch": 5.657698056801196, + "grad_norm": 0.7322259545326233, + "learning_rate": 0.00029278774289985055, + "loss": 3.6835, + "step": 83270 + }, + { + "epoch": 5.658037776871858, + "grad_norm": 0.6387506723403931, + "learning_rate": 0.0002927452778910178, + "loss": 3.1596, + "step": 83275 + }, + { + "epoch": 5.658377496942519, + "grad_norm": 0.8852750062942505, + "learning_rate": 0.0002927028128821851, + "loss": 3.3895, + "step": 83280 + }, + { + "epoch": 5.658717217013181, + "grad_norm": 1.015770673751831, + "learning_rate": 0.0002926603478733524, + "loss": 3.249, + "step": 83285 + }, + { + "epoch": 5.659056937083843, + "grad_norm": 0.9906133413314819, + "learning_rate": 0.0002926178828645196, + "loss": 3.519, + "step": 83290 + }, + { + "epoch": 5.659396657154504, + "grad_norm": 0.9971023201942444, + "learning_rate": 0.0002925754178556869, + "loss": 3.0304, + "step": 83295 + }, + { + "epoch": 5.659736377225166, + "grad_norm": 0.9242939949035645, + "learning_rate": 0.00029253295284685423, + "loss": 3.3278, + "step": 83300 + }, + { + "epoch": 5.660076097295828, + "grad_norm": 0.8084380030632019, + "learning_rate": 0.00029249048783802146, + "loss": 3.2948, + "step": 83305 + }, + { + "epoch": 5.66041581736649, + "grad_norm": 0.9707916975021362, + "learning_rate": 0.00029244802282918874, + "loss": 3.231, + "step": 83310 + }, + { + "epoch": 5.660755537437152, + "grad_norm": 0.7198179960250854, + "learning_rate": 0.00029240555782035607, + "loss": 3.3629, + "step": 83315 + }, + { + "epoch": 5.661095257507814, + "grad_norm": 1.0670853853225708, + "learning_rate": 0.0002923630928115233, + "loss": 3.5328, + "step": 83320 + }, + { + "epoch": 5.661434977578475, + "grad_norm": 0.9983049035072327, + "learning_rate": 0.0002923206278026906, + "loss": 3.4819, + "step": 83325 + }, + { + "epoch": 5.661774697649137, + "grad_norm": 0.8845999836921692, + "learning_rate": 0.00029227816279385786, + "loss": 3.6444, + "step": 83330 + }, + { + "epoch": 5.662114417719799, + "grad_norm": 0.780309796333313, + "learning_rate": 0.00029223569778502514, + "loss": 3.2808, + "step": 83335 + }, + { + "epoch": 5.66245413779046, + "grad_norm": 0.7731478214263916, + "learning_rate": 0.0002921932327761924, + "loss": 3.5712, + "step": 83340 + }, + { + "epoch": 5.662793857861122, + "grad_norm": 0.9698320031166077, + "learning_rate": 0.0002921507677673597, + "loss": 3.5264, + "step": 83345 + }, + { + "epoch": 5.663133577931784, + "grad_norm": 1.28143310546875, + "learning_rate": 0.000292108302758527, + "loss": 3.3969, + "step": 83350 + }, + { + "epoch": 5.663473298002446, + "grad_norm": 0.8373331427574158, + "learning_rate": 0.00029206583774969426, + "loss": 3.5784, + "step": 83355 + }, + { + "epoch": 5.663813018073108, + "grad_norm": 0.8787707686424255, + "learning_rate": 0.00029202337274086154, + "loss": 3.578, + "step": 83360 + }, + { + "epoch": 5.66415273814377, + "grad_norm": 1.591157078742981, + "learning_rate": 0.00029198090773202876, + "loss": 3.3844, + "step": 83365 + }, + { + "epoch": 5.664492458214431, + "grad_norm": 0.9360778331756592, + "learning_rate": 0.0002919384427231961, + "loss": 3.2784, + "step": 83370 + }, + { + "epoch": 5.664832178285093, + "grad_norm": 0.7707211971282959, + "learning_rate": 0.0002918959777143634, + "loss": 3.3723, + "step": 83375 + }, + { + "epoch": 5.665171898355755, + "grad_norm": 0.8232119679450989, + "learning_rate": 0.0002918535127055306, + "loss": 3.3123, + "step": 83380 + }, + { + "epoch": 5.665511618426416, + "grad_norm": 0.907619833946228, + "learning_rate": 0.00029181104769669794, + "loss": 3.2408, + "step": 83385 + }, + { + "epoch": 5.665851338497078, + "grad_norm": 0.7776772379875183, + "learning_rate": 0.0002917685826878652, + "loss": 3.6471, + "step": 83390 + }, + { + "epoch": 5.66619105856774, + "grad_norm": 0.8521826267242432, + "learning_rate": 0.0002917261176790325, + "loss": 3.4327, + "step": 83395 + }, + { + "epoch": 5.666530778638402, + "grad_norm": 1.1201339960098267, + "learning_rate": 0.0002916836526701997, + "loss": 3.1758, + "step": 83400 + }, + { + "epoch": 5.666870498709064, + "grad_norm": 0.9783396124839783, + "learning_rate": 0.00029164118766136706, + "loss": 3.336, + "step": 83405 + }, + { + "epoch": 5.667210218779726, + "grad_norm": 0.8204969763755798, + "learning_rate": 0.00029159872265253434, + "loss": 3.2709, + "step": 83410 + }, + { + "epoch": 5.667549938850387, + "grad_norm": 0.793096661567688, + "learning_rate": 0.00029155625764370156, + "loss": 3.6273, + "step": 83415 + }, + { + "epoch": 5.667889658921049, + "grad_norm": 0.9060960412025452, + "learning_rate": 0.0002915137926348689, + "loss": 3.4827, + "step": 83420 + }, + { + "epoch": 5.668229378991711, + "grad_norm": 0.9079042077064514, + "learning_rate": 0.0002914713276260362, + "loss": 3.3715, + "step": 83425 + }, + { + "epoch": 5.668569099062372, + "grad_norm": 1.024854302406311, + "learning_rate": 0.0002914288626172034, + "loss": 3.4902, + "step": 83430 + }, + { + "epoch": 5.668908819133034, + "grad_norm": 0.8753691911697388, + "learning_rate": 0.00029138639760837074, + "loss": 3.6279, + "step": 83435 + }, + { + "epoch": 5.669248539203696, + "grad_norm": 1.3571269512176514, + "learning_rate": 0.000291343932599538, + "loss": 3.4124, + "step": 83440 + }, + { + "epoch": 5.669588259274358, + "grad_norm": 0.9249932169914246, + "learning_rate": 0.00029130146759070525, + "loss": 3.5921, + "step": 83445 + }, + { + "epoch": 5.66992797934502, + "grad_norm": 0.8799342513084412, + "learning_rate": 0.0002912590025818725, + "loss": 3.243, + "step": 83450 + }, + { + "epoch": 5.670267699415682, + "grad_norm": 0.947935938835144, + "learning_rate": 0.00029121653757303986, + "loss": 3.4729, + "step": 83455 + }, + { + "epoch": 5.670607419486343, + "grad_norm": 1.3947057723999023, + "learning_rate": 0.0002911740725642071, + "loss": 3.5035, + "step": 83460 + }, + { + "epoch": 5.670947139557005, + "grad_norm": 0.8693006038665771, + "learning_rate": 0.00029113160755537437, + "loss": 3.6941, + "step": 83465 + }, + { + "epoch": 5.671286859627667, + "grad_norm": 1.843916893005371, + "learning_rate": 0.0002910891425465417, + "loss": 3.4625, + "step": 83470 + }, + { + "epoch": 5.671626579698328, + "grad_norm": 0.9616157412528992, + "learning_rate": 0.0002910466775377089, + "loss": 3.1795, + "step": 83475 + }, + { + "epoch": 5.67196629976899, + "grad_norm": 1.0847221612930298, + "learning_rate": 0.0002910042125288762, + "loss": 3.237, + "step": 83480 + }, + { + "epoch": 5.672306019839652, + "grad_norm": 0.9928697347640991, + "learning_rate": 0.0002909617475200435, + "loss": 3.4038, + "step": 83485 + }, + { + "epoch": 5.672645739910314, + "grad_norm": 0.8215965628623962, + "learning_rate": 0.00029091928251121077, + "loss": 3.6302, + "step": 83490 + }, + { + "epoch": 5.672985459980976, + "grad_norm": 1.1190478801727295, + "learning_rate": 0.00029087681750237805, + "loss": 3.384, + "step": 83495 + }, + { + "epoch": 5.673325180051638, + "grad_norm": 0.7992681860923767, + "learning_rate": 0.0002908343524935453, + "loss": 3.4899, + "step": 83500 + }, + { + "epoch": 5.673664900122299, + "grad_norm": 1.1197290420532227, + "learning_rate": 0.0002907918874847126, + "loss": 3.3736, + "step": 83505 + }, + { + "epoch": 5.674004620192961, + "grad_norm": 0.9365699887275696, + "learning_rate": 0.0002907494224758799, + "loss": 3.4236, + "step": 83510 + }, + { + "epoch": 5.674344340263623, + "grad_norm": 1.0207115411758423, + "learning_rate": 0.00029070695746704717, + "loss": 3.5253, + "step": 83515 + }, + { + "epoch": 5.674684060334284, + "grad_norm": 0.7943755984306335, + "learning_rate": 0.0002906644924582144, + "loss": 3.507, + "step": 83520 + }, + { + "epoch": 5.675023780404946, + "grad_norm": 0.694978654384613, + "learning_rate": 0.0002906220274493817, + "loss": 3.201, + "step": 83525 + }, + { + "epoch": 5.6753635004756084, + "grad_norm": 0.8559492230415344, + "learning_rate": 0.000290579562440549, + "loss": 3.6477, + "step": 83530 + }, + { + "epoch": 5.67570322054627, + "grad_norm": 1.0569499731063843, + "learning_rate": 0.00029053709743171623, + "loss": 3.4724, + "step": 83535 + }, + { + "epoch": 5.676042940616932, + "grad_norm": 0.947289228439331, + "learning_rate": 0.00029049463242288357, + "loss": 3.5101, + "step": 83540 + }, + { + "epoch": 5.676382660687594, + "grad_norm": 1.0206961631774902, + "learning_rate": 0.00029045216741405085, + "loss": 3.364, + "step": 83545 + }, + { + "epoch": 5.676722380758255, + "grad_norm": 1.0054506063461304, + "learning_rate": 0.00029040970240521807, + "loss": 3.1569, + "step": 83550 + }, + { + "epoch": 5.677062100828917, + "grad_norm": 0.8027119636535645, + "learning_rate": 0.00029036723739638535, + "loss": 3.3997, + "step": 83555 + }, + { + "epoch": 5.677401820899579, + "grad_norm": 0.8249996304512024, + "learning_rate": 0.0002903247723875527, + "loss": 3.3876, + "step": 83560 + }, + { + "epoch": 5.67774154097024, + "grad_norm": 0.9305970668792725, + "learning_rate": 0.00029028230737871997, + "loss": 3.4435, + "step": 83565 + }, + { + "epoch": 5.678081261040902, + "grad_norm": 1.16923189163208, + "learning_rate": 0.0002902398423698872, + "loss": 3.4565, + "step": 83570 + }, + { + "epoch": 5.6784209811115645, + "grad_norm": 0.9335669875144958, + "learning_rate": 0.00029019737736105453, + "loss": 3.4037, + "step": 83575 + }, + { + "epoch": 5.678760701182226, + "grad_norm": 0.7997127771377563, + "learning_rate": 0.0002901549123522218, + "loss": 3.3241, + "step": 83580 + }, + { + "epoch": 5.679100421252888, + "grad_norm": 0.9237399101257324, + "learning_rate": 0.00029011244734338903, + "loss": 3.3929, + "step": 83585 + }, + { + "epoch": 5.67944014132355, + "grad_norm": 0.8259170651435852, + "learning_rate": 0.0002900699823345563, + "loss": 3.4499, + "step": 83590 + }, + { + "epoch": 5.679779861394211, + "grad_norm": 0.8382492065429688, + "learning_rate": 0.00029002751732572365, + "loss": 3.3591, + "step": 83595 + }, + { + "epoch": 5.680119581464873, + "grad_norm": 0.9118777513504028, + "learning_rate": 0.0002899850523168909, + "loss": 3.4997, + "step": 83600 + }, + { + "epoch": 5.680459301535535, + "grad_norm": 0.9826738238334656, + "learning_rate": 0.00028994258730805815, + "loss": 3.4835, + "step": 83605 + }, + { + "epoch": 5.680799021606196, + "grad_norm": 0.8213613629341125, + "learning_rate": 0.0002899001222992255, + "loss": 3.63, + "step": 83610 + }, + { + "epoch": 5.681138741676858, + "grad_norm": 0.856124222278595, + "learning_rate": 0.0002898576572903927, + "loss": 3.6009, + "step": 83615 + }, + { + "epoch": 5.6814784617475205, + "grad_norm": 0.713871419429779, + "learning_rate": 0.00028981519228156, + "loss": 3.3987, + "step": 83620 + }, + { + "epoch": 5.681818181818182, + "grad_norm": 0.7922143936157227, + "learning_rate": 0.0002897727272727273, + "loss": 3.5835, + "step": 83625 + }, + { + "epoch": 5.682157901888844, + "grad_norm": 0.9641343951225281, + "learning_rate": 0.00028973026226389455, + "loss": 3.4919, + "step": 83630 + }, + { + "epoch": 5.682497621959506, + "grad_norm": 0.8714730143547058, + "learning_rate": 0.00028968779725506183, + "loss": 3.3982, + "step": 83635 + }, + { + "epoch": 5.682837342030167, + "grad_norm": 1.1657137870788574, + "learning_rate": 0.0002896453322462291, + "loss": 3.7124, + "step": 83640 + }, + { + "epoch": 5.683177062100829, + "grad_norm": 0.7330372333526611, + "learning_rate": 0.0002896028672373964, + "loss": 3.541, + "step": 83645 + }, + { + "epoch": 5.683516782171491, + "grad_norm": 0.8605338335037231, + "learning_rate": 0.0002895604022285637, + "loss": 3.0809, + "step": 83650 + }, + { + "epoch": 5.683856502242152, + "grad_norm": 1.101027250289917, + "learning_rate": 0.00028951793721973095, + "loss": 3.1177, + "step": 83655 + }, + { + "epoch": 5.684196222312814, + "grad_norm": 0.6473367214202881, + "learning_rate": 0.0002894754722108982, + "loss": 3.6106, + "step": 83660 + }, + { + "epoch": 5.6845359423834765, + "grad_norm": 0.9344337582588196, + "learning_rate": 0.0002894330072020655, + "loss": 3.2412, + "step": 83665 + }, + { + "epoch": 5.684875662454138, + "grad_norm": 0.8824185132980347, + "learning_rate": 0.0002893905421932328, + "loss": 3.4567, + "step": 83670 + }, + { + "epoch": 5.6852153825248, + "grad_norm": 0.8569685220718384, + "learning_rate": 0.0002893480771844, + "loss": 3.3529, + "step": 83675 + }, + { + "epoch": 5.685555102595462, + "grad_norm": 1.3566203117370605, + "learning_rate": 0.00028930561217556735, + "loss": 3.6026, + "step": 83680 + }, + { + "epoch": 5.685894822666123, + "grad_norm": 0.7413835525512695, + "learning_rate": 0.00028926314716673463, + "loss": 3.5377, + "step": 83685 + }, + { + "epoch": 5.686234542736785, + "grad_norm": 1.0775066614151, + "learning_rate": 0.00028922068215790186, + "loss": 3.4886, + "step": 83690 + }, + { + "epoch": 5.686574262807447, + "grad_norm": 1.0997930765151978, + "learning_rate": 0.00028917821714906914, + "loss": 3.4644, + "step": 83695 + }, + { + "epoch": 5.686913982878108, + "grad_norm": 0.7739960551261902, + "learning_rate": 0.0002891357521402365, + "loss": 3.5414, + "step": 83700 + }, + { + "epoch": 5.68725370294877, + "grad_norm": 0.9319838285446167, + "learning_rate": 0.0002890932871314037, + "loss": 3.3576, + "step": 83705 + }, + { + "epoch": 5.687593423019432, + "grad_norm": 0.8731487989425659, + "learning_rate": 0.000289050822122571, + "loss": 3.1742, + "step": 83710 + }, + { + "epoch": 5.687933143090094, + "grad_norm": 0.9512740969657898, + "learning_rate": 0.0002890083571137383, + "loss": 3.3325, + "step": 83715 + }, + { + "epoch": 5.688272863160756, + "grad_norm": 0.8766419887542725, + "learning_rate": 0.00028896589210490554, + "loss": 3.4934, + "step": 83720 + }, + { + "epoch": 5.688612583231417, + "grad_norm": 1.0418423414230347, + "learning_rate": 0.0002889234270960728, + "loss": 3.5246, + "step": 83725 + }, + { + "epoch": 5.688952303302079, + "grad_norm": 0.9664839506149292, + "learning_rate": 0.00028888096208724016, + "loss": 3.6018, + "step": 83730 + }, + { + "epoch": 5.689292023372741, + "grad_norm": 0.8024872541427612, + "learning_rate": 0.00028883849707840744, + "loss": 3.3275, + "step": 83735 + }, + { + "epoch": 5.689631743443402, + "grad_norm": 0.853101909160614, + "learning_rate": 0.00028879603206957466, + "loss": 3.4575, + "step": 83740 + }, + { + "epoch": 5.689971463514064, + "grad_norm": 0.8780530095100403, + "learning_rate": 0.00028875356706074194, + "loss": 3.5338, + "step": 83745 + }, + { + "epoch": 5.690311183584726, + "grad_norm": 0.7772284150123596, + "learning_rate": 0.0002887111020519093, + "loss": 3.4744, + "step": 83750 + }, + { + "epoch": 5.690650903655388, + "grad_norm": 0.9718529582023621, + "learning_rate": 0.0002886686370430765, + "loss": 3.5455, + "step": 83755 + }, + { + "epoch": 5.69099062372605, + "grad_norm": 0.7898495197296143, + "learning_rate": 0.0002886261720342438, + "loss": 3.492, + "step": 83760 + }, + { + "epoch": 5.691330343796712, + "grad_norm": 0.9698404669761658, + "learning_rate": 0.0002885837070254111, + "loss": 3.526, + "step": 83765 + }, + { + "epoch": 5.691670063867373, + "grad_norm": 0.9017671346664429, + "learning_rate": 0.00028854124201657834, + "loss": 3.3497, + "step": 83770 + }, + { + "epoch": 5.692009783938035, + "grad_norm": 0.8094810247421265, + "learning_rate": 0.0002884987770077456, + "loss": 3.4283, + "step": 83775 + }, + { + "epoch": 5.692349504008697, + "grad_norm": 0.8384556174278259, + "learning_rate": 0.0002884563119989129, + "loss": 3.4772, + "step": 83780 + }, + { + "epoch": 5.692689224079358, + "grad_norm": 0.9147753715515137, + "learning_rate": 0.0002884138469900802, + "loss": 3.4849, + "step": 83785 + }, + { + "epoch": 5.69302894415002, + "grad_norm": 1.1709409952163696, + "learning_rate": 0.00028837138198124746, + "loss": 3.1604, + "step": 83790 + }, + { + "epoch": 5.693368664220682, + "grad_norm": 0.877680242061615, + "learning_rate": 0.00028832891697241474, + "loss": 3.5849, + "step": 83795 + }, + { + "epoch": 5.693708384291344, + "grad_norm": 0.9595133662223816, + "learning_rate": 0.000288286451963582, + "loss": 3.4772, + "step": 83800 + }, + { + "epoch": 5.694048104362006, + "grad_norm": 0.8448981642723083, + "learning_rate": 0.0002882439869547493, + "loss": 3.2734, + "step": 83805 + }, + { + "epoch": 5.694387824432668, + "grad_norm": 1.5855991840362549, + "learning_rate": 0.0002882015219459166, + "loss": 3.4545, + "step": 83810 + }, + { + "epoch": 5.694727544503329, + "grad_norm": 1.0869280099868774, + "learning_rate": 0.0002881590569370838, + "loss": 3.5631, + "step": 83815 + }, + { + "epoch": 5.695067264573991, + "grad_norm": 0.9549832344055176, + "learning_rate": 0.00028811659192825114, + "loss": 3.4173, + "step": 83820 + }, + { + "epoch": 5.695406984644653, + "grad_norm": 0.9095388054847717, + "learning_rate": 0.0002880741269194184, + "loss": 3.2966, + "step": 83825 + }, + { + "epoch": 5.695746704715314, + "grad_norm": 1.0798704624176025, + "learning_rate": 0.00028803166191058565, + "loss": 3.5168, + "step": 83830 + }, + { + "epoch": 5.696086424785976, + "grad_norm": 0.8289989829063416, + "learning_rate": 0.000287989196901753, + "loss": 3.7472, + "step": 83835 + }, + { + "epoch": 5.6964261448566385, + "grad_norm": 0.8996767997741699, + "learning_rate": 0.00028794673189292026, + "loss": 3.3144, + "step": 83840 + }, + { + "epoch": 5.6967658649273, + "grad_norm": 1.0706709623336792, + "learning_rate": 0.0002879042668840875, + "loss": 3.3853, + "step": 83845 + }, + { + "epoch": 5.697105584997962, + "grad_norm": 0.8196334838867188, + "learning_rate": 0.00028786180187525477, + "loss": 3.434, + "step": 83850 + }, + { + "epoch": 5.697445305068624, + "grad_norm": 1.0109295845031738, + "learning_rate": 0.0002878193368664221, + "loss": 3.4064, + "step": 83855 + }, + { + "epoch": 5.697785025139285, + "grad_norm": 1.053186058998108, + "learning_rate": 0.00028777687185758933, + "loss": 3.4234, + "step": 83860 + }, + { + "epoch": 5.698124745209947, + "grad_norm": 1.072319507598877, + "learning_rate": 0.0002877344068487566, + "loss": 3.4044, + "step": 83865 + }, + { + "epoch": 5.698464465280609, + "grad_norm": 1.0606738328933716, + "learning_rate": 0.00028769194183992394, + "loss": 3.3686, + "step": 83870 + }, + { + "epoch": 5.69880418535127, + "grad_norm": 1.1386884450912476, + "learning_rate": 0.00028764947683109117, + "loss": 3.5223, + "step": 83875 + }, + { + "epoch": 5.699143905421932, + "grad_norm": 1.0773673057556152, + "learning_rate": 0.00028760701182225845, + "loss": 3.5415, + "step": 83880 + }, + { + "epoch": 5.6994836254925945, + "grad_norm": 1.0632858276367188, + "learning_rate": 0.00028756454681342573, + "loss": 3.301, + "step": 83885 + }, + { + "epoch": 5.699823345563256, + "grad_norm": 0.8226331472396851, + "learning_rate": 0.000287522081804593, + "loss": 3.3034, + "step": 83890 + }, + { + "epoch": 5.700163065633918, + "grad_norm": 0.6676120162010193, + "learning_rate": 0.0002874796167957603, + "loss": 3.3912, + "step": 83895 + }, + { + "epoch": 5.70050278570458, + "grad_norm": 0.8970184326171875, + "learning_rate": 0.00028743715178692757, + "loss": 3.3206, + "step": 83900 + }, + { + "epoch": 5.700842505775241, + "grad_norm": 0.9870097041130066, + "learning_rate": 0.0002873946867780949, + "loss": 3.4802, + "step": 83905 + }, + { + "epoch": 5.701182225845903, + "grad_norm": 0.9885196685791016, + "learning_rate": 0.00028735222176926213, + "loss": 3.5585, + "step": 83910 + }, + { + "epoch": 5.701521945916565, + "grad_norm": 0.7322256565093994, + "learning_rate": 0.0002873097567604294, + "loss": 3.5418, + "step": 83915 + }, + { + "epoch": 5.701861665987226, + "grad_norm": 1.0818402767181396, + "learning_rate": 0.0002872672917515967, + "loss": 3.3878, + "step": 83920 + }, + { + "epoch": 5.702201386057888, + "grad_norm": 0.8178190588951111, + "learning_rate": 0.00028722482674276397, + "loss": 3.3554, + "step": 83925 + }, + { + "epoch": 5.70254110612855, + "grad_norm": 0.8919997215270996, + "learning_rate": 0.00028718236173393125, + "loss": 3.4222, + "step": 83930 + }, + { + "epoch": 5.702880826199212, + "grad_norm": 1.0931775569915771, + "learning_rate": 0.00028713989672509853, + "loss": 3.3344, + "step": 83935 + }, + { + "epoch": 5.703220546269874, + "grad_norm": 0.9028447270393372, + "learning_rate": 0.0002870974317162658, + "loss": 3.1894, + "step": 83940 + }, + { + "epoch": 5.703560266340535, + "grad_norm": 0.9204580783843994, + "learning_rate": 0.0002870549667074331, + "loss": 3.6648, + "step": 83945 + }, + { + "epoch": 5.703899986411197, + "grad_norm": 1.0741089582443237, + "learning_rate": 0.00028701250169860037, + "loss": 3.2224, + "step": 83950 + }, + { + "epoch": 5.704239706481859, + "grad_norm": 0.9022413492202759, + "learning_rate": 0.0002869700366897676, + "loss": 3.4032, + "step": 83955 + }, + { + "epoch": 5.70457942655252, + "grad_norm": 1.0687668323516846, + "learning_rate": 0.00028692757168093493, + "loss": 3.303, + "step": 83960 + }, + { + "epoch": 5.704919146623182, + "grad_norm": 0.8772162795066833, + "learning_rate": 0.0002868851066721022, + "loss": 3.3588, + "step": 83965 + }, + { + "epoch": 5.705258866693844, + "grad_norm": 1.0311598777770996, + "learning_rate": 0.00028684264166326944, + "loss": 3.4788, + "step": 83970 + }, + { + "epoch": 5.705598586764506, + "grad_norm": 1.0922843217849731, + "learning_rate": 0.00028680017665443677, + "loss": 3.5475, + "step": 83975 + }, + { + "epoch": 5.705938306835168, + "grad_norm": 0.798789381980896, + "learning_rate": 0.00028675771164560405, + "loss": 3.6184, + "step": 83980 + }, + { + "epoch": 5.70627802690583, + "grad_norm": 0.8494711518287659, + "learning_rate": 0.0002867152466367713, + "loss": 3.3918, + "step": 83985 + }, + { + "epoch": 5.706617746976491, + "grad_norm": 0.8345401287078857, + "learning_rate": 0.00028667278162793856, + "loss": 3.2829, + "step": 83990 + }, + { + "epoch": 5.706957467047153, + "grad_norm": 0.9448081851005554, + "learning_rate": 0.0002866303166191059, + "loss": 3.5346, + "step": 83995 + }, + { + "epoch": 5.707297187117815, + "grad_norm": 0.8514798879623413, + "learning_rate": 0.0002865878516102731, + "loss": 3.3315, + "step": 84000 + }, + { + "epoch": 5.707636907188476, + "grad_norm": 0.9723079800605774, + "learning_rate": 0.0002865453866014404, + "loss": 3.4747, + "step": 84005 + }, + { + "epoch": 5.707976627259138, + "grad_norm": 0.8446614146232605, + "learning_rate": 0.00028650292159260773, + "loss": 3.5318, + "step": 84010 + }, + { + "epoch": 5.7083163473298, + "grad_norm": 0.9322373270988464, + "learning_rate": 0.00028646045658377496, + "loss": 3.2718, + "step": 84015 + }, + { + "epoch": 5.708656067400462, + "grad_norm": 0.9467368125915527, + "learning_rate": 0.00028641799157494224, + "loss": 3.4314, + "step": 84020 + }, + { + "epoch": 5.708995787471124, + "grad_norm": 0.8080707788467407, + "learning_rate": 0.00028637552656610957, + "loss": 3.2791, + "step": 84025 + }, + { + "epoch": 5.709335507541786, + "grad_norm": 1.0159244537353516, + "learning_rate": 0.0002863330615572768, + "loss": 3.2554, + "step": 84030 + }, + { + "epoch": 5.709675227612447, + "grad_norm": 0.8671897649765015, + "learning_rate": 0.0002862905965484441, + "loss": 3.4122, + "step": 84035 + }, + { + "epoch": 5.710014947683109, + "grad_norm": 0.9156672954559326, + "learning_rate": 0.00028624813153961136, + "loss": 3.5787, + "step": 84040 + }, + { + "epoch": 5.710354667753771, + "grad_norm": 0.9359909892082214, + "learning_rate": 0.00028620566653077864, + "loss": 3.573, + "step": 84045 + }, + { + "epoch": 5.710694387824432, + "grad_norm": 0.7729071974754333, + "learning_rate": 0.0002861632015219459, + "loss": 3.6875, + "step": 84050 + }, + { + "epoch": 5.711034107895094, + "grad_norm": 0.9090317487716675, + "learning_rate": 0.0002861207365131132, + "loss": 3.5562, + "step": 84055 + }, + { + "epoch": 5.711373827965756, + "grad_norm": 0.9269712567329407, + "learning_rate": 0.0002860782715042805, + "loss": 3.4594, + "step": 84060 + }, + { + "epoch": 5.711713548036418, + "grad_norm": 1.371489405632019, + "learning_rate": 0.00028603580649544776, + "loss": 3.3133, + "step": 84065 + }, + { + "epoch": 5.71205326810708, + "grad_norm": 1.0069010257720947, + "learning_rate": 0.00028599334148661504, + "loss": 3.5283, + "step": 84070 + }, + { + "epoch": 5.712392988177742, + "grad_norm": 1.005874752998352, + "learning_rate": 0.0002859508764777823, + "loss": 3.5683, + "step": 84075 + }, + { + "epoch": 5.712732708248403, + "grad_norm": 0.804942786693573, + "learning_rate": 0.0002859084114689496, + "loss": 3.6965, + "step": 84080 + }, + { + "epoch": 5.713072428319065, + "grad_norm": 0.8429903984069824, + "learning_rate": 0.0002858659464601169, + "loss": 3.3661, + "step": 84085 + }, + { + "epoch": 5.713412148389727, + "grad_norm": 0.9244166612625122, + "learning_rate": 0.00028582348145128416, + "loss": 3.2561, + "step": 84090 + }, + { + "epoch": 5.713751868460388, + "grad_norm": 0.7285259366035461, + "learning_rate": 0.00028578101644245144, + "loss": 3.4882, + "step": 84095 + }, + { + "epoch": 5.71409158853105, + "grad_norm": 0.7426473498344421, + "learning_rate": 0.0002857385514336187, + "loss": 3.3409, + "step": 84100 + }, + { + "epoch": 5.7144313086017124, + "grad_norm": 0.9187553524971008, + "learning_rate": 0.000285696086424786, + "loss": 3.4546, + "step": 84105 + }, + { + "epoch": 5.714771028672374, + "grad_norm": 0.9087973237037659, + "learning_rate": 0.0002856536214159532, + "loss": 3.3987, + "step": 84110 + }, + { + "epoch": 5.715110748743036, + "grad_norm": 0.9256554841995239, + "learning_rate": 0.00028561115640712056, + "loss": 3.4038, + "step": 84115 + }, + { + "epoch": 5.715450468813698, + "grad_norm": 0.8801407217979431, + "learning_rate": 0.00028556869139828784, + "loss": 3.5577, + "step": 84120 + }, + { + "epoch": 5.715790188884359, + "grad_norm": 1.1676042079925537, + "learning_rate": 0.00028552622638945506, + "loss": 3.4637, + "step": 84125 + }, + { + "epoch": 5.716129908955021, + "grad_norm": 1.0718412399291992, + "learning_rate": 0.0002854837613806224, + "loss": 3.2875, + "step": 84130 + }, + { + "epoch": 5.716469629025683, + "grad_norm": 0.887121319770813, + "learning_rate": 0.0002854412963717897, + "loss": 3.5386, + "step": 84135 + }, + { + "epoch": 5.716809349096344, + "grad_norm": 0.8285028338432312, + "learning_rate": 0.0002853988313629569, + "loss": 3.6977, + "step": 84140 + }, + { + "epoch": 5.717149069167006, + "grad_norm": 1.046920895576477, + "learning_rate": 0.0002853563663541242, + "loss": 3.612, + "step": 84145 + }, + { + "epoch": 5.7174887892376685, + "grad_norm": 0.7470901608467102, + "learning_rate": 0.0002853139013452915, + "loss": 3.4974, + "step": 84150 + }, + { + "epoch": 5.71782850930833, + "grad_norm": 0.7823852896690369, + "learning_rate": 0.00028527143633645875, + "loss": 3.7629, + "step": 84155 + }, + { + "epoch": 5.718168229378992, + "grad_norm": 0.9573970437049866, + "learning_rate": 0.000285228971327626, + "loss": 3.3715, + "step": 84160 + }, + { + "epoch": 5.718507949449654, + "grad_norm": 0.9378623366355896, + "learning_rate": 0.00028518650631879336, + "loss": 3.5928, + "step": 84165 + }, + { + "epoch": 5.718847669520315, + "grad_norm": 0.8206192255020142, + "learning_rate": 0.0002851440413099606, + "loss": 3.6126, + "step": 84170 + }, + { + "epoch": 5.719187389590977, + "grad_norm": 1.2133420705795288, + "learning_rate": 0.00028510157630112787, + "loss": 3.1956, + "step": 84175 + }, + { + "epoch": 5.719527109661639, + "grad_norm": 0.8665253520011902, + "learning_rate": 0.00028505911129229515, + "loss": 3.2586, + "step": 84180 + }, + { + "epoch": 5.7198668297323, + "grad_norm": 0.8450859189033508, + "learning_rate": 0.0002850166462834624, + "loss": 3.535, + "step": 84185 + }, + { + "epoch": 5.720206549802962, + "grad_norm": 0.7366791367530823, + "learning_rate": 0.0002849741812746297, + "loss": 3.3075, + "step": 84190 + }, + { + "epoch": 5.7205462698736245, + "grad_norm": 0.9411636590957642, + "learning_rate": 0.000284931716265797, + "loss": 3.3706, + "step": 84195 + }, + { + "epoch": 5.720885989944286, + "grad_norm": 1.2940138578414917, + "learning_rate": 0.00028488925125696427, + "loss": 3.4005, + "step": 84200 + }, + { + "epoch": 5.721225710014948, + "grad_norm": 0.7453096508979797, + "learning_rate": 0.00028484678624813155, + "loss": 3.3809, + "step": 84205 + }, + { + "epoch": 5.72156543008561, + "grad_norm": 0.8241920471191406, + "learning_rate": 0.0002848043212392988, + "loss": 3.494, + "step": 84210 + }, + { + "epoch": 5.721905150156271, + "grad_norm": 0.7937658429145813, + "learning_rate": 0.00028476185623046605, + "loss": 3.5053, + "step": 84215 + }, + { + "epoch": 5.722244870226933, + "grad_norm": 1.191135048866272, + "learning_rate": 0.0002847193912216334, + "loss": 3.5977, + "step": 84220 + }, + { + "epoch": 5.722584590297595, + "grad_norm": 0.8079379200935364, + "learning_rate": 0.00028467692621280067, + "loss": 3.4345, + "step": 84225 + }, + { + "epoch": 5.722924310368256, + "grad_norm": 0.7136418223381042, + "learning_rate": 0.0002846344612039679, + "loss": 3.4114, + "step": 84230 + }, + { + "epoch": 5.723264030438918, + "grad_norm": 0.7274147868156433, + "learning_rate": 0.0002845919961951352, + "loss": 3.2753, + "step": 84235 + }, + { + "epoch": 5.7236037505095805, + "grad_norm": 0.7918118238449097, + "learning_rate": 0.0002845495311863025, + "loss": 3.5911, + "step": 84240 + }, + { + "epoch": 5.723943470580242, + "grad_norm": 0.9004654884338379, + "learning_rate": 0.0002845070661774698, + "loss": 3.4257, + "step": 84245 + }, + { + "epoch": 5.724283190650904, + "grad_norm": 0.9150378704071045, + "learning_rate": 0.000284464601168637, + "loss": 3.5169, + "step": 84250 + }, + { + "epoch": 5.724622910721566, + "grad_norm": 0.9237611889839172, + "learning_rate": 0.00028442213615980435, + "loss": 3.2728, + "step": 84255 + }, + { + "epoch": 5.724962630792227, + "grad_norm": 1.181609869003296, + "learning_rate": 0.0002843796711509716, + "loss": 3.0971, + "step": 84260 + }, + { + "epoch": 5.725302350862889, + "grad_norm": 0.7847892642021179, + "learning_rate": 0.00028433720614213885, + "loss": 3.4945, + "step": 84265 + }, + { + "epoch": 5.725642070933551, + "grad_norm": 0.9372137784957886, + "learning_rate": 0.0002842947411333062, + "loss": 3.4669, + "step": 84270 + }, + { + "epoch": 5.725981791004212, + "grad_norm": 0.8832663893699646, + "learning_rate": 0.00028425227612447347, + "loss": 3.5059, + "step": 84275 + }, + { + "epoch": 5.726321511074874, + "grad_norm": 0.9269825220108032, + "learning_rate": 0.0002842098111156407, + "loss": 3.422, + "step": 84280 + }, + { + "epoch": 5.7266612311455365, + "grad_norm": 0.8080892562866211, + "learning_rate": 0.00028416734610680803, + "loss": 3.2264, + "step": 84285 + }, + { + "epoch": 5.727000951216198, + "grad_norm": 0.7354010343551636, + "learning_rate": 0.0002841248810979753, + "loss": 3.3669, + "step": 84290 + }, + { + "epoch": 5.72734067128686, + "grad_norm": 0.8802018761634827, + "learning_rate": 0.00028408241608914253, + "loss": 3.1606, + "step": 84295 + }, + { + "epoch": 5.727680391357522, + "grad_norm": 1.4239526987075806, + "learning_rate": 0.0002840399510803098, + "loss": 3.4503, + "step": 84300 + }, + { + "epoch": 5.728020111428183, + "grad_norm": 0.8080979585647583, + "learning_rate": 0.00028399748607147715, + "loss": 3.4527, + "step": 84305 + }, + { + "epoch": 5.728359831498845, + "grad_norm": 1.1819610595703125, + "learning_rate": 0.0002839550210626444, + "loss": 3.3812, + "step": 84310 + }, + { + "epoch": 5.728699551569507, + "grad_norm": 2.1501505374908447, + "learning_rate": 0.00028391255605381165, + "loss": 3.4002, + "step": 84315 + }, + { + "epoch": 5.729039271640168, + "grad_norm": 1.0254167318344116, + "learning_rate": 0.000283870091044979, + "loss": 3.38, + "step": 84320 + }, + { + "epoch": 5.72937899171083, + "grad_norm": 0.8922179341316223, + "learning_rate": 0.0002838276260361462, + "loss": 3.169, + "step": 84325 + }, + { + "epoch": 5.7297187117814925, + "grad_norm": 0.90912264585495, + "learning_rate": 0.0002837851610273135, + "loss": 3.2973, + "step": 84330 + }, + { + "epoch": 5.730058431852154, + "grad_norm": 1.118115782737732, + "learning_rate": 0.0002837426960184808, + "loss": 3.2776, + "step": 84335 + }, + { + "epoch": 5.730398151922816, + "grad_norm": 0.7984672784805298, + "learning_rate": 0.00028370023100964805, + "loss": 3.3826, + "step": 84340 + }, + { + "epoch": 5.730737871993478, + "grad_norm": 1.1564778089523315, + "learning_rate": 0.00028365776600081533, + "loss": 3.5756, + "step": 84345 + }, + { + "epoch": 5.731077592064139, + "grad_norm": 0.8630908727645874, + "learning_rate": 0.0002836153009919826, + "loss": 3.3204, + "step": 84350 + }, + { + "epoch": 5.731417312134801, + "grad_norm": 0.8502505421638489, + "learning_rate": 0.0002835728359831499, + "loss": 3.5247, + "step": 84355 + }, + { + "epoch": 5.731757032205463, + "grad_norm": 0.9214577674865723, + "learning_rate": 0.0002835303709743172, + "loss": 3.2942, + "step": 84360 + }, + { + "epoch": 5.732096752276124, + "grad_norm": 1.0285985469818115, + "learning_rate": 0.00028348790596548445, + "loss": 3.3868, + "step": 84365 + }, + { + "epoch": 5.732436472346786, + "grad_norm": 0.9314363598823547, + "learning_rate": 0.0002834454409566517, + "loss": 3.3335, + "step": 84370 + }, + { + "epoch": 5.7327761924174485, + "grad_norm": 0.7926965951919556, + "learning_rate": 0.000283402975947819, + "loss": 3.2671, + "step": 84375 + }, + { + "epoch": 5.73311591248811, + "grad_norm": 0.9742127060890198, + "learning_rate": 0.0002833605109389863, + "loss": 3.3638, + "step": 84380 + }, + { + "epoch": 5.733455632558772, + "grad_norm": 0.9291610717773438, + "learning_rate": 0.0002833180459301535, + "loss": 3.3129, + "step": 84385 + }, + { + "epoch": 5.733795352629433, + "grad_norm": 0.9314466118812561, + "learning_rate": 0.00028327558092132085, + "loss": 3.5246, + "step": 84390 + }, + { + "epoch": 5.734135072700095, + "grad_norm": 1.0701738595962524, + "learning_rate": 0.00028323311591248813, + "loss": 3.8348, + "step": 84395 + }, + { + "epoch": 5.734474792770757, + "grad_norm": 0.8623260259628296, + "learning_rate": 0.00028319065090365536, + "loss": 3.3738, + "step": 84400 + }, + { + "epoch": 5.734814512841418, + "grad_norm": 1.0144683122634888, + "learning_rate": 0.00028314818589482264, + "loss": 3.1789, + "step": 84405 + }, + { + "epoch": 5.73515423291208, + "grad_norm": 0.7138890027999878, + "learning_rate": 0.00028310572088599, + "loss": 3.6743, + "step": 84410 + }, + { + "epoch": 5.7354939529827424, + "grad_norm": 0.83744215965271, + "learning_rate": 0.00028306325587715725, + "loss": 3.4298, + "step": 84415 + }, + { + "epoch": 5.735833673053404, + "grad_norm": 1.1548726558685303, + "learning_rate": 0.0002830207908683245, + "loss": 3.5322, + "step": 84420 + }, + { + "epoch": 5.736173393124066, + "grad_norm": 0.8313130736351013, + "learning_rate": 0.0002829783258594918, + "loss": 3.5726, + "step": 84425 + }, + { + "epoch": 5.736513113194728, + "grad_norm": 0.8563076257705688, + "learning_rate": 0.0002829358608506591, + "loss": 3.4189, + "step": 84430 + }, + { + "epoch": 5.736852833265389, + "grad_norm": 1.0885120630264282, + "learning_rate": 0.0002828933958418263, + "loss": 3.3929, + "step": 84435 + }, + { + "epoch": 5.737192553336051, + "grad_norm": 0.7187303304672241, + "learning_rate": 0.0002828509308329936, + "loss": 3.5041, + "step": 84440 + }, + { + "epoch": 5.737532273406713, + "grad_norm": 0.9216341972351074, + "learning_rate": 0.00028280846582416094, + "loss": 3.1429, + "step": 84445 + }, + { + "epoch": 5.737871993477374, + "grad_norm": 0.9204725027084351, + "learning_rate": 0.00028276600081532816, + "loss": 3.591, + "step": 84450 + }, + { + "epoch": 5.738211713548036, + "grad_norm": 0.7729880809783936, + "learning_rate": 0.00028272353580649544, + "loss": 3.3622, + "step": 84455 + }, + { + "epoch": 5.7385514336186985, + "grad_norm": 0.889034628868103, + "learning_rate": 0.0002826810707976628, + "loss": 3.3659, + "step": 84460 + }, + { + "epoch": 5.73889115368936, + "grad_norm": 1.0555063486099243, + "learning_rate": 0.00028263860578883, + "loss": 3.4863, + "step": 84465 + }, + { + "epoch": 5.739230873760022, + "grad_norm": 1.2278649806976318, + "learning_rate": 0.0002825961407799973, + "loss": 3.6559, + "step": 84470 + }, + { + "epoch": 5.739570593830684, + "grad_norm": 0.8109133839607239, + "learning_rate": 0.00028255367577116456, + "loss": 3.5988, + "step": 84475 + }, + { + "epoch": 5.739910313901345, + "grad_norm": 1.1261653900146484, + "learning_rate": 0.00028251121076233184, + "loss": 3.407, + "step": 84480 + }, + { + "epoch": 5.740250033972007, + "grad_norm": 0.796308696269989, + "learning_rate": 0.0002824687457534991, + "loss": 3.4035, + "step": 84485 + }, + { + "epoch": 5.740589754042669, + "grad_norm": 0.9848004579544067, + "learning_rate": 0.0002824262807446664, + "loss": 3.6228, + "step": 84490 + }, + { + "epoch": 5.74092947411333, + "grad_norm": 0.6940716505050659, + "learning_rate": 0.0002823838157358337, + "loss": 3.5295, + "step": 84495 + }, + { + "epoch": 5.741269194183992, + "grad_norm": 0.9717973470687866, + "learning_rate": 0.00028234135072700096, + "loss": 3.4517, + "step": 84500 + }, + { + "epoch": 5.7416089142546545, + "grad_norm": 0.8272750377655029, + "learning_rate": 0.00028229888571816824, + "loss": 3.08, + "step": 84505 + }, + { + "epoch": 5.741948634325316, + "grad_norm": 0.8971176743507385, + "learning_rate": 0.00028225642070933547, + "loss": 3.1578, + "step": 84510 + }, + { + "epoch": 5.742288354395978, + "grad_norm": 0.8350014090538025, + "learning_rate": 0.0002822139557005028, + "loss": 3.474, + "step": 84515 + }, + { + "epoch": 5.74262807446664, + "grad_norm": 0.9874699115753174, + "learning_rate": 0.0002821714906916701, + "loss": 3.5072, + "step": 84520 + }, + { + "epoch": 5.742967794537301, + "grad_norm": 0.8679519295692444, + "learning_rate": 0.0002821290256828373, + "loss": 3.5679, + "step": 84525 + }, + { + "epoch": 5.743307514607963, + "grad_norm": 0.7251117825508118, + "learning_rate": 0.00028208656067400464, + "loss": 3.5299, + "step": 84530 + }, + { + "epoch": 5.743647234678625, + "grad_norm": 1.0036567449569702, + "learning_rate": 0.0002820440956651719, + "loss": 3.4496, + "step": 84535 + }, + { + "epoch": 5.743986954749286, + "grad_norm": 0.8619116544723511, + "learning_rate": 0.00028200163065633915, + "loss": 3.5888, + "step": 84540 + }, + { + "epoch": 5.744326674819948, + "grad_norm": 1.024381160736084, + "learning_rate": 0.00028195916564750643, + "loss": 3.4979, + "step": 84545 + }, + { + "epoch": 5.7446663948906105, + "grad_norm": 0.9617547392845154, + "learning_rate": 0.00028191670063867376, + "loss": 3.4353, + "step": 84550 + }, + { + "epoch": 5.745006114961272, + "grad_norm": 1.1640223264694214, + "learning_rate": 0.000281874235629841, + "loss": 3.3353, + "step": 84555 + }, + { + "epoch": 5.745345835031934, + "grad_norm": 0.7917426824569702, + "learning_rate": 0.00028183177062100827, + "loss": 3.5303, + "step": 84560 + }, + { + "epoch": 5.745685555102596, + "grad_norm": 1.0829036235809326, + "learning_rate": 0.0002817893056121756, + "loss": 3.3918, + "step": 84565 + }, + { + "epoch": 5.746025275173257, + "grad_norm": 0.8225312232971191, + "learning_rate": 0.00028174684060334283, + "loss": 3.5657, + "step": 84570 + }, + { + "epoch": 5.746364995243919, + "grad_norm": 0.9610084295272827, + "learning_rate": 0.0002817043755945101, + "loss": 3.5816, + "step": 84575 + }, + { + "epoch": 5.746704715314581, + "grad_norm": 0.9116448163986206, + "learning_rate": 0.00028166191058567744, + "loss": 3.2602, + "step": 84580 + }, + { + "epoch": 5.747044435385242, + "grad_norm": 0.9722238183021545, + "learning_rate": 0.0002816194455768447, + "loss": 3.5461, + "step": 84585 + }, + { + "epoch": 5.747384155455904, + "grad_norm": 0.9347169995307922, + "learning_rate": 0.00028157698056801195, + "loss": 3.6777, + "step": 84590 + }, + { + "epoch": 5.7477238755265665, + "grad_norm": 0.9015772938728333, + "learning_rate": 0.00028153451555917923, + "loss": 3.3544, + "step": 84595 + }, + { + "epoch": 5.748063595597228, + "grad_norm": 0.8130584955215454, + "learning_rate": 0.00028149205055034656, + "loss": 3.1738, + "step": 84600 + }, + { + "epoch": 5.74840331566789, + "grad_norm": 0.7512129545211792, + "learning_rate": 0.0002814495855415138, + "loss": 3.5225, + "step": 84605 + }, + { + "epoch": 5.748743035738551, + "grad_norm": 1.1360578536987305, + "learning_rate": 0.00028140712053268107, + "loss": 3.4459, + "step": 84610 + }, + { + "epoch": 5.749082755809213, + "grad_norm": 0.7968753576278687, + "learning_rate": 0.0002813646555238484, + "loss": 3.3326, + "step": 84615 + }, + { + "epoch": 5.749422475879875, + "grad_norm": 0.8719421625137329, + "learning_rate": 0.00028132219051501563, + "loss": 3.4085, + "step": 84620 + }, + { + "epoch": 5.749762195950536, + "grad_norm": 0.7664800882339478, + "learning_rate": 0.0002812797255061829, + "loss": 3.4528, + "step": 84625 + }, + { + "epoch": 5.750101916021198, + "grad_norm": 1.0982551574707031, + "learning_rate": 0.0002812372604973502, + "loss": 3.3311, + "step": 84630 + }, + { + "epoch": 5.75044163609186, + "grad_norm": 0.9693065881729126, + "learning_rate": 0.00028119479548851747, + "loss": 3.3093, + "step": 84635 + }, + { + "epoch": 5.750781356162522, + "grad_norm": 0.8258723616600037, + "learning_rate": 0.00028115233047968475, + "loss": 3.353, + "step": 84640 + }, + { + "epoch": 5.751121076233184, + "grad_norm": 0.6775805950164795, + "learning_rate": 0.00028110986547085203, + "loss": 3.3371, + "step": 84645 + }, + { + "epoch": 5.751460796303846, + "grad_norm": 0.7492602467536926, + "learning_rate": 0.0002810674004620193, + "loss": 3.3218, + "step": 84650 + }, + { + "epoch": 5.751800516374507, + "grad_norm": 0.8571633696556091, + "learning_rate": 0.0002810249354531866, + "loss": 3.4108, + "step": 84655 + }, + { + "epoch": 5.752140236445169, + "grad_norm": 0.7492689490318298, + "learning_rate": 0.00028098247044435387, + "loss": 3.6325, + "step": 84660 + }, + { + "epoch": 5.752479956515831, + "grad_norm": 0.8897091746330261, + "learning_rate": 0.0002809400054355211, + "loss": 3.3558, + "step": 84665 + }, + { + "epoch": 5.752819676586492, + "grad_norm": 1.013622522354126, + "learning_rate": 0.00028089754042668843, + "loss": 3.4798, + "step": 84670 + }, + { + "epoch": 5.753159396657154, + "grad_norm": 0.9857528805732727, + "learning_rate": 0.0002808550754178557, + "loss": 3.3363, + "step": 84675 + }, + { + "epoch": 5.753499116727816, + "grad_norm": 1.493955373764038, + "learning_rate": 0.00028081261040902294, + "loss": 3.585, + "step": 84680 + }, + { + "epoch": 5.753838836798478, + "grad_norm": 0.8143149614334106, + "learning_rate": 0.00028077014540019027, + "loss": 3.232, + "step": 84685 + }, + { + "epoch": 5.75417855686914, + "grad_norm": 1.0209901332855225, + "learning_rate": 0.00028072768039135755, + "loss": 3.4786, + "step": 84690 + }, + { + "epoch": 5.754518276939802, + "grad_norm": 1.0197370052337646, + "learning_rate": 0.0002806852153825248, + "loss": 3.5541, + "step": 84695 + }, + { + "epoch": 5.754857997010463, + "grad_norm": 1.0470458269119263, + "learning_rate": 0.00028064275037369206, + "loss": 3.1511, + "step": 84700 + }, + { + "epoch": 5.755197717081125, + "grad_norm": 0.9887160658836365, + "learning_rate": 0.0002806002853648594, + "loss": 3.3463, + "step": 84705 + }, + { + "epoch": 5.755537437151787, + "grad_norm": 1.0099499225616455, + "learning_rate": 0.0002805578203560266, + "loss": 3.3159, + "step": 84710 + }, + { + "epoch": 5.755877157222448, + "grad_norm": 1.040313959121704, + "learning_rate": 0.0002805153553471939, + "loss": 2.9884, + "step": 84715 + }, + { + "epoch": 5.75621687729311, + "grad_norm": 0.985922634601593, + "learning_rate": 0.00028047289033836123, + "loss": 3.3712, + "step": 84720 + }, + { + "epoch": 5.7565565973637725, + "grad_norm": 1.0711133480072021, + "learning_rate": 0.00028043042532952846, + "loss": 3.3369, + "step": 84725 + }, + { + "epoch": 5.756896317434434, + "grad_norm": 0.9990330338478088, + "learning_rate": 0.00028038796032069574, + "loss": 3.2786, + "step": 84730 + }, + { + "epoch": 5.757236037505096, + "grad_norm": 0.819372832775116, + "learning_rate": 0.000280345495311863, + "loss": 3.2619, + "step": 84735 + }, + { + "epoch": 5.757575757575758, + "grad_norm": 0.8904964923858643, + "learning_rate": 0.0002803030303030303, + "loss": 3.4609, + "step": 84740 + }, + { + "epoch": 5.757915477646419, + "grad_norm": 1.4461841583251953, + "learning_rate": 0.0002802605652941976, + "loss": 3.242, + "step": 84745 + }, + { + "epoch": 5.758255197717081, + "grad_norm": 0.7628354430198669, + "learning_rate": 0.00028021810028536486, + "loss": 3.4557, + "step": 84750 + }, + { + "epoch": 5.758594917787743, + "grad_norm": 0.9564782381057739, + "learning_rate": 0.0002801756352765322, + "loss": 3.717, + "step": 84755 + }, + { + "epoch": 5.758934637858404, + "grad_norm": 1.012934684753418, + "learning_rate": 0.0002801331702676994, + "loss": 3.338, + "step": 84760 + }, + { + "epoch": 5.759274357929066, + "grad_norm": 1.0095961093902588, + "learning_rate": 0.0002800907052588667, + "loss": 3.7653, + "step": 84765 + }, + { + "epoch": 5.7596140779997285, + "grad_norm": 0.9408828616142273, + "learning_rate": 0.000280048240250034, + "loss": 3.3584, + "step": 84770 + }, + { + "epoch": 5.75995379807039, + "grad_norm": 0.9987916946411133, + "learning_rate": 0.00028000577524120126, + "loss": 3.3689, + "step": 84775 + }, + { + "epoch": 5.760293518141052, + "grad_norm": 0.9112595319747925, + "learning_rate": 0.00027996331023236854, + "loss": 3.65, + "step": 84780 + }, + { + "epoch": 5.760633238211714, + "grad_norm": 0.9903378486633301, + "learning_rate": 0.0002799208452235358, + "loss": 3.4764, + "step": 84785 + }, + { + "epoch": 5.760972958282375, + "grad_norm": 0.8739778995513916, + "learning_rate": 0.0002798783802147031, + "loss": 3.3431, + "step": 84790 + }, + { + "epoch": 5.761312678353037, + "grad_norm": 0.941175639629364, + "learning_rate": 0.0002798359152058704, + "loss": 3.4752, + "step": 84795 + }, + { + "epoch": 5.761652398423699, + "grad_norm": 0.9631157517433167, + "learning_rate": 0.00027979345019703766, + "loss": 3.3741, + "step": 84800 + }, + { + "epoch": 5.76199211849436, + "grad_norm": 0.8805457949638367, + "learning_rate": 0.0002797509851882049, + "loss": 3.0029, + "step": 84805 + }, + { + "epoch": 5.762331838565022, + "grad_norm": 0.8393582701683044, + "learning_rate": 0.0002797085201793722, + "loss": 3.3259, + "step": 84810 + }, + { + "epoch": 5.7626715586356845, + "grad_norm": 0.9385985732078552, + "learning_rate": 0.0002796660551705395, + "loss": 3.6487, + "step": 84815 + }, + { + "epoch": 5.763011278706346, + "grad_norm": 0.9378626346588135, + "learning_rate": 0.0002796235901617067, + "loss": 3.6157, + "step": 84820 + }, + { + "epoch": 5.763350998777008, + "grad_norm": 0.7376296520233154, + "learning_rate": 0.00027958112515287406, + "loss": 3.4403, + "step": 84825 + }, + { + "epoch": 5.76369071884767, + "grad_norm": 1.0918469429016113, + "learning_rate": 0.00027953866014404134, + "loss": 3.5284, + "step": 84830 + }, + { + "epoch": 5.764030438918331, + "grad_norm": 0.8113758563995361, + "learning_rate": 0.00027949619513520856, + "loss": 3.4687, + "step": 84835 + }, + { + "epoch": 5.764370158988993, + "grad_norm": 0.8024821877479553, + "learning_rate": 0.00027945373012637584, + "loss": 3.259, + "step": 84840 + }, + { + "epoch": 5.764709879059655, + "grad_norm": 1.0318459272384644, + "learning_rate": 0.0002794112651175432, + "loss": 3.4653, + "step": 84845 + }, + { + "epoch": 5.765049599130316, + "grad_norm": 0.9057720303535461, + "learning_rate": 0.0002793688001087104, + "loss": 3.2315, + "step": 84850 + }, + { + "epoch": 5.765389319200978, + "grad_norm": 0.9232569932937622, + "learning_rate": 0.0002793263350998777, + "loss": 3.5287, + "step": 84855 + }, + { + "epoch": 5.7657290392716405, + "grad_norm": 0.8076748847961426, + "learning_rate": 0.000279283870091045, + "loss": 3.4549, + "step": 84860 + }, + { + "epoch": 5.766068759342302, + "grad_norm": 0.8412861227989197, + "learning_rate": 0.00027924140508221225, + "loss": 3.4378, + "step": 84865 + }, + { + "epoch": 5.766408479412964, + "grad_norm": 0.7184168696403503, + "learning_rate": 0.0002791989400733795, + "loss": 3.5305, + "step": 84870 + }, + { + "epoch": 5.766748199483626, + "grad_norm": 0.9143257141113281, + "learning_rate": 0.00027915647506454686, + "loss": 3.4888, + "step": 84875 + }, + { + "epoch": 5.767087919554287, + "grad_norm": 0.8010311126708984, + "learning_rate": 0.0002791140100557141, + "loss": 3.3894, + "step": 84880 + }, + { + "epoch": 5.767427639624949, + "grad_norm": 0.824298083782196, + "learning_rate": 0.00027907154504688137, + "loss": 3.3623, + "step": 84885 + }, + { + "epoch": 5.767767359695611, + "grad_norm": 0.9923129677772522, + "learning_rate": 0.00027902908003804865, + "loss": 3.6162, + "step": 84890 + }, + { + "epoch": 5.768107079766272, + "grad_norm": 0.9098342061042786, + "learning_rate": 0.0002789866150292159, + "loss": 3.4115, + "step": 84895 + }, + { + "epoch": 5.768446799836934, + "grad_norm": 0.7669000029563904, + "learning_rate": 0.0002789441500203832, + "loss": 3.543, + "step": 84900 + }, + { + "epoch": 5.7687865199075965, + "grad_norm": 0.7977213263511658, + "learning_rate": 0.0002789016850115505, + "loss": 3.6517, + "step": 84905 + }, + { + "epoch": 5.769126239978258, + "grad_norm": 0.7996681332588196, + "learning_rate": 0.00027885922000271777, + "loss": 3.5461, + "step": 84910 + }, + { + "epoch": 5.76946596004892, + "grad_norm": 1.1067283153533936, + "learning_rate": 0.00027881675499388505, + "loss": 3.4234, + "step": 84915 + }, + { + "epoch": 5.769805680119582, + "grad_norm": 0.9065130352973938, + "learning_rate": 0.0002787742899850523, + "loss": 3.3491, + "step": 84920 + }, + { + "epoch": 5.770145400190243, + "grad_norm": 1.1819778680801392, + "learning_rate": 0.0002787318249762196, + "loss": 3.5745, + "step": 84925 + }, + { + "epoch": 5.770485120260905, + "grad_norm": 0.7864854335784912, + "learning_rate": 0.0002786893599673869, + "loss": 3.6009, + "step": 84930 + }, + { + "epoch": 5.770824840331567, + "grad_norm": 0.8967171311378479, + "learning_rate": 0.00027864689495855417, + "loss": 3.6565, + "step": 84935 + }, + { + "epoch": 5.771164560402228, + "grad_norm": 0.8215663433074951, + "learning_rate": 0.00027860442994972145, + "loss": 3.473, + "step": 84940 + }, + { + "epoch": 5.77150428047289, + "grad_norm": 0.7403429746627808, + "learning_rate": 0.0002785619649408887, + "loss": 3.4333, + "step": 84945 + }, + { + "epoch": 5.7718440005435525, + "grad_norm": 1.0635896921157837, + "learning_rate": 0.000278519499932056, + "loss": 3.4566, + "step": 84950 + }, + { + "epoch": 5.772183720614214, + "grad_norm": 0.6742855906486511, + "learning_rate": 0.0002784770349232233, + "loss": 3.5216, + "step": 84955 + }, + { + "epoch": 5.772523440684876, + "grad_norm": 0.8137953281402588, + "learning_rate": 0.0002784345699143905, + "loss": 3.4716, + "step": 84960 + }, + { + "epoch": 5.772863160755538, + "grad_norm": 0.884519100189209, + "learning_rate": 0.00027839210490555785, + "loss": 3.3545, + "step": 84965 + }, + { + "epoch": 5.773202880826199, + "grad_norm": 0.7287304401397705, + "learning_rate": 0.0002783496398967251, + "loss": 3.407, + "step": 84970 + }, + { + "epoch": 5.773542600896861, + "grad_norm": 0.9417672753334045, + "learning_rate": 0.00027830717488789235, + "loss": 3.4167, + "step": 84975 + }, + { + "epoch": 5.773882320967523, + "grad_norm": 0.9843248128890991, + "learning_rate": 0.0002782647098790597, + "loss": 3.3122, + "step": 84980 + }, + { + "epoch": 5.774222041038184, + "grad_norm": 0.814001739025116, + "learning_rate": 0.00027822224487022697, + "loss": 3.2245, + "step": 84985 + }, + { + "epoch": 5.7745617611088464, + "grad_norm": 0.8859785199165344, + "learning_rate": 0.0002781797798613942, + "loss": 3.6028, + "step": 84990 + }, + { + "epoch": 5.7749014811795085, + "grad_norm": 1.0831137895584106, + "learning_rate": 0.00027813731485256147, + "loss": 3.5136, + "step": 84995 + }, + { + "epoch": 5.77524120125017, + "grad_norm": 0.6970549821853638, + "learning_rate": 0.0002780948498437288, + "loss": 3.4374, + "step": 85000 + }, + { + "epoch": 5.775580921320832, + "grad_norm": 0.8316484689712524, + "learning_rate": 0.00027805238483489603, + "loss": 3.2437, + "step": 85005 + }, + { + "epoch": 5.775920641391494, + "grad_norm": 1.0049972534179688, + "learning_rate": 0.0002780099198260633, + "loss": 3.46, + "step": 85010 + }, + { + "epoch": 5.776260361462155, + "grad_norm": 0.7613423466682434, + "learning_rate": 0.00027796745481723065, + "loss": 3.6322, + "step": 85015 + }, + { + "epoch": 5.776600081532817, + "grad_norm": 0.8280127644538879, + "learning_rate": 0.0002779249898083979, + "loss": 3.4453, + "step": 85020 + }, + { + "epoch": 5.776939801603479, + "grad_norm": 0.8864409327507019, + "learning_rate": 0.00027788252479956515, + "loss": 3.3708, + "step": 85025 + }, + { + "epoch": 5.77727952167414, + "grad_norm": 0.8070069551467896, + "learning_rate": 0.00027784005979073243, + "loss": 3.4299, + "step": 85030 + }, + { + "epoch": 5.7776192417448025, + "grad_norm": 0.8413402438163757, + "learning_rate": 0.0002777975947818997, + "loss": 3.5294, + "step": 85035 + }, + { + "epoch": 5.7779589618154645, + "grad_norm": 0.9970145225524902, + "learning_rate": 0.000277755129773067, + "loss": 3.3096, + "step": 85040 + }, + { + "epoch": 5.778298681886126, + "grad_norm": 0.8554643392562866, + "learning_rate": 0.0002777126647642343, + "loss": 3.7774, + "step": 85045 + }, + { + "epoch": 5.778638401956788, + "grad_norm": 1.105699062347412, + "learning_rate": 0.00027767019975540155, + "loss": 3.3027, + "step": 85050 + }, + { + "epoch": 5.77897812202745, + "grad_norm": 0.784903347492218, + "learning_rate": 0.00027762773474656883, + "loss": 3.4916, + "step": 85055 + }, + { + "epoch": 5.779317842098111, + "grad_norm": 1.0892821550369263, + "learning_rate": 0.0002775852697377361, + "loss": 3.2721, + "step": 85060 + }, + { + "epoch": 5.779657562168773, + "grad_norm": 0.8801475167274475, + "learning_rate": 0.00027754280472890334, + "loss": 3.6024, + "step": 85065 + }, + { + "epoch": 5.779997282239434, + "grad_norm": 0.7461523413658142, + "learning_rate": 0.0002775003397200707, + "loss": 3.2954, + "step": 85070 + }, + { + "epoch": 5.780337002310096, + "grad_norm": 1.2777671813964844, + "learning_rate": 0.00027745787471123795, + "loss": 3.3099, + "step": 85075 + }, + { + "epoch": 5.7806767223807585, + "grad_norm": 0.8527796864509583, + "learning_rate": 0.0002774154097024052, + "loss": 3.5161, + "step": 85080 + }, + { + "epoch": 5.78101644245142, + "grad_norm": 0.9096696972846985, + "learning_rate": 0.0002773729446935725, + "loss": 3.3977, + "step": 85085 + }, + { + "epoch": 5.781356162522082, + "grad_norm": 0.9232884645462036, + "learning_rate": 0.0002773304796847398, + "loss": 3.4829, + "step": 85090 + }, + { + "epoch": 5.781695882592744, + "grad_norm": 0.8620707988739014, + "learning_rate": 0.0002772880146759071, + "loss": 3.5398, + "step": 85095 + }, + { + "epoch": 5.782035602663405, + "grad_norm": 0.8821724057197571, + "learning_rate": 0.0002772455496670743, + "loss": 3.3148, + "step": 85100 + }, + { + "epoch": 5.782375322734067, + "grad_norm": 1.108919620513916, + "learning_rate": 0.00027720308465824163, + "loss": 3.1349, + "step": 85105 + }, + { + "epoch": 5.782715042804729, + "grad_norm": 0.7483648657798767, + "learning_rate": 0.0002771606196494089, + "loss": 3.4792, + "step": 85110 + }, + { + "epoch": 5.78305476287539, + "grad_norm": 0.8516687154769897, + "learning_rate": 0.00027711815464057614, + "loss": 3.5186, + "step": 85115 + }, + { + "epoch": 5.783394482946052, + "grad_norm": 1.0146435499191284, + "learning_rate": 0.0002770756896317435, + "loss": 3.1839, + "step": 85120 + }, + { + "epoch": 5.7837342030167145, + "grad_norm": 0.7424401640892029, + "learning_rate": 0.00027703322462291075, + "loss": 3.4199, + "step": 85125 + }, + { + "epoch": 5.784073923087376, + "grad_norm": 1.0545909404754639, + "learning_rate": 0.000276990759614078, + "loss": 3.3169, + "step": 85130 + }, + { + "epoch": 5.784413643158038, + "grad_norm": 0.9177398681640625, + "learning_rate": 0.00027694829460524526, + "loss": 3.7042, + "step": 85135 + }, + { + "epoch": 5.7847533632287, + "grad_norm": 0.8496518731117249, + "learning_rate": 0.0002769058295964126, + "loss": 3.3361, + "step": 85140 + }, + { + "epoch": 5.785093083299361, + "grad_norm": 0.8929536938667297, + "learning_rate": 0.0002768633645875798, + "loss": 3.4954, + "step": 85145 + }, + { + "epoch": 5.785432803370023, + "grad_norm": 0.8576931953430176, + "learning_rate": 0.0002768208995787471, + "loss": 3.4054, + "step": 85150 + }, + { + "epoch": 5.785772523440685, + "grad_norm": 0.9505293369293213, + "learning_rate": 0.00027677843456991444, + "loss": 3.382, + "step": 85155 + }, + { + "epoch": 5.786112243511346, + "grad_norm": 0.9586608409881592, + "learning_rate": 0.00027673596956108166, + "loss": 3.4526, + "step": 85160 + }, + { + "epoch": 5.786451963582008, + "grad_norm": 0.8742292523384094, + "learning_rate": 0.00027669350455224894, + "loss": 3.4869, + "step": 85165 + }, + { + "epoch": 5.7867916836526705, + "grad_norm": 0.8352279663085938, + "learning_rate": 0.0002766510395434163, + "loss": 3.518, + "step": 85170 + }, + { + "epoch": 5.787131403723332, + "grad_norm": 0.7971248626708984, + "learning_rate": 0.0002766085745345835, + "loss": 3.4276, + "step": 85175 + }, + { + "epoch": 5.787471123793994, + "grad_norm": 0.7960358262062073, + "learning_rate": 0.0002765661095257508, + "loss": 3.7624, + "step": 85180 + }, + { + "epoch": 5.787810843864656, + "grad_norm": 0.7510971426963806, + "learning_rate": 0.00027652364451691806, + "loss": 3.4266, + "step": 85185 + }, + { + "epoch": 5.788150563935317, + "grad_norm": 0.8277126550674438, + "learning_rate": 0.00027648117950808534, + "loss": 3.6447, + "step": 85190 + }, + { + "epoch": 5.788490284005979, + "grad_norm": 0.8296245336532593, + "learning_rate": 0.0002764387144992526, + "loss": 3.4857, + "step": 85195 + }, + { + "epoch": 5.788830004076641, + "grad_norm": 1.1003093719482422, + "learning_rate": 0.0002763962494904199, + "loss": 3.3522, + "step": 85200 + }, + { + "epoch": 5.789169724147302, + "grad_norm": 0.9212766885757446, + "learning_rate": 0.0002763537844815872, + "loss": 3.5218, + "step": 85205 + }, + { + "epoch": 5.789509444217964, + "grad_norm": 0.9671154022216797, + "learning_rate": 0.00027631131947275446, + "loss": 3.5837, + "step": 85210 + }, + { + "epoch": 5.7898491642886265, + "grad_norm": 0.9475633502006531, + "learning_rate": 0.00027626885446392174, + "loss": 3.3139, + "step": 85215 + }, + { + "epoch": 5.790188884359288, + "grad_norm": 0.9362192153930664, + "learning_rate": 0.00027622638945508897, + "loss": 3.5035, + "step": 85220 + }, + { + "epoch": 5.79052860442995, + "grad_norm": 1.1112134456634521, + "learning_rate": 0.0002761839244462563, + "loss": 3.2641, + "step": 85225 + }, + { + "epoch": 5.790868324500612, + "grad_norm": 1.018168568611145, + "learning_rate": 0.0002761414594374236, + "loss": 3.5804, + "step": 85230 + }, + { + "epoch": 5.791208044571273, + "grad_norm": 1.0218136310577393, + "learning_rate": 0.0002760989944285908, + "loss": 3.3278, + "step": 85235 + }, + { + "epoch": 5.791547764641935, + "grad_norm": 0.9222137928009033, + "learning_rate": 0.00027605652941975814, + "loss": 3.3777, + "step": 85240 + }, + { + "epoch": 5.791887484712597, + "grad_norm": 0.9385232329368591, + "learning_rate": 0.0002760140644109254, + "loss": 3.5547, + "step": 85245 + }, + { + "epoch": 5.792227204783258, + "grad_norm": 0.9414511919021606, + "learning_rate": 0.00027597159940209265, + "loss": 3.4844, + "step": 85250 + }, + { + "epoch": 5.79256692485392, + "grad_norm": 0.8871060609817505, + "learning_rate": 0.00027592913439325993, + "loss": 3.3282, + "step": 85255 + }, + { + "epoch": 5.7929066449245825, + "grad_norm": 1.0003222227096558, + "learning_rate": 0.00027588666938442726, + "loss": 3.6555, + "step": 85260 + }, + { + "epoch": 5.793246364995244, + "grad_norm": 0.8678728938102722, + "learning_rate": 0.00027584420437559454, + "loss": 3.3329, + "step": 85265 + }, + { + "epoch": 5.793586085065906, + "grad_norm": 1.3389482498168945, + "learning_rate": 0.00027580173936676177, + "loss": 3.322, + "step": 85270 + }, + { + "epoch": 5.793925805136568, + "grad_norm": 0.7082506418228149, + "learning_rate": 0.0002757592743579291, + "loss": 3.5262, + "step": 85275 + }, + { + "epoch": 5.794265525207229, + "grad_norm": 0.8104579448699951, + "learning_rate": 0.0002757168093490964, + "loss": 3.7101, + "step": 85280 + }, + { + "epoch": 5.794605245277891, + "grad_norm": 0.7900086641311646, + "learning_rate": 0.0002756743443402636, + "loss": 3.5779, + "step": 85285 + }, + { + "epoch": 5.794944965348552, + "grad_norm": 0.7708951830863953, + "learning_rate": 0.0002756318793314309, + "loss": 3.4201, + "step": 85290 + }, + { + "epoch": 5.795284685419214, + "grad_norm": 0.9626030325889587, + "learning_rate": 0.0002755894143225982, + "loss": 3.5494, + "step": 85295 + }, + { + "epoch": 5.7956244054898765, + "grad_norm": 0.8902991414070129, + "learning_rate": 0.00027554694931376545, + "loss": 3.4811, + "step": 85300 + }, + { + "epoch": 5.795964125560538, + "grad_norm": 1.0814568996429443, + "learning_rate": 0.00027550448430493273, + "loss": 3.5386, + "step": 85305 + }, + { + "epoch": 5.7963038456312, + "grad_norm": 0.8665685653686523, + "learning_rate": 0.00027546201929610006, + "loss": 3.637, + "step": 85310 + }, + { + "epoch": 5.796643565701862, + "grad_norm": 0.771484911441803, + "learning_rate": 0.0002754195542872673, + "loss": 3.3651, + "step": 85315 + }, + { + "epoch": 5.796983285772523, + "grad_norm": 0.9619312286376953, + "learning_rate": 0.00027537708927843457, + "loss": 3.1736, + "step": 85320 + }, + { + "epoch": 5.797323005843185, + "grad_norm": 0.7134144902229309, + "learning_rate": 0.00027533462426960185, + "loss": 3.2355, + "step": 85325 + }, + { + "epoch": 5.797662725913847, + "grad_norm": 1.018972396850586, + "learning_rate": 0.00027529215926076913, + "loss": 3.5379, + "step": 85330 + }, + { + "epoch": 5.798002445984508, + "grad_norm": 0.6907059550285339, + "learning_rate": 0.0002752496942519364, + "loss": 3.5623, + "step": 85335 + }, + { + "epoch": 5.79834216605517, + "grad_norm": 1.268068552017212, + "learning_rate": 0.0002752072292431037, + "loss": 3.3848, + "step": 85340 + }, + { + "epoch": 5.7986818861258325, + "grad_norm": 1.4232944250106812, + "learning_rate": 0.00027516476423427097, + "loss": 3.3481, + "step": 85345 + }, + { + "epoch": 5.799021606196494, + "grad_norm": 1.2240643501281738, + "learning_rate": 0.00027512229922543825, + "loss": 3.4591, + "step": 85350 + }, + { + "epoch": 5.799361326267156, + "grad_norm": 0.8978531360626221, + "learning_rate": 0.00027507983421660553, + "loss": 3.4038, + "step": 85355 + }, + { + "epoch": 5.799701046337818, + "grad_norm": 0.9523215889930725, + "learning_rate": 0.00027503736920777276, + "loss": 3.4074, + "step": 85360 + }, + { + "epoch": 5.800040766408479, + "grad_norm": 0.9013262391090393, + "learning_rate": 0.0002749949041989401, + "loss": 3.3333, + "step": 85365 + }, + { + "epoch": 5.800380486479141, + "grad_norm": 1.0248736143112183, + "learning_rate": 0.00027495243919010737, + "loss": 3.0011, + "step": 85370 + }, + { + "epoch": 5.800720206549803, + "grad_norm": 1.0579249858856201, + "learning_rate": 0.0002749099741812746, + "loss": 3.4031, + "step": 85375 + }, + { + "epoch": 5.801059926620464, + "grad_norm": 0.8195279836654663, + "learning_rate": 0.00027486750917244193, + "loss": 3.3102, + "step": 85380 + }, + { + "epoch": 5.801399646691126, + "grad_norm": 0.8672236800193787, + "learning_rate": 0.0002748250441636092, + "loss": 3.4618, + "step": 85385 + }, + { + "epoch": 5.8017393667617885, + "grad_norm": 0.8514004945755005, + "learning_rate": 0.00027478257915477644, + "loss": 3.3222, + "step": 85390 + }, + { + "epoch": 5.80207908683245, + "grad_norm": 0.9608245491981506, + "learning_rate": 0.0002747401141459437, + "loss": 3.7151, + "step": 85395 + }, + { + "epoch": 5.802418806903112, + "grad_norm": 0.9496618509292603, + "learning_rate": 0.00027469764913711105, + "loss": 3.5064, + "step": 85400 + }, + { + "epoch": 5.802758526973774, + "grad_norm": 0.9411175847053528, + "learning_rate": 0.0002746551841282783, + "loss": 3.5027, + "step": 85405 + }, + { + "epoch": 5.803098247044435, + "grad_norm": 0.998494029045105, + "learning_rate": 0.00027461271911944556, + "loss": 3.4552, + "step": 85410 + }, + { + "epoch": 5.803437967115097, + "grad_norm": 0.9135532975196838, + "learning_rate": 0.0002745702541106129, + "loss": 3.2788, + "step": 85415 + }, + { + "epoch": 5.803777687185759, + "grad_norm": 0.7087032198905945, + "learning_rate": 0.0002745277891017801, + "loss": 3.3046, + "step": 85420 + }, + { + "epoch": 5.80411740725642, + "grad_norm": 1.1888831853866577, + "learning_rate": 0.0002744853240929474, + "loss": 3.2063, + "step": 85425 + }, + { + "epoch": 5.804457127327082, + "grad_norm": 1.206748366355896, + "learning_rate": 0.00027444285908411473, + "loss": 3.2164, + "step": 85430 + }, + { + "epoch": 5.8047968473977445, + "grad_norm": 1.1295928955078125, + "learning_rate": 0.000274400394075282, + "loss": 3.7114, + "step": 85435 + }, + { + "epoch": 5.805136567468406, + "grad_norm": 0.7978733777999878, + "learning_rate": 0.00027435792906644924, + "loss": 3.284, + "step": 85440 + }, + { + "epoch": 5.805476287539068, + "grad_norm": 0.8604005575180054, + "learning_rate": 0.0002743154640576165, + "loss": 3.5307, + "step": 85445 + }, + { + "epoch": 5.80581600760973, + "grad_norm": 1.2144805192947388, + "learning_rate": 0.00027427299904878385, + "loss": 3.4398, + "step": 85450 + }, + { + "epoch": 5.806155727680391, + "grad_norm": 0.992076575756073, + "learning_rate": 0.0002742305340399511, + "loss": 3.5094, + "step": 85455 + }, + { + "epoch": 5.806495447751053, + "grad_norm": 0.8304373025894165, + "learning_rate": 0.00027418806903111836, + "loss": 3.3764, + "step": 85460 + }, + { + "epoch": 5.806835167821715, + "grad_norm": 0.7868382930755615, + "learning_rate": 0.0002741456040222857, + "loss": 3.5291, + "step": 85465 + }, + { + "epoch": 5.807174887892376, + "grad_norm": 0.9606617093086243, + "learning_rate": 0.0002741031390134529, + "loss": 3.5315, + "step": 85470 + }, + { + "epoch": 5.807514607963038, + "grad_norm": 0.8384671807289124, + "learning_rate": 0.0002740606740046202, + "loss": 3.4263, + "step": 85475 + }, + { + "epoch": 5.8078543280337005, + "grad_norm": 0.7339631915092468, + "learning_rate": 0.0002740182089957875, + "loss": 3.4639, + "step": 85480 + }, + { + "epoch": 5.808194048104362, + "grad_norm": 0.9288833141326904, + "learning_rate": 0.00027397574398695476, + "loss": 3.5329, + "step": 85485 + }, + { + "epoch": 5.808533768175024, + "grad_norm": 0.8140423893928528, + "learning_rate": 0.00027393327897812204, + "loss": 3.5551, + "step": 85490 + }, + { + "epoch": 5.808873488245686, + "grad_norm": 0.9124660491943359, + "learning_rate": 0.0002738908139692893, + "loss": 3.3516, + "step": 85495 + }, + { + "epoch": 5.809213208316347, + "grad_norm": 0.9237702488899231, + "learning_rate": 0.0002738483489604566, + "loss": 3.4477, + "step": 85500 + }, + { + "epoch": 5.809552928387009, + "grad_norm": 1.0749143362045288, + "learning_rate": 0.0002738058839516239, + "loss": 3.3363, + "step": 85505 + }, + { + "epoch": 5.809892648457671, + "grad_norm": 0.7769162058830261, + "learning_rate": 0.00027376341894279116, + "loss": 3.3781, + "step": 85510 + }, + { + "epoch": 5.810232368528332, + "grad_norm": 1.0294381380081177, + "learning_rate": 0.0002737209539339584, + "loss": 3.4317, + "step": 85515 + }, + { + "epoch": 5.810572088598994, + "grad_norm": 0.7882481813430786, + "learning_rate": 0.0002736784889251257, + "loss": 3.4383, + "step": 85520 + }, + { + "epoch": 5.8109118086696565, + "grad_norm": 0.9073072671890259, + "learning_rate": 0.000273636023916293, + "loss": 3.7361, + "step": 85525 + }, + { + "epoch": 5.811251528740318, + "grad_norm": 1.0004278421401978, + "learning_rate": 0.0002735935589074602, + "loss": 3.5537, + "step": 85530 + }, + { + "epoch": 5.81159124881098, + "grad_norm": 0.849044680595398, + "learning_rate": 0.00027355109389862756, + "loss": 3.6018, + "step": 85535 + }, + { + "epoch": 5.811930968881642, + "grad_norm": 1.3402457237243652, + "learning_rate": 0.00027350862888979484, + "loss": 3.2131, + "step": 85540 + }, + { + "epoch": 5.812270688952303, + "grad_norm": 0.7930300235748291, + "learning_rate": 0.00027346616388096206, + "loss": 3.4557, + "step": 85545 + }, + { + "epoch": 5.812610409022965, + "grad_norm": 0.8929469585418701, + "learning_rate": 0.00027342369887212934, + "loss": 3.3156, + "step": 85550 + }, + { + "epoch": 5.812950129093627, + "grad_norm": 0.9994937777519226, + "learning_rate": 0.0002733812338632967, + "loss": 3.5438, + "step": 85555 + }, + { + "epoch": 5.813289849164288, + "grad_norm": 0.7199445962905884, + "learning_rate": 0.0002733387688544639, + "loss": 3.4977, + "step": 85560 + }, + { + "epoch": 5.8136295692349504, + "grad_norm": 0.6542371511459351, + "learning_rate": 0.0002732963038456312, + "loss": 3.5139, + "step": 85565 + }, + { + "epoch": 5.8139692893056125, + "grad_norm": 0.8727741837501526, + "learning_rate": 0.0002732538388367985, + "loss": 3.3795, + "step": 85570 + }, + { + "epoch": 5.814309009376274, + "grad_norm": 0.9861164689064026, + "learning_rate": 0.00027321137382796574, + "loss": 3.5103, + "step": 85575 + }, + { + "epoch": 5.814648729446936, + "grad_norm": 0.9636548757553101, + "learning_rate": 0.000273168908819133, + "loss": 3.57, + "step": 85580 + }, + { + "epoch": 5.814988449517598, + "grad_norm": 0.8657228350639343, + "learning_rate": 0.0002731264438103003, + "loss": 3.5232, + "step": 85585 + }, + { + "epoch": 5.815328169588259, + "grad_norm": 1.0021485090255737, + "learning_rate": 0.0002730839788014676, + "loss": 3.2022, + "step": 85590 + }, + { + "epoch": 5.815667889658921, + "grad_norm": 1.1353603601455688, + "learning_rate": 0.00027304151379263487, + "loss": 3.4679, + "step": 85595 + }, + { + "epoch": 5.816007609729583, + "grad_norm": 1.0067800283432007, + "learning_rate": 0.00027299904878380215, + "loss": 3.4992, + "step": 85600 + }, + { + "epoch": 5.816347329800244, + "grad_norm": 0.7749165296554565, + "learning_rate": 0.0002729565837749695, + "loss": 3.3805, + "step": 85605 + }, + { + "epoch": 5.8166870498709065, + "grad_norm": 1.0942476987838745, + "learning_rate": 0.0002729141187661367, + "loss": 3.3111, + "step": 85610 + }, + { + "epoch": 5.8170267699415685, + "grad_norm": 0.965944766998291, + "learning_rate": 0.000272871653757304, + "loss": 3.342, + "step": 85615 + }, + { + "epoch": 5.81736649001223, + "grad_norm": 0.890684187412262, + "learning_rate": 0.00027282918874847127, + "loss": 3.2842, + "step": 85620 + }, + { + "epoch": 5.817706210082892, + "grad_norm": 1.121200680732727, + "learning_rate": 0.00027278672373963855, + "loss": 3.3264, + "step": 85625 + }, + { + "epoch": 5.818045930153554, + "grad_norm": 0.794026255607605, + "learning_rate": 0.0002727442587308058, + "loss": 3.4603, + "step": 85630 + }, + { + "epoch": 5.818385650224215, + "grad_norm": 0.9887314438819885, + "learning_rate": 0.0002727017937219731, + "loss": 3.4582, + "step": 85635 + }, + { + "epoch": 5.818725370294877, + "grad_norm": 0.8179646134376526, + "learning_rate": 0.0002726593287131404, + "loss": 3.456, + "step": 85640 + }, + { + "epoch": 5.819065090365539, + "grad_norm": 1.057542085647583, + "learning_rate": 0.00027261686370430767, + "loss": 3.3893, + "step": 85645 + }, + { + "epoch": 5.8194048104362, + "grad_norm": 1.0698884725570679, + "learning_rate": 0.00027257439869547495, + "loss": 3.5208, + "step": 85650 + }, + { + "epoch": 5.8197445305068625, + "grad_norm": 0.8251063227653503, + "learning_rate": 0.00027253193368664217, + "loss": 3.5329, + "step": 85655 + }, + { + "epoch": 5.8200842505775245, + "grad_norm": 1.224420428276062, + "learning_rate": 0.0002724894686778095, + "loss": 3.5612, + "step": 85660 + }, + { + "epoch": 5.820423970648186, + "grad_norm": 0.8822425603866577, + "learning_rate": 0.0002724470036689768, + "loss": 3.5652, + "step": 85665 + }, + { + "epoch": 5.820763690718848, + "grad_norm": 0.888246476650238, + "learning_rate": 0.000272404538660144, + "loss": 3.5334, + "step": 85670 + }, + { + "epoch": 5.82110341078951, + "grad_norm": 0.84879070520401, + "learning_rate": 0.00027236207365131135, + "loss": 3.4567, + "step": 85675 + }, + { + "epoch": 5.821443130860171, + "grad_norm": 1.0686910152435303, + "learning_rate": 0.0002723196086424786, + "loss": 3.5227, + "step": 85680 + }, + { + "epoch": 5.821782850930833, + "grad_norm": 0.7883431315422058, + "learning_rate": 0.00027227714363364585, + "loss": 3.2565, + "step": 85685 + }, + { + "epoch": 5.822122571001495, + "grad_norm": 0.9040455222129822, + "learning_rate": 0.00027223467862481313, + "loss": 3.3286, + "step": 85690 + }, + { + "epoch": 5.822462291072156, + "grad_norm": 0.9537210464477539, + "learning_rate": 0.00027219221361598047, + "loss": 3.6048, + "step": 85695 + }, + { + "epoch": 5.8228020111428185, + "grad_norm": 1.1601464748382568, + "learning_rate": 0.0002721497486071477, + "loss": 3.4285, + "step": 85700 + }, + { + "epoch": 5.8231417312134806, + "grad_norm": 1.0626112222671509, + "learning_rate": 0.00027210728359831497, + "loss": 3.4138, + "step": 85705 + }, + { + "epoch": 5.823481451284142, + "grad_norm": 0.8661244511604309, + "learning_rate": 0.0002720648185894823, + "loss": 3.3547, + "step": 85710 + }, + { + "epoch": 5.823821171354804, + "grad_norm": 0.8774874210357666, + "learning_rate": 0.00027202235358064953, + "loss": 3.225, + "step": 85715 + }, + { + "epoch": 5.824160891425466, + "grad_norm": 0.9934190511703491, + "learning_rate": 0.0002719798885718168, + "loss": 3.5714, + "step": 85720 + }, + { + "epoch": 5.824500611496127, + "grad_norm": 1.0539610385894775, + "learning_rate": 0.00027193742356298415, + "loss": 3.4123, + "step": 85725 + }, + { + "epoch": 5.824840331566789, + "grad_norm": 1.0325329303741455, + "learning_rate": 0.0002718949585541514, + "loss": 3.5002, + "step": 85730 + }, + { + "epoch": 5.825180051637451, + "grad_norm": 1.203180193901062, + "learning_rate": 0.00027185249354531865, + "loss": 3.2958, + "step": 85735 + }, + { + "epoch": 5.825519771708112, + "grad_norm": 0.8976393938064575, + "learning_rate": 0.00027181002853648593, + "loss": 3.4176, + "step": 85740 + }, + { + "epoch": 5.8258594917787745, + "grad_norm": 0.9067656397819519, + "learning_rate": 0.0002717675635276532, + "loss": 3.4575, + "step": 85745 + }, + { + "epoch": 5.826199211849436, + "grad_norm": 0.9201294779777527, + "learning_rate": 0.0002717250985188205, + "loss": 3.5453, + "step": 85750 + }, + { + "epoch": 5.826538931920098, + "grad_norm": 1.0084196329116821, + "learning_rate": 0.0002716826335099878, + "loss": 3.3215, + "step": 85755 + }, + { + "epoch": 5.82687865199076, + "grad_norm": 0.9513784646987915, + "learning_rate": 0.00027164016850115505, + "loss": 3.175, + "step": 85760 + }, + { + "epoch": 5.827218372061421, + "grad_norm": 1.898790717124939, + "learning_rate": 0.00027159770349232233, + "loss": 3.7159, + "step": 85765 + }, + { + "epoch": 5.827558092132083, + "grad_norm": 1.0209416151046753, + "learning_rate": 0.0002715552384834896, + "loss": 3.3393, + "step": 85770 + }, + { + "epoch": 5.827897812202745, + "grad_norm": 0.968442440032959, + "learning_rate": 0.0002715127734746569, + "loss": 3.173, + "step": 85775 + }, + { + "epoch": 5.828237532273406, + "grad_norm": 0.784275233745575, + "learning_rate": 0.0002714703084658242, + "loss": 3.4726, + "step": 85780 + }, + { + "epoch": 5.828577252344068, + "grad_norm": 0.9052839875221252, + "learning_rate": 0.00027142784345699145, + "loss": 3.5144, + "step": 85785 + }, + { + "epoch": 5.8289169724147305, + "grad_norm": 0.9009918570518494, + "learning_rate": 0.00027138537844815873, + "loss": 3.5782, + "step": 85790 + }, + { + "epoch": 5.829256692485392, + "grad_norm": 0.7110707759857178, + "learning_rate": 0.000271342913439326, + "loss": 3.5126, + "step": 85795 + }, + { + "epoch": 5.829596412556054, + "grad_norm": 0.8679285645484924, + "learning_rate": 0.0002713004484304933, + "loss": 3.2702, + "step": 85800 + }, + { + "epoch": 5.829936132626716, + "grad_norm": 1.185449242591858, + "learning_rate": 0.0002712579834216606, + "loss": 3.3249, + "step": 85805 + }, + { + "epoch": 5.830275852697377, + "grad_norm": 0.8555469512939453, + "learning_rate": 0.0002712155184128278, + "loss": 3.5668, + "step": 85810 + }, + { + "epoch": 5.830615572768039, + "grad_norm": 0.9341093301773071, + "learning_rate": 0.00027117305340399513, + "loss": 3.2412, + "step": 85815 + }, + { + "epoch": 5.830955292838701, + "grad_norm": 0.8699649572372437, + "learning_rate": 0.0002711305883951624, + "loss": 3.4434, + "step": 85820 + }, + { + "epoch": 5.831295012909362, + "grad_norm": 0.855373740196228, + "learning_rate": 0.00027108812338632964, + "loss": 3.6179, + "step": 85825 + }, + { + "epoch": 5.831634732980024, + "grad_norm": 0.9894996881484985, + "learning_rate": 0.000271045658377497, + "loss": 3.4534, + "step": 85830 + }, + { + "epoch": 5.8319744530506865, + "grad_norm": 0.7085078954696655, + "learning_rate": 0.00027100319336866425, + "loss": 3.7151, + "step": 85835 + }, + { + "epoch": 5.832314173121348, + "grad_norm": 0.8549900054931641, + "learning_rate": 0.0002709607283598315, + "loss": 3.6032, + "step": 85840 + }, + { + "epoch": 5.83265389319201, + "grad_norm": 0.9382591247558594, + "learning_rate": 0.00027091826335099876, + "loss": 3.4548, + "step": 85845 + }, + { + "epoch": 5.832993613262672, + "grad_norm": 0.9508116841316223, + "learning_rate": 0.0002708757983421661, + "loss": 3.3739, + "step": 85850 + }, + { + "epoch": 5.833333333333333, + "grad_norm": 0.8172245025634766, + "learning_rate": 0.0002708333333333333, + "loss": 3.6782, + "step": 85855 + }, + { + "epoch": 5.833673053403995, + "grad_norm": 0.8330725431442261, + "learning_rate": 0.0002707908683245006, + "loss": 3.5548, + "step": 85860 + }, + { + "epoch": 5.834012773474657, + "grad_norm": 0.9419605135917664, + "learning_rate": 0.00027074840331566794, + "loss": 3.3676, + "step": 85865 + }, + { + "epoch": 5.834352493545318, + "grad_norm": 0.8327247500419617, + "learning_rate": 0.00027070593830683516, + "loss": 3.5301, + "step": 85870 + }, + { + "epoch": 5.8346922136159804, + "grad_norm": 1.0354814529418945, + "learning_rate": 0.00027066347329800244, + "loss": 3.2892, + "step": 85875 + }, + { + "epoch": 5.8350319336866425, + "grad_norm": 0.8013768196105957, + "learning_rate": 0.0002706210082891697, + "loss": 3.4534, + "step": 85880 + }, + { + "epoch": 5.835371653757304, + "grad_norm": 0.8095645904541016, + "learning_rate": 0.000270578543280337, + "loss": 3.3046, + "step": 85885 + }, + { + "epoch": 5.835711373827966, + "grad_norm": 1.0984495878219604, + "learning_rate": 0.0002705360782715043, + "loss": 3.6689, + "step": 85890 + }, + { + "epoch": 5.836051093898628, + "grad_norm": 0.942044198513031, + "learning_rate": 0.00027049361326267156, + "loss": 3.4435, + "step": 85895 + }, + { + "epoch": 5.836390813969289, + "grad_norm": 0.9594922661781311, + "learning_rate": 0.00027045114825383884, + "loss": 3.3226, + "step": 85900 + }, + { + "epoch": 5.836730534039951, + "grad_norm": 0.992332398891449, + "learning_rate": 0.0002704086832450061, + "loss": 3.4097, + "step": 85905 + }, + { + "epoch": 5.837070254110613, + "grad_norm": 0.993399977684021, + "learning_rate": 0.0002703662182361734, + "loss": 3.4901, + "step": 85910 + }, + { + "epoch": 5.837409974181274, + "grad_norm": 0.9158075451850891, + "learning_rate": 0.00027032375322734063, + "loss": 3.3584, + "step": 85915 + }, + { + "epoch": 5.8377496942519365, + "grad_norm": 0.9005814790725708, + "learning_rate": 0.00027028128821850796, + "loss": 3.5667, + "step": 85920 + }, + { + "epoch": 5.8380894143225985, + "grad_norm": 0.692868173122406, + "learning_rate": 0.00027023882320967524, + "loss": 3.4387, + "step": 85925 + }, + { + "epoch": 5.83842913439326, + "grad_norm": 0.9008669853210449, + "learning_rate": 0.00027019635820084247, + "loss": 3.6579, + "step": 85930 + }, + { + "epoch": 5.838768854463922, + "grad_norm": 1.2968273162841797, + "learning_rate": 0.0002701538931920098, + "loss": 3.2465, + "step": 85935 + }, + { + "epoch": 5.839108574534584, + "grad_norm": 0.7343127131462097, + "learning_rate": 0.0002701114281831771, + "loss": 3.2198, + "step": 85940 + }, + { + "epoch": 5.839448294605245, + "grad_norm": 0.8652023673057556, + "learning_rate": 0.00027006896317434436, + "loss": 3.4782, + "step": 85945 + }, + { + "epoch": 5.839788014675907, + "grad_norm": 0.7940728664398193, + "learning_rate": 0.0002700264981655116, + "loss": 3.6836, + "step": 85950 + }, + { + "epoch": 5.840127734746569, + "grad_norm": 0.8610370755195618, + "learning_rate": 0.0002699840331566789, + "loss": 3.5552, + "step": 85955 + }, + { + "epoch": 5.84046745481723, + "grad_norm": 0.8352829813957214, + "learning_rate": 0.0002699415681478462, + "loss": 3.3348, + "step": 85960 + }, + { + "epoch": 5.8408071748878925, + "grad_norm": 0.8985799551010132, + "learning_rate": 0.00026989910313901343, + "loss": 3.3019, + "step": 85965 + }, + { + "epoch": 5.841146894958554, + "grad_norm": 1.1118210554122925, + "learning_rate": 0.00026985663813018076, + "loss": 3.1513, + "step": 85970 + }, + { + "epoch": 5.841486615029216, + "grad_norm": 0.8033086061477661, + "learning_rate": 0.00026981417312134804, + "loss": 3.5659, + "step": 85975 + }, + { + "epoch": 5.841826335099878, + "grad_norm": 0.8585745692253113, + "learning_rate": 0.00026977170811251527, + "loss": 3.3805, + "step": 85980 + }, + { + "epoch": 5.842166055170539, + "grad_norm": 1.4445583820343018, + "learning_rate": 0.00026972924310368255, + "loss": 3.8648, + "step": 85985 + }, + { + "epoch": 5.842505775241201, + "grad_norm": 1.1817612648010254, + "learning_rate": 0.0002696867780948499, + "loss": 3.4753, + "step": 85990 + }, + { + "epoch": 5.842845495311863, + "grad_norm": 0.7003769874572754, + "learning_rate": 0.0002696443130860171, + "loss": 3.4123, + "step": 85995 + }, + { + "epoch": 5.843185215382524, + "grad_norm": 0.9787039756774902, + "learning_rate": 0.0002696018480771844, + "loss": 3.1584, + "step": 86000 + }, + { + "epoch": 5.843524935453186, + "grad_norm": 0.9772524833679199, + "learning_rate": 0.0002695593830683517, + "loss": 3.3732, + "step": 86005 + }, + { + "epoch": 5.8438646555238485, + "grad_norm": 0.9459580183029175, + "learning_rate": 0.00026951691805951895, + "loss": 3.5668, + "step": 86010 + }, + { + "epoch": 5.84420437559451, + "grad_norm": 1.0837233066558838, + "learning_rate": 0.00026947445305068623, + "loss": 3.434, + "step": 86015 + }, + { + "epoch": 5.844544095665172, + "grad_norm": 0.8834533095359802, + "learning_rate": 0.00026943198804185356, + "loss": 3.375, + "step": 86020 + }, + { + "epoch": 5.844883815735834, + "grad_norm": 0.9668422341346741, + "learning_rate": 0.0002693895230330208, + "loss": 3.5324, + "step": 86025 + }, + { + "epoch": 5.845223535806495, + "grad_norm": 1.0426534414291382, + "learning_rate": 0.00026934705802418807, + "loss": 3.5978, + "step": 86030 + }, + { + "epoch": 5.845563255877157, + "grad_norm": 1.7355109453201294, + "learning_rate": 0.00026930459301535535, + "loss": 3.4554, + "step": 86035 + }, + { + "epoch": 5.845902975947819, + "grad_norm": 0.9771691560745239, + "learning_rate": 0.00026926212800652263, + "loss": 3.4304, + "step": 86040 + }, + { + "epoch": 5.84624269601848, + "grad_norm": 0.9578458666801453, + "learning_rate": 0.0002692196629976899, + "loss": 3.4267, + "step": 86045 + }, + { + "epoch": 5.846582416089142, + "grad_norm": 0.9304202795028687, + "learning_rate": 0.0002691771979888572, + "loss": 3.2788, + "step": 86050 + }, + { + "epoch": 5.8469221361598045, + "grad_norm": 1.0265262126922607, + "learning_rate": 0.00026913473298002447, + "loss": 3.1988, + "step": 86055 + }, + { + "epoch": 5.847261856230466, + "grad_norm": 0.8590037226676941, + "learning_rate": 0.00026909226797119175, + "loss": 3.4472, + "step": 86060 + }, + { + "epoch": 5.847601576301128, + "grad_norm": 0.8802750706672668, + "learning_rate": 0.00026904980296235903, + "loss": 3.226, + "step": 86065 + }, + { + "epoch": 5.84794129637179, + "grad_norm": 0.9002315402030945, + "learning_rate": 0.00026900733795352626, + "loss": 3.2233, + "step": 86070 + }, + { + "epoch": 5.848281016442451, + "grad_norm": 0.8698520660400391, + "learning_rate": 0.0002689648729446936, + "loss": 3.4278, + "step": 86075 + }, + { + "epoch": 5.848620736513113, + "grad_norm": 1.0070141553878784, + "learning_rate": 0.00026892240793586087, + "loss": 3.2801, + "step": 86080 + }, + { + "epoch": 5.848960456583775, + "grad_norm": 0.9052004218101501, + "learning_rate": 0.0002688799429270281, + "loss": 3.374, + "step": 86085 + }, + { + "epoch": 5.849300176654436, + "grad_norm": 0.8487313985824585, + "learning_rate": 0.00026883747791819543, + "loss": 3.5366, + "step": 86090 + }, + { + "epoch": 5.849639896725098, + "grad_norm": 0.8681938648223877, + "learning_rate": 0.0002687950129093627, + "loss": 3.3462, + "step": 86095 + }, + { + "epoch": 5.8499796167957605, + "grad_norm": 1.2344688177108765, + "learning_rate": 0.00026875254790052994, + "loss": 3.238, + "step": 86100 + }, + { + "epoch": 5.850319336866422, + "grad_norm": 0.7453005313873291, + "learning_rate": 0.0002687100828916972, + "loss": 3.4132, + "step": 86105 + }, + { + "epoch": 5.850659056937084, + "grad_norm": 0.7762342095375061, + "learning_rate": 0.00026866761788286455, + "loss": 3.3458, + "step": 86110 + }, + { + "epoch": 5.850998777007746, + "grad_norm": 0.8245344758033752, + "learning_rate": 0.00026862515287403183, + "loss": 3.263, + "step": 86115 + }, + { + "epoch": 5.851338497078407, + "grad_norm": 1.153594970703125, + "learning_rate": 0.00026858268786519906, + "loss": 3.4077, + "step": 86120 + }, + { + "epoch": 5.851678217149069, + "grad_norm": 0.8248249292373657, + "learning_rate": 0.0002685402228563664, + "loss": 3.2343, + "step": 86125 + }, + { + "epoch": 5.852017937219731, + "grad_norm": 0.7167032957077026, + "learning_rate": 0.00026849775784753367, + "loss": 3.4012, + "step": 86130 + }, + { + "epoch": 5.852357657290392, + "grad_norm": 1.0033481121063232, + "learning_rate": 0.0002684552928387009, + "loss": 3.3844, + "step": 86135 + }, + { + "epoch": 5.852697377361054, + "grad_norm": 0.9826151728630066, + "learning_rate": 0.0002684128278298682, + "loss": 3.4357, + "step": 86140 + }, + { + "epoch": 5.8530370974317165, + "grad_norm": 0.7771371006965637, + "learning_rate": 0.0002683703628210355, + "loss": 3.5191, + "step": 86145 + }, + { + "epoch": 5.853376817502378, + "grad_norm": 0.7295055985450745, + "learning_rate": 0.00026832789781220274, + "loss": 3.4548, + "step": 86150 + }, + { + "epoch": 5.85371653757304, + "grad_norm": 0.858648955821991, + "learning_rate": 0.00026828543280337, + "loss": 3.4194, + "step": 86155 + }, + { + "epoch": 5.854056257643702, + "grad_norm": 0.7147545218467712, + "learning_rate": 0.00026824296779453735, + "loss": 3.1952, + "step": 86160 + }, + { + "epoch": 5.854395977714363, + "grad_norm": 0.8920286297798157, + "learning_rate": 0.0002682005027857046, + "loss": 3.6288, + "step": 86165 + }, + { + "epoch": 5.854735697785025, + "grad_norm": 1.2790724039077759, + "learning_rate": 0.00026815803777687186, + "loss": 3.5598, + "step": 86170 + }, + { + "epoch": 5.855075417855687, + "grad_norm": 0.9622620940208435, + "learning_rate": 0.00026811557276803914, + "loss": 3.378, + "step": 86175 + }, + { + "epoch": 5.855415137926348, + "grad_norm": 0.8299192786216736, + "learning_rate": 0.0002680731077592064, + "loss": 3.556, + "step": 86180 + }, + { + "epoch": 5.8557548579970105, + "grad_norm": 0.8470836877822876, + "learning_rate": 0.0002680306427503737, + "loss": 3.5028, + "step": 86185 + }, + { + "epoch": 5.8560945780676725, + "grad_norm": 0.8886511325836182, + "learning_rate": 0.000267988177741541, + "loss": 3.4517, + "step": 86190 + }, + { + "epoch": 5.856434298138334, + "grad_norm": 0.7392469048500061, + "learning_rate": 0.00026794571273270826, + "loss": 3.4395, + "step": 86195 + }, + { + "epoch": 5.856774018208996, + "grad_norm": 0.8894254565238953, + "learning_rate": 0.00026790324772387554, + "loss": 3.3444, + "step": 86200 + }, + { + "epoch": 5.857113738279658, + "grad_norm": 1.0719871520996094, + "learning_rate": 0.0002678607827150428, + "loss": 3.4834, + "step": 86205 + }, + { + "epoch": 5.857453458350319, + "grad_norm": 1.5331881046295166, + "learning_rate": 0.00026781831770621004, + "loss": 3.4178, + "step": 86210 + }, + { + "epoch": 5.857793178420981, + "grad_norm": 1.074005365371704, + "learning_rate": 0.0002677758526973774, + "loss": 3.3821, + "step": 86215 + }, + { + "epoch": 5.858132898491643, + "grad_norm": 0.776703953742981, + "learning_rate": 0.00026773338768854466, + "loss": 3.4549, + "step": 86220 + }, + { + "epoch": 5.858472618562304, + "grad_norm": 0.948914647102356, + "learning_rate": 0.0002676909226797119, + "loss": 3.2463, + "step": 86225 + }, + { + "epoch": 5.8588123386329665, + "grad_norm": 0.8874778747558594, + "learning_rate": 0.0002676484576708792, + "loss": 3.5903, + "step": 86230 + }, + { + "epoch": 5.8591520587036285, + "grad_norm": 0.8263382315635681, + "learning_rate": 0.0002676059926620465, + "loss": 3.2143, + "step": 86235 + }, + { + "epoch": 5.85949177877429, + "grad_norm": 0.7853386998176575, + "learning_rate": 0.0002675635276532137, + "loss": 3.3434, + "step": 86240 + }, + { + "epoch": 5.859831498844952, + "grad_norm": 0.824016273021698, + "learning_rate": 0.000267521062644381, + "loss": 3.5763, + "step": 86245 + }, + { + "epoch": 5.860171218915614, + "grad_norm": 1.3476810455322266, + "learning_rate": 0.00026747859763554834, + "loss": 3.3429, + "step": 86250 + }, + { + "epoch": 5.860510938986275, + "grad_norm": 0.8506530523300171, + "learning_rate": 0.00026743613262671556, + "loss": 3.2041, + "step": 86255 + }, + { + "epoch": 5.860850659056937, + "grad_norm": 1.1093811988830566, + "learning_rate": 0.00026739366761788284, + "loss": 3.4016, + "step": 86260 + }, + { + "epoch": 5.861190379127599, + "grad_norm": 0.8833467364311218, + "learning_rate": 0.0002673512026090502, + "loss": 3.3799, + "step": 86265 + }, + { + "epoch": 5.86153009919826, + "grad_norm": 1.0075180530548096, + "learning_rate": 0.0002673087376002174, + "loss": 3.5542, + "step": 86270 + }, + { + "epoch": 5.8618698192689225, + "grad_norm": 0.9027162194252014, + "learning_rate": 0.0002672662725913847, + "loss": 3.5145, + "step": 86275 + }, + { + "epoch": 5.8622095393395846, + "grad_norm": 0.7770543694496155, + "learning_rate": 0.000267223807582552, + "loss": 3.4825, + "step": 86280 + }, + { + "epoch": 5.862549259410246, + "grad_norm": 0.9155957102775574, + "learning_rate": 0.0002671813425737193, + "loss": 3.0356, + "step": 86285 + }, + { + "epoch": 5.862888979480908, + "grad_norm": 0.79121333360672, + "learning_rate": 0.0002671388775648865, + "loss": 3.2668, + "step": 86290 + }, + { + "epoch": 5.86322869955157, + "grad_norm": 1.051538109779358, + "learning_rate": 0.0002670964125560538, + "loss": 3.6053, + "step": 86295 + }, + { + "epoch": 5.863568419622231, + "grad_norm": 0.7549198865890503, + "learning_rate": 0.00026705394754722114, + "loss": 3.2647, + "step": 86300 + }, + { + "epoch": 5.863908139692893, + "grad_norm": 0.7794956564903259, + "learning_rate": 0.00026701148253838837, + "loss": 3.6437, + "step": 86305 + }, + { + "epoch": 5.864247859763555, + "grad_norm": 1.0626429319381714, + "learning_rate": 0.00026696901752955565, + "loss": 3.7291, + "step": 86310 + }, + { + "epoch": 5.864587579834216, + "grad_norm": 0.8733043670654297, + "learning_rate": 0.000266926552520723, + "loss": 3.7, + "step": 86315 + }, + { + "epoch": 5.8649272999048785, + "grad_norm": 0.8317854404449463, + "learning_rate": 0.0002668840875118902, + "loss": 3.6175, + "step": 86320 + }, + { + "epoch": 5.865267019975541, + "grad_norm": 0.8838074803352356, + "learning_rate": 0.0002668416225030575, + "loss": 3.3105, + "step": 86325 + }, + { + "epoch": 5.865606740046202, + "grad_norm": 0.8156776428222656, + "learning_rate": 0.00026679915749422477, + "loss": 3.5435, + "step": 86330 + }, + { + "epoch": 5.865946460116864, + "grad_norm": 1.0593010187149048, + "learning_rate": 0.00026675669248539205, + "loss": 3.4216, + "step": 86335 + }, + { + "epoch": 5.866286180187526, + "grad_norm": 0.9067009091377258, + "learning_rate": 0.0002667142274765593, + "loss": 3.6014, + "step": 86340 + }, + { + "epoch": 5.866625900258187, + "grad_norm": 1.1322778463363647, + "learning_rate": 0.0002666717624677266, + "loss": 3.1432, + "step": 86345 + }, + { + "epoch": 5.866965620328849, + "grad_norm": 0.7949217557907104, + "learning_rate": 0.0002666292974588939, + "loss": 3.3763, + "step": 86350 + }, + { + "epoch": 5.867305340399511, + "grad_norm": 0.8863410353660583, + "learning_rate": 0.00026658683245006117, + "loss": 3.3348, + "step": 86355 + }, + { + "epoch": 5.867645060470172, + "grad_norm": 0.8905214667320251, + "learning_rate": 0.00026654436744122845, + "loss": 3.2738, + "step": 86360 + }, + { + "epoch": 5.8679847805408345, + "grad_norm": 0.8957790732383728, + "learning_rate": 0.00026650190243239567, + "loss": 3.5278, + "step": 86365 + }, + { + "epoch": 5.868324500611497, + "grad_norm": 0.7183364629745483, + "learning_rate": 0.000266459437423563, + "loss": 3.4109, + "step": 86370 + }, + { + "epoch": 5.868664220682158, + "grad_norm": 0.9085429310798645, + "learning_rate": 0.0002664169724147303, + "loss": 3.7691, + "step": 86375 + }, + { + "epoch": 5.86900394075282, + "grad_norm": 1.0386015176773071, + "learning_rate": 0.0002663745074058975, + "loss": 3.3952, + "step": 86380 + }, + { + "epoch": 5.869343660823482, + "grad_norm": 0.9703196287155151, + "learning_rate": 0.00026633204239706485, + "loss": 3.3524, + "step": 86385 + }, + { + "epoch": 5.869683380894143, + "grad_norm": 0.8766586184501648, + "learning_rate": 0.0002662895773882321, + "loss": 3.5959, + "step": 86390 + }, + { + "epoch": 5.870023100964805, + "grad_norm": 0.9453226327896118, + "learning_rate": 0.00026624711237939935, + "loss": 3.3516, + "step": 86395 + }, + { + "epoch": 5.870362821035467, + "grad_norm": 0.8838901519775391, + "learning_rate": 0.00026620464737056663, + "loss": 3.5589, + "step": 86400 + }, + { + "epoch": 5.870702541106128, + "grad_norm": 1.005314826965332, + "learning_rate": 0.00026616218236173397, + "loss": 3.45, + "step": 86405 + }, + { + "epoch": 5.8710422611767905, + "grad_norm": 0.9389330148696899, + "learning_rate": 0.0002661197173529012, + "loss": 3.3672, + "step": 86410 + }, + { + "epoch": 5.871381981247453, + "grad_norm": 0.736731231212616, + "learning_rate": 0.00026607725234406847, + "loss": 3.5602, + "step": 86415 + }, + { + "epoch": 5.871721701318114, + "grad_norm": 0.8334453701972961, + "learning_rate": 0.0002660347873352358, + "loss": 3.4295, + "step": 86420 + }, + { + "epoch": 5.872061421388776, + "grad_norm": 1.2921814918518066, + "learning_rate": 0.00026599232232640303, + "loss": 3.2611, + "step": 86425 + }, + { + "epoch": 5.872401141459437, + "grad_norm": 0.9349871277809143, + "learning_rate": 0.0002659498573175703, + "loss": 3.4054, + "step": 86430 + }, + { + "epoch": 5.872740861530099, + "grad_norm": 0.8525412082672119, + "learning_rate": 0.0002659073923087376, + "loss": 3.5783, + "step": 86435 + }, + { + "epoch": 5.873080581600761, + "grad_norm": 0.9275960922241211, + "learning_rate": 0.0002658649272999049, + "loss": 3.1863, + "step": 86440 + }, + { + "epoch": 5.873420301671422, + "grad_norm": 0.8481304049491882, + "learning_rate": 0.00026582246229107215, + "loss": 3.4148, + "step": 86445 + }, + { + "epoch": 5.8737600217420844, + "grad_norm": 2.259323835372925, + "learning_rate": 0.00026577999728223943, + "loss": 3.7092, + "step": 86450 + }, + { + "epoch": 5.8740997418127465, + "grad_norm": 1.061008095741272, + "learning_rate": 0.00026573753227340677, + "loss": 3.4568, + "step": 86455 + }, + { + "epoch": 5.874439461883408, + "grad_norm": 0.8969730138778687, + "learning_rate": 0.000265695067264574, + "loss": 3.4913, + "step": 86460 + }, + { + "epoch": 5.87477918195407, + "grad_norm": 0.9192014932632446, + "learning_rate": 0.0002656526022557413, + "loss": 3.2825, + "step": 86465 + }, + { + "epoch": 5.875118902024732, + "grad_norm": 0.9566301107406616, + "learning_rate": 0.00026561013724690855, + "loss": 3.4987, + "step": 86470 + }, + { + "epoch": 5.875458622095393, + "grad_norm": 0.7487375736236572, + "learning_rate": 0.00026556767223807583, + "loss": 3.6183, + "step": 86475 + }, + { + "epoch": 5.875798342166055, + "grad_norm": 0.7771413922309875, + "learning_rate": 0.0002655252072292431, + "loss": 3.455, + "step": 86480 + }, + { + "epoch": 5.876138062236717, + "grad_norm": 0.8373425602912903, + "learning_rate": 0.0002654827422204104, + "loss": 3.3749, + "step": 86485 + }, + { + "epoch": 5.876477782307378, + "grad_norm": 1.1413938999176025, + "learning_rate": 0.0002654402772115777, + "loss": 3.3825, + "step": 86490 + }, + { + "epoch": 5.8768175023780405, + "grad_norm": 0.8059937953948975, + "learning_rate": 0.00026539781220274495, + "loss": 3.3389, + "step": 86495 + }, + { + "epoch": 5.8771572224487025, + "grad_norm": 1.1402323246002197, + "learning_rate": 0.00026535534719391223, + "loss": 3.4738, + "step": 86500 + }, + { + "epoch": 5.877496942519364, + "grad_norm": 1.1138890981674194, + "learning_rate": 0.00026531288218507946, + "loss": 3.2426, + "step": 86505 + }, + { + "epoch": 5.877836662590026, + "grad_norm": 0.8819926977157593, + "learning_rate": 0.0002652704171762468, + "loss": 3.4258, + "step": 86510 + }, + { + "epoch": 5.878176382660688, + "grad_norm": 0.6907956004142761, + "learning_rate": 0.0002652279521674141, + "loss": 3.4881, + "step": 86515 + }, + { + "epoch": 5.878516102731349, + "grad_norm": 1.0524836778640747, + "learning_rate": 0.0002651854871585813, + "loss": 3.5037, + "step": 86520 + }, + { + "epoch": 5.878855822802011, + "grad_norm": 1.0506043434143066, + "learning_rate": 0.00026514302214974863, + "loss": 3.5152, + "step": 86525 + }, + { + "epoch": 5.879195542872673, + "grad_norm": 0.9730178713798523, + "learning_rate": 0.0002651005571409159, + "loss": 3.014, + "step": 86530 + }, + { + "epoch": 5.879535262943334, + "grad_norm": 1.3492637872695923, + "learning_rate": 0.00026505809213208314, + "loss": 3.4695, + "step": 86535 + }, + { + "epoch": 5.8798749830139965, + "grad_norm": 0.9881723523139954, + "learning_rate": 0.0002650156271232504, + "loss": 3.4573, + "step": 86540 + }, + { + "epoch": 5.8802147030846585, + "grad_norm": 0.8586419224739075, + "learning_rate": 0.00026497316211441775, + "loss": 3.2759, + "step": 86545 + }, + { + "epoch": 5.88055442315532, + "grad_norm": 0.9406324028968811, + "learning_rate": 0.000264930697105585, + "loss": 3.5943, + "step": 86550 + }, + { + "epoch": 5.880894143225982, + "grad_norm": 1.0977766513824463, + "learning_rate": 0.00026488823209675226, + "loss": 3.5053, + "step": 86555 + }, + { + "epoch": 5.881233863296644, + "grad_norm": 1.0892994403839111, + "learning_rate": 0.0002648457670879196, + "loss": 3.5523, + "step": 86560 + }, + { + "epoch": 5.881573583367305, + "grad_norm": 1.0148600339889526, + "learning_rate": 0.0002648033020790868, + "loss": 3.4684, + "step": 86565 + }, + { + "epoch": 5.881913303437967, + "grad_norm": 0.8273751735687256, + "learning_rate": 0.0002647608370702541, + "loss": 3.4998, + "step": 86570 + }, + { + "epoch": 5.882253023508629, + "grad_norm": 0.9430851340293884, + "learning_rate": 0.00026471837206142144, + "loss": 3.3568, + "step": 86575 + }, + { + "epoch": 5.88259274357929, + "grad_norm": 0.8044989109039307, + "learning_rate": 0.00026467590705258866, + "loss": 3.4841, + "step": 86580 + }, + { + "epoch": 5.8829324636499525, + "grad_norm": 0.921524703502655, + "learning_rate": 0.00026463344204375594, + "loss": 3.6414, + "step": 86585 + }, + { + "epoch": 5.883272183720615, + "grad_norm": 1.083160161972046, + "learning_rate": 0.0002645909770349232, + "loss": 3.5524, + "step": 86590 + }, + { + "epoch": 5.883611903791276, + "grad_norm": 1.1565852165222168, + "learning_rate": 0.0002645485120260905, + "loss": 3.1195, + "step": 86595 + }, + { + "epoch": 5.883951623861938, + "grad_norm": 0.927943766117096, + "learning_rate": 0.0002645060470172578, + "loss": 3.4799, + "step": 86600 + }, + { + "epoch": 5.8842913439326, + "grad_norm": 1.0399051904678345, + "learning_rate": 0.00026446358200842506, + "loss": 3.3207, + "step": 86605 + }, + { + "epoch": 5.884631064003261, + "grad_norm": 1.3866750001907349, + "learning_rate": 0.00026442111699959234, + "loss": 3.1983, + "step": 86610 + }, + { + "epoch": 5.884970784073923, + "grad_norm": 0.8107403516769409, + "learning_rate": 0.0002643786519907596, + "loss": 3.5514, + "step": 86615 + }, + { + "epoch": 5.885310504144585, + "grad_norm": 0.9815630912780762, + "learning_rate": 0.0002643361869819269, + "loss": 3.4712, + "step": 86620 + }, + { + "epoch": 5.885650224215246, + "grad_norm": 0.892328679561615, + "learning_rate": 0.0002642937219730942, + "loss": 3.2969, + "step": 86625 + }, + { + "epoch": 5.8859899442859085, + "grad_norm": 0.9207344055175781, + "learning_rate": 0.00026425125696426146, + "loss": 3.4592, + "step": 86630 + }, + { + "epoch": 5.886329664356571, + "grad_norm": 0.7865584492683411, + "learning_rate": 0.00026420879195542874, + "loss": 3.357, + "step": 86635 + }, + { + "epoch": 5.886669384427232, + "grad_norm": 1.0817041397094727, + "learning_rate": 0.000264166326946596, + "loss": 3.4243, + "step": 86640 + }, + { + "epoch": 5.887009104497894, + "grad_norm": 1.0846365690231323, + "learning_rate": 0.0002641238619377633, + "loss": 3.3632, + "step": 86645 + }, + { + "epoch": 5.887348824568555, + "grad_norm": 0.9023453593254089, + "learning_rate": 0.0002640813969289306, + "loss": 3.5626, + "step": 86650 + }, + { + "epoch": 5.887688544639217, + "grad_norm": 0.8457932472229004, + "learning_rate": 0.00026403893192009786, + "loss": 3.5144, + "step": 86655 + }, + { + "epoch": 5.888028264709879, + "grad_norm": 0.973366916179657, + "learning_rate": 0.0002639964669112651, + "loss": 3.5001, + "step": 86660 + }, + { + "epoch": 5.88836798478054, + "grad_norm": 0.7693150639533997, + "learning_rate": 0.0002639540019024324, + "loss": 3.6408, + "step": 86665 + }, + { + "epoch": 5.888707704851202, + "grad_norm": 1.0492876768112183, + "learning_rate": 0.0002639115368935997, + "loss": 3.391, + "step": 86670 + }, + { + "epoch": 5.8890474249218645, + "grad_norm": 1.0527756214141846, + "learning_rate": 0.00026386907188476693, + "loss": 3.2772, + "step": 86675 + }, + { + "epoch": 5.889387144992526, + "grad_norm": 0.9608847498893738, + "learning_rate": 0.00026382660687593426, + "loss": 3.2477, + "step": 86680 + }, + { + "epoch": 5.889726865063188, + "grad_norm": 0.9003650546073914, + "learning_rate": 0.00026378414186710154, + "loss": 3.5715, + "step": 86685 + }, + { + "epoch": 5.89006658513385, + "grad_norm": 0.7668390870094299, + "learning_rate": 0.00026374167685826877, + "loss": 3.2785, + "step": 86690 + }, + { + "epoch": 5.890406305204511, + "grad_norm": 0.9139150381088257, + "learning_rate": 0.00026369921184943605, + "loss": 3.464, + "step": 86695 + }, + { + "epoch": 5.890746025275173, + "grad_norm": 0.8657779693603516, + "learning_rate": 0.0002636567468406034, + "loss": 3.2761, + "step": 86700 + }, + { + "epoch": 5.891085745345835, + "grad_norm": 1.1397814750671387, + "learning_rate": 0.0002636142818317706, + "loss": 3.3309, + "step": 86705 + }, + { + "epoch": 5.891425465416496, + "grad_norm": 1.1026175022125244, + "learning_rate": 0.0002635718168229379, + "loss": 3.5543, + "step": 86710 + }, + { + "epoch": 5.891765185487158, + "grad_norm": 0.9933289885520935, + "learning_rate": 0.0002635293518141052, + "loss": 3.3644, + "step": 86715 + }, + { + "epoch": 5.8921049055578205, + "grad_norm": 0.8410524129867554, + "learning_rate": 0.00026348688680527245, + "loss": 3.3458, + "step": 86720 + }, + { + "epoch": 5.892444625628482, + "grad_norm": 0.9390279650688171, + "learning_rate": 0.00026344442179643973, + "loss": 3.7889, + "step": 86725 + }, + { + "epoch": 5.892784345699144, + "grad_norm": 0.8043792843818665, + "learning_rate": 0.000263401956787607, + "loss": 3.4558, + "step": 86730 + }, + { + "epoch": 5.893124065769806, + "grad_norm": 0.8868273496627808, + "learning_rate": 0.0002633594917787743, + "loss": 3.3326, + "step": 86735 + }, + { + "epoch": 5.893463785840467, + "grad_norm": 0.8429461121559143, + "learning_rate": 0.00026331702676994157, + "loss": 3.6529, + "step": 86740 + }, + { + "epoch": 5.893803505911129, + "grad_norm": 1.2867234945297241, + "learning_rate": 0.00026327456176110885, + "loss": 3.2304, + "step": 86745 + }, + { + "epoch": 5.894143225981791, + "grad_norm": 0.9196484684944153, + "learning_rate": 0.00026323209675227613, + "loss": 3.4653, + "step": 86750 + }, + { + "epoch": 5.894482946052452, + "grad_norm": 0.9040078520774841, + "learning_rate": 0.0002631896317434434, + "loss": 3.4672, + "step": 86755 + }, + { + "epoch": 5.8948226661231145, + "grad_norm": 0.8999331593513489, + "learning_rate": 0.0002631471667346107, + "loss": 3.4039, + "step": 86760 + }, + { + "epoch": 5.8951623861937765, + "grad_norm": 0.947849452495575, + "learning_rate": 0.0002631047017257779, + "loss": 3.1984, + "step": 86765 + }, + { + "epoch": 5.895502106264438, + "grad_norm": 0.8114612102508545, + "learning_rate": 0.00026306223671694525, + "loss": 3.5397, + "step": 86770 + }, + { + "epoch": 5.8958418263351, + "grad_norm": 0.9328486323356628, + "learning_rate": 0.00026301977170811253, + "loss": 3.5195, + "step": 86775 + }, + { + "epoch": 5.896181546405762, + "grad_norm": 1.0835790634155273, + "learning_rate": 0.00026297730669927976, + "loss": 3.3179, + "step": 86780 + }, + { + "epoch": 5.896521266476423, + "grad_norm": 1.0040569305419922, + "learning_rate": 0.0002629348416904471, + "loss": 3.5962, + "step": 86785 + }, + { + "epoch": 5.896860986547085, + "grad_norm": 1.104378581047058, + "learning_rate": 0.00026289237668161437, + "loss": 3.464, + "step": 86790 + }, + { + "epoch": 5.897200706617747, + "grad_norm": 0.9844347834587097, + "learning_rate": 0.00026284991167278165, + "loss": 3.317, + "step": 86795 + }, + { + "epoch": 5.897540426688408, + "grad_norm": 0.8422542810440063, + "learning_rate": 0.0002628074466639489, + "loss": 3.3706, + "step": 86800 + }, + { + "epoch": 5.8978801467590705, + "grad_norm": 1.1742855310440063, + "learning_rate": 0.0002627649816551162, + "loss": 3.4955, + "step": 86805 + }, + { + "epoch": 5.8982198668297325, + "grad_norm": 0.8861384391784668, + "learning_rate": 0.0002627225166462835, + "loss": 3.4515, + "step": 86810 + }, + { + "epoch": 5.898559586900394, + "grad_norm": 0.8072331547737122, + "learning_rate": 0.0002626800516374507, + "loss": 3.3623, + "step": 86815 + }, + { + "epoch": 5.898899306971056, + "grad_norm": 0.8211284279823303, + "learning_rate": 0.00026263758662861805, + "loss": 3.5454, + "step": 86820 + }, + { + "epoch": 5.899239027041718, + "grad_norm": 0.7291365265846252, + "learning_rate": 0.00026259512161978533, + "loss": 3.505, + "step": 86825 + }, + { + "epoch": 5.899578747112379, + "grad_norm": 0.9404595494270325, + "learning_rate": 0.00026255265661095256, + "loss": 3.2433, + "step": 86830 + }, + { + "epoch": 5.899918467183041, + "grad_norm": 0.830746054649353, + "learning_rate": 0.00026251019160211984, + "loss": 3.3945, + "step": 86835 + }, + { + "epoch": 5.900258187253703, + "grad_norm": 0.9141514301300049, + "learning_rate": 0.00026246772659328717, + "loss": 3.4052, + "step": 86840 + }, + { + "epoch": 5.900597907324364, + "grad_norm": 1.0112676620483398, + "learning_rate": 0.0002624252615844544, + "loss": 3.6389, + "step": 86845 + }, + { + "epoch": 5.9009376273950265, + "grad_norm": 0.8237084150314331, + "learning_rate": 0.0002623827965756217, + "loss": 3.3747, + "step": 86850 + }, + { + "epoch": 5.9012773474656885, + "grad_norm": 1.1113238334655762, + "learning_rate": 0.000262340331566789, + "loss": 3.2057, + "step": 86855 + }, + { + "epoch": 5.90161706753635, + "grad_norm": 0.7811151146888733, + "learning_rate": 0.00026229786655795624, + "loss": 3.6354, + "step": 86860 + }, + { + "epoch": 5.901956787607012, + "grad_norm": 0.8714499473571777, + "learning_rate": 0.0002622554015491235, + "loss": 3.4251, + "step": 86865 + }, + { + "epoch": 5.902296507677674, + "grad_norm": 579.6832275390625, + "learning_rate": 0.00026221293654029085, + "loss": 3.6363, + "step": 86870 + }, + { + "epoch": 5.902636227748335, + "grad_norm": 0.8753916025161743, + "learning_rate": 0.0002621704715314581, + "loss": 3.4464, + "step": 86875 + }, + { + "epoch": 5.902975947818997, + "grad_norm": 0.8919311761856079, + "learning_rate": 0.00026212800652262536, + "loss": 3.4541, + "step": 86880 + }, + { + "epoch": 5.903315667889659, + "grad_norm": 0.9172821640968323, + "learning_rate": 0.00026208554151379264, + "loss": 3.4275, + "step": 86885 + }, + { + "epoch": 5.90365538796032, + "grad_norm": 0.878645122051239, + "learning_rate": 0.0002620430765049599, + "loss": 3.5749, + "step": 86890 + }, + { + "epoch": 5.9039951080309825, + "grad_norm": 0.9559105634689331, + "learning_rate": 0.0002620006114961272, + "loss": 3.4506, + "step": 86895 + }, + { + "epoch": 5.904334828101645, + "grad_norm": 0.9417399764060974, + "learning_rate": 0.0002619581464872945, + "loss": 3.454, + "step": 86900 + }, + { + "epoch": 5.904674548172306, + "grad_norm": 0.9088181257247925, + "learning_rate": 0.00026191568147846176, + "loss": 3.4495, + "step": 86905 + }, + { + "epoch": 5.905014268242968, + "grad_norm": 0.7918052077293396, + "learning_rate": 0.00026187321646962904, + "loss": 3.1733, + "step": 86910 + }, + { + "epoch": 5.90535398831363, + "grad_norm": 0.9783101677894592, + "learning_rate": 0.0002618307514607963, + "loss": 3.3448, + "step": 86915 + }, + { + "epoch": 5.905693708384291, + "grad_norm": 0.8716035485267639, + "learning_rate": 0.00026178828645196354, + "loss": 3.2641, + "step": 86920 + }, + { + "epoch": 5.906033428454953, + "grad_norm": 1.1220875978469849, + "learning_rate": 0.0002617458214431309, + "loss": 3.423, + "step": 86925 + }, + { + "epoch": 5.906373148525615, + "grad_norm": 0.9642802476882935, + "learning_rate": 0.00026170335643429816, + "loss": 3.3641, + "step": 86930 + }, + { + "epoch": 5.906712868596276, + "grad_norm": 0.7496830821037292, + "learning_rate": 0.0002616608914254654, + "loss": 3.1032, + "step": 86935 + }, + { + "epoch": 5.9070525886669385, + "grad_norm": 1.0065748691558838, + "learning_rate": 0.0002616184264166327, + "loss": 3.0922, + "step": 86940 + }, + { + "epoch": 5.907392308737601, + "grad_norm": 0.858755886554718, + "learning_rate": 0.0002615759614078, + "loss": 3.3356, + "step": 86945 + }, + { + "epoch": 5.907732028808262, + "grad_norm": 1.0335866212844849, + "learning_rate": 0.0002615334963989672, + "loss": 3.7257, + "step": 86950 + }, + { + "epoch": 5.908071748878924, + "grad_norm": 0.8041394948959351, + "learning_rate": 0.0002614910313901345, + "loss": 3.4732, + "step": 86955 + }, + { + "epoch": 5.908411468949586, + "grad_norm": 0.8124260306358337, + "learning_rate": 0.00026144856638130184, + "loss": 3.3198, + "step": 86960 + }, + { + "epoch": 5.908751189020247, + "grad_norm": 1.1435471773147583, + "learning_rate": 0.0002614061013724691, + "loss": 3.5829, + "step": 86965 + }, + { + "epoch": 5.909090909090909, + "grad_norm": 0.8497371673583984, + "learning_rate": 0.00026136363636363634, + "loss": 3.4526, + "step": 86970 + }, + { + "epoch": 5.909430629161571, + "grad_norm": 0.8618152737617493, + "learning_rate": 0.0002613211713548037, + "loss": 3.742, + "step": 86975 + }, + { + "epoch": 5.909770349232232, + "grad_norm": 0.9107972979545593, + "learning_rate": 0.00026127870634597096, + "loss": 3.4161, + "step": 86980 + }, + { + "epoch": 5.9101100693028945, + "grad_norm": 0.7411313056945801, + "learning_rate": 0.0002612362413371382, + "loss": 3.3177, + "step": 86985 + }, + { + "epoch": 5.910449789373557, + "grad_norm": 0.9005715250968933, + "learning_rate": 0.00026119377632830546, + "loss": 3.4154, + "step": 86990 + }, + { + "epoch": 5.910789509444218, + "grad_norm": 0.8922621011734009, + "learning_rate": 0.0002611513113194728, + "loss": 3.5609, + "step": 86995 + }, + { + "epoch": 5.91112922951488, + "grad_norm": 0.9250367879867554, + "learning_rate": 0.00026110884631064, + "loss": 3.4561, + "step": 87000 + }, + { + "epoch": 5.911468949585542, + "grad_norm": 0.8878572583198547, + "learning_rate": 0.0002610663813018073, + "loss": 3.5806, + "step": 87005 + }, + { + "epoch": 5.911808669656203, + "grad_norm": 1.0487000942230225, + "learning_rate": 0.00026102391629297464, + "loss": 3.2145, + "step": 87010 + }, + { + "epoch": 5.912148389726865, + "grad_norm": 0.8766903281211853, + "learning_rate": 0.00026098145128414187, + "loss": 3.354, + "step": 87015 + }, + { + "epoch": 5.912488109797527, + "grad_norm": 0.9046432971954346, + "learning_rate": 0.00026093898627530915, + "loss": 3.4835, + "step": 87020 + }, + { + "epoch": 5.9128278298681884, + "grad_norm": 0.9405030012130737, + "learning_rate": 0.0002608965212664764, + "loss": 3.6776, + "step": 87025 + }, + { + "epoch": 5.9131675499388505, + "grad_norm": 0.8741708397865295, + "learning_rate": 0.0002608540562576437, + "loss": 3.7468, + "step": 87030 + }, + { + "epoch": 5.913507270009513, + "grad_norm": 0.8123577237129211, + "learning_rate": 0.000260811591248811, + "loss": 3.587, + "step": 87035 + }, + { + "epoch": 5.913846990080174, + "grad_norm": 0.9039894342422485, + "learning_rate": 0.00026076912623997827, + "loss": 3.4001, + "step": 87040 + }, + { + "epoch": 5.914186710150836, + "grad_norm": 0.8513646721839905, + "learning_rate": 0.00026072666123114555, + "loss": 3.1548, + "step": 87045 + }, + { + "epoch": 5.914526430221498, + "grad_norm": 0.8191611766815186, + "learning_rate": 0.0002606841962223128, + "loss": 3.4523, + "step": 87050 + }, + { + "epoch": 5.914866150292159, + "grad_norm": 1.1811246871948242, + "learning_rate": 0.0002606417312134801, + "loss": 3.1899, + "step": 87055 + }, + { + "epoch": 5.915205870362821, + "grad_norm": 0.859306275844574, + "learning_rate": 0.00026059926620464733, + "loss": 3.3489, + "step": 87060 + }, + { + "epoch": 5.915545590433483, + "grad_norm": 0.8048444986343384, + "learning_rate": 0.00026055680119581467, + "loss": 3.4524, + "step": 87065 + }, + { + "epoch": 5.9158853105041445, + "grad_norm": 0.8702788949012756, + "learning_rate": 0.00026051433618698195, + "loss": 3.3381, + "step": 87070 + }, + { + "epoch": 5.9162250305748065, + "grad_norm": 0.7615591287612915, + "learning_rate": 0.00026047187117814917, + "loss": 3.5811, + "step": 87075 + }, + { + "epoch": 5.916564750645469, + "grad_norm": 0.9778746962547302, + "learning_rate": 0.0002604294061693165, + "loss": 3.5486, + "step": 87080 + }, + { + "epoch": 5.91690447071613, + "grad_norm": 1.114913821220398, + "learning_rate": 0.0002603869411604838, + "loss": 3.2842, + "step": 87085 + }, + { + "epoch": 5.917244190786792, + "grad_norm": 0.7593933343887329, + "learning_rate": 0.000260344476151651, + "loss": 3.7841, + "step": 87090 + }, + { + "epoch": 5.917583910857454, + "grad_norm": 0.946738600730896, + "learning_rate": 0.0002603020111428183, + "loss": 3.2942, + "step": 87095 + }, + { + "epoch": 5.917923630928115, + "grad_norm": 0.8523774147033691, + "learning_rate": 0.0002602595461339856, + "loss": 3.5062, + "step": 87100 + }, + { + "epoch": 5.918263350998777, + "grad_norm": 0.9060896635055542, + "learning_rate": 0.00026021708112515285, + "loss": 3.59, + "step": 87105 + }, + { + "epoch": 5.918603071069439, + "grad_norm": 0.8794929385185242, + "learning_rate": 0.00026017461611632013, + "loss": 3.3847, + "step": 87110 + }, + { + "epoch": 5.9189427911401005, + "grad_norm": 1.282698392868042, + "learning_rate": 0.00026013215110748747, + "loss": 3.5109, + "step": 87115 + }, + { + "epoch": 5.9192825112107625, + "grad_norm": 0.8819625377655029, + "learning_rate": 0.0002600896860986547, + "loss": 3.2781, + "step": 87120 + }, + { + "epoch": 5.919622231281424, + "grad_norm": 1.569460153579712, + "learning_rate": 0.00026004722108982197, + "loss": 3.5024, + "step": 87125 + }, + { + "epoch": 5.919961951352086, + "grad_norm": 0.9171451330184937, + "learning_rate": 0.00026000475608098925, + "loss": 3.1204, + "step": 87130 + }, + { + "epoch": 5.920301671422748, + "grad_norm": 1.0037691593170166, + "learning_rate": 0.0002599622910721566, + "loss": 3.5264, + "step": 87135 + }, + { + "epoch": 5.920641391493409, + "grad_norm": 0.8538153171539307, + "learning_rate": 0.0002599198260633238, + "loss": 3.4451, + "step": 87140 + }, + { + "epoch": 5.920981111564071, + "grad_norm": 1.0676628351211548, + "learning_rate": 0.0002598773610544911, + "loss": 3.5566, + "step": 87145 + }, + { + "epoch": 5.921320831634733, + "grad_norm": 0.8805878758430481, + "learning_rate": 0.00025983489604565843, + "loss": 3.4972, + "step": 87150 + }, + { + "epoch": 5.921660551705394, + "grad_norm": 0.9083634614944458, + "learning_rate": 0.00025979243103682565, + "loss": 3.387, + "step": 87155 + }, + { + "epoch": 5.9220002717760565, + "grad_norm": 0.9054036736488342, + "learning_rate": 0.00025974996602799293, + "loss": 3.3981, + "step": 87160 + }, + { + "epoch": 5.9223399918467186, + "grad_norm": 1.0472131967544556, + "learning_rate": 0.00025970750101916027, + "loss": 3.7891, + "step": 87165 + }, + { + "epoch": 5.92267971191738, + "grad_norm": 0.9835063219070435, + "learning_rate": 0.0002596650360103275, + "loss": 3.531, + "step": 87170 + }, + { + "epoch": 5.923019431988042, + "grad_norm": 0.8201888203620911, + "learning_rate": 0.0002596225710014948, + "loss": 3.3871, + "step": 87175 + }, + { + "epoch": 5.923359152058704, + "grad_norm": 0.8548210263252258, + "learning_rate": 0.00025958010599266205, + "loss": 3.5423, + "step": 87180 + }, + { + "epoch": 5.923698872129365, + "grad_norm": 0.7816788554191589, + "learning_rate": 0.00025953764098382933, + "loss": 3.5296, + "step": 87185 + }, + { + "epoch": 5.924038592200027, + "grad_norm": 0.8344810009002686, + "learning_rate": 0.0002594951759749966, + "loss": 3.4779, + "step": 87190 + }, + { + "epoch": 5.924378312270689, + "grad_norm": 0.8944981098175049, + "learning_rate": 0.0002594527109661639, + "loss": 3.6964, + "step": 87195 + }, + { + "epoch": 5.92471803234135, + "grad_norm": 0.6969605088233948, + "learning_rate": 0.0002594102459573312, + "loss": 3.7256, + "step": 87200 + }, + { + "epoch": 5.9250577524120125, + "grad_norm": 0.8397440910339355, + "learning_rate": 0.00025936778094849845, + "loss": 3.3472, + "step": 87205 + }, + { + "epoch": 5.925397472482675, + "grad_norm": 1.1327563524246216, + "learning_rate": 0.00025932531593966573, + "loss": 3.3953, + "step": 87210 + }, + { + "epoch": 5.925737192553336, + "grad_norm": 1.0135266780853271, + "learning_rate": 0.00025928285093083296, + "loss": 3.5325, + "step": 87215 + }, + { + "epoch": 5.926076912623998, + "grad_norm": 0.9305192828178406, + "learning_rate": 0.0002592403859220003, + "loss": 3.3799, + "step": 87220 + }, + { + "epoch": 5.92641663269466, + "grad_norm": 0.9032133221626282, + "learning_rate": 0.0002591979209131676, + "loss": 3.2738, + "step": 87225 + }, + { + "epoch": 5.926756352765321, + "grad_norm": 1.152830719947815, + "learning_rate": 0.0002591554559043348, + "loss": 3.4199, + "step": 87230 + }, + { + "epoch": 5.927096072835983, + "grad_norm": 0.7360384464263916, + "learning_rate": 0.00025911299089550213, + "loss": 3.6978, + "step": 87235 + }, + { + "epoch": 5.927435792906645, + "grad_norm": 0.77264004945755, + "learning_rate": 0.0002590705258866694, + "loss": 3.3866, + "step": 87240 + }, + { + "epoch": 5.927775512977306, + "grad_norm": 0.7775171399116516, + "learning_rate": 0.00025902806087783664, + "loss": 3.0679, + "step": 87245 + }, + { + "epoch": 5.9281152330479685, + "grad_norm": 0.827771008014679, + "learning_rate": 0.0002589855958690039, + "loss": 3.4941, + "step": 87250 + }, + { + "epoch": 5.928454953118631, + "grad_norm": 1.0844062566757202, + "learning_rate": 0.00025894313086017125, + "loss": 3.382, + "step": 87255 + }, + { + "epoch": 5.928794673189292, + "grad_norm": 1.0564496517181396, + "learning_rate": 0.0002589006658513385, + "loss": 3.5151, + "step": 87260 + }, + { + "epoch": 5.929134393259954, + "grad_norm": 1.0422168970108032, + "learning_rate": 0.00025885820084250576, + "loss": 3.2498, + "step": 87265 + }, + { + "epoch": 5.929474113330616, + "grad_norm": 0.9066182971000671, + "learning_rate": 0.0002588157358336731, + "loss": 3.6542, + "step": 87270 + }, + { + "epoch": 5.929813833401277, + "grad_norm": 0.9588635563850403, + "learning_rate": 0.0002587732708248403, + "loss": 3.3069, + "step": 87275 + }, + { + "epoch": 5.930153553471939, + "grad_norm": 0.8084071278572083, + "learning_rate": 0.0002587308058160076, + "loss": 3.3211, + "step": 87280 + }, + { + "epoch": 5.930493273542601, + "grad_norm": 0.8942273259162903, + "learning_rate": 0.0002586883408071749, + "loss": 3.4001, + "step": 87285 + }, + { + "epoch": 5.930832993613262, + "grad_norm": 0.8593006730079651, + "learning_rate": 0.00025864587579834216, + "loss": 3.4387, + "step": 87290 + }, + { + "epoch": 5.9311727136839245, + "grad_norm": 0.9233217835426331, + "learning_rate": 0.00025860341078950944, + "loss": 3.4869, + "step": 87295 + }, + { + "epoch": 5.931512433754587, + "grad_norm": 0.9641635417938232, + "learning_rate": 0.0002585609457806767, + "loss": 3.3964, + "step": 87300 + }, + { + "epoch": 5.931852153825248, + "grad_norm": 0.9919977188110352, + "learning_rate": 0.00025851848077184406, + "loss": 3.4627, + "step": 87305 + }, + { + "epoch": 5.93219187389591, + "grad_norm": 1.0040230751037598, + "learning_rate": 0.0002584760157630113, + "loss": 3.4442, + "step": 87310 + }, + { + "epoch": 5.932531593966572, + "grad_norm": 0.8807284235954285, + "learning_rate": 0.00025843355075417856, + "loss": 3.5739, + "step": 87315 + }, + { + "epoch": 5.932871314037233, + "grad_norm": 0.7402127385139465, + "learning_rate": 0.00025839108574534584, + "loss": 3.4237, + "step": 87320 + }, + { + "epoch": 5.933211034107895, + "grad_norm": 0.9314968585968018, + "learning_rate": 0.0002583486207365131, + "loss": 3.4594, + "step": 87325 + }, + { + "epoch": 5.933550754178556, + "grad_norm": 1.0152008533477783, + "learning_rate": 0.0002583061557276804, + "loss": 3.4363, + "step": 87330 + }, + { + "epoch": 5.9338904742492184, + "grad_norm": 1.0330173969268799, + "learning_rate": 0.0002582636907188477, + "loss": 3.5134, + "step": 87335 + }, + { + "epoch": 5.9342301943198805, + "grad_norm": 0.8797692656517029, + "learning_rate": 0.00025822122571001496, + "loss": 3.7714, + "step": 87340 + }, + { + "epoch": 5.934569914390542, + "grad_norm": 0.8708392977714539, + "learning_rate": 0.00025817876070118224, + "loss": 3.3061, + "step": 87345 + }, + { + "epoch": 5.934909634461204, + "grad_norm": 0.8700308799743652, + "learning_rate": 0.0002581362956923495, + "loss": 3.418, + "step": 87350 + }, + { + "epoch": 5.935249354531866, + "grad_norm": 0.8791851997375488, + "learning_rate": 0.00025809383068351675, + "loss": 3.4571, + "step": 87355 + }, + { + "epoch": 5.935589074602527, + "grad_norm": 1.0171502828598022, + "learning_rate": 0.0002580513656746841, + "loss": 3.478, + "step": 87360 + }, + { + "epoch": 5.935928794673189, + "grad_norm": 0.823566734790802, + "learning_rate": 0.00025800890066585136, + "loss": 3.3414, + "step": 87365 + }, + { + "epoch": 5.936268514743851, + "grad_norm": 0.7822336554527283, + "learning_rate": 0.0002579664356570186, + "loss": 3.4607, + "step": 87370 + }, + { + "epoch": 5.936608234814512, + "grad_norm": 0.9780349731445312, + "learning_rate": 0.0002579239706481859, + "loss": 3.176, + "step": 87375 + }, + { + "epoch": 5.9369479548851745, + "grad_norm": 0.8756744861602783, + "learning_rate": 0.0002578815056393532, + "loss": 3.4159, + "step": 87380 + }, + { + "epoch": 5.9372876749558365, + "grad_norm": 1.0665007829666138, + "learning_rate": 0.00025783904063052043, + "loss": 3.4949, + "step": 87385 + }, + { + "epoch": 5.937627395026498, + "grad_norm": 0.8300045728683472, + "learning_rate": 0.0002577965756216877, + "loss": 3.3136, + "step": 87390 + }, + { + "epoch": 5.93796711509716, + "grad_norm": 1.0675162076950073, + "learning_rate": 0.00025775411061285504, + "loss": 3.1168, + "step": 87395 + }, + { + "epoch": 5.938306835167822, + "grad_norm": 0.9599024057388306, + "learning_rate": 0.00025771164560402227, + "loss": 3.0581, + "step": 87400 + }, + { + "epoch": 5.938646555238483, + "grad_norm": 0.9952685832977295, + "learning_rate": 0.00025766918059518955, + "loss": 3.392, + "step": 87405 + }, + { + "epoch": 5.938986275309145, + "grad_norm": 1.0593780279159546, + "learning_rate": 0.0002576267155863569, + "loss": 3.5892, + "step": 87410 + }, + { + "epoch": 5.939325995379807, + "grad_norm": 0.8815662264823914, + "learning_rate": 0.0002575842505775241, + "loss": 3.2007, + "step": 87415 + }, + { + "epoch": 5.939665715450468, + "grad_norm": 0.7948782444000244, + "learning_rate": 0.0002575417855686914, + "loss": 3.3731, + "step": 87420 + }, + { + "epoch": 5.9400054355211305, + "grad_norm": 0.7677506804466248, + "learning_rate": 0.0002574993205598587, + "loss": 3.2166, + "step": 87425 + }, + { + "epoch": 5.9403451555917925, + "grad_norm": 0.6899620294570923, + "learning_rate": 0.00025745685555102595, + "loss": 3.2498, + "step": 87430 + }, + { + "epoch": 5.940684875662454, + "grad_norm": 0.8404735326766968, + "learning_rate": 0.00025741439054219323, + "loss": 3.5508, + "step": 87435 + }, + { + "epoch": 5.941024595733116, + "grad_norm": 0.8968085050582886, + "learning_rate": 0.0002573719255333605, + "loss": 3.3986, + "step": 87440 + }, + { + "epoch": 5.941364315803778, + "grad_norm": 0.851989209651947, + "learning_rate": 0.0002573294605245278, + "loss": 3.2774, + "step": 87445 + }, + { + "epoch": 5.941704035874439, + "grad_norm": 0.9279588460922241, + "learning_rate": 0.00025728699551569507, + "loss": 3.4749, + "step": 87450 + }, + { + "epoch": 5.942043755945101, + "grad_norm": 0.9167598485946655, + "learning_rate": 0.00025724453050686235, + "loss": 3.4838, + "step": 87455 + }, + { + "epoch": 5.942383476015763, + "grad_norm": 0.7974656224250793, + "learning_rate": 0.00025720206549802963, + "loss": 3.5046, + "step": 87460 + }, + { + "epoch": 5.942723196086424, + "grad_norm": 0.9780206084251404, + "learning_rate": 0.0002571596004891969, + "loss": 3.2816, + "step": 87465 + }, + { + "epoch": 5.9430629161570865, + "grad_norm": 0.7584128975868225, + "learning_rate": 0.0002571171354803642, + "loss": 3.3014, + "step": 87470 + }, + { + "epoch": 5.943402636227749, + "grad_norm": 0.9550994634628296, + "learning_rate": 0.00025707467047153147, + "loss": 3.5415, + "step": 87475 + }, + { + "epoch": 5.94374235629841, + "grad_norm": 0.9063881635665894, + "learning_rate": 0.00025703220546269875, + "loss": 3.5972, + "step": 87480 + }, + { + "epoch": 5.944082076369072, + "grad_norm": 0.7920548319816589, + "learning_rate": 0.00025698974045386603, + "loss": 3.5448, + "step": 87485 + }, + { + "epoch": 5.944421796439734, + "grad_norm": 0.857038140296936, + "learning_rate": 0.0002569472754450333, + "loss": 3.3626, + "step": 87490 + }, + { + "epoch": 5.944761516510395, + "grad_norm": 0.9470358490943909, + "learning_rate": 0.0002569048104362006, + "loss": 3.2626, + "step": 87495 + }, + { + "epoch": 5.945101236581057, + "grad_norm": 1.0175539255142212, + "learning_rate": 0.00025686234542736787, + "loss": 3.5909, + "step": 87500 + }, + { + "epoch": 5.945440956651719, + "grad_norm": 0.7661160230636597, + "learning_rate": 0.00025681988041853515, + "loss": 3.4494, + "step": 87505 + }, + { + "epoch": 5.94578067672238, + "grad_norm": 0.8242473602294922, + "learning_rate": 0.0002567774154097024, + "loss": 3.2919, + "step": 87510 + }, + { + "epoch": 5.9461203967930425, + "grad_norm": 0.7870598435401917, + "learning_rate": 0.0002567349504008697, + "loss": 3.5833, + "step": 87515 + }, + { + "epoch": 5.946460116863705, + "grad_norm": 1.0757640600204468, + "learning_rate": 0.000256692485392037, + "loss": 3.6027, + "step": 87520 + }, + { + "epoch": 5.946799836934366, + "grad_norm": 0.8482693433761597, + "learning_rate": 0.0002566500203832042, + "loss": 3.643, + "step": 87525 + }, + { + "epoch": 5.947139557005028, + "grad_norm": 1.0831378698349, + "learning_rate": 0.00025660755537437155, + "loss": 3.2219, + "step": 87530 + }, + { + "epoch": 5.94747927707569, + "grad_norm": 0.9210177659988403, + "learning_rate": 0.00025656509036553883, + "loss": 3.586, + "step": 87535 + }, + { + "epoch": 5.947818997146351, + "grad_norm": 0.8165308833122253, + "learning_rate": 0.00025652262535670606, + "loss": 3.4304, + "step": 87540 + }, + { + "epoch": 5.948158717217013, + "grad_norm": 1.0155463218688965, + "learning_rate": 0.00025648016034787334, + "loss": 3.3828, + "step": 87545 + }, + { + "epoch": 5.948498437287675, + "grad_norm": 0.9909318089485168, + "learning_rate": 0.00025643769533904067, + "loss": 3.247, + "step": 87550 + }, + { + "epoch": 5.948838157358336, + "grad_norm": 1.0564830303192139, + "learning_rate": 0.0002563952303302079, + "loss": 3.6905, + "step": 87555 + }, + { + "epoch": 5.9491778774289985, + "grad_norm": 0.7387513518333435, + "learning_rate": 0.0002563527653213752, + "loss": 3.1883, + "step": 87560 + }, + { + "epoch": 5.949517597499661, + "grad_norm": 0.8919906616210938, + "learning_rate": 0.0002563103003125425, + "loss": 3.3275, + "step": 87565 + }, + { + "epoch": 5.949857317570322, + "grad_norm": 1.1173657178878784, + "learning_rate": 0.00025626783530370974, + "loss": 3.493, + "step": 87570 + }, + { + "epoch": 5.950197037640984, + "grad_norm": 0.8686089515686035, + "learning_rate": 0.000256225370294877, + "loss": 3.2703, + "step": 87575 + }, + { + "epoch": 5.950536757711646, + "grad_norm": 1.40020751953125, + "learning_rate": 0.0002561829052860443, + "loss": 3.4626, + "step": 87580 + }, + { + "epoch": 5.950876477782307, + "grad_norm": 0.7526452541351318, + "learning_rate": 0.0002561404402772116, + "loss": 3.618, + "step": 87585 + }, + { + "epoch": 5.951216197852969, + "grad_norm": 1.0939830541610718, + "learning_rate": 0.00025609797526837886, + "loss": 3.3209, + "step": 87590 + }, + { + "epoch": 5.951555917923631, + "grad_norm": 1.2953850030899048, + "learning_rate": 0.00025605551025954614, + "loss": 3.5557, + "step": 87595 + }, + { + "epoch": 5.951895637994292, + "grad_norm": 0.8172399401664734, + "learning_rate": 0.0002560130452507134, + "loss": 3.4038, + "step": 87600 + }, + { + "epoch": 5.9522353580649545, + "grad_norm": 1.1040102243423462, + "learning_rate": 0.0002559705802418807, + "loss": 3.5745, + "step": 87605 + }, + { + "epoch": 5.952575078135617, + "grad_norm": 0.8998822569847107, + "learning_rate": 0.000255928115233048, + "loss": 3.6826, + "step": 87610 + }, + { + "epoch": 5.952914798206278, + "grad_norm": 0.8187810778617859, + "learning_rate": 0.0002558856502242152, + "loss": 3.6587, + "step": 87615 + }, + { + "epoch": 5.95325451827694, + "grad_norm": 0.9343876242637634, + "learning_rate": 0.00025584318521538254, + "loss": 3.3489, + "step": 87620 + }, + { + "epoch": 5.953594238347602, + "grad_norm": 1.072891354560852, + "learning_rate": 0.0002558007202065498, + "loss": 3.2811, + "step": 87625 + }, + { + "epoch": 5.953933958418263, + "grad_norm": 0.8833297491073608, + "learning_rate": 0.00025575825519771704, + "loss": 3.6804, + "step": 87630 + }, + { + "epoch": 5.954273678488925, + "grad_norm": 0.9450722932815552, + "learning_rate": 0.0002557157901888844, + "loss": 3.4501, + "step": 87635 + }, + { + "epoch": 5.954613398559587, + "grad_norm": 0.7401144504547119, + "learning_rate": 0.00025567332518005166, + "loss": 3.2962, + "step": 87640 + }, + { + "epoch": 5.9549531186302485, + "grad_norm": 0.8565064072608948, + "learning_rate": 0.00025563086017121894, + "loss": 3.4133, + "step": 87645 + }, + { + "epoch": 5.9552928387009105, + "grad_norm": 0.8214325308799744, + "learning_rate": 0.00025558839516238616, + "loss": 3.5129, + "step": 87650 + }, + { + "epoch": 5.955632558771573, + "grad_norm": 1.1307978630065918, + "learning_rate": 0.0002555459301535535, + "loss": 3.4319, + "step": 87655 + }, + { + "epoch": 5.955972278842234, + "grad_norm": 0.7849156260490417, + "learning_rate": 0.0002555034651447208, + "loss": 3.4021, + "step": 87660 + }, + { + "epoch": 5.956311998912896, + "grad_norm": 1.1903949975967407, + "learning_rate": 0.000255461000135888, + "loss": 3.5383, + "step": 87665 + }, + { + "epoch": 5.956651718983558, + "grad_norm": 0.953140139579773, + "learning_rate": 0.00025541853512705534, + "loss": 3.3946, + "step": 87670 + }, + { + "epoch": 5.956991439054219, + "grad_norm": 1.0677319765090942, + "learning_rate": 0.0002553760701182226, + "loss": 3.4506, + "step": 87675 + }, + { + "epoch": 5.957331159124881, + "grad_norm": 0.9141011238098145, + "learning_rate": 0.00025533360510938984, + "loss": 3.6847, + "step": 87680 + }, + { + "epoch": 5.957670879195543, + "grad_norm": 1.1151607036590576, + "learning_rate": 0.0002552911401005571, + "loss": 3.5215, + "step": 87685 + }, + { + "epoch": 5.9580105992662045, + "grad_norm": 0.7576051950454712, + "learning_rate": 0.00025524867509172446, + "loss": 3.3482, + "step": 87690 + }, + { + "epoch": 5.9583503193368665, + "grad_norm": 0.7513319849967957, + "learning_rate": 0.0002552062100828917, + "loss": 3.3793, + "step": 87695 + }, + { + "epoch": 5.958690039407529, + "grad_norm": 0.8699908256530762, + "learning_rate": 0.00025516374507405896, + "loss": 3.6958, + "step": 87700 + }, + { + "epoch": 5.95902975947819, + "grad_norm": 0.7480401992797852, + "learning_rate": 0.0002551212800652263, + "loss": 3.5752, + "step": 87705 + }, + { + "epoch": 5.959369479548852, + "grad_norm": 0.9115986227989197, + "learning_rate": 0.0002550788150563935, + "loss": 3.4835, + "step": 87710 + }, + { + "epoch": 5.959709199619514, + "grad_norm": 0.8467034101486206, + "learning_rate": 0.0002550363500475608, + "loss": 3.3964, + "step": 87715 + }, + { + "epoch": 5.960048919690175, + "grad_norm": 0.8195412755012512, + "learning_rate": 0.00025499388503872814, + "loss": 3.4874, + "step": 87720 + }, + { + "epoch": 5.960388639760837, + "grad_norm": 0.8298139572143555, + "learning_rate": 0.00025495142002989537, + "loss": 3.445, + "step": 87725 + }, + { + "epoch": 5.960728359831499, + "grad_norm": 0.8372587561607361, + "learning_rate": 0.00025490895502106265, + "loss": 3.7782, + "step": 87730 + }, + { + "epoch": 5.9610680799021605, + "grad_norm": 0.8053191304206848, + "learning_rate": 0.0002548664900122299, + "loss": 3.5352, + "step": 87735 + }, + { + "epoch": 5.9614077999728226, + "grad_norm": 0.9222981929779053, + "learning_rate": 0.0002548240250033972, + "loss": 3.3226, + "step": 87740 + }, + { + "epoch": 5.961747520043485, + "grad_norm": 1.3866087198257446, + "learning_rate": 0.0002547815599945645, + "loss": 3.1954, + "step": 87745 + }, + { + "epoch": 5.962087240114146, + "grad_norm": 0.9325369596481323, + "learning_rate": 0.00025473909498573177, + "loss": 3.8943, + "step": 87750 + }, + { + "epoch": 5.962426960184808, + "grad_norm": 0.7968777418136597, + "learning_rate": 0.00025469662997689905, + "loss": 3.4862, + "step": 87755 + }, + { + "epoch": 5.96276668025547, + "grad_norm": 0.9184139966964722, + "learning_rate": 0.0002546541649680663, + "loss": 3.546, + "step": 87760 + }, + { + "epoch": 5.963106400326131, + "grad_norm": 0.8017551898956299, + "learning_rate": 0.0002546116999592336, + "loss": 3.7562, + "step": 87765 + }, + { + "epoch": 5.963446120396793, + "grad_norm": 0.9906613230705261, + "learning_rate": 0.00025456923495040083, + "loss": 3.5998, + "step": 87770 + }, + { + "epoch": 5.963785840467455, + "grad_norm": 0.9008069634437561, + "learning_rate": 0.00025452676994156817, + "loss": 3.3407, + "step": 87775 + }, + { + "epoch": 5.9641255605381165, + "grad_norm": 0.7324938774108887, + "learning_rate": 0.00025448430493273545, + "loss": 3.4865, + "step": 87780 + }, + { + "epoch": 5.964465280608779, + "grad_norm": 0.8969598412513733, + "learning_rate": 0.00025444183992390267, + "loss": 3.6578, + "step": 87785 + }, + { + "epoch": 5.964805000679441, + "grad_norm": 0.9423465728759766, + "learning_rate": 0.00025439937491507, + "loss": 3.547, + "step": 87790 + }, + { + "epoch": 5.965144720750102, + "grad_norm": 0.7941276431083679, + "learning_rate": 0.0002543569099062373, + "loss": 3.099, + "step": 87795 + }, + { + "epoch": 5.965484440820764, + "grad_norm": 0.9506189823150635, + "learning_rate": 0.0002543144448974045, + "loss": 3.546, + "step": 87800 + }, + { + "epoch": 5.965824160891425, + "grad_norm": 1.1353613138198853, + "learning_rate": 0.0002542719798885718, + "loss": 3.2872, + "step": 87805 + }, + { + "epoch": 5.966163880962087, + "grad_norm": 0.8470731377601624, + "learning_rate": 0.0002542295148797391, + "loss": 3.6516, + "step": 87810 + }, + { + "epoch": 5.966503601032749, + "grad_norm": 0.961182713508606, + "learning_rate": 0.0002541870498709064, + "loss": 3.7925, + "step": 87815 + }, + { + "epoch": 5.96684332110341, + "grad_norm": 0.8845279216766357, + "learning_rate": 0.00025414458486207363, + "loss": 3.3264, + "step": 87820 + }, + { + "epoch": 5.9671830411740725, + "grad_norm": 0.8230481743812561, + "learning_rate": 0.00025410211985324097, + "loss": 3.481, + "step": 87825 + }, + { + "epoch": 5.967522761244735, + "grad_norm": 0.8327860832214355, + "learning_rate": 0.00025405965484440825, + "loss": 3.6837, + "step": 87830 + }, + { + "epoch": 5.967862481315396, + "grad_norm": 0.9442246556282043, + "learning_rate": 0.00025401718983557547, + "loss": 3.491, + "step": 87835 + }, + { + "epoch": 5.968202201386058, + "grad_norm": 0.9645861983299255, + "learning_rate": 0.00025397472482674275, + "loss": 3.438, + "step": 87840 + }, + { + "epoch": 5.96854192145672, + "grad_norm": 0.7841061353683472, + "learning_rate": 0.0002539322598179101, + "loss": 3.5916, + "step": 87845 + }, + { + "epoch": 5.968881641527381, + "grad_norm": 0.8801999092102051, + "learning_rate": 0.0002538897948090773, + "loss": 3.3937, + "step": 87850 + }, + { + "epoch": 5.969221361598043, + "grad_norm": 0.8848710060119629, + "learning_rate": 0.0002538473298002446, + "loss": 3.5727, + "step": 87855 + }, + { + "epoch": 5.969561081668705, + "grad_norm": 0.9078718423843384, + "learning_rate": 0.00025380486479141193, + "loss": 3.4323, + "step": 87860 + }, + { + "epoch": 5.969900801739366, + "grad_norm": 0.8893359899520874, + "learning_rate": 0.00025376239978257915, + "loss": 3.5389, + "step": 87865 + }, + { + "epoch": 5.9702405218100285, + "grad_norm": 0.8817317485809326, + "learning_rate": 0.00025371993477374643, + "loss": 3.4542, + "step": 87870 + }, + { + "epoch": 5.970580241880691, + "grad_norm": 0.9319194555282593, + "learning_rate": 0.0002536774697649137, + "loss": 3.3732, + "step": 87875 + }, + { + "epoch": 5.970919961951352, + "grad_norm": 0.9150902628898621, + "learning_rate": 0.000253635004756081, + "loss": 3.366, + "step": 87880 + }, + { + "epoch": 5.971259682022014, + "grad_norm": 0.7564211487770081, + "learning_rate": 0.0002535925397472483, + "loss": 3.6287, + "step": 87885 + }, + { + "epoch": 5.971599402092676, + "grad_norm": 1.2268205881118774, + "learning_rate": 0.00025355007473841555, + "loss": 3.1855, + "step": 87890 + }, + { + "epoch": 5.971939122163337, + "grad_norm": 0.902304470539093, + "learning_rate": 0.00025350760972958283, + "loss": 3.554, + "step": 87895 + }, + { + "epoch": 5.972278842233999, + "grad_norm": 0.9127666354179382, + "learning_rate": 0.0002534651447207501, + "loss": 3.2878, + "step": 87900 + }, + { + "epoch": 5.972618562304661, + "grad_norm": 0.7556972503662109, + "learning_rate": 0.0002534226797119174, + "loss": 3.3472, + "step": 87905 + }, + { + "epoch": 5.9729582823753224, + "grad_norm": 0.9941595792770386, + "learning_rate": 0.0002533802147030846, + "loss": 3.5895, + "step": 87910 + }, + { + "epoch": 5.9732980024459845, + "grad_norm": 0.7425792813301086, + "learning_rate": 0.00025333774969425195, + "loss": 3.4895, + "step": 87915 + }, + { + "epoch": 5.973637722516647, + "grad_norm": 1.030763030052185, + "learning_rate": 0.00025329528468541923, + "loss": 3.4746, + "step": 87920 + }, + { + "epoch": 5.973977442587308, + "grad_norm": 0.9264202117919922, + "learning_rate": 0.00025325281967658646, + "loss": 3.4153, + "step": 87925 + }, + { + "epoch": 5.97431716265797, + "grad_norm": 0.8829048871994019, + "learning_rate": 0.0002532103546677538, + "loss": 3.5429, + "step": 87930 + }, + { + "epoch": 5.974656882728632, + "grad_norm": 0.8642759323120117, + "learning_rate": 0.0002531678896589211, + "loss": 3.5575, + "step": 87935 + }, + { + "epoch": 5.974996602799293, + "grad_norm": 0.7715226411819458, + "learning_rate": 0.0002531254246500883, + "loss": 3.2366, + "step": 87940 + }, + { + "epoch": 5.975336322869955, + "grad_norm": 0.7336623072624207, + "learning_rate": 0.0002530829596412556, + "loss": 3.5344, + "step": 87945 + }, + { + "epoch": 5.975676042940617, + "grad_norm": 0.7674343585968018, + "learning_rate": 0.0002530404946324229, + "loss": 3.2711, + "step": 87950 + }, + { + "epoch": 5.9760157630112785, + "grad_norm": 0.8075256943702698, + "learning_rate": 0.00025299802962359014, + "loss": 3.2749, + "step": 87955 + }, + { + "epoch": 5.9763554830819405, + "grad_norm": 0.9380080699920654, + "learning_rate": 0.0002529555646147574, + "loss": 3.6459, + "step": 87960 + }, + { + "epoch": 5.976695203152603, + "grad_norm": 0.8172228932380676, + "learning_rate": 0.00025291309960592475, + "loss": 3.3484, + "step": 87965 + }, + { + "epoch": 5.977034923223264, + "grad_norm": 0.8264505863189697, + "learning_rate": 0.000252870634597092, + "loss": 3.5681, + "step": 87970 + }, + { + "epoch": 5.977374643293926, + "grad_norm": 0.8286378383636475, + "learning_rate": 0.00025282816958825926, + "loss": 3.7217, + "step": 87975 + }, + { + "epoch": 5.977714363364588, + "grad_norm": 0.866304337978363, + "learning_rate": 0.00025278570457942654, + "loss": 3.4658, + "step": 87980 + }, + { + "epoch": 5.978054083435249, + "grad_norm": 0.6695724129676819, + "learning_rate": 0.0002527432395705939, + "loss": 3.6369, + "step": 87985 + }, + { + "epoch": 5.978393803505911, + "grad_norm": 0.8788052797317505, + "learning_rate": 0.0002527007745617611, + "loss": 3.2139, + "step": 87990 + }, + { + "epoch": 5.978733523576573, + "grad_norm": 0.9803169965744019, + "learning_rate": 0.0002526583095529284, + "loss": 3.6267, + "step": 87995 + }, + { + "epoch": 5.9790732436472345, + "grad_norm": 1.0306413173675537, + "learning_rate": 0.0002526158445440957, + "loss": 3.4312, + "step": 88000 + }, + { + "epoch": 5.9794129637178965, + "grad_norm": 0.9144333004951477, + "learning_rate": 0.00025257337953526294, + "loss": 3.3391, + "step": 88005 + }, + { + "epoch": 5.979752683788558, + "grad_norm": 0.9094138145446777, + "learning_rate": 0.0002525309145264302, + "loss": 3.2381, + "step": 88010 + }, + { + "epoch": 5.98009240385922, + "grad_norm": 1.0916175842285156, + "learning_rate": 0.00025248844951759756, + "loss": 3.4422, + "step": 88015 + }, + { + "epoch": 5.980432123929882, + "grad_norm": 0.7795401215553284, + "learning_rate": 0.0002524459845087648, + "loss": 3.5418, + "step": 88020 + }, + { + "epoch": 5.980771844000543, + "grad_norm": 0.8213106989860535, + "learning_rate": 0.00025240351949993206, + "loss": 3.3352, + "step": 88025 + }, + { + "epoch": 5.981111564071205, + "grad_norm": 0.8686926364898682, + "learning_rate": 0.00025236105449109934, + "loss": 3.5608, + "step": 88030 + }, + { + "epoch": 5.981451284141867, + "grad_norm": 0.9499101042747498, + "learning_rate": 0.0002523185894822666, + "loss": 3.6518, + "step": 88035 + }, + { + "epoch": 5.981791004212528, + "grad_norm": 0.9914563894271851, + "learning_rate": 0.0002522761244734339, + "loss": 3.5408, + "step": 88040 + }, + { + "epoch": 5.9821307242831905, + "grad_norm": 1.2207297086715698, + "learning_rate": 0.0002522336594646012, + "loss": 3.4204, + "step": 88045 + }, + { + "epoch": 5.982470444353853, + "grad_norm": 0.8428399562835693, + "learning_rate": 0.00025219119445576846, + "loss": 3.4549, + "step": 88050 + }, + { + "epoch": 5.982810164424514, + "grad_norm": 0.854900062084198, + "learning_rate": 0.00025214872944693574, + "loss": 3.3504, + "step": 88055 + }, + { + "epoch": 5.983149884495176, + "grad_norm": 0.7577596306800842, + "learning_rate": 0.000252106264438103, + "loss": 3.5553, + "step": 88060 + }, + { + "epoch": 5.983489604565838, + "grad_norm": 0.8783277273178101, + "learning_rate": 0.00025206379942927025, + "loss": 3.5637, + "step": 88065 + }, + { + "epoch": 5.983829324636499, + "grad_norm": 0.8838988542556763, + "learning_rate": 0.0002520213344204376, + "loss": 3.6819, + "step": 88070 + }, + { + "epoch": 5.984169044707161, + "grad_norm": 0.839244544506073, + "learning_rate": 0.00025197886941160486, + "loss": 3.2292, + "step": 88075 + }, + { + "epoch": 5.984508764777823, + "grad_norm": 1.4854435920715332, + "learning_rate": 0.0002519364044027721, + "loss": 3.6048, + "step": 88080 + }, + { + "epoch": 5.984848484848484, + "grad_norm": 1.1846059560775757, + "learning_rate": 0.0002518939393939394, + "loss": 3.1716, + "step": 88085 + }, + { + "epoch": 5.9851882049191465, + "grad_norm": 0.8549778461456299, + "learning_rate": 0.0002518514743851067, + "loss": 3.5488, + "step": 88090 + }, + { + "epoch": 5.985527924989809, + "grad_norm": 0.9296994209289551, + "learning_rate": 0.00025180900937627393, + "loss": 3.183, + "step": 88095 + }, + { + "epoch": 5.98586764506047, + "grad_norm": 0.9364610910415649, + "learning_rate": 0.0002517665443674412, + "loss": 3.5484, + "step": 88100 + }, + { + "epoch": 5.986207365131132, + "grad_norm": 0.7126706838607788, + "learning_rate": 0.00025172407935860854, + "loss": 3.3379, + "step": 88105 + }, + { + "epoch": 5.986547085201794, + "grad_norm": 0.9163709282875061, + "learning_rate": 0.00025168161434977577, + "loss": 3.6132, + "step": 88110 + }, + { + "epoch": 5.986886805272455, + "grad_norm": 0.7285540103912354, + "learning_rate": 0.00025163914934094305, + "loss": 3.5996, + "step": 88115 + }, + { + "epoch": 5.987226525343117, + "grad_norm": 0.7629046440124512, + "learning_rate": 0.0002515966843321104, + "loss": 3.3475, + "step": 88120 + }, + { + "epoch": 5.987566245413779, + "grad_norm": 0.9181415438652039, + "learning_rate": 0.0002515542193232776, + "loss": 3.3149, + "step": 88125 + }, + { + "epoch": 5.98790596548444, + "grad_norm": 0.9522302746772766, + "learning_rate": 0.0002515117543144449, + "loss": 3.1457, + "step": 88130 + }, + { + "epoch": 5.9882456855551025, + "grad_norm": 0.9400190114974976, + "learning_rate": 0.00025146928930561217, + "loss": 3.2017, + "step": 88135 + }, + { + "epoch": 5.988585405625765, + "grad_norm": 1.0296279191970825, + "learning_rate": 0.00025142682429677945, + "loss": 3.4343, + "step": 88140 + }, + { + "epoch": 5.988925125696426, + "grad_norm": 1.015599250793457, + "learning_rate": 0.00025138435928794673, + "loss": 3.3007, + "step": 88145 + }, + { + "epoch": 5.989264845767088, + "grad_norm": 0.9581639170646667, + "learning_rate": 0.000251341894279114, + "loss": 3.3513, + "step": 88150 + }, + { + "epoch": 5.98960456583775, + "grad_norm": 0.8747519850730896, + "learning_rate": 0.00025129942927028134, + "loss": 3.4879, + "step": 88155 + }, + { + "epoch": 5.989944285908411, + "grad_norm": 0.880652129650116, + "learning_rate": 0.00025125696426144857, + "loss": 3.3327, + "step": 88160 + }, + { + "epoch": 5.990284005979073, + "grad_norm": 0.9219487309455872, + "learning_rate": 0.00025121449925261585, + "loss": 3.2546, + "step": 88165 + }, + { + "epoch": 5.990623726049735, + "grad_norm": 1.0514140129089355, + "learning_rate": 0.00025117203424378313, + "loss": 3.7368, + "step": 88170 + }, + { + "epoch": 5.990963446120396, + "grad_norm": 1.0886567831039429, + "learning_rate": 0.0002511295692349504, + "loss": 3.3659, + "step": 88175 + }, + { + "epoch": 5.9913031661910585, + "grad_norm": 1.203183650970459, + "learning_rate": 0.0002510871042261177, + "loss": 3.6497, + "step": 88180 + }, + { + "epoch": 5.991642886261721, + "grad_norm": 0.9404119849205017, + "learning_rate": 0.00025104463921728497, + "loss": 3.486, + "step": 88185 + }, + { + "epoch": 5.991982606332382, + "grad_norm": 0.90493243932724, + "learning_rate": 0.00025100217420845225, + "loss": 3.3324, + "step": 88190 + }, + { + "epoch": 5.992322326403044, + "grad_norm": 1.397676706314087, + "learning_rate": 0.00025095970919961953, + "loss": 3.1644, + "step": 88195 + }, + { + "epoch": 5.992662046473706, + "grad_norm": 0.861950695514679, + "learning_rate": 0.0002509172441907868, + "loss": 3.5544, + "step": 88200 + }, + { + "epoch": 5.993001766544367, + "grad_norm": 0.8270180821418762, + "learning_rate": 0.00025087477918195404, + "loss": 3.3112, + "step": 88205 + }, + { + "epoch": 5.993341486615029, + "grad_norm": 0.8583129048347473, + "learning_rate": 0.00025083231417312137, + "loss": 3.4928, + "step": 88210 + }, + { + "epoch": 5.993681206685691, + "grad_norm": 0.8138909935951233, + "learning_rate": 0.00025078984916428865, + "loss": 3.2606, + "step": 88215 + }, + { + "epoch": 5.9940209267563525, + "grad_norm": 0.8053900599479675, + "learning_rate": 0.0002507473841554559, + "loss": 3.3827, + "step": 88220 + }, + { + "epoch": 5.9943606468270145, + "grad_norm": 0.8470441102981567, + "learning_rate": 0.0002507049191466232, + "loss": 3.5933, + "step": 88225 + }, + { + "epoch": 5.994700366897677, + "grad_norm": 1.1569738388061523, + "learning_rate": 0.0002506624541377905, + "loss": 3.6711, + "step": 88230 + }, + { + "epoch": 5.995040086968338, + "grad_norm": 1.0644514560699463, + "learning_rate": 0.0002506199891289577, + "loss": 3.6686, + "step": 88235 + }, + { + "epoch": 5.995379807039, + "grad_norm": 0.8666078448295593, + "learning_rate": 0.000250577524120125, + "loss": 3.5863, + "step": 88240 + }, + { + "epoch": 5.995719527109662, + "grad_norm": 0.9564222693443298, + "learning_rate": 0.00025053505911129233, + "loss": 3.4059, + "step": 88245 + }, + { + "epoch": 5.996059247180323, + "grad_norm": 1.125313639640808, + "learning_rate": 0.00025049259410245956, + "loss": 3.6172, + "step": 88250 + }, + { + "epoch": 5.996398967250985, + "grad_norm": 1.02402663230896, + "learning_rate": 0.00025045012909362684, + "loss": 3.2075, + "step": 88255 + }, + { + "epoch": 5.996738687321647, + "grad_norm": 0.8217767477035522, + "learning_rate": 0.00025040766408479417, + "loss": 3.4003, + "step": 88260 + }, + { + "epoch": 5.9970784073923085, + "grad_norm": 1.0333209037780762, + "learning_rate": 0.0002503651990759614, + "loss": 3.5562, + "step": 88265 + }, + { + "epoch": 5.9974181274629705, + "grad_norm": 1.0327162742614746, + "learning_rate": 0.0002503227340671287, + "loss": 3.3522, + "step": 88270 + }, + { + "epoch": 5.997757847533633, + "grad_norm": 0.7804272174835205, + "learning_rate": 0.000250280269058296, + "loss": 3.3501, + "step": 88275 + }, + { + "epoch": 5.998097567604294, + "grad_norm": 0.8569813966751099, + "learning_rate": 0.00025023780404946324, + "loss": 3.5995, + "step": 88280 + }, + { + "epoch": 5.998437287674956, + "grad_norm": 0.8795720934867859, + "learning_rate": 0.0002501953390406305, + "loss": 3.5056, + "step": 88285 + }, + { + "epoch": 5.998777007745618, + "grad_norm": 0.8322656750679016, + "learning_rate": 0.0002501528740317978, + "loss": 3.5322, + "step": 88290 + }, + { + "epoch": 5.999116727816279, + "grad_norm": 0.8637499213218689, + "learning_rate": 0.0002501104090229651, + "loss": 3.3911, + "step": 88295 + }, + { + "epoch": 5.999456447886941, + "grad_norm": 1.3256415128707886, + "learning_rate": 0.00025006794401413236, + "loss": 3.5186, + "step": 88300 + }, + { + "epoch": 5.999796167957603, + "grad_norm": 1.140984058380127, + "learning_rate": 0.00025002547900529964, + "loss": 3.556, + "step": 88305 + }, + { + "epoch": 6.0, + "eval_bertscore": { + "f1": 0.8412353278812206, + "precision": 0.8421136805375586, + "recall": 0.8411571180327178 + }, + "eval_bleu_4": 0.021544713242252534, + "eval_exact_match": 0.00038763446070355656, + "eval_loss": 3.4019320011138916, + "eval_meteor": 0.09107590758767604, + "eval_rouge": { + "rouge1": 0.1306312601958517, + "rouge2": 0.021282046414223994, + "rougeL": 0.11342475627469248, + "rougeLsum": 0.11345516014889859 + }, + "eval_runtime": 2390.6995, + "eval_samples_per_second": 4.316, + "eval_steps_per_second": 0.54, + "step": 88308 + }, + { + "epoch": 6.0001358880282645, + "grad_norm": 0.796630322933197, + "learning_rate": 0.0002499830139964669, + "loss": 3.3529, + "step": 88310 + }, + { + "epoch": 6.0004756080989265, + "grad_norm": 0.8437119126319885, + "learning_rate": 0.0002499405489876342, + "loss": 3.3227, + "step": 88315 + }, + { + "epoch": 6.000815328169589, + "grad_norm": 0.7918819785118103, + "learning_rate": 0.0002498980839788015, + "loss": 3.3793, + "step": 88320 + }, + { + "epoch": 6.00115504824025, + "grad_norm": 0.9082744121551514, + "learning_rate": 0.00024985561896996876, + "loss": 3.2301, + "step": 88325 + }, + { + "epoch": 6.001494768310912, + "grad_norm": 0.8444698452949524, + "learning_rate": 0.00024981315396113604, + "loss": 3.3678, + "step": 88330 + }, + { + "epoch": 6.001834488381574, + "grad_norm": 0.7588802576065063, + "learning_rate": 0.0002497706889523033, + "loss": 3.4236, + "step": 88335 + }, + { + "epoch": 6.002174208452235, + "grad_norm": 1.0127778053283691, + "learning_rate": 0.0002497282239434706, + "loss": 3.4373, + "step": 88340 + }, + { + "epoch": 6.002513928522897, + "grad_norm": 0.8195062279701233, + "learning_rate": 0.0002496857589346379, + "loss": 3.526, + "step": 88345 + }, + { + "epoch": 6.002853648593559, + "grad_norm": 0.9148167371749878, + "learning_rate": 0.00024964329392580516, + "loss": 3.4321, + "step": 88350 + }, + { + "epoch": 6.0031933686642205, + "grad_norm": 1.148107647895813, + "learning_rate": 0.00024960082891697244, + "loss": 3.5192, + "step": 88355 + }, + { + "epoch": 6.003533088734883, + "grad_norm": 0.8534746170043945, + "learning_rate": 0.00024955836390813966, + "loss": 3.2948, + "step": 88360 + }, + { + "epoch": 6.003872808805545, + "grad_norm": 0.8700295686721802, + "learning_rate": 0.000249515898899307, + "loss": 3.1201, + "step": 88365 + }, + { + "epoch": 6.004212528876206, + "grad_norm": 0.9279048442840576, + "learning_rate": 0.0002494734338904742, + "loss": 2.9832, + "step": 88370 + }, + { + "epoch": 6.004552248946868, + "grad_norm": 0.787178635597229, + "learning_rate": 0.00024943096888164156, + "loss": 3.4281, + "step": 88375 + }, + { + "epoch": 6.00489196901753, + "grad_norm": 0.9760860204696655, + "learning_rate": 0.00024938850387280884, + "loss": 3.3565, + "step": 88380 + }, + { + "epoch": 6.005231689088191, + "grad_norm": 0.9710628986358643, + "learning_rate": 0.00024934603886397606, + "loss": 3.5832, + "step": 88385 + }, + { + "epoch": 6.005571409158853, + "grad_norm": 0.8416259288787842, + "learning_rate": 0.0002493035738551434, + "loss": 3.4266, + "step": 88390 + }, + { + "epoch": 6.005911129229515, + "grad_norm": 1.0236973762512207, + "learning_rate": 0.0002492611088463106, + "loss": 3.5295, + "step": 88395 + }, + { + "epoch": 6.0062508493001765, + "grad_norm": 0.9624733924865723, + "learning_rate": 0.0002492186438374779, + "loss": 3.3926, + "step": 88400 + }, + { + "epoch": 6.006590569370839, + "grad_norm": 1.0399904251098633, + "learning_rate": 0.0002491761788286452, + "loss": 3.5875, + "step": 88405 + }, + { + "epoch": 6.006930289441501, + "grad_norm": 1.0013476610183716, + "learning_rate": 0.00024913371381981246, + "loss": 3.3269, + "step": 88410 + }, + { + "epoch": 6.007270009512162, + "grad_norm": 0.9076151847839355, + "learning_rate": 0.00024909124881097974, + "loss": 3.3152, + "step": 88415 + }, + { + "epoch": 6.007609729582824, + "grad_norm": 0.8646873831748962, + "learning_rate": 0.000249048783802147, + "loss": 3.4431, + "step": 88420 + }, + { + "epoch": 6.007949449653485, + "grad_norm": 1.0981229543685913, + "learning_rate": 0.0002490063187933143, + "loss": 3.4269, + "step": 88425 + }, + { + "epoch": 6.008289169724147, + "grad_norm": 1.1867728233337402, + "learning_rate": 0.0002489638537844816, + "loss": 3.5109, + "step": 88430 + }, + { + "epoch": 6.008628889794809, + "grad_norm": 0.9142845273017883, + "learning_rate": 0.00024892138877564886, + "loss": 3.2442, + "step": 88435 + }, + { + "epoch": 6.00896860986547, + "grad_norm": 1.1163314580917358, + "learning_rate": 0.00024887892376681615, + "loss": 3.2964, + "step": 88440 + }, + { + "epoch": 6.0093083299361325, + "grad_norm": 1.0559029579162598, + "learning_rate": 0.0002488364587579834, + "loss": 3.5992, + "step": 88445 + }, + { + "epoch": 6.009648050006795, + "grad_norm": 0.8932525515556335, + "learning_rate": 0.0002487939937491507, + "loss": 3.2314, + "step": 88450 + }, + { + "epoch": 6.009987770077456, + "grad_norm": 0.9444791078567505, + "learning_rate": 0.000248751528740318, + "loss": 3.4098, + "step": 88455 + }, + { + "epoch": 6.010327490148118, + "grad_norm": 0.9295849204063416, + "learning_rate": 0.00024870906373148527, + "loss": 3.1616, + "step": 88460 + }, + { + "epoch": 6.01066721021878, + "grad_norm": 0.7798036336898804, + "learning_rate": 0.00024866659872265255, + "loss": 3.4547, + "step": 88465 + }, + { + "epoch": 6.011006930289441, + "grad_norm": 0.9205248951911926, + "learning_rate": 0.0002486241337138198, + "loss": 3.5647, + "step": 88470 + }, + { + "epoch": 6.011346650360103, + "grad_norm": 0.727745771408081, + "learning_rate": 0.0002485816687049871, + "loss": 3.6683, + "step": 88475 + }, + { + "epoch": 6.011686370430765, + "grad_norm": 1.0261657238006592, + "learning_rate": 0.0002485392036961544, + "loss": 3.2732, + "step": 88480 + }, + { + "epoch": 6.0120260905014264, + "grad_norm": 1.2219383716583252, + "learning_rate": 0.00024849673868732167, + "loss": 3.3516, + "step": 88485 + }, + { + "epoch": 6.0123658105720885, + "grad_norm": 1.0045169591903687, + "learning_rate": 0.00024845427367848895, + "loss": 3.2921, + "step": 88490 + }, + { + "epoch": 6.012705530642751, + "grad_norm": 0.918348491191864, + "learning_rate": 0.0002484118086696562, + "loss": 3.5195, + "step": 88495 + }, + { + "epoch": 6.013045250713412, + "grad_norm": 0.8692794442176819, + "learning_rate": 0.00024836934366082345, + "loss": 3.3956, + "step": 88500 + }, + { + "epoch": 6.013384970784074, + "grad_norm": 1.0888947248458862, + "learning_rate": 0.0002483268786519908, + "loss": 3.4296, + "step": 88505 + }, + { + "epoch": 6.013724690854736, + "grad_norm": 1.0747720003128052, + "learning_rate": 0.00024828441364315807, + "loss": 3.4165, + "step": 88510 + }, + { + "epoch": 6.014064410925397, + "grad_norm": 1.1175966262817383, + "learning_rate": 0.0002482419486343253, + "loss": 3.3456, + "step": 88515 + }, + { + "epoch": 6.014404130996059, + "grad_norm": 0.9356057643890381, + "learning_rate": 0.0002481994836254926, + "loss": 3.4295, + "step": 88520 + }, + { + "epoch": 6.014743851066721, + "grad_norm": 1.0473543405532837, + "learning_rate": 0.00024815701861665985, + "loss": 3.4161, + "step": 88525 + }, + { + "epoch": 6.0150835711373825, + "grad_norm": 0.9900091886520386, + "learning_rate": 0.00024811455360782713, + "loss": 3.4136, + "step": 88530 + }, + { + "epoch": 6.0154232912080445, + "grad_norm": 0.9443422555923462, + "learning_rate": 0.0002480720885989944, + "loss": 3.4391, + "step": 88535 + }, + { + "epoch": 6.015763011278707, + "grad_norm": 0.8819485306739807, + "learning_rate": 0.0002480296235901617, + "loss": 3.2144, + "step": 88540 + }, + { + "epoch": 6.016102731349368, + "grad_norm": 0.8331897258758545, + "learning_rate": 0.000247987158581329, + "loss": 3.4407, + "step": 88545 + }, + { + "epoch": 6.01644245142003, + "grad_norm": 0.9211376309394836, + "learning_rate": 0.00024794469357249625, + "loss": 3.5111, + "step": 88550 + }, + { + "epoch": 6.016782171490692, + "grad_norm": 1.1135050058364868, + "learning_rate": 0.00024790222856366353, + "loss": 3.2559, + "step": 88555 + }, + { + "epoch": 6.017121891561353, + "grad_norm": 0.9495705366134644, + "learning_rate": 0.0002478597635548308, + "loss": 3.0686, + "step": 88560 + }, + { + "epoch": 6.017461611632015, + "grad_norm": 0.9664967656135559, + "learning_rate": 0.0002478172985459981, + "loss": 3.4148, + "step": 88565 + }, + { + "epoch": 6.017801331702677, + "grad_norm": 0.9518519043922424, + "learning_rate": 0.00024777483353716537, + "loss": 3.4547, + "step": 88570 + }, + { + "epoch": 6.0181410517733385, + "grad_norm": 0.7431796789169312, + "learning_rate": 0.00024773236852833265, + "loss": 3.438, + "step": 88575 + }, + { + "epoch": 6.0184807718440005, + "grad_norm": 0.8046507239341736, + "learning_rate": 0.00024768990351949993, + "loss": 3.5718, + "step": 88580 + }, + { + "epoch": 6.018820491914663, + "grad_norm": 0.7348518967628479, + "learning_rate": 0.0002476474385106672, + "loss": 3.3737, + "step": 88585 + }, + { + "epoch": 6.019160211985324, + "grad_norm": 1.0576852560043335, + "learning_rate": 0.0002476049735018345, + "loss": 3.535, + "step": 88590 + }, + { + "epoch": 6.019499932055986, + "grad_norm": 0.9384962916374207, + "learning_rate": 0.0002475625084930018, + "loss": 3.0076, + "step": 88595 + }, + { + "epoch": 6.019839652126648, + "grad_norm": 0.9129181504249573, + "learning_rate": 0.00024752004348416905, + "loss": 3.2396, + "step": 88600 + }, + { + "epoch": 6.020179372197309, + "grad_norm": 0.9567431211471558, + "learning_rate": 0.00024747757847533633, + "loss": 3.4615, + "step": 88605 + }, + { + "epoch": 6.020519092267971, + "grad_norm": 0.9321610331535339, + "learning_rate": 0.0002474351134665036, + "loss": 3.3304, + "step": 88610 + }, + { + "epoch": 6.020858812338633, + "grad_norm": 0.9863654971122742, + "learning_rate": 0.0002473926484576709, + "loss": 3.4429, + "step": 88615 + }, + { + "epoch": 6.0211985324092945, + "grad_norm": 0.8258249759674072, + "learning_rate": 0.0002473501834488382, + "loss": 3.4032, + "step": 88620 + }, + { + "epoch": 6.0215382524799566, + "grad_norm": 1.0859414339065552, + "learning_rate": 0.00024730771844000545, + "loss": 3.3105, + "step": 88625 + }, + { + "epoch": 6.021877972550619, + "grad_norm": 1.0607225894927979, + "learning_rate": 0.00024726525343117273, + "loss": 3.5009, + "step": 88630 + }, + { + "epoch": 6.02221769262128, + "grad_norm": 0.8315865993499756, + "learning_rate": 0.00024722278842234, + "loss": 3.371, + "step": 88635 + }, + { + "epoch": 6.022557412691942, + "grad_norm": 0.9273335933685303, + "learning_rate": 0.0002471803234135073, + "loss": 3.4339, + "step": 88640 + }, + { + "epoch": 6.022897132762604, + "grad_norm": 0.8942989110946655, + "learning_rate": 0.0002471378584046746, + "loss": 3.4538, + "step": 88645 + }, + { + "epoch": 6.023236852833265, + "grad_norm": 1.0561470985412598, + "learning_rate": 0.00024709539339584185, + "loss": 3.2869, + "step": 88650 + }, + { + "epoch": 6.023576572903927, + "grad_norm": 0.908456027507782, + "learning_rate": 0.0002470529283870091, + "loss": 3.2228, + "step": 88655 + }, + { + "epoch": 6.023916292974589, + "grad_norm": 0.93304842710495, + "learning_rate": 0.0002470104633781764, + "loss": 3.4016, + "step": 88660 + }, + { + "epoch": 6.0242560130452505, + "grad_norm": 0.8405183553695679, + "learning_rate": 0.00024696799836934364, + "loss": 3.114, + "step": 88665 + }, + { + "epoch": 6.024595733115913, + "grad_norm": 1.0090446472167969, + "learning_rate": 0.0002469255333605109, + "loss": 3.6521, + "step": 88670 + }, + { + "epoch": 6.024935453186575, + "grad_norm": 0.7723715305328369, + "learning_rate": 0.00024688306835167825, + "loss": 3.4578, + "step": 88675 + }, + { + "epoch": 6.025275173257236, + "grad_norm": 0.9237025380134583, + "learning_rate": 0.0002468406033428455, + "loss": 3.4888, + "step": 88680 + }, + { + "epoch": 6.025614893327898, + "grad_norm": 1.0106652975082397, + "learning_rate": 0.00024679813833401276, + "loss": 3.3312, + "step": 88685 + }, + { + "epoch": 6.02595461339856, + "grad_norm": 0.9861084222793579, + "learning_rate": 0.00024675567332518004, + "loss": 3.5458, + "step": 88690 + }, + { + "epoch": 6.026294333469221, + "grad_norm": 0.7567527890205383, + "learning_rate": 0.0002467132083163473, + "loss": 3.2004, + "step": 88695 + }, + { + "epoch": 6.026634053539883, + "grad_norm": 0.7753181457519531, + "learning_rate": 0.0002466707433075146, + "loss": 3.7683, + "step": 88700 + }, + { + "epoch": 6.026973773610545, + "grad_norm": 0.9352250695228577, + "learning_rate": 0.0002466282782986819, + "loss": 3.2153, + "step": 88705 + }, + { + "epoch": 6.0273134936812065, + "grad_norm": 1.3977853059768677, + "learning_rate": 0.00024658581328984916, + "loss": 3.2418, + "step": 88710 + }, + { + "epoch": 6.027653213751869, + "grad_norm": 0.9408144354820251, + "learning_rate": 0.00024654334828101644, + "loss": 3.5238, + "step": 88715 + }, + { + "epoch": 6.027992933822531, + "grad_norm": 0.9628551006317139, + "learning_rate": 0.0002465008832721837, + "loss": 3.4154, + "step": 88720 + }, + { + "epoch": 6.028332653893192, + "grad_norm": 0.8153097033500671, + "learning_rate": 0.000246458418263351, + "loss": 3.5282, + "step": 88725 + }, + { + "epoch": 6.028672373963854, + "grad_norm": 0.770617663860321, + "learning_rate": 0.0002464159532545183, + "loss": 3.6552, + "step": 88730 + }, + { + "epoch": 6.029012094034516, + "grad_norm": 0.8381354212760925, + "learning_rate": 0.00024637348824568556, + "loss": 3.3728, + "step": 88735 + }, + { + "epoch": 6.029351814105177, + "grad_norm": 0.9300280809402466, + "learning_rate": 0.00024633102323685284, + "loss": 3.6506, + "step": 88740 + }, + { + "epoch": 6.029691534175839, + "grad_norm": 0.8082245588302612, + "learning_rate": 0.0002462885582280201, + "loss": 3.4628, + "step": 88745 + }, + { + "epoch": 6.0300312542465, + "grad_norm": 7.8649702072143555, + "learning_rate": 0.0002462460932191874, + "loss": 3.3466, + "step": 88750 + }, + { + "epoch": 6.0303709743171625, + "grad_norm": 1.1054326295852661, + "learning_rate": 0.0002462036282103547, + "loss": 3.5422, + "step": 88755 + }, + { + "epoch": 6.030710694387825, + "grad_norm": 0.8808524012565613, + "learning_rate": 0.00024616116320152196, + "loss": 3.2217, + "step": 88760 + }, + { + "epoch": 6.031050414458486, + "grad_norm": 1.1410502195358276, + "learning_rate": 0.00024611869819268924, + "loss": 3.41, + "step": 88765 + }, + { + "epoch": 6.031390134529148, + "grad_norm": 0.9590603709220886, + "learning_rate": 0.0002460762331838565, + "loss": 3.0533, + "step": 88770 + }, + { + "epoch": 6.03172985459981, + "grad_norm": 0.7441473007202148, + "learning_rate": 0.0002460337681750238, + "loss": 3.4724, + "step": 88775 + }, + { + "epoch": 6.032069574670471, + "grad_norm": 0.933161735534668, + "learning_rate": 0.0002459913031661911, + "loss": 3.2703, + "step": 88780 + }, + { + "epoch": 6.032409294741133, + "grad_norm": 1.0800739526748657, + "learning_rate": 0.0002459488381573583, + "loss": 3.2545, + "step": 88785 + }, + { + "epoch": 6.032749014811795, + "grad_norm": 0.9666741490364075, + "learning_rate": 0.00024590637314852564, + "loss": 3.4437, + "step": 88790 + }, + { + "epoch": 6.0330887348824564, + "grad_norm": 0.9822800755500793, + "learning_rate": 0.00024586390813969287, + "loss": 3.5068, + "step": 88795 + }, + { + "epoch": 6.0334284549531185, + "grad_norm": 1.0650913715362549, + "learning_rate": 0.0002458214431308602, + "loss": 3.4599, + "step": 88800 + }, + { + "epoch": 6.033768175023781, + "grad_norm": 1.0070991516113281, + "learning_rate": 0.0002457789781220275, + "loss": 3.5054, + "step": 88805 + }, + { + "epoch": 6.034107895094442, + "grad_norm": 1.036076545715332, + "learning_rate": 0.0002457365131131947, + "loss": 3.4976, + "step": 88810 + }, + { + "epoch": 6.034447615165104, + "grad_norm": 1.2247867584228516, + "learning_rate": 0.00024569404810436204, + "loss": 3.2631, + "step": 88815 + }, + { + "epoch": 6.034787335235766, + "grad_norm": 1.0721193552017212, + "learning_rate": 0.00024565158309552927, + "loss": 3.5988, + "step": 88820 + }, + { + "epoch": 6.035127055306427, + "grad_norm": 0.8092731833457947, + "learning_rate": 0.00024560911808669655, + "loss": 3.4233, + "step": 88825 + }, + { + "epoch": 6.035466775377089, + "grad_norm": 0.9987058043479919, + "learning_rate": 0.00024556665307786383, + "loss": 3.5003, + "step": 88830 + }, + { + "epoch": 6.035806495447751, + "grad_norm": 0.9242515563964844, + "learning_rate": 0.0002455241880690311, + "loss": 3.3138, + "step": 88835 + }, + { + "epoch": 6.0361462155184125, + "grad_norm": 0.8274189829826355, + "learning_rate": 0.0002454817230601984, + "loss": 3.2505, + "step": 88840 + }, + { + "epoch": 6.0364859355890745, + "grad_norm": 1.4141457080841064, + "learning_rate": 0.00024543925805136567, + "loss": 3.1841, + "step": 88845 + }, + { + "epoch": 6.036825655659737, + "grad_norm": 0.8867030143737793, + "learning_rate": 0.00024539679304253295, + "loss": 3.5033, + "step": 88850 + }, + { + "epoch": 6.037165375730398, + "grad_norm": 0.8536878824234009, + "learning_rate": 0.00024535432803370023, + "loss": 3.4905, + "step": 88855 + }, + { + "epoch": 6.03750509580106, + "grad_norm": 0.8773537278175354, + "learning_rate": 0.0002453118630248675, + "loss": 3.7687, + "step": 88860 + }, + { + "epoch": 6.037844815871722, + "grad_norm": 1.0221110582351685, + "learning_rate": 0.0002452693980160348, + "loss": 3.1463, + "step": 88865 + }, + { + "epoch": 6.038184535942383, + "grad_norm": 0.9372830390930176, + "learning_rate": 0.00024522693300720207, + "loss": 3.5497, + "step": 88870 + }, + { + "epoch": 6.038524256013045, + "grad_norm": 0.9079229235649109, + "learning_rate": 0.00024518446799836935, + "loss": 3.2768, + "step": 88875 + }, + { + "epoch": 6.038863976083707, + "grad_norm": 0.8554083108901978, + "learning_rate": 0.00024514200298953663, + "loss": 3.6529, + "step": 88880 + }, + { + "epoch": 6.0392036961543685, + "grad_norm": 0.9396691918373108, + "learning_rate": 0.0002450995379807039, + "loss": 3.3993, + "step": 88885 + }, + { + "epoch": 6.0395434162250305, + "grad_norm": 0.9670695066452026, + "learning_rate": 0.0002450570729718712, + "loss": 3.1865, + "step": 88890 + }, + { + "epoch": 6.039883136295693, + "grad_norm": 0.9635370373725891, + "learning_rate": 0.00024501460796303847, + "loss": 3.2543, + "step": 88895 + }, + { + "epoch": 6.040222856366354, + "grad_norm": 1.0157456398010254, + "learning_rate": 0.00024497214295420575, + "loss": 3.484, + "step": 88900 + }, + { + "epoch": 6.040562576437016, + "grad_norm": 1.019752025604248, + "learning_rate": 0.00024492967794537303, + "loss": 3.5005, + "step": 88905 + }, + { + "epoch": 6.040902296507678, + "grad_norm": 1.3231010437011719, + "learning_rate": 0.0002448872129365403, + "loss": 3.3754, + "step": 88910 + }, + { + "epoch": 6.041242016578339, + "grad_norm": 1.0389431715011597, + "learning_rate": 0.0002448447479277076, + "loss": 3.4486, + "step": 88915 + }, + { + "epoch": 6.041581736649001, + "grad_norm": 1.212103247642517, + "learning_rate": 0.00024480228291887487, + "loss": 3.3149, + "step": 88920 + }, + { + "epoch": 6.041921456719663, + "grad_norm": 0.9145174622535706, + "learning_rate": 0.0002447598179100421, + "loss": 3.2902, + "step": 88925 + }, + { + "epoch": 6.0422611767903245, + "grad_norm": 0.8497671484947205, + "learning_rate": 0.00024471735290120943, + "loss": 3.4799, + "step": 88930 + }, + { + "epoch": 6.042600896860987, + "grad_norm": 0.7781046032905579, + "learning_rate": 0.0002446748878923767, + "loss": 3.647, + "step": 88935 + }, + { + "epoch": 6.042940616931649, + "grad_norm": 0.9547692537307739, + "learning_rate": 0.00024463242288354394, + "loss": 3.2677, + "step": 88940 + }, + { + "epoch": 6.04328033700231, + "grad_norm": 0.8554035425186157, + "learning_rate": 0.00024458995787471127, + "loss": 3.3818, + "step": 88945 + }, + { + "epoch": 6.043620057072972, + "grad_norm": 0.8663648366928101, + "learning_rate": 0.0002445474928658785, + "loss": 3.5146, + "step": 88950 + }, + { + "epoch": 6.043959777143634, + "grad_norm": 1.0172157287597656, + "learning_rate": 0.0002445050278570458, + "loss": 3.4976, + "step": 88955 + }, + { + "epoch": 6.044299497214295, + "grad_norm": 0.97449791431427, + "learning_rate": 0.00024446256284821306, + "loss": 3.672, + "step": 88960 + }, + { + "epoch": 6.044639217284957, + "grad_norm": 0.910797119140625, + "learning_rate": 0.00024442009783938034, + "loss": 3.3802, + "step": 88965 + }, + { + "epoch": 6.044978937355619, + "grad_norm": 1.1318457126617432, + "learning_rate": 0.00024437763283054767, + "loss": 3.6854, + "step": 88970 + }, + { + "epoch": 6.0453186574262805, + "grad_norm": 0.7808481454849243, + "learning_rate": 0.0002443351678217149, + "loss": 3.2249, + "step": 88975 + }, + { + "epoch": 6.045658377496943, + "grad_norm": 0.864913821220398, + "learning_rate": 0.0002442927028128822, + "loss": 3.4333, + "step": 88980 + }, + { + "epoch": 6.045998097567605, + "grad_norm": 1.0262689590454102, + "learning_rate": 0.00024425023780404946, + "loss": 3.4359, + "step": 88985 + }, + { + "epoch": 6.046337817638266, + "grad_norm": 0.8884408473968506, + "learning_rate": 0.00024420777279521674, + "loss": 3.105, + "step": 88990 + }, + { + "epoch": 6.046677537708928, + "grad_norm": 1.0499542951583862, + "learning_rate": 0.000244165307786384, + "loss": 3.3534, + "step": 88995 + }, + { + "epoch": 6.04701725777959, + "grad_norm": 0.927367091178894, + "learning_rate": 0.00024412284277755132, + "loss": 3.0436, + "step": 89000 + }, + { + "epoch": 6.047356977850251, + "grad_norm": 1.0168758630752563, + "learning_rate": 0.00024408037776871858, + "loss": 3.395, + "step": 89005 + }, + { + "epoch": 6.047696697920913, + "grad_norm": 1.0082067251205444, + "learning_rate": 0.00024403791275988586, + "loss": 3.3164, + "step": 89010 + }, + { + "epoch": 6.048036417991575, + "grad_norm": 0.9303625822067261, + "learning_rate": 0.00024399544775105314, + "loss": 3.3134, + "step": 89015 + }, + { + "epoch": 6.0483761380622365, + "grad_norm": 0.7705940008163452, + "learning_rate": 0.00024395298274222042, + "loss": 3.1042, + "step": 89020 + }, + { + "epoch": 6.048715858132899, + "grad_norm": 1.333645224571228, + "learning_rate": 0.00024391051773338767, + "loss": 3.3914, + "step": 89025 + }, + { + "epoch": 6.049055578203561, + "grad_norm": 0.7887675166130066, + "learning_rate": 0.00024386805272455498, + "loss": 3.3755, + "step": 89030 + }, + { + "epoch": 6.049395298274222, + "grad_norm": 0.8471320867538452, + "learning_rate": 0.00024382558771572226, + "loss": 3.47, + "step": 89035 + }, + { + "epoch": 6.049735018344884, + "grad_norm": 1.054848551750183, + "learning_rate": 0.0002437831227068895, + "loss": 3.5282, + "step": 89040 + }, + { + "epoch": 6.050074738415546, + "grad_norm": 0.9673366546630859, + "learning_rate": 0.00024374065769805682, + "loss": 3.4267, + "step": 89045 + }, + { + "epoch": 6.050414458486207, + "grad_norm": 0.8275158405303955, + "learning_rate": 0.00024369819268922407, + "loss": 3.3382, + "step": 89050 + }, + { + "epoch": 6.050754178556869, + "grad_norm": 0.9961436986923218, + "learning_rate": 0.00024365572768039138, + "loss": 3.4944, + "step": 89055 + }, + { + "epoch": 6.051093898627531, + "grad_norm": 0.8575528860092163, + "learning_rate": 0.00024361326267155863, + "loss": 3.3952, + "step": 89060 + }, + { + "epoch": 6.0514336186981925, + "grad_norm": 1.018876075744629, + "learning_rate": 0.0002435707976627259, + "loss": 3.4235, + "step": 89065 + }, + { + "epoch": 6.051773338768855, + "grad_norm": 1.0575248003005981, + "learning_rate": 0.00024352833265389322, + "loss": 3.5268, + "step": 89070 + }, + { + "epoch": 6.052113058839517, + "grad_norm": 1.096799373626709, + "learning_rate": 0.00024348586764506047, + "loss": 3.1757, + "step": 89075 + }, + { + "epoch": 6.052452778910178, + "grad_norm": 0.9118776917457581, + "learning_rate": 0.00024344340263622775, + "loss": 3.4698, + "step": 89080 + }, + { + "epoch": 6.05279249898084, + "grad_norm": 0.9961947202682495, + "learning_rate": 0.00024340093762739503, + "loss": 3.3597, + "step": 89085 + }, + { + "epoch": 6.053132219051502, + "grad_norm": 0.7569659352302551, + "learning_rate": 0.0002433584726185623, + "loss": 3.3362, + "step": 89090 + }, + { + "epoch": 6.053471939122163, + "grad_norm": 0.8684404492378235, + "learning_rate": 0.0002433160076097296, + "loss": 3.3033, + "step": 89095 + }, + { + "epoch": 6.053811659192825, + "grad_norm": 0.8340876698493958, + "learning_rate": 0.00024327354260089687, + "loss": 3.3393, + "step": 89100 + }, + { + "epoch": 6.0541513792634865, + "grad_norm": 0.7909108996391296, + "learning_rate": 0.00024323107759206415, + "loss": 3.4961, + "step": 89105 + }, + { + "epoch": 6.0544910993341485, + "grad_norm": 0.8043861985206604, + "learning_rate": 0.0002431886125832314, + "loss": 3.388, + "step": 89110 + }, + { + "epoch": 6.054830819404811, + "grad_norm": 0.8000028729438782, + "learning_rate": 0.0002431461475743987, + "loss": 2.8677, + "step": 89115 + }, + { + "epoch": 6.055170539475472, + "grad_norm": 0.798330545425415, + "learning_rate": 0.00024310368256556596, + "loss": 3.607, + "step": 89120 + }, + { + "epoch": 6.055510259546134, + "grad_norm": 1.0592607259750366, + "learning_rate": 0.00024306121755673324, + "loss": 3.4386, + "step": 89125 + }, + { + "epoch": 6.055849979616796, + "grad_norm": 0.8507229685783386, + "learning_rate": 0.00024301875254790055, + "loss": 3.0949, + "step": 89130 + }, + { + "epoch": 6.056189699687457, + "grad_norm": 0.7841441035270691, + "learning_rate": 0.0002429762875390678, + "loss": 3.56, + "step": 89135 + }, + { + "epoch": 6.056529419758119, + "grad_norm": 0.6671336889266968, + "learning_rate": 0.0002429338225302351, + "loss": 3.2357, + "step": 89140 + }, + { + "epoch": 6.056869139828781, + "grad_norm": 0.8808044791221619, + "learning_rate": 0.00024289135752140236, + "loss": 3.3065, + "step": 89145 + }, + { + "epoch": 6.0572088598994425, + "grad_norm": 0.8722726702690125, + "learning_rate": 0.00024284889251256964, + "loss": 3.3063, + "step": 89150 + }, + { + "epoch": 6.0575485799701045, + "grad_norm": 0.8668311238288879, + "learning_rate": 0.00024280642750373693, + "loss": 3.4823, + "step": 89155 + }, + { + "epoch": 6.057888300040767, + "grad_norm": 0.9184710383415222, + "learning_rate": 0.0002427639624949042, + "loss": 3.1972, + "step": 89160 + }, + { + "epoch": 6.058228020111428, + "grad_norm": 1.1044507026672363, + "learning_rate": 0.00024272149748607149, + "loss": 3.4794, + "step": 89165 + }, + { + "epoch": 6.05856774018209, + "grad_norm": 0.9794785976409912, + "learning_rate": 0.00024267903247723877, + "loss": 3.3973, + "step": 89170 + }, + { + "epoch": 6.058907460252752, + "grad_norm": 0.8939027786254883, + "learning_rate": 0.00024263656746840605, + "loss": 3.7392, + "step": 89175 + }, + { + "epoch": 6.059247180323413, + "grad_norm": 0.9125053882598877, + "learning_rate": 0.0002425941024595733, + "loss": 3.4527, + "step": 89180 + }, + { + "epoch": 6.059586900394075, + "grad_norm": 0.979253888130188, + "learning_rate": 0.0002425516374507406, + "loss": 3.4813, + "step": 89185 + }, + { + "epoch": 6.059926620464737, + "grad_norm": 0.842002809047699, + "learning_rate": 0.00024250917244190786, + "loss": 3.5036, + "step": 89190 + }, + { + "epoch": 6.0602663405353985, + "grad_norm": 0.8517590761184692, + "learning_rate": 0.00024246670743307514, + "loss": 3.3076, + "step": 89195 + }, + { + "epoch": 6.0606060606060606, + "grad_norm": 1.1651798486709595, + "learning_rate": 0.00024242424242424245, + "loss": 3.5058, + "step": 89200 + }, + { + "epoch": 6.060945780676723, + "grad_norm": 0.6713446974754333, + "learning_rate": 0.0002423817774154097, + "loss": 3.2238, + "step": 89205 + }, + { + "epoch": 6.061285500747384, + "grad_norm": 0.9533427357673645, + "learning_rate": 0.00024233931240657698, + "loss": 3.344, + "step": 89210 + }, + { + "epoch": 6.061625220818046, + "grad_norm": 0.8362655639648438, + "learning_rate": 0.00024229684739774426, + "loss": 3.1009, + "step": 89215 + }, + { + "epoch": 6.061964940888708, + "grad_norm": 0.8972848653793335, + "learning_rate": 0.00024225438238891154, + "loss": 3.4631, + "step": 89220 + }, + { + "epoch": 6.062304660959369, + "grad_norm": 1.0554741621017456, + "learning_rate": 0.00024221191738007885, + "loss": 3.3889, + "step": 89225 + }, + { + "epoch": 6.062644381030031, + "grad_norm": 0.8202165961265564, + "learning_rate": 0.0002421694523712461, + "loss": 3.5048, + "step": 89230 + }, + { + "epoch": 6.062984101100693, + "grad_norm": 0.847058892250061, + "learning_rate": 0.00024212698736241338, + "loss": 3.6726, + "step": 89235 + }, + { + "epoch": 6.0633238211713545, + "grad_norm": 1.0399094820022583, + "learning_rate": 0.00024208452235358066, + "loss": 3.3139, + "step": 89240 + }, + { + "epoch": 6.063663541242017, + "grad_norm": 0.8836455941200256, + "learning_rate": 0.00024204205734474794, + "loss": 3.508, + "step": 89245 + }, + { + "epoch": 6.064003261312679, + "grad_norm": 0.9140763282775879, + "learning_rate": 0.0002419995923359152, + "loss": 3.5849, + "step": 89250 + }, + { + "epoch": 6.06434298138334, + "grad_norm": 0.7426170110702515, + "learning_rate": 0.0002419571273270825, + "loss": 3.3808, + "step": 89255 + }, + { + "epoch": 6.064682701454002, + "grad_norm": 0.778454065322876, + "learning_rate": 0.00024191466231824978, + "loss": 3.0881, + "step": 89260 + }, + { + "epoch": 6.065022421524664, + "grad_norm": 0.7027047276496887, + "learning_rate": 0.00024187219730941703, + "loss": 3.4707, + "step": 89265 + }, + { + "epoch": 6.065362141595325, + "grad_norm": 0.7255326509475708, + "learning_rate": 0.00024182973230058434, + "loss": 3.6035, + "step": 89270 + }, + { + "epoch": 6.065701861665987, + "grad_norm": 1.2504273653030396, + "learning_rate": 0.0002417872672917516, + "loss": 3.3604, + "step": 89275 + }, + { + "epoch": 6.066041581736649, + "grad_norm": 1.0556514263153076, + "learning_rate": 0.00024174480228291887, + "loss": 3.2635, + "step": 89280 + }, + { + "epoch": 6.0663813018073105, + "grad_norm": 0.8731863498687744, + "learning_rate": 0.00024170233727408615, + "loss": 3.3795, + "step": 89285 + }, + { + "epoch": 6.066721021877973, + "grad_norm": 1.3120312690734863, + "learning_rate": 0.00024165987226525343, + "loss": 3.7312, + "step": 89290 + }, + { + "epoch": 6.067060741948635, + "grad_norm": 0.9159737229347229, + "learning_rate": 0.0002416174072564207, + "loss": 3.7588, + "step": 89295 + }, + { + "epoch": 6.067400462019296, + "grad_norm": 0.6884854435920715, + "learning_rate": 0.000241574942247588, + "loss": 3.5437, + "step": 89300 + }, + { + "epoch": 6.067740182089958, + "grad_norm": 0.9283587336540222, + "learning_rate": 0.00024153247723875527, + "loss": 3.3181, + "step": 89305 + }, + { + "epoch": 6.06807990216062, + "grad_norm": 1.0496129989624023, + "learning_rate": 0.00024149001222992255, + "loss": 3.4097, + "step": 89310 + }, + { + "epoch": 6.068419622231281, + "grad_norm": 1.1161385774612427, + "learning_rate": 0.00024144754722108983, + "loss": 3.3568, + "step": 89315 + }, + { + "epoch": 6.068759342301943, + "grad_norm": 1.5862756967544556, + "learning_rate": 0.00024140508221225709, + "loss": 3.1617, + "step": 89320 + }, + { + "epoch": 6.069099062372605, + "grad_norm": 0.826882004737854, + "learning_rate": 0.0002413626172034244, + "loss": 3.7457, + "step": 89325 + }, + { + "epoch": 6.0694387824432665, + "grad_norm": 1.15510094165802, + "learning_rate": 0.00024132015219459167, + "loss": 3.4718, + "step": 89330 + }, + { + "epoch": 6.069778502513929, + "grad_norm": 0.8577184677124023, + "learning_rate": 0.00024127768718575893, + "loss": 3.5618, + "step": 89335 + }, + { + "epoch": 6.070118222584591, + "grad_norm": 0.8503198027610779, + "learning_rate": 0.00024123522217692623, + "loss": 3.6982, + "step": 89340 + }, + { + "epoch": 6.070457942655252, + "grad_norm": 0.9446349143981934, + "learning_rate": 0.0002411927571680935, + "loss": 3.1911, + "step": 89345 + }, + { + "epoch": 6.070797662725914, + "grad_norm": 0.9949141144752502, + "learning_rate": 0.00024115029215926077, + "loss": 3.4352, + "step": 89350 + }, + { + "epoch": 6.071137382796576, + "grad_norm": 1.071476936340332, + "learning_rate": 0.00024110782715042807, + "loss": 3.3979, + "step": 89355 + }, + { + "epoch": 6.071477102867237, + "grad_norm": 0.9801374673843384, + "learning_rate": 0.00024106536214159533, + "loss": 3.3683, + "step": 89360 + }, + { + "epoch": 6.071816822937899, + "grad_norm": 0.8223042488098145, + "learning_rate": 0.0002410228971327626, + "loss": 3.5827, + "step": 89365 + }, + { + "epoch": 6.072156543008561, + "grad_norm": 0.8270096182823181, + "learning_rate": 0.0002409804321239299, + "loss": 3.5156, + "step": 89370 + }, + { + "epoch": 6.0724962630792225, + "grad_norm": 0.9069366455078125, + "learning_rate": 0.00024093796711509717, + "loss": 3.3182, + "step": 89375 + }, + { + "epoch": 6.072835983149885, + "grad_norm": 0.7479249835014343, + "learning_rate": 0.00024089550210626442, + "loss": 3.4049, + "step": 89380 + }, + { + "epoch": 6.073175703220547, + "grad_norm": 0.858173131942749, + "learning_rate": 0.00024085303709743173, + "loss": 3.4638, + "step": 89385 + }, + { + "epoch": 6.073515423291208, + "grad_norm": 0.7299743294715881, + "learning_rate": 0.000240810572088599, + "loss": 3.6008, + "step": 89390 + }, + { + "epoch": 6.07385514336187, + "grad_norm": 1.2593039274215698, + "learning_rate": 0.0002407681070797663, + "loss": 3.5816, + "step": 89395 + }, + { + "epoch": 6.074194863432532, + "grad_norm": 0.9869457483291626, + "learning_rate": 0.00024072564207093357, + "loss": 3.0555, + "step": 89400 + }, + { + "epoch": 6.074534583503193, + "grad_norm": 0.9097234606742859, + "learning_rate": 0.00024068317706210082, + "loss": 3.5277, + "step": 89405 + }, + { + "epoch": 6.074874303573855, + "grad_norm": 0.9457623958587646, + "learning_rate": 0.00024064071205326813, + "loss": 3.5904, + "step": 89410 + }, + { + "epoch": 6.075214023644517, + "grad_norm": 0.9601683616638184, + "learning_rate": 0.00024059824704443538, + "loss": 3.433, + "step": 89415 + }, + { + "epoch": 6.0755537437151785, + "grad_norm": 0.809941291809082, + "learning_rate": 0.00024055578203560266, + "loss": 3.3049, + "step": 89420 + }, + { + "epoch": 6.075893463785841, + "grad_norm": 0.9254326820373535, + "learning_rate": 0.00024051331702676997, + "loss": 3.5131, + "step": 89425 + }, + { + "epoch": 6.076233183856502, + "grad_norm": 0.7775248289108276, + "learning_rate": 0.00024047085201793722, + "loss": 3.6325, + "step": 89430 + }, + { + "epoch": 6.076572903927164, + "grad_norm": 0.9692682027816772, + "learning_rate": 0.0002404283870091045, + "loss": 3.5123, + "step": 89435 + }, + { + "epoch": 6.076912623997826, + "grad_norm": 0.9628472328186035, + "learning_rate": 0.00024038592200027178, + "loss": 3.6272, + "step": 89440 + }, + { + "epoch": 6.077252344068487, + "grad_norm": 0.9200848340988159, + "learning_rate": 0.00024034345699143906, + "loss": 3.6366, + "step": 89445 + }, + { + "epoch": 6.077592064139149, + "grad_norm": 1.0391849279403687, + "learning_rate": 0.00024030099198260631, + "loss": 3.3281, + "step": 89450 + }, + { + "epoch": 6.077931784209811, + "grad_norm": 0.9352471828460693, + "learning_rate": 0.00024025852697377362, + "loss": 3.4297, + "step": 89455 + }, + { + "epoch": 6.0782715042804725, + "grad_norm": 1.046221375465393, + "learning_rate": 0.0002402160619649409, + "loss": 3.549, + "step": 89460 + }, + { + "epoch": 6.0786112243511345, + "grad_norm": 1.1687648296356201, + "learning_rate": 0.00024017359695610815, + "loss": 3.2802, + "step": 89465 + }, + { + "epoch": 6.078950944421797, + "grad_norm": 1.0716989040374756, + "learning_rate": 0.00024013113194727546, + "loss": 3.5767, + "step": 89470 + }, + { + "epoch": 6.079290664492458, + "grad_norm": 0.9735480546951294, + "learning_rate": 0.00024008866693844271, + "loss": 3.298, + "step": 89475 + }, + { + "epoch": 6.07963038456312, + "grad_norm": 0.8288670182228088, + "learning_rate": 0.00024004620192961002, + "loss": 3.4705, + "step": 89480 + }, + { + "epoch": 6.079970104633782, + "grad_norm": 0.9510166049003601, + "learning_rate": 0.00024000373692077727, + "loss": 3.4363, + "step": 89485 + }, + { + "epoch": 6.080309824704443, + "grad_norm": 0.9233943223953247, + "learning_rate": 0.00023996127191194455, + "loss": 3.4316, + "step": 89490 + }, + { + "epoch": 6.080649544775105, + "grad_norm": 1.009620189666748, + "learning_rate": 0.00023991880690311186, + "loss": 3.4537, + "step": 89495 + }, + { + "epoch": 6.080989264845767, + "grad_norm": 0.9265052080154419, + "learning_rate": 0.00023987634189427911, + "loss": 3.3913, + "step": 89500 + }, + { + "epoch": 6.0813289849164285, + "grad_norm": 0.9221977591514587, + "learning_rate": 0.0002398338768854464, + "loss": 3.5063, + "step": 89505 + }, + { + "epoch": 6.081668704987091, + "grad_norm": 1.019503116607666, + "learning_rate": 0.00023979141187661367, + "loss": 3.6605, + "step": 89510 + }, + { + "epoch": 6.082008425057753, + "grad_norm": 1.3186135292053223, + "learning_rate": 0.00023974894686778096, + "loss": 3.4421, + "step": 89515 + }, + { + "epoch": 6.082348145128414, + "grad_norm": 0.7029008269309998, + "learning_rate": 0.00023970648185894824, + "loss": 3.4251, + "step": 89520 + }, + { + "epoch": 6.082687865199076, + "grad_norm": 0.8604206442832947, + "learning_rate": 0.00023966401685011552, + "loss": 3.4309, + "step": 89525 + }, + { + "epoch": 6.083027585269738, + "grad_norm": 1.0671781301498413, + "learning_rate": 0.0002396215518412828, + "loss": 3.4794, + "step": 89530 + }, + { + "epoch": 6.083367305340399, + "grad_norm": 0.9547089338302612, + "learning_rate": 0.00023957908683245005, + "loss": 3.4886, + "step": 89535 + }, + { + "epoch": 6.083707025411061, + "grad_norm": 0.9834766983985901, + "learning_rate": 0.00023953662182361736, + "loss": 3.125, + "step": 89540 + }, + { + "epoch": 6.084046745481723, + "grad_norm": 1.1925398111343384, + "learning_rate": 0.0002394941568147846, + "loss": 3.3794, + "step": 89545 + }, + { + "epoch": 6.0843864655523845, + "grad_norm": 0.9400942325592041, + "learning_rate": 0.0002394516918059519, + "loss": 3.3345, + "step": 89550 + }, + { + "epoch": 6.084726185623047, + "grad_norm": 0.7922540307044983, + "learning_rate": 0.0002394092267971192, + "loss": 3.2687, + "step": 89555 + }, + { + "epoch": 6.085065905693709, + "grad_norm": 0.8292868137359619, + "learning_rate": 0.00023936676178828645, + "loss": 3.0784, + "step": 89560 + }, + { + "epoch": 6.08540562576437, + "grad_norm": 1.1590956449508667, + "learning_rate": 0.00023932429677945376, + "loss": 3.2806, + "step": 89565 + }, + { + "epoch": 6.085745345835032, + "grad_norm": 0.8257320523262024, + "learning_rate": 0.000239281831770621, + "loss": 3.4389, + "step": 89570 + }, + { + "epoch": 6.086085065905694, + "grad_norm": 1.0026909112930298, + "learning_rate": 0.0002392393667617883, + "loss": 3.401, + "step": 89575 + }, + { + "epoch": 6.086424785976355, + "grad_norm": 0.9434211254119873, + "learning_rate": 0.00023919690175295557, + "loss": 3.4479, + "step": 89580 + }, + { + "epoch": 6.086764506047017, + "grad_norm": 0.8856450915336609, + "learning_rate": 0.00023915443674412285, + "loss": 3.1343, + "step": 89585 + }, + { + "epoch": 6.087104226117679, + "grad_norm": 0.832743227481842, + "learning_rate": 0.00023911197173529013, + "loss": 3.3018, + "step": 89590 + }, + { + "epoch": 6.0874439461883405, + "grad_norm": 0.8996881246566772, + "learning_rate": 0.0002390695067264574, + "loss": 3.5109, + "step": 89595 + }, + { + "epoch": 6.087783666259003, + "grad_norm": 0.8773401379585266, + "learning_rate": 0.0002390270417176247, + "loss": 3.5513, + "step": 89600 + }, + { + "epoch": 6.088123386329665, + "grad_norm": 0.9596992135047913, + "learning_rate": 0.00023898457670879194, + "loss": 3.166, + "step": 89605 + }, + { + "epoch": 6.088463106400326, + "grad_norm": 0.9178014993667603, + "learning_rate": 0.00023894211169995925, + "loss": 3.5138, + "step": 89610 + }, + { + "epoch": 6.088802826470988, + "grad_norm": 0.9914410710334778, + "learning_rate": 0.0002388996466911265, + "loss": 3.4228, + "step": 89615 + }, + { + "epoch": 6.08914254654165, + "grad_norm": 0.9182834029197693, + "learning_rate": 0.00023885718168229378, + "loss": 3.5753, + "step": 89620 + }, + { + "epoch": 6.089482266612311, + "grad_norm": 1.0113338232040405, + "learning_rate": 0.0002388147166734611, + "loss": 3.6668, + "step": 89625 + }, + { + "epoch": 6.089821986682973, + "grad_norm": 1.071702480316162, + "learning_rate": 0.00023877225166462834, + "loss": 3.4788, + "step": 89630 + }, + { + "epoch": 6.090161706753635, + "grad_norm": 0.8054448366165161, + "learning_rate": 0.00023872978665579562, + "loss": 3.0753, + "step": 89635 + }, + { + "epoch": 6.0905014268242965, + "grad_norm": 0.7624775767326355, + "learning_rate": 0.0002386873216469629, + "loss": 3.3737, + "step": 89640 + }, + { + "epoch": 6.090841146894959, + "grad_norm": 1.0980470180511475, + "learning_rate": 0.00023864485663813018, + "loss": 3.3093, + "step": 89645 + }, + { + "epoch": 6.091180866965621, + "grad_norm": 0.9021480679512024, + "learning_rate": 0.0002386023916292975, + "loss": 3.2531, + "step": 89650 + }, + { + "epoch": 6.091520587036282, + "grad_norm": 0.793542206287384, + "learning_rate": 0.00023855992662046474, + "loss": 3.5148, + "step": 89655 + }, + { + "epoch": 6.091860307106944, + "grad_norm": 1.146060585975647, + "learning_rate": 0.00023851746161163202, + "loss": 3.5163, + "step": 89660 + }, + { + "epoch": 6.092200027177606, + "grad_norm": 0.9407854080200195, + "learning_rate": 0.0002384749966027993, + "loss": 3.1856, + "step": 89665 + }, + { + "epoch": 6.092539747248267, + "grad_norm": 0.9086285829544067, + "learning_rate": 0.00023843253159396658, + "loss": 3.5872, + "step": 89670 + }, + { + "epoch": 6.092879467318929, + "grad_norm": 0.9575817584991455, + "learning_rate": 0.00023839006658513384, + "loss": 3.2784, + "step": 89675 + }, + { + "epoch": 6.093219187389591, + "grad_norm": 0.9392192363739014, + "learning_rate": 0.00023834760157630114, + "loss": 3.4147, + "step": 89680 + }, + { + "epoch": 6.0935589074602525, + "grad_norm": 0.9434480667114258, + "learning_rate": 0.00023830513656746842, + "loss": 3.3388, + "step": 89685 + }, + { + "epoch": 6.093898627530915, + "grad_norm": 0.7922220230102539, + "learning_rate": 0.00023826267155863568, + "loss": 3.3362, + "step": 89690 + }, + { + "epoch": 6.094238347601577, + "grad_norm": 0.9328005313873291, + "learning_rate": 0.00023822020654980298, + "loss": 3.618, + "step": 89695 + }, + { + "epoch": 6.094578067672238, + "grad_norm": 1.1631463766098022, + "learning_rate": 0.00023817774154097024, + "loss": 3.38, + "step": 89700 + }, + { + "epoch": 6.0949177877429, + "grad_norm": 0.9439435601234436, + "learning_rate": 0.00023813527653213752, + "loss": 3.3579, + "step": 89705 + }, + { + "epoch": 6.095257507813562, + "grad_norm": 0.9873977303504944, + "learning_rate": 0.0002380928115233048, + "loss": 3.4534, + "step": 89710 + }, + { + "epoch": 6.095597227884223, + "grad_norm": 0.8384420275688171, + "learning_rate": 0.00023805034651447208, + "loss": 3.3921, + "step": 89715 + }, + { + "epoch": 6.095936947954885, + "grad_norm": 1.018181562423706, + "learning_rate": 0.00023800788150563936, + "loss": 3.6651, + "step": 89720 + }, + { + "epoch": 6.096276668025547, + "grad_norm": 1.0983372926712036, + "learning_rate": 0.00023796541649680664, + "loss": 3.5898, + "step": 89725 + }, + { + "epoch": 6.0966163880962085, + "grad_norm": 0.8391343951225281, + "learning_rate": 0.00023792295148797392, + "loss": 3.1235, + "step": 89730 + }, + { + "epoch": 6.096956108166871, + "grad_norm": 0.7298493981361389, + "learning_rate": 0.0002378804864791412, + "loss": 3.1618, + "step": 89735 + }, + { + "epoch": 6.097295828237533, + "grad_norm": 1.2733367681503296, + "learning_rate": 0.00023783802147030848, + "loss": 3.3387, + "step": 89740 + }, + { + "epoch": 6.097635548308194, + "grad_norm": 1.0613369941711426, + "learning_rate": 0.00023779555646147573, + "loss": 3.5222, + "step": 89745 + }, + { + "epoch": 6.097975268378856, + "grad_norm": 1.0554680824279785, + "learning_rate": 0.00023775309145264304, + "loss": 3.3089, + "step": 89750 + }, + { + "epoch": 6.098314988449518, + "grad_norm": 1.0433984994888306, + "learning_rate": 0.00023771062644381032, + "loss": 3.2612, + "step": 89755 + }, + { + "epoch": 6.098654708520179, + "grad_norm": 0.7245514988899231, + "learning_rate": 0.00023766816143497757, + "loss": 3.519, + "step": 89760 + }, + { + "epoch": 6.098994428590841, + "grad_norm": 0.8663594722747803, + "learning_rate": 0.00023762569642614488, + "loss": 3.5375, + "step": 89765 + }, + { + "epoch": 6.099334148661503, + "grad_norm": 0.9176639318466187, + "learning_rate": 0.00023758323141731213, + "loss": 3.3943, + "step": 89770 + }, + { + "epoch": 6.0996738687321646, + "grad_norm": 0.9776946902275085, + "learning_rate": 0.0002375407664084794, + "loss": 3.5684, + "step": 89775 + }, + { + "epoch": 6.100013588802827, + "grad_norm": 1.091001033782959, + "learning_rate": 0.00023749830139964672, + "loss": 3.5416, + "step": 89780 + }, + { + "epoch": 6.100353308873488, + "grad_norm": 1.0127372741699219, + "learning_rate": 0.00023745583639081397, + "loss": 3.3878, + "step": 89785 + }, + { + "epoch": 6.10069302894415, + "grad_norm": 0.8447362184524536, + "learning_rate": 0.00023741337138198125, + "loss": 3.3061, + "step": 89790 + }, + { + "epoch": 6.101032749014812, + "grad_norm": 0.6997787952423096, + "learning_rate": 0.00023737090637314853, + "loss": 3.4077, + "step": 89795 + }, + { + "epoch": 6.101372469085473, + "grad_norm": 1.1032695770263672, + "learning_rate": 0.0002373284413643158, + "loss": 3.3725, + "step": 89800 + }, + { + "epoch": 6.101712189156135, + "grad_norm": 0.9287084937095642, + "learning_rate": 0.00023728597635548306, + "loss": 3.5327, + "step": 89805 + }, + { + "epoch": 6.102051909226797, + "grad_norm": 0.9722607135772705, + "learning_rate": 0.00023724351134665037, + "loss": 3.2338, + "step": 89810 + }, + { + "epoch": 6.1023916292974585, + "grad_norm": 0.9047874808311462, + "learning_rate": 0.00023720104633781765, + "loss": 3.3315, + "step": 89815 + }, + { + "epoch": 6.102731349368121, + "grad_norm": 0.9178425669670105, + "learning_rate": 0.00023715858132898493, + "loss": 3.4827, + "step": 89820 + }, + { + "epoch": 6.103071069438783, + "grad_norm": 1.0871332883834839, + "learning_rate": 0.0002371161163201522, + "loss": 3.7032, + "step": 89825 + }, + { + "epoch": 6.103410789509444, + "grad_norm": 0.9227219820022583, + "learning_rate": 0.00023707365131131946, + "loss": 3.3957, + "step": 89830 + }, + { + "epoch": 6.103750509580106, + "grad_norm": 0.9181329607963562, + "learning_rate": 0.00023703118630248677, + "loss": 3.4744, + "step": 89835 + }, + { + "epoch": 6.104090229650768, + "grad_norm": 0.8841462135314941, + "learning_rate": 0.00023698872129365402, + "loss": 3.1555, + "step": 89840 + }, + { + "epoch": 6.104429949721429, + "grad_norm": 0.9821570515632629, + "learning_rate": 0.0002369462562848213, + "loss": 3.3941, + "step": 89845 + }, + { + "epoch": 6.104769669792091, + "grad_norm": 0.8757491111755371, + "learning_rate": 0.0002369037912759886, + "loss": 3.343, + "step": 89850 + }, + { + "epoch": 6.105109389862753, + "grad_norm": 0.7711431384086609, + "learning_rate": 0.00023686132626715586, + "loss": 3.3029, + "step": 89855 + }, + { + "epoch": 6.1054491099334145, + "grad_norm": 1.0460692644119263, + "learning_rate": 0.00023681886125832314, + "loss": 3.4889, + "step": 89860 + }, + { + "epoch": 6.105788830004077, + "grad_norm": 1.2256931066513062, + "learning_rate": 0.00023677639624949042, + "loss": 3.4003, + "step": 89865 + }, + { + "epoch": 6.106128550074739, + "grad_norm": 1.106864333152771, + "learning_rate": 0.0002367339312406577, + "loss": 3.6101, + "step": 89870 + }, + { + "epoch": 6.1064682701454, + "grad_norm": 0.8199442028999329, + "learning_rate": 0.00023669146623182496, + "loss": 3.5755, + "step": 89875 + }, + { + "epoch": 6.106807990216062, + "grad_norm": 0.9966558218002319, + "learning_rate": 0.00023664900122299227, + "loss": 3.471, + "step": 89880 + }, + { + "epoch": 6.107147710286724, + "grad_norm": 1.1259863376617432, + "learning_rate": 0.00023660653621415955, + "loss": 3.5596, + "step": 89885 + }, + { + "epoch": 6.107487430357385, + "grad_norm": 0.728262186050415, + "learning_rate": 0.0002365640712053268, + "loss": 3.371, + "step": 89890 + }, + { + "epoch": 6.107827150428047, + "grad_norm": 0.8186905384063721, + "learning_rate": 0.0002365216061964941, + "loss": 3.3853, + "step": 89895 + }, + { + "epoch": 6.108166870498709, + "grad_norm": 1.1975466012954712, + "learning_rate": 0.00023647914118766136, + "loss": 3.435, + "step": 89900 + }, + { + "epoch": 6.1085065905693705, + "grad_norm": 0.902852475643158, + "learning_rate": 0.00023643667617882867, + "loss": 3.5523, + "step": 89905 + }, + { + "epoch": 6.108846310640033, + "grad_norm": 1.040503740310669, + "learning_rate": 0.00023639421116999592, + "loss": 3.1976, + "step": 89910 + }, + { + "epoch": 6.109186030710695, + "grad_norm": 0.9125615358352661, + "learning_rate": 0.0002363517461611632, + "loss": 3.495, + "step": 89915 + }, + { + "epoch": 6.109525750781356, + "grad_norm": 1.1379427909851074, + "learning_rate": 0.0002363092811523305, + "loss": 3.5525, + "step": 89920 + }, + { + "epoch": 6.109865470852018, + "grad_norm": 0.8789522051811218, + "learning_rate": 0.00023626681614349776, + "loss": 3.2999, + "step": 89925 + }, + { + "epoch": 6.11020519092268, + "grad_norm": 0.7650138139724731, + "learning_rate": 0.00023622435113466504, + "loss": 3.3598, + "step": 89930 + }, + { + "epoch": 6.110544910993341, + "grad_norm": 0.7818825840950012, + "learning_rate": 0.00023618188612583232, + "loss": 3.2485, + "step": 89935 + }, + { + "epoch": 6.110884631064003, + "grad_norm": 0.920985996723175, + "learning_rate": 0.0002361394211169996, + "loss": 3.6962, + "step": 89940 + }, + { + "epoch": 6.111224351134665, + "grad_norm": 0.8646937012672424, + "learning_rate": 0.00023609695610816688, + "loss": 3.4434, + "step": 89945 + }, + { + "epoch": 6.1115640712053265, + "grad_norm": 0.8254199028015137, + "learning_rate": 0.00023605449109933416, + "loss": 3.533, + "step": 89950 + }, + { + "epoch": 6.111903791275989, + "grad_norm": 0.8395771980285645, + "learning_rate": 0.00023601202609050144, + "loss": 3.3314, + "step": 89955 + }, + { + "epoch": 6.112243511346651, + "grad_norm": 1.1409871578216553, + "learning_rate": 0.0002359695610816687, + "loss": 3.3654, + "step": 89960 + }, + { + "epoch": 6.112583231417312, + "grad_norm": 0.843114972114563, + "learning_rate": 0.000235927096072836, + "loss": 3.3947, + "step": 89965 + }, + { + "epoch": 6.112922951487974, + "grad_norm": 0.7661665081977844, + "learning_rate": 0.00023588463106400325, + "loss": 3.69, + "step": 89970 + }, + { + "epoch": 6.113262671558636, + "grad_norm": 0.8964381814002991, + "learning_rate": 0.00023584216605517053, + "loss": 3.5845, + "step": 89975 + }, + { + "epoch": 6.113602391629297, + "grad_norm": 0.7211498618125916, + "learning_rate": 0.00023579970104633784, + "loss": 3.4176, + "step": 89980 + }, + { + "epoch": 6.113942111699959, + "grad_norm": 1.1485949754714966, + "learning_rate": 0.0002357572360375051, + "loss": 3.3832, + "step": 89985 + }, + { + "epoch": 6.114281831770621, + "grad_norm": 0.953254759311676, + "learning_rate": 0.0002357147710286724, + "loss": 3.2367, + "step": 89990 + }, + { + "epoch": 6.1146215518412825, + "grad_norm": 1.1115642786026, + "learning_rate": 0.00023567230601983965, + "loss": 3.4254, + "step": 89995 + }, + { + "epoch": 6.114961271911945, + "grad_norm": 1.0170844793319702, + "learning_rate": 0.00023562984101100693, + "loss": 3.5521, + "step": 90000 + }, + { + "epoch": 6.115300991982607, + "grad_norm": 0.7317707538604736, + "learning_rate": 0.0002355873760021742, + "loss": 3.4283, + "step": 90005 + }, + { + "epoch": 6.115640712053268, + "grad_norm": 0.8784629106521606, + "learning_rate": 0.0002355449109933415, + "loss": 3.2437, + "step": 90010 + }, + { + "epoch": 6.11598043212393, + "grad_norm": 0.8254554271697998, + "learning_rate": 0.00023550244598450877, + "loss": 3.667, + "step": 90015 + }, + { + "epoch": 6.116320152194592, + "grad_norm": 1.0225645303726196, + "learning_rate": 0.00023545998097567605, + "loss": 3.2894, + "step": 90020 + }, + { + "epoch": 6.116659872265253, + "grad_norm": 0.6905179619789124, + "learning_rate": 0.00023541751596684333, + "loss": 3.607, + "step": 90025 + }, + { + "epoch": 6.116999592335915, + "grad_norm": 0.8879868388175964, + "learning_rate": 0.00023537505095801059, + "loss": 3.2633, + "step": 90030 + }, + { + "epoch": 6.117339312406577, + "grad_norm": 1.1156305074691772, + "learning_rate": 0.0002353325859491779, + "loss": 3.4961, + "step": 90035 + }, + { + "epoch": 6.1176790324772385, + "grad_norm": 0.8649004697799683, + "learning_rate": 0.00023529012094034515, + "loss": 3.5278, + "step": 90040 + }, + { + "epoch": 6.118018752547901, + "grad_norm": 0.9062698483467102, + "learning_rate": 0.00023524765593151243, + "loss": 3.5302, + "step": 90045 + }, + { + "epoch": 6.118358472618563, + "grad_norm": 0.9589974880218506, + "learning_rate": 0.00023520519092267973, + "loss": 3.8051, + "step": 90050 + }, + { + "epoch": 6.118698192689224, + "grad_norm": 0.8206161856651306, + "learning_rate": 0.000235162725913847, + "loss": 3.5343, + "step": 90055 + }, + { + "epoch": 6.119037912759886, + "grad_norm": 1.0611627101898193, + "learning_rate": 0.00023512026090501427, + "loss": 3.5641, + "step": 90060 + }, + { + "epoch": 6.119377632830548, + "grad_norm": 0.8985006809234619, + "learning_rate": 0.00023507779589618155, + "loss": 3.4346, + "step": 90065 + }, + { + "epoch": 6.119717352901209, + "grad_norm": 0.9736484885215759, + "learning_rate": 0.00023503533088734883, + "loss": 3.25, + "step": 90070 + }, + { + "epoch": 6.120057072971871, + "grad_norm": 0.8587360382080078, + "learning_rate": 0.00023499286587851613, + "loss": 3.4738, + "step": 90075 + }, + { + "epoch": 6.120396793042533, + "grad_norm": 0.9343883395195007, + "learning_rate": 0.0002349504008696834, + "loss": 3.4002, + "step": 90080 + }, + { + "epoch": 6.1207365131131946, + "grad_norm": 1.0145918130874634, + "learning_rate": 0.00023490793586085067, + "loss": 3.1762, + "step": 90085 + }, + { + "epoch": 6.121076233183857, + "grad_norm": 0.7589127421379089, + "learning_rate": 0.00023486547085201795, + "loss": 3.3236, + "step": 90090 + }, + { + "epoch": 6.121415953254519, + "grad_norm": 0.8311502933502197, + "learning_rate": 0.00023482300584318523, + "loss": 3.4427, + "step": 90095 + }, + { + "epoch": 6.12175567332518, + "grad_norm": 0.8670511841773987, + "learning_rate": 0.00023478054083435248, + "loss": 3.4542, + "step": 90100 + }, + { + "epoch": 6.122095393395842, + "grad_norm": 0.9201338291168213, + "learning_rate": 0.0002347380758255198, + "loss": 3.4039, + "step": 90105 + }, + { + "epoch": 6.122435113466503, + "grad_norm": 1.0310992002487183, + "learning_rate": 0.00023469561081668707, + "loss": 3.2484, + "step": 90110 + }, + { + "epoch": 6.122774833537165, + "grad_norm": 0.9368970990180969, + "learning_rate": 0.00023465314580785432, + "loss": 3.4819, + "step": 90115 + }, + { + "epoch": 6.123114553607827, + "grad_norm": 1.1574337482452393, + "learning_rate": 0.00023461068079902163, + "loss": 3.4495, + "step": 90120 + }, + { + "epoch": 6.1234542736784885, + "grad_norm": 0.972122073173523, + "learning_rate": 0.00023456821579018888, + "loss": 3.2226, + "step": 90125 + }, + { + "epoch": 6.123793993749151, + "grad_norm": 0.8691845536231995, + "learning_rate": 0.00023452575078135616, + "loss": 3.4626, + "step": 90130 + }, + { + "epoch": 6.124133713819813, + "grad_norm": 0.9759337902069092, + "learning_rate": 0.00023448328577252344, + "loss": 3.2612, + "step": 90135 + }, + { + "epoch": 6.124473433890474, + "grad_norm": 0.9591606259346008, + "learning_rate": 0.00023444082076369072, + "loss": 3.5352, + "step": 90140 + }, + { + "epoch": 6.124813153961136, + "grad_norm": 0.805439293384552, + "learning_rate": 0.000234398355754858, + "loss": 3.459, + "step": 90145 + }, + { + "epoch": 6.125152874031798, + "grad_norm": 1.0369882583618164, + "learning_rate": 0.00023435589074602528, + "loss": 3.4972, + "step": 90150 + }, + { + "epoch": 6.125492594102459, + "grad_norm": 1.0262166261672974, + "learning_rate": 0.00023431342573719256, + "loss": 3.597, + "step": 90155 + }, + { + "epoch": 6.125832314173121, + "grad_norm": 1.150357723236084, + "learning_rate": 0.00023427096072835984, + "loss": 3.3248, + "step": 90160 + }, + { + "epoch": 6.126172034243783, + "grad_norm": 0.9391384720802307, + "learning_rate": 0.00023422849571952712, + "loss": 3.5338, + "step": 90165 + }, + { + "epoch": 6.1265117543144445, + "grad_norm": 1.0490968227386475, + "learning_rate": 0.00023418603071069437, + "loss": 3.4735, + "step": 90170 + }, + { + "epoch": 6.126851474385107, + "grad_norm": 0.9139017462730408, + "learning_rate": 0.00023414356570186168, + "loss": 3.4525, + "step": 90175 + }, + { + "epoch": 6.127191194455769, + "grad_norm": 0.9436922669410706, + "learning_rate": 0.00023410110069302896, + "loss": 3.293, + "step": 90180 + }, + { + "epoch": 6.12753091452643, + "grad_norm": 0.8887547254562378, + "learning_rate": 0.00023405863568419621, + "loss": 3.4136, + "step": 90185 + }, + { + "epoch": 6.127870634597092, + "grad_norm": 0.8381767272949219, + "learning_rate": 0.00023401617067536352, + "loss": 3.6038, + "step": 90190 + }, + { + "epoch": 6.128210354667754, + "grad_norm": 1.076701045036316, + "learning_rate": 0.00023397370566653077, + "loss": 3.6098, + "step": 90195 + }, + { + "epoch": 6.128550074738415, + "grad_norm": 0.7767021059989929, + "learning_rate": 0.00023393124065769805, + "loss": 3.712, + "step": 90200 + }, + { + "epoch": 6.128889794809077, + "grad_norm": 0.9297806620597839, + "learning_rate": 0.00023388877564886536, + "loss": 3.6333, + "step": 90205 + }, + { + "epoch": 6.129229514879739, + "grad_norm": 0.9899222254753113, + "learning_rate": 0.00023384631064003261, + "loss": 3.4069, + "step": 90210 + }, + { + "epoch": 6.1295692349504005, + "grad_norm": 0.7825217247009277, + "learning_rate": 0.0002338038456311999, + "loss": 3.3325, + "step": 90215 + }, + { + "epoch": 6.129908955021063, + "grad_norm": 0.7858249545097351, + "learning_rate": 0.00023376138062236717, + "loss": 3.204, + "step": 90220 + }, + { + "epoch": 6.130248675091725, + "grad_norm": 0.8473266959190369, + "learning_rate": 0.00023371891561353445, + "loss": 3.274, + "step": 90225 + }, + { + "epoch": 6.130588395162386, + "grad_norm": 0.9434612393379211, + "learning_rate": 0.0002336764506047017, + "loss": 3.1238, + "step": 90230 + }, + { + "epoch": 6.130928115233048, + "grad_norm": 0.8879968523979187, + "learning_rate": 0.00023363398559586902, + "loss": 3.3918, + "step": 90235 + }, + { + "epoch": 6.13126783530371, + "grad_norm": 0.9099522829055786, + "learning_rate": 0.0002335915205870363, + "loss": 3.5622, + "step": 90240 + }, + { + "epoch": 6.131607555374371, + "grad_norm": 0.7805164456367493, + "learning_rate": 0.00023354905557820358, + "loss": 3.4654, + "step": 90245 + }, + { + "epoch": 6.131947275445033, + "grad_norm": 1.0828574895858765, + "learning_rate": 0.00023350659056937086, + "loss": 3.1572, + "step": 90250 + }, + { + "epoch": 6.132286995515695, + "grad_norm": 0.8798744082450867, + "learning_rate": 0.0002334641255605381, + "loss": 3.3928, + "step": 90255 + }, + { + "epoch": 6.1326267155863565, + "grad_norm": 0.921140193939209, + "learning_rate": 0.00023342166055170542, + "loss": 3.5706, + "step": 90260 + }, + { + "epoch": 6.132966435657019, + "grad_norm": 0.97506183385849, + "learning_rate": 0.00023337919554287267, + "loss": 3.4805, + "step": 90265 + }, + { + "epoch": 6.133306155727681, + "grad_norm": 0.8897194266319275, + "learning_rate": 0.00023333673053403995, + "loss": 3.1114, + "step": 90270 + }, + { + "epoch": 6.133645875798342, + "grad_norm": 1.329687237739563, + "learning_rate": 0.00023329426552520726, + "loss": 3.6342, + "step": 90275 + }, + { + "epoch": 6.133985595869004, + "grad_norm": 0.9062702655792236, + "learning_rate": 0.0002332518005163745, + "loss": 3.6071, + "step": 90280 + }, + { + "epoch": 6.134325315939666, + "grad_norm": 1.3279105424880981, + "learning_rate": 0.0002332093355075418, + "loss": 3.682, + "step": 90285 + }, + { + "epoch": 6.134665036010327, + "grad_norm": 0.9675519466400146, + "learning_rate": 0.00023316687049870907, + "loss": 3.3997, + "step": 90290 + }, + { + "epoch": 6.135004756080989, + "grad_norm": 0.9161233901977539, + "learning_rate": 0.00023312440548987635, + "loss": 3.5569, + "step": 90295 + }, + { + "epoch": 6.135344476151651, + "grad_norm": 0.8284622430801392, + "learning_rate": 0.0002330819404810436, + "loss": 3.2154, + "step": 90300 + }, + { + "epoch": 6.1356841962223125, + "grad_norm": 0.8195561170578003, + "learning_rate": 0.0002330394754722109, + "loss": 3.4147, + "step": 90305 + }, + { + "epoch": 6.136023916292975, + "grad_norm": 1.296570062637329, + "learning_rate": 0.0002329970104633782, + "loss": 3.4302, + "step": 90310 + }, + { + "epoch": 6.136363636363637, + "grad_norm": 1.1554843187332153, + "learning_rate": 0.00023295454545454544, + "loss": 3.4028, + "step": 90315 + }, + { + "epoch": 6.136703356434298, + "grad_norm": 0.9567162990570068, + "learning_rate": 0.00023291208044571275, + "loss": 3.3419, + "step": 90320 + }, + { + "epoch": 6.13704307650496, + "grad_norm": 0.9426801204681396, + "learning_rate": 0.00023286961543688, + "loss": 3.2378, + "step": 90325 + }, + { + "epoch": 6.137382796575622, + "grad_norm": 0.8989987373352051, + "learning_rate": 0.0002328271504280473, + "loss": 3.454, + "step": 90330 + }, + { + "epoch": 6.137722516646283, + "grad_norm": 0.8077263236045837, + "learning_rate": 0.00023278468541921456, + "loss": 3.6107, + "step": 90335 + }, + { + "epoch": 6.138062236716945, + "grad_norm": 0.8330694437026978, + "learning_rate": 0.00023274222041038184, + "loss": 3.4252, + "step": 90340 + }, + { + "epoch": 6.138401956787607, + "grad_norm": 1.0547561645507812, + "learning_rate": 0.00023269975540154915, + "loss": 3.7351, + "step": 90345 + }, + { + "epoch": 6.1387416768582685, + "grad_norm": 0.9197625517845154, + "learning_rate": 0.0002326572903927164, + "loss": 3.5248, + "step": 90350 + }, + { + "epoch": 6.139081396928931, + "grad_norm": 1.0010108947753906, + "learning_rate": 0.00023261482538388368, + "loss": 3.4987, + "step": 90355 + }, + { + "epoch": 6.139421116999593, + "grad_norm": 1.081394910812378, + "learning_rate": 0.00023257236037505096, + "loss": 3.5094, + "step": 90360 + }, + { + "epoch": 6.139760837070254, + "grad_norm": 1.1292688846588135, + "learning_rate": 0.00023252989536621824, + "loss": 3.4228, + "step": 90365 + }, + { + "epoch": 6.140100557140916, + "grad_norm": 1.003387451171875, + "learning_rate": 0.00023248743035738552, + "loss": 3.3309, + "step": 90370 + }, + { + "epoch": 6.140440277211578, + "grad_norm": 1.052046537399292, + "learning_rate": 0.0002324449653485528, + "loss": 3.3096, + "step": 90375 + }, + { + "epoch": 6.140779997282239, + "grad_norm": 1.1048333644866943, + "learning_rate": 0.00023240250033972008, + "loss": 3.6364, + "step": 90380 + }, + { + "epoch": 6.141119717352901, + "grad_norm": 1.084534764289856, + "learning_rate": 0.00023236003533088734, + "loss": 3.5557, + "step": 90385 + }, + { + "epoch": 6.141459437423563, + "grad_norm": 0.9606860280036926, + "learning_rate": 0.00023231757032205464, + "loss": 3.4435, + "step": 90390 + }, + { + "epoch": 6.141799157494225, + "grad_norm": 1.3729310035705566, + "learning_rate": 0.0002322751053132219, + "loss": 3.2231, + "step": 90395 + }, + { + "epoch": 6.142138877564887, + "grad_norm": 1.0723402500152588, + "learning_rate": 0.00023223264030438918, + "loss": 3.4962, + "step": 90400 + }, + { + "epoch": 6.142478597635549, + "grad_norm": 0.8187320828437805, + "learning_rate": 0.00023219017529555648, + "loss": 3.5095, + "step": 90405 + }, + { + "epoch": 6.14281831770621, + "grad_norm": 0.9083629250526428, + "learning_rate": 0.00023214771028672374, + "loss": 3.3738, + "step": 90410 + }, + { + "epoch": 6.143158037776872, + "grad_norm": 1.0968647003173828, + "learning_rate": 0.00023210524527789104, + "loss": 3.6175, + "step": 90415 + }, + { + "epoch": 6.143497757847534, + "grad_norm": 0.8757907748222351, + "learning_rate": 0.0002320627802690583, + "loss": 3.4618, + "step": 90420 + }, + { + "epoch": 6.143837477918195, + "grad_norm": 0.8823025822639465, + "learning_rate": 0.00023202031526022558, + "loss": 3.4798, + "step": 90425 + }, + { + "epoch": 6.144177197988857, + "grad_norm": 0.7558712959289551, + "learning_rate": 0.00023197785025139286, + "loss": 3.3364, + "step": 90430 + }, + { + "epoch": 6.144516918059519, + "grad_norm": 0.7783344984054565, + "learning_rate": 0.00023193538524256014, + "loss": 3.4183, + "step": 90435 + }, + { + "epoch": 6.144856638130181, + "grad_norm": 0.958996057510376, + "learning_rate": 0.00023189292023372742, + "loss": 3.6831, + "step": 90440 + }, + { + "epoch": 6.145196358200843, + "grad_norm": 0.8586955666542053, + "learning_rate": 0.0002318504552248947, + "loss": 3.4982, + "step": 90445 + }, + { + "epoch": 6.145536078271505, + "grad_norm": 0.9256868958473206, + "learning_rate": 0.00023180799021606198, + "loss": 3.3451, + "step": 90450 + }, + { + "epoch": 6.145875798342166, + "grad_norm": 0.9507012963294983, + "learning_rate": 0.00023176552520722923, + "loss": 3.5928, + "step": 90455 + }, + { + "epoch": 6.146215518412828, + "grad_norm": 0.9384951591491699, + "learning_rate": 0.00023172306019839654, + "loss": 3.5885, + "step": 90460 + }, + { + "epoch": 6.14655523848349, + "grad_norm": 0.6857055425643921, + "learning_rate": 0.0002316805951895638, + "loss": 3.5401, + "step": 90465 + }, + { + "epoch": 6.146894958554151, + "grad_norm": 0.8285214304924011, + "learning_rate": 0.00023163813018073107, + "loss": 3.4532, + "step": 90470 + }, + { + "epoch": 6.147234678624813, + "grad_norm": 0.8974959254264832, + "learning_rate": 0.00023159566517189838, + "loss": 3.3727, + "step": 90475 + }, + { + "epoch": 6.1475743986954745, + "grad_norm": 1.0732169151306152, + "learning_rate": 0.00023155320016306563, + "loss": 3.3871, + "step": 90480 + }, + { + "epoch": 6.147914118766137, + "grad_norm": 1.0348073244094849, + "learning_rate": 0.0002315107351542329, + "loss": 3.6356, + "step": 90485 + }, + { + "epoch": 6.148253838836799, + "grad_norm": 1.392376184463501, + "learning_rate": 0.0002314682701454002, + "loss": 3.3328, + "step": 90490 + }, + { + "epoch": 6.14859355890746, + "grad_norm": 0.8257189393043518, + "learning_rate": 0.00023142580513656747, + "loss": 3.4451, + "step": 90495 + }, + { + "epoch": 6.148933278978122, + "grad_norm": 1.054093837738037, + "learning_rate": 0.00023138334012773478, + "loss": 3.3334, + "step": 90500 + }, + { + "epoch": 6.149272999048784, + "grad_norm": 1.1818761825561523, + "learning_rate": 0.00023134087511890203, + "loss": 3.4195, + "step": 90505 + }, + { + "epoch": 6.149612719119445, + "grad_norm": 0.8716172575950623, + "learning_rate": 0.0002312984101100693, + "loss": 3.259, + "step": 90510 + }, + { + "epoch": 6.149952439190107, + "grad_norm": 0.9813249111175537, + "learning_rate": 0.0002312559451012366, + "loss": 3.5279, + "step": 90515 + }, + { + "epoch": 6.150292159260769, + "grad_norm": 1.0199986696243286, + "learning_rate": 0.00023121348009240387, + "loss": 3.409, + "step": 90520 + }, + { + "epoch": 6.1506318793314305, + "grad_norm": 1.0102208852767944, + "learning_rate": 0.00023117101508357112, + "loss": 3.5675, + "step": 90525 + }, + { + "epoch": 6.150971599402093, + "grad_norm": 1.0130239725112915, + "learning_rate": 0.00023112855007473843, + "loss": 3.6191, + "step": 90530 + }, + { + "epoch": 6.151311319472755, + "grad_norm": 0.8974916934967041, + "learning_rate": 0.0002310860850659057, + "loss": 3.5496, + "step": 90535 + }, + { + "epoch": 6.151651039543416, + "grad_norm": 1.0729035139083862, + "learning_rate": 0.00023104362005707296, + "loss": 3.5772, + "step": 90540 + }, + { + "epoch": 6.151990759614078, + "grad_norm": 1.0161757469177246, + "learning_rate": 0.00023100115504824027, + "loss": 3.426, + "step": 90545 + }, + { + "epoch": 6.15233047968474, + "grad_norm": 1.1313872337341309, + "learning_rate": 0.00023095869003940752, + "loss": 3.4528, + "step": 90550 + }, + { + "epoch": 6.152670199755401, + "grad_norm": 1.1670204401016235, + "learning_rate": 0.0002309162250305748, + "loss": 3.5912, + "step": 90555 + }, + { + "epoch": 6.153009919826063, + "grad_norm": 0.9091718792915344, + "learning_rate": 0.00023087376002174208, + "loss": 3.499, + "step": 90560 + }, + { + "epoch": 6.153349639896725, + "grad_norm": 0.8918125629425049, + "learning_rate": 0.00023083129501290936, + "loss": 3.52, + "step": 90565 + }, + { + "epoch": 6.1536893599673865, + "grad_norm": 1.0750137567520142, + "learning_rate": 0.00023078883000407664, + "loss": 3.5275, + "step": 90570 + }, + { + "epoch": 6.154029080038049, + "grad_norm": 0.9216564297676086, + "learning_rate": 0.00023074636499524392, + "loss": 3.3549, + "step": 90575 + }, + { + "epoch": 6.154368800108711, + "grad_norm": 0.8564386367797852, + "learning_rate": 0.0002307038999864112, + "loss": 3.4522, + "step": 90580 + }, + { + "epoch": 6.154708520179372, + "grad_norm": 1.0826627016067505, + "learning_rate": 0.00023066143497757848, + "loss": 3.1514, + "step": 90585 + }, + { + "epoch": 6.155048240250034, + "grad_norm": 0.9433484673500061, + "learning_rate": 0.00023061896996874577, + "loss": 3.4562, + "step": 90590 + }, + { + "epoch": 6.155387960320696, + "grad_norm": 0.8191636800765991, + "learning_rate": 0.00023057650495991302, + "loss": 3.2048, + "step": 90595 + }, + { + "epoch": 6.155727680391357, + "grad_norm": 0.9451192617416382, + "learning_rate": 0.00023053403995108033, + "loss": 3.5608, + "step": 90600 + }, + { + "epoch": 6.156067400462019, + "grad_norm": 0.9147023558616638, + "learning_rate": 0.0002304915749422476, + "loss": 3.5174, + "step": 90605 + }, + { + "epoch": 6.156407120532681, + "grad_norm": 1.1863701343536377, + "learning_rate": 0.00023044910993341486, + "loss": 3.6168, + "step": 90610 + }, + { + "epoch": 6.1567468406033425, + "grad_norm": 0.7214667797088623, + "learning_rate": 0.00023040664492458217, + "loss": 3.3949, + "step": 90615 + }, + { + "epoch": 6.157086560674005, + "grad_norm": 0.8681305050849915, + "learning_rate": 0.00023036417991574942, + "loss": 3.6502, + "step": 90620 + }, + { + "epoch": 6.157426280744667, + "grad_norm": 0.8936669826507568, + "learning_rate": 0.0002303217149069167, + "loss": 3.466, + "step": 90625 + }, + { + "epoch": 6.157766000815328, + "grad_norm": 0.9390856027603149, + "learning_rate": 0.000230279249898084, + "loss": 3.4292, + "step": 90630 + }, + { + "epoch": 6.15810572088599, + "grad_norm": 1.0567810535430908, + "learning_rate": 0.00023023678488925126, + "loss": 3.3658, + "step": 90635 + }, + { + "epoch": 6.158445440956652, + "grad_norm": 0.8712486624717712, + "learning_rate": 0.00023019431988041854, + "loss": 3.3836, + "step": 90640 + }, + { + "epoch": 6.158785161027313, + "grad_norm": 0.8283782005310059, + "learning_rate": 0.00023015185487158582, + "loss": 3.2742, + "step": 90645 + }, + { + "epoch": 6.159124881097975, + "grad_norm": 0.8462200164794922, + "learning_rate": 0.0002301093898627531, + "loss": 3.5401, + "step": 90650 + }, + { + "epoch": 6.159464601168637, + "grad_norm": 0.7539302110671997, + "learning_rate": 0.00023006692485392035, + "loss": 3.3086, + "step": 90655 + }, + { + "epoch": 6.1598043212392986, + "grad_norm": 0.8114352822303772, + "learning_rate": 0.00023002445984508766, + "loss": 3.6745, + "step": 90660 + }, + { + "epoch": 6.160144041309961, + "grad_norm": 1.1245791912078857, + "learning_rate": 0.00022998199483625494, + "loss": 3.0953, + "step": 90665 + }, + { + "epoch": 6.160483761380623, + "grad_norm": 0.8448067307472229, + "learning_rate": 0.00022993952982742222, + "loss": 3.6617, + "step": 90670 + }, + { + "epoch": 6.160823481451284, + "grad_norm": 0.8309692144393921, + "learning_rate": 0.0002298970648185895, + "loss": 3.374, + "step": 90675 + }, + { + "epoch": 6.161163201521946, + "grad_norm": 0.9471685290336609, + "learning_rate": 0.00022985459980975675, + "loss": 3.7397, + "step": 90680 + }, + { + "epoch": 6.161502921592608, + "grad_norm": 1.2656900882720947, + "learning_rate": 0.00022981213480092406, + "loss": 3.5527, + "step": 90685 + }, + { + "epoch": 6.161842641663269, + "grad_norm": 0.9192420840263367, + "learning_rate": 0.0002297696697920913, + "loss": 3.3707, + "step": 90690 + }, + { + "epoch": 6.162182361733931, + "grad_norm": 0.9169350266456604, + "learning_rate": 0.0002297272047832586, + "loss": 3.6093, + "step": 90695 + }, + { + "epoch": 6.162522081804593, + "grad_norm": 0.900912344455719, + "learning_rate": 0.0002296847397744259, + "loss": 3.3636, + "step": 90700 + }, + { + "epoch": 6.162861801875255, + "grad_norm": 0.8392032980918884, + "learning_rate": 0.00022964227476559315, + "loss": 3.6349, + "step": 90705 + }, + { + "epoch": 6.163201521945917, + "grad_norm": 0.9424541592597961, + "learning_rate": 0.00022959980975676043, + "loss": 3.4648, + "step": 90710 + }, + { + "epoch": 6.163541242016579, + "grad_norm": 0.8518842458724976, + "learning_rate": 0.0002295573447479277, + "loss": 3.4512, + "step": 90715 + }, + { + "epoch": 6.16388096208724, + "grad_norm": 0.9782639145851135, + "learning_rate": 0.000229514879739095, + "loss": 3.3613, + "step": 90720 + }, + { + "epoch": 6.164220682157902, + "grad_norm": 1.1959002017974854, + "learning_rate": 0.00022947241473026225, + "loss": 3.4911, + "step": 90725 + }, + { + "epoch": 6.164560402228564, + "grad_norm": 1.138426661491394, + "learning_rate": 0.00022942994972142955, + "loss": 3.4599, + "step": 90730 + }, + { + "epoch": 6.164900122299225, + "grad_norm": 0.7962737679481506, + "learning_rate": 0.00022938748471259683, + "loss": 3.5655, + "step": 90735 + }, + { + "epoch": 6.165239842369887, + "grad_norm": 1.031675934791565, + "learning_rate": 0.00022934501970376409, + "loss": 3.4642, + "step": 90740 + }, + { + "epoch": 6.165579562440549, + "grad_norm": 1.185836911201477, + "learning_rate": 0.0002293025546949314, + "loss": 3.434, + "step": 90745 + }, + { + "epoch": 6.165919282511211, + "grad_norm": 1.2029790878295898, + "learning_rate": 0.00022926008968609865, + "loss": 3.4586, + "step": 90750 + }, + { + "epoch": 6.166259002581873, + "grad_norm": 0.7992639541625977, + "learning_rate": 0.00022921762467726595, + "loss": 3.4228, + "step": 90755 + }, + { + "epoch": 6.166598722652535, + "grad_norm": 0.7933395504951477, + "learning_rate": 0.0002291751596684332, + "loss": 3.192, + "step": 90760 + }, + { + "epoch": 6.166938442723196, + "grad_norm": 0.8864936232566833, + "learning_rate": 0.0002291326946596005, + "loss": 3.2854, + "step": 90765 + }, + { + "epoch": 6.167278162793858, + "grad_norm": 0.9408644437789917, + "learning_rate": 0.0002290902296507678, + "loss": 3.3999, + "step": 90770 + }, + { + "epoch": 6.16761788286452, + "grad_norm": 0.9514845013618469, + "learning_rate": 0.00022904776464193505, + "loss": 3.3415, + "step": 90775 + }, + { + "epoch": 6.167957602935181, + "grad_norm": 1.0231291055679321, + "learning_rate": 0.00022900529963310233, + "loss": 3.6515, + "step": 90780 + }, + { + "epoch": 6.168297323005843, + "grad_norm": 0.638744056224823, + "learning_rate": 0.0002289628346242696, + "loss": 3.4954, + "step": 90785 + }, + { + "epoch": 6.1686370430765045, + "grad_norm": 1.0557352304458618, + "learning_rate": 0.0002289203696154369, + "loss": 3.7117, + "step": 90790 + }, + { + "epoch": 6.168976763147167, + "grad_norm": 1.7130573987960815, + "learning_rate": 0.00022887790460660417, + "loss": 3.4191, + "step": 90795 + }, + { + "epoch": 6.169316483217829, + "grad_norm": 0.8598589897155762, + "learning_rate": 0.00022883543959777145, + "loss": 3.4111, + "step": 90800 + }, + { + "epoch": 6.16965620328849, + "grad_norm": 0.6965853571891785, + "learning_rate": 0.00022879297458893873, + "loss": 3.3909, + "step": 90805 + }, + { + "epoch": 6.169995923359152, + "grad_norm": 1.0191088914871216, + "learning_rate": 0.00022875050958010598, + "loss": 3.6145, + "step": 90810 + }, + { + "epoch": 6.170335643429814, + "grad_norm": 0.8858039975166321, + "learning_rate": 0.0002287080445712733, + "loss": 3.5533, + "step": 90815 + }, + { + "epoch": 6.170675363500475, + "grad_norm": 0.9053271412849426, + "learning_rate": 0.00022866557956244054, + "loss": 3.4141, + "step": 90820 + }, + { + "epoch": 6.171015083571137, + "grad_norm": 0.7783395051956177, + "learning_rate": 0.00022862311455360782, + "loss": 3.2244, + "step": 90825 + }, + { + "epoch": 6.171354803641799, + "grad_norm": 0.8646712899208069, + "learning_rate": 0.00022858064954477513, + "loss": 3.3025, + "step": 90830 + }, + { + "epoch": 6.1716945237124605, + "grad_norm": 1.0673201084136963, + "learning_rate": 0.00022853818453594238, + "loss": 3.314, + "step": 90835 + }, + { + "epoch": 6.172034243783123, + "grad_norm": 1.1085801124572754, + "learning_rate": 0.0002284957195271097, + "loss": 3.4229, + "step": 90840 + }, + { + "epoch": 6.172373963853785, + "grad_norm": 0.9971343874931335, + "learning_rate": 0.00022845325451827694, + "loss": 3.2454, + "step": 90845 + }, + { + "epoch": 6.172713683924446, + "grad_norm": 1.0795583724975586, + "learning_rate": 0.00022841078950944422, + "loss": 3.3957, + "step": 90850 + }, + { + "epoch": 6.173053403995108, + "grad_norm": 1.0345350503921509, + "learning_rate": 0.0002283683245006115, + "loss": 3.4007, + "step": 90855 + }, + { + "epoch": 6.17339312406577, + "grad_norm": 0.8854944705963135, + "learning_rate": 0.00022832585949177878, + "loss": 3.2382, + "step": 90860 + }, + { + "epoch": 6.173732844136431, + "grad_norm": 0.889509379863739, + "learning_rate": 0.00022828339448294606, + "loss": 3.4371, + "step": 90865 + }, + { + "epoch": 6.174072564207093, + "grad_norm": 0.8219030499458313, + "learning_rate": 0.00022824092947411334, + "loss": 3.3391, + "step": 90870 + }, + { + "epoch": 6.174412284277755, + "grad_norm": 0.9163932204246521, + "learning_rate": 0.00022819846446528062, + "loss": 3.0954, + "step": 90875 + }, + { + "epoch": 6.1747520043484165, + "grad_norm": 0.8960126042366028, + "learning_rate": 0.00022815599945644787, + "loss": 3.4239, + "step": 90880 + }, + { + "epoch": 6.175091724419079, + "grad_norm": 1.0875799655914307, + "learning_rate": 0.00022811353444761518, + "loss": 3.4378, + "step": 90885 + }, + { + "epoch": 6.175431444489741, + "grad_norm": 0.8141494393348694, + "learning_rate": 0.00022807106943878243, + "loss": 3.3901, + "step": 90890 + }, + { + "epoch": 6.175771164560402, + "grad_norm": 0.9403805136680603, + "learning_rate": 0.00022802860442994971, + "loss": 3.4544, + "step": 90895 + }, + { + "epoch": 6.176110884631064, + "grad_norm": 1.0195300579071045, + "learning_rate": 0.00022798613942111702, + "loss": 3.3991, + "step": 90900 + }, + { + "epoch": 6.176450604701726, + "grad_norm": 0.9390950202941895, + "learning_rate": 0.00022794367441228427, + "loss": 3.2223, + "step": 90905 + }, + { + "epoch": 6.176790324772387, + "grad_norm": 0.7327580451965332, + "learning_rate": 0.00022790120940345155, + "loss": 3.355, + "step": 90910 + }, + { + "epoch": 6.177130044843049, + "grad_norm": 0.837429404258728, + "learning_rate": 0.00022785874439461883, + "loss": 3.3359, + "step": 90915 + }, + { + "epoch": 6.177469764913711, + "grad_norm": 1.0164953470230103, + "learning_rate": 0.00022781627938578611, + "loss": 3.4653, + "step": 90920 + }, + { + "epoch": 6.1778094849843725, + "grad_norm": 0.9160313606262207, + "learning_rate": 0.00022777381437695342, + "loss": 3.3909, + "step": 90925 + }, + { + "epoch": 6.178149205055035, + "grad_norm": 0.8425294160842896, + "learning_rate": 0.00022773134936812067, + "loss": 3.5108, + "step": 90930 + }, + { + "epoch": 6.178488925125697, + "grad_norm": 1.0116581916809082, + "learning_rate": 0.00022768888435928795, + "loss": 3.5183, + "step": 90935 + }, + { + "epoch": 6.178828645196358, + "grad_norm": 0.7998234629631042, + "learning_rate": 0.00022764641935045523, + "loss": 3.4933, + "step": 90940 + }, + { + "epoch": 6.17916836526702, + "grad_norm": 1.0491366386413574, + "learning_rate": 0.00022760395434162252, + "loss": 3.3222, + "step": 90945 + }, + { + "epoch": 6.179508085337682, + "grad_norm": 0.9984050393104553, + "learning_rate": 0.00022756148933278977, + "loss": 3.2522, + "step": 90950 + }, + { + "epoch": 6.179847805408343, + "grad_norm": 1.2336381673812866, + "learning_rate": 0.00022751902432395708, + "loss": 3.35, + "step": 90955 + }, + { + "epoch": 6.180187525479005, + "grad_norm": 0.8273404836654663, + "learning_rate": 0.00022747655931512436, + "loss": 3.3811, + "step": 90960 + }, + { + "epoch": 6.180527245549667, + "grad_norm": 1.0590180158615112, + "learning_rate": 0.0002274340943062916, + "loss": 3.312, + "step": 90965 + }, + { + "epoch": 6.180866965620329, + "grad_norm": 1.165709137916565, + "learning_rate": 0.00022739162929745892, + "loss": 3.6538, + "step": 90970 + }, + { + "epoch": 6.181206685690991, + "grad_norm": 0.898287296295166, + "learning_rate": 0.00022734916428862617, + "loss": 3.4789, + "step": 90975 + }, + { + "epoch": 6.181546405761653, + "grad_norm": 0.97523033618927, + "learning_rate": 0.00022730669927979345, + "loss": 3.7121, + "step": 90980 + }, + { + "epoch": 6.181886125832314, + "grad_norm": 0.8981620073318481, + "learning_rate": 0.00022726423427096073, + "loss": 3.1741, + "step": 90985 + }, + { + "epoch": 6.182225845902976, + "grad_norm": 0.9335720539093018, + "learning_rate": 0.000227221769262128, + "loss": 3.5065, + "step": 90990 + }, + { + "epoch": 6.182565565973638, + "grad_norm": 0.9235891699790955, + "learning_rate": 0.0002271793042532953, + "loss": 3.499, + "step": 90995 + }, + { + "epoch": 6.182905286044299, + "grad_norm": 1.0540515184402466, + "learning_rate": 0.00022713683924446257, + "loss": 3.6062, + "step": 91000 + }, + { + "epoch": 6.183245006114961, + "grad_norm": 0.8692999482154846, + "learning_rate": 0.00022709437423562985, + "loss": 3.3175, + "step": 91005 + }, + { + "epoch": 6.183584726185623, + "grad_norm": 0.9033063650131226, + "learning_rate": 0.00022705190922679713, + "loss": 3.3768, + "step": 91010 + }, + { + "epoch": 6.183924446256285, + "grad_norm": 0.8082554340362549, + "learning_rate": 0.0002270094442179644, + "loss": 3.516, + "step": 91015 + }, + { + "epoch": 6.184264166326947, + "grad_norm": 0.7425940632820129, + "learning_rate": 0.00022696697920913166, + "loss": 3.2641, + "step": 91020 + }, + { + "epoch": 6.184603886397609, + "grad_norm": 0.9870772957801819, + "learning_rate": 0.00022692451420029897, + "loss": 3.5529, + "step": 91025 + }, + { + "epoch": 6.18494360646827, + "grad_norm": 1.1142958402633667, + "learning_rate": 0.00022688204919146625, + "loss": 3.3655, + "step": 91030 + }, + { + "epoch": 6.185283326538932, + "grad_norm": 0.839480459690094, + "learning_rate": 0.0002268395841826335, + "loss": 3.3542, + "step": 91035 + }, + { + "epoch": 6.185623046609594, + "grad_norm": 0.8901545405387878, + "learning_rate": 0.0002267971191738008, + "loss": 3.2371, + "step": 91040 + }, + { + "epoch": 6.185962766680255, + "grad_norm": 0.8993017673492432, + "learning_rate": 0.00022675465416496806, + "loss": 3.4604, + "step": 91045 + }, + { + "epoch": 6.186302486750917, + "grad_norm": 0.6864322423934937, + "learning_rate": 0.00022671218915613534, + "loss": 3.6256, + "step": 91050 + }, + { + "epoch": 6.186642206821579, + "grad_norm": 0.9046564102172852, + "learning_rate": 0.00022666972414730265, + "loss": 3.7288, + "step": 91055 + }, + { + "epoch": 6.186981926892241, + "grad_norm": 0.8494551777839661, + "learning_rate": 0.0002266272591384699, + "loss": 3.1177, + "step": 91060 + }, + { + "epoch": 6.187321646962903, + "grad_norm": 1.1661748886108398, + "learning_rate": 0.00022658479412963718, + "loss": 3.5509, + "step": 91065 + }, + { + "epoch": 6.187661367033565, + "grad_norm": 0.7532400488853455, + "learning_rate": 0.00022654232912080446, + "loss": 3.3007, + "step": 91070 + }, + { + "epoch": 6.188001087104226, + "grad_norm": 0.9591653347015381, + "learning_rate": 0.00022649986411197174, + "loss": 3.3345, + "step": 91075 + }, + { + "epoch": 6.188340807174888, + "grad_norm": 0.7847570776939392, + "learning_rate": 0.000226457399103139, + "loss": 3.4716, + "step": 91080 + }, + { + "epoch": 6.18868052724555, + "grad_norm": 1.1805249452590942, + "learning_rate": 0.0002264149340943063, + "loss": 3.3652, + "step": 91085 + }, + { + "epoch": 6.189020247316211, + "grad_norm": 0.9418165683746338, + "learning_rate": 0.00022637246908547358, + "loss": 3.5069, + "step": 91090 + }, + { + "epoch": 6.189359967386873, + "grad_norm": 0.7655673027038574, + "learning_rate": 0.00022633000407664086, + "loss": 3.6229, + "step": 91095 + }, + { + "epoch": 6.189699687457535, + "grad_norm": 1.2213689088821411, + "learning_rate": 0.00022628753906780814, + "loss": 3.3499, + "step": 91100 + }, + { + "epoch": 6.190039407528197, + "grad_norm": 0.9939951300621033, + "learning_rate": 0.0002262450740589754, + "loss": 3.3511, + "step": 91105 + }, + { + "epoch": 6.190379127598859, + "grad_norm": 0.7482156753540039, + "learning_rate": 0.0002262026090501427, + "loss": 3.6355, + "step": 91110 + }, + { + "epoch": 6.190718847669521, + "grad_norm": 0.9241912961006165, + "learning_rate": 0.00022616014404130996, + "loss": 3.6398, + "step": 91115 + }, + { + "epoch": 6.191058567740182, + "grad_norm": 1.2100588083267212, + "learning_rate": 0.00022611767903247724, + "loss": 3.522, + "step": 91120 + }, + { + "epoch": 6.191398287810844, + "grad_norm": 1.1023330688476562, + "learning_rate": 0.00022607521402364454, + "loss": 3.4859, + "step": 91125 + }, + { + "epoch": 6.191738007881506, + "grad_norm": 0.9313455820083618, + "learning_rate": 0.0002260327490148118, + "loss": 3.2776, + "step": 91130 + }, + { + "epoch": 6.192077727952167, + "grad_norm": 1.4443272352218628, + "learning_rate": 0.00022599028400597908, + "loss": 3.6161, + "step": 91135 + }, + { + "epoch": 6.192417448022829, + "grad_norm": 0.8198590874671936, + "learning_rate": 0.00022594781899714636, + "loss": 3.6906, + "step": 91140 + }, + { + "epoch": 6.192757168093491, + "grad_norm": 1.0384960174560547, + "learning_rate": 0.00022590535398831364, + "loss": 3.0457, + "step": 91145 + }, + { + "epoch": 6.193096888164153, + "grad_norm": 0.7394150495529175, + "learning_rate": 0.0002258628889794809, + "loss": 3.3104, + "step": 91150 + }, + { + "epoch": 6.193436608234815, + "grad_norm": 0.9200577139854431, + "learning_rate": 0.0002258204239706482, + "loss": 3.5112, + "step": 91155 + }, + { + "epoch": 6.193776328305476, + "grad_norm": 0.9335869550704956, + "learning_rate": 0.00022577795896181548, + "loss": 3.4409, + "step": 91160 + }, + { + "epoch": 6.194116048376138, + "grad_norm": 0.7920849323272705, + "learning_rate": 0.00022573549395298273, + "loss": 3.3839, + "step": 91165 + }, + { + "epoch": 6.1944557684468, + "grad_norm": 0.8127315044403076, + "learning_rate": 0.00022569302894415004, + "loss": 3.5069, + "step": 91170 + }, + { + "epoch": 6.194795488517461, + "grad_norm": 1.008046269416809, + "learning_rate": 0.0002256505639353173, + "loss": 3.3093, + "step": 91175 + }, + { + "epoch": 6.195135208588123, + "grad_norm": 0.9313398599624634, + "learning_rate": 0.0002256080989264846, + "loss": 3.6415, + "step": 91180 + }, + { + "epoch": 6.195474928658785, + "grad_norm": 1.0969114303588867, + "learning_rate": 0.00022556563391765185, + "loss": 3.4093, + "step": 91185 + }, + { + "epoch": 6.1958146487294465, + "grad_norm": 0.8970260620117188, + "learning_rate": 0.00022552316890881913, + "loss": 3.4278, + "step": 91190 + }, + { + "epoch": 6.196154368800109, + "grad_norm": 0.8236124515533447, + "learning_rate": 0.00022548070389998644, + "loss": 3.4042, + "step": 91195 + }, + { + "epoch": 6.196494088870771, + "grad_norm": 1.1214083433151245, + "learning_rate": 0.0002254382388911537, + "loss": 3.6678, + "step": 91200 + }, + { + "epoch": 6.196833808941432, + "grad_norm": 0.8696040511131287, + "learning_rate": 0.00022539577388232097, + "loss": 3.3446, + "step": 91205 + }, + { + "epoch": 6.197173529012094, + "grad_norm": 1.143045425415039, + "learning_rate": 0.00022535330887348825, + "loss": 3.2963, + "step": 91210 + }, + { + "epoch": 6.197513249082756, + "grad_norm": 0.8145077228546143, + "learning_rate": 0.00022531084386465553, + "loss": 3.4541, + "step": 91215 + }, + { + "epoch": 6.197852969153417, + "grad_norm": 0.9721153378486633, + "learning_rate": 0.0002252683788558228, + "loss": 3.2641, + "step": 91220 + }, + { + "epoch": 6.198192689224079, + "grad_norm": 0.8590301275253296, + "learning_rate": 0.0002252259138469901, + "loss": 3.5149, + "step": 91225 + }, + { + "epoch": 6.198532409294741, + "grad_norm": 0.9187374114990234, + "learning_rate": 0.00022518344883815737, + "loss": 3.4783, + "step": 91230 + }, + { + "epoch": 6.1988721293654026, + "grad_norm": 1.1660041809082031, + "learning_rate": 0.00022514098382932462, + "loss": 3.3745, + "step": 91235 + }, + { + "epoch": 6.199211849436065, + "grad_norm": 0.9120582938194275, + "learning_rate": 0.00022509851882049193, + "loss": 3.4491, + "step": 91240 + }, + { + "epoch": 6.199551569506727, + "grad_norm": 0.9263826012611389, + "learning_rate": 0.00022505605381165918, + "loss": 3.4554, + "step": 91245 + }, + { + "epoch": 6.199891289577388, + "grad_norm": 0.9913333058357239, + "learning_rate": 0.00022501358880282646, + "loss": 3.4298, + "step": 91250 + }, + { + "epoch": 6.20023100964805, + "grad_norm": 0.9374046921730042, + "learning_rate": 0.00022497112379399377, + "loss": 3.1481, + "step": 91255 + }, + { + "epoch": 6.200570729718712, + "grad_norm": 0.9598096609115601, + "learning_rate": 0.00022492865878516102, + "loss": 3.2823, + "step": 91260 + }, + { + "epoch": 6.200910449789373, + "grad_norm": 0.8446417450904846, + "learning_rate": 0.00022488619377632833, + "loss": 3.4577, + "step": 91265 + }, + { + "epoch": 6.201250169860035, + "grad_norm": 1.1102107763290405, + "learning_rate": 0.00022484372876749558, + "loss": 3.3218, + "step": 91270 + }, + { + "epoch": 6.201589889930697, + "grad_norm": 0.9262752532958984, + "learning_rate": 0.00022480126375866286, + "loss": 3.3688, + "step": 91275 + }, + { + "epoch": 6.201929610001359, + "grad_norm": 0.7228497862815857, + "learning_rate": 0.00022475879874983014, + "loss": 3.4714, + "step": 91280 + }, + { + "epoch": 6.202269330072021, + "grad_norm": 1.0269438028335571, + "learning_rate": 0.00022471633374099742, + "loss": 3.147, + "step": 91285 + }, + { + "epoch": 6.202609050142683, + "grad_norm": 0.8682957291603088, + "learning_rate": 0.0002246738687321647, + "loss": 3.4442, + "step": 91290 + }, + { + "epoch": 6.202948770213344, + "grad_norm": 0.9754287600517273, + "learning_rate": 0.00022463140372333198, + "loss": 3.6056, + "step": 91295 + }, + { + "epoch": 6.203288490284006, + "grad_norm": 1.0042673349380493, + "learning_rate": 0.00022458893871449926, + "loss": 2.9748, + "step": 91300 + }, + { + "epoch": 6.203628210354668, + "grad_norm": 0.8220917582511902, + "learning_rate": 0.00022454647370566652, + "loss": 3.4315, + "step": 91305 + }, + { + "epoch": 6.203967930425329, + "grad_norm": 0.8285524249076843, + "learning_rate": 0.00022450400869683383, + "loss": 3.3342, + "step": 91310 + }, + { + "epoch": 6.204307650495991, + "grad_norm": 0.8249390125274658, + "learning_rate": 0.00022446154368800108, + "loss": 3.2956, + "step": 91315 + }, + { + "epoch": 6.204647370566653, + "grad_norm": 1.0120559930801392, + "learning_rate": 0.00022441907867916836, + "loss": 3.5639, + "step": 91320 + }, + { + "epoch": 6.204987090637315, + "grad_norm": 1.2241082191467285, + "learning_rate": 0.00022437661367033567, + "loss": 3.2062, + "step": 91325 + }, + { + "epoch": 6.205326810707977, + "grad_norm": 0.7561842203140259, + "learning_rate": 0.00022433414866150292, + "loss": 3.4201, + "step": 91330 + }, + { + "epoch": 6.205666530778639, + "grad_norm": 0.8201459050178528, + "learning_rate": 0.0002242916836526702, + "loss": 3.458, + "step": 91335 + }, + { + "epoch": 6.2060062508493, + "grad_norm": 0.7808039784431458, + "learning_rate": 0.00022424921864383748, + "loss": 3.3441, + "step": 91340 + }, + { + "epoch": 6.206345970919962, + "grad_norm": 1.287811517715454, + "learning_rate": 0.00022420675363500476, + "loss": 3.4764, + "step": 91345 + }, + { + "epoch": 6.206685690990624, + "grad_norm": 1.1327370405197144, + "learning_rate": 0.00022416428862617207, + "loss": 3.1685, + "step": 91350 + }, + { + "epoch": 6.207025411061285, + "grad_norm": 1.134243130683899, + "learning_rate": 0.00022412182361733932, + "loss": 3.3701, + "step": 91355 + }, + { + "epoch": 6.207365131131947, + "grad_norm": 0.8373261094093323, + "learning_rate": 0.0002240793586085066, + "loss": 3.3237, + "step": 91360 + }, + { + "epoch": 6.207704851202609, + "grad_norm": 1.1181309223175049, + "learning_rate": 0.00022403689359967388, + "loss": 3.3053, + "step": 91365 + }, + { + "epoch": 6.208044571273271, + "grad_norm": 1.0359162092208862, + "learning_rate": 0.00022399442859084116, + "loss": 3.6289, + "step": 91370 + }, + { + "epoch": 6.208384291343933, + "grad_norm": 0.8373446464538574, + "learning_rate": 0.0002239519635820084, + "loss": 3.6685, + "step": 91375 + }, + { + "epoch": 6.208724011414595, + "grad_norm": 0.9910107851028442, + "learning_rate": 0.00022390949857317572, + "loss": 3.4436, + "step": 91380 + }, + { + "epoch": 6.209063731485256, + "grad_norm": 0.8144149780273438, + "learning_rate": 0.000223867033564343, + "loss": 3.3888, + "step": 91385 + }, + { + "epoch": 6.209403451555918, + "grad_norm": 0.9367064833641052, + "learning_rate": 0.00022382456855551025, + "loss": 3.3902, + "step": 91390 + }, + { + "epoch": 6.20974317162658, + "grad_norm": 0.9591310620307922, + "learning_rate": 0.00022378210354667756, + "loss": 3.3647, + "step": 91395 + }, + { + "epoch": 6.210082891697241, + "grad_norm": 0.8278966546058655, + "learning_rate": 0.0002237396385378448, + "loss": 3.3412, + "step": 91400 + }, + { + "epoch": 6.210422611767903, + "grad_norm": 0.7016304731369019, + "learning_rate": 0.0002236971735290121, + "loss": 3.4699, + "step": 91405 + }, + { + "epoch": 6.210762331838565, + "grad_norm": 0.8163431882858276, + "learning_rate": 0.00022365470852017937, + "loss": 3.2997, + "step": 91410 + }, + { + "epoch": 6.211102051909227, + "grad_norm": 1.4049125909805298, + "learning_rate": 0.00022361224351134665, + "loss": 3.4555, + "step": 91415 + }, + { + "epoch": 6.211441771979889, + "grad_norm": 0.896393358707428, + "learning_rate": 0.00022356977850251393, + "loss": 3.5206, + "step": 91420 + }, + { + "epoch": 6.211781492050551, + "grad_norm": 0.8502756953239441, + "learning_rate": 0.0002235273134936812, + "loss": 3.4468, + "step": 91425 + }, + { + "epoch": 6.212121212121212, + "grad_norm": 1.234155535697937, + "learning_rate": 0.0002234848484848485, + "loss": 3.2751, + "step": 91430 + }, + { + "epoch": 6.212460932191874, + "grad_norm": 0.8009254932403564, + "learning_rate": 0.00022344238347601577, + "loss": 3.579, + "step": 91435 + }, + { + "epoch": 6.212800652262536, + "grad_norm": 0.8655442595481873, + "learning_rate": 0.00022339991846718305, + "loss": 3.6737, + "step": 91440 + }, + { + "epoch": 6.213140372333197, + "grad_norm": 0.8145853281021118, + "learning_rate": 0.0002233574534583503, + "loss": 3.45, + "step": 91445 + }, + { + "epoch": 6.213480092403859, + "grad_norm": 0.9187182188034058, + "learning_rate": 0.0002233149884495176, + "loss": 3.4603, + "step": 91450 + }, + { + "epoch": 6.213819812474521, + "grad_norm": 1.1248862743377686, + "learning_rate": 0.0002232725234406849, + "loss": 3.5702, + "step": 91455 + }, + { + "epoch": 6.214159532545183, + "grad_norm": 0.7840197682380676, + "learning_rate": 0.00022323005843185215, + "loss": 3.6902, + "step": 91460 + }, + { + "epoch": 6.214499252615845, + "grad_norm": 0.900150716304779, + "learning_rate": 0.00022318759342301945, + "loss": 3.4504, + "step": 91465 + }, + { + "epoch": 6.214838972686506, + "grad_norm": 1.0406229496002197, + "learning_rate": 0.0002231451284141867, + "loss": 3.3143, + "step": 91470 + }, + { + "epoch": 6.215178692757168, + "grad_norm": 0.8114564418792725, + "learning_rate": 0.00022310266340535399, + "loss": 3.3584, + "step": 91475 + }, + { + "epoch": 6.21551841282783, + "grad_norm": 0.8749794960021973, + "learning_rate": 0.00022306019839652127, + "loss": 3.2073, + "step": 91480 + }, + { + "epoch": 6.215858132898491, + "grad_norm": 1.0953396558761597, + "learning_rate": 0.00022301773338768855, + "loss": 3.4483, + "step": 91485 + }, + { + "epoch": 6.216197852969153, + "grad_norm": 1.0054417848587036, + "learning_rate": 0.00022297526837885583, + "loss": 3.6976, + "step": 91490 + }, + { + "epoch": 6.216537573039815, + "grad_norm": 0.8592808842658997, + "learning_rate": 0.0002229328033700231, + "loss": 3.5867, + "step": 91495 + }, + { + "epoch": 6.2168772931104765, + "grad_norm": 0.8345815539360046, + "learning_rate": 0.0002228903383611904, + "loss": 3.4214, + "step": 91500 + }, + { + "epoch": 6.217217013181139, + "grad_norm": 0.8264532089233398, + "learning_rate": 0.00022284787335235764, + "loss": 3.384, + "step": 91505 + }, + { + "epoch": 6.217556733251801, + "grad_norm": 1.0752284526824951, + "learning_rate": 0.00022280540834352495, + "loss": 3.1018, + "step": 91510 + }, + { + "epoch": 6.217896453322462, + "grad_norm": 0.9747252464294434, + "learning_rate": 0.00022276294333469223, + "loss": 3.1022, + "step": 91515 + }, + { + "epoch": 6.218236173393124, + "grad_norm": 0.9268384575843811, + "learning_rate": 0.0002227204783258595, + "loss": 3.1776, + "step": 91520 + }, + { + "epoch": 6.218575893463786, + "grad_norm": 0.9454618692398071, + "learning_rate": 0.0002226780133170268, + "loss": 3.5255, + "step": 91525 + }, + { + "epoch": 6.218915613534447, + "grad_norm": 0.9710923433303833, + "learning_rate": 0.00022263554830819404, + "loss": 3.4867, + "step": 91530 + }, + { + "epoch": 6.219255333605109, + "grad_norm": 0.9882373809814453, + "learning_rate": 0.00022259308329936135, + "loss": 3.2497, + "step": 91535 + }, + { + "epoch": 6.219595053675771, + "grad_norm": 0.88743656873703, + "learning_rate": 0.0002225506182905286, + "loss": 3.5218, + "step": 91540 + }, + { + "epoch": 6.2199347737464326, + "grad_norm": 1.0199642181396484, + "learning_rate": 0.00022250815328169588, + "loss": 3.3273, + "step": 91545 + }, + { + "epoch": 6.220274493817095, + "grad_norm": 0.8071062564849854, + "learning_rate": 0.0002224656882728632, + "loss": 3.3973, + "step": 91550 + }, + { + "epoch": 6.220614213887757, + "grad_norm": 0.9354631900787354, + "learning_rate": 0.00022242322326403044, + "loss": 3.5713, + "step": 91555 + }, + { + "epoch": 6.220953933958418, + "grad_norm": 0.8057290315628052, + "learning_rate": 0.00022238075825519772, + "loss": 3.4866, + "step": 91560 + }, + { + "epoch": 6.22129365402908, + "grad_norm": 0.9645929336547852, + "learning_rate": 0.000222338293246365, + "loss": 3.4213, + "step": 91565 + }, + { + "epoch": 6.221633374099742, + "grad_norm": 0.9268535375595093, + "learning_rate": 0.00022229582823753228, + "loss": 3.3116, + "step": 91570 + }, + { + "epoch": 6.221973094170403, + "grad_norm": 0.7922458648681641, + "learning_rate": 0.00022225336322869953, + "loss": 3.3985, + "step": 91575 + }, + { + "epoch": 6.222312814241065, + "grad_norm": 1.0156108140945435, + "learning_rate": 0.00022221089821986684, + "loss": 3.5712, + "step": 91580 + }, + { + "epoch": 6.222652534311727, + "grad_norm": 0.9524393081665039, + "learning_rate": 0.00022216843321103412, + "loss": 3.3631, + "step": 91585 + }, + { + "epoch": 6.222992254382389, + "grad_norm": 0.7943010330200195, + "learning_rate": 0.00022212596820220137, + "loss": 3.3095, + "step": 91590 + }, + { + "epoch": 6.223331974453051, + "grad_norm": 0.7677335739135742, + "learning_rate": 0.00022208350319336868, + "loss": 3.3102, + "step": 91595 + }, + { + "epoch": 6.223671694523713, + "grad_norm": 1.0518152713775635, + "learning_rate": 0.00022204103818453593, + "loss": 3.4361, + "step": 91600 + }, + { + "epoch": 6.224011414594374, + "grad_norm": 0.9104008674621582, + "learning_rate": 0.00022199857317570324, + "loss": 3.3847, + "step": 91605 + }, + { + "epoch": 6.224351134665036, + "grad_norm": 1.0499447584152222, + "learning_rate": 0.0002219561081668705, + "loss": 3.2034, + "step": 91610 + }, + { + "epoch": 6.224690854735698, + "grad_norm": 0.9243391156196594, + "learning_rate": 0.00022191364315803777, + "loss": 3.6383, + "step": 91615 + }, + { + "epoch": 6.225030574806359, + "grad_norm": 1.0231773853302002, + "learning_rate": 0.00022187117814920508, + "loss": 3.3763, + "step": 91620 + }, + { + "epoch": 6.225370294877021, + "grad_norm": 0.9385827779769897, + "learning_rate": 0.00022182871314037233, + "loss": 3.6743, + "step": 91625 + }, + { + "epoch": 6.225710014947683, + "grad_norm": 0.9642002582550049, + "learning_rate": 0.00022178624813153961, + "loss": 3.3, + "step": 91630 + }, + { + "epoch": 6.226049735018345, + "grad_norm": 0.7249485850334167, + "learning_rate": 0.0002217437831227069, + "loss": 3.4741, + "step": 91635 + }, + { + "epoch": 6.226389455089007, + "grad_norm": 1.19260835647583, + "learning_rate": 0.00022170131811387417, + "loss": 3.3307, + "step": 91640 + }, + { + "epoch": 6.226729175159669, + "grad_norm": 0.9525484442710876, + "learning_rate": 0.00022165885310504145, + "loss": 3.7355, + "step": 91645 + }, + { + "epoch": 6.22706889523033, + "grad_norm": 0.9441640973091125, + "learning_rate": 0.00022161638809620873, + "loss": 3.1363, + "step": 91650 + }, + { + "epoch": 6.227408615300992, + "grad_norm": 0.946280300617218, + "learning_rate": 0.00022157392308737601, + "loss": 3.1707, + "step": 91655 + }, + { + "epoch": 6.227748335371654, + "grad_norm": 0.9746332168579102, + "learning_rate": 0.00022153145807854327, + "loss": 3.3953, + "step": 91660 + }, + { + "epoch": 6.228088055442315, + "grad_norm": 0.8614193201065063, + "learning_rate": 0.00022148899306971058, + "loss": 3.4249, + "step": 91665 + }, + { + "epoch": 6.228427775512977, + "grad_norm": 0.8505099415779114, + "learning_rate": 0.00022144652806087783, + "loss": 3.5004, + "step": 91670 + }, + { + "epoch": 6.228767495583639, + "grad_norm": 1.219322681427002, + "learning_rate": 0.0002214040630520451, + "loss": 3.3302, + "step": 91675 + }, + { + "epoch": 6.229107215654301, + "grad_norm": 0.9070805907249451, + "learning_rate": 0.00022136159804321242, + "loss": 3.3401, + "step": 91680 + }, + { + "epoch": 6.229446935724963, + "grad_norm": 0.8754664659500122, + "learning_rate": 0.00022131913303437967, + "loss": 3.4832, + "step": 91685 + }, + { + "epoch": 6.229786655795625, + "grad_norm": 0.998661458492279, + "learning_rate": 0.00022127666802554698, + "loss": 3.4368, + "step": 91690 + }, + { + "epoch": 6.230126375866286, + "grad_norm": 0.9930835962295532, + "learning_rate": 0.00022123420301671423, + "loss": 3.3363, + "step": 91695 + }, + { + "epoch": 6.230466095936948, + "grad_norm": 0.7878478169441223, + "learning_rate": 0.0002211917380078815, + "loss": 3.3223, + "step": 91700 + }, + { + "epoch": 6.23080581600761, + "grad_norm": 0.9926546216011047, + "learning_rate": 0.0002211492729990488, + "loss": 3.2744, + "step": 91705 + }, + { + "epoch": 6.231145536078271, + "grad_norm": 0.9886942505836487, + "learning_rate": 0.00022110680799021607, + "loss": 3.3246, + "step": 91710 + }, + { + "epoch": 6.231485256148933, + "grad_norm": 0.8542454838752747, + "learning_rate": 0.00022106434298138335, + "loss": 3.7848, + "step": 91715 + }, + { + "epoch": 6.231824976219595, + "grad_norm": 1.3048243522644043, + "learning_rate": 0.00022102187797255063, + "loss": 3.4775, + "step": 91720 + }, + { + "epoch": 6.232164696290257, + "grad_norm": 0.910533607006073, + "learning_rate": 0.0002209794129637179, + "loss": 3.4094, + "step": 91725 + }, + { + "epoch": 6.232504416360919, + "grad_norm": 0.7731325626373291, + "learning_rate": 0.00022093694795488516, + "loss": 3.2392, + "step": 91730 + }, + { + "epoch": 6.232844136431581, + "grad_norm": 0.8041125535964966, + "learning_rate": 0.00022089448294605247, + "loss": 3.6308, + "step": 91735 + }, + { + "epoch": 6.233183856502242, + "grad_norm": 1.095589518547058, + "learning_rate": 0.00022085201793721972, + "loss": 3.4272, + "step": 91740 + }, + { + "epoch": 6.233523576572904, + "grad_norm": 1.0329113006591797, + "learning_rate": 0.000220809552928387, + "loss": 3.4022, + "step": 91745 + }, + { + "epoch": 6.233863296643566, + "grad_norm": 1.029699444770813, + "learning_rate": 0.0002207670879195543, + "loss": 3.4269, + "step": 91750 + }, + { + "epoch": 6.234203016714227, + "grad_norm": 0.9072460532188416, + "learning_rate": 0.00022072462291072156, + "loss": 3.3618, + "step": 91755 + }, + { + "epoch": 6.234542736784889, + "grad_norm": 0.8524041771888733, + "learning_rate": 0.00022068215790188884, + "loss": 3.4504, + "step": 91760 + }, + { + "epoch": 6.234882456855551, + "grad_norm": 0.8986486196517944, + "learning_rate": 0.00022063969289305612, + "loss": 3.2714, + "step": 91765 + }, + { + "epoch": 6.235222176926213, + "grad_norm": 0.7959215641021729, + "learning_rate": 0.0002205972278842234, + "loss": 3.4752, + "step": 91770 + }, + { + "epoch": 6.235561896996875, + "grad_norm": 1.1015453338623047, + "learning_rate": 0.0002205547628753907, + "loss": 3.4254, + "step": 91775 + }, + { + "epoch": 6.235901617067537, + "grad_norm": 1.0226341485977173, + "learning_rate": 0.00022051229786655796, + "loss": 3.6876, + "step": 91780 + }, + { + "epoch": 6.236241337138198, + "grad_norm": 0.8979068994522095, + "learning_rate": 0.00022046983285772524, + "loss": 3.2894, + "step": 91785 + }, + { + "epoch": 6.23658105720886, + "grad_norm": 0.9316333532333374, + "learning_rate": 0.00022042736784889252, + "loss": 3.4334, + "step": 91790 + }, + { + "epoch": 6.236920777279522, + "grad_norm": 0.8743883967399597, + "learning_rate": 0.0002203849028400598, + "loss": 3.3604, + "step": 91795 + }, + { + "epoch": 6.237260497350183, + "grad_norm": 0.9224259853363037, + "learning_rate": 0.00022034243783122706, + "loss": 3.738, + "step": 91800 + }, + { + "epoch": 6.237600217420845, + "grad_norm": 0.9478924870491028, + "learning_rate": 0.00022029997282239436, + "loss": 3.5785, + "step": 91805 + }, + { + "epoch": 6.237939937491507, + "grad_norm": 1.113447904586792, + "learning_rate": 0.00022025750781356164, + "loss": 3.3689, + "step": 91810 + }, + { + "epoch": 6.238279657562169, + "grad_norm": 1.0497957468032837, + "learning_rate": 0.0002202150428047289, + "loss": 3.5037, + "step": 91815 + }, + { + "epoch": 6.238619377632831, + "grad_norm": 1.0561506748199463, + "learning_rate": 0.0002201725777958962, + "loss": 3.3556, + "step": 91820 + }, + { + "epoch": 6.238959097703493, + "grad_norm": 0.7771998643875122, + "learning_rate": 0.00022013011278706346, + "loss": 3.2789, + "step": 91825 + }, + { + "epoch": 6.239298817774154, + "grad_norm": 0.7058184146881104, + "learning_rate": 0.00022008764777823074, + "loss": 3.0453, + "step": 91830 + }, + { + "epoch": 6.239638537844816, + "grad_norm": 1.0058231353759766, + "learning_rate": 0.00022004518276939802, + "loss": 3.2044, + "step": 91835 + }, + { + "epoch": 6.239978257915477, + "grad_norm": 1.0189650058746338, + "learning_rate": 0.0002200027177605653, + "loss": 3.2437, + "step": 91840 + }, + { + "epoch": 6.240317977986139, + "grad_norm": 0.9322975277900696, + "learning_rate": 0.00021996025275173258, + "loss": 3.5832, + "step": 91845 + }, + { + "epoch": 6.240657698056801, + "grad_norm": 0.9141377210617065, + "learning_rate": 0.00021991778774289986, + "loss": 3.4389, + "step": 91850 + }, + { + "epoch": 6.240997418127463, + "grad_norm": 3.0145485401153564, + "learning_rate": 0.00021987532273406714, + "loss": 3.4083, + "step": 91855 + }, + { + "epoch": 6.241337138198125, + "grad_norm": 1.0799916982650757, + "learning_rate": 0.00021983285772523442, + "loss": 3.5544, + "step": 91860 + }, + { + "epoch": 6.241676858268787, + "grad_norm": 1.024052381515503, + "learning_rate": 0.0002197903927164017, + "loss": 3.425, + "step": 91865 + }, + { + "epoch": 6.242016578339448, + "grad_norm": 1.0668745040893555, + "learning_rate": 0.00021974792770756895, + "loss": 3.2078, + "step": 91870 + }, + { + "epoch": 6.24235629841011, + "grad_norm": 1.1407063007354736, + "learning_rate": 0.00021970546269873626, + "loss": 3.2761, + "step": 91875 + }, + { + "epoch": 6.242696018480772, + "grad_norm": 0.7392059564590454, + "learning_rate": 0.00021966299768990354, + "loss": 3.4185, + "step": 91880 + }, + { + "epoch": 6.243035738551433, + "grad_norm": 0.9124923944473267, + "learning_rate": 0.0002196205326810708, + "loss": 3.2877, + "step": 91885 + }, + { + "epoch": 6.243375458622095, + "grad_norm": 0.8157771825790405, + "learning_rate": 0.0002195780676722381, + "loss": 3.282, + "step": 91890 + }, + { + "epoch": 6.243715178692757, + "grad_norm": 0.863594651222229, + "learning_rate": 0.00021953560266340535, + "loss": 3.4949, + "step": 91895 + }, + { + "epoch": 6.244054898763419, + "grad_norm": 0.889578640460968, + "learning_rate": 0.00021949313765457263, + "loss": 3.4267, + "step": 91900 + }, + { + "epoch": 6.244394618834081, + "grad_norm": 0.8833943009376526, + "learning_rate": 0.0002194506726457399, + "loss": 3.4565, + "step": 91905 + }, + { + "epoch": 6.244734338904743, + "grad_norm": 1.0835720300674438, + "learning_rate": 0.0002194082076369072, + "loss": 3.5937, + "step": 91910 + }, + { + "epoch": 6.245074058975404, + "grad_norm": 0.816899299621582, + "learning_rate": 0.00021936574262807447, + "loss": 3.371, + "step": 91915 + }, + { + "epoch": 6.245413779046066, + "grad_norm": 1.044680118560791, + "learning_rate": 0.00021932327761924175, + "loss": 3.527, + "step": 91920 + }, + { + "epoch": 6.245753499116728, + "grad_norm": 1.053371787071228, + "learning_rate": 0.00021928081261040903, + "loss": 3.5182, + "step": 91925 + }, + { + "epoch": 6.246093219187389, + "grad_norm": 0.9631611704826355, + "learning_rate": 0.00021923834760157628, + "loss": 3.5297, + "step": 91930 + }, + { + "epoch": 6.246432939258051, + "grad_norm": 1.0250746011734009, + "learning_rate": 0.0002191958825927436, + "loss": 3.1605, + "step": 91935 + }, + { + "epoch": 6.246772659328713, + "grad_norm": 0.8924435973167419, + "learning_rate": 0.00021915341758391087, + "loss": 3.4404, + "step": 91940 + }, + { + "epoch": 6.247112379399375, + "grad_norm": 1.1601213216781616, + "learning_rate": 0.00021911095257507815, + "loss": 3.6469, + "step": 91945 + }, + { + "epoch": 6.247452099470037, + "grad_norm": 0.899371862411499, + "learning_rate": 0.00021906848756624543, + "loss": 3.2738, + "step": 91950 + }, + { + "epoch": 6.247791819540699, + "grad_norm": 0.8973756432533264, + "learning_rate": 0.00021902602255741268, + "loss": 3.4871, + "step": 91955 + }, + { + "epoch": 6.24813153961136, + "grad_norm": 0.7477234601974487, + "learning_rate": 0.00021898355754858, + "loss": 3.479, + "step": 91960 + }, + { + "epoch": 6.248471259682022, + "grad_norm": 0.935004711151123, + "learning_rate": 0.00021894109253974724, + "loss": 3.4664, + "step": 91965 + }, + { + "epoch": 6.248810979752684, + "grad_norm": 0.9005734920501709, + "learning_rate": 0.00021889862753091452, + "loss": 3.3401, + "step": 91970 + }, + { + "epoch": 6.249150699823345, + "grad_norm": 0.772993803024292, + "learning_rate": 0.00021885616252208183, + "loss": 3.4551, + "step": 91975 + }, + { + "epoch": 6.249490419894007, + "grad_norm": 0.9154747128486633, + "learning_rate": 0.00021881369751324908, + "loss": 3.2756, + "step": 91980 + }, + { + "epoch": 6.249830139964669, + "grad_norm": 0.8963324427604675, + "learning_rate": 0.00021877123250441636, + "loss": 3.1769, + "step": 91985 + }, + { + "epoch": 6.250169860035331, + "grad_norm": 0.8433915972709656, + "learning_rate": 0.00021872876749558364, + "loss": 3.3453, + "step": 91990 + }, + { + "epoch": 6.250509580105993, + "grad_norm": 0.8479928374290466, + "learning_rate": 0.00021868630248675092, + "loss": 3.4732, + "step": 91995 + }, + { + "epoch": 6.250849300176655, + "grad_norm": 0.955699622631073, + "learning_rate": 0.00021864383747791818, + "loss": 3.4451, + "step": 92000 + }, + { + "epoch": 6.251189020247316, + "grad_norm": 0.664086639881134, + "learning_rate": 0.00021860137246908548, + "loss": 3.6881, + "step": 92005 + }, + { + "epoch": 6.251528740317978, + "grad_norm": 0.8215911984443665, + "learning_rate": 0.00021855890746025276, + "loss": 3.3922, + "step": 92010 + }, + { + "epoch": 6.25186846038864, + "grad_norm": 0.7532089948654175, + "learning_rate": 0.00021851644245142002, + "loss": 3.1202, + "step": 92015 + }, + { + "epoch": 6.252208180459301, + "grad_norm": 0.9377472400665283, + "learning_rate": 0.00021847397744258733, + "loss": 3.3684, + "step": 92020 + }, + { + "epoch": 6.252547900529963, + "grad_norm": 1.0643583536148071, + "learning_rate": 0.00021843151243375458, + "loss": 3.3294, + "step": 92025 + }, + { + "epoch": 6.252887620600625, + "grad_norm": 0.8668105006217957, + "learning_rate": 0.00021838904742492186, + "loss": 3.3151, + "step": 92030 + }, + { + "epoch": 6.253227340671287, + "grad_norm": 0.7217879295349121, + "learning_rate": 0.00021834658241608914, + "loss": 3.4574, + "step": 92035 + }, + { + "epoch": 6.253567060741949, + "grad_norm": 0.8615267872810364, + "learning_rate": 0.00021830411740725642, + "loss": 3.2898, + "step": 92040 + }, + { + "epoch": 6.253906780812611, + "grad_norm": 0.8358634114265442, + "learning_rate": 0.00021826165239842373, + "loss": 3.5225, + "step": 92045 + }, + { + "epoch": 6.254246500883272, + "grad_norm": 0.69049072265625, + "learning_rate": 0.00021821918738959098, + "loss": 3.4924, + "step": 92050 + }, + { + "epoch": 6.254586220953934, + "grad_norm": 1.0215718746185303, + "learning_rate": 0.00021817672238075826, + "loss": 3.3757, + "step": 92055 + }, + { + "epoch": 6.254925941024596, + "grad_norm": 0.8576117157936096, + "learning_rate": 0.00021813425737192554, + "loss": 3.4395, + "step": 92060 + }, + { + "epoch": 6.255265661095257, + "grad_norm": 1.0028754472732544, + "learning_rate": 0.00021809179236309282, + "loss": 3.289, + "step": 92065 + }, + { + "epoch": 6.255605381165919, + "grad_norm": 0.9595955014228821, + "learning_rate": 0.0002180493273542601, + "loss": 3.2085, + "step": 92070 + }, + { + "epoch": 6.255945101236581, + "grad_norm": 0.9853459000587463, + "learning_rate": 0.00021800686234542738, + "loss": 3.7005, + "step": 92075 + }, + { + "epoch": 6.256284821307243, + "grad_norm": 1.05899977684021, + "learning_rate": 0.00021796439733659466, + "loss": 3.3484, + "step": 92080 + }, + { + "epoch": 6.256624541377905, + "grad_norm": 1.1980276107788086, + "learning_rate": 0.0002179219323277619, + "loss": 3.4384, + "step": 92085 + }, + { + "epoch": 6.256964261448567, + "grad_norm": 0.7114601135253906, + "learning_rate": 0.00021787946731892922, + "loss": 3.4012, + "step": 92090 + }, + { + "epoch": 6.257303981519228, + "grad_norm": 1.1075538396835327, + "learning_rate": 0.00021783700231009647, + "loss": 3.1269, + "step": 92095 + }, + { + "epoch": 6.25764370158989, + "grad_norm": 0.8443933725357056, + "learning_rate": 0.00021779453730126375, + "loss": 3.4316, + "step": 92100 + }, + { + "epoch": 6.257983421660552, + "grad_norm": 0.9339787364006042, + "learning_rate": 0.00021775207229243106, + "loss": 3.3219, + "step": 92105 + }, + { + "epoch": 6.258323141731213, + "grad_norm": 1.1698670387268066, + "learning_rate": 0.0002177096072835983, + "loss": 3.4648, + "step": 92110 + }, + { + "epoch": 6.258662861801875, + "grad_norm": 0.9238146543502808, + "learning_rate": 0.0002176671422747656, + "loss": 3.5188, + "step": 92115 + }, + { + "epoch": 6.259002581872537, + "grad_norm": 0.7683802247047424, + "learning_rate": 0.00021762467726593287, + "loss": 3.5454, + "step": 92120 + }, + { + "epoch": 6.259342301943199, + "grad_norm": 1.375771403312683, + "learning_rate": 0.00021758221225710015, + "loss": 3.2382, + "step": 92125 + }, + { + "epoch": 6.259682022013861, + "grad_norm": 0.8927508592605591, + "learning_rate": 0.00021753974724826743, + "loss": 3.61, + "step": 92130 + }, + { + "epoch": 6.260021742084522, + "grad_norm": 0.7364124655723572, + "learning_rate": 0.0002174972822394347, + "loss": 3.3884, + "step": 92135 + }, + { + "epoch": 6.260361462155184, + "grad_norm": 0.8245034217834473, + "learning_rate": 0.000217454817230602, + "loss": 3.3097, + "step": 92140 + }, + { + "epoch": 6.260701182225846, + "grad_norm": 0.859221875667572, + "learning_rate": 0.00021741235222176927, + "loss": 3.4707, + "step": 92145 + }, + { + "epoch": 6.261040902296507, + "grad_norm": 0.8121516704559326, + "learning_rate": 0.00021736988721293655, + "loss": 3.4889, + "step": 92150 + }, + { + "epoch": 6.261380622367169, + "grad_norm": 1.101943850517273, + "learning_rate": 0.0002173274222041038, + "loss": 3.3823, + "step": 92155 + }, + { + "epoch": 6.261720342437831, + "grad_norm": 0.8484485149383545, + "learning_rate": 0.0002172849571952711, + "loss": 3.6052, + "step": 92160 + }, + { + "epoch": 6.262060062508493, + "grad_norm": 1.1712725162506104, + "learning_rate": 0.00021724249218643837, + "loss": 3.3422, + "step": 92165 + }, + { + "epoch": 6.262399782579155, + "grad_norm": 1.0730161666870117, + "learning_rate": 0.00021720002717760565, + "loss": 3.4552, + "step": 92170 + }, + { + "epoch": 6.262739502649817, + "grad_norm": 1.121092438697815, + "learning_rate": 0.00021715756216877295, + "loss": 3.4983, + "step": 92175 + }, + { + "epoch": 6.263079222720478, + "grad_norm": 1.587396264076233, + "learning_rate": 0.0002171150971599402, + "loss": 3.2967, + "step": 92180 + }, + { + "epoch": 6.26341894279114, + "grad_norm": 0.9430939555168152, + "learning_rate": 0.00021707263215110749, + "loss": 3.4775, + "step": 92185 + }, + { + "epoch": 6.263758662861802, + "grad_norm": 0.9521841406822205, + "learning_rate": 0.00021703016714227477, + "loss": 3.4971, + "step": 92190 + }, + { + "epoch": 6.264098382932463, + "grad_norm": 0.9053028225898743, + "learning_rate": 0.00021698770213344205, + "loss": 3.4886, + "step": 92195 + }, + { + "epoch": 6.264438103003125, + "grad_norm": 0.949072539806366, + "learning_rate": 0.0002169452371246093, + "loss": 3.4676, + "step": 92200 + }, + { + "epoch": 6.264777823073787, + "grad_norm": 0.8761289715766907, + "learning_rate": 0.0002169027721157766, + "loss": 3.2368, + "step": 92205 + }, + { + "epoch": 6.265117543144449, + "grad_norm": 0.9133915305137634, + "learning_rate": 0.0002168603071069439, + "loss": 3.4404, + "step": 92210 + }, + { + "epoch": 6.265457263215111, + "grad_norm": 0.9532087445259094, + "learning_rate": 0.00021681784209811117, + "loss": 3.5517, + "step": 92215 + }, + { + "epoch": 6.265796983285773, + "grad_norm": 1.0880122184753418, + "learning_rate": 0.00021677537708927845, + "loss": 3.4124, + "step": 92220 + }, + { + "epoch": 6.266136703356434, + "grad_norm": 0.8730844259262085, + "learning_rate": 0.0002167329120804457, + "loss": 3.2403, + "step": 92225 + }, + { + "epoch": 6.266476423427096, + "grad_norm": 0.8323755264282227, + "learning_rate": 0.000216690447071613, + "loss": 3.0602, + "step": 92230 + }, + { + "epoch": 6.266816143497758, + "grad_norm": 1.3349214792251587, + "learning_rate": 0.0002166479820627803, + "loss": 3.2801, + "step": 92235 + }, + { + "epoch": 6.267155863568419, + "grad_norm": 0.933864176273346, + "learning_rate": 0.00021660551705394754, + "loss": 3.3493, + "step": 92240 + }, + { + "epoch": 6.267495583639081, + "grad_norm": 0.8932079076766968, + "learning_rate": 0.00021656305204511485, + "loss": 3.5967, + "step": 92245 + }, + { + "epoch": 6.267835303709743, + "grad_norm": 1.3251349925994873, + "learning_rate": 0.0002165205870362821, + "loss": 3.5112, + "step": 92250 + }, + { + "epoch": 6.268175023780405, + "grad_norm": 0.9830455780029297, + "learning_rate": 0.00021647812202744938, + "loss": 3.549, + "step": 92255 + }, + { + "epoch": 6.268514743851067, + "grad_norm": 0.8174699544906616, + "learning_rate": 0.00021643565701861666, + "loss": 3.6839, + "step": 92260 + }, + { + "epoch": 6.268854463921729, + "grad_norm": 0.8530420660972595, + "learning_rate": 0.00021639319200978394, + "loss": 3.4143, + "step": 92265 + }, + { + "epoch": 6.26919418399239, + "grad_norm": 1.0627332925796509, + "learning_rate": 0.00021635072700095122, + "loss": 3.3338, + "step": 92270 + }, + { + "epoch": 6.269533904063052, + "grad_norm": 1.0777802467346191, + "learning_rate": 0.0002163082619921185, + "loss": 3.3799, + "step": 92275 + }, + { + "epoch": 6.269873624133714, + "grad_norm": 0.9872119426727295, + "learning_rate": 0.00021626579698328578, + "loss": 3.1729, + "step": 92280 + }, + { + "epoch": 6.270213344204375, + "grad_norm": 0.9032939076423645, + "learning_rate": 0.00021622333197445303, + "loss": 3.5678, + "step": 92285 + }, + { + "epoch": 6.270553064275037, + "grad_norm": 1.2597192525863647, + "learning_rate": 0.00021618086696562034, + "loss": 3.5358, + "step": 92290 + }, + { + "epoch": 6.270892784345699, + "grad_norm": 0.9028797149658203, + "learning_rate": 0.0002161384019567876, + "loss": 3.27, + "step": 92295 + }, + { + "epoch": 6.271232504416361, + "grad_norm": 0.7938652634620667, + "learning_rate": 0.0002160959369479549, + "loss": 3.2229, + "step": 92300 + }, + { + "epoch": 6.271572224487023, + "grad_norm": 0.7583000063896179, + "learning_rate": 0.00021605347193912218, + "loss": 3.313, + "step": 92305 + }, + { + "epoch": 6.271911944557685, + "grad_norm": 0.7683283686637878, + "learning_rate": 0.00021601100693028943, + "loss": 3.7488, + "step": 92310 + }, + { + "epoch": 6.272251664628346, + "grad_norm": 1.0146740674972534, + "learning_rate": 0.00021596854192145674, + "loss": 3.4832, + "step": 92315 + }, + { + "epoch": 6.272591384699008, + "grad_norm": 1.0343672037124634, + "learning_rate": 0.000215926076912624, + "loss": 3.4372, + "step": 92320 + }, + { + "epoch": 6.27293110476967, + "grad_norm": 0.8925562500953674, + "learning_rate": 0.00021588361190379127, + "loss": 3.496, + "step": 92325 + }, + { + "epoch": 6.273270824840331, + "grad_norm": 0.8888325691223145, + "learning_rate": 0.00021584114689495855, + "loss": 3.5261, + "step": 92330 + }, + { + "epoch": 6.273610544910993, + "grad_norm": 0.8983575105667114, + "learning_rate": 0.00021579868188612583, + "loss": 3.3819, + "step": 92335 + }, + { + "epoch": 6.273950264981655, + "grad_norm": 0.9812121391296387, + "learning_rate": 0.00021575621687729311, + "loss": 3.6367, + "step": 92340 + }, + { + "epoch": 6.274289985052317, + "grad_norm": 0.7401511073112488, + "learning_rate": 0.0002157137518684604, + "loss": 3.2963, + "step": 92345 + }, + { + "epoch": 6.274629705122979, + "grad_norm": 1.0156750679016113, + "learning_rate": 0.00021567128685962767, + "loss": 3.4777, + "step": 92350 + }, + { + "epoch": 6.274969425193641, + "grad_norm": 1.0728609561920166, + "learning_rate": 0.00021562882185079493, + "loss": 3.1786, + "step": 92355 + }, + { + "epoch": 6.275309145264302, + "grad_norm": 0.8794326782226562, + "learning_rate": 0.00021558635684196223, + "loss": 3.2898, + "step": 92360 + }, + { + "epoch": 6.275648865334964, + "grad_norm": 0.7802878618240356, + "learning_rate": 0.00021554389183312951, + "loss": 3.3, + "step": 92365 + }, + { + "epoch": 6.275988585405626, + "grad_norm": 0.864422619342804, + "learning_rate": 0.00021550142682429677, + "loss": 3.1875, + "step": 92370 + }, + { + "epoch": 6.276328305476287, + "grad_norm": 1.0695279836654663, + "learning_rate": 0.00021545896181546408, + "loss": 3.356, + "step": 92375 + }, + { + "epoch": 6.276668025546949, + "grad_norm": 0.8857941627502441, + "learning_rate": 0.00021541649680663133, + "loss": 3.298, + "step": 92380 + }, + { + "epoch": 6.277007745617611, + "grad_norm": 0.7683687806129456, + "learning_rate": 0.00021537403179779864, + "loss": 3.456, + "step": 92385 + }, + { + "epoch": 6.277347465688273, + "grad_norm": 1.1676340103149414, + "learning_rate": 0.0002153315667889659, + "loss": 3.321, + "step": 92390 + }, + { + "epoch": 6.277687185758935, + "grad_norm": 0.6604769229888916, + "learning_rate": 0.00021528910178013317, + "loss": 3.2654, + "step": 92395 + }, + { + "epoch": 6.278026905829597, + "grad_norm": 0.869402289390564, + "learning_rate": 0.00021524663677130048, + "loss": 3.177, + "step": 92400 + }, + { + "epoch": 6.278366625900258, + "grad_norm": 0.8769462704658508, + "learning_rate": 0.00021520417176246773, + "loss": 3.5102, + "step": 92405 + }, + { + "epoch": 6.27870634597092, + "grad_norm": 0.8544986248016357, + "learning_rate": 0.000215161706753635, + "loss": 3.339, + "step": 92410 + }, + { + "epoch": 6.279046066041582, + "grad_norm": 1.3053003549575806, + "learning_rate": 0.0002151192417448023, + "loss": 3.3343, + "step": 92415 + }, + { + "epoch": 6.279385786112243, + "grad_norm": 0.9459705352783203, + "learning_rate": 0.00021507677673596957, + "loss": 3.2336, + "step": 92420 + }, + { + "epoch": 6.279725506182905, + "grad_norm": 0.7853270173072815, + "learning_rate": 0.00021503431172713682, + "loss": 3.6328, + "step": 92425 + }, + { + "epoch": 6.2800652262535674, + "grad_norm": 0.9455561637878418, + "learning_rate": 0.00021499184671830413, + "loss": 3.2919, + "step": 92430 + }, + { + "epoch": 6.280404946324229, + "grad_norm": 1.0764596462249756, + "learning_rate": 0.0002149493817094714, + "loss": 3.6109, + "step": 92435 + }, + { + "epoch": 6.280744666394891, + "grad_norm": 0.9076644778251648, + "learning_rate": 0.00021490691670063866, + "loss": 3.4284, + "step": 92440 + }, + { + "epoch": 6.281084386465553, + "grad_norm": 1.0764981508255005, + "learning_rate": 0.00021486445169180597, + "loss": 3.5571, + "step": 92445 + }, + { + "epoch": 6.281424106536214, + "grad_norm": 0.9228456020355225, + "learning_rate": 0.00021482198668297322, + "loss": 3.3568, + "step": 92450 + }, + { + "epoch": 6.281763826606876, + "grad_norm": 0.9463489651679993, + "learning_rate": 0.0002147795216741405, + "loss": 3.5044, + "step": 92455 + }, + { + "epoch": 6.282103546677538, + "grad_norm": 0.9241097569465637, + "learning_rate": 0.00021473705666530778, + "loss": 3.6773, + "step": 92460 + }, + { + "epoch": 6.282443266748199, + "grad_norm": 1.2108047008514404, + "learning_rate": 0.00021469459165647506, + "loss": 3.2322, + "step": 92465 + }, + { + "epoch": 6.282782986818861, + "grad_norm": 0.9172243475914001, + "learning_rate": 0.00021465212664764237, + "loss": 3.4751, + "step": 92470 + }, + { + "epoch": 6.2831227068895235, + "grad_norm": 1.082959532737732, + "learning_rate": 0.00021460966163880962, + "loss": 3.4316, + "step": 92475 + }, + { + "epoch": 6.283462426960185, + "grad_norm": 0.9732356071472168, + "learning_rate": 0.0002145671966299769, + "loss": 3.5282, + "step": 92480 + }, + { + "epoch": 6.283802147030847, + "grad_norm": 0.801630973815918, + "learning_rate": 0.00021452473162114418, + "loss": 3.3278, + "step": 92485 + }, + { + "epoch": 6.284141867101509, + "grad_norm": 1.0341492891311646, + "learning_rate": 0.00021448226661231146, + "loss": 3.0105, + "step": 92490 + }, + { + "epoch": 6.28448158717217, + "grad_norm": 1.0753453969955444, + "learning_rate": 0.00021443980160347874, + "loss": 3.3254, + "step": 92495 + }, + { + "epoch": 6.284821307242832, + "grad_norm": 1.3121623992919922, + "learning_rate": 0.00021439733659464602, + "loss": 3.3865, + "step": 92500 + }, + { + "epoch": 6.285161027313494, + "grad_norm": 1.005827784538269, + "learning_rate": 0.0002143548715858133, + "loss": 3.6222, + "step": 92505 + }, + { + "epoch": 6.285500747384155, + "grad_norm": 0.9281402230262756, + "learning_rate": 0.00021431240657698056, + "loss": 3.3211, + "step": 92510 + }, + { + "epoch": 6.285840467454817, + "grad_norm": 0.7568219304084778, + "learning_rate": 0.00021426994156814786, + "loss": 3.3384, + "step": 92515 + }, + { + "epoch": 6.2861801875254795, + "grad_norm": 0.7436158657073975, + "learning_rate": 0.00021422747655931512, + "loss": 3.2806, + "step": 92520 + }, + { + "epoch": 6.286519907596141, + "grad_norm": 1.251130223274231, + "learning_rate": 0.0002141850115504824, + "loss": 3.4067, + "step": 92525 + }, + { + "epoch": 6.286859627666803, + "grad_norm": 0.9107165336608887, + "learning_rate": 0.0002141425465416497, + "loss": 3.6129, + "step": 92530 + }, + { + "epoch": 6.287199347737464, + "grad_norm": 0.876348614692688, + "learning_rate": 0.00021410008153281696, + "loss": 3.2992, + "step": 92535 + }, + { + "epoch": 6.287539067808126, + "grad_norm": 1.2684186697006226, + "learning_rate": 0.00021405761652398424, + "loss": 3.6067, + "step": 92540 + }, + { + "epoch": 6.287878787878788, + "grad_norm": 0.9386810660362244, + "learning_rate": 0.00021401515151515152, + "loss": 3.3434, + "step": 92545 + }, + { + "epoch": 6.288218507949449, + "grad_norm": 0.7590407729148865, + "learning_rate": 0.0002139726865063188, + "loss": 3.1638, + "step": 92550 + }, + { + "epoch": 6.288558228020111, + "grad_norm": 0.7799517512321472, + "learning_rate": 0.00021393022149748608, + "loss": 3.4473, + "step": 92555 + }, + { + "epoch": 6.288897948090773, + "grad_norm": 0.9008297920227051, + "learning_rate": 0.00021388775648865336, + "loss": 3.5112, + "step": 92560 + }, + { + "epoch": 6.289237668161435, + "grad_norm": 0.9019983410835266, + "learning_rate": 0.00021384529147982064, + "loss": 3.3545, + "step": 92565 + }, + { + "epoch": 6.289577388232097, + "grad_norm": 1.082118034362793, + "learning_rate": 0.00021380282647098792, + "loss": 3.5786, + "step": 92570 + }, + { + "epoch": 6.289917108302759, + "grad_norm": 1.1226024627685547, + "learning_rate": 0.0002137603614621552, + "loss": 3.5436, + "step": 92575 + }, + { + "epoch": 6.29025682837342, + "grad_norm": 1.164933681488037, + "learning_rate": 0.00021371789645332245, + "loss": 3.2394, + "step": 92580 + }, + { + "epoch": 6.290596548444082, + "grad_norm": 1.0059500932693481, + "learning_rate": 0.00021367543144448976, + "loss": 3.4235, + "step": 92585 + }, + { + "epoch": 6.290936268514744, + "grad_norm": 0.8223692178726196, + "learning_rate": 0.000213632966435657, + "loss": 3.6283, + "step": 92590 + }, + { + "epoch": 6.291275988585405, + "grad_norm": 0.9923964142799377, + "learning_rate": 0.0002135905014268243, + "loss": 3.3039, + "step": 92595 + }, + { + "epoch": 6.291615708656067, + "grad_norm": 1.3390709161758423, + "learning_rate": 0.0002135480364179916, + "loss": 3.5589, + "step": 92600 + }, + { + "epoch": 6.291955428726729, + "grad_norm": 1.0106472969055176, + "learning_rate": 0.00021350557140915885, + "loss": 3.6194, + "step": 92605 + }, + { + "epoch": 6.292295148797391, + "grad_norm": 0.8877888321876526, + "learning_rate": 0.00021346310640032613, + "loss": 3.3291, + "step": 92610 + }, + { + "epoch": 6.292634868868053, + "grad_norm": 1.0763236284255981, + "learning_rate": 0.0002134206413914934, + "loss": 3.5287, + "step": 92615 + }, + { + "epoch": 6.292974588938715, + "grad_norm": 0.9156359434127808, + "learning_rate": 0.0002133781763826607, + "loss": 3.4005, + "step": 92620 + }, + { + "epoch": 6.293314309009376, + "grad_norm": 0.8959893584251404, + "learning_rate": 0.00021333571137382794, + "loss": 3.4009, + "step": 92625 + }, + { + "epoch": 6.293654029080038, + "grad_norm": 1.0773513317108154, + "learning_rate": 0.00021329324636499525, + "loss": 3.3605, + "step": 92630 + }, + { + "epoch": 6.2939937491507, + "grad_norm": 0.8629525303840637, + "learning_rate": 0.00021325078135616253, + "loss": 3.4692, + "step": 92635 + }, + { + "epoch": 6.294333469221361, + "grad_norm": 0.8022880554199219, + "learning_rate": 0.0002132083163473298, + "loss": 3.3537, + "step": 92640 + }, + { + "epoch": 6.294673189292023, + "grad_norm": 0.7768376469612122, + "learning_rate": 0.0002131658513384971, + "loss": 3.3552, + "step": 92645 + }, + { + "epoch": 6.295012909362685, + "grad_norm": 0.8091172575950623, + "learning_rate": 0.00021312338632966434, + "loss": 3.4781, + "step": 92650 + }, + { + "epoch": 6.295352629433347, + "grad_norm": 0.7924095988273621, + "learning_rate": 0.00021308092132083165, + "loss": 3.5715, + "step": 92655 + }, + { + "epoch": 6.295692349504009, + "grad_norm": 0.7695227861404419, + "learning_rate": 0.00021303845631199893, + "loss": 3.4082, + "step": 92660 + }, + { + "epoch": 6.296032069574671, + "grad_norm": 0.7917152643203735, + "learning_rate": 0.00021299599130316618, + "loss": 3.3587, + "step": 92665 + }, + { + "epoch": 6.296371789645332, + "grad_norm": 0.8382953405380249, + "learning_rate": 0.0002129535262943335, + "loss": 3.2999, + "step": 92670 + }, + { + "epoch": 6.296711509715994, + "grad_norm": 0.73267662525177, + "learning_rate": 0.00021291106128550074, + "loss": 3.3486, + "step": 92675 + }, + { + "epoch": 6.297051229786656, + "grad_norm": 0.9308411478996277, + "learning_rate": 0.00021286859627666802, + "loss": 3.536, + "step": 92680 + }, + { + "epoch": 6.297390949857317, + "grad_norm": 1.0506831407546997, + "learning_rate": 0.0002128261312678353, + "loss": 3.3558, + "step": 92685 + }, + { + "epoch": 6.297730669927979, + "grad_norm": 0.8237034678459167, + "learning_rate": 0.00021278366625900258, + "loss": 3.428, + "step": 92690 + }, + { + "epoch": 6.298070389998641, + "grad_norm": 0.9582799673080444, + "learning_rate": 0.00021274120125016986, + "loss": 3.4008, + "step": 92695 + }, + { + "epoch": 6.298410110069303, + "grad_norm": 1.026798963546753, + "learning_rate": 0.00021269873624133714, + "loss": 3.3788, + "step": 92700 + }, + { + "epoch": 6.298749830139965, + "grad_norm": 1.1342514753341675, + "learning_rate": 0.00021265627123250442, + "loss": 3.1901, + "step": 92705 + }, + { + "epoch": 6.299089550210627, + "grad_norm": 0.998609185218811, + "learning_rate": 0.00021261380622367168, + "loss": 3.4195, + "step": 92710 + }, + { + "epoch": 6.299429270281288, + "grad_norm": 0.9667478799819946, + "learning_rate": 0.00021257134121483898, + "loss": 3.8401, + "step": 92715 + }, + { + "epoch": 6.29976899035195, + "grad_norm": 0.9975641965866089, + "learning_rate": 0.00021252887620600624, + "loss": 3.6461, + "step": 92720 + }, + { + "epoch": 6.300108710422612, + "grad_norm": 0.8454297780990601, + "learning_rate": 0.00021248641119717354, + "loss": 3.5022, + "step": 92725 + }, + { + "epoch": 6.300448430493273, + "grad_norm": 0.7456380724906921, + "learning_rate": 0.00021244394618834082, + "loss": 3.5167, + "step": 92730 + }, + { + "epoch": 6.300788150563935, + "grad_norm": 0.8359859585762024, + "learning_rate": 0.00021240148117950808, + "loss": 3.4952, + "step": 92735 + }, + { + "epoch": 6.3011278706345974, + "grad_norm": 1.0109202861785889, + "learning_rate": 0.00021235901617067539, + "loss": 3.4466, + "step": 92740 + }, + { + "epoch": 6.301467590705259, + "grad_norm": 0.8100677132606506, + "learning_rate": 0.00021231655116184264, + "loss": 3.323, + "step": 92745 + }, + { + "epoch": 6.301807310775921, + "grad_norm": 0.7144304513931274, + "learning_rate": 0.00021227408615300992, + "loss": 3.1259, + "step": 92750 + }, + { + "epoch": 6.302147030846583, + "grad_norm": 0.8077279925346375, + "learning_rate": 0.0002122316211441772, + "loss": 3.1372, + "step": 92755 + }, + { + "epoch": 6.302486750917244, + "grad_norm": 0.8964887857437134, + "learning_rate": 0.00021218915613534448, + "loss": 3.3884, + "step": 92760 + }, + { + "epoch": 6.302826470987906, + "grad_norm": 0.8109945058822632, + "learning_rate": 0.00021214669112651176, + "loss": 3.3146, + "step": 92765 + }, + { + "epoch": 6.303166191058568, + "grad_norm": 0.7090948224067688, + "learning_rate": 0.00021210422611767904, + "loss": 3.4876, + "step": 92770 + }, + { + "epoch": 6.303505911129229, + "grad_norm": 1.0680859088897705, + "learning_rate": 0.00021206176110884632, + "loss": 3.5312, + "step": 92775 + }, + { + "epoch": 6.303845631199891, + "grad_norm": 0.9900736808776855, + "learning_rate": 0.00021201929610001357, + "loss": 3.4874, + "step": 92780 + }, + { + "epoch": 6.3041853512705535, + "grad_norm": 1.3722587823867798, + "learning_rate": 0.00021197683109118088, + "loss": 3.334, + "step": 92785 + }, + { + "epoch": 6.304525071341215, + "grad_norm": 0.8592213988304138, + "learning_rate": 0.00021193436608234816, + "loss": 3.4845, + "step": 92790 + }, + { + "epoch": 6.304864791411877, + "grad_norm": 0.7497922778129578, + "learning_rate": 0.0002118919010735154, + "loss": 3.4588, + "step": 92795 + }, + { + "epoch": 6.305204511482539, + "grad_norm": 1.0801632404327393, + "learning_rate": 0.00021184943606468272, + "loss": 3.3887, + "step": 92800 + }, + { + "epoch": 6.3055442315532, + "grad_norm": 0.8127753734588623, + "learning_rate": 0.00021180697105584997, + "loss": 3.4262, + "step": 92805 + }, + { + "epoch": 6.305883951623862, + "grad_norm": 0.9302627444267273, + "learning_rate": 0.00021176450604701728, + "loss": 3.3717, + "step": 92810 + }, + { + "epoch": 6.306223671694523, + "grad_norm": 1.207848072052002, + "learning_rate": 0.00021172204103818453, + "loss": 3.47, + "step": 92815 + }, + { + "epoch": 6.306563391765185, + "grad_norm": 0.7383628487586975, + "learning_rate": 0.0002116795760293518, + "loss": 3.2055, + "step": 92820 + }, + { + "epoch": 6.306903111835847, + "grad_norm": 0.7703129053115845, + "learning_rate": 0.00021163711102051912, + "loss": 3.5551, + "step": 92825 + }, + { + "epoch": 6.307242831906509, + "grad_norm": 0.9326669573783875, + "learning_rate": 0.00021159464601168637, + "loss": 3.5753, + "step": 92830 + }, + { + "epoch": 6.307582551977171, + "grad_norm": 1.016476035118103, + "learning_rate": 0.00021155218100285365, + "loss": 3.5272, + "step": 92835 + }, + { + "epoch": 6.307922272047833, + "grad_norm": 0.9228595495223999, + "learning_rate": 0.00021150971599402093, + "loss": 3.5386, + "step": 92840 + }, + { + "epoch": 6.308261992118494, + "grad_norm": 1.3569319248199463, + "learning_rate": 0.0002114672509851882, + "loss": 3.3373, + "step": 92845 + }, + { + "epoch": 6.308601712189156, + "grad_norm": 0.9067007303237915, + "learning_rate": 0.00021142478597635547, + "loss": 3.4181, + "step": 92850 + }, + { + "epoch": 6.308941432259818, + "grad_norm": 1.0169674158096313, + "learning_rate": 0.00021138232096752277, + "loss": 3.4683, + "step": 92855 + }, + { + "epoch": 6.309281152330479, + "grad_norm": 1.021612286567688, + "learning_rate": 0.00021133985595869005, + "loss": 3.4254, + "step": 92860 + }, + { + "epoch": 6.309620872401141, + "grad_norm": 1.2254984378814697, + "learning_rate": 0.0002112973909498573, + "loss": 3.588, + "step": 92865 + }, + { + "epoch": 6.309960592471803, + "grad_norm": 0.9217529892921448, + "learning_rate": 0.0002112549259410246, + "loss": 3.3321, + "step": 92870 + }, + { + "epoch": 6.310300312542465, + "grad_norm": 0.9056695699691772, + "learning_rate": 0.00021121246093219187, + "loss": 3.4954, + "step": 92875 + }, + { + "epoch": 6.310640032613127, + "grad_norm": 0.9107279181480408, + "learning_rate": 0.00021116999592335915, + "loss": 3.2387, + "step": 92880 + }, + { + "epoch": 6.310979752683789, + "grad_norm": 1.2281490564346313, + "learning_rate": 0.00021112753091452643, + "loss": 3.4541, + "step": 92885 + }, + { + "epoch": 6.31131947275445, + "grad_norm": 0.8052440285682678, + "learning_rate": 0.0002110850659056937, + "loss": 3.338, + "step": 92890 + }, + { + "epoch": 6.311659192825112, + "grad_norm": 1.0549678802490234, + "learning_rate": 0.000211042600896861, + "loss": 3.3562, + "step": 92895 + }, + { + "epoch": 6.311998912895774, + "grad_norm": 0.833828866481781, + "learning_rate": 0.00021100013588802827, + "loss": 3.4928, + "step": 92900 + }, + { + "epoch": 6.312338632966435, + "grad_norm": 0.8759251236915588, + "learning_rate": 0.00021095767087919555, + "loss": 3.3014, + "step": 92905 + }, + { + "epoch": 6.312678353037097, + "grad_norm": 0.9012020826339722, + "learning_rate": 0.00021091520587036283, + "loss": 3.3711, + "step": 92910 + }, + { + "epoch": 6.313018073107759, + "grad_norm": 1.251889944076538, + "learning_rate": 0.0002108727408615301, + "loss": 3.3642, + "step": 92915 + }, + { + "epoch": 6.313357793178421, + "grad_norm": 0.8887602090835571, + "learning_rate": 0.00021083027585269736, + "loss": 3.3508, + "step": 92920 + }, + { + "epoch": 6.313697513249083, + "grad_norm": 0.9437685012817383, + "learning_rate": 0.00021078781084386467, + "loss": 3.3346, + "step": 92925 + }, + { + "epoch": 6.314037233319745, + "grad_norm": 0.8199687600135803, + "learning_rate": 0.00021074534583503195, + "loss": 3.3863, + "step": 92930 + }, + { + "epoch": 6.314376953390406, + "grad_norm": 1.0552566051483154, + "learning_rate": 0.0002107028808261992, + "loss": 3.3185, + "step": 92935 + }, + { + "epoch": 6.314716673461068, + "grad_norm": 0.8494763970375061, + "learning_rate": 0.0002106604158173665, + "loss": 3.3014, + "step": 92940 + }, + { + "epoch": 6.31505639353173, + "grad_norm": 0.7918281555175781, + "learning_rate": 0.00021061795080853376, + "loss": 3.5222, + "step": 92945 + }, + { + "epoch": 6.315396113602391, + "grad_norm": 0.8768267035484314, + "learning_rate": 0.00021057548579970104, + "loss": 3.3369, + "step": 92950 + }, + { + "epoch": 6.315735833673053, + "grad_norm": 0.8506472706794739, + "learning_rate": 0.00021053302079086835, + "loss": 3.4126, + "step": 92955 + }, + { + "epoch": 6.316075553743715, + "grad_norm": 0.7600993514060974, + "learning_rate": 0.0002104905557820356, + "loss": 3.537, + "step": 92960 + }, + { + "epoch": 6.316415273814377, + "grad_norm": 0.7509888410568237, + "learning_rate": 0.00021044809077320288, + "loss": 3.2806, + "step": 92965 + }, + { + "epoch": 6.316754993885039, + "grad_norm": 0.9349867105484009, + "learning_rate": 0.00021040562576437016, + "loss": 3.3286, + "step": 92970 + }, + { + "epoch": 6.317094713955701, + "grad_norm": 0.8032007813453674, + "learning_rate": 0.00021036316075553744, + "loss": 3.4489, + "step": 92975 + }, + { + "epoch": 6.317434434026362, + "grad_norm": 1.0763477087020874, + "learning_rate": 0.00021032069574670472, + "loss": 3.4126, + "step": 92980 + }, + { + "epoch": 6.317774154097024, + "grad_norm": 0.927708625793457, + "learning_rate": 0.000210278230737872, + "loss": 3.314, + "step": 92985 + }, + { + "epoch": 6.318113874167686, + "grad_norm": 0.7462497353553772, + "learning_rate": 0.00021023576572903928, + "loss": 3.6401, + "step": 92990 + }, + { + "epoch": 6.318453594238347, + "grad_norm": 1.0594457387924194, + "learning_rate": 0.00021019330072020656, + "loss": 3.4659, + "step": 92995 + }, + { + "epoch": 6.318793314309009, + "grad_norm": 1.11244797706604, + "learning_rate": 0.00021015083571137384, + "loss": 3.3078, + "step": 93000 + }, + { + "epoch": 6.319133034379671, + "grad_norm": 0.8586558699607849, + "learning_rate": 0.0002101083707025411, + "loss": 3.5166, + "step": 93005 + }, + { + "epoch": 6.319472754450333, + "grad_norm": 0.7933627963066101, + "learning_rate": 0.0002100659056937084, + "loss": 3.5189, + "step": 93010 + }, + { + "epoch": 6.319812474520995, + "grad_norm": 0.9431900978088379, + "learning_rate": 0.00021002344068487565, + "loss": 3.1853, + "step": 93015 + }, + { + "epoch": 6.320152194591657, + "grad_norm": 1.3083829879760742, + "learning_rate": 0.00020998097567604293, + "loss": 3.3265, + "step": 93020 + }, + { + "epoch": 6.320491914662318, + "grad_norm": 1.0301471948623657, + "learning_rate": 0.00020993851066721024, + "loss": 3.3989, + "step": 93025 + }, + { + "epoch": 6.32083163473298, + "grad_norm": 1.0037345886230469, + "learning_rate": 0.0002098960456583775, + "loss": 3.3877, + "step": 93030 + }, + { + "epoch": 6.321171354803642, + "grad_norm": 0.8912409543991089, + "learning_rate": 0.00020985358064954477, + "loss": 3.3336, + "step": 93035 + }, + { + "epoch": 6.321511074874303, + "grad_norm": 0.7959961295127869, + "learning_rate": 0.00020981111564071205, + "loss": 3.2868, + "step": 93040 + }, + { + "epoch": 6.321850794944965, + "grad_norm": 0.9019456505775452, + "learning_rate": 0.00020976865063187933, + "loss": 3.2662, + "step": 93045 + }, + { + "epoch": 6.3221905150156275, + "grad_norm": 1.313879132270813, + "learning_rate": 0.0002097261856230466, + "loss": 3.234, + "step": 93050 + }, + { + "epoch": 6.322530235086289, + "grad_norm": 1.0549626350402832, + "learning_rate": 0.0002096837206142139, + "loss": 3.1793, + "step": 93055 + }, + { + "epoch": 6.322869955156951, + "grad_norm": 1.1045031547546387, + "learning_rate": 0.00020964125560538117, + "loss": 3.2014, + "step": 93060 + }, + { + "epoch": 6.323209675227613, + "grad_norm": 1.2126115560531616, + "learning_rate": 0.00020959879059654845, + "loss": 3.274, + "step": 93065 + }, + { + "epoch": 6.323549395298274, + "grad_norm": 0.9573268890380859, + "learning_rate": 0.00020955632558771573, + "loss": 3.4238, + "step": 93070 + }, + { + "epoch": 6.323889115368936, + "grad_norm": 1.0978518724441528, + "learning_rate": 0.000209513860578883, + "loss": 3.5386, + "step": 93075 + }, + { + "epoch": 6.324228835439598, + "grad_norm": 0.9228789806365967, + "learning_rate": 0.0002094713955700503, + "loss": 3.4592, + "step": 93080 + }, + { + "epoch": 6.324568555510259, + "grad_norm": 0.7471944093704224, + "learning_rate": 0.00020942893056121757, + "loss": 3.2953, + "step": 93085 + }, + { + "epoch": 6.324908275580921, + "grad_norm": 0.9422825574874878, + "learning_rate": 0.00020938646555238483, + "loss": 3.7767, + "step": 93090 + }, + { + "epoch": 6.3252479956515835, + "grad_norm": 0.9584270119667053, + "learning_rate": 0.00020934400054355214, + "loss": 3.3635, + "step": 93095 + }, + { + "epoch": 6.325587715722245, + "grad_norm": 0.9281519651412964, + "learning_rate": 0.0002093015355347194, + "loss": 3.2187, + "step": 93100 + }, + { + "epoch": 6.325927435792907, + "grad_norm": 0.7711881995201111, + "learning_rate": 0.00020925907052588667, + "loss": 3.4525, + "step": 93105 + }, + { + "epoch": 6.326267155863569, + "grad_norm": 1.0408729314804077, + "learning_rate": 0.00020921660551705395, + "loss": 3.4577, + "step": 93110 + }, + { + "epoch": 6.32660687593423, + "grad_norm": 0.8527374267578125, + "learning_rate": 0.00020917414050822123, + "loss": 3.3573, + "step": 93115 + }, + { + "epoch": 6.326946596004892, + "grad_norm": 0.8087122440338135, + "learning_rate": 0.0002091316754993885, + "loss": 3.5317, + "step": 93120 + }, + { + "epoch": 6.327286316075554, + "grad_norm": 0.7332887649536133, + "learning_rate": 0.0002090892104905558, + "loss": 3.404, + "step": 93125 + }, + { + "epoch": 6.327626036146215, + "grad_norm": 1.2621077299118042, + "learning_rate": 0.00020904674548172307, + "loss": 3.4994, + "step": 93130 + }, + { + "epoch": 6.327965756216877, + "grad_norm": 1.3128325939178467, + "learning_rate": 0.00020900428047289032, + "loss": 3.3778, + "step": 93135 + }, + { + "epoch": 6.3283054762875395, + "grad_norm": 1.1470468044281006, + "learning_rate": 0.00020896181546405763, + "loss": 3.4444, + "step": 93140 + }, + { + "epoch": 6.328645196358201, + "grad_norm": 1.169789433479309, + "learning_rate": 0.00020891935045522488, + "loss": 3.1228, + "step": 93145 + }, + { + "epoch": 6.328984916428863, + "grad_norm": 0.8558716773986816, + "learning_rate": 0.0002088768854463922, + "loss": 3.4084, + "step": 93150 + }, + { + "epoch": 6.329324636499525, + "grad_norm": 0.8265169262886047, + "learning_rate": 0.00020883442043755947, + "loss": 3.4278, + "step": 93155 + }, + { + "epoch": 6.329664356570186, + "grad_norm": 0.9496740698814392, + "learning_rate": 0.00020879195542872672, + "loss": 3.5505, + "step": 93160 + }, + { + "epoch": 6.330004076640848, + "grad_norm": 0.9115471839904785, + "learning_rate": 0.00020874949041989403, + "loss": 3.3597, + "step": 93165 + }, + { + "epoch": 6.33034379671151, + "grad_norm": 0.9049580097198486, + "learning_rate": 0.00020870702541106128, + "loss": 3.2882, + "step": 93170 + }, + { + "epoch": 6.330683516782171, + "grad_norm": 0.9282203316688538, + "learning_rate": 0.00020866456040222856, + "loss": 3.116, + "step": 93175 + }, + { + "epoch": 6.331023236852833, + "grad_norm": 0.9392063021659851, + "learning_rate": 0.00020862209539339584, + "loss": 3.4154, + "step": 93180 + }, + { + "epoch": 6.3313629569234955, + "grad_norm": 0.8577257990837097, + "learning_rate": 0.00020857963038456312, + "loss": 3.2864, + "step": 93185 + }, + { + "epoch": 6.331702676994157, + "grad_norm": 0.9449401497840881, + "learning_rate": 0.0002085371653757304, + "loss": 3.4316, + "step": 93190 + }, + { + "epoch": 6.332042397064819, + "grad_norm": 0.8295576572418213, + "learning_rate": 0.00020849470036689768, + "loss": 3.6906, + "step": 93195 + }, + { + "epoch": 6.332382117135481, + "grad_norm": 1.1219544410705566, + "learning_rate": 0.00020845223535806496, + "loss": 3.3563, + "step": 93200 + }, + { + "epoch": 6.332721837206142, + "grad_norm": 0.7571269869804382, + "learning_rate": 0.00020840977034923222, + "loss": 3.1303, + "step": 93205 + }, + { + "epoch": 6.333061557276804, + "grad_norm": 0.7990916967391968, + "learning_rate": 0.00020836730534039952, + "loss": 3.3591, + "step": 93210 + }, + { + "epoch": 6.333401277347465, + "grad_norm": 0.8608622550964355, + "learning_rate": 0.0002083248403315668, + "loss": 3.4729, + "step": 93215 + }, + { + "epoch": 6.333740997418127, + "grad_norm": 0.951915442943573, + "learning_rate": 0.00020828237532273406, + "loss": 3.4133, + "step": 93220 + }, + { + "epoch": 6.334080717488789, + "grad_norm": 0.8623780012130737, + "learning_rate": 0.00020823991031390136, + "loss": 3.219, + "step": 93225 + }, + { + "epoch": 6.334420437559451, + "grad_norm": 0.8072280287742615, + "learning_rate": 0.00020819744530506862, + "loss": 3.3657, + "step": 93230 + }, + { + "epoch": 6.334760157630113, + "grad_norm": 1.162879228591919, + "learning_rate": 0.00020815498029623592, + "loss": 3.748, + "step": 93235 + }, + { + "epoch": 6.335099877700775, + "grad_norm": 0.9772855639457703, + "learning_rate": 0.00020811251528740318, + "loss": 3.414, + "step": 93240 + }, + { + "epoch": 6.335439597771436, + "grad_norm": 0.7952073812484741, + "learning_rate": 0.00020807005027857046, + "loss": 3.4048, + "step": 93245 + }, + { + "epoch": 6.335779317842098, + "grad_norm": 0.8229575753211975, + "learning_rate": 0.00020802758526973776, + "loss": 3.5183, + "step": 93250 + }, + { + "epoch": 6.33611903791276, + "grad_norm": 0.9559911489486694, + "learning_rate": 0.00020798512026090502, + "loss": 3.3616, + "step": 93255 + }, + { + "epoch": 6.336458757983421, + "grad_norm": 1.1874074935913086, + "learning_rate": 0.0002079426552520723, + "loss": 3.4556, + "step": 93260 + }, + { + "epoch": 6.336798478054083, + "grad_norm": 0.8314729928970337, + "learning_rate": 0.00020790019024323958, + "loss": 3.4872, + "step": 93265 + }, + { + "epoch": 6.337138198124745, + "grad_norm": 1.1051291227340698, + "learning_rate": 0.00020785772523440686, + "loss": 3.2684, + "step": 93270 + }, + { + "epoch": 6.337477918195407, + "grad_norm": 0.9532017707824707, + "learning_rate": 0.0002078152602255741, + "loss": 3.6478, + "step": 93275 + }, + { + "epoch": 6.337817638266069, + "grad_norm": 1.2911181449890137, + "learning_rate": 0.00020777279521674142, + "loss": 3.5507, + "step": 93280 + }, + { + "epoch": 6.338157358336731, + "grad_norm": 1.0194748640060425, + "learning_rate": 0.0002077303302079087, + "loss": 3.3473, + "step": 93285 + }, + { + "epoch": 6.338497078407392, + "grad_norm": 0.9114698767662048, + "learning_rate": 0.00020768786519907595, + "loss": 3.5028, + "step": 93290 + }, + { + "epoch": 6.338836798478054, + "grad_norm": 0.8540284633636475, + "learning_rate": 0.00020764540019024326, + "loss": 3.4809, + "step": 93295 + }, + { + "epoch": 6.339176518548716, + "grad_norm": 0.7690562009811401, + "learning_rate": 0.0002076029351814105, + "loss": 3.2859, + "step": 93300 + }, + { + "epoch": 6.339516238619377, + "grad_norm": 1.0112966299057007, + "learning_rate": 0.0002075604701725778, + "loss": 3.5502, + "step": 93305 + }, + { + "epoch": 6.339855958690039, + "grad_norm": 0.9176316261291504, + "learning_rate": 0.00020751800516374507, + "loss": 3.3735, + "step": 93310 + }, + { + "epoch": 6.3401956787607014, + "grad_norm": 1.0724854469299316, + "learning_rate": 0.00020747554015491235, + "loss": 3.1104, + "step": 93315 + }, + { + "epoch": 6.340535398831363, + "grad_norm": 1.0300065279006958, + "learning_rate": 0.00020743307514607966, + "loss": 3.5864, + "step": 93320 + }, + { + "epoch": 6.340875118902025, + "grad_norm": 0.8555862903594971, + "learning_rate": 0.0002073906101372469, + "loss": 3.3918, + "step": 93325 + }, + { + "epoch": 6.341214838972687, + "grad_norm": 0.9722951650619507, + "learning_rate": 0.0002073481451284142, + "loss": 3.4647, + "step": 93330 + }, + { + "epoch": 6.341554559043348, + "grad_norm": 0.8005595803260803, + "learning_rate": 0.00020730568011958147, + "loss": 3.184, + "step": 93335 + }, + { + "epoch": 6.34189427911401, + "grad_norm": 1.3127795457839966, + "learning_rate": 0.00020726321511074875, + "loss": 3.2958, + "step": 93340 + }, + { + "epoch": 6.342233999184672, + "grad_norm": 0.7469251751899719, + "learning_rate": 0.000207220750101916, + "loss": 3.5091, + "step": 93345 + }, + { + "epoch": 6.342573719255333, + "grad_norm": 0.8643844127655029, + "learning_rate": 0.0002071782850930833, + "loss": 3.2851, + "step": 93350 + }, + { + "epoch": 6.342913439325995, + "grad_norm": 0.8339781761169434, + "learning_rate": 0.0002071358200842506, + "loss": 3.4669, + "step": 93355 + }, + { + "epoch": 6.3432531593966575, + "grad_norm": 0.7458248734474182, + "learning_rate": 0.00020709335507541784, + "loss": 3.2202, + "step": 93360 + }, + { + "epoch": 6.343592879467319, + "grad_norm": 1.006633996963501, + "learning_rate": 0.00020705089006658515, + "loss": 3.3883, + "step": 93365 + }, + { + "epoch": 6.343932599537981, + "grad_norm": 1.1080385446548462, + "learning_rate": 0.0002070084250577524, + "loss": 3.3964, + "step": 93370 + }, + { + "epoch": 6.344272319608643, + "grad_norm": 0.9366435408592224, + "learning_rate": 0.00020696596004891968, + "loss": 3.5193, + "step": 93375 + }, + { + "epoch": 6.344612039679304, + "grad_norm": 0.9472304582595825, + "learning_rate": 0.000206923495040087, + "loss": 3.4431, + "step": 93380 + }, + { + "epoch": 6.344951759749966, + "grad_norm": 1.1042567491531372, + "learning_rate": 0.00020688103003125424, + "loss": 3.5745, + "step": 93385 + }, + { + "epoch": 6.345291479820628, + "grad_norm": 0.9909352660179138, + "learning_rate": 0.00020683856502242152, + "loss": 3.3532, + "step": 93390 + }, + { + "epoch": 6.345631199891289, + "grad_norm": 0.7646176218986511, + "learning_rate": 0.0002067961000135888, + "loss": 3.5027, + "step": 93395 + }, + { + "epoch": 6.345970919961951, + "grad_norm": 1.1099528074264526, + "learning_rate": 0.00020675363500475608, + "loss": 3.25, + "step": 93400 + }, + { + "epoch": 6.3463106400326135, + "grad_norm": 1.1553760766983032, + "learning_rate": 0.00020671116999592336, + "loss": 3.3692, + "step": 93405 + }, + { + "epoch": 6.346650360103275, + "grad_norm": 1.1822328567504883, + "learning_rate": 0.00020666870498709064, + "loss": 3.4193, + "step": 93410 + }, + { + "epoch": 6.346990080173937, + "grad_norm": 1.0182394981384277, + "learning_rate": 0.00020662623997825792, + "loss": 3.3917, + "step": 93415 + }, + { + "epoch": 6.347329800244599, + "grad_norm": 1.038223147392273, + "learning_rate": 0.0002065837749694252, + "loss": 3.375, + "step": 93420 + }, + { + "epoch": 6.34766952031526, + "grad_norm": 0.8603542447090149, + "learning_rate": 0.00020654130996059248, + "loss": 3.5199, + "step": 93425 + }, + { + "epoch": 6.348009240385922, + "grad_norm": 1.2962745428085327, + "learning_rate": 0.00020649884495175974, + "loss": 3.2823, + "step": 93430 + }, + { + "epoch": 6.348348960456584, + "grad_norm": 0.986595094203949, + "learning_rate": 0.00020645637994292704, + "loss": 3.2192, + "step": 93435 + }, + { + "epoch": 6.348688680527245, + "grad_norm": 0.8476473689079285, + "learning_rate": 0.0002064139149340943, + "loss": 3.4166, + "step": 93440 + }, + { + "epoch": 6.349028400597907, + "grad_norm": 0.9261736273765564, + "learning_rate": 0.00020637144992526158, + "loss": 3.4966, + "step": 93445 + }, + { + "epoch": 6.3493681206685695, + "grad_norm": 0.886756181716919, + "learning_rate": 0.00020632898491642889, + "loss": 3.5677, + "step": 93450 + }, + { + "epoch": 6.349707840739231, + "grad_norm": 0.86305832862854, + "learning_rate": 0.00020628651990759614, + "loss": 3.5894, + "step": 93455 + }, + { + "epoch": 6.350047560809893, + "grad_norm": 0.9117932915687561, + "learning_rate": 0.00020624405489876342, + "loss": 3.2428, + "step": 93460 + }, + { + "epoch": 6.350387280880555, + "grad_norm": 1.0597732067108154, + "learning_rate": 0.0002062015898899307, + "loss": 3.3908, + "step": 93465 + }, + { + "epoch": 6.350727000951216, + "grad_norm": 0.9991332292556763, + "learning_rate": 0.00020615912488109798, + "loss": 3.4515, + "step": 93470 + }, + { + "epoch": 6.351066721021878, + "grad_norm": 1.0112873315811157, + "learning_rate": 0.00020611665987226523, + "loss": 3.7104, + "step": 93475 + }, + { + "epoch": 6.35140644109254, + "grad_norm": 1.1035487651824951, + "learning_rate": 0.00020607419486343254, + "loss": 3.4464, + "step": 93480 + }, + { + "epoch": 6.351746161163201, + "grad_norm": 0.9429889917373657, + "learning_rate": 0.00020603172985459982, + "loss": 3.4006, + "step": 93485 + }, + { + "epoch": 6.352085881233863, + "grad_norm": 0.8171404004096985, + "learning_rate": 0.0002059892648457671, + "loss": 3.1409, + "step": 93490 + }, + { + "epoch": 6.352425601304525, + "grad_norm": 0.9287579655647278, + "learning_rate": 0.00020594679983693438, + "loss": 3.3017, + "step": 93495 + }, + { + "epoch": 6.352765321375187, + "grad_norm": 0.826914370059967, + "learning_rate": 0.00020590433482810163, + "loss": 3.5343, + "step": 93500 + }, + { + "epoch": 6.353105041445849, + "grad_norm": 0.7676225304603577, + "learning_rate": 0.00020586186981926894, + "loss": 3.2706, + "step": 93505 + }, + { + "epoch": 6.35344476151651, + "grad_norm": 0.9659386873245239, + "learning_rate": 0.00020581940481043622, + "loss": 3.3683, + "step": 93510 + }, + { + "epoch": 6.353784481587172, + "grad_norm": 0.9111468195915222, + "learning_rate": 0.00020577693980160347, + "loss": 3.4701, + "step": 93515 + }, + { + "epoch": 6.354124201657834, + "grad_norm": 1.0626252889633179, + "learning_rate": 0.00020573447479277078, + "loss": 3.1697, + "step": 93520 + }, + { + "epoch": 6.354463921728495, + "grad_norm": 0.7348642349243164, + "learning_rate": 0.00020569200978393803, + "loss": 3.48, + "step": 93525 + }, + { + "epoch": 6.354803641799157, + "grad_norm": 0.8672160506248474, + "learning_rate": 0.0002056495447751053, + "loss": 3.518, + "step": 93530 + }, + { + "epoch": 6.355143361869819, + "grad_norm": 1.0207653045654297, + "learning_rate": 0.0002056070797662726, + "loss": 3.63, + "step": 93535 + }, + { + "epoch": 6.355483081940481, + "grad_norm": 1.0631635189056396, + "learning_rate": 0.00020556461475743987, + "loss": 3.3081, + "step": 93540 + }, + { + "epoch": 6.355822802011143, + "grad_norm": 1.506929636001587, + "learning_rate": 0.00020552214974860715, + "loss": 3.7309, + "step": 93545 + }, + { + "epoch": 6.356162522081805, + "grad_norm": 0.9792369604110718, + "learning_rate": 0.00020547968473977443, + "loss": 3.4332, + "step": 93550 + }, + { + "epoch": 6.356502242152466, + "grad_norm": 0.9839656352996826, + "learning_rate": 0.0002054372197309417, + "loss": 3.4698, + "step": 93555 + }, + { + "epoch": 6.356841962223128, + "grad_norm": 1.0484446287155151, + "learning_rate": 0.00020539475472210897, + "loss": 3.1608, + "step": 93560 + }, + { + "epoch": 6.35718168229379, + "grad_norm": 0.956450343132019, + "learning_rate": 0.00020535228971327627, + "loss": 3.1835, + "step": 93565 + }, + { + "epoch": 6.357521402364451, + "grad_norm": 0.7095959186553955, + "learning_rate": 0.00020530982470444353, + "loss": 3.3862, + "step": 93570 + }, + { + "epoch": 6.357861122435113, + "grad_norm": 1.105338215827942, + "learning_rate": 0.00020526735969561083, + "loss": 3.4894, + "step": 93575 + }, + { + "epoch": 6.358200842505775, + "grad_norm": 1.1284514665603638, + "learning_rate": 0.0002052248946867781, + "loss": 3.4887, + "step": 93580 + }, + { + "epoch": 6.358540562576437, + "grad_norm": 0.7850402593612671, + "learning_rate": 0.00020518242967794537, + "loss": 3.4181, + "step": 93585 + }, + { + "epoch": 6.358880282647099, + "grad_norm": 1.0426077842712402, + "learning_rate": 0.00020513996466911267, + "loss": 3.3546, + "step": 93590 + }, + { + "epoch": 6.359220002717761, + "grad_norm": 0.8507510423660278, + "learning_rate": 0.00020509749966027993, + "loss": 3.3571, + "step": 93595 + }, + { + "epoch": 6.359559722788422, + "grad_norm": 0.8204994201660156, + "learning_rate": 0.0002050550346514472, + "loss": 3.6818, + "step": 93600 + }, + { + "epoch": 6.359899442859084, + "grad_norm": 1.0001808404922485, + "learning_rate": 0.00020501256964261449, + "loss": 3.1644, + "step": 93605 + }, + { + "epoch": 6.360239162929746, + "grad_norm": 0.8016472458839417, + "learning_rate": 0.00020497010463378177, + "loss": 3.5513, + "step": 93610 + }, + { + "epoch": 6.360578883000407, + "grad_norm": 0.8873401880264282, + "learning_rate": 0.00020492763962494905, + "loss": 3.4867, + "step": 93615 + }, + { + "epoch": 6.360918603071069, + "grad_norm": 0.7872211337089539, + "learning_rate": 0.00020488517461611633, + "loss": 3.3584, + "step": 93620 + }, + { + "epoch": 6.3612583231417315, + "grad_norm": 0.8319951891899109, + "learning_rate": 0.0002048427096072836, + "loss": 3.4804, + "step": 93625 + }, + { + "epoch": 6.361598043212393, + "grad_norm": 0.8324245810508728, + "learning_rate": 0.00020480024459845086, + "loss": 3.2218, + "step": 93630 + }, + { + "epoch": 6.361937763283055, + "grad_norm": 0.7450134754180908, + "learning_rate": 0.00020475777958961817, + "loss": 3.5067, + "step": 93635 + }, + { + "epoch": 6.362277483353717, + "grad_norm": 1.001420259475708, + "learning_rate": 0.00020471531458078545, + "loss": 3.4165, + "step": 93640 + }, + { + "epoch": 6.362617203424378, + "grad_norm": 0.9672594666481018, + "learning_rate": 0.0002046728495719527, + "loss": 3.5956, + "step": 93645 + }, + { + "epoch": 6.36295692349504, + "grad_norm": 0.8559045791625977, + "learning_rate": 0.00020463038456312, + "loss": 3.293, + "step": 93650 + }, + { + "epoch": 6.363296643565702, + "grad_norm": 0.8664975166320801, + "learning_rate": 0.00020458791955428726, + "loss": 3.2147, + "step": 93655 + }, + { + "epoch": 6.363636363636363, + "grad_norm": 0.7858191728591919, + "learning_rate": 0.00020454545454545457, + "loss": 3.2088, + "step": 93660 + }, + { + "epoch": 6.363976083707025, + "grad_norm": 0.8549580574035645, + "learning_rate": 0.00020450298953662182, + "loss": 3.3774, + "step": 93665 + }, + { + "epoch": 6.3643158037776875, + "grad_norm": 1.1617403030395508, + "learning_rate": 0.0002044605245277891, + "loss": 3.4874, + "step": 93670 + }, + { + "epoch": 6.364655523848349, + "grad_norm": 0.8117837309837341, + "learning_rate": 0.0002044180595189564, + "loss": 3.466, + "step": 93675 + }, + { + "epoch": 6.364995243919011, + "grad_norm": 1.1622360944747925, + "learning_rate": 0.00020437559451012366, + "loss": 3.1324, + "step": 93680 + }, + { + "epoch": 6.365334963989673, + "grad_norm": 1.0495998859405518, + "learning_rate": 0.00020433312950129094, + "loss": 3.6885, + "step": 93685 + }, + { + "epoch": 6.365674684060334, + "grad_norm": 0.9093348979949951, + "learning_rate": 0.00020429066449245822, + "loss": 3.6279, + "step": 93690 + }, + { + "epoch": 6.366014404130996, + "grad_norm": 0.9214874505996704, + "learning_rate": 0.0002042481994836255, + "loss": 3.3643, + "step": 93695 + }, + { + "epoch": 6.366354124201658, + "grad_norm": 0.7569080591201782, + "learning_rate": 0.00020420573447479275, + "loss": 3.4816, + "step": 93700 + }, + { + "epoch": 6.366693844272319, + "grad_norm": 0.8083691000938416, + "learning_rate": 0.00020416326946596006, + "loss": 3.0548, + "step": 93705 + }, + { + "epoch": 6.367033564342981, + "grad_norm": 1.0062192678451538, + "learning_rate": 0.00020412080445712734, + "loss": 3.5413, + "step": 93710 + }, + { + "epoch": 6.3673732844136435, + "grad_norm": 0.8926637172698975, + "learning_rate": 0.0002040783394482946, + "loss": 3.4394, + "step": 93715 + }, + { + "epoch": 6.367713004484305, + "grad_norm": 0.8056049942970276, + "learning_rate": 0.0002040358744394619, + "loss": 3.1883, + "step": 93720 + }, + { + "epoch": 6.368052724554967, + "grad_norm": 0.9057061672210693, + "learning_rate": 0.00020399340943062915, + "loss": 3.5683, + "step": 93725 + }, + { + "epoch": 6.368392444625629, + "grad_norm": 0.8892221450805664, + "learning_rate": 0.00020395094442179643, + "loss": 3.4879, + "step": 93730 + }, + { + "epoch": 6.36873216469629, + "grad_norm": 0.8191994428634644, + "learning_rate": 0.00020390847941296371, + "loss": 3.339, + "step": 93735 + }, + { + "epoch": 6.369071884766952, + "grad_norm": 0.8208190202713013, + "learning_rate": 0.000203866014404131, + "loss": 3.3284, + "step": 93740 + }, + { + "epoch": 6.369411604837614, + "grad_norm": 0.9544496536254883, + "learning_rate": 0.0002038235493952983, + "loss": 3.6595, + "step": 93745 + }, + { + "epoch": 6.369751324908275, + "grad_norm": 0.9484769105911255, + "learning_rate": 0.00020378108438646555, + "loss": 3.5248, + "step": 93750 + }, + { + "epoch": 6.370091044978937, + "grad_norm": 0.8739914894104004, + "learning_rate": 0.00020373861937763283, + "loss": 3.4515, + "step": 93755 + }, + { + "epoch": 6.3704307650495995, + "grad_norm": 0.7641595602035522, + "learning_rate": 0.00020369615436880011, + "loss": 3.5378, + "step": 93760 + }, + { + "epoch": 6.370770485120261, + "grad_norm": 0.8846280574798584, + "learning_rate": 0.0002036536893599674, + "loss": 3.5274, + "step": 93765 + }, + { + "epoch": 6.371110205190923, + "grad_norm": 0.774095892906189, + "learning_rate": 0.00020361122435113465, + "loss": 3.5279, + "step": 93770 + }, + { + "epoch": 6.371449925261585, + "grad_norm": 0.9133212566375732, + "learning_rate": 0.00020356875934230195, + "loss": 3.3956, + "step": 93775 + }, + { + "epoch": 6.371789645332246, + "grad_norm": 1.027191400527954, + "learning_rate": 0.00020352629433346923, + "loss": 3.4433, + "step": 93780 + }, + { + "epoch": 6.372129365402908, + "grad_norm": 0.9296512603759766, + "learning_rate": 0.0002034838293246365, + "loss": 3.3768, + "step": 93785 + }, + { + "epoch": 6.37246908547357, + "grad_norm": 0.8677007555961609, + "learning_rate": 0.0002034413643158038, + "loss": 3.3521, + "step": 93790 + }, + { + "epoch": 6.372808805544231, + "grad_norm": 0.7557551860809326, + "learning_rate": 0.00020339889930697105, + "loss": 3.525, + "step": 93795 + }, + { + "epoch": 6.373148525614893, + "grad_norm": 0.8307666778564453, + "learning_rate": 0.00020335643429813833, + "loss": 3.232, + "step": 93800 + }, + { + "epoch": 6.3734882456855555, + "grad_norm": 0.8950627446174622, + "learning_rate": 0.00020331396928930563, + "loss": 3.5116, + "step": 93805 + }, + { + "epoch": 6.373827965756217, + "grad_norm": 0.8702715635299683, + "learning_rate": 0.0002032715042804729, + "loss": 3.5914, + "step": 93810 + }, + { + "epoch": 6.374167685826879, + "grad_norm": 1.0832160711288452, + "learning_rate": 0.00020322903927164017, + "loss": 3.4421, + "step": 93815 + }, + { + "epoch": 6.374507405897541, + "grad_norm": 0.7675866484642029, + "learning_rate": 0.00020318657426280745, + "loss": 3.452, + "step": 93820 + }, + { + "epoch": 6.374847125968202, + "grad_norm": 0.9426397681236267, + "learning_rate": 0.00020314410925397473, + "loss": 3.3412, + "step": 93825 + }, + { + "epoch": 6.375186846038864, + "grad_norm": 1.0613439083099365, + "learning_rate": 0.000203101644245142, + "loss": 3.5292, + "step": 93830 + }, + { + "epoch": 6.375526566109526, + "grad_norm": 0.8843908309936523, + "learning_rate": 0.0002030591792363093, + "loss": 3.3265, + "step": 93835 + }, + { + "epoch": 6.375866286180187, + "grad_norm": 0.8755523562431335, + "learning_rate": 0.00020301671422747657, + "loss": 3.4459, + "step": 93840 + }, + { + "epoch": 6.376206006250849, + "grad_norm": 1.0048447847366333, + "learning_rate": 0.00020297424921864385, + "loss": 3.438, + "step": 93845 + }, + { + "epoch": 6.3765457263215115, + "grad_norm": 0.874752938747406, + "learning_rate": 0.00020293178420981113, + "loss": 3.3767, + "step": 93850 + }, + { + "epoch": 6.376885446392173, + "grad_norm": 0.9518517255783081, + "learning_rate": 0.00020288931920097838, + "loss": 3.3378, + "step": 93855 + }, + { + "epoch": 6.377225166462835, + "grad_norm": 0.6359298825263977, + "learning_rate": 0.0002028468541921457, + "loss": 3.6595, + "step": 93860 + }, + { + "epoch": 6.377564886533497, + "grad_norm": 0.930454671382904, + "learning_rate": 0.00020280438918331294, + "loss": 3.3698, + "step": 93865 + }, + { + "epoch": 6.377904606604158, + "grad_norm": 0.8082962036132812, + "learning_rate": 0.00020276192417448022, + "loss": 3.3674, + "step": 93870 + }, + { + "epoch": 6.37824432667482, + "grad_norm": 0.889962375164032, + "learning_rate": 0.00020271945916564753, + "loss": 3.3517, + "step": 93875 + }, + { + "epoch": 6.378584046745482, + "grad_norm": 1.0820988416671753, + "learning_rate": 0.00020267699415681478, + "loss": 3.4093, + "step": 93880 + }, + { + "epoch": 6.378923766816143, + "grad_norm": 1.0812885761260986, + "learning_rate": 0.00020263452914798206, + "loss": 3.1576, + "step": 93885 + }, + { + "epoch": 6.3792634868868054, + "grad_norm": 1.0768516063690186, + "learning_rate": 0.00020259206413914934, + "loss": 3.2329, + "step": 93890 + }, + { + "epoch": 6.379603206957467, + "grad_norm": 0.9297894835472107, + "learning_rate": 0.00020254959913031662, + "loss": 3.0526, + "step": 93895 + }, + { + "epoch": 6.379942927028129, + "grad_norm": 1.0292409658432007, + "learning_rate": 0.00020250713412148388, + "loss": 3.4109, + "step": 93900 + }, + { + "epoch": 6.380282647098791, + "grad_norm": 0.9477608799934387, + "learning_rate": 0.00020246466911265118, + "loss": 3.3807, + "step": 93905 + }, + { + "epoch": 6.380622367169452, + "grad_norm": 0.8944919109344482, + "learning_rate": 0.00020242220410381846, + "loss": 3.066, + "step": 93910 + }, + { + "epoch": 6.380962087240114, + "grad_norm": 0.9279345870018005, + "learning_rate": 0.00020237973909498574, + "loss": 3.6535, + "step": 93915 + }, + { + "epoch": 6.381301807310776, + "grad_norm": 0.9191113710403442, + "learning_rate": 0.00020233727408615302, + "loss": 2.6949, + "step": 93920 + }, + { + "epoch": 6.381641527381437, + "grad_norm": 0.8488415479660034, + "learning_rate": 0.00020229480907732028, + "loss": 3.4084, + "step": 93925 + }, + { + "epoch": 6.381981247452099, + "grad_norm": 0.7851270437240601, + "learning_rate": 0.00020225234406848758, + "loss": 3.6746, + "step": 93930 + }, + { + "epoch": 6.3823209675227615, + "grad_norm": 0.7941996455192566, + "learning_rate": 0.00020220987905965486, + "loss": 3.195, + "step": 93935 + }, + { + "epoch": 6.382660687593423, + "grad_norm": 0.8481007814407349, + "learning_rate": 0.00020216741405082212, + "loss": 3.4492, + "step": 93940 + }, + { + "epoch": 6.383000407664085, + "grad_norm": 0.8437857031822205, + "learning_rate": 0.00020212494904198942, + "loss": 3.4781, + "step": 93945 + }, + { + "epoch": 6.383340127734747, + "grad_norm": 0.9256895184516907, + "learning_rate": 0.00020208248403315668, + "loss": 3.2574, + "step": 93950 + }, + { + "epoch": 6.383679847805408, + "grad_norm": 0.960946261882782, + "learning_rate": 0.00020204001902432396, + "loss": 3.451, + "step": 93955 + }, + { + "epoch": 6.38401956787607, + "grad_norm": 1.1019169092178345, + "learning_rate": 0.00020199755401549124, + "loss": 3.3219, + "step": 93960 + }, + { + "epoch": 6.384359287946732, + "grad_norm": 1.0506983995437622, + "learning_rate": 0.00020195508900665852, + "loss": 3.4946, + "step": 93965 + }, + { + "epoch": 6.384699008017393, + "grad_norm": 1.5714099407196045, + "learning_rate": 0.0002019126239978258, + "loss": 3.6872, + "step": 93970 + }, + { + "epoch": 6.385038728088055, + "grad_norm": 0.8229495286941528, + "learning_rate": 0.00020187015898899308, + "loss": 3.3947, + "step": 93975 + }, + { + "epoch": 6.3853784481587175, + "grad_norm": 0.9928074479103088, + "learning_rate": 0.00020182769398016036, + "loss": 3.4647, + "step": 93980 + }, + { + "epoch": 6.385718168229379, + "grad_norm": 0.9613772630691528, + "learning_rate": 0.0002017852289713276, + "loss": 3.4007, + "step": 93985 + }, + { + "epoch": 6.386057888300041, + "grad_norm": 0.7350130081176758, + "learning_rate": 0.00020174276396249492, + "loss": 3.3665, + "step": 93990 + }, + { + "epoch": 6.386397608370703, + "grad_norm": 1.1244192123413086, + "learning_rate": 0.00020170029895366217, + "loss": 3.287, + "step": 93995 + }, + { + "epoch": 6.386737328441364, + "grad_norm": 0.9433171153068542, + "learning_rate": 0.00020165783394482948, + "loss": 3.2382, + "step": 94000 + }, + { + "epoch": 6.387077048512026, + "grad_norm": 1.0225480794906616, + "learning_rate": 0.00020161536893599676, + "loss": 3.1762, + "step": 94005 + }, + { + "epoch": 6.387416768582688, + "grad_norm": 1.0853766202926636, + "learning_rate": 0.000201572903927164, + "loss": 3.6434, + "step": 94010 + }, + { + "epoch": 6.387756488653349, + "grad_norm": 0.841166079044342, + "learning_rate": 0.00020153043891833132, + "loss": 3.5388, + "step": 94015 + }, + { + "epoch": 6.388096208724011, + "grad_norm": 0.8185452818870544, + "learning_rate": 0.00020148797390949857, + "loss": 3.4021, + "step": 94020 + }, + { + "epoch": 6.3884359287946735, + "grad_norm": 0.8220205307006836, + "learning_rate": 0.00020144550890066585, + "loss": 3.5441, + "step": 94025 + }, + { + "epoch": 6.388775648865335, + "grad_norm": 1.0135780572891235, + "learning_rate": 0.00020140304389183313, + "loss": 3.5098, + "step": 94030 + }, + { + "epoch": 6.389115368935997, + "grad_norm": 1.133387565612793, + "learning_rate": 0.0002013605788830004, + "loss": 3.4509, + "step": 94035 + }, + { + "epoch": 6.389455089006659, + "grad_norm": 0.9448578953742981, + "learning_rate": 0.0002013181138741677, + "loss": 3.5134, + "step": 94040 + }, + { + "epoch": 6.38979480907732, + "grad_norm": 0.8928435444831848, + "learning_rate": 0.00020127564886533497, + "loss": 3.5691, + "step": 94045 + }, + { + "epoch": 6.390134529147982, + "grad_norm": 0.8090105056762695, + "learning_rate": 0.00020123318385650225, + "loss": 3.5787, + "step": 94050 + }, + { + "epoch": 6.390474249218644, + "grad_norm": 0.8108289241790771, + "learning_rate": 0.0002011907188476695, + "loss": 3.1055, + "step": 94055 + }, + { + "epoch": 6.390813969289305, + "grad_norm": 1.0403434038162231, + "learning_rate": 0.0002011482538388368, + "loss": 3.2732, + "step": 94060 + }, + { + "epoch": 6.391153689359967, + "grad_norm": 0.7698134183883667, + "learning_rate": 0.0002011057888300041, + "loss": 3.4272, + "step": 94065 + }, + { + "epoch": 6.3914934094306295, + "grad_norm": 0.9463909268379211, + "learning_rate": 0.00020106332382117134, + "loss": 3.7031, + "step": 94070 + }, + { + "epoch": 6.391833129501291, + "grad_norm": 1.1491811275482178, + "learning_rate": 0.00020102085881233865, + "loss": 3.5149, + "step": 94075 + }, + { + "epoch": 6.392172849571953, + "grad_norm": 0.8719490170478821, + "learning_rate": 0.0002009783938035059, + "loss": 3.532, + "step": 94080 + }, + { + "epoch": 6.392512569642615, + "grad_norm": 0.8672534823417664, + "learning_rate": 0.0002009359287946732, + "loss": 3.5393, + "step": 94085 + }, + { + "epoch": 6.392852289713276, + "grad_norm": 0.8841899037361145, + "learning_rate": 0.00020089346378584046, + "loss": 3.6435, + "step": 94090 + }, + { + "epoch": 6.393192009783938, + "grad_norm": 0.8846684694290161, + "learning_rate": 0.00020085099877700774, + "loss": 3.3087, + "step": 94095 + }, + { + "epoch": 6.3935317298546, + "grad_norm": 0.7281980514526367, + "learning_rate": 0.00020080853376817505, + "loss": 3.3524, + "step": 94100 + }, + { + "epoch": 6.393871449925261, + "grad_norm": 1.019934058189392, + "learning_rate": 0.0002007660687593423, + "loss": 3.5568, + "step": 94105 + }, + { + "epoch": 6.394211169995923, + "grad_norm": 0.72243332862854, + "learning_rate": 0.00020072360375050958, + "loss": 3.5558, + "step": 94110 + }, + { + "epoch": 6.3945508900665855, + "grad_norm": 1.370438814163208, + "learning_rate": 0.00020068113874167686, + "loss": 3.2035, + "step": 94115 + }, + { + "epoch": 6.394890610137247, + "grad_norm": 0.9669160842895508, + "learning_rate": 0.00020063867373284414, + "loss": 3.2901, + "step": 94120 + }, + { + "epoch": 6.395230330207909, + "grad_norm": 1.0289109945297241, + "learning_rate": 0.0002005962087240114, + "loss": 3.3835, + "step": 94125 + }, + { + "epoch": 6.395570050278571, + "grad_norm": 0.9330375790596008, + "learning_rate": 0.0002005537437151787, + "loss": 3.6151, + "step": 94130 + }, + { + "epoch": 6.395909770349232, + "grad_norm": 0.858307421207428, + "learning_rate": 0.00020051127870634598, + "loss": 3.3143, + "step": 94135 + }, + { + "epoch": 6.396249490419894, + "grad_norm": 0.9446781277656555, + "learning_rate": 0.00020046881369751324, + "loss": 3.3237, + "step": 94140 + }, + { + "epoch": 6.396589210490556, + "grad_norm": 1.0568711757659912, + "learning_rate": 0.00020042634868868054, + "loss": 3.2019, + "step": 94145 + }, + { + "epoch": 6.396928930561217, + "grad_norm": 0.884634256362915, + "learning_rate": 0.0002003838836798478, + "loss": 3.4508, + "step": 94150 + }, + { + "epoch": 6.397268650631879, + "grad_norm": 0.9149097204208374, + "learning_rate": 0.00020034141867101508, + "loss": 3.395, + "step": 94155 + }, + { + "epoch": 6.3976083707025415, + "grad_norm": 1.0309501886367798, + "learning_rate": 0.00020029895366218236, + "loss": 3.4189, + "step": 94160 + }, + { + "epoch": 6.397948090773203, + "grad_norm": 0.8115800619125366, + "learning_rate": 0.00020025648865334964, + "loss": 3.2884, + "step": 94165 + }, + { + "epoch": 6.398287810843865, + "grad_norm": 0.823860228061676, + "learning_rate": 0.00020021402364451695, + "loss": 3.4775, + "step": 94170 + }, + { + "epoch": 6.398627530914526, + "grad_norm": 0.8368425369262695, + "learning_rate": 0.0002001715586356842, + "loss": 3.3078, + "step": 94175 + }, + { + "epoch": 6.398967250985188, + "grad_norm": 0.8600596785545349, + "learning_rate": 0.00020012909362685148, + "loss": 3.3654, + "step": 94180 + }, + { + "epoch": 6.39930697105585, + "grad_norm": 0.7504297494888306, + "learning_rate": 0.00020008662861801876, + "loss": 3.3804, + "step": 94185 + }, + { + "epoch": 6.399646691126511, + "grad_norm": 0.8712162375450134, + "learning_rate": 0.00020004416360918604, + "loss": 3.6945, + "step": 94190 + }, + { + "epoch": 6.399986411197173, + "grad_norm": 0.8880426287651062, + "learning_rate": 0.0002000016986003533, + "loss": 3.501, + "step": 94195 + }, + { + "epoch": 6.4003261312678354, + "grad_norm": 0.8655920624732971, + "learning_rate": 0.0001999592335915206, + "loss": 3.3497, + "step": 94200 + }, + { + "epoch": 6.400665851338497, + "grad_norm": 0.9679503440856934, + "learning_rate": 0.00019991676858268788, + "loss": 3.385, + "step": 94205 + }, + { + "epoch": 6.401005571409159, + "grad_norm": 0.9058688879013062, + "learning_rate": 0.00019987430357385513, + "loss": 3.2638, + "step": 94210 + }, + { + "epoch": 6.401345291479821, + "grad_norm": 1.003571629524231, + "learning_rate": 0.00019983183856502244, + "loss": 3.4701, + "step": 94215 + }, + { + "epoch": 6.401685011550482, + "grad_norm": 0.7694059014320374, + "learning_rate": 0.0001997893735561897, + "loss": 3.5808, + "step": 94220 + }, + { + "epoch": 6.402024731621144, + "grad_norm": 1.0196306705474854, + "learning_rate": 0.00019974690854735697, + "loss": 3.2855, + "step": 94225 + }, + { + "epoch": 6.402364451691806, + "grad_norm": 1.01668119430542, + "learning_rate": 0.00019970444353852428, + "loss": 3.2497, + "step": 94230 + }, + { + "epoch": 6.402704171762467, + "grad_norm": 1.2193078994750977, + "learning_rate": 0.00019966197852969153, + "loss": 3.5322, + "step": 94235 + }, + { + "epoch": 6.403043891833129, + "grad_norm": 0.839046835899353, + "learning_rate": 0.0001996195135208588, + "loss": 3.2552, + "step": 94240 + }, + { + "epoch": 6.4033836119037915, + "grad_norm": 1.0382131338119507, + "learning_rate": 0.0001995770485120261, + "loss": 3.2842, + "step": 94245 + }, + { + "epoch": 6.403723331974453, + "grad_norm": 0.9707229733467102, + "learning_rate": 0.00019953458350319337, + "loss": 3.602, + "step": 94250 + }, + { + "epoch": 6.404063052045115, + "grad_norm": 1.0003660917282104, + "learning_rate": 0.00019949211849436065, + "loss": 3.4761, + "step": 94255 + }, + { + "epoch": 6.404402772115777, + "grad_norm": 0.9458271265029907, + "learning_rate": 0.00019944965348552793, + "loss": 3.2583, + "step": 94260 + }, + { + "epoch": 6.404742492186438, + "grad_norm": 0.9275177717208862, + "learning_rate": 0.0001994071884766952, + "loss": 3.4681, + "step": 94265 + }, + { + "epoch": 6.4050822122571, + "grad_norm": 1.6994671821594238, + "learning_rate": 0.0001993647234678625, + "loss": 3.4693, + "step": 94270 + }, + { + "epoch": 6.405421932327762, + "grad_norm": 0.840797483921051, + "learning_rate": 0.00019932225845902977, + "loss": 3.2549, + "step": 94275 + }, + { + "epoch": 6.405761652398423, + "grad_norm": 0.8338384032249451, + "learning_rate": 0.00019927979345019703, + "loss": 3.3768, + "step": 94280 + }, + { + "epoch": 6.406101372469085, + "grad_norm": 0.8536479473114014, + "learning_rate": 0.00019923732844136433, + "loss": 3.4236, + "step": 94285 + }, + { + "epoch": 6.4064410925397475, + "grad_norm": 0.8568475842475891, + "learning_rate": 0.00019919486343253159, + "loss": 3.4768, + "step": 94290 + }, + { + "epoch": 6.406780812610409, + "grad_norm": 0.9291400909423828, + "learning_rate": 0.00019915239842369887, + "loss": 3.2524, + "step": 94295 + }, + { + "epoch": 6.407120532681071, + "grad_norm": 0.9569466710090637, + "learning_rate": 0.00019910993341486617, + "loss": 3.4752, + "step": 94300 + }, + { + "epoch": 6.407460252751733, + "grad_norm": 0.8599242568016052, + "learning_rate": 0.00019906746840603343, + "loss": 3.4168, + "step": 94305 + }, + { + "epoch": 6.407799972822394, + "grad_norm": 1.1547737121582031, + "learning_rate": 0.0001990250033972007, + "loss": 3.218, + "step": 94310 + }, + { + "epoch": 6.408139692893056, + "grad_norm": 0.9053438901901245, + "learning_rate": 0.00019898253838836799, + "loss": 3.2541, + "step": 94315 + }, + { + "epoch": 6.408479412963718, + "grad_norm": 0.9830126762390137, + "learning_rate": 0.00019894007337953527, + "loss": 3.4808, + "step": 94320 + }, + { + "epoch": 6.408819133034379, + "grad_norm": 0.8952814340591431, + "learning_rate": 0.00019889760837070252, + "loss": 3.273, + "step": 94325 + }, + { + "epoch": 6.409158853105041, + "grad_norm": 0.9692132472991943, + "learning_rate": 0.00019885514336186983, + "loss": 3.5366, + "step": 94330 + }, + { + "epoch": 6.4094985731757035, + "grad_norm": 0.9083878993988037, + "learning_rate": 0.0001988126783530371, + "loss": 3.385, + "step": 94335 + }, + { + "epoch": 6.409838293246365, + "grad_norm": 1.1134544610977173, + "learning_rate": 0.00019877021334420439, + "loss": 3.281, + "step": 94340 + }, + { + "epoch": 6.410178013317027, + "grad_norm": 0.8345342874526978, + "learning_rate": 0.00019872774833537167, + "loss": 3.6168, + "step": 94345 + }, + { + "epoch": 6.410517733387689, + "grad_norm": 0.9090805053710938, + "learning_rate": 0.00019868528332653892, + "loss": 3.5559, + "step": 94350 + }, + { + "epoch": 6.41085745345835, + "grad_norm": 0.9290177226066589, + "learning_rate": 0.00019864281831770623, + "loss": 3.3687, + "step": 94355 + }, + { + "epoch": 6.411197173529012, + "grad_norm": 0.8726564645767212, + "learning_rate": 0.0001986003533088735, + "loss": 3.4904, + "step": 94360 + }, + { + "epoch": 6.411536893599674, + "grad_norm": 1.0579537153244019, + "learning_rate": 0.00019855788830004076, + "loss": 3.4182, + "step": 94365 + }, + { + "epoch": 6.411876613670335, + "grad_norm": 0.8563729524612427, + "learning_rate": 0.00019851542329120807, + "loss": 3.2878, + "step": 94370 + }, + { + "epoch": 6.412216333740997, + "grad_norm": 1.0745298862457275, + "learning_rate": 0.00019847295828237532, + "loss": 3.4511, + "step": 94375 + }, + { + "epoch": 6.4125560538116595, + "grad_norm": 0.7604897022247314, + "learning_rate": 0.0001984304932735426, + "loss": 3.4696, + "step": 94380 + }, + { + "epoch": 6.412895773882321, + "grad_norm": 0.7326841950416565, + "learning_rate": 0.00019838802826470988, + "loss": 3.3344, + "step": 94385 + }, + { + "epoch": 6.413235493952983, + "grad_norm": 0.7123711109161377, + "learning_rate": 0.00019834556325587716, + "loss": 3.4377, + "step": 94390 + }, + { + "epoch": 6.413575214023645, + "grad_norm": 1.1824698448181152, + "learning_rate": 0.00019830309824704444, + "loss": 3.3657, + "step": 94395 + }, + { + "epoch": 6.413914934094306, + "grad_norm": 0.7063915133476257, + "learning_rate": 0.00019826063323821172, + "loss": 3.261, + "step": 94400 + }, + { + "epoch": 6.414254654164968, + "grad_norm": 1.0879038572311401, + "learning_rate": 0.000198218168229379, + "loss": 3.4669, + "step": 94405 + }, + { + "epoch": 6.41459437423563, + "grad_norm": 0.7581556439399719, + "learning_rate": 0.00019817570322054625, + "loss": 3.4453, + "step": 94410 + }, + { + "epoch": 6.414934094306291, + "grad_norm": 0.9701844453811646, + "learning_rate": 0.00019813323821171356, + "loss": 3.5449, + "step": 94415 + }, + { + "epoch": 6.415273814376953, + "grad_norm": 1.177895188331604, + "learning_rate": 0.0001980907732028808, + "loss": 3.3933, + "step": 94420 + }, + { + "epoch": 6.4156135344476155, + "grad_norm": 0.8684411644935608, + "learning_rate": 0.00019804830819404812, + "loss": 3.4479, + "step": 94425 + }, + { + "epoch": 6.415953254518277, + "grad_norm": 0.8927183747291565, + "learning_rate": 0.0001980058431852154, + "loss": 3.6145, + "step": 94430 + }, + { + "epoch": 6.416292974588939, + "grad_norm": 0.8759338855743408, + "learning_rate": 0.00019796337817638265, + "loss": 3.131, + "step": 94435 + }, + { + "epoch": 6.416632694659601, + "grad_norm": 0.9693296551704407, + "learning_rate": 0.00019792091316754996, + "loss": 3.3141, + "step": 94440 + }, + { + "epoch": 6.416972414730262, + "grad_norm": 0.9389493465423584, + "learning_rate": 0.00019787844815871721, + "loss": 3.5583, + "step": 94445 + }, + { + "epoch": 6.417312134800924, + "grad_norm": 0.913906455039978, + "learning_rate": 0.0001978359831498845, + "loss": 3.3257, + "step": 94450 + }, + { + "epoch": 6.417651854871586, + "grad_norm": 0.9906608462333679, + "learning_rate": 0.00019779351814105177, + "loss": 3.486, + "step": 94455 + }, + { + "epoch": 6.417991574942247, + "grad_norm": 0.8540285229682922, + "learning_rate": 0.00019775105313221905, + "loss": 3.4777, + "step": 94460 + }, + { + "epoch": 6.418331295012909, + "grad_norm": 1.1110198497772217, + "learning_rate": 0.00019770858812338633, + "loss": 3.4592, + "step": 94465 + }, + { + "epoch": 6.4186710150835715, + "grad_norm": 0.8503866791725159, + "learning_rate": 0.00019766612311455361, + "loss": 3.4121, + "step": 94470 + }, + { + "epoch": 6.419010735154233, + "grad_norm": 0.8738013505935669, + "learning_rate": 0.0001976236581057209, + "loss": 3.2833, + "step": 94475 + }, + { + "epoch": 6.419350455224895, + "grad_norm": 0.8284906148910522, + "learning_rate": 0.00019758119309688815, + "loss": 3.616, + "step": 94480 + }, + { + "epoch": 6.419690175295557, + "grad_norm": 0.8850225210189819, + "learning_rate": 0.00019753872808805545, + "loss": 3.4138, + "step": 94485 + }, + { + "epoch": 6.420029895366218, + "grad_norm": 0.9518073201179504, + "learning_rate": 0.00019749626307922273, + "loss": 3.3692, + "step": 94490 + }, + { + "epoch": 6.42036961543688, + "grad_norm": 0.9258936643600464, + "learning_rate": 0.00019745379807039, + "loss": 3.2634, + "step": 94495 + }, + { + "epoch": 6.420709335507542, + "grad_norm": 1.0055289268493652, + "learning_rate": 0.0001974113330615573, + "loss": 3.5236, + "step": 94500 + }, + { + "epoch": 6.421049055578203, + "grad_norm": 0.9624795913696289, + "learning_rate": 0.00019736886805272455, + "loss": 3.2647, + "step": 94505 + }, + { + "epoch": 6.4213887756488655, + "grad_norm": 0.9884502291679382, + "learning_rate": 0.00019732640304389185, + "loss": 3.0977, + "step": 94510 + }, + { + "epoch": 6.4217284957195275, + "grad_norm": 0.9902674555778503, + "learning_rate": 0.0001972839380350591, + "loss": 3.6251, + "step": 94515 + }, + { + "epoch": 6.422068215790189, + "grad_norm": 0.8017335534095764, + "learning_rate": 0.0001972414730262264, + "loss": 3.2871, + "step": 94520 + }, + { + "epoch": 6.422407935860851, + "grad_norm": 0.829951822757721, + "learning_rate": 0.0001971990080173937, + "loss": 3.1032, + "step": 94525 + }, + { + "epoch": 6.422747655931513, + "grad_norm": 0.8138710856437683, + "learning_rate": 0.00019715654300856095, + "loss": 3.4624, + "step": 94530 + }, + { + "epoch": 6.423087376002174, + "grad_norm": 0.8226576447486877, + "learning_rate": 0.00019711407799972823, + "loss": 3.4828, + "step": 94535 + }, + { + "epoch": 6.423427096072836, + "grad_norm": 0.9317386150360107, + "learning_rate": 0.0001970716129908955, + "loss": 3.5891, + "step": 94540 + }, + { + "epoch": 6.423766816143498, + "grad_norm": 0.7941151857376099, + "learning_rate": 0.0001970291479820628, + "loss": 3.3421, + "step": 94545 + }, + { + "epoch": 6.424106536214159, + "grad_norm": 0.9000148773193359, + "learning_rate": 0.00019698668297323004, + "loss": 3.2123, + "step": 94550 + }, + { + "epoch": 6.4244462562848215, + "grad_norm": 1.0708335638046265, + "learning_rate": 0.00019694421796439735, + "loss": 3.5312, + "step": 94555 + }, + { + "epoch": 6.4247859763554835, + "grad_norm": 0.7891896963119507, + "learning_rate": 0.00019690175295556463, + "loss": 3.6844, + "step": 94560 + }, + { + "epoch": 6.425125696426145, + "grad_norm": 1.101039171218872, + "learning_rate": 0.00019685928794673188, + "loss": 3.453, + "step": 94565 + }, + { + "epoch": 6.425465416496807, + "grad_norm": 0.9672911167144775, + "learning_rate": 0.0001968168229378992, + "loss": 3.4399, + "step": 94570 + }, + { + "epoch": 6.425805136567468, + "grad_norm": 0.9769676327705383, + "learning_rate": 0.00019677435792906644, + "loss": 3.0398, + "step": 94575 + }, + { + "epoch": 6.42614485663813, + "grad_norm": 0.9298335313796997, + "learning_rate": 0.00019673189292023372, + "loss": 3.3332, + "step": 94580 + }, + { + "epoch": 6.426484576708792, + "grad_norm": 0.869197428226471, + "learning_rate": 0.000196689427911401, + "loss": 3.5765, + "step": 94585 + }, + { + "epoch": 6.426824296779453, + "grad_norm": 1.049764633178711, + "learning_rate": 0.00019664696290256828, + "loss": 3.4547, + "step": 94590 + }, + { + "epoch": 6.427164016850115, + "grad_norm": 0.9074998497962952, + "learning_rate": 0.0001966044978937356, + "loss": 3.6607, + "step": 94595 + }, + { + "epoch": 6.4275037369207775, + "grad_norm": 0.955600917339325, + "learning_rate": 0.00019656203288490284, + "loss": 3.5641, + "step": 94600 + }, + { + "epoch": 6.427843456991439, + "grad_norm": 0.9366847276687622, + "learning_rate": 0.00019651956787607012, + "loss": 3.4263, + "step": 94605 + }, + { + "epoch": 6.428183177062101, + "grad_norm": 0.765015721321106, + "learning_rate": 0.0001964771028672374, + "loss": 3.3937, + "step": 94610 + }, + { + "epoch": 6.428522897132763, + "grad_norm": 0.9642559289932251, + "learning_rate": 0.00019643463785840468, + "loss": 3.323, + "step": 94615 + }, + { + "epoch": 6.428862617203424, + "grad_norm": 1.0486899614334106, + "learning_rate": 0.00019639217284957194, + "loss": 3.6157, + "step": 94620 + }, + { + "epoch": 6.429202337274086, + "grad_norm": 1.053512454032898, + "learning_rate": 0.00019634970784073924, + "loss": 3.3851, + "step": 94625 + }, + { + "epoch": 6.429542057344748, + "grad_norm": 0.9264383316040039, + "learning_rate": 0.00019630724283190652, + "loss": 3.7381, + "step": 94630 + }, + { + "epoch": 6.429881777415409, + "grad_norm": 1.2092671394348145, + "learning_rate": 0.00019626477782307378, + "loss": 3.4388, + "step": 94635 + }, + { + "epoch": 6.430221497486071, + "grad_norm": 0.8404533267021179, + "learning_rate": 0.00019622231281424108, + "loss": 3.503, + "step": 94640 + }, + { + "epoch": 6.4305612175567335, + "grad_norm": 1.1857317686080933, + "learning_rate": 0.00019617984780540834, + "loss": 3.4363, + "step": 94645 + }, + { + "epoch": 6.430900937627395, + "grad_norm": 0.9531431794166565, + "learning_rate": 0.00019613738279657562, + "loss": 3.607, + "step": 94650 + }, + { + "epoch": 6.431240657698057, + "grad_norm": 1.0560015439987183, + "learning_rate": 0.00019609491778774292, + "loss": 3.4874, + "step": 94655 + }, + { + "epoch": 6.431580377768719, + "grad_norm": 0.8443254232406616, + "learning_rate": 0.00019605245277891018, + "loss": 3.4091, + "step": 94660 + }, + { + "epoch": 6.43192009783938, + "grad_norm": 0.851940393447876, + "learning_rate": 0.00019600998777007746, + "loss": 3.5828, + "step": 94665 + }, + { + "epoch": 6.432259817910042, + "grad_norm": 0.8949359655380249, + "learning_rate": 0.00019596752276124474, + "loss": 3.3021, + "step": 94670 + }, + { + "epoch": 6.432599537980704, + "grad_norm": 0.802013099193573, + "learning_rate": 0.00019592505775241202, + "loss": 3.7003, + "step": 94675 + }, + { + "epoch": 6.432939258051365, + "grad_norm": 0.8400716781616211, + "learning_rate": 0.0001958825927435793, + "loss": 3.2994, + "step": 94680 + }, + { + "epoch": 6.433278978122027, + "grad_norm": 1.0812069177627563, + "learning_rate": 0.00019584012773474658, + "loss": 3.4401, + "step": 94685 + }, + { + "epoch": 6.4336186981926895, + "grad_norm": 0.9094770550727844, + "learning_rate": 0.00019579766272591386, + "loss": 3.4575, + "step": 94690 + }, + { + "epoch": 6.433958418263351, + "grad_norm": 0.8289616703987122, + "learning_rate": 0.00019575519771708114, + "loss": 3.3892, + "step": 94695 + }, + { + "epoch": 6.434298138334013, + "grad_norm": 0.9435230493545532, + "learning_rate": 0.00019571273270824842, + "loss": 3.2765, + "step": 94700 + }, + { + "epoch": 6.434637858404675, + "grad_norm": 0.9430993795394897, + "learning_rate": 0.00019567026769941567, + "loss": 3.531, + "step": 94705 + }, + { + "epoch": 6.434977578475336, + "grad_norm": 1.0194718837738037, + "learning_rate": 0.00019562780269058298, + "loss": 3.4765, + "step": 94710 + }, + { + "epoch": 6.435317298545998, + "grad_norm": 1.0916880369186401, + "learning_rate": 0.00019558533768175023, + "loss": 3.2493, + "step": 94715 + }, + { + "epoch": 6.43565701861666, + "grad_norm": 1.0255666971206665, + "learning_rate": 0.0001955428726729175, + "loss": 3.5068, + "step": 94720 + }, + { + "epoch": 6.435996738687321, + "grad_norm": 0.8067401051521301, + "learning_rate": 0.00019550040766408482, + "loss": 3.4109, + "step": 94725 + }, + { + "epoch": 6.436336458757983, + "grad_norm": 1.130969524383545, + "learning_rate": 0.00019545794265525207, + "loss": 3.3244, + "step": 94730 + }, + { + "epoch": 6.4366761788286455, + "grad_norm": 0.9053440690040588, + "learning_rate": 0.00019541547764641935, + "loss": 3.3311, + "step": 94735 + }, + { + "epoch": 6.437015898899307, + "grad_norm": 1.0420185327529907, + "learning_rate": 0.00019537301263758663, + "loss": 3.4312, + "step": 94740 + }, + { + "epoch": 6.437355618969969, + "grad_norm": 1.119447112083435, + "learning_rate": 0.0001953305476287539, + "loss": 3.3684, + "step": 94745 + }, + { + "epoch": 6.437695339040631, + "grad_norm": 0.9045239686965942, + "learning_rate": 0.00019528808261992116, + "loss": 3.4766, + "step": 94750 + }, + { + "epoch": 6.438035059111292, + "grad_norm": 1.0380254983901978, + "learning_rate": 0.00019524561761108847, + "loss": 3.4412, + "step": 94755 + }, + { + "epoch": 6.438374779181954, + "grad_norm": 0.7977120280265808, + "learning_rate": 0.00019520315260225575, + "loss": 3.6239, + "step": 94760 + }, + { + "epoch": 6.438714499252616, + "grad_norm": 1.0360493659973145, + "learning_rate": 0.00019516068759342303, + "loss": 3.4699, + "step": 94765 + }, + { + "epoch": 6.439054219323277, + "grad_norm": 1.131766438484192, + "learning_rate": 0.0001951182225845903, + "loss": 3.3391, + "step": 94770 + }, + { + "epoch": 6.4393939393939394, + "grad_norm": 0.7502501606941223, + "learning_rate": 0.00019507575757575756, + "loss": 3.4139, + "step": 94775 + }, + { + "epoch": 6.4397336594646015, + "grad_norm": 0.9408470988273621, + "learning_rate": 0.00019503329256692487, + "loss": 3.4577, + "step": 94780 + }, + { + "epoch": 6.440073379535263, + "grad_norm": 0.9874077439308167, + "learning_rate": 0.00019499082755809215, + "loss": 3.4623, + "step": 94785 + }, + { + "epoch": 6.440413099605925, + "grad_norm": 0.916574239730835, + "learning_rate": 0.0001949483625492594, + "loss": 3.1786, + "step": 94790 + }, + { + "epoch": 6.440752819676587, + "grad_norm": 1.1151025295257568, + "learning_rate": 0.0001949058975404267, + "loss": 3.2429, + "step": 94795 + }, + { + "epoch": 6.441092539747248, + "grad_norm": 0.9891525506973267, + "learning_rate": 0.00019486343253159396, + "loss": 3.4087, + "step": 94800 + }, + { + "epoch": 6.44143225981791, + "grad_norm": 1.0969865322113037, + "learning_rate": 0.00019482096752276124, + "loss": 3.385, + "step": 94805 + }, + { + "epoch": 6.441771979888572, + "grad_norm": 0.6811586618423462, + "learning_rate": 0.00019477850251392852, + "loss": 3.3306, + "step": 94810 + }, + { + "epoch": 6.442111699959233, + "grad_norm": 0.9426028728485107, + "learning_rate": 0.0001947360375050958, + "loss": 3.4868, + "step": 94815 + }, + { + "epoch": 6.4424514200298955, + "grad_norm": 0.9948224425315857, + "learning_rate": 0.00019469357249626308, + "loss": 3.3776, + "step": 94820 + }, + { + "epoch": 6.4427911401005575, + "grad_norm": 0.8450108170509338, + "learning_rate": 0.00019465110748743036, + "loss": 3.3709, + "step": 94825 + }, + { + "epoch": 6.443130860171219, + "grad_norm": 0.9250946044921875, + "learning_rate": 0.00019460864247859764, + "loss": 3.5423, + "step": 94830 + }, + { + "epoch": 6.443470580241881, + "grad_norm": 1.1400418281555176, + "learning_rate": 0.0001945661774697649, + "loss": 3.2293, + "step": 94835 + }, + { + "epoch": 6.443810300312543, + "grad_norm": 1.0790576934814453, + "learning_rate": 0.0001945237124609322, + "loss": 3.5786, + "step": 94840 + }, + { + "epoch": 6.444150020383204, + "grad_norm": 0.9465042948722839, + "learning_rate": 0.00019448124745209946, + "loss": 3.42, + "step": 94845 + }, + { + "epoch": 6.444489740453866, + "grad_norm": 0.8735162615776062, + "learning_rate": 0.00019443878244326676, + "loss": 3.3253, + "step": 94850 + }, + { + "epoch": 6.444829460524527, + "grad_norm": 0.9346691966056824, + "learning_rate": 0.00019439631743443404, + "loss": 3.1273, + "step": 94855 + }, + { + "epoch": 6.445169180595189, + "grad_norm": 1.109722375869751, + "learning_rate": 0.0001943538524256013, + "loss": 3.388, + "step": 94860 + }, + { + "epoch": 6.4455089006658515, + "grad_norm": 0.9078435301780701, + "learning_rate": 0.0001943113874167686, + "loss": 3.6035, + "step": 94865 + }, + { + "epoch": 6.445848620736513, + "grad_norm": 0.7827392220497131, + "learning_rate": 0.00019426892240793586, + "loss": 3.5061, + "step": 94870 + }, + { + "epoch": 6.446188340807175, + "grad_norm": 0.8930590152740479, + "learning_rate": 0.00019422645739910314, + "loss": 3.3889, + "step": 94875 + }, + { + "epoch": 6.446528060877837, + "grad_norm": 0.7855969071388245, + "learning_rate": 0.00019418399239027042, + "loss": 3.4453, + "step": 94880 + }, + { + "epoch": 6.446867780948498, + "grad_norm": 0.9665852189064026, + "learning_rate": 0.0001941415273814377, + "loss": 3.5371, + "step": 94885 + }, + { + "epoch": 6.44720750101916, + "grad_norm": 1.1645790338516235, + "learning_rate": 0.00019409906237260498, + "loss": 3.2318, + "step": 94890 + }, + { + "epoch": 6.447547221089822, + "grad_norm": 0.8539092540740967, + "learning_rate": 0.00019405659736377226, + "loss": 3.3267, + "step": 94895 + }, + { + "epoch": 6.447886941160483, + "grad_norm": 0.9220376014709473, + "learning_rate": 0.00019401413235493954, + "loss": 3.3193, + "step": 94900 + }, + { + "epoch": 6.448226661231145, + "grad_norm": 0.8782495856285095, + "learning_rate": 0.0001939716673461068, + "loss": 3.2406, + "step": 94905 + }, + { + "epoch": 6.4485663813018075, + "grad_norm": 6.23012113571167, + "learning_rate": 0.0001939292023372741, + "loss": 3.4303, + "step": 94910 + }, + { + "epoch": 6.448906101372469, + "grad_norm": 1.0711983442306519, + "learning_rate": 0.00019388673732844138, + "loss": 3.3153, + "step": 94915 + }, + { + "epoch": 6.449245821443131, + "grad_norm": 0.8846371173858643, + "learning_rate": 0.00019384427231960863, + "loss": 3.5846, + "step": 94920 + }, + { + "epoch": 6.449585541513793, + "grad_norm": 0.8021688461303711, + "learning_rate": 0.00019380180731077594, + "loss": 3.2039, + "step": 94925 + }, + { + "epoch": 6.449925261584454, + "grad_norm": 1.0134023427963257, + "learning_rate": 0.0001937593423019432, + "loss": 3.2387, + "step": 94930 + }, + { + "epoch": 6.450264981655116, + "grad_norm": 1.0150816440582275, + "learning_rate": 0.0001937168772931105, + "loss": 3.4593, + "step": 94935 + }, + { + "epoch": 6.450604701725778, + "grad_norm": 1.1467565298080444, + "learning_rate": 0.00019367441228427775, + "loss": 3.601, + "step": 94940 + }, + { + "epoch": 6.450944421796439, + "grad_norm": 0.8157227635383606, + "learning_rate": 0.00019363194727544503, + "loss": 3.42, + "step": 94945 + }, + { + "epoch": 6.451284141867101, + "grad_norm": 0.8637373447418213, + "learning_rate": 0.00019358948226661234, + "loss": 3.4066, + "step": 94950 + }, + { + "epoch": 6.4516238619377635, + "grad_norm": 0.9858822226524353, + "learning_rate": 0.0001935470172577796, + "loss": 3.4702, + "step": 94955 + }, + { + "epoch": 6.451963582008425, + "grad_norm": 0.9928544759750366, + "learning_rate": 0.00019350455224894687, + "loss": 3.4732, + "step": 94960 + }, + { + "epoch": 6.452303302079087, + "grad_norm": 0.9684414267539978, + "learning_rate": 0.00019346208724011415, + "loss": 3.4266, + "step": 94965 + }, + { + "epoch": 6.452643022149749, + "grad_norm": 1.0708383321762085, + "learning_rate": 0.00019341962223128143, + "loss": 3.7005, + "step": 94970 + }, + { + "epoch": 6.45298274222041, + "grad_norm": 0.9552847743034363, + "learning_rate": 0.00019337715722244869, + "loss": 3.3537, + "step": 94975 + }, + { + "epoch": 6.453322462291072, + "grad_norm": 0.8269797563552856, + "learning_rate": 0.000193334692213616, + "loss": 3.189, + "step": 94980 + }, + { + "epoch": 6.453662182361734, + "grad_norm": 1.8431400060653687, + "learning_rate": 0.00019329222720478327, + "loss": 3.38, + "step": 94985 + }, + { + "epoch": 6.454001902432395, + "grad_norm": 0.8474747538566589, + "learning_rate": 0.00019324976219595053, + "loss": 3.3757, + "step": 94990 + }, + { + "epoch": 6.454341622503057, + "grad_norm": 0.9332036972045898, + "learning_rate": 0.00019320729718711783, + "loss": 3.4071, + "step": 94995 + }, + { + "epoch": 6.4546813425737195, + "grad_norm": 0.7695759534835815, + "learning_rate": 0.00019316483217828509, + "loss": 3.1941, + "step": 95000 + }, + { + "epoch": 6.455021062644381, + "grad_norm": 0.8575145602226257, + "learning_rate": 0.00019312236716945237, + "loss": 3.5128, + "step": 95005 + }, + { + "epoch": 6.455360782715043, + "grad_norm": 1.0620280504226685, + "learning_rate": 0.00019307990216061965, + "loss": 3.3793, + "step": 95010 + }, + { + "epoch": 6.455700502785705, + "grad_norm": 0.7979390621185303, + "learning_rate": 0.00019303743715178693, + "loss": 3.3536, + "step": 95015 + }, + { + "epoch": 6.456040222856366, + "grad_norm": 0.9156166315078735, + "learning_rate": 0.00019299497214295423, + "loss": 3.5454, + "step": 95020 + }, + { + "epoch": 6.456379942927028, + "grad_norm": 0.7706505656242371, + "learning_rate": 0.00019295250713412149, + "loss": 3.6619, + "step": 95025 + }, + { + "epoch": 6.45671966299769, + "grad_norm": 0.9085488319396973, + "learning_rate": 0.00019291004212528877, + "loss": 3.358, + "step": 95030 + }, + { + "epoch": 6.457059383068351, + "grad_norm": 0.858361005783081, + "learning_rate": 0.00019286757711645605, + "loss": 3.2603, + "step": 95035 + }, + { + "epoch": 6.457399103139013, + "grad_norm": 1.1374613046646118, + "learning_rate": 0.00019282511210762333, + "loss": 3.3747, + "step": 95040 + }, + { + "epoch": 6.4577388232096755, + "grad_norm": 0.7841351628303528, + "learning_rate": 0.00019278264709879058, + "loss": 3.3399, + "step": 95045 + }, + { + "epoch": 6.458078543280337, + "grad_norm": 0.8317552804946899, + "learning_rate": 0.00019274018208995789, + "loss": 3.4064, + "step": 95050 + }, + { + "epoch": 6.458418263350999, + "grad_norm": 1.1384576559066772, + "learning_rate": 0.00019269771708112517, + "loss": 3.0975, + "step": 95055 + }, + { + "epoch": 6.458757983421661, + "grad_norm": 1.0989350080490112, + "learning_rate": 0.00019265525207229242, + "loss": 3.6299, + "step": 95060 + }, + { + "epoch": 6.459097703492322, + "grad_norm": 0.8346503376960754, + "learning_rate": 0.00019261278706345973, + "loss": 3.3691, + "step": 95065 + }, + { + "epoch": 6.459437423562984, + "grad_norm": 0.9950863122940063, + "learning_rate": 0.00019257032205462698, + "loss": 3.2859, + "step": 95070 + }, + { + "epoch": 6.459777143633646, + "grad_norm": 1.0674313306808472, + "learning_rate": 0.00019252785704579426, + "loss": 3.6398, + "step": 95075 + }, + { + "epoch": 6.460116863704307, + "grad_norm": 1.002909779548645, + "learning_rate": 0.00019248539203696157, + "loss": 3.626, + "step": 95080 + }, + { + "epoch": 6.4604565837749695, + "grad_norm": 0.9905146956443787, + "learning_rate": 0.00019244292702812882, + "loss": 3.3864, + "step": 95085 + }, + { + "epoch": 6.4607963038456315, + "grad_norm": 0.9290003776550293, + "learning_rate": 0.0001924004620192961, + "loss": 3.5311, + "step": 95090 + }, + { + "epoch": 6.461136023916293, + "grad_norm": 0.9002822041511536, + "learning_rate": 0.00019235799701046338, + "loss": 3.4796, + "step": 95095 + }, + { + "epoch": 6.461475743986955, + "grad_norm": 0.8358800411224365, + "learning_rate": 0.00019231553200163066, + "loss": 3.2546, + "step": 95100 + }, + { + "epoch": 6.461815464057617, + "grad_norm": 0.9306395649909973, + "learning_rate": 0.00019227306699279794, + "loss": 3.2798, + "step": 95105 + }, + { + "epoch": 6.462155184128278, + "grad_norm": 0.8266304731369019, + "learning_rate": 0.00019223060198396522, + "loss": 3.1465, + "step": 95110 + }, + { + "epoch": 6.46249490419894, + "grad_norm": 0.7120440602302551, + "learning_rate": 0.0001921881369751325, + "loss": 3.5627, + "step": 95115 + }, + { + "epoch": 6.462834624269602, + "grad_norm": 0.7785124182701111, + "learning_rate": 0.00019214567196629978, + "loss": 3.3027, + "step": 95120 + }, + { + "epoch": 6.463174344340263, + "grad_norm": 0.7845525145530701, + "learning_rate": 0.00019210320695746706, + "loss": 3.6024, + "step": 95125 + }, + { + "epoch": 6.4635140644109255, + "grad_norm": 0.9143544435501099, + "learning_rate": 0.0001920607419486343, + "loss": 3.3432, + "step": 95130 + }, + { + "epoch": 6.4638537844815875, + "grad_norm": 0.8623539209365845, + "learning_rate": 0.00019201827693980162, + "loss": 3.247, + "step": 95135 + }, + { + "epoch": 6.464193504552249, + "grad_norm": 1.0073165893554688, + "learning_rate": 0.00019197581193096887, + "loss": 3.5771, + "step": 95140 + }, + { + "epoch": 6.464533224622911, + "grad_norm": 0.7703071236610413, + "learning_rate": 0.00019193334692213615, + "loss": 3.5843, + "step": 95145 + }, + { + "epoch": 6.464872944693573, + "grad_norm": 0.7825093865394592, + "learning_rate": 0.00019189088191330346, + "loss": 3.5386, + "step": 95150 + }, + { + "epoch": 6.465212664764234, + "grad_norm": 0.7102939486503601, + "learning_rate": 0.00019184841690447071, + "loss": 3.4272, + "step": 95155 + }, + { + "epoch": 6.465552384834896, + "grad_norm": 1.005266785621643, + "learning_rate": 0.000191805951895638, + "loss": 3.5465, + "step": 95160 + }, + { + "epoch": 6.465892104905558, + "grad_norm": 1.029145359992981, + "learning_rate": 0.00019176348688680527, + "loss": 3.3546, + "step": 95165 + }, + { + "epoch": 6.466231824976219, + "grad_norm": 0.8069216012954712, + "learning_rate": 0.00019172102187797255, + "loss": 3.3391, + "step": 95170 + }, + { + "epoch": 6.4665715450468815, + "grad_norm": 0.8937439322471619, + "learning_rate": 0.0001916785568691398, + "loss": 3.5037, + "step": 95175 + }, + { + "epoch": 6.4669112651175436, + "grad_norm": 1.2379796504974365, + "learning_rate": 0.00019163609186030711, + "loss": 3.2405, + "step": 95180 + }, + { + "epoch": 6.467250985188205, + "grad_norm": 0.8497258424758911, + "learning_rate": 0.0001915936268514744, + "loss": 3.2973, + "step": 95185 + }, + { + "epoch": 6.467590705258867, + "grad_norm": 1.0921845436096191, + "learning_rate": 0.00019155116184264167, + "loss": 3.386, + "step": 95190 + }, + { + "epoch": 6.467930425329529, + "grad_norm": 1.0079916715621948, + "learning_rate": 0.00019150869683380895, + "loss": 3.4616, + "step": 95195 + }, + { + "epoch": 6.46827014540019, + "grad_norm": 1.0753599405288696, + "learning_rate": 0.0001914662318249762, + "loss": 3.6156, + "step": 95200 + }, + { + "epoch": 6.468609865470852, + "grad_norm": 0.9207011461257935, + "learning_rate": 0.00019142376681614351, + "loss": 3.4517, + "step": 95205 + }, + { + "epoch": 6.468949585541514, + "grad_norm": 1.1252244710922241, + "learning_rate": 0.0001913813018073108, + "loss": 3.3694, + "step": 95210 + }, + { + "epoch": 6.469289305612175, + "grad_norm": 0.9290748238563538, + "learning_rate": 0.00019133883679847805, + "loss": 3.418, + "step": 95215 + }, + { + "epoch": 6.4696290256828375, + "grad_norm": 1.0528316497802734, + "learning_rate": 0.00019129637178964535, + "loss": 3.3088, + "step": 95220 + }, + { + "epoch": 6.4699687457535, + "grad_norm": 1.0448590517044067, + "learning_rate": 0.0001912539067808126, + "loss": 3.2162, + "step": 95225 + }, + { + "epoch": 6.470308465824161, + "grad_norm": 0.9971960783004761, + "learning_rate": 0.0001912114417719799, + "loss": 3.6543, + "step": 95230 + }, + { + "epoch": 6.470648185894823, + "grad_norm": 0.9716728329658508, + "learning_rate": 0.00019116897676314717, + "loss": 3.8108, + "step": 95235 + }, + { + "epoch": 6.470987905965485, + "grad_norm": 0.8605753183364868, + "learning_rate": 0.00019112651175431445, + "loss": 3.6845, + "step": 95240 + }, + { + "epoch": 6.471327626036146, + "grad_norm": 1.0101613998413086, + "learning_rate": 0.00019108404674548173, + "loss": 3.6752, + "step": 95245 + }, + { + "epoch": 6.471667346106808, + "grad_norm": 0.8103004097938538, + "learning_rate": 0.000191041581736649, + "loss": 3.3529, + "step": 95250 + }, + { + "epoch": 6.47200706617747, + "grad_norm": 1.1060494184494019, + "learning_rate": 0.0001909991167278163, + "loss": 3.3827, + "step": 95255 + }, + { + "epoch": 6.472346786248131, + "grad_norm": 0.8938196301460266, + "learning_rate": 0.00019095665171898354, + "loss": 3.6132, + "step": 95260 + }, + { + "epoch": 6.4726865063187935, + "grad_norm": 0.752816379070282, + "learning_rate": 0.00019091418671015085, + "loss": 3.2712, + "step": 95265 + }, + { + "epoch": 6.473026226389455, + "grad_norm": 0.7576449513435364, + "learning_rate": 0.0001908717217013181, + "loss": 3.401, + "step": 95270 + }, + { + "epoch": 6.473365946460117, + "grad_norm": 0.8788657784461975, + "learning_rate": 0.0001908292566924854, + "loss": 3.253, + "step": 95275 + }, + { + "epoch": 6.473705666530779, + "grad_norm": 0.8597740530967712, + "learning_rate": 0.0001907867916836527, + "loss": 3.6346, + "step": 95280 + }, + { + "epoch": 6.47404538660144, + "grad_norm": 0.7713863849639893, + "learning_rate": 0.00019074432667481994, + "loss": 3.3617, + "step": 95285 + }, + { + "epoch": 6.474385106672102, + "grad_norm": 1.0054175853729248, + "learning_rate": 0.00019070186166598725, + "loss": 3.4897, + "step": 95290 + }, + { + "epoch": 6.474724826742764, + "grad_norm": 1.0332468748092651, + "learning_rate": 0.0001906593966571545, + "loss": 3.2662, + "step": 95295 + }, + { + "epoch": 6.475064546813425, + "grad_norm": 0.9180460572242737, + "learning_rate": 0.00019061693164832178, + "loss": 3.4336, + "step": 95300 + }, + { + "epoch": 6.475404266884087, + "grad_norm": 0.6963275671005249, + "learning_rate": 0.00019057446663948906, + "loss": 3.5349, + "step": 95305 + }, + { + "epoch": 6.4757439869547495, + "grad_norm": 0.8127650022506714, + "learning_rate": 0.00019053200163065634, + "loss": 3.6239, + "step": 95310 + }, + { + "epoch": 6.476083707025411, + "grad_norm": 0.827316164970398, + "learning_rate": 0.00019048953662182362, + "loss": 3.3207, + "step": 95315 + }, + { + "epoch": 6.476423427096073, + "grad_norm": 0.864655613899231, + "learning_rate": 0.0001904470716129909, + "loss": 3.3099, + "step": 95320 + }, + { + "epoch": 6.476763147166735, + "grad_norm": 0.896900475025177, + "learning_rate": 0.00019040460660415818, + "loss": 3.2888, + "step": 95325 + }, + { + "epoch": 6.477102867237396, + "grad_norm": 0.8422818183898926, + "learning_rate": 0.00019036214159532544, + "loss": 3.3993, + "step": 95330 + }, + { + "epoch": 6.477442587308058, + "grad_norm": 0.7142874002456665, + "learning_rate": 0.00019031967658649274, + "loss": 3.593, + "step": 95335 + }, + { + "epoch": 6.47778230737872, + "grad_norm": 1.0662217140197754, + "learning_rate": 0.00019027721157766, + "loss": 3.3577, + "step": 95340 + }, + { + "epoch": 6.478122027449381, + "grad_norm": 0.9148504137992859, + "learning_rate": 0.00019023474656882728, + "loss": 3.3646, + "step": 95345 + }, + { + "epoch": 6.4784617475200434, + "grad_norm": 0.733856201171875, + "learning_rate": 0.00019019228155999458, + "loss": 3.4268, + "step": 95350 + }, + { + "epoch": 6.4788014675907055, + "grad_norm": 1.1984648704528809, + "learning_rate": 0.00019014981655116184, + "loss": 3.2778, + "step": 95355 + }, + { + "epoch": 6.479141187661367, + "grad_norm": 1.173652172088623, + "learning_rate": 0.00019010735154232914, + "loss": 3.1849, + "step": 95360 + }, + { + "epoch": 6.479480907732029, + "grad_norm": 0.9604526162147522, + "learning_rate": 0.0001900648865334964, + "loss": 3.2485, + "step": 95365 + }, + { + "epoch": 6.479820627802691, + "grad_norm": 1.0382074117660522, + "learning_rate": 0.00019002242152466368, + "loss": 3.461, + "step": 95370 + }, + { + "epoch": 6.480160347873352, + "grad_norm": 1.1888231039047241, + "learning_rate": 0.00018997995651583098, + "loss": 2.9572, + "step": 95375 + }, + { + "epoch": 6.480500067944014, + "grad_norm": 0.8243505358695984, + "learning_rate": 0.00018993749150699824, + "loss": 3.5724, + "step": 95380 + }, + { + "epoch": 6.480839788014676, + "grad_norm": 0.7733266949653625, + "learning_rate": 0.00018989502649816552, + "loss": 3.4638, + "step": 95385 + }, + { + "epoch": 6.481179508085337, + "grad_norm": 1.0880565643310547, + "learning_rate": 0.0001898525614893328, + "loss": 3.4037, + "step": 95390 + }, + { + "epoch": 6.4815192281559995, + "grad_norm": 0.811475396156311, + "learning_rate": 0.00018981009648050008, + "loss": 3.3227, + "step": 95395 + }, + { + "epoch": 6.4818589482266615, + "grad_norm": 1.081324815750122, + "learning_rate": 0.00018976763147166733, + "loss": 3.5735, + "step": 95400 + }, + { + "epoch": 6.482198668297323, + "grad_norm": 0.9490269422531128, + "learning_rate": 0.00018972516646283464, + "loss": 3.4608, + "step": 95405 + }, + { + "epoch": 6.482538388367985, + "grad_norm": 0.8235070109367371, + "learning_rate": 0.00018968270145400192, + "loss": 3.5049, + "step": 95410 + }, + { + "epoch": 6.482878108438647, + "grad_norm": 0.9324278831481934, + "learning_rate": 0.00018964023644516917, + "loss": 3.5182, + "step": 95415 + }, + { + "epoch": 6.483217828509308, + "grad_norm": 0.9732624888420105, + "learning_rate": 0.00018959777143633648, + "loss": 3.5709, + "step": 95420 + }, + { + "epoch": 6.48355754857997, + "grad_norm": 0.7981487512588501, + "learning_rate": 0.00018955530642750373, + "loss": 3.3477, + "step": 95425 + }, + { + "epoch": 6.483897268650632, + "grad_norm": 0.9485097527503967, + "learning_rate": 0.000189512841418671, + "loss": 3.3976, + "step": 95430 + }, + { + "epoch": 6.484236988721293, + "grad_norm": 1.074361801147461, + "learning_rate": 0.0001894703764098383, + "loss": 3.4917, + "step": 95435 + }, + { + "epoch": 6.4845767087919555, + "grad_norm": 0.8492256999015808, + "learning_rate": 0.00018942791140100557, + "loss": 3.3628, + "step": 95440 + }, + { + "epoch": 6.4849164288626175, + "grad_norm": 0.884450376033783, + "learning_rate": 0.00018938544639217288, + "loss": 3.4664, + "step": 95445 + }, + { + "epoch": 6.485256148933279, + "grad_norm": 1.1263154745101929, + "learning_rate": 0.00018934298138334013, + "loss": 3.5433, + "step": 95450 + }, + { + "epoch": 6.485595869003941, + "grad_norm": 0.8710302114486694, + "learning_rate": 0.0001893005163745074, + "loss": 3.3598, + "step": 95455 + }, + { + "epoch": 6.485935589074603, + "grad_norm": 0.8459093570709229, + "learning_rate": 0.0001892580513656747, + "loss": 3.497, + "step": 95460 + }, + { + "epoch": 6.486275309145264, + "grad_norm": 1.0175800323486328, + "learning_rate": 0.00018921558635684197, + "loss": 3.6753, + "step": 95465 + }, + { + "epoch": 6.486615029215926, + "grad_norm": 0.8568399548530579, + "learning_rate": 0.00018917312134800922, + "loss": 3.5451, + "step": 95470 + }, + { + "epoch": 6.486954749286588, + "grad_norm": 1.0013428926467896, + "learning_rate": 0.00018913065633917653, + "loss": 3.6358, + "step": 95475 + }, + { + "epoch": 6.487294469357249, + "grad_norm": 0.8600124716758728, + "learning_rate": 0.0001890881913303438, + "loss": 3.2178, + "step": 95480 + }, + { + "epoch": 6.4876341894279115, + "grad_norm": 0.956333339214325, + "learning_rate": 0.00018904572632151106, + "loss": 3.556, + "step": 95485 + }, + { + "epoch": 6.4879739094985736, + "grad_norm": 0.9985502362251282, + "learning_rate": 0.00018900326131267837, + "loss": 3.486, + "step": 95490 + }, + { + "epoch": 6.488313629569235, + "grad_norm": 0.7734472155570984, + "learning_rate": 0.00018896079630384562, + "loss": 3.3437, + "step": 95495 + }, + { + "epoch": 6.488653349639897, + "grad_norm": 1.0701208114624023, + "learning_rate": 0.0001889183312950129, + "loss": 3.3698, + "step": 95500 + }, + { + "epoch": 6.488993069710559, + "grad_norm": 0.9310901761054993, + "learning_rate": 0.0001888758662861802, + "loss": 3.4246, + "step": 95505 + }, + { + "epoch": 6.48933278978122, + "grad_norm": 0.7174954414367676, + "learning_rate": 0.00018883340127734746, + "loss": 3.332, + "step": 95510 + }, + { + "epoch": 6.489672509851882, + "grad_norm": 0.91680508852005, + "learning_rate": 0.00018879093626851474, + "loss": 3.5775, + "step": 95515 + }, + { + "epoch": 6.490012229922544, + "grad_norm": 0.8848182559013367, + "learning_rate": 0.00018874847125968202, + "loss": 3.6643, + "step": 95520 + }, + { + "epoch": 6.490351949993205, + "grad_norm": 1.1256928443908691, + "learning_rate": 0.0001887060062508493, + "loss": 3.1881, + "step": 95525 + }, + { + "epoch": 6.4906916700638675, + "grad_norm": 0.8040698766708374, + "learning_rate": 0.00018866354124201658, + "loss": 3.2495, + "step": 95530 + }, + { + "epoch": 6.491031390134529, + "grad_norm": 0.920427143573761, + "learning_rate": 0.00018862107623318386, + "loss": 3.1746, + "step": 95535 + }, + { + "epoch": 6.491371110205191, + "grad_norm": 0.9053266644477844, + "learning_rate": 0.00018857861122435114, + "loss": 3.6003, + "step": 95540 + }, + { + "epoch": 6.491710830275853, + "grad_norm": 1.1021244525909424, + "learning_rate": 0.00018853614621551842, + "loss": 3.3476, + "step": 95545 + }, + { + "epoch": 6.492050550346514, + "grad_norm": 0.8730857968330383, + "learning_rate": 0.0001884936812066857, + "loss": 3.186, + "step": 95550 + }, + { + "epoch": 6.492390270417176, + "grad_norm": 0.8883584141731262, + "learning_rate": 0.00018845121619785296, + "loss": 3.4165, + "step": 95555 + }, + { + "epoch": 6.492729990487838, + "grad_norm": 1.0450048446655273, + "learning_rate": 0.00018840875118902026, + "loss": 3.4978, + "step": 95560 + }, + { + "epoch": 6.493069710558499, + "grad_norm": 0.8854564428329468, + "learning_rate": 0.00018836628618018752, + "loss": 3.587, + "step": 95565 + }, + { + "epoch": 6.493409430629161, + "grad_norm": 1.1046572923660278, + "learning_rate": 0.0001883238211713548, + "loss": 3.2816, + "step": 95570 + }, + { + "epoch": 6.4937491506998235, + "grad_norm": 0.809381902217865, + "learning_rate": 0.0001882813561625221, + "loss": 3.5675, + "step": 95575 + }, + { + "epoch": 6.494088870770485, + "grad_norm": 1.047095775604248, + "learning_rate": 0.00018823889115368936, + "loss": 3.3445, + "step": 95580 + }, + { + "epoch": 6.494428590841147, + "grad_norm": 0.9509291648864746, + "learning_rate": 0.00018819642614485664, + "loss": 3.4342, + "step": 95585 + }, + { + "epoch": 6.494768310911809, + "grad_norm": 1.0215983390808105, + "learning_rate": 0.00018815396113602392, + "loss": 3.2021, + "step": 95590 + }, + { + "epoch": 6.49510803098247, + "grad_norm": 0.7788615822792053, + "learning_rate": 0.0001881114961271912, + "loss": 3.2194, + "step": 95595 + }, + { + "epoch": 6.495447751053132, + "grad_norm": 0.7209200859069824, + "learning_rate": 0.00018806903111835845, + "loss": 3.3303, + "step": 95600 + }, + { + "epoch": 6.495787471123794, + "grad_norm": 0.8314764499664307, + "learning_rate": 0.00018802656610952576, + "loss": 3.3451, + "step": 95605 + }, + { + "epoch": 6.496127191194455, + "grad_norm": 1.0790159702301025, + "learning_rate": 0.00018798410110069304, + "loss": 3.3196, + "step": 95610 + }, + { + "epoch": 6.496466911265117, + "grad_norm": 0.6468613743782043, + "learning_rate": 0.00018794163609186032, + "loss": 3.5733, + "step": 95615 + }, + { + "epoch": 6.4968066313357795, + "grad_norm": 0.8813697099685669, + "learning_rate": 0.0001878991710830276, + "loss": 3.5026, + "step": 95620 + }, + { + "epoch": 6.497146351406441, + "grad_norm": 0.7603367567062378, + "learning_rate": 0.00018785670607419485, + "loss": 3.4674, + "step": 95625 + }, + { + "epoch": 6.497486071477103, + "grad_norm": 0.932537317276001, + "learning_rate": 0.00018781424106536216, + "loss": 3.5178, + "step": 95630 + }, + { + "epoch": 6.497825791547765, + "grad_norm": 0.7723372578620911, + "learning_rate": 0.00018777177605652944, + "loss": 3.6392, + "step": 95635 + }, + { + "epoch": 6.498165511618426, + "grad_norm": 0.8769289255142212, + "learning_rate": 0.0001877293110476967, + "loss": 3.3713, + "step": 95640 + }, + { + "epoch": 6.498505231689088, + "grad_norm": 0.828498125076294, + "learning_rate": 0.000187686846038864, + "loss": 3.4371, + "step": 95645 + }, + { + "epoch": 6.49884495175975, + "grad_norm": 1.1076728105545044, + "learning_rate": 0.00018764438103003125, + "loss": 3.6802, + "step": 95650 + }, + { + "epoch": 6.499184671830411, + "grad_norm": 0.9243494868278503, + "learning_rate": 0.00018760191602119853, + "loss": 3.5227, + "step": 95655 + }, + { + "epoch": 6.4995243919010735, + "grad_norm": 0.995291531085968, + "learning_rate": 0.0001875594510123658, + "loss": 3.3456, + "step": 95660 + }, + { + "epoch": 6.4998641119717355, + "grad_norm": 0.9217178225517273, + "learning_rate": 0.0001875169860035331, + "loss": 3.5933, + "step": 95665 + }, + { + "epoch": 6.500203832042397, + "grad_norm": 1.4218069314956665, + "learning_rate": 0.00018747452099470037, + "loss": 3.4191, + "step": 95670 + }, + { + "epoch": 6.500543552113059, + "grad_norm": 1.3256723880767822, + "learning_rate": 0.00018743205598586765, + "loss": 3.5083, + "step": 95675 + }, + { + "epoch": 6.500883272183721, + "grad_norm": 0.9958829283714294, + "learning_rate": 0.00018738959097703493, + "loss": 3.007, + "step": 95680 + }, + { + "epoch": 6.501222992254382, + "grad_norm": 0.7901955842971802, + "learning_rate": 0.00018734712596820219, + "loss": 3.2928, + "step": 95685 + }, + { + "epoch": 6.501562712325044, + "grad_norm": 0.8954638838768005, + "learning_rate": 0.0001873046609593695, + "loss": 3.5434, + "step": 95690 + }, + { + "epoch": 6.501902432395706, + "grad_norm": 0.927442193031311, + "learning_rate": 0.00018726219595053675, + "loss": 3.5184, + "step": 95695 + }, + { + "epoch": 6.502242152466367, + "grad_norm": 0.8968055248260498, + "learning_rate": 0.00018721973094170405, + "loss": 3.5824, + "step": 95700 + }, + { + "epoch": 6.5025818725370295, + "grad_norm": 1.0732568502426147, + "learning_rate": 0.00018717726593287133, + "loss": 3.5274, + "step": 95705 + }, + { + "epoch": 6.5029215926076915, + "grad_norm": 0.8949770331382751, + "learning_rate": 0.00018713480092403859, + "loss": 3.4575, + "step": 95710 + }, + { + "epoch": 6.503261312678353, + "grad_norm": 0.9830872416496277, + "learning_rate": 0.0001870923359152059, + "loss": 3.3577, + "step": 95715 + }, + { + "epoch": 6.503601032749015, + "grad_norm": 1.0713967084884644, + "learning_rate": 0.00018704987090637315, + "loss": 3.4814, + "step": 95720 + }, + { + "epoch": 6.503940752819677, + "grad_norm": 0.8332688808441162, + "learning_rate": 0.00018700740589754043, + "loss": 3.5869, + "step": 95725 + }, + { + "epoch": 6.504280472890338, + "grad_norm": 1.1593533754348755, + "learning_rate": 0.0001869649408887077, + "loss": 3.4487, + "step": 95730 + }, + { + "epoch": 6.504620192961, + "grad_norm": 0.758775532245636, + "learning_rate": 0.00018692247587987499, + "loss": 3.2661, + "step": 95735 + }, + { + "epoch": 6.504959913031662, + "grad_norm": 0.874485433101654, + "learning_rate": 0.00018688001087104227, + "loss": 3.4712, + "step": 95740 + }, + { + "epoch": 6.505299633102323, + "grad_norm": 0.8545040488243103, + "learning_rate": 0.00018683754586220955, + "loss": 3.5125, + "step": 95745 + }, + { + "epoch": 6.5056393531729855, + "grad_norm": 1.0239518880844116, + "learning_rate": 0.00018679508085337683, + "loss": 3.2425, + "step": 95750 + }, + { + "epoch": 6.5059790732436475, + "grad_norm": 1.364113450050354, + "learning_rate": 0.00018675261584454408, + "loss": 3.4685, + "step": 95755 + }, + { + "epoch": 6.506318793314309, + "grad_norm": 0.9102060198783875, + "learning_rate": 0.00018671015083571139, + "loss": 3.4737, + "step": 95760 + }, + { + "epoch": 6.506658513384971, + "grad_norm": 0.8163856863975525, + "learning_rate": 0.00018666768582687864, + "loss": 3.4305, + "step": 95765 + }, + { + "epoch": 6.506998233455633, + "grad_norm": 0.9099118113517761, + "learning_rate": 0.00018662522081804592, + "loss": 3.6705, + "step": 95770 + }, + { + "epoch": 6.507337953526294, + "grad_norm": 0.9486088156700134, + "learning_rate": 0.00018658275580921323, + "loss": 3.3578, + "step": 95775 + }, + { + "epoch": 6.507677673596956, + "grad_norm": 1.1574320793151855, + "learning_rate": 0.00018654029080038048, + "loss": 3.6332, + "step": 95780 + }, + { + "epoch": 6.508017393667618, + "grad_norm": 0.6993726491928101, + "learning_rate": 0.0001864978257915478, + "loss": 3.656, + "step": 95785 + }, + { + "epoch": 6.508357113738279, + "grad_norm": 0.8355573415756226, + "learning_rate": 0.00018645536078271504, + "loss": 3.3977, + "step": 95790 + }, + { + "epoch": 6.5086968338089415, + "grad_norm": 0.9206762313842773, + "learning_rate": 0.00018641289577388232, + "loss": 3.393, + "step": 95795 + }, + { + "epoch": 6.509036553879604, + "grad_norm": 0.7680445313453674, + "learning_rate": 0.00018637043076504963, + "loss": 3.3506, + "step": 95800 + }, + { + "epoch": 6.509376273950265, + "grad_norm": 1.094101071357727, + "learning_rate": 0.00018632796575621688, + "loss": 3.3931, + "step": 95805 + }, + { + "epoch": 6.509715994020927, + "grad_norm": 1.0647644996643066, + "learning_rate": 0.00018628550074738416, + "loss": 3.3738, + "step": 95810 + }, + { + "epoch": 6.510055714091589, + "grad_norm": 0.793809175491333, + "learning_rate": 0.00018624303573855144, + "loss": 3.3349, + "step": 95815 + }, + { + "epoch": 6.51039543416225, + "grad_norm": 0.856489360332489, + "learning_rate": 0.00018620057072971872, + "loss": 3.6436, + "step": 95820 + }, + { + "epoch": 6.510735154232912, + "grad_norm": 0.863967776298523, + "learning_rate": 0.00018615810572088597, + "loss": 3.2587, + "step": 95825 + }, + { + "epoch": 6.511074874303574, + "grad_norm": 0.9875862002372742, + "learning_rate": 0.00018611564071205328, + "loss": 3.3778, + "step": 95830 + }, + { + "epoch": 6.511414594374235, + "grad_norm": 0.7928885221481323, + "learning_rate": 0.00018607317570322056, + "loss": 3.3062, + "step": 95835 + }, + { + "epoch": 6.5117543144448975, + "grad_norm": 1.1737507581710815, + "learning_rate": 0.0001860307106943878, + "loss": 3.3303, + "step": 95840 + }, + { + "epoch": 6.51209403451556, + "grad_norm": 0.7638118863105774, + "learning_rate": 0.00018598824568555512, + "loss": 3.4606, + "step": 95845 + }, + { + "epoch": 6.512433754586221, + "grad_norm": 0.728661060333252, + "learning_rate": 0.00018594578067672237, + "loss": 3.6624, + "step": 95850 + }, + { + "epoch": 6.512773474656883, + "grad_norm": 1.0224533081054688, + "learning_rate": 0.00018590331566788965, + "loss": 3.4587, + "step": 95855 + }, + { + "epoch": 6.513113194727545, + "grad_norm": 1.0485289096832275, + "learning_rate": 0.00018586085065905693, + "loss": 3.376, + "step": 95860 + }, + { + "epoch": 6.513452914798206, + "grad_norm": 0.860866367816925, + "learning_rate": 0.00018581838565022421, + "loss": 3.2909, + "step": 95865 + }, + { + "epoch": 6.513792634868868, + "grad_norm": 0.855871319770813, + "learning_rate": 0.00018577592064139152, + "loss": 3.3095, + "step": 95870 + }, + { + "epoch": 6.51413235493953, + "grad_norm": 0.9089747071266174, + "learning_rate": 0.00018573345563255877, + "loss": 3.3029, + "step": 95875 + }, + { + "epoch": 6.514472075010191, + "grad_norm": 0.9991668462753296, + "learning_rate": 0.00018569099062372605, + "loss": 3.5482, + "step": 95880 + }, + { + "epoch": 6.5148117950808535, + "grad_norm": 0.9563205242156982, + "learning_rate": 0.00018564852561489333, + "loss": 3.242, + "step": 95885 + }, + { + "epoch": 6.515151515151516, + "grad_norm": 0.9731966853141785, + "learning_rate": 0.00018560606060606061, + "loss": 3.545, + "step": 95890 + }, + { + "epoch": 6.515491235222177, + "grad_norm": 1.108031988143921, + "learning_rate": 0.00018556359559722787, + "loss": 3.4671, + "step": 95895 + }, + { + "epoch": 6.515830955292839, + "grad_norm": 0.8586388826370239, + "learning_rate": 0.00018552113058839517, + "loss": 3.6352, + "step": 95900 + }, + { + "epoch": 6.516170675363501, + "grad_norm": 3.8086116313934326, + "learning_rate": 0.00018547866557956245, + "loss": 3.3686, + "step": 95905 + }, + { + "epoch": 6.516510395434162, + "grad_norm": 1.242133617401123, + "learning_rate": 0.0001854362005707297, + "loss": 3.5487, + "step": 95910 + }, + { + "epoch": 6.516850115504824, + "grad_norm": 0.9252604246139526, + "learning_rate": 0.00018539373556189701, + "loss": 3.7432, + "step": 95915 + }, + { + "epoch": 6.517189835575486, + "grad_norm": 0.8174116611480713, + "learning_rate": 0.00018535127055306427, + "loss": 3.4629, + "step": 95920 + }, + { + "epoch": 6.517529555646147, + "grad_norm": 0.9442908763885498, + "learning_rate": 0.00018530880554423155, + "loss": 3.342, + "step": 95925 + }, + { + "epoch": 6.5178692757168095, + "grad_norm": 0.9604387283325195, + "learning_rate": 0.00018526634053539885, + "loss": 3.2073, + "step": 95930 + }, + { + "epoch": 6.518208995787472, + "grad_norm": 0.8393046259880066, + "learning_rate": 0.0001852238755265661, + "loss": 3.4116, + "step": 95935 + }, + { + "epoch": 6.518548715858133, + "grad_norm": 1.293305516242981, + "learning_rate": 0.0001851814105177334, + "loss": 3.4915, + "step": 95940 + }, + { + "epoch": 6.518888435928795, + "grad_norm": 0.8400179743766785, + "learning_rate": 0.00018513894550890067, + "loss": 3.6969, + "step": 95945 + }, + { + "epoch": 6.519228155999457, + "grad_norm": 0.8703471422195435, + "learning_rate": 0.00018509648050006795, + "loss": 3.5516, + "step": 95950 + }, + { + "epoch": 6.519567876070118, + "grad_norm": 0.9292824864387512, + "learning_rate": 0.00018505401549123523, + "loss": 3.4302, + "step": 95955 + }, + { + "epoch": 6.51990759614078, + "grad_norm": 0.8844501972198486, + "learning_rate": 0.0001850115504824025, + "loss": 3.4187, + "step": 95960 + }, + { + "epoch": 6.520247316211442, + "grad_norm": 1.0458048582077026, + "learning_rate": 0.0001849690854735698, + "loss": 3.1834, + "step": 95965 + }, + { + "epoch": 6.5205870362821035, + "grad_norm": 0.8219287991523743, + "learning_rate": 0.00018492662046473707, + "loss": 3.5045, + "step": 95970 + }, + { + "epoch": 6.5209267563527655, + "grad_norm": 0.8267305493354797, + "learning_rate": 0.00018488415545590435, + "loss": 3.2585, + "step": 95975 + }, + { + "epoch": 6.521266476423427, + "grad_norm": 1.2666661739349365, + "learning_rate": 0.0001848416904470716, + "loss": 3.2145, + "step": 95980 + }, + { + "epoch": 6.521606196494089, + "grad_norm": 0.6847668290138245, + "learning_rate": 0.0001847992254382389, + "loss": 3.3865, + "step": 95985 + }, + { + "epoch": 6.521945916564751, + "grad_norm": 0.9170442223548889, + "learning_rate": 0.00018475676042940616, + "loss": 3.2551, + "step": 95990 + }, + { + "epoch": 6.522285636635412, + "grad_norm": 0.8472086191177368, + "learning_rate": 0.00018471429542057344, + "loss": 3.4932, + "step": 95995 + }, + { + "epoch": 6.522625356706074, + "grad_norm": 0.8066442012786865, + "learning_rate": 0.00018467183041174075, + "loss": 3.2708, + "step": 96000 + }, + { + "epoch": 6.522965076776736, + "grad_norm": 0.7546025514602661, + "learning_rate": 0.000184629365402908, + "loss": 3.199, + "step": 96005 + }, + { + "epoch": 6.523304796847397, + "grad_norm": 0.8436610698699951, + "learning_rate": 0.00018458690039407528, + "loss": 3.2906, + "step": 96010 + }, + { + "epoch": 6.5236445169180595, + "grad_norm": 1.0117905139923096, + "learning_rate": 0.00018454443538524256, + "loss": 3.4779, + "step": 96015 + }, + { + "epoch": 6.5239842369887215, + "grad_norm": 0.9203060269355774, + "learning_rate": 0.00018450197037640984, + "loss": 3.3492, + "step": 96020 + }, + { + "epoch": 6.524323957059383, + "grad_norm": 0.8746620416641235, + "learning_rate": 0.0001844595053675771, + "loss": 3.5957, + "step": 96025 + }, + { + "epoch": 6.524663677130045, + "grad_norm": 0.8685802817344666, + "learning_rate": 0.0001844170403587444, + "loss": 3.418, + "step": 96030 + }, + { + "epoch": 6.525003397200707, + "grad_norm": 1.0191702842712402, + "learning_rate": 0.00018437457534991168, + "loss": 3.368, + "step": 96035 + }, + { + "epoch": 6.525343117271368, + "grad_norm": 1.1413466930389404, + "learning_rate": 0.00018433211034107896, + "loss": 3.3236, + "step": 96040 + }, + { + "epoch": 6.52568283734203, + "grad_norm": 0.9895147085189819, + "learning_rate": 0.00018428964533224624, + "loss": 3.5922, + "step": 96045 + }, + { + "epoch": 6.526022557412692, + "grad_norm": 1.032096266746521, + "learning_rate": 0.0001842471803234135, + "loss": 3.4093, + "step": 96050 + }, + { + "epoch": 6.526362277483353, + "grad_norm": 1.0359653234481812, + "learning_rate": 0.0001842047153145808, + "loss": 3.719, + "step": 96055 + }, + { + "epoch": 6.5267019975540155, + "grad_norm": 0.9334611296653748, + "learning_rate": 0.00018416225030574808, + "loss": 3.5916, + "step": 96060 + }, + { + "epoch": 6.5270417176246776, + "grad_norm": 0.852807343006134, + "learning_rate": 0.00018411978529691534, + "loss": 3.2924, + "step": 96065 + }, + { + "epoch": 6.527381437695339, + "grad_norm": 0.8230688571929932, + "learning_rate": 0.00018407732028808264, + "loss": 3.4333, + "step": 96070 + }, + { + "epoch": 6.527721157766001, + "grad_norm": 0.818594753742218, + "learning_rate": 0.0001840348552792499, + "loss": 3.4036, + "step": 96075 + }, + { + "epoch": 6.528060877836663, + "grad_norm": 1.2974421977996826, + "learning_rate": 0.00018399239027041718, + "loss": 3.6257, + "step": 96080 + }, + { + "epoch": 6.528400597907324, + "grad_norm": 0.9404606819152832, + "learning_rate": 0.00018394992526158446, + "loss": 3.6362, + "step": 96085 + }, + { + "epoch": 6.528740317977986, + "grad_norm": 1.0041372776031494, + "learning_rate": 0.00018390746025275174, + "loss": 3.2908, + "step": 96090 + }, + { + "epoch": 6.529080038048648, + "grad_norm": 0.8028071522712708, + "learning_rate": 0.00018386499524391902, + "loss": 3.3609, + "step": 96095 + }, + { + "epoch": 6.529419758119309, + "grad_norm": 0.9531019330024719, + "learning_rate": 0.0001838225302350863, + "loss": 3.4212, + "step": 96100 + }, + { + "epoch": 6.5297594781899715, + "grad_norm": 0.822538435459137, + "learning_rate": 0.00018378006522625358, + "loss": 3.3213, + "step": 96105 + }, + { + "epoch": 6.530099198260634, + "grad_norm": 1.0606689453125, + "learning_rate": 0.00018373760021742083, + "loss": 3.6586, + "step": 96110 + }, + { + "epoch": 6.530438918331295, + "grad_norm": 1.0199307203292847, + "learning_rate": 0.00018369513520858814, + "loss": 3.6389, + "step": 96115 + }, + { + "epoch": 6.530778638401957, + "grad_norm": 0.8051803708076477, + "learning_rate": 0.0001836526701997554, + "loss": 3.4889, + "step": 96120 + }, + { + "epoch": 6.531118358472619, + "grad_norm": 0.8674982786178589, + "learning_rate": 0.0001836102051909227, + "loss": 3.4121, + "step": 96125 + }, + { + "epoch": 6.53145807854328, + "grad_norm": 1.0286844968795776, + "learning_rate": 0.00018356774018208998, + "loss": 3.513, + "step": 96130 + }, + { + "epoch": 6.531797798613942, + "grad_norm": 0.97404944896698, + "learning_rate": 0.00018352527517325723, + "loss": 3.565, + "step": 96135 + }, + { + "epoch": 6.532137518684604, + "grad_norm": 0.8620907068252563, + "learning_rate": 0.00018348281016442454, + "loss": 3.0239, + "step": 96140 + }, + { + "epoch": 6.532477238755265, + "grad_norm": 0.7672960162162781, + "learning_rate": 0.0001834403451555918, + "loss": 3.5258, + "step": 96145 + }, + { + "epoch": 6.5328169588259275, + "grad_norm": 0.8723746538162231, + "learning_rate": 0.00018339788014675907, + "loss": 3.457, + "step": 96150 + }, + { + "epoch": 6.53315667889659, + "grad_norm": 0.7748669981956482, + "learning_rate": 0.00018335541513792635, + "loss": 3.3638, + "step": 96155 + }, + { + "epoch": 6.533496398967251, + "grad_norm": 0.7602195143699646, + "learning_rate": 0.00018331295012909363, + "loss": 3.46, + "step": 96160 + }, + { + "epoch": 6.533836119037913, + "grad_norm": 0.7579171657562256, + "learning_rate": 0.0001832704851202609, + "loss": 3.4407, + "step": 96165 + }, + { + "epoch": 6.534175839108575, + "grad_norm": 1.313528299331665, + "learning_rate": 0.0001832280201114282, + "loss": 3.2686, + "step": 96170 + }, + { + "epoch": 6.534515559179236, + "grad_norm": 0.7333289384841919, + "learning_rate": 0.00018318555510259547, + "loss": 3.6491, + "step": 96175 + }, + { + "epoch": 6.534855279249898, + "grad_norm": 0.8281060457229614, + "learning_rate": 0.00018314309009376272, + "loss": 3.2306, + "step": 96180 + }, + { + "epoch": 6.535194999320559, + "grad_norm": 0.8094376921653748, + "learning_rate": 0.00018310062508493003, + "loss": 3.5752, + "step": 96185 + }, + { + "epoch": 6.535534719391221, + "grad_norm": 0.914044976234436, + "learning_rate": 0.00018305816007609728, + "loss": 3.4868, + "step": 96190 + }, + { + "epoch": 6.5358744394618835, + "grad_norm": 1.0869314670562744, + "learning_rate": 0.00018301569506726456, + "loss": 3.3284, + "step": 96195 + }, + { + "epoch": 6.536214159532545, + "grad_norm": 1.1346052885055542, + "learning_rate": 0.00018297323005843187, + "loss": 3.185, + "step": 96200 + }, + { + "epoch": 6.536553879603207, + "grad_norm": 0.8431175947189331, + "learning_rate": 0.00018293076504959912, + "loss": 3.5696, + "step": 96205 + }, + { + "epoch": 6.536893599673869, + "grad_norm": 0.9523182511329651, + "learning_rate": 0.00018288830004076643, + "loss": 3.432, + "step": 96210 + }, + { + "epoch": 6.53723331974453, + "grad_norm": 0.967414140701294, + "learning_rate": 0.00018284583503193368, + "loss": 3.5201, + "step": 96215 + }, + { + "epoch": 6.537573039815192, + "grad_norm": 0.9264104962348938, + "learning_rate": 0.00018280337002310096, + "loss": 3.2958, + "step": 96220 + }, + { + "epoch": 6.537912759885854, + "grad_norm": 1.212496042251587, + "learning_rate": 0.00018276090501426827, + "loss": 3.247, + "step": 96225 + }, + { + "epoch": 6.538252479956515, + "grad_norm": 1.0725003480911255, + "learning_rate": 0.00018271844000543552, + "loss": 3.1692, + "step": 96230 + }, + { + "epoch": 6.5385922000271774, + "grad_norm": 0.8422450423240662, + "learning_rate": 0.0001826759749966028, + "loss": 3.3385, + "step": 96235 + }, + { + "epoch": 6.5389319200978395, + "grad_norm": 0.9508202075958252, + "learning_rate": 0.00018263350998777008, + "loss": 3.3481, + "step": 96240 + }, + { + "epoch": 6.539271640168501, + "grad_norm": 0.9718259572982788, + "learning_rate": 0.00018259104497893736, + "loss": 3.5718, + "step": 96245 + }, + { + "epoch": 6.539611360239163, + "grad_norm": 0.8119010329246521, + "learning_rate": 0.00018254857997010462, + "loss": 3.1814, + "step": 96250 + }, + { + "epoch": 6.539951080309825, + "grad_norm": 0.837174117565155, + "learning_rate": 0.00018250611496127192, + "loss": 3.3066, + "step": 96255 + }, + { + "epoch": 6.540290800380486, + "grad_norm": 0.9051608443260193, + "learning_rate": 0.0001824636499524392, + "loss": 3.4143, + "step": 96260 + }, + { + "epoch": 6.540630520451148, + "grad_norm": 0.7958584427833557, + "learning_rate": 0.00018242118494360646, + "loss": 3.6045, + "step": 96265 + }, + { + "epoch": 6.54097024052181, + "grad_norm": 0.7727922797203064, + "learning_rate": 0.00018237871993477376, + "loss": 3.3074, + "step": 96270 + }, + { + "epoch": 6.541309960592471, + "grad_norm": 0.8252001404762268, + "learning_rate": 0.00018233625492594102, + "loss": 3.5392, + "step": 96275 + }, + { + "epoch": 6.5416496806631335, + "grad_norm": 0.9295417070388794, + "learning_rate": 0.0001822937899171083, + "loss": 3.0066, + "step": 96280 + }, + { + "epoch": 6.5419894007337955, + "grad_norm": 0.7995486855506897, + "learning_rate": 0.00018225132490827558, + "loss": 3.3325, + "step": 96285 + }, + { + "epoch": 6.542329120804457, + "grad_norm": 0.7768329977989197, + "learning_rate": 0.00018220885989944286, + "loss": 3.6758, + "step": 96290 + }, + { + "epoch": 6.542668840875119, + "grad_norm": 1.0452044010162354, + "learning_rate": 0.00018216639489061016, + "loss": 3.5726, + "step": 96295 + }, + { + "epoch": 6.543008560945781, + "grad_norm": 0.7729411125183105, + "learning_rate": 0.00018212392988177742, + "loss": 3.4331, + "step": 96300 + }, + { + "epoch": 6.543348281016442, + "grad_norm": 0.9688342213630676, + "learning_rate": 0.0001820814648729447, + "loss": 3.3579, + "step": 96305 + }, + { + "epoch": 6.543688001087104, + "grad_norm": 0.8764786720275879, + "learning_rate": 0.00018203899986411198, + "loss": 3.3274, + "step": 96310 + }, + { + "epoch": 6.544027721157766, + "grad_norm": 1.4430091381072998, + "learning_rate": 0.00018199653485527926, + "loss": 3.7318, + "step": 96315 + }, + { + "epoch": 6.544367441228427, + "grad_norm": 0.8831747174263, + "learning_rate": 0.0001819540698464465, + "loss": 3.2517, + "step": 96320 + }, + { + "epoch": 6.5447071612990895, + "grad_norm": 1.793178677558899, + "learning_rate": 0.00018191160483761382, + "loss": 3.424, + "step": 96325 + }, + { + "epoch": 6.5450468813697515, + "grad_norm": 1.0286811590194702, + "learning_rate": 0.0001818691398287811, + "loss": 3.1488, + "step": 96330 + }, + { + "epoch": 6.545386601440413, + "grad_norm": 0.8510566353797913, + "learning_rate": 0.00018182667481994835, + "loss": 3.4564, + "step": 96335 + }, + { + "epoch": 6.545726321511075, + "grad_norm": 0.7581299543380737, + "learning_rate": 0.00018178420981111566, + "loss": 3.6529, + "step": 96340 + }, + { + "epoch": 6.546066041581737, + "grad_norm": 0.9651316404342651, + "learning_rate": 0.0001817417448022829, + "loss": 3.407, + "step": 96345 + }, + { + "epoch": 6.546405761652398, + "grad_norm": 1.0034099817276, + "learning_rate": 0.0001816992797934502, + "loss": 3.3473, + "step": 96350 + }, + { + "epoch": 6.54674548172306, + "grad_norm": 0.8370539546012878, + "learning_rate": 0.0001816568147846175, + "loss": 3.4498, + "step": 96355 + }, + { + "epoch": 6.547085201793722, + "grad_norm": 1.0059490203857422, + "learning_rate": 0.00018161434977578475, + "loss": 3.3152, + "step": 96360 + }, + { + "epoch": 6.547424921864383, + "grad_norm": 0.9495205879211426, + "learning_rate": 0.00018157188476695203, + "loss": 3.4323, + "step": 96365 + }, + { + "epoch": 6.5477646419350455, + "grad_norm": 0.8184398412704468, + "learning_rate": 0.0001815294197581193, + "loss": 3.5395, + "step": 96370 + }, + { + "epoch": 6.548104362005708, + "grad_norm": 1.1394977569580078, + "learning_rate": 0.0001814869547492866, + "loss": 3.2633, + "step": 96375 + }, + { + "epoch": 6.548444082076369, + "grad_norm": 0.8378457427024841, + "learning_rate": 0.00018144448974045387, + "loss": 3.3935, + "step": 96380 + }, + { + "epoch": 6.548783802147031, + "grad_norm": 0.8346160650253296, + "learning_rate": 0.00018140202473162115, + "loss": 3.2957, + "step": 96385 + }, + { + "epoch": 6.549123522217693, + "grad_norm": 0.893260657787323, + "learning_rate": 0.00018135955972278843, + "loss": 3.4471, + "step": 96390 + }, + { + "epoch": 6.549463242288354, + "grad_norm": 0.8777986168861389, + "learning_rate": 0.0001813170947139557, + "loss": 3.4238, + "step": 96395 + }, + { + "epoch": 6.549802962359016, + "grad_norm": 1.0411789417266846, + "learning_rate": 0.000181274629705123, + "loss": 3.4382, + "step": 96400 + }, + { + "epoch": 6.550142682429678, + "grad_norm": 0.8662761449813843, + "learning_rate": 0.00018123216469629025, + "loss": 3.3216, + "step": 96405 + }, + { + "epoch": 6.550482402500339, + "grad_norm": 1.3751085996627808, + "learning_rate": 0.00018118969968745755, + "loss": 3.5311, + "step": 96410 + }, + { + "epoch": 6.5508221225710015, + "grad_norm": 1.0227241516113281, + "learning_rate": 0.0001811472346786248, + "loss": 3.5494, + "step": 96415 + }, + { + "epoch": 6.551161842641664, + "grad_norm": 0.8639829158782959, + "learning_rate": 0.00018110476966979209, + "loss": 3.3302, + "step": 96420 + }, + { + "epoch": 6.551501562712325, + "grad_norm": 1.1059315204620361, + "learning_rate": 0.0001810623046609594, + "loss": 3.5717, + "step": 96425 + }, + { + "epoch": 6.551841282782987, + "grad_norm": 0.702799379825592, + "learning_rate": 0.00018101983965212665, + "loss": 3.4817, + "step": 96430 + }, + { + "epoch": 6.552181002853649, + "grad_norm": 1.0107479095458984, + "learning_rate": 0.00018097737464329393, + "loss": 3.4792, + "step": 96435 + }, + { + "epoch": 6.55252072292431, + "grad_norm": 0.7449522018432617, + "learning_rate": 0.0001809349096344612, + "loss": 3.2505, + "step": 96440 + }, + { + "epoch": 6.552860442994972, + "grad_norm": 1.0005202293395996, + "learning_rate": 0.00018089244462562849, + "loss": 3.2074, + "step": 96445 + }, + { + "epoch": 6.553200163065634, + "grad_norm": 1.139012336730957, + "learning_rate": 0.00018084997961679574, + "loss": 3.4631, + "step": 96450 + }, + { + "epoch": 6.553539883136295, + "grad_norm": 0.982383668422699, + "learning_rate": 0.00018080751460796305, + "loss": 3.4828, + "step": 96455 + }, + { + "epoch": 6.5538796032069575, + "grad_norm": 0.8553029298782349, + "learning_rate": 0.00018076504959913033, + "loss": 3.1523, + "step": 96460 + }, + { + "epoch": 6.55421932327762, + "grad_norm": 0.8360076546669006, + "learning_rate": 0.0001807225845902976, + "loss": 3.2841, + "step": 96465 + }, + { + "epoch": 6.554559043348281, + "grad_norm": 0.7418019771575928, + "learning_rate": 0.00018068011958146489, + "loss": 3.6524, + "step": 96470 + }, + { + "epoch": 6.554898763418943, + "grad_norm": 1.0690518617630005, + "learning_rate": 0.00018063765457263214, + "loss": 3.5251, + "step": 96475 + }, + { + "epoch": 6.555238483489605, + "grad_norm": 0.8379611968994141, + "learning_rate": 0.00018059518956379945, + "loss": 3.4103, + "step": 96480 + }, + { + "epoch": 6.555578203560266, + "grad_norm": 0.7836893796920776, + "learning_rate": 0.00018055272455496673, + "loss": 3.403, + "step": 96485 + }, + { + "epoch": 6.555917923630928, + "grad_norm": 1.0476809740066528, + "learning_rate": 0.00018051025954613398, + "loss": 3.4315, + "step": 96490 + }, + { + "epoch": 6.55625764370159, + "grad_norm": 1.204714298248291, + "learning_rate": 0.0001804677945373013, + "loss": 3.0426, + "step": 96495 + }, + { + "epoch": 6.556597363772251, + "grad_norm": 1.3424108028411865, + "learning_rate": 0.00018042532952846854, + "loss": 3.4637, + "step": 96500 + }, + { + "epoch": 6.5569370838429135, + "grad_norm": 0.9598059058189392, + "learning_rate": 0.00018038286451963582, + "loss": 3.2847, + "step": 96505 + }, + { + "epoch": 6.557276803913576, + "grad_norm": 0.892426609992981, + "learning_rate": 0.0001803403995108031, + "loss": 3.3879, + "step": 96510 + }, + { + "epoch": 6.557616523984237, + "grad_norm": 0.9444541335105896, + "learning_rate": 0.00018029793450197038, + "loss": 3.3837, + "step": 96515 + }, + { + "epoch": 6.557956244054899, + "grad_norm": 0.799524188041687, + "learning_rate": 0.00018025546949313766, + "loss": 3.4511, + "step": 96520 + }, + { + "epoch": 6.558295964125561, + "grad_norm": 1.108907699584961, + "learning_rate": 0.00018021300448430494, + "loss": 3.5128, + "step": 96525 + }, + { + "epoch": 6.558635684196222, + "grad_norm": 0.8416382074356079, + "learning_rate": 0.00018017053947547222, + "loss": 3.5781, + "step": 96530 + }, + { + "epoch": 6.558975404266884, + "grad_norm": 1.1393036842346191, + "learning_rate": 0.00018012807446663947, + "loss": 3.6635, + "step": 96535 + }, + { + "epoch": 6.559315124337546, + "grad_norm": 0.7632871270179749, + "learning_rate": 0.00018008560945780678, + "loss": 3.3651, + "step": 96540 + }, + { + "epoch": 6.5596548444082075, + "grad_norm": 0.8355586528778076, + "learning_rate": 0.00018004314444897403, + "loss": 3.4651, + "step": 96545 + }, + { + "epoch": 6.5599945644788695, + "grad_norm": 0.9064339995384216, + "learning_rate": 0.00018000067944014134, + "loss": 3.5642, + "step": 96550 + }, + { + "epoch": 6.560334284549532, + "grad_norm": 0.9710269570350647, + "learning_rate": 0.00017995821443130862, + "loss": 3.3034, + "step": 96555 + }, + { + "epoch": 6.560674004620193, + "grad_norm": 1.068237066268921, + "learning_rate": 0.00017991574942247587, + "loss": 3.4274, + "step": 96560 + }, + { + "epoch": 6.561013724690855, + "grad_norm": 1.1141470670700073, + "learning_rate": 0.00017987328441364318, + "loss": 3.3955, + "step": 96565 + }, + { + "epoch": 6.561353444761517, + "grad_norm": 0.8380339741706848, + "learning_rate": 0.00017983081940481043, + "loss": 3.1719, + "step": 96570 + }, + { + "epoch": 6.561693164832178, + "grad_norm": 0.879740297794342, + "learning_rate": 0.00017978835439597771, + "loss": 3.3552, + "step": 96575 + }, + { + "epoch": 6.56203288490284, + "grad_norm": 0.7422114610671997, + "learning_rate": 0.000179745889387145, + "loss": 3.4146, + "step": 96580 + }, + { + "epoch": 6.562372604973502, + "grad_norm": 1.0993226766586304, + "learning_rate": 0.00017970342437831227, + "loss": 3.2413, + "step": 96585 + }, + { + "epoch": 6.5627123250441635, + "grad_norm": 1.0552394390106201, + "learning_rate": 0.00017966095936947955, + "loss": 3.29, + "step": 96590 + }, + { + "epoch": 6.5630520451148255, + "grad_norm": 0.9912132024765015, + "learning_rate": 0.00017961849436064683, + "loss": 3.3263, + "step": 96595 + }, + { + "epoch": 6.563391765185488, + "grad_norm": 1.0826581716537476, + "learning_rate": 0.00017957602935181411, + "loss": 3.4181, + "step": 96600 + }, + { + "epoch": 6.563731485256149, + "grad_norm": 1.0164145231246948, + "learning_rate": 0.00017953356434298137, + "loss": 3.282, + "step": 96605 + }, + { + "epoch": 6.564071205326811, + "grad_norm": 0.8137017488479614, + "learning_rate": 0.00017949109933414867, + "loss": 3.411, + "step": 96610 + }, + { + "epoch": 6.564410925397473, + "grad_norm": 0.9394779205322266, + "learning_rate": 0.00017944863432531593, + "loss": 3.6155, + "step": 96615 + }, + { + "epoch": 6.564750645468134, + "grad_norm": 0.9173466563224792, + "learning_rate": 0.0001794061693164832, + "loss": 3.206, + "step": 96620 + }, + { + "epoch": 6.565090365538796, + "grad_norm": 1.2332847118377686, + "learning_rate": 0.00017936370430765051, + "loss": 3.4542, + "step": 96625 + }, + { + "epoch": 6.565430085609458, + "grad_norm": 0.8610777854919434, + "learning_rate": 0.00017932123929881777, + "loss": 3.3322, + "step": 96630 + }, + { + "epoch": 6.5657698056801195, + "grad_norm": 0.9394741058349609, + "learning_rate": 0.00017927877428998507, + "loss": 3.3994, + "step": 96635 + }, + { + "epoch": 6.5661095257507816, + "grad_norm": 0.9544647932052612, + "learning_rate": 0.00017923630928115233, + "loss": 3.2107, + "step": 96640 + }, + { + "epoch": 6.566449245821444, + "grad_norm": 1.1223320960998535, + "learning_rate": 0.0001791938442723196, + "loss": 3.3461, + "step": 96645 + }, + { + "epoch": 6.566788965892105, + "grad_norm": 0.9003207683563232, + "learning_rate": 0.00017915137926348691, + "loss": 3.6896, + "step": 96650 + }, + { + "epoch": 6.567128685962767, + "grad_norm": 0.8407727479934692, + "learning_rate": 0.00017910891425465417, + "loss": 3.5955, + "step": 96655 + }, + { + "epoch": 6.567468406033428, + "grad_norm": 0.8986020684242249, + "learning_rate": 0.00017906644924582145, + "loss": 3.5315, + "step": 96660 + }, + { + "epoch": 6.56780812610409, + "grad_norm": 1.1519970893859863, + "learning_rate": 0.00017902398423698873, + "loss": 3.5169, + "step": 96665 + }, + { + "epoch": 6.568147846174752, + "grad_norm": 1.0017104148864746, + "learning_rate": 0.000178981519228156, + "loss": 3.5117, + "step": 96670 + }, + { + "epoch": 6.568487566245413, + "grad_norm": 0.9840829372406006, + "learning_rate": 0.00017893905421932326, + "loss": 3.6065, + "step": 96675 + }, + { + "epoch": 6.5688272863160755, + "grad_norm": 0.8146967887878418, + "learning_rate": 0.00017889658921049057, + "loss": 3.4699, + "step": 96680 + }, + { + "epoch": 6.569167006386738, + "grad_norm": 1.0082557201385498, + "learning_rate": 0.00017885412420165785, + "loss": 3.5117, + "step": 96685 + }, + { + "epoch": 6.569506726457399, + "grad_norm": 0.9970901608467102, + "learning_rate": 0.0001788116591928251, + "loss": 3.2072, + "step": 96690 + }, + { + "epoch": 6.569846446528061, + "grad_norm": 1.1105859279632568, + "learning_rate": 0.0001787691941839924, + "loss": 3.2322, + "step": 96695 + }, + { + "epoch": 6.570186166598723, + "grad_norm": 0.7650870084762573, + "learning_rate": 0.00017872672917515966, + "loss": 3.3295, + "step": 96700 + }, + { + "epoch": 6.570525886669384, + "grad_norm": 1.094420075416565, + "learning_rate": 0.00017868426416632694, + "loss": 3.1102, + "step": 96705 + }, + { + "epoch": 6.570865606740046, + "grad_norm": 0.8585089445114136, + "learning_rate": 0.00017864179915749422, + "loss": 3.3181, + "step": 96710 + }, + { + "epoch": 6.571205326810708, + "grad_norm": 1.038527488708496, + "learning_rate": 0.0001785993341486615, + "loss": 3.3718, + "step": 96715 + }, + { + "epoch": 6.571545046881369, + "grad_norm": 0.9779325723648071, + "learning_rate": 0.0001785568691398288, + "loss": 3.4474, + "step": 96720 + }, + { + "epoch": 6.5718847669520315, + "grad_norm": 0.817451000213623, + "learning_rate": 0.00017851440413099606, + "loss": 3.152, + "step": 96725 + }, + { + "epoch": 6.572224487022694, + "grad_norm": 1.1775193214416504, + "learning_rate": 0.00017847193912216334, + "loss": 3.2355, + "step": 96730 + }, + { + "epoch": 6.572564207093355, + "grad_norm": 0.9945552945137024, + "learning_rate": 0.00017842947411333062, + "loss": 3.3415, + "step": 96735 + }, + { + "epoch": 6.572903927164017, + "grad_norm": 0.8823405504226685, + "learning_rate": 0.0001783870091044979, + "loss": 3.39, + "step": 96740 + }, + { + "epoch": 6.573243647234679, + "grad_norm": 0.9376872181892395, + "learning_rate": 0.00017834454409566515, + "loss": 3.2239, + "step": 96745 + }, + { + "epoch": 6.57358336730534, + "grad_norm": 0.9345604777336121, + "learning_rate": 0.00017830207908683246, + "loss": 3.7012, + "step": 96750 + }, + { + "epoch": 6.573923087376002, + "grad_norm": 0.8919379115104675, + "learning_rate": 0.00017825961407799974, + "loss": 3.3716, + "step": 96755 + }, + { + "epoch": 6.574262807446664, + "grad_norm": 1.1227861642837524, + "learning_rate": 0.000178217149069167, + "loss": 3.3507, + "step": 96760 + }, + { + "epoch": 6.574602527517325, + "grad_norm": 0.8192927837371826, + "learning_rate": 0.0001781746840603343, + "loss": 3.5682, + "step": 96765 + }, + { + "epoch": 6.5749422475879875, + "grad_norm": 1.1087735891342163, + "learning_rate": 0.00017813221905150156, + "loss": 3.4697, + "step": 96770 + }, + { + "epoch": 6.57528196765865, + "grad_norm": 0.8533098101615906, + "learning_rate": 0.00017808975404266884, + "loss": 3.3559, + "step": 96775 + }, + { + "epoch": 6.575621687729311, + "grad_norm": 1.1783359050750732, + "learning_rate": 0.00017804728903383614, + "loss": 3.2492, + "step": 96780 + }, + { + "epoch": 6.575961407799973, + "grad_norm": 0.8158478140830994, + "learning_rate": 0.0001780048240250034, + "loss": 3.3771, + "step": 96785 + }, + { + "epoch": 6.576301127870635, + "grad_norm": 0.9620354175567627, + "learning_rate": 0.00017796235901617068, + "loss": 3.4077, + "step": 96790 + }, + { + "epoch": 6.576640847941296, + "grad_norm": 0.9056695699691772, + "learning_rate": 0.00017791989400733796, + "loss": 3.4942, + "step": 96795 + }, + { + "epoch": 6.576980568011958, + "grad_norm": 2.6155855655670166, + "learning_rate": 0.00017787742899850524, + "loss": 3.5389, + "step": 96800 + }, + { + "epoch": 6.57732028808262, + "grad_norm": 1.388078212738037, + "learning_rate": 0.00017783496398967252, + "loss": 3.5287, + "step": 96805 + }, + { + "epoch": 6.5776600081532814, + "grad_norm": 0.7088529467582703, + "learning_rate": 0.0001777924989808398, + "loss": 3.6379, + "step": 96810 + }, + { + "epoch": 6.5779997282239435, + "grad_norm": 0.957036018371582, + "learning_rate": 0.00017775003397200708, + "loss": 3.4873, + "step": 96815 + }, + { + "epoch": 6.578339448294606, + "grad_norm": 0.9193865060806274, + "learning_rate": 0.00017770756896317436, + "loss": 3.3159, + "step": 96820 + }, + { + "epoch": 6.578679168365267, + "grad_norm": 0.9519356489181519, + "learning_rate": 0.00017766510395434164, + "loss": 3.6776, + "step": 96825 + }, + { + "epoch": 6.579018888435929, + "grad_norm": 1.0563308000564575, + "learning_rate": 0.0001776226389455089, + "loss": 3.182, + "step": 96830 + }, + { + "epoch": 6.579358608506591, + "grad_norm": 0.9857468008995056, + "learning_rate": 0.0001775801739366762, + "loss": 3.4982, + "step": 96835 + }, + { + "epoch": 6.579698328577252, + "grad_norm": 0.9621128439903259, + "learning_rate": 0.00017753770892784345, + "loss": 3.3874, + "step": 96840 + }, + { + "epoch": 6.580038048647914, + "grad_norm": 1.0218464136123657, + "learning_rate": 0.00017749524391901073, + "loss": 3.2821, + "step": 96845 + }, + { + "epoch": 6.580377768718576, + "grad_norm": 0.7325925827026367, + "learning_rate": 0.00017745277891017804, + "loss": 3.4563, + "step": 96850 + }, + { + "epoch": 6.5807174887892375, + "grad_norm": 1.0075844526290894, + "learning_rate": 0.0001774103139013453, + "loss": 3.3229, + "step": 96855 + }, + { + "epoch": 6.5810572088598995, + "grad_norm": 0.7121471762657166, + "learning_rate": 0.00017736784889251257, + "loss": 3.6009, + "step": 96860 + }, + { + "epoch": 6.581396928930561, + "grad_norm": 0.86198890209198, + "learning_rate": 0.00017732538388367985, + "loss": 3.5832, + "step": 96865 + }, + { + "epoch": 6.581736649001223, + "grad_norm": 1.0362322330474854, + "learning_rate": 0.00017728291887484713, + "loss": 3.232, + "step": 96870 + }, + { + "epoch": 6.582076369071885, + "grad_norm": 0.7891597747802734, + "learning_rate": 0.00017724045386601438, + "loss": 3.6037, + "step": 96875 + }, + { + "epoch": 6.582416089142546, + "grad_norm": 0.8380394577980042, + "learning_rate": 0.0001771979888571817, + "loss": 3.5148, + "step": 96880 + }, + { + "epoch": 6.582755809213208, + "grad_norm": 1.3050543069839478, + "learning_rate": 0.00017715552384834897, + "loss": 3.7364, + "step": 96885 + }, + { + "epoch": 6.58309552928387, + "grad_norm": 1.142210602760315, + "learning_rate": 0.00017711305883951625, + "loss": 3.1994, + "step": 96890 + }, + { + "epoch": 6.583435249354531, + "grad_norm": 1.025530457496643, + "learning_rate": 0.00017707059383068353, + "loss": 3.5795, + "step": 96895 + }, + { + "epoch": 6.5837749694251935, + "grad_norm": 0.976513147354126, + "learning_rate": 0.00017702812882185078, + "loss": 3.5733, + "step": 96900 + }, + { + "epoch": 6.5841146894958555, + "grad_norm": 0.7609542012214661, + "learning_rate": 0.0001769856638130181, + "loss": 3.4576, + "step": 96905 + }, + { + "epoch": 6.584454409566517, + "grad_norm": 0.8577448129653931, + "learning_rate": 0.00017694319880418537, + "loss": 3.4666, + "step": 96910 + }, + { + "epoch": 6.584794129637179, + "grad_norm": 0.8393151164054871, + "learning_rate": 0.00017690073379535262, + "loss": 3.3919, + "step": 96915 + }, + { + "epoch": 6.585133849707841, + "grad_norm": 1.2872567176818848, + "learning_rate": 0.00017685826878651993, + "loss": 3.4852, + "step": 96920 + }, + { + "epoch": 6.585473569778502, + "grad_norm": 0.9215535521507263, + "learning_rate": 0.00017681580377768718, + "loss": 3.2103, + "step": 96925 + }, + { + "epoch": 6.585813289849164, + "grad_norm": 0.7153884172439575, + "learning_rate": 0.00017677333876885446, + "loss": 3.5369, + "step": 96930 + }, + { + "epoch": 6.586153009919826, + "grad_norm": 1.169394612312317, + "learning_rate": 0.00017673087376002174, + "loss": 3.1736, + "step": 96935 + }, + { + "epoch": 6.586492729990487, + "grad_norm": 1.040228247642517, + "learning_rate": 0.00017668840875118902, + "loss": 3.4918, + "step": 96940 + }, + { + "epoch": 6.5868324500611495, + "grad_norm": 0.8365182280540466, + "learning_rate": 0.0001766459437423563, + "loss": 3.2675, + "step": 96945 + }, + { + "epoch": 6.5871721701318116, + "grad_norm": 0.8340921401977539, + "learning_rate": 0.00017660347873352358, + "loss": 3.2253, + "step": 96950 + }, + { + "epoch": 6.587511890202473, + "grad_norm": 1.0617649555206299, + "learning_rate": 0.00017656101372469086, + "loss": 3.244, + "step": 96955 + }, + { + "epoch": 6.587851610273135, + "grad_norm": 0.8880401849746704, + "learning_rate": 0.00017651854871585812, + "loss": 3.6268, + "step": 96960 + }, + { + "epoch": 6.588191330343797, + "grad_norm": 0.9446393847465515, + "learning_rate": 0.00017647608370702542, + "loss": 3.4175, + "step": 96965 + }, + { + "epoch": 6.588531050414458, + "grad_norm": 0.8900627493858337, + "learning_rate": 0.00017643361869819268, + "loss": 3.6855, + "step": 96970 + }, + { + "epoch": 6.58887077048512, + "grad_norm": 0.967086672782898, + "learning_rate": 0.00017639115368935998, + "loss": 3.0015, + "step": 96975 + }, + { + "epoch": 6.589210490555782, + "grad_norm": 1.095318078994751, + "learning_rate": 0.00017634868868052726, + "loss": 3.3785, + "step": 96980 + }, + { + "epoch": 6.589550210626443, + "grad_norm": 1.1031166315078735, + "learning_rate": 0.00017630622367169452, + "loss": 3.4277, + "step": 96985 + }, + { + "epoch": 6.5898899306971055, + "grad_norm": 0.9443245530128479, + "learning_rate": 0.00017626375866286182, + "loss": 3.513, + "step": 96990 + }, + { + "epoch": 6.590229650767768, + "grad_norm": 0.8057238459587097, + "learning_rate": 0.00017622129365402908, + "loss": 3.5998, + "step": 96995 + }, + { + "epoch": 6.590569370838429, + "grad_norm": 1.0880097150802612, + "learning_rate": 0.00017617882864519636, + "loss": 3.4988, + "step": 97000 + }, + { + "epoch": 6.590909090909091, + "grad_norm": 0.8588536381721497, + "learning_rate": 0.00017613636363636364, + "loss": 3.46, + "step": 97005 + }, + { + "epoch": 6.591248810979753, + "grad_norm": 0.7398130893707275, + "learning_rate": 0.00017609389862753092, + "loss": 3.298, + "step": 97010 + }, + { + "epoch": 6.591588531050414, + "grad_norm": 0.934594988822937, + "learning_rate": 0.0001760514336186982, + "loss": 3.3907, + "step": 97015 + }, + { + "epoch": 6.591928251121076, + "grad_norm": 0.9283747673034668, + "learning_rate": 0.00017600896860986548, + "loss": 3.4374, + "step": 97020 + }, + { + "epoch": 6.592267971191738, + "grad_norm": 1.0539137125015259, + "learning_rate": 0.00017596650360103276, + "loss": 3.3823, + "step": 97025 + }, + { + "epoch": 6.592607691262399, + "grad_norm": 1.143457293510437, + "learning_rate": 0.0001759240385922, + "loss": 3.515, + "step": 97030 + }, + { + "epoch": 6.5929474113330615, + "grad_norm": 1.0699450969696045, + "learning_rate": 0.00017588157358336732, + "loss": 3.1947, + "step": 97035 + }, + { + "epoch": 6.593287131403724, + "grad_norm": 0.9379249811172485, + "learning_rate": 0.00017583910857453457, + "loss": 3.1163, + "step": 97040 + }, + { + "epoch": 6.593626851474385, + "grad_norm": 0.8506780862808228, + "learning_rate": 0.00017579664356570185, + "loss": 3.3005, + "step": 97045 + }, + { + "epoch": 6.593966571545047, + "grad_norm": 1.0411192178726196, + "learning_rate": 0.00017575417855686916, + "loss": 3.293, + "step": 97050 + }, + { + "epoch": 6.594306291615709, + "grad_norm": 1.0176706314086914, + "learning_rate": 0.0001757117135480364, + "loss": 3.3373, + "step": 97055 + }, + { + "epoch": 6.59464601168637, + "grad_norm": 1.2964448928833008, + "learning_rate": 0.00017566924853920372, + "loss": 3.3327, + "step": 97060 + }, + { + "epoch": 6.594985731757032, + "grad_norm": 0.9258418083190918, + "learning_rate": 0.00017562678353037097, + "loss": 3.2672, + "step": 97065 + }, + { + "epoch": 6.595325451827694, + "grad_norm": 0.8606317043304443, + "learning_rate": 0.00017558431852153825, + "loss": 3.3155, + "step": 97070 + }, + { + "epoch": 6.595665171898355, + "grad_norm": 0.9827470779418945, + "learning_rate": 0.00017554185351270556, + "loss": 3.5252, + "step": 97075 + }, + { + "epoch": 6.5960048919690175, + "grad_norm": 1.1102116107940674, + "learning_rate": 0.0001754993885038728, + "loss": 3.2725, + "step": 97080 + }, + { + "epoch": 6.59634461203968, + "grad_norm": 0.8371499180793762, + "learning_rate": 0.0001754569234950401, + "loss": 3.4715, + "step": 97085 + }, + { + "epoch": 6.596684332110341, + "grad_norm": 0.7522786259651184, + "learning_rate": 0.00017541445848620737, + "loss": 3.4239, + "step": 97090 + }, + { + "epoch": 6.597024052181003, + "grad_norm": 0.9403442740440369, + "learning_rate": 0.00017537199347737465, + "loss": 3.5409, + "step": 97095 + }, + { + "epoch": 6.597363772251665, + "grad_norm": 0.9758060574531555, + "learning_rate": 0.0001753295284685419, + "loss": 3.4078, + "step": 97100 + }, + { + "epoch": 6.597703492322326, + "grad_norm": 1.1014878749847412, + "learning_rate": 0.0001752870634597092, + "loss": 3.4691, + "step": 97105 + }, + { + "epoch": 6.598043212392988, + "grad_norm": 0.9790297746658325, + "learning_rate": 0.0001752445984508765, + "loss": 3.5173, + "step": 97110 + }, + { + "epoch": 6.59838293246365, + "grad_norm": 1.202134370803833, + "learning_rate": 0.00017520213344204375, + "loss": 3.4041, + "step": 97115 + }, + { + "epoch": 6.5987226525343115, + "grad_norm": 0.8870344161987305, + "learning_rate": 0.00017515966843321105, + "loss": 3.3593, + "step": 97120 + }, + { + "epoch": 6.5990623726049735, + "grad_norm": 0.8611522912979126, + "learning_rate": 0.0001751172034243783, + "loss": 3.2736, + "step": 97125 + }, + { + "epoch": 6.599402092675636, + "grad_norm": 0.8605720400810242, + "learning_rate": 0.00017507473841554559, + "loss": 3.3223, + "step": 97130 + }, + { + "epoch": 6.599741812746297, + "grad_norm": 0.7668940424919128, + "learning_rate": 0.00017503227340671287, + "loss": 3.2286, + "step": 97135 + }, + { + "epoch": 6.600081532816959, + "grad_norm": 0.9660331010818481, + "learning_rate": 0.00017498980839788015, + "loss": 3.3517, + "step": 97140 + }, + { + "epoch": 6.600421252887621, + "grad_norm": 1.0815320014953613, + "learning_rate": 0.00017494734338904745, + "loss": 3.3717, + "step": 97145 + }, + { + "epoch": 6.600760972958282, + "grad_norm": 0.9496144652366638, + "learning_rate": 0.0001749048783802147, + "loss": 3.428, + "step": 97150 + }, + { + "epoch": 6.601100693028944, + "grad_norm": 0.7161312699317932, + "learning_rate": 0.00017486241337138199, + "loss": 3.334, + "step": 97155 + }, + { + "epoch": 6.601440413099606, + "grad_norm": 0.9579046368598938, + "learning_rate": 0.00017481994836254927, + "loss": 3.7254, + "step": 97160 + }, + { + "epoch": 6.6017801331702675, + "grad_norm": 0.9717501401901245, + "learning_rate": 0.00017477748335371655, + "loss": 3.7231, + "step": 97165 + }, + { + "epoch": 6.6021198532409295, + "grad_norm": 0.9098544716835022, + "learning_rate": 0.0001747350183448838, + "loss": 3.5238, + "step": 97170 + }, + { + "epoch": 6.602459573311592, + "grad_norm": 0.7948932647705078, + "learning_rate": 0.0001746925533360511, + "loss": 3.2513, + "step": 97175 + }, + { + "epoch": 6.602799293382253, + "grad_norm": 1.3976343870162964, + "learning_rate": 0.00017465008832721839, + "loss": 3.4779, + "step": 97180 + }, + { + "epoch": 6.603139013452915, + "grad_norm": 0.7515347003936768, + "learning_rate": 0.00017460762331838564, + "loss": 3.4871, + "step": 97185 + }, + { + "epoch": 6.603478733523577, + "grad_norm": 1.3301080465316772, + "learning_rate": 0.00017456515830955295, + "loss": 3.5431, + "step": 97190 + }, + { + "epoch": 6.603818453594238, + "grad_norm": 0.999260663986206, + "learning_rate": 0.0001745226933007202, + "loss": 3.53, + "step": 97195 + }, + { + "epoch": 6.6041581736649, + "grad_norm": 1.024856448173523, + "learning_rate": 0.00017448022829188748, + "loss": 3.762, + "step": 97200 + }, + { + "epoch": 6.604497893735562, + "grad_norm": 0.9875078201293945, + "learning_rate": 0.00017443776328305479, + "loss": 3.3218, + "step": 97205 + }, + { + "epoch": 6.6048376138062235, + "grad_norm": 1.0391463041305542, + "learning_rate": 0.00017439529827422204, + "loss": 3.1963, + "step": 97210 + }, + { + "epoch": 6.6051773338768855, + "grad_norm": 0.8712729811668396, + "learning_rate": 0.00017435283326538932, + "loss": 3.4108, + "step": 97215 + }, + { + "epoch": 6.605517053947548, + "grad_norm": 1.0027925968170166, + "learning_rate": 0.0001743103682565566, + "loss": 3.4871, + "step": 97220 + }, + { + "epoch": 6.605856774018209, + "grad_norm": 0.9626083374023438, + "learning_rate": 0.00017426790324772388, + "loss": 3.3567, + "step": 97225 + }, + { + "epoch": 6.606196494088871, + "grad_norm": 0.9243528842926025, + "learning_rate": 0.00017422543823889116, + "loss": 3.4419, + "step": 97230 + }, + { + "epoch": 6.606536214159533, + "grad_norm": 0.777763843536377, + "learning_rate": 0.00017418297323005844, + "loss": 3.4713, + "step": 97235 + }, + { + "epoch": 6.606875934230194, + "grad_norm": 0.9687936305999756, + "learning_rate": 0.00017414050822122572, + "loss": 3.4523, + "step": 97240 + }, + { + "epoch": 6.607215654300856, + "grad_norm": 1.1212396621704102, + "learning_rate": 0.000174098043212393, + "loss": 3.4206, + "step": 97245 + }, + { + "epoch": 6.607555374371518, + "grad_norm": 0.8235111236572266, + "learning_rate": 0.00017405557820356028, + "loss": 3.0957, + "step": 97250 + }, + { + "epoch": 6.6078950944421795, + "grad_norm": 0.9137499928474426, + "learning_rate": 0.00017401311319472753, + "loss": 3.2126, + "step": 97255 + }, + { + "epoch": 6.608234814512842, + "grad_norm": 0.8260185718536377, + "learning_rate": 0.00017397064818589484, + "loss": 3.6001, + "step": 97260 + }, + { + "epoch": 6.608574534583504, + "grad_norm": 0.756996214389801, + "learning_rate": 0.0001739281831770621, + "loss": 3.4617, + "step": 97265 + }, + { + "epoch": 6.608914254654165, + "grad_norm": 0.9281120300292969, + "learning_rate": 0.00017388571816822937, + "loss": 3.1534, + "step": 97270 + }, + { + "epoch": 6.609253974724827, + "grad_norm": 1.5400605201721191, + "learning_rate": 0.00017384325315939668, + "loss": 3.6344, + "step": 97275 + }, + { + "epoch": 6.609593694795489, + "grad_norm": 0.7828205227851868, + "learning_rate": 0.00017380078815056393, + "loss": 3.6083, + "step": 97280 + }, + { + "epoch": 6.60993341486615, + "grad_norm": 1.0113838911056519, + "learning_rate": 0.0001737583231417312, + "loss": 3.59, + "step": 97285 + }, + { + "epoch": 6.610273134936812, + "grad_norm": 0.809722900390625, + "learning_rate": 0.0001737158581328985, + "loss": 3.3597, + "step": 97290 + }, + { + "epoch": 6.610612855007474, + "grad_norm": 0.8679875135421753, + "learning_rate": 0.00017367339312406577, + "loss": 3.2865, + "step": 97295 + }, + { + "epoch": 6.6109525750781355, + "grad_norm": 0.8548592925071716, + "learning_rate": 0.00017363092811523303, + "loss": 3.6948, + "step": 97300 + }, + { + "epoch": 6.611292295148798, + "grad_norm": 1.0086027383804321, + "learning_rate": 0.00017358846310640033, + "loss": 3.2164, + "step": 97305 + }, + { + "epoch": 6.61163201521946, + "grad_norm": 0.9731512069702148, + "learning_rate": 0.00017354599809756761, + "loss": 3.3526, + "step": 97310 + }, + { + "epoch": 6.611971735290121, + "grad_norm": 0.8902407884597778, + "learning_rate": 0.0001735035330887349, + "loss": 3.544, + "step": 97315 + }, + { + "epoch": 6.612311455360783, + "grad_norm": 1.0919443368911743, + "learning_rate": 0.00017346106807990217, + "loss": 3.3923, + "step": 97320 + }, + { + "epoch": 6.612651175431445, + "grad_norm": 0.9543624520301819, + "learning_rate": 0.00017341860307106943, + "loss": 3.4573, + "step": 97325 + }, + { + "epoch": 6.612990895502106, + "grad_norm": 0.9324690103530884, + "learning_rate": 0.00017337613806223673, + "loss": 3.2852, + "step": 97330 + }, + { + "epoch": 6.613330615572768, + "grad_norm": 0.8224936723709106, + "learning_rate": 0.00017333367305340401, + "loss": 3.2928, + "step": 97335 + }, + { + "epoch": 6.613670335643429, + "grad_norm": 1.0577610731124878, + "learning_rate": 0.00017329120804457127, + "loss": 3.4361, + "step": 97340 + }, + { + "epoch": 6.6140100557140915, + "grad_norm": 0.9444671869277954, + "learning_rate": 0.00017324874303573857, + "loss": 3.4876, + "step": 97345 + }, + { + "epoch": 6.614349775784754, + "grad_norm": 1.4498974084854126, + "learning_rate": 0.00017320627802690583, + "loss": 3.3596, + "step": 97350 + }, + { + "epoch": 6.614689495855415, + "grad_norm": 0.9004837870597839, + "learning_rate": 0.0001731638130180731, + "loss": 3.1726, + "step": 97355 + }, + { + "epoch": 6.615029215926077, + "grad_norm": 1.0981498956680298, + "learning_rate": 0.0001731213480092404, + "loss": 3.0868, + "step": 97360 + }, + { + "epoch": 6.615368935996739, + "grad_norm": 1.0047904253005981, + "learning_rate": 0.00017307888300040767, + "loss": 3.4206, + "step": 97365 + }, + { + "epoch": 6.6157086560674, + "grad_norm": 0.9675589203834534, + "learning_rate": 0.00017303641799157495, + "loss": 3.46, + "step": 97370 + }, + { + "epoch": 6.616048376138062, + "grad_norm": 0.8424542546272278, + "learning_rate": 0.00017299395298274223, + "loss": 3.0176, + "step": 97375 + }, + { + "epoch": 6.616388096208724, + "grad_norm": 0.823799192905426, + "learning_rate": 0.0001729514879739095, + "loss": 3.2791, + "step": 97380 + }, + { + "epoch": 6.616727816279385, + "grad_norm": 0.9271381497383118, + "learning_rate": 0.00017290902296507676, + "loss": 3.59, + "step": 97385 + }, + { + "epoch": 6.6170675363500475, + "grad_norm": 0.9733273983001709, + "learning_rate": 0.00017286655795624407, + "loss": 3.1146, + "step": 97390 + }, + { + "epoch": 6.61740725642071, + "grad_norm": 0.793690025806427, + "learning_rate": 0.00017282409294741132, + "loss": 3.4336, + "step": 97395 + }, + { + "epoch": 6.617746976491371, + "grad_norm": 0.9721079468727112, + "learning_rate": 0.00017278162793857863, + "loss": 3.4053, + "step": 97400 + }, + { + "epoch": 6.618086696562033, + "grad_norm": 0.8432226777076721, + "learning_rate": 0.0001727391629297459, + "loss": 3.5304, + "step": 97405 + }, + { + "epoch": 6.618426416632695, + "grad_norm": 0.9247578382492065, + "learning_rate": 0.00017269669792091316, + "loss": 3.5286, + "step": 97410 + }, + { + "epoch": 6.618766136703356, + "grad_norm": 0.9321267604827881, + "learning_rate": 0.00017265423291208047, + "loss": 3.4086, + "step": 97415 + }, + { + "epoch": 6.619105856774018, + "grad_norm": 0.9049462676048279, + "learning_rate": 0.00017261176790324772, + "loss": 3.3728, + "step": 97420 + }, + { + "epoch": 6.61944557684468, + "grad_norm": 0.8023772835731506, + "learning_rate": 0.000172569302894415, + "loss": 3.6074, + "step": 97425 + }, + { + "epoch": 6.6197852969153415, + "grad_norm": 1.0134303569793701, + "learning_rate": 0.00017252683788558228, + "loss": 3.5517, + "step": 97430 + }, + { + "epoch": 6.6201250169860035, + "grad_norm": 0.859363853931427, + "learning_rate": 0.00017248437287674956, + "loss": 3.2292, + "step": 97435 + }, + { + "epoch": 6.620464737056666, + "grad_norm": 0.946462094783783, + "learning_rate": 0.00017244190786791684, + "loss": 3.4178, + "step": 97440 + }, + { + "epoch": 6.620804457127327, + "grad_norm": 0.8268222808837891, + "learning_rate": 0.00017239944285908412, + "loss": 3.1238, + "step": 97445 + }, + { + "epoch": 6.621144177197989, + "grad_norm": 0.7915638089179993, + "learning_rate": 0.0001723569778502514, + "loss": 3.6888, + "step": 97450 + }, + { + "epoch": 6.621483897268651, + "grad_norm": 1.0485198497772217, + "learning_rate": 0.00017231451284141865, + "loss": 3.3734, + "step": 97455 + }, + { + "epoch": 6.621823617339312, + "grad_norm": 0.9868941307067871, + "learning_rate": 0.00017227204783258596, + "loss": 3.5827, + "step": 97460 + }, + { + "epoch": 6.622163337409974, + "grad_norm": 0.9920569658279419, + "learning_rate": 0.00017222958282375321, + "loss": 3.3572, + "step": 97465 + }, + { + "epoch": 6.622503057480636, + "grad_norm": 0.9414002895355225, + "learning_rate": 0.0001721871178149205, + "loss": 3.432, + "step": 97470 + }, + { + "epoch": 6.6228427775512975, + "grad_norm": 0.9690113663673401, + "learning_rate": 0.0001721446528060878, + "loss": 3.3414, + "step": 97475 + }, + { + "epoch": 6.6231824976219595, + "grad_norm": 0.6124287843704224, + "learning_rate": 0.00017210218779725506, + "loss": 3.2697, + "step": 97480 + }, + { + "epoch": 6.623522217692622, + "grad_norm": 0.7516734004020691, + "learning_rate": 0.00017205972278842236, + "loss": 3.7023, + "step": 97485 + }, + { + "epoch": 6.623861937763283, + "grad_norm": 1.1481750011444092, + "learning_rate": 0.00017201725777958962, + "loss": 3.4669, + "step": 97490 + }, + { + "epoch": 6.624201657833945, + "grad_norm": 1.1928808689117432, + "learning_rate": 0.0001719747927707569, + "loss": 3.4481, + "step": 97495 + }, + { + "epoch": 6.624541377904607, + "grad_norm": 1.1480422019958496, + "learning_rate": 0.0001719323277619242, + "loss": 3.2602, + "step": 97500 + }, + { + "epoch": 6.624881097975268, + "grad_norm": 1.102179765701294, + "learning_rate": 0.00017188986275309146, + "loss": 3.4345, + "step": 97505 + }, + { + "epoch": 6.62522081804593, + "grad_norm": 0.9814446568489075, + "learning_rate": 0.00017184739774425874, + "loss": 3.5167, + "step": 97510 + }, + { + "epoch": 6.625560538116592, + "grad_norm": 0.957133948802948, + "learning_rate": 0.00017180493273542602, + "loss": 3.5274, + "step": 97515 + }, + { + "epoch": 6.6259002581872535, + "grad_norm": 0.8214372396469116, + "learning_rate": 0.0001717624677265933, + "loss": 3.7891, + "step": 97520 + }, + { + "epoch": 6.6262399782579156, + "grad_norm": 1.002238392829895, + "learning_rate": 0.00017172000271776055, + "loss": 3.1828, + "step": 97525 + }, + { + "epoch": 6.626579698328578, + "grad_norm": 1.4513875246047974, + "learning_rate": 0.00017167753770892786, + "loss": 3.6567, + "step": 97530 + }, + { + "epoch": 6.626919418399239, + "grad_norm": 0.9507254958152771, + "learning_rate": 0.00017163507270009514, + "loss": 3.5081, + "step": 97535 + }, + { + "epoch": 6.627259138469901, + "grad_norm": 0.9331368207931519, + "learning_rate": 0.0001715926076912624, + "loss": 3.4851, + "step": 97540 + }, + { + "epoch": 6.627598858540563, + "grad_norm": 0.9124776124954224, + "learning_rate": 0.0001715501426824297, + "loss": 3.3098, + "step": 97545 + }, + { + "epoch": 6.627938578611224, + "grad_norm": 1.278019666671753, + "learning_rate": 0.00017150767767359695, + "loss": 3.3929, + "step": 97550 + }, + { + "epoch": 6.628278298681886, + "grad_norm": 0.8858926296234131, + "learning_rate": 0.00017146521266476423, + "loss": 3.4578, + "step": 97555 + }, + { + "epoch": 6.628618018752547, + "grad_norm": 0.8165700435638428, + "learning_rate": 0.0001714227476559315, + "loss": 3.2665, + "step": 97560 + }, + { + "epoch": 6.6289577388232095, + "grad_norm": 0.8125775456428528, + "learning_rate": 0.0001713802826470988, + "loss": 3.2713, + "step": 97565 + }, + { + "epoch": 6.629297458893872, + "grad_norm": 0.9125362634658813, + "learning_rate": 0.0001713378176382661, + "loss": 3.3286, + "step": 97570 + }, + { + "epoch": 6.629637178964533, + "grad_norm": 1.0096575021743774, + "learning_rate": 0.00017129535262943335, + "loss": 3.4077, + "step": 97575 + }, + { + "epoch": 6.629976899035195, + "grad_norm": 0.8703675270080566, + "learning_rate": 0.00017125288762060063, + "loss": 3.5286, + "step": 97580 + }, + { + "epoch": 6.630316619105857, + "grad_norm": 1.1517874002456665, + "learning_rate": 0.0001712104226117679, + "loss": 3.2722, + "step": 97585 + }, + { + "epoch": 6.630656339176518, + "grad_norm": 1.0827560424804688, + "learning_rate": 0.0001711679576029352, + "loss": 3.2492, + "step": 97590 + }, + { + "epoch": 6.63099605924718, + "grad_norm": 0.8561388850212097, + "learning_rate": 0.00017112549259410244, + "loss": 3.5292, + "step": 97595 + }, + { + "epoch": 6.631335779317842, + "grad_norm": 0.8867644667625427, + "learning_rate": 0.00017108302758526975, + "loss": 3.441, + "step": 97600 + }, + { + "epoch": 6.631675499388503, + "grad_norm": 1.012528419494629, + "learning_rate": 0.00017104056257643703, + "loss": 3.7671, + "step": 97605 + }, + { + "epoch": 6.6320152194591655, + "grad_norm": 1.2963308095932007, + "learning_rate": 0.00017099809756760428, + "loss": 3.6241, + "step": 97610 + }, + { + "epoch": 6.632354939529828, + "grad_norm": 0.9203546643257141, + "learning_rate": 0.0001709556325587716, + "loss": 3.3777, + "step": 97615 + }, + { + "epoch": 6.632694659600489, + "grad_norm": 0.8354291915893555, + "learning_rate": 0.00017091316754993884, + "loss": 3.1723, + "step": 97620 + }, + { + "epoch": 6.633034379671151, + "grad_norm": 1.1190476417541504, + "learning_rate": 0.00017087070254110612, + "loss": 3.211, + "step": 97625 + }, + { + "epoch": 6.633374099741813, + "grad_norm": 0.8410468101501465, + "learning_rate": 0.00017082823753227343, + "loss": 3.2589, + "step": 97630 + }, + { + "epoch": 6.633713819812474, + "grad_norm": 0.8552782535552979, + "learning_rate": 0.00017078577252344068, + "loss": 3.2862, + "step": 97635 + }, + { + "epoch": 6.634053539883136, + "grad_norm": 0.8392249345779419, + "learning_rate": 0.00017074330751460796, + "loss": 3.3627, + "step": 97640 + }, + { + "epoch": 6.634393259953798, + "grad_norm": 0.7977872490882874, + "learning_rate": 0.00017070084250577524, + "loss": 3.5542, + "step": 97645 + }, + { + "epoch": 6.634732980024459, + "grad_norm": 0.9388447999954224, + "learning_rate": 0.00017065837749694252, + "loss": 3.487, + "step": 97650 + }, + { + "epoch": 6.6350727000951215, + "grad_norm": 0.7804733514785767, + "learning_rate": 0.0001706159124881098, + "loss": 3.5152, + "step": 97655 + }, + { + "epoch": 6.635412420165784, + "grad_norm": 0.7862626314163208, + "learning_rate": 0.00017057344747927708, + "loss": 3.5668, + "step": 97660 + }, + { + "epoch": 6.635752140236445, + "grad_norm": 0.7736045718193054, + "learning_rate": 0.00017053098247044436, + "loss": 3.2019, + "step": 97665 + }, + { + "epoch": 6.636091860307107, + "grad_norm": 0.9774294495582581, + "learning_rate": 0.00017048851746161164, + "loss": 3.4918, + "step": 97670 + }, + { + "epoch": 6.636431580377769, + "grad_norm": 0.8121858239173889, + "learning_rate": 0.00017044605245277892, + "loss": 3.7203, + "step": 97675 + }, + { + "epoch": 6.63677130044843, + "grad_norm": 0.8712082505226135, + "learning_rate": 0.00017040358744394618, + "loss": 3.4775, + "step": 97680 + }, + { + "epoch": 6.637111020519092, + "grad_norm": 0.8872144818305969, + "learning_rate": 0.00017036112243511348, + "loss": 3.119, + "step": 97685 + }, + { + "epoch": 6.637450740589754, + "grad_norm": 0.819141685962677, + "learning_rate": 0.00017031865742628074, + "loss": 3.6822, + "step": 97690 + }, + { + "epoch": 6.6377904606604154, + "grad_norm": 0.8558163046836853, + "learning_rate": 0.00017027619241744802, + "loss": 3.652, + "step": 97695 + }, + { + "epoch": 6.6381301807310775, + "grad_norm": 1.3472166061401367, + "learning_rate": 0.00017023372740861532, + "loss": 3.3726, + "step": 97700 + }, + { + "epoch": 6.63846990080174, + "grad_norm": 0.7964611649513245, + "learning_rate": 0.00017019126239978258, + "loss": 3.2461, + "step": 97705 + }, + { + "epoch": 6.638809620872401, + "grad_norm": 0.7839168310165405, + "learning_rate": 0.00017014879739094986, + "loss": 3.321, + "step": 97710 + }, + { + "epoch": 6.639149340943063, + "grad_norm": 0.7143064737319946, + "learning_rate": 0.00017010633238211714, + "loss": 3.5427, + "step": 97715 + }, + { + "epoch": 6.639489061013725, + "grad_norm": 0.8498361706733704, + "learning_rate": 0.00017006386737328442, + "loss": 3.426, + "step": 97720 + }, + { + "epoch": 6.639828781084386, + "grad_norm": 1.011250615119934, + "learning_rate": 0.00017002140236445167, + "loss": 3.2305, + "step": 97725 + }, + { + "epoch": 6.640168501155048, + "grad_norm": 0.8560449481010437, + "learning_rate": 0.00016997893735561898, + "loss": 3.5129, + "step": 97730 + }, + { + "epoch": 6.64050822122571, + "grad_norm": 1.1112430095672607, + "learning_rate": 0.00016993647234678626, + "loss": 3.4084, + "step": 97735 + }, + { + "epoch": 6.6408479412963715, + "grad_norm": 1.0720210075378418, + "learning_rate": 0.00016989400733795354, + "loss": 3.4684, + "step": 97740 + }, + { + "epoch": 6.6411876613670335, + "grad_norm": 0.8710058331489563, + "learning_rate": 0.00016985154232912082, + "loss": 3.5861, + "step": 97745 + }, + { + "epoch": 6.641527381437696, + "grad_norm": 0.8212557435035706, + "learning_rate": 0.00016980907732028807, + "loss": 3.4995, + "step": 97750 + }, + { + "epoch": 6.641867101508357, + "grad_norm": 0.9875242114067078, + "learning_rate": 0.00016976661231145538, + "loss": 3.368, + "step": 97755 + }, + { + "epoch": 6.642206821579019, + "grad_norm": 1.0137239694595337, + "learning_rate": 0.00016972414730262263, + "loss": 3.3653, + "step": 97760 + }, + { + "epoch": 6.642546541649681, + "grad_norm": 0.9343211054801941, + "learning_rate": 0.0001696816822937899, + "loss": 3.5679, + "step": 97765 + }, + { + "epoch": 6.642886261720342, + "grad_norm": 0.7763845920562744, + "learning_rate": 0.00016963921728495722, + "loss": 3.3021, + "step": 97770 + }, + { + "epoch": 6.643225981791004, + "grad_norm": 0.9112868309020996, + "learning_rate": 0.00016959675227612447, + "loss": 3.3739, + "step": 97775 + }, + { + "epoch": 6.643565701861666, + "grad_norm": 0.9177730679512024, + "learning_rate": 0.00016955428726729175, + "loss": 3.2359, + "step": 97780 + }, + { + "epoch": 6.6439054219323275, + "grad_norm": 0.8904754519462585, + "learning_rate": 0.00016951182225845903, + "loss": 3.5126, + "step": 97785 + }, + { + "epoch": 6.6442451420029895, + "grad_norm": 0.9986740350723267, + "learning_rate": 0.0001694693572496263, + "loss": 3.4741, + "step": 97790 + }, + { + "epoch": 6.644584862073652, + "grad_norm": 0.7718408703804016, + "learning_rate": 0.0001694268922407936, + "loss": 3.322, + "step": 97795 + }, + { + "epoch": 6.644924582144313, + "grad_norm": 1.255194067955017, + "learning_rate": 0.00016938442723196087, + "loss": 3.2962, + "step": 97800 + }, + { + "epoch": 6.645264302214975, + "grad_norm": 0.9872162342071533, + "learning_rate": 0.00016934196222312815, + "loss": 3.4387, + "step": 97805 + }, + { + "epoch": 6.645604022285637, + "grad_norm": 1.2399612665176392, + "learning_rate": 0.0001692994972142954, + "loss": 3.295, + "step": 97810 + }, + { + "epoch": 6.645943742356298, + "grad_norm": 0.9782564640045166, + "learning_rate": 0.0001692570322054627, + "loss": 3.31, + "step": 97815 + }, + { + "epoch": 6.64628346242696, + "grad_norm": 0.8603094220161438, + "learning_rate": 0.00016921456719662996, + "loss": 3.2038, + "step": 97820 + }, + { + "epoch": 6.646623182497622, + "grad_norm": 0.8670868873596191, + "learning_rate": 0.00016917210218779727, + "loss": 3.4582, + "step": 97825 + }, + { + "epoch": 6.6469629025682835, + "grad_norm": 1.1208715438842773, + "learning_rate": 0.00016912963717896455, + "loss": 3.2104, + "step": 97830 + }, + { + "epoch": 6.647302622638946, + "grad_norm": 0.9903820753097534, + "learning_rate": 0.0001690871721701318, + "loss": 3.5296, + "step": 97835 + }, + { + "epoch": 6.647642342709608, + "grad_norm": 0.8403297066688538, + "learning_rate": 0.0001690447071612991, + "loss": 3.6109, + "step": 97840 + }, + { + "epoch": 6.647982062780269, + "grad_norm": 0.7690026164054871, + "learning_rate": 0.00016900224215246637, + "loss": 3.3585, + "step": 97845 + }, + { + "epoch": 6.648321782850931, + "grad_norm": 0.8832000494003296, + "learning_rate": 0.00016895977714363365, + "loss": 3.467, + "step": 97850 + }, + { + "epoch": 6.648661502921593, + "grad_norm": 0.9940364956855774, + "learning_rate": 0.00016891731213480093, + "loss": 3.1582, + "step": 97855 + }, + { + "epoch": 6.649001222992254, + "grad_norm": 0.9801106452941895, + "learning_rate": 0.0001688748471259682, + "loss": 3.5045, + "step": 97860 + }, + { + "epoch": 6.649340943062916, + "grad_norm": 0.9503993988037109, + "learning_rate": 0.00016883238211713549, + "loss": 3.6056, + "step": 97865 + }, + { + "epoch": 6.649680663133578, + "grad_norm": 0.9465503096580505, + "learning_rate": 0.00016878991710830277, + "loss": 3.2774, + "step": 97870 + }, + { + "epoch": 6.6500203832042395, + "grad_norm": 0.9064024686813354, + "learning_rate": 0.00016874745209947005, + "loss": 3.1446, + "step": 97875 + }, + { + "epoch": 6.650360103274902, + "grad_norm": 1.075541377067566, + "learning_rate": 0.0001687049870906373, + "loss": 3.28, + "step": 97880 + }, + { + "epoch": 6.650699823345564, + "grad_norm": 0.9695895910263062, + "learning_rate": 0.0001686625220818046, + "loss": 3.5558, + "step": 97885 + }, + { + "epoch": 6.651039543416225, + "grad_norm": 0.9236860871315002, + "learning_rate": 0.00016862005707297186, + "loss": 3.5909, + "step": 97890 + }, + { + "epoch": 6.651379263486887, + "grad_norm": 1.0298969745635986, + "learning_rate": 0.00016857759206413914, + "loss": 3.5322, + "step": 97895 + }, + { + "epoch": 6.651718983557549, + "grad_norm": 0.917732834815979, + "learning_rate": 0.00016853512705530645, + "loss": 3.3665, + "step": 97900 + }, + { + "epoch": 6.65205870362821, + "grad_norm": 1.0932207107543945, + "learning_rate": 0.0001684926620464737, + "loss": 3.3746, + "step": 97905 + }, + { + "epoch": 6.652398423698872, + "grad_norm": 1.0874580144882202, + "learning_rate": 0.000168450197037641, + "loss": 3.4578, + "step": 97910 + }, + { + "epoch": 6.652738143769534, + "grad_norm": 0.8795223832130432, + "learning_rate": 0.00016840773202880826, + "loss": 3.2969, + "step": 97915 + }, + { + "epoch": 6.6530778638401955, + "grad_norm": 1.0609577894210815, + "learning_rate": 0.00016836526701997554, + "loss": 3.3834, + "step": 97920 + }, + { + "epoch": 6.653417583910858, + "grad_norm": 0.9489551186561584, + "learning_rate": 0.00016832280201114285, + "loss": 3.5226, + "step": 97925 + }, + { + "epoch": 6.65375730398152, + "grad_norm": 0.6742945313453674, + "learning_rate": 0.0001682803370023101, + "loss": 3.2807, + "step": 97930 + }, + { + "epoch": 6.654097024052181, + "grad_norm": 0.9157310128211975, + "learning_rate": 0.00016823787199347738, + "loss": 3.3887, + "step": 97935 + }, + { + "epoch": 6.654436744122843, + "grad_norm": 0.8461849689483643, + "learning_rate": 0.00016819540698464466, + "loss": 3.2651, + "step": 97940 + }, + { + "epoch": 6.654776464193505, + "grad_norm": 1.1872966289520264, + "learning_rate": 0.00016815294197581194, + "loss": 3.3922, + "step": 97945 + }, + { + "epoch": 6.655116184264166, + "grad_norm": 0.9898809790611267, + "learning_rate": 0.0001681104769669792, + "loss": 3.4992, + "step": 97950 + }, + { + "epoch": 6.655455904334828, + "grad_norm": 0.9593959450721741, + "learning_rate": 0.0001680680119581465, + "loss": 3.5086, + "step": 97955 + }, + { + "epoch": 6.65579562440549, + "grad_norm": 0.8375208377838135, + "learning_rate": 0.00016802554694931378, + "loss": 3.051, + "step": 97960 + }, + { + "epoch": 6.6561353444761515, + "grad_norm": 0.9702267646789551, + "learning_rate": 0.00016798308194048103, + "loss": 3.4573, + "step": 97965 + }, + { + "epoch": 6.656475064546814, + "grad_norm": 0.9078837037086487, + "learning_rate": 0.00016794061693164834, + "loss": 3.4935, + "step": 97970 + }, + { + "epoch": 6.656814784617476, + "grad_norm": 1.0676743984222412, + "learning_rate": 0.0001678981519228156, + "loss": 3.6612, + "step": 97975 + }, + { + "epoch": 6.657154504688137, + "grad_norm": 0.788857638835907, + "learning_rate": 0.00016785568691398287, + "loss": 3.3015, + "step": 97980 + }, + { + "epoch": 6.657494224758799, + "grad_norm": 0.7431078553199768, + "learning_rate": 0.00016781322190515015, + "loss": 3.317, + "step": 97985 + }, + { + "epoch": 6.657833944829461, + "grad_norm": 0.7656048536300659, + "learning_rate": 0.00016777075689631743, + "loss": 3.5001, + "step": 97990 + }, + { + "epoch": 6.658173664900122, + "grad_norm": 0.9151297211647034, + "learning_rate": 0.00016772829188748474, + "loss": 3.4467, + "step": 97995 + }, + { + "epoch": 6.658513384970784, + "grad_norm": 0.9288700222969055, + "learning_rate": 0.000167685826878652, + "loss": 3.3485, + "step": 98000 + }, + { + "epoch": 6.658853105041446, + "grad_norm": 2.0917770862579346, + "learning_rate": 0.00016764336186981927, + "loss": 3.4767, + "step": 98005 + }, + { + "epoch": 6.6591928251121075, + "grad_norm": 0.8727841973304749, + "learning_rate": 0.00016760089686098655, + "loss": 3.1634, + "step": 98010 + }, + { + "epoch": 6.65953254518277, + "grad_norm": 0.7722344994544983, + "learning_rate": 0.00016755843185215383, + "loss": 3.708, + "step": 98015 + }, + { + "epoch": 6.659872265253431, + "grad_norm": 1.1908748149871826, + "learning_rate": 0.0001675159668433211, + "loss": 3.4302, + "step": 98020 + }, + { + "epoch": 6.660211985324093, + "grad_norm": 0.8323476314544678, + "learning_rate": 0.0001674735018344884, + "loss": 3.4241, + "step": 98025 + }, + { + "epoch": 6.660551705394755, + "grad_norm": 1.0906380414962769, + "learning_rate": 0.00016743103682565567, + "loss": 3.5773, + "step": 98030 + }, + { + "epoch": 6.660891425465416, + "grad_norm": 0.8112308979034424, + "learning_rate": 0.00016738857181682293, + "loss": 3.6435, + "step": 98035 + }, + { + "epoch": 6.661231145536078, + "grad_norm": 0.9612371921539307, + "learning_rate": 0.00016734610680799023, + "loss": 3.4213, + "step": 98040 + }, + { + "epoch": 6.66157086560674, + "grad_norm": 1.0858843326568604, + "learning_rate": 0.0001673036417991575, + "loss": 3.359, + "step": 98045 + }, + { + "epoch": 6.6619105856774015, + "grad_norm": 0.7724815607070923, + "learning_rate": 0.00016726117679032477, + "loss": 3.3713, + "step": 98050 + }, + { + "epoch": 6.6622503057480635, + "grad_norm": 0.916249692440033, + "learning_rate": 0.00016721871178149207, + "loss": 3.5027, + "step": 98055 + }, + { + "epoch": 6.662590025818726, + "grad_norm": 0.9715247750282288, + "learning_rate": 0.00016717624677265933, + "loss": 3.4791, + "step": 98060 + }, + { + "epoch": 6.662929745889387, + "grad_norm": 0.9973054528236389, + "learning_rate": 0.0001671337817638266, + "loss": 3.228, + "step": 98065 + }, + { + "epoch": 6.663269465960049, + "grad_norm": 0.8792285919189453, + "learning_rate": 0.0001670913167549939, + "loss": 3.3826, + "step": 98070 + }, + { + "epoch": 6.663609186030711, + "grad_norm": 0.9451937675476074, + "learning_rate": 0.00016704885174616117, + "loss": 3.4915, + "step": 98075 + }, + { + "epoch": 6.663948906101372, + "grad_norm": 1.0245307683944702, + "learning_rate": 0.00016700638673732845, + "loss": 3.253, + "step": 98080 + }, + { + "epoch": 6.664288626172034, + "grad_norm": 1.0046247243881226, + "learning_rate": 0.00016696392172849573, + "loss": 3.3737, + "step": 98085 + }, + { + "epoch": 6.664628346242696, + "grad_norm": 0.8351115584373474, + "learning_rate": 0.000166921456719663, + "loss": 3.4537, + "step": 98090 + }, + { + "epoch": 6.6649680663133575, + "grad_norm": 0.9152930378913879, + "learning_rate": 0.0001668789917108303, + "loss": 3.3453, + "step": 98095 + }, + { + "epoch": 6.6653077863840196, + "grad_norm": 0.7720099687576294, + "learning_rate": 0.00016683652670199757, + "loss": 3.2661, + "step": 98100 + }, + { + "epoch": 6.665647506454682, + "grad_norm": 0.8018552660942078, + "learning_rate": 0.00016679406169316482, + "loss": 3.4261, + "step": 98105 + }, + { + "epoch": 6.665987226525343, + "grad_norm": 0.8648019433021545, + "learning_rate": 0.00016675159668433213, + "loss": 3.1563, + "step": 98110 + }, + { + "epoch": 6.666326946596005, + "grad_norm": 0.8071718811988831, + "learning_rate": 0.00016670913167549938, + "loss": 3.3976, + "step": 98115 + }, + { + "epoch": 6.666666666666667, + "grad_norm": 1.0491482019424438, + "learning_rate": 0.00016666666666666666, + "loss": 3.6159, + "step": 98120 + }, + { + "epoch": 6.667006386737328, + "grad_norm": 0.6909235119819641, + "learning_rate": 0.00016662420165783397, + "loss": 3.3828, + "step": 98125 + }, + { + "epoch": 6.66734610680799, + "grad_norm": 0.7966781854629517, + "learning_rate": 0.00016658173664900122, + "loss": 3.5563, + "step": 98130 + }, + { + "epoch": 6.667685826878652, + "grad_norm": 0.8739886283874512, + "learning_rate": 0.0001665392716401685, + "loss": 3.3712, + "step": 98135 + }, + { + "epoch": 6.6680255469493135, + "grad_norm": 0.868000864982605, + "learning_rate": 0.00016649680663133578, + "loss": 3.2817, + "step": 98140 + }, + { + "epoch": 6.668365267019976, + "grad_norm": 0.9625754952430725, + "learning_rate": 0.00016645434162250306, + "loss": 3.3901, + "step": 98145 + }, + { + "epoch": 6.668704987090638, + "grad_norm": 1.0675256252288818, + "learning_rate": 0.00016641187661367031, + "loss": 3.6323, + "step": 98150 + }, + { + "epoch": 6.669044707161299, + "grad_norm": 0.9916807413101196, + "learning_rate": 0.00016636941160483762, + "loss": 3.3948, + "step": 98155 + }, + { + "epoch": 6.669384427231961, + "grad_norm": 1.1934154033660889, + "learning_rate": 0.0001663269465960049, + "loss": 3.4385, + "step": 98160 + }, + { + "epoch": 6.669724147302623, + "grad_norm": 0.9237998723983765, + "learning_rate": 0.00016628448158717218, + "loss": 3.2442, + "step": 98165 + }, + { + "epoch": 6.670063867373284, + "grad_norm": 0.7651429176330566, + "learning_rate": 0.00016624201657833946, + "loss": 3.5318, + "step": 98170 + }, + { + "epoch": 6.670403587443946, + "grad_norm": 1.1509166955947876, + "learning_rate": 0.00016619955156950671, + "loss": 3.3184, + "step": 98175 + }, + { + "epoch": 6.670743307514608, + "grad_norm": 0.9995712637901306, + "learning_rate": 0.00016615708656067402, + "loss": 3.6023, + "step": 98180 + }, + { + "epoch": 6.6710830275852695, + "grad_norm": 1.2081350088119507, + "learning_rate": 0.00016611462155184127, + "loss": 3.3532, + "step": 98185 + }, + { + "epoch": 6.671422747655932, + "grad_norm": 0.9198182225227356, + "learning_rate": 0.00016607215654300856, + "loss": 3.4782, + "step": 98190 + }, + { + "epoch": 6.671762467726594, + "grad_norm": 0.8567647933959961, + "learning_rate": 0.00016602969153417586, + "loss": 3.3669, + "step": 98195 + }, + { + "epoch": 6.672102187797255, + "grad_norm": 0.8976579308509827, + "learning_rate": 0.00016598722652534312, + "loss": 3.4571, + "step": 98200 + }, + { + "epoch": 6.672441907867917, + "grad_norm": 0.788441002368927, + "learning_rate": 0.0001659447615165104, + "loss": 3.3014, + "step": 98205 + }, + { + "epoch": 6.672781627938579, + "grad_norm": 1.2006947994232178, + "learning_rate": 0.00016590229650767768, + "loss": 3.6216, + "step": 98210 + }, + { + "epoch": 6.67312134800924, + "grad_norm": 0.9174500107765198, + "learning_rate": 0.00016585983149884496, + "loss": 3.2705, + "step": 98215 + }, + { + "epoch": 6.673461068079902, + "grad_norm": 0.886371374130249, + "learning_rate": 0.00016581736649001224, + "loss": 3.4378, + "step": 98220 + }, + { + "epoch": 6.673800788150564, + "grad_norm": 0.8850720524787903, + "learning_rate": 0.00016577490148117952, + "loss": 3.1971, + "step": 98225 + }, + { + "epoch": 6.6741405082212255, + "grad_norm": 0.87474125623703, + "learning_rate": 0.0001657324364723468, + "loss": 3.2724, + "step": 98230 + }, + { + "epoch": 6.674480228291888, + "grad_norm": 0.8662728667259216, + "learning_rate": 0.00016568997146351405, + "loss": 3.5206, + "step": 98235 + }, + { + "epoch": 6.674819948362549, + "grad_norm": 1.2235602140426636, + "learning_rate": 0.00016564750645468136, + "loss": 3.1462, + "step": 98240 + }, + { + "epoch": 6.675159668433211, + "grad_norm": 0.7906063795089722, + "learning_rate": 0.0001656050414458486, + "loss": 3.4825, + "step": 98245 + }, + { + "epoch": 6.675499388503873, + "grad_norm": 0.982133686542511, + "learning_rate": 0.00016556257643701592, + "loss": 3.5641, + "step": 98250 + }, + { + "epoch": 6.675839108574534, + "grad_norm": 0.9806733131408691, + "learning_rate": 0.0001655201114281832, + "loss": 3.0311, + "step": 98255 + }, + { + "epoch": 6.676178828645196, + "grad_norm": 0.9009264707565308, + "learning_rate": 0.00016547764641935045, + "loss": 3.4962, + "step": 98260 + }, + { + "epoch": 6.676518548715858, + "grad_norm": 0.9585974216461182, + "learning_rate": 0.00016543518141051776, + "loss": 3.3709, + "step": 98265 + }, + { + "epoch": 6.6768582687865194, + "grad_norm": 0.8510173559188843, + "learning_rate": 0.000165392716401685, + "loss": 3.2771, + "step": 98270 + }, + { + "epoch": 6.6771979888571815, + "grad_norm": 1.1167799234390259, + "learning_rate": 0.0001653502513928523, + "loss": 3.136, + "step": 98275 + }, + { + "epoch": 6.677537708927844, + "grad_norm": 0.7295570373535156, + "learning_rate": 0.00016530778638401957, + "loss": 3.5368, + "step": 98280 + }, + { + "epoch": 6.677877428998505, + "grad_norm": 0.7594197392463684, + "learning_rate": 0.00016526532137518685, + "loss": 3.6076, + "step": 98285 + }, + { + "epoch": 6.678217149069167, + "grad_norm": 0.9619186520576477, + "learning_rate": 0.00016522285636635413, + "loss": 3.469, + "step": 98290 + }, + { + "epoch": 6.678556869139829, + "grad_norm": 1.0190736055374146, + "learning_rate": 0.0001651803913575214, + "loss": 3.3561, + "step": 98295 + }, + { + "epoch": 6.67889658921049, + "grad_norm": 0.8615480661392212, + "learning_rate": 0.0001651379263486887, + "loss": 3.5176, + "step": 98300 + }, + { + "epoch": 6.679236309281152, + "grad_norm": 0.8286004066467285, + "learning_rate": 0.00016509546133985594, + "loss": 3.5416, + "step": 98305 + }, + { + "epoch": 6.679576029351814, + "grad_norm": 0.7458638548851013, + "learning_rate": 0.00016505299633102325, + "loss": 3.3095, + "step": 98310 + }, + { + "epoch": 6.6799157494224755, + "grad_norm": 0.7598650455474854, + "learning_rate": 0.0001650105313221905, + "loss": 3.5673, + "step": 98315 + }, + { + "epoch": 6.6802554694931375, + "grad_norm": 0.9053853750228882, + "learning_rate": 0.00016496806631335778, + "loss": 3.6509, + "step": 98320 + }, + { + "epoch": 6.6805951895638, + "grad_norm": 1.027963399887085, + "learning_rate": 0.0001649256013045251, + "loss": 3.3882, + "step": 98325 + }, + { + "epoch": 6.680934909634461, + "grad_norm": 0.8867194056510925, + "learning_rate": 0.00016488313629569234, + "loss": 3.5519, + "step": 98330 + }, + { + "epoch": 6.681274629705123, + "grad_norm": 1.1156399250030518, + "learning_rate": 0.00016484067128685965, + "loss": 3.2986, + "step": 98335 + }, + { + "epoch": 6.681614349775785, + "grad_norm": 0.7441017627716064, + "learning_rate": 0.0001647982062780269, + "loss": 3.5647, + "step": 98340 + }, + { + "epoch": 6.681954069846446, + "grad_norm": 0.888398289680481, + "learning_rate": 0.00016475574126919418, + "loss": 3.5477, + "step": 98345 + }, + { + "epoch": 6.682293789917108, + "grad_norm": 0.9858867526054382, + "learning_rate": 0.0001647132762603615, + "loss": 3.4639, + "step": 98350 + }, + { + "epoch": 6.68263350998777, + "grad_norm": 0.895205020904541, + "learning_rate": 0.00016467081125152874, + "loss": 3.5608, + "step": 98355 + }, + { + "epoch": 6.6829732300584315, + "grad_norm": 1.345379114151001, + "learning_rate": 0.00016462834624269602, + "loss": 3.5771, + "step": 98360 + }, + { + "epoch": 6.6833129501290935, + "grad_norm": 1.2730047702789307, + "learning_rate": 0.0001645858812338633, + "loss": 3.4551, + "step": 98365 + }, + { + "epoch": 6.683652670199756, + "grad_norm": 0.746798574924469, + "learning_rate": 0.00016454341622503058, + "loss": 3.1932, + "step": 98370 + }, + { + "epoch": 6.683992390270417, + "grad_norm": 1.0989958047866821, + "learning_rate": 0.00016450095121619784, + "loss": 3.2504, + "step": 98375 + }, + { + "epoch": 6.684332110341079, + "grad_norm": 1.1013727188110352, + "learning_rate": 0.00016445848620736514, + "loss": 3.4277, + "step": 98380 + }, + { + "epoch": 6.684671830411741, + "grad_norm": 1.0964738130569458, + "learning_rate": 0.00016441602119853242, + "loss": 3.7291, + "step": 98385 + }, + { + "epoch": 6.685011550482402, + "grad_norm": 1.2039554119110107, + "learning_rate": 0.00016437355618969968, + "loss": 3.2758, + "step": 98390 + }, + { + "epoch": 6.685351270553064, + "grad_norm": 0.8873413801193237, + "learning_rate": 0.00016433109118086698, + "loss": 3.5989, + "step": 98395 + }, + { + "epoch": 6.685690990623726, + "grad_norm": 1.310025691986084, + "learning_rate": 0.00016428862617203424, + "loss": 3.3078, + "step": 98400 + }, + { + "epoch": 6.6860307106943875, + "grad_norm": 1.0929789543151855, + "learning_rate": 0.00016424616116320152, + "loss": 3.5417, + "step": 98405 + }, + { + "epoch": 6.6863704307650496, + "grad_norm": 0.7772908806800842, + "learning_rate": 0.0001642036961543688, + "loss": 3.2857, + "step": 98410 + }, + { + "epoch": 6.686710150835712, + "grad_norm": 1.139809489250183, + "learning_rate": 0.00016416123114553608, + "loss": 3.4764, + "step": 98415 + }, + { + "epoch": 6.687049870906373, + "grad_norm": 0.7878756523132324, + "learning_rate": 0.00016411876613670338, + "loss": 3.293, + "step": 98420 + }, + { + "epoch": 6.687389590977035, + "grad_norm": 0.760002076625824, + "learning_rate": 0.00016407630112787064, + "loss": 3.3901, + "step": 98425 + }, + { + "epoch": 6.687729311047697, + "grad_norm": 1.0513997077941895, + "learning_rate": 0.00016403383611903792, + "loss": 3.2002, + "step": 98430 + }, + { + "epoch": 6.688069031118358, + "grad_norm": 0.945249617099762, + "learning_rate": 0.0001639913711102052, + "loss": 3.433, + "step": 98435 + }, + { + "epoch": 6.68840875118902, + "grad_norm": 0.902946949005127, + "learning_rate": 0.00016394890610137248, + "loss": 3.385, + "step": 98440 + }, + { + "epoch": 6.688748471259682, + "grad_norm": 0.9585364460945129, + "learning_rate": 0.00016390644109253973, + "loss": 3.4385, + "step": 98445 + }, + { + "epoch": 6.6890881913303435, + "grad_norm": 0.8166432976722717, + "learning_rate": 0.00016386397608370704, + "loss": 3.8471, + "step": 98450 + }, + { + "epoch": 6.689427911401006, + "grad_norm": 0.8916035890579224, + "learning_rate": 0.00016382151107487432, + "loss": 3.3899, + "step": 98455 + }, + { + "epoch": 6.689767631471668, + "grad_norm": 0.8209661841392517, + "learning_rate": 0.00016377904606604157, + "loss": 3.5519, + "step": 98460 + }, + { + "epoch": 6.690107351542329, + "grad_norm": 0.8757140636444092, + "learning_rate": 0.00016373658105720888, + "loss": 3.4573, + "step": 98465 + }, + { + "epoch": 6.690447071612991, + "grad_norm": 0.7047562003135681, + "learning_rate": 0.00016369411604837613, + "loss": 3.2192, + "step": 98470 + }, + { + "epoch": 6.690786791683653, + "grad_norm": 0.7235922813415527, + "learning_rate": 0.0001636516510395434, + "loss": 3.4014, + "step": 98475 + }, + { + "epoch": 6.691126511754314, + "grad_norm": 0.8449955582618713, + "learning_rate": 0.00016360918603071072, + "loss": 3.476, + "step": 98480 + }, + { + "epoch": 6.691466231824976, + "grad_norm": 0.7932359576225281, + "learning_rate": 0.00016356672102187797, + "loss": 3.6212, + "step": 98485 + }, + { + "epoch": 6.691805951895638, + "grad_norm": 1.1313536167144775, + "learning_rate": 0.00016352425601304525, + "loss": 3.2809, + "step": 98490 + }, + { + "epoch": 6.6921456719662995, + "grad_norm": 0.7325586676597595, + "learning_rate": 0.00016348179100421253, + "loss": 3.4094, + "step": 98495 + }, + { + "epoch": 6.692485392036962, + "grad_norm": 1.1796579360961914, + "learning_rate": 0.0001634393259953798, + "loss": 3.6041, + "step": 98500 + }, + { + "epoch": 6.692825112107624, + "grad_norm": 0.87030428647995, + "learning_rate": 0.0001633968609865471, + "loss": 3.184, + "step": 98505 + }, + { + "epoch": 6.693164832178285, + "grad_norm": 1.3303197622299194, + "learning_rate": 0.00016335439597771437, + "loss": 3.4962, + "step": 98510 + }, + { + "epoch": 6.693504552248947, + "grad_norm": 0.9986035227775574, + "learning_rate": 0.00016331193096888165, + "loss": 3.5129, + "step": 98515 + }, + { + "epoch": 6.693844272319609, + "grad_norm": 0.9769651889801025, + "learning_rate": 0.00016326946596004893, + "loss": 3.1848, + "step": 98520 + }, + { + "epoch": 6.69418399239027, + "grad_norm": 0.7941696047782898, + "learning_rate": 0.0001632270009512162, + "loss": 3.0457, + "step": 98525 + }, + { + "epoch": 6.694523712460932, + "grad_norm": 1.054927945137024, + "learning_rate": 0.00016318453594238346, + "loss": 3.4729, + "step": 98530 + }, + { + "epoch": 6.694863432531594, + "grad_norm": 1.0164052248001099, + "learning_rate": 0.00016314207093355077, + "loss": 3.3369, + "step": 98535 + }, + { + "epoch": 6.6952031526022555, + "grad_norm": 0.979567289352417, + "learning_rate": 0.00016309960592471802, + "loss": 3.4113, + "step": 98540 + }, + { + "epoch": 6.695542872672918, + "grad_norm": 0.719219982624054, + "learning_rate": 0.0001630571409158853, + "loss": 3.5766, + "step": 98545 + }, + { + "epoch": 6.69588259274358, + "grad_norm": 1.2375397682189941, + "learning_rate": 0.0001630146759070526, + "loss": 3.4738, + "step": 98550 + }, + { + "epoch": 6.696222312814241, + "grad_norm": 1.1438812017440796, + "learning_rate": 0.00016297221089821987, + "loss": 3.3594, + "step": 98555 + }, + { + "epoch": 6.696562032884903, + "grad_norm": 0.9718422889709473, + "learning_rate": 0.00016292974588938715, + "loss": 3.4889, + "step": 98560 + }, + { + "epoch": 6.696901752955565, + "grad_norm": 1.0074540376663208, + "learning_rate": 0.00016288728088055443, + "loss": 3.423, + "step": 98565 + }, + { + "epoch": 6.697241473026226, + "grad_norm": 1.008644461631775, + "learning_rate": 0.0001628448158717217, + "loss": 3.3451, + "step": 98570 + }, + { + "epoch": 6.697581193096888, + "grad_norm": 0.7972521185874939, + "learning_rate": 0.00016280235086288896, + "loss": 3.3158, + "step": 98575 + }, + { + "epoch": 6.69792091316755, + "grad_norm": 1.1516525745391846, + "learning_rate": 0.00016275988585405627, + "loss": 3.3738, + "step": 98580 + }, + { + "epoch": 6.6982606332382115, + "grad_norm": 3.286757230758667, + "learning_rate": 0.00016271742084522355, + "loss": 3.2527, + "step": 98585 + }, + { + "epoch": 6.698600353308874, + "grad_norm": 0.9683800935745239, + "learning_rate": 0.00016267495583639083, + "loss": 3.2279, + "step": 98590 + }, + { + "epoch": 6.698940073379536, + "grad_norm": 0.818374752998352, + "learning_rate": 0.0001626324908275581, + "loss": 3.1175, + "step": 98595 + }, + { + "epoch": 6.699279793450197, + "grad_norm": 1.0047860145568848, + "learning_rate": 0.00016259002581872536, + "loss": 3.4622, + "step": 98600 + }, + { + "epoch": 6.699619513520859, + "grad_norm": 0.7941889762878418, + "learning_rate": 0.00016254756080989267, + "loss": 3.5446, + "step": 98605 + }, + { + "epoch": 6.699959233591521, + "grad_norm": 1.2283183336257935, + "learning_rate": 0.00016250509580105992, + "loss": 3.3428, + "step": 98610 + }, + { + "epoch": 6.700298953662182, + "grad_norm": 0.9395216107368469, + "learning_rate": 0.0001624626307922272, + "loss": 3.2785, + "step": 98615 + }, + { + "epoch": 6.700638673732844, + "grad_norm": 1.0770148038864136, + "learning_rate": 0.0001624201657833945, + "loss": 3.4937, + "step": 98620 + }, + { + "epoch": 6.700978393803506, + "grad_norm": 0.912502110004425, + "learning_rate": 0.00016237770077456176, + "loss": 3.3356, + "step": 98625 + }, + { + "epoch": 6.7013181138741675, + "grad_norm": 0.8245057463645935, + "learning_rate": 0.00016233523576572904, + "loss": 3.3062, + "step": 98630 + }, + { + "epoch": 6.70165783394483, + "grad_norm": 0.9143803119659424, + "learning_rate": 0.00016229277075689632, + "loss": 3.4017, + "step": 98635 + }, + { + "epoch": 6.701997554015492, + "grad_norm": 1.201717734336853, + "learning_rate": 0.0001622503057480636, + "loss": 3.6408, + "step": 98640 + }, + { + "epoch": 6.702337274086153, + "grad_norm": 0.8823744058609009, + "learning_rate": 0.00016220784073923088, + "loss": 3.3644, + "step": 98645 + }, + { + "epoch": 6.702676994156815, + "grad_norm": 0.938220739364624, + "learning_rate": 0.00016216537573039816, + "loss": 3.3108, + "step": 98650 + }, + { + "epoch": 6.703016714227477, + "grad_norm": 0.8770846128463745, + "learning_rate": 0.00016212291072156544, + "loss": 3.2519, + "step": 98655 + }, + { + "epoch": 6.703356434298138, + "grad_norm": 0.8435313105583191, + "learning_rate": 0.0001620804457127327, + "loss": 3.4498, + "step": 98660 + }, + { + "epoch": 6.7036961543688, + "grad_norm": 0.8561041355133057, + "learning_rate": 0.0001620379807039, + "loss": 3.4536, + "step": 98665 + }, + { + "epoch": 6.704035874439462, + "grad_norm": 0.828154444694519, + "learning_rate": 0.00016199551569506725, + "loss": 3.4274, + "step": 98670 + }, + { + "epoch": 6.7043755945101235, + "grad_norm": 0.8490538597106934, + "learning_rate": 0.00016195305068623456, + "loss": 3.3161, + "step": 98675 + }, + { + "epoch": 6.704715314580786, + "grad_norm": 0.8369812965393066, + "learning_rate": 0.00016191058567740184, + "loss": 3.3717, + "step": 98680 + }, + { + "epoch": 6.705055034651448, + "grad_norm": 0.7257477641105652, + "learning_rate": 0.0001618681206685691, + "loss": 3.39, + "step": 98685 + }, + { + "epoch": 6.705394754722109, + "grad_norm": 0.9297597408294678, + "learning_rate": 0.0001618256556597364, + "loss": 3.3216, + "step": 98690 + }, + { + "epoch": 6.705734474792771, + "grad_norm": 1.0446336269378662, + "learning_rate": 0.00016178319065090365, + "loss": 3.5106, + "step": 98695 + }, + { + "epoch": 6.706074194863432, + "grad_norm": 1.429106593132019, + "learning_rate": 0.00016174072564207093, + "loss": 3.3439, + "step": 98700 + }, + { + "epoch": 6.706413914934094, + "grad_norm": 1.1665338277816772, + "learning_rate": 0.0001616982606332382, + "loss": 3.4741, + "step": 98705 + }, + { + "epoch": 6.706753635004756, + "grad_norm": 0.8283876180648804, + "learning_rate": 0.0001616557956244055, + "loss": 3.4838, + "step": 98710 + }, + { + "epoch": 6.7070933550754175, + "grad_norm": 0.9966378808021545, + "learning_rate": 0.00016161333061557277, + "loss": 3.4436, + "step": 98715 + }, + { + "epoch": 6.70743307514608, + "grad_norm": 0.8640484809875488, + "learning_rate": 0.00016157086560674005, + "loss": 3.4366, + "step": 98720 + }, + { + "epoch": 6.707772795216742, + "grad_norm": 0.8649706840515137, + "learning_rate": 0.00016152840059790733, + "loss": 3.377, + "step": 98725 + }, + { + "epoch": 6.708112515287403, + "grad_norm": 1.0464508533477783, + "learning_rate": 0.0001614859355890746, + "loss": 3.4688, + "step": 98730 + }, + { + "epoch": 6.708452235358065, + "grad_norm": 0.9635571241378784, + "learning_rate": 0.0001614434705802419, + "loss": 3.3284, + "step": 98735 + }, + { + "epoch": 6.708791955428727, + "grad_norm": 1.0976835489273071, + "learning_rate": 0.00016140100557140915, + "loss": 3.2723, + "step": 98740 + }, + { + "epoch": 6.709131675499388, + "grad_norm": 0.7682868242263794, + "learning_rate": 0.00016135854056257643, + "loss": 3.4166, + "step": 98745 + }, + { + "epoch": 6.70947139557005, + "grad_norm": 0.8764656186103821, + "learning_rate": 0.00016131607555374373, + "loss": 3.4398, + "step": 98750 + }, + { + "epoch": 6.709811115640712, + "grad_norm": 0.7481299042701721, + "learning_rate": 0.000161273610544911, + "loss": 3.7501, + "step": 98755 + }, + { + "epoch": 6.7101508357113735, + "grad_norm": 0.9635537266731262, + "learning_rate": 0.0001612311455360783, + "loss": 3.5712, + "step": 98760 + }, + { + "epoch": 6.710490555782036, + "grad_norm": 0.8888024091720581, + "learning_rate": 0.00016118868052724555, + "loss": 3.7231, + "step": 98765 + }, + { + "epoch": 6.710830275852698, + "grad_norm": 0.8020325303077698, + "learning_rate": 0.00016114621551841283, + "loss": 3.5963, + "step": 98770 + }, + { + "epoch": 6.711169995923359, + "grad_norm": 0.9146431684494019, + "learning_rate": 0.00016110375050958013, + "loss": 3.3041, + "step": 98775 + }, + { + "epoch": 6.711509715994021, + "grad_norm": 1.0397709608078003, + "learning_rate": 0.0001610612855007474, + "loss": 3.4637, + "step": 98780 + }, + { + "epoch": 6.711849436064683, + "grad_norm": 0.9316775798797607, + "learning_rate": 0.00016101882049191467, + "loss": 3.3969, + "step": 98785 + }, + { + "epoch": 6.712189156135344, + "grad_norm": 0.9348915815353394, + "learning_rate": 0.00016097635548308195, + "loss": 3.3976, + "step": 98790 + }, + { + "epoch": 6.712528876206006, + "grad_norm": 1.0540988445281982, + "learning_rate": 0.00016093389047424923, + "loss": 3.4517, + "step": 98795 + }, + { + "epoch": 6.712868596276668, + "grad_norm": 0.8583312630653381, + "learning_rate": 0.00016089142546541648, + "loss": 3.352, + "step": 98800 + }, + { + "epoch": 6.7132083163473295, + "grad_norm": 0.8585736155509949, + "learning_rate": 0.0001608489604565838, + "loss": 3.375, + "step": 98805 + }, + { + "epoch": 6.713548036417992, + "grad_norm": 1.1303921937942505, + "learning_rate": 0.00016080649544775107, + "loss": 3.4667, + "step": 98810 + }, + { + "epoch": 6.713887756488654, + "grad_norm": 0.9784711599349976, + "learning_rate": 0.00016076403043891832, + "loss": 3.4703, + "step": 98815 + }, + { + "epoch": 6.714227476559315, + "grad_norm": 0.928087055683136, + "learning_rate": 0.00016072156543008563, + "loss": 3.5231, + "step": 98820 + }, + { + "epoch": 6.714567196629977, + "grad_norm": 0.9439919590950012, + "learning_rate": 0.00016067910042125288, + "loss": 3.4182, + "step": 98825 + }, + { + "epoch": 6.714906916700639, + "grad_norm": 0.782618522644043, + "learning_rate": 0.00016063663541242016, + "loss": 3.3069, + "step": 98830 + }, + { + "epoch": 6.7152466367713, + "grad_norm": 0.8889277577400208, + "learning_rate": 0.00016059417040358744, + "loss": 3.4738, + "step": 98835 + }, + { + "epoch": 6.715586356841962, + "grad_norm": 0.9240031242370605, + "learning_rate": 0.00016055170539475472, + "loss": 3.4911, + "step": 98840 + }, + { + "epoch": 6.715926076912624, + "grad_norm": 1.0841857194900513, + "learning_rate": 0.00016050924038592203, + "loss": 3.035, + "step": 98845 + }, + { + "epoch": 6.7162657969832855, + "grad_norm": 1.4616115093231201, + "learning_rate": 0.00016046677537708928, + "loss": 3.1898, + "step": 98850 + }, + { + "epoch": 6.716605517053948, + "grad_norm": 1.3437278270721436, + "learning_rate": 0.00016042431036825656, + "loss": 3.4602, + "step": 98855 + }, + { + "epoch": 6.71694523712461, + "grad_norm": 0.8307734727859497, + "learning_rate": 0.00016038184535942384, + "loss": 3.2844, + "step": 98860 + }, + { + "epoch": 6.717284957195271, + "grad_norm": 0.7899867296218872, + "learning_rate": 0.00016033938035059112, + "loss": 3.6887, + "step": 98865 + }, + { + "epoch": 6.717624677265933, + "grad_norm": 0.823029100894928, + "learning_rate": 0.00016029691534175837, + "loss": 3.5462, + "step": 98870 + }, + { + "epoch": 6.717964397336595, + "grad_norm": 0.7253313064575195, + "learning_rate": 0.00016025445033292568, + "loss": 3.2789, + "step": 98875 + }, + { + "epoch": 6.718304117407256, + "grad_norm": 1.0103795528411865, + "learning_rate": 0.00016021198532409296, + "loss": 3.6321, + "step": 98880 + }, + { + "epoch": 6.718643837477918, + "grad_norm": 0.8535981178283691, + "learning_rate": 0.00016016952031526021, + "loss": 3.2313, + "step": 98885 + }, + { + "epoch": 6.71898355754858, + "grad_norm": 0.8998686075210571, + "learning_rate": 0.00016012705530642752, + "loss": 3.4221, + "step": 98890 + }, + { + "epoch": 6.7193232776192415, + "grad_norm": 0.845862865447998, + "learning_rate": 0.00016008459029759477, + "loss": 3.1784, + "step": 98895 + }, + { + "epoch": 6.719662997689904, + "grad_norm": 0.9197079539299011, + "learning_rate": 0.00016004212528876205, + "loss": 3.3326, + "step": 98900 + }, + { + "epoch": 6.720002717760566, + "grad_norm": 0.8116263747215271, + "learning_rate": 0.00015999966027992936, + "loss": 3.3147, + "step": 98905 + }, + { + "epoch": 6.720342437831227, + "grad_norm": 0.8154120445251465, + "learning_rate": 0.00015995719527109662, + "loss": 3.4958, + "step": 98910 + }, + { + "epoch": 6.720682157901889, + "grad_norm": 1.0729119777679443, + "learning_rate": 0.0001599147302622639, + "loss": 3.3104, + "step": 98915 + }, + { + "epoch": 6.72102187797255, + "grad_norm": 0.7671745419502258, + "learning_rate": 0.00015987226525343118, + "loss": 3.5198, + "step": 98920 + }, + { + "epoch": 6.721361598043212, + "grad_norm": 0.8769633173942566, + "learning_rate": 0.00015982980024459846, + "loss": 3.5287, + "step": 98925 + }, + { + "epoch": 6.721701318113874, + "grad_norm": 0.766202986240387, + "learning_rate": 0.00015978733523576574, + "loss": 3.5148, + "step": 98930 + }, + { + "epoch": 6.7220410381845355, + "grad_norm": 1.0282957553863525, + "learning_rate": 0.00015974487022693302, + "loss": 3.5726, + "step": 98935 + }, + { + "epoch": 6.7223807582551975, + "grad_norm": 0.953100323677063, + "learning_rate": 0.0001597024052181003, + "loss": 3.464, + "step": 98940 + }, + { + "epoch": 6.72272047832586, + "grad_norm": 0.8177765011787415, + "learning_rate": 0.00015965994020926758, + "loss": 3.4094, + "step": 98945 + }, + { + "epoch": 6.723060198396521, + "grad_norm": 0.8799620270729065, + "learning_rate": 0.00015961747520043486, + "loss": 3.4807, + "step": 98950 + }, + { + "epoch": 6.723399918467183, + "grad_norm": 0.7455402612686157, + "learning_rate": 0.0001595750101916021, + "loss": 3.462, + "step": 98955 + }, + { + "epoch": 6.723739638537845, + "grad_norm": 0.8464288115501404, + "learning_rate": 0.00015953254518276942, + "loss": 3.3744, + "step": 98960 + }, + { + "epoch": 6.724079358608506, + "grad_norm": 0.7706100940704346, + "learning_rate": 0.00015949008017393667, + "loss": 3.6032, + "step": 98965 + }, + { + "epoch": 6.724419078679168, + "grad_norm": 1.0172011852264404, + "learning_rate": 0.00015944761516510395, + "loss": 3.5016, + "step": 98970 + }, + { + "epoch": 6.72475879874983, + "grad_norm": 0.9133962988853455, + "learning_rate": 0.00015940515015627126, + "loss": 3.1873, + "step": 98975 + }, + { + "epoch": 6.7250985188204915, + "grad_norm": 0.8787901401519775, + "learning_rate": 0.0001593626851474385, + "loss": 3.4418, + "step": 98980 + }, + { + "epoch": 6.7254382388911536, + "grad_norm": 0.7707837224006653, + "learning_rate": 0.0001593202201386058, + "loss": 3.366, + "step": 98985 + }, + { + "epoch": 6.725777958961816, + "grad_norm": 1.1194208860397339, + "learning_rate": 0.00015927775512977307, + "loss": 3.6229, + "step": 98990 + }, + { + "epoch": 6.726117679032477, + "grad_norm": 0.8416011929512024, + "learning_rate": 0.00015923529012094035, + "loss": 3.6009, + "step": 98995 + }, + { + "epoch": 6.726457399103139, + "grad_norm": 0.7884020209312439, + "learning_rate": 0.0001591928251121076, + "loss": 3.4343, + "step": 99000 + }, + { + "epoch": 6.726797119173801, + "grad_norm": 0.7559952139854431, + "learning_rate": 0.0001591503601032749, + "loss": 3.6023, + "step": 99005 + }, + { + "epoch": 6.727136839244462, + "grad_norm": 1.0300061702728271, + "learning_rate": 0.0001591078950944422, + "loss": 3.2328, + "step": 99010 + }, + { + "epoch": 6.727476559315124, + "grad_norm": 0.9809158444404602, + "learning_rate": 0.00015906543008560947, + "loss": 3.5773, + "step": 99015 + }, + { + "epoch": 6.727816279385786, + "grad_norm": 0.9537068605422974, + "learning_rate": 0.00015902296507677675, + "loss": 3.3722, + "step": 99020 + }, + { + "epoch": 6.7281559994564475, + "grad_norm": 1.047251582145691, + "learning_rate": 0.000158980500067944, + "loss": 3.5564, + "step": 99025 + }, + { + "epoch": 6.72849571952711, + "grad_norm": 1.0032260417938232, + "learning_rate": 0.0001589380350591113, + "loss": 3.5207, + "step": 99030 + }, + { + "epoch": 6.728835439597772, + "grad_norm": 0.9746182560920715, + "learning_rate": 0.00015889557005027856, + "loss": 3.3264, + "step": 99035 + }, + { + "epoch": 6.729175159668433, + "grad_norm": 0.9321110248565674, + "learning_rate": 0.00015885310504144584, + "loss": 3.5463, + "step": 99040 + }, + { + "epoch": 6.729514879739095, + "grad_norm": 0.8984461426734924, + "learning_rate": 0.00015881064003261315, + "loss": 3.3981, + "step": 99045 + }, + { + "epoch": 6.729854599809757, + "grad_norm": 0.9931296110153198, + "learning_rate": 0.0001587681750237804, + "loss": 3.3252, + "step": 99050 + }, + { + "epoch": 6.730194319880418, + "grad_norm": 0.8704235553741455, + "learning_rate": 0.00015872571001494768, + "loss": 3.436, + "step": 99055 + }, + { + "epoch": 6.73053403995108, + "grad_norm": 1.1477922201156616, + "learning_rate": 0.00015868324500611496, + "loss": 3.3313, + "step": 99060 + }, + { + "epoch": 6.730873760021742, + "grad_norm": 0.9511985182762146, + "learning_rate": 0.00015864077999728224, + "loss": 3.1581, + "step": 99065 + }, + { + "epoch": 6.7312134800924035, + "grad_norm": 0.6877484321594238, + "learning_rate": 0.00015859831498844952, + "loss": 3.7375, + "step": 99070 + }, + { + "epoch": 6.731553200163066, + "grad_norm": 0.978493869304657, + "learning_rate": 0.0001585558499796168, + "loss": 3.3143, + "step": 99075 + }, + { + "epoch": 6.731892920233728, + "grad_norm": 0.873456597328186, + "learning_rate": 0.00015851338497078408, + "loss": 3.4101, + "step": 99080 + }, + { + "epoch": 6.732232640304389, + "grad_norm": 0.7577624917030334, + "learning_rate": 0.00015847091996195134, + "loss": 3.4635, + "step": 99085 + }, + { + "epoch": 6.732572360375051, + "grad_norm": 0.9603385329246521, + "learning_rate": 0.00015842845495311864, + "loss": 3.437, + "step": 99090 + }, + { + "epoch": 6.732912080445713, + "grad_norm": 0.9179843068122864, + "learning_rate": 0.0001583859899442859, + "loss": 3.338, + "step": 99095 + }, + { + "epoch": 6.733251800516374, + "grad_norm": 0.8067453503608704, + "learning_rate": 0.0001583435249354532, + "loss": 3.3285, + "step": 99100 + }, + { + "epoch": 6.733591520587036, + "grad_norm": 0.8273066878318787, + "learning_rate": 0.00015830105992662048, + "loss": 3.2436, + "step": 99105 + }, + { + "epoch": 6.733931240657698, + "grad_norm": 0.7473201751708984, + "learning_rate": 0.00015825859491778774, + "loss": 3.2903, + "step": 99110 + }, + { + "epoch": 6.7342709607283595, + "grad_norm": 0.9552164673805237, + "learning_rate": 0.00015821612990895504, + "loss": 3.4038, + "step": 99115 + }, + { + "epoch": 6.734610680799022, + "grad_norm": 0.9579547643661499, + "learning_rate": 0.0001581736649001223, + "loss": 3.3109, + "step": 99120 + }, + { + "epoch": 6.734950400869684, + "grad_norm": 0.8588147759437561, + "learning_rate": 0.00015813119989128958, + "loss": 3.2811, + "step": 99125 + }, + { + "epoch": 6.735290120940345, + "grad_norm": 0.771340012550354, + "learning_rate": 0.00015808873488245686, + "loss": 3.6911, + "step": 99130 + }, + { + "epoch": 6.735629841011007, + "grad_norm": 0.8099253177642822, + "learning_rate": 0.00015804626987362414, + "loss": 3.2782, + "step": 99135 + }, + { + "epoch": 6.735969561081669, + "grad_norm": 0.7799479365348816, + "learning_rate": 0.00015800380486479142, + "loss": 3.4099, + "step": 99140 + }, + { + "epoch": 6.73630928115233, + "grad_norm": 0.9630786180496216, + "learning_rate": 0.0001579613398559587, + "loss": 3.3546, + "step": 99145 + }, + { + "epoch": 6.736649001222992, + "grad_norm": 0.8047103881835938, + "learning_rate": 0.00015791887484712598, + "loss": 3.3906, + "step": 99150 + }, + { + "epoch": 6.736988721293654, + "grad_norm": 1.1070650815963745, + "learning_rate": 0.00015787640983829323, + "loss": 3.3626, + "step": 99155 + }, + { + "epoch": 6.7373284413643155, + "grad_norm": 0.8413215279579163, + "learning_rate": 0.00015783394482946054, + "loss": 3.3432, + "step": 99160 + }, + { + "epoch": 6.737668161434978, + "grad_norm": 0.908247172832489, + "learning_rate": 0.0001577914798206278, + "loss": 3.4784, + "step": 99165 + }, + { + "epoch": 6.73800788150564, + "grad_norm": 1.0514072179794312, + "learning_rate": 0.00015774901481179507, + "loss": 3.2066, + "step": 99170 + }, + { + "epoch": 6.738347601576301, + "grad_norm": 0.8807491064071655, + "learning_rate": 0.00015770654980296238, + "loss": 3.4516, + "step": 99175 + }, + { + "epoch": 6.738687321646963, + "grad_norm": 0.9485208988189697, + "learning_rate": 0.00015766408479412963, + "loss": 3.4672, + "step": 99180 + }, + { + "epoch": 6.739027041717625, + "grad_norm": 0.9426895976066589, + "learning_rate": 0.00015762161978529694, + "loss": 3.4479, + "step": 99185 + }, + { + "epoch": 6.739366761788286, + "grad_norm": 1.1083801984786987, + "learning_rate": 0.0001575791547764642, + "loss": 3.4926, + "step": 99190 + }, + { + "epoch": 6.739706481858948, + "grad_norm": 0.9270679950714111, + "learning_rate": 0.00015753668976763147, + "loss": 3.4702, + "step": 99195 + }, + { + "epoch": 6.74004620192961, + "grad_norm": 0.8887948989868164, + "learning_rate": 0.00015749422475879878, + "loss": 3.5439, + "step": 99200 + }, + { + "epoch": 6.7403859220002715, + "grad_norm": 0.9404715299606323, + "learning_rate": 0.00015745175974996603, + "loss": 3.6696, + "step": 99205 + }, + { + "epoch": 6.740725642070934, + "grad_norm": 1.0021692514419556, + "learning_rate": 0.0001574092947411333, + "loss": 3.5114, + "step": 99210 + }, + { + "epoch": 6.741065362141596, + "grad_norm": 0.9406409859657288, + "learning_rate": 0.0001573668297323006, + "loss": 3.2796, + "step": 99215 + }, + { + "epoch": 6.741405082212257, + "grad_norm": 0.9603568315505981, + "learning_rate": 0.00015732436472346787, + "loss": 3.3923, + "step": 99220 + }, + { + "epoch": 6.741744802282919, + "grad_norm": 0.7589702606201172, + "learning_rate": 0.00015728189971463512, + "loss": 3.297, + "step": 99225 + }, + { + "epoch": 6.742084522353581, + "grad_norm": 0.9933352470397949, + "learning_rate": 0.00015723943470580243, + "loss": 3.4646, + "step": 99230 + }, + { + "epoch": 6.742424242424242, + "grad_norm": 0.8444876670837402, + "learning_rate": 0.0001571969696969697, + "loss": 3.4444, + "step": 99235 + }, + { + "epoch": 6.742763962494904, + "grad_norm": 1.0127308368682861, + "learning_rate": 0.00015715450468813696, + "loss": 3.5526, + "step": 99240 + }, + { + "epoch": 6.743103682565566, + "grad_norm": 0.9706007838249207, + "learning_rate": 0.00015711203967930427, + "loss": 3.2996, + "step": 99245 + }, + { + "epoch": 6.7434434026362275, + "grad_norm": 0.799593448638916, + "learning_rate": 0.00015706957467047152, + "loss": 3.3443, + "step": 99250 + }, + { + "epoch": 6.74378312270689, + "grad_norm": 0.9249041080474854, + "learning_rate": 0.0001570271096616388, + "loss": 3.6821, + "step": 99255 + }, + { + "epoch": 6.744122842777552, + "grad_norm": 0.8527132868766785, + "learning_rate": 0.00015698464465280609, + "loss": 3.4674, + "step": 99260 + }, + { + "epoch": 6.744462562848213, + "grad_norm": 0.8478440046310425, + "learning_rate": 0.00015694217964397337, + "loss": 3.2863, + "step": 99265 + }, + { + "epoch": 6.744802282918875, + "grad_norm": 0.9996679425239563, + "learning_rate": 0.00015689971463514067, + "loss": 3.2996, + "step": 99270 + }, + { + "epoch": 6.745142002989537, + "grad_norm": 0.9850682020187378, + "learning_rate": 0.00015685724962630793, + "loss": 3.2737, + "step": 99275 + }, + { + "epoch": 6.745481723060198, + "grad_norm": 0.7644358277320862, + "learning_rate": 0.0001568147846174752, + "loss": 3.3969, + "step": 99280 + }, + { + "epoch": 6.74582144313086, + "grad_norm": 0.9883807897567749, + "learning_rate": 0.00015677231960864249, + "loss": 3.3918, + "step": 99285 + }, + { + "epoch": 6.746161163201522, + "grad_norm": 0.8288475871086121, + "learning_rate": 0.00015672985459980977, + "loss": 3.1479, + "step": 99290 + }, + { + "epoch": 6.746500883272184, + "grad_norm": 0.9836732745170593, + "learning_rate": 0.00015668738959097702, + "loss": 3.2021, + "step": 99295 + }, + { + "epoch": 6.746840603342846, + "grad_norm": 0.8206176161766052, + "learning_rate": 0.00015664492458214433, + "loss": 3.3835, + "step": 99300 + }, + { + "epoch": 6.747180323413508, + "grad_norm": 0.7885094881057739, + "learning_rate": 0.0001566024595733116, + "loss": 3.5296, + "step": 99305 + }, + { + "epoch": 6.747520043484169, + "grad_norm": 1.1571476459503174, + "learning_rate": 0.00015655999456447886, + "loss": 3.1853, + "step": 99310 + }, + { + "epoch": 6.747859763554831, + "grad_norm": 0.9673506617546082, + "learning_rate": 0.00015651752955564617, + "loss": 3.294, + "step": 99315 + }, + { + "epoch": 6.748199483625493, + "grad_norm": 0.8363659977912903, + "learning_rate": 0.00015647506454681342, + "loss": 3.4402, + "step": 99320 + }, + { + "epoch": 6.748539203696154, + "grad_norm": 0.8963209390640259, + "learning_rate": 0.0001564325995379807, + "loss": 3.614, + "step": 99325 + }, + { + "epoch": 6.748878923766816, + "grad_norm": 0.9322326183319092, + "learning_rate": 0.000156390134529148, + "loss": 3.5427, + "step": 99330 + }, + { + "epoch": 6.749218643837478, + "grad_norm": 1.0387027263641357, + "learning_rate": 0.00015634766952031526, + "loss": 3.3277, + "step": 99335 + }, + { + "epoch": 6.74955836390814, + "grad_norm": 1.0139487981796265, + "learning_rate": 0.00015630520451148254, + "loss": 3.1374, + "step": 99340 + }, + { + "epoch": 6.749898083978802, + "grad_norm": 1.00033700466156, + "learning_rate": 0.00015626273950264982, + "loss": 3.7285, + "step": 99345 + }, + { + "epoch": 6.750237804049464, + "grad_norm": 0.9665691256523132, + "learning_rate": 0.0001562202744938171, + "loss": 3.3824, + "step": 99350 + }, + { + "epoch": 6.750577524120125, + "grad_norm": 0.8024733662605286, + "learning_rate": 0.00015617780948498435, + "loss": 3.2926, + "step": 99355 + }, + { + "epoch": 6.750917244190787, + "grad_norm": 0.8492614030838013, + "learning_rate": 0.00015613534447615166, + "loss": 3.3834, + "step": 99360 + }, + { + "epoch": 6.751256964261449, + "grad_norm": 0.7802373766899109, + "learning_rate": 0.00015609287946731894, + "loss": 3.4515, + "step": 99365 + }, + { + "epoch": 6.75159668433211, + "grad_norm": 0.8119608759880066, + "learning_rate": 0.00015605041445848622, + "loss": 3.4682, + "step": 99370 + }, + { + "epoch": 6.751936404402772, + "grad_norm": 1.0602524280548096, + "learning_rate": 0.0001560079494496535, + "loss": 3.3894, + "step": 99375 + }, + { + "epoch": 6.7522761244734335, + "grad_norm": 1.0069350004196167, + "learning_rate": 0.00015596548444082075, + "loss": 3.3941, + "step": 99380 + }, + { + "epoch": 6.752615844544096, + "grad_norm": 0.7759472131729126, + "learning_rate": 0.00015592301943198806, + "loss": 3.5038, + "step": 99385 + }, + { + "epoch": 6.752955564614758, + "grad_norm": 1.1187303066253662, + "learning_rate": 0.0001558805544231553, + "loss": 3.5269, + "step": 99390 + }, + { + "epoch": 6.753295284685419, + "grad_norm": 0.8150542974472046, + "learning_rate": 0.0001558380894143226, + "loss": 3.3218, + "step": 99395 + }, + { + "epoch": 6.753635004756081, + "grad_norm": 0.9629438519477844, + "learning_rate": 0.0001557956244054899, + "loss": 3.5151, + "step": 99400 + }, + { + "epoch": 6.753974724826743, + "grad_norm": 0.8733299970626831, + "learning_rate": 0.00015575315939665715, + "loss": 3.4309, + "step": 99405 + }, + { + "epoch": 6.754314444897404, + "grad_norm": 0.8583925366401672, + "learning_rate": 0.00015571069438782443, + "loss": 3.4179, + "step": 99410 + }, + { + "epoch": 6.754654164968066, + "grad_norm": 0.93126380443573, + "learning_rate": 0.0001556682293789917, + "loss": 3.4512, + "step": 99415 + }, + { + "epoch": 6.754993885038728, + "grad_norm": 1.0962668657302856, + "learning_rate": 0.000155625764370159, + "loss": 3.3005, + "step": 99420 + }, + { + "epoch": 6.7553336051093895, + "grad_norm": 1.083292007446289, + "learning_rate": 0.00015558329936132625, + "loss": 3.6887, + "step": 99425 + }, + { + "epoch": 6.755673325180052, + "grad_norm": 0.9044144749641418, + "learning_rate": 0.00015554083435249355, + "loss": 3.451, + "step": 99430 + }, + { + "epoch": 6.756013045250714, + "grad_norm": 0.8609837889671326, + "learning_rate": 0.00015549836934366083, + "loss": 3.47, + "step": 99435 + }, + { + "epoch": 6.756352765321375, + "grad_norm": 0.8209478855133057, + "learning_rate": 0.0001554559043348281, + "loss": 3.4512, + "step": 99440 + }, + { + "epoch": 6.756692485392037, + "grad_norm": 0.9232655763626099, + "learning_rate": 0.0001554134393259954, + "loss": 3.5651, + "step": 99445 + }, + { + "epoch": 6.757032205462699, + "grad_norm": 0.6874026656150818, + "learning_rate": 0.00015537097431716265, + "loss": 3.5951, + "step": 99450 + }, + { + "epoch": 6.75737192553336, + "grad_norm": 0.7462101578712463, + "learning_rate": 0.00015532850930832995, + "loss": 3.5096, + "step": 99455 + }, + { + "epoch": 6.757711645604022, + "grad_norm": 0.8444480299949646, + "learning_rate": 0.0001552860442994972, + "loss": 3.4984, + "step": 99460 + }, + { + "epoch": 6.758051365674684, + "grad_norm": 1.2011293172836304, + "learning_rate": 0.0001552435792906645, + "loss": 3.3675, + "step": 99465 + }, + { + "epoch": 6.7583910857453455, + "grad_norm": 1.0360223054885864, + "learning_rate": 0.0001552011142818318, + "loss": 3.3399, + "step": 99470 + }, + { + "epoch": 6.758730805816008, + "grad_norm": 0.8913591504096985, + "learning_rate": 0.00015515864927299905, + "loss": 3.2486, + "step": 99475 + }, + { + "epoch": 6.75907052588667, + "grad_norm": 0.7857750654220581, + "learning_rate": 0.00015511618426416633, + "loss": 3.3083, + "step": 99480 + }, + { + "epoch": 6.759410245957331, + "grad_norm": 0.7855299115180969, + "learning_rate": 0.0001550737192553336, + "loss": 3.6492, + "step": 99485 + }, + { + "epoch": 6.759749966027993, + "grad_norm": 1.3507615327835083, + "learning_rate": 0.0001550312542465009, + "loss": 3.6018, + "step": 99490 + }, + { + "epoch": 6.760089686098655, + "grad_norm": 0.9530397057533264, + "learning_rate": 0.00015498878923766817, + "loss": 3.3512, + "step": 99495 + }, + { + "epoch": 6.760429406169316, + "grad_norm": 1.5734102725982666, + "learning_rate": 0.00015494632422883545, + "loss": 3.4982, + "step": 99500 + }, + { + "epoch": 6.760769126239978, + "grad_norm": 0.8163034915924072, + "learning_rate": 0.00015490385922000273, + "loss": 3.6466, + "step": 99505 + }, + { + "epoch": 6.76110884631064, + "grad_norm": 0.8135191798210144, + "learning_rate": 0.00015486139421116998, + "loss": 3.3615, + "step": 99510 + }, + { + "epoch": 6.7614485663813015, + "grad_norm": 0.8595287799835205, + "learning_rate": 0.0001548189292023373, + "loss": 3.5014, + "step": 99515 + }, + { + "epoch": 6.761788286451964, + "grad_norm": 0.9657663702964783, + "learning_rate": 0.00015477646419350454, + "loss": 3.3086, + "step": 99520 + }, + { + "epoch": 6.762128006522626, + "grad_norm": 0.9085947871208191, + "learning_rate": 0.00015473399918467182, + "loss": 3.552, + "step": 99525 + }, + { + "epoch": 6.762467726593287, + "grad_norm": 0.8908163905143738, + "learning_rate": 0.00015469153417583913, + "loss": 3.5079, + "step": 99530 + }, + { + "epoch": 6.762807446663949, + "grad_norm": 0.9615597128868103, + "learning_rate": 0.00015464906916700638, + "loss": 3.7726, + "step": 99535 + }, + { + "epoch": 6.763147166734611, + "grad_norm": 0.8835120797157288, + "learning_rate": 0.0001546066041581737, + "loss": 3.5855, + "step": 99540 + }, + { + "epoch": 6.763486886805272, + "grad_norm": 0.7877165675163269, + "learning_rate": 0.00015456413914934094, + "loss": 3.4978, + "step": 99545 + }, + { + "epoch": 6.763826606875934, + "grad_norm": 0.9773317575454712, + "learning_rate": 0.00015452167414050822, + "loss": 3.291, + "step": 99550 + }, + { + "epoch": 6.764166326946596, + "grad_norm": 0.8295761346817017, + "learning_rate": 0.0001544792091316755, + "loss": 3.4661, + "step": 99555 + }, + { + "epoch": 6.7645060470172576, + "grad_norm": 0.9688109755516052, + "learning_rate": 0.00015443674412284278, + "loss": 3.0631, + "step": 99560 + }, + { + "epoch": 6.76484576708792, + "grad_norm": 0.8193382024765015, + "learning_rate": 0.00015439427911401006, + "loss": 3.5563, + "step": 99565 + }, + { + "epoch": 6.765185487158582, + "grad_norm": 1.0294445753097534, + "learning_rate": 0.00015435181410517734, + "loss": 3.3035, + "step": 99570 + }, + { + "epoch": 6.765525207229243, + "grad_norm": 0.7206056118011475, + "learning_rate": 0.00015430934909634462, + "loss": 3.4467, + "step": 99575 + }, + { + "epoch": 6.765864927299905, + "grad_norm": 0.9589874148368835, + "learning_rate": 0.00015426688408751187, + "loss": 3.327, + "step": 99580 + }, + { + "epoch": 6.766204647370567, + "grad_norm": 1.0590951442718506, + "learning_rate": 0.00015422441907867918, + "loss": 3.1902, + "step": 99585 + }, + { + "epoch": 6.766544367441228, + "grad_norm": 1.0660455226898193, + "learning_rate": 0.00015418195406984643, + "loss": 3.2957, + "step": 99590 + }, + { + "epoch": 6.76688408751189, + "grad_norm": 0.9801725745201111, + "learning_rate": 0.00015413948906101371, + "loss": 3.5038, + "step": 99595 + }, + { + "epoch": 6.7672238075825515, + "grad_norm": 0.9071020483970642, + "learning_rate": 0.00015409702405218102, + "loss": 3.5563, + "step": 99600 + }, + { + "epoch": 6.767563527653214, + "grad_norm": 0.7007171511650085, + "learning_rate": 0.00015405455904334827, + "loss": 3.2975, + "step": 99605 + }, + { + "epoch": 6.767903247723876, + "grad_norm": 0.960843026638031, + "learning_rate": 0.00015401209403451555, + "loss": 3.3242, + "step": 99610 + }, + { + "epoch": 6.768242967794537, + "grad_norm": 1.2630194425582886, + "learning_rate": 0.00015396962902568283, + "loss": 3.4377, + "step": 99615 + }, + { + "epoch": 6.768582687865199, + "grad_norm": 1.115818738937378, + "learning_rate": 0.00015392716401685012, + "loss": 3.3232, + "step": 99620 + }, + { + "epoch": 6.768922407935861, + "grad_norm": 0.7125521898269653, + "learning_rate": 0.00015388469900801742, + "loss": 3.2265, + "step": 99625 + }, + { + "epoch": 6.769262128006522, + "grad_norm": 1.066498041152954, + "learning_rate": 0.00015384223399918468, + "loss": 3.462, + "step": 99630 + }, + { + "epoch": 6.769601848077184, + "grad_norm": 0.7662089467048645, + "learning_rate": 0.00015379976899035196, + "loss": 3.2683, + "step": 99635 + }, + { + "epoch": 6.769941568147846, + "grad_norm": 0.8437575697898865, + "learning_rate": 0.00015375730398151924, + "loss": 3.5209, + "step": 99640 + }, + { + "epoch": 6.7702812882185075, + "grad_norm": 0.9816077947616577, + "learning_rate": 0.00015371483897268652, + "loss": 3.3507, + "step": 99645 + }, + { + "epoch": 6.77062100828917, + "grad_norm": 0.7291998863220215, + "learning_rate": 0.00015367237396385377, + "loss": 3.5195, + "step": 99650 + }, + { + "epoch": 6.770960728359832, + "grad_norm": 0.926530122756958, + "learning_rate": 0.00015362990895502108, + "loss": 3.5913, + "step": 99655 + }, + { + "epoch": 6.771300448430493, + "grad_norm": 0.9219086170196533, + "learning_rate": 0.00015358744394618836, + "loss": 3.4519, + "step": 99660 + }, + { + "epoch": 6.771640168501155, + "grad_norm": 0.7228702306747437, + "learning_rate": 0.0001535449789373556, + "loss": 3.1878, + "step": 99665 + }, + { + "epoch": 6.771979888571817, + "grad_norm": 0.8624316453933716, + "learning_rate": 0.00015350251392852292, + "loss": 3.43, + "step": 99670 + }, + { + "epoch": 6.772319608642478, + "grad_norm": 1.0738073587417603, + "learning_rate": 0.00015346004891969017, + "loss": 3.5331, + "step": 99675 + }, + { + "epoch": 6.77265932871314, + "grad_norm": 0.9521868228912354, + "learning_rate": 0.00015341758391085745, + "loss": 3.4311, + "step": 99680 + }, + { + "epoch": 6.772999048783802, + "grad_norm": 0.7913209795951843, + "learning_rate": 0.00015337511890202473, + "loss": 3.4324, + "step": 99685 + }, + { + "epoch": 6.7733387688544635, + "grad_norm": 0.8949770331382751, + "learning_rate": 0.000153332653893192, + "loss": 3.5552, + "step": 99690 + }, + { + "epoch": 6.773678488925126, + "grad_norm": 0.9246948957443237, + "learning_rate": 0.0001532901888843593, + "loss": 3.4747, + "step": 99695 + }, + { + "epoch": 6.774018208995788, + "grad_norm": 0.6874966621398926, + "learning_rate": 0.00015324772387552657, + "loss": 3.4551, + "step": 99700 + }, + { + "epoch": 6.774357929066449, + "grad_norm": 0.8899385333061218, + "learning_rate": 0.00015320525886669385, + "loss": 3.5549, + "step": 99705 + }, + { + "epoch": 6.774697649137111, + "grad_norm": 1.0086959600448608, + "learning_rate": 0.00015316279385786113, + "loss": 3.2559, + "step": 99710 + }, + { + "epoch": 6.775037369207773, + "grad_norm": 0.8510286808013916, + "learning_rate": 0.0001531203288490284, + "loss": 3.2841, + "step": 99715 + }, + { + "epoch": 6.775377089278434, + "grad_norm": 0.7927818894386292, + "learning_rate": 0.00015307786384019566, + "loss": 3.2082, + "step": 99720 + }, + { + "epoch": 6.775716809349096, + "grad_norm": 0.8934504985809326, + "learning_rate": 0.00015303539883136297, + "loss": 3.4826, + "step": 99725 + }, + { + "epoch": 6.776056529419758, + "grad_norm": 1.10349702835083, + "learning_rate": 0.00015299293382253025, + "loss": 3.5363, + "step": 99730 + }, + { + "epoch": 6.7763962494904195, + "grad_norm": 0.9643020629882812, + "learning_rate": 0.0001529504688136975, + "loss": 3.463, + "step": 99735 + }, + { + "epoch": 6.776735969561082, + "grad_norm": 1.0113599300384521, + "learning_rate": 0.0001529080038048648, + "loss": 3.0655, + "step": 99740 + }, + { + "epoch": 6.777075689631744, + "grad_norm": 0.9481931924819946, + "learning_rate": 0.00015286553879603206, + "loss": 3.045, + "step": 99745 + }, + { + "epoch": 6.777415409702405, + "grad_norm": 0.9736223220825195, + "learning_rate": 0.00015282307378719934, + "loss": 3.086, + "step": 99750 + }, + { + "epoch": 6.777755129773067, + "grad_norm": 0.6941520571708679, + "learning_rate": 0.00015278060877836665, + "loss": 3.4232, + "step": 99755 + }, + { + "epoch": 6.778094849843729, + "grad_norm": 0.9495178461074829, + "learning_rate": 0.0001527381437695339, + "loss": 3.3548, + "step": 99760 + }, + { + "epoch": 6.77843456991439, + "grad_norm": 0.7712280750274658, + "learning_rate": 0.00015269567876070118, + "loss": 3.3122, + "step": 99765 + }, + { + "epoch": 6.778774289985052, + "grad_norm": 1.1348013877868652, + "learning_rate": 0.00015265321375186846, + "loss": 3.2676, + "step": 99770 + }, + { + "epoch": 6.779114010055714, + "grad_norm": 1.1072458028793335, + "learning_rate": 0.00015261074874303574, + "loss": 3.2638, + "step": 99775 + }, + { + "epoch": 6.7794537301263755, + "grad_norm": 0.9069634079933167, + "learning_rate": 0.000152568283734203, + "loss": 3.3378, + "step": 99780 + }, + { + "epoch": 6.779793450197038, + "grad_norm": 1.0267560482025146, + "learning_rate": 0.0001525258187253703, + "loss": 3.2608, + "step": 99785 + }, + { + "epoch": 6.7801331702677, + "grad_norm": 0.9481503367424011, + "learning_rate": 0.00015248335371653758, + "loss": 3.36, + "step": 99790 + }, + { + "epoch": 6.780472890338361, + "grad_norm": 1.0087605714797974, + "learning_rate": 0.00015244088870770486, + "loss": 3.7108, + "step": 99795 + }, + { + "epoch": 6.780812610409023, + "grad_norm": 1.0535352230072021, + "learning_rate": 0.00015239842369887214, + "loss": 3.4181, + "step": 99800 + }, + { + "epoch": 6.781152330479685, + "grad_norm": 1.218019962310791, + "learning_rate": 0.0001523559586900394, + "loss": 3.3924, + "step": 99805 + }, + { + "epoch": 6.781492050550346, + "grad_norm": 0.8821643590927124, + "learning_rate": 0.0001523134936812067, + "loss": 3.2516, + "step": 99810 + }, + { + "epoch": 6.781831770621008, + "grad_norm": 0.9434052109718323, + "learning_rate": 0.00015227102867237396, + "loss": 3.2397, + "step": 99815 + }, + { + "epoch": 6.78217149069167, + "grad_norm": 0.9311332702636719, + "learning_rate": 0.00015222856366354124, + "loss": 3.5625, + "step": 99820 + }, + { + "epoch": 6.7825112107623315, + "grad_norm": 0.9318792223930359, + "learning_rate": 0.00015218609865470854, + "loss": 3.4343, + "step": 99825 + }, + { + "epoch": 6.782850930832994, + "grad_norm": 0.99784255027771, + "learning_rate": 0.0001521436336458758, + "loss": 3.4202, + "step": 99830 + }, + { + "epoch": 6.783190650903656, + "grad_norm": 0.8228357434272766, + "learning_rate": 0.00015210116863704308, + "loss": 3.3909, + "step": 99835 + }, + { + "epoch": 6.783530370974317, + "grad_norm": 0.9715644717216492, + "learning_rate": 0.00015205870362821036, + "loss": 3.4198, + "step": 99840 + }, + { + "epoch": 6.783870091044979, + "grad_norm": 0.8504847884178162, + "learning_rate": 0.00015201623861937764, + "loss": 3.4327, + "step": 99845 + }, + { + "epoch": 6.784209811115641, + "grad_norm": 0.901361346244812, + "learning_rate": 0.0001519737736105449, + "loss": 3.5248, + "step": 99850 + }, + { + "epoch": 6.784549531186302, + "grad_norm": 0.656367301940918, + "learning_rate": 0.0001519313086017122, + "loss": 3.4148, + "step": 99855 + }, + { + "epoch": 6.784889251256964, + "grad_norm": 0.8693908452987671, + "learning_rate": 0.00015188884359287948, + "loss": 3.3153, + "step": 99860 + }, + { + "epoch": 6.785228971327626, + "grad_norm": 0.8020505309104919, + "learning_rate": 0.00015184637858404673, + "loss": 3.5313, + "step": 99865 + }, + { + "epoch": 6.7855686913982876, + "grad_norm": 1.2236961126327515, + "learning_rate": 0.00015180391357521404, + "loss": 3.4842, + "step": 99870 + }, + { + "epoch": 6.78590841146895, + "grad_norm": 0.9227713346481323, + "learning_rate": 0.0001517614485663813, + "loss": 3.6687, + "step": 99875 + }, + { + "epoch": 6.786248131539612, + "grad_norm": 0.9256808757781982, + "learning_rate": 0.0001517189835575486, + "loss": 3.4526, + "step": 99880 + }, + { + "epoch": 6.786587851610273, + "grad_norm": 0.9269837737083435, + "learning_rate": 0.00015167651854871585, + "loss": 3.3246, + "step": 99885 + }, + { + "epoch": 6.786927571680935, + "grad_norm": 0.902163028717041, + "learning_rate": 0.00015163405353988313, + "loss": 3.4058, + "step": 99890 + }, + { + "epoch": 6.787267291751597, + "grad_norm": 0.8992892503738403, + "learning_rate": 0.00015159158853105044, + "loss": 3.764, + "step": 99895 + }, + { + "epoch": 6.787607011822258, + "grad_norm": 0.8953142762184143, + "learning_rate": 0.0001515491235222177, + "loss": 3.4786, + "step": 99900 + }, + { + "epoch": 6.78794673189292, + "grad_norm": 0.7373257279396057, + "learning_rate": 0.00015150665851338497, + "loss": 3.2808, + "step": 99905 + }, + { + "epoch": 6.788286451963582, + "grad_norm": 0.8006570339202881, + "learning_rate": 0.00015146419350455225, + "loss": 3.6323, + "step": 99910 + }, + { + "epoch": 6.788626172034244, + "grad_norm": 0.9359864592552185, + "learning_rate": 0.00015142172849571953, + "loss": 3.2072, + "step": 99915 + }, + { + "epoch": 6.788965892104906, + "grad_norm": 0.8773544430732727, + "learning_rate": 0.0001513792634868868, + "loss": 3.474, + "step": 99920 + }, + { + "epoch": 6.789305612175568, + "grad_norm": 1.06246018409729, + "learning_rate": 0.0001513367984780541, + "loss": 3.1991, + "step": 99925 + }, + { + "epoch": 6.789645332246229, + "grad_norm": 0.796363115310669, + "learning_rate": 0.00015129433346922137, + "loss": 3.43, + "step": 99930 + }, + { + "epoch": 6.789985052316891, + "grad_norm": 0.8031945824623108, + "learning_rate": 0.00015125186846038862, + "loss": 3.41, + "step": 99935 + }, + { + "epoch": 6.790324772387553, + "grad_norm": 1.0307410955429077, + "learning_rate": 0.00015120940345155593, + "loss": 3.5347, + "step": 99940 + }, + { + "epoch": 6.790664492458214, + "grad_norm": 1.311097502708435, + "learning_rate": 0.00015116693844272318, + "loss": 2.7546, + "step": 99945 + }, + { + "epoch": 6.791004212528876, + "grad_norm": 1.0082018375396729, + "learning_rate": 0.00015112447343389046, + "loss": 3.2832, + "step": 99950 + }, + { + "epoch": 6.791343932599538, + "grad_norm": 0.9035312533378601, + "learning_rate": 0.00015108200842505777, + "loss": 3.5014, + "step": 99955 + }, + { + "epoch": 6.7916836526702, + "grad_norm": 0.9561468958854675, + "learning_rate": 0.00015103954341622502, + "loss": 3.4317, + "step": 99960 + }, + { + "epoch": 6.792023372740862, + "grad_norm": 0.9281104207038879, + "learning_rate": 0.00015099707840739233, + "loss": 3.1988, + "step": 99965 + }, + { + "epoch": 6.792363092811524, + "grad_norm": 0.9586379528045654, + "learning_rate": 0.00015095461339855958, + "loss": 3.4013, + "step": 99970 + }, + { + "epoch": 6.792702812882185, + "grad_norm": 1.0212427377700806, + "learning_rate": 0.00015091214838972687, + "loss": 3.3891, + "step": 99975 + }, + { + "epoch": 6.793042532952847, + "grad_norm": 0.992225170135498, + "learning_rate": 0.00015086968338089415, + "loss": 3.5361, + "step": 99980 + }, + { + "epoch": 6.793382253023509, + "grad_norm": 0.8166058659553528, + "learning_rate": 0.00015082721837206143, + "loss": 3.4137, + "step": 99985 + }, + { + "epoch": 6.79372197309417, + "grad_norm": 1.0598595142364502, + "learning_rate": 0.0001507847533632287, + "loss": 3.2166, + "step": 99990 + }, + { + "epoch": 6.794061693164832, + "grad_norm": 0.9296268224716187, + "learning_rate": 0.00015074228835439599, + "loss": 3.0655, + "step": 99995 + }, + { + "epoch": 6.794401413235494, + "grad_norm": 0.913037121295929, + "learning_rate": 0.00015069982334556327, + "loss": 3.593, + "step": 100000 + }, + { + "epoch": 6.794741133306156, + "grad_norm": 0.8575137257575989, + "learning_rate": 0.00015065735833673052, + "loss": 3.4516, + "step": 100005 + }, + { + "epoch": 6.795080853376818, + "grad_norm": 0.8964600563049316, + "learning_rate": 0.00015061489332789783, + "loss": 3.2929, + "step": 100010 + }, + { + "epoch": 6.79542057344748, + "grad_norm": 0.9258711934089661, + "learning_rate": 0.00015057242831906508, + "loss": 3.5581, + "step": 100015 + }, + { + "epoch": 6.795760293518141, + "grad_norm": 0.9007853865623474, + "learning_rate": 0.00015052996331023236, + "loss": 3.3735, + "step": 100020 + }, + { + "epoch": 6.796100013588803, + "grad_norm": 0.8424878716468811, + "learning_rate": 0.00015048749830139967, + "loss": 3.4538, + "step": 100025 + }, + { + "epoch": 6.796439733659465, + "grad_norm": 0.8854019641876221, + "learning_rate": 0.00015044503329256692, + "loss": 3.3118, + "step": 100030 + }, + { + "epoch": 6.796779453730126, + "grad_norm": 1.1265170574188232, + "learning_rate": 0.0001504025682837342, + "loss": 3.4999, + "step": 100035 + }, + { + "epoch": 6.797119173800788, + "grad_norm": 1.0287039279937744, + "learning_rate": 0.00015036010327490148, + "loss": 3.0802, + "step": 100040 + }, + { + "epoch": 6.79745889387145, + "grad_norm": 0.7567350268363953, + "learning_rate": 0.00015031763826606876, + "loss": 3.5051, + "step": 100045 + }, + { + "epoch": 6.797798613942112, + "grad_norm": 0.8356943726539612, + "learning_rate": 0.00015027517325723607, + "loss": 3.4666, + "step": 100050 + }, + { + "epoch": 6.798138334012774, + "grad_norm": 0.851801872253418, + "learning_rate": 0.00015023270824840332, + "loss": 3.4712, + "step": 100055 + }, + { + "epoch": 6.798478054083435, + "grad_norm": 1.043276309967041, + "learning_rate": 0.0001501902432395706, + "loss": 3.5415, + "step": 100060 + }, + { + "epoch": 6.798817774154097, + "grad_norm": 0.796796977519989, + "learning_rate": 0.00015014777823073788, + "loss": 3.7055, + "step": 100065 + }, + { + "epoch": 6.799157494224759, + "grad_norm": 0.9193426966667175, + "learning_rate": 0.00015010531322190516, + "loss": 3.5252, + "step": 100070 + }, + { + "epoch": 6.79949721429542, + "grad_norm": 1.0381145477294922, + "learning_rate": 0.0001500628482130724, + "loss": 3.29, + "step": 100075 + }, + { + "epoch": 6.799836934366082, + "grad_norm": 0.9987073540687561, + "learning_rate": 0.00015002038320423972, + "loss": 3.3965, + "step": 100080 + }, + { + "epoch": 6.800176654436744, + "grad_norm": 0.9647864699363708, + "learning_rate": 0.000149977918195407, + "loss": 3.355, + "step": 100085 + }, + { + "epoch": 6.8005163745074055, + "grad_norm": 0.8205980062484741, + "learning_rate": 0.00014993545318657425, + "loss": 3.3669, + "step": 100090 + }, + { + "epoch": 6.800856094578068, + "grad_norm": 0.7873755693435669, + "learning_rate": 0.00014989298817774156, + "loss": 3.5047, + "step": 100095 + }, + { + "epoch": 6.80119581464873, + "grad_norm": 0.8553416132926941, + "learning_rate": 0.0001498505231689088, + "loss": 3.3453, + "step": 100100 + }, + { + "epoch": 6.801535534719391, + "grad_norm": 1.2499926090240479, + "learning_rate": 0.0001498080581600761, + "loss": 3.5693, + "step": 100105 + }, + { + "epoch": 6.801875254790053, + "grad_norm": 1.1078156232833862, + "learning_rate": 0.00014976559315124337, + "loss": 3.2617, + "step": 100110 + }, + { + "epoch": 6.802214974860715, + "grad_norm": 0.9178809523582458, + "learning_rate": 0.00014972312814241065, + "loss": 3.516, + "step": 100115 + }, + { + "epoch": 6.802554694931376, + "grad_norm": 0.9593670964241028, + "learning_rate": 0.00014968066313357793, + "loss": 3.2317, + "step": 100120 + }, + { + "epoch": 6.802894415002038, + "grad_norm": 0.9572060704231262, + "learning_rate": 0.0001496381981247452, + "loss": 3.7028, + "step": 100125 + }, + { + "epoch": 6.8032341350727, + "grad_norm": 0.8555945158004761, + "learning_rate": 0.0001495957331159125, + "loss": 3.3364, + "step": 100130 + }, + { + "epoch": 6.8035738551433615, + "grad_norm": 1.0601768493652344, + "learning_rate": 0.00014955326810707977, + "loss": 3.3416, + "step": 100135 + }, + { + "epoch": 6.803913575214024, + "grad_norm": 0.8893862962722778, + "learning_rate": 0.00014951080309824705, + "loss": 3.5226, + "step": 100140 + }, + { + "epoch": 6.804253295284686, + "grad_norm": 1.0829765796661377, + "learning_rate": 0.0001494683380894143, + "loss": 3.298, + "step": 100145 + }, + { + "epoch": 6.804593015355347, + "grad_norm": 1.0734409093856812, + "learning_rate": 0.0001494258730805816, + "loss": 3.4083, + "step": 100150 + }, + { + "epoch": 6.804932735426009, + "grad_norm": 0.8647913336753845, + "learning_rate": 0.0001493834080717489, + "loss": 3.662, + "step": 100155 + }, + { + "epoch": 6.805272455496671, + "grad_norm": 0.868672251701355, + "learning_rate": 0.00014934094306291615, + "loss": 3.2605, + "step": 100160 + }, + { + "epoch": 6.805612175567332, + "grad_norm": 1.0270127058029175, + "learning_rate": 0.00014929847805408345, + "loss": 3.5679, + "step": 100165 + }, + { + "epoch": 6.805951895637994, + "grad_norm": 0.964471161365509, + "learning_rate": 0.0001492560130452507, + "loss": 3.0583, + "step": 100170 + }, + { + "epoch": 6.806291615708656, + "grad_norm": 0.8278515934944153, + "learning_rate": 0.000149213548036418, + "loss": 3.4016, + "step": 100175 + }, + { + "epoch": 6.806631335779318, + "grad_norm": 1.074007511138916, + "learning_rate": 0.00014917108302758527, + "loss": 3.5742, + "step": 100180 + }, + { + "epoch": 6.80697105584998, + "grad_norm": 0.9369561672210693, + "learning_rate": 0.00014912861801875255, + "loss": 3.6651, + "step": 100185 + }, + { + "epoch": 6.807310775920642, + "grad_norm": 0.9747997522354126, + "learning_rate": 0.00014908615300991983, + "loss": 3.3609, + "step": 100190 + }, + { + "epoch": 6.807650495991303, + "grad_norm": 1.1763815879821777, + "learning_rate": 0.0001490436880010871, + "loss": 3.6564, + "step": 100195 + }, + { + "epoch": 6.807990216061965, + "grad_norm": 1.1907552480697632, + "learning_rate": 0.0001490012229922544, + "loss": 3.3302, + "step": 100200 + }, + { + "epoch": 6.808329936132627, + "grad_norm": 0.9265353679656982, + "learning_rate": 0.00014895875798342164, + "loss": 3.3753, + "step": 100205 + }, + { + "epoch": 6.808669656203288, + "grad_norm": 0.8883565068244934, + "learning_rate": 0.00014891629297458895, + "loss": 3.2999, + "step": 100210 + }, + { + "epoch": 6.80900937627395, + "grad_norm": 0.8970712423324585, + "learning_rate": 0.00014887382796575623, + "loss": 3.3962, + "step": 100215 + }, + { + "epoch": 6.809349096344612, + "grad_norm": 1.0163397789001465, + "learning_rate": 0.0001488313629569235, + "loss": 3.3997, + "step": 100220 + }, + { + "epoch": 6.809688816415274, + "grad_norm": 0.8363779187202454, + "learning_rate": 0.0001487888979480908, + "loss": 3.2527, + "step": 100225 + }, + { + "epoch": 6.810028536485936, + "grad_norm": 1.0404610633850098, + "learning_rate": 0.00014874643293925804, + "loss": 3.3091, + "step": 100230 + }, + { + "epoch": 6.810368256556598, + "grad_norm": 1.456931710243225, + "learning_rate": 0.00014870396793042535, + "loss": 3.4955, + "step": 100235 + }, + { + "epoch": 6.810707976627259, + "grad_norm": 1.0122010707855225, + "learning_rate": 0.0001486615029215926, + "loss": 3.4173, + "step": 100240 + }, + { + "epoch": 6.811047696697921, + "grad_norm": 0.9294032454490662, + "learning_rate": 0.00014861903791275988, + "loss": 3.69, + "step": 100245 + }, + { + "epoch": 6.811387416768583, + "grad_norm": 1.2338829040527344, + "learning_rate": 0.0001485765729039272, + "loss": 3.0954, + "step": 100250 + }, + { + "epoch": 6.811727136839244, + "grad_norm": 0.9054112434387207, + "learning_rate": 0.00014853410789509444, + "loss": 3.3347, + "step": 100255 + }, + { + "epoch": 6.812066856909906, + "grad_norm": 0.9830452799797058, + "learning_rate": 0.00014849164288626172, + "loss": 3.1962, + "step": 100260 + }, + { + "epoch": 6.812406576980568, + "grad_norm": 0.903194010257721, + "learning_rate": 0.000148449177877429, + "loss": 3.5, + "step": 100265 + }, + { + "epoch": 6.81274629705123, + "grad_norm": 0.8654564023017883, + "learning_rate": 0.00014840671286859628, + "loss": 3.3625, + "step": 100270 + }, + { + "epoch": 6.813086017121892, + "grad_norm": 1.2019495964050293, + "learning_rate": 0.00014836424785976353, + "loss": 3.4559, + "step": 100275 + }, + { + "epoch": 6.813425737192553, + "grad_norm": 1.4092931747436523, + "learning_rate": 0.00014832178285093084, + "loss": 3.4223, + "step": 100280 + }, + { + "epoch": 6.813765457263215, + "grad_norm": 1.2316879034042358, + "learning_rate": 0.00014827931784209812, + "loss": 3.1193, + "step": 100285 + }, + { + "epoch": 6.814105177333877, + "grad_norm": 0.916812002658844, + "learning_rate": 0.00014823685283326537, + "loss": 3.2433, + "step": 100290 + }, + { + "epoch": 6.814444897404538, + "grad_norm": 0.9034768342971802, + "learning_rate": 0.00014819438782443268, + "loss": 3.3563, + "step": 100295 + }, + { + "epoch": 6.8147846174752, + "grad_norm": 0.8853583931922913, + "learning_rate": 0.00014815192281559993, + "loss": 3.6059, + "step": 100300 + }, + { + "epoch": 6.815124337545862, + "grad_norm": 1.1311098337173462, + "learning_rate": 0.00014810945780676724, + "loss": 2.9079, + "step": 100305 + }, + { + "epoch": 6.8154640576165235, + "grad_norm": 1.983076810836792, + "learning_rate": 0.0001480669927979345, + "loss": 3.3532, + "step": 100310 + }, + { + "epoch": 6.815803777687186, + "grad_norm": 0.7817500233650208, + "learning_rate": 0.00014802452778910177, + "loss": 3.3615, + "step": 100315 + }, + { + "epoch": 6.816143497757848, + "grad_norm": 0.8985015749931335, + "learning_rate": 0.00014798206278026908, + "loss": 3.2826, + "step": 100320 + }, + { + "epoch": 6.816483217828509, + "grad_norm": 0.9918767213821411, + "learning_rate": 0.00014793959777143633, + "loss": 3.3861, + "step": 100325 + }, + { + "epoch": 6.816822937899171, + "grad_norm": 0.8523836135864258, + "learning_rate": 0.00014789713276260361, + "loss": 3.2539, + "step": 100330 + }, + { + "epoch": 6.817162657969833, + "grad_norm": 0.9988969564437866, + "learning_rate": 0.0001478546677537709, + "loss": 3.5585, + "step": 100335 + }, + { + "epoch": 6.817502378040494, + "grad_norm": 1.0139954090118408, + "learning_rate": 0.00014781220274493818, + "loss": 3.4551, + "step": 100340 + }, + { + "epoch": 6.817842098111156, + "grad_norm": 1.3267606496810913, + "learning_rate": 0.00014776973773610546, + "loss": 3.3382, + "step": 100345 + }, + { + "epoch": 6.818181818181818, + "grad_norm": 0.7996826171875, + "learning_rate": 0.00014772727272727274, + "loss": 3.5286, + "step": 100350 + }, + { + "epoch": 6.8185215382524795, + "grad_norm": 0.9354169368743896, + "learning_rate": 0.00014768480771844002, + "loss": 3.3639, + "step": 100355 + }, + { + "epoch": 6.818861258323142, + "grad_norm": 0.8256972432136536, + "learning_rate": 0.00014764234270960727, + "loss": 3.5612, + "step": 100360 + }, + { + "epoch": 6.819200978393804, + "grad_norm": 0.8026857376098633, + "learning_rate": 0.00014759987770077458, + "loss": 3.5555, + "step": 100365 + }, + { + "epoch": 6.819540698464465, + "grad_norm": 1.170589804649353, + "learning_rate": 0.00014755741269194183, + "loss": 3.4873, + "step": 100370 + }, + { + "epoch": 6.819880418535127, + "grad_norm": 1.1481497287750244, + "learning_rate": 0.0001475149476831091, + "loss": 3.2395, + "step": 100375 + }, + { + "epoch": 6.820220138605789, + "grad_norm": 0.9609932899475098, + "learning_rate": 0.00014747248267427642, + "loss": 3.602, + "step": 100380 + }, + { + "epoch": 6.82055985867645, + "grad_norm": 0.8649572134017944, + "learning_rate": 0.00014743001766544367, + "loss": 3.4154, + "step": 100385 + }, + { + "epoch": 6.820899578747112, + "grad_norm": 0.8478183150291443, + "learning_rate": 0.00014738755265661098, + "loss": 3.4007, + "step": 100390 + }, + { + "epoch": 6.821239298817774, + "grad_norm": 0.821980357170105, + "learning_rate": 0.00014734508764777823, + "loss": 3.5888, + "step": 100395 + }, + { + "epoch": 6.8215790188884355, + "grad_norm": 0.7931752800941467, + "learning_rate": 0.0001473026226389455, + "loss": 3.434, + "step": 100400 + }, + { + "epoch": 6.821918738959098, + "grad_norm": 0.9506734013557434, + "learning_rate": 0.0001472601576301128, + "loss": 3.4195, + "step": 100405 + }, + { + "epoch": 6.82225845902976, + "grad_norm": 0.9139347672462463, + "learning_rate": 0.00014721769262128007, + "loss": 3.5722, + "step": 100410 + }, + { + "epoch": 6.822598179100421, + "grad_norm": 0.756956934928894, + "learning_rate": 0.00014717522761244735, + "loss": 3.5824, + "step": 100415 + }, + { + "epoch": 6.822937899171083, + "grad_norm": 0.8547456860542297, + "learning_rate": 0.00014713276260361463, + "loss": 3.4591, + "step": 100420 + }, + { + "epoch": 6.823277619241745, + "grad_norm": 0.8208205103874207, + "learning_rate": 0.0001470902975947819, + "loss": 3.3559, + "step": 100425 + }, + { + "epoch": 6.823617339312406, + "grad_norm": 0.9858450889587402, + "learning_rate": 0.00014704783258594916, + "loss": 3.4136, + "step": 100430 + }, + { + "epoch": 6.823957059383068, + "grad_norm": 0.8814415335655212, + "learning_rate": 0.00014700536757711647, + "loss": 3.3346, + "step": 100435 + }, + { + "epoch": 6.82429677945373, + "grad_norm": 0.9527665376663208, + "learning_rate": 0.00014696290256828372, + "loss": 3.2548, + "step": 100440 + }, + { + "epoch": 6.8246364995243916, + "grad_norm": 0.8286465406417847, + "learning_rate": 0.000146920437559451, + "loss": 3.5349, + "step": 100445 + }, + { + "epoch": 6.824976219595054, + "grad_norm": 0.9038989543914795, + "learning_rate": 0.0001468779725506183, + "loss": 3.3297, + "step": 100450 + }, + { + "epoch": 6.825315939665716, + "grad_norm": 0.9671839475631714, + "learning_rate": 0.00014683550754178556, + "loss": 3.4728, + "step": 100455 + }, + { + "epoch": 6.825655659736377, + "grad_norm": 0.7053710222244263, + "learning_rate": 0.00014679304253295284, + "loss": 3.5216, + "step": 100460 + }, + { + "epoch": 6.825995379807039, + "grad_norm": 0.8481890559196472, + "learning_rate": 0.00014675057752412012, + "loss": 3.5968, + "step": 100465 + }, + { + "epoch": 6.826335099877701, + "grad_norm": 1.16195809841156, + "learning_rate": 0.0001467081125152874, + "loss": 3.4479, + "step": 100470 + }, + { + "epoch": 6.826674819948362, + "grad_norm": 0.8493369817733765, + "learning_rate": 0.0001466656475064547, + "loss": 3.5959, + "step": 100475 + }, + { + "epoch": 6.827014540019024, + "grad_norm": 0.9425905346870422, + "learning_rate": 0.00014662318249762196, + "loss": 3.377, + "step": 100480 + }, + { + "epoch": 6.827354260089686, + "grad_norm": 0.8706255555152893, + "learning_rate": 0.00014658071748878924, + "loss": 3.2139, + "step": 100485 + }, + { + "epoch": 6.827693980160348, + "grad_norm": 0.9197949767112732, + "learning_rate": 0.00014653825247995652, + "loss": 3.2202, + "step": 100490 + }, + { + "epoch": 6.82803370023101, + "grad_norm": 0.7884461879730225, + "learning_rate": 0.0001464957874711238, + "loss": 3.4946, + "step": 100495 + }, + { + "epoch": 6.828373420301672, + "grad_norm": 1.0084640979766846, + "learning_rate": 0.00014645332246229106, + "loss": 3.4014, + "step": 100500 + }, + { + "epoch": 6.828713140372333, + "grad_norm": 0.9196860194206238, + "learning_rate": 0.00014641085745345836, + "loss": 3.6542, + "step": 100505 + }, + { + "epoch": 6.829052860442995, + "grad_norm": 1.5915594100952148, + "learning_rate": 0.00014636839244462564, + "loss": 3.5378, + "step": 100510 + }, + { + "epoch": 6.829392580513657, + "grad_norm": 0.9608041644096375, + "learning_rate": 0.0001463259274357929, + "loss": 3.4732, + "step": 100515 + }, + { + "epoch": 6.829732300584318, + "grad_norm": 0.9159142971038818, + "learning_rate": 0.0001462834624269602, + "loss": 3.4008, + "step": 100520 + }, + { + "epoch": 6.83007202065498, + "grad_norm": 1.448606252670288, + "learning_rate": 0.00014624099741812746, + "loss": 3.4153, + "step": 100525 + }, + { + "epoch": 6.830411740725642, + "grad_norm": 0.8458374738693237, + "learning_rate": 0.00014619853240929474, + "loss": 3.4307, + "step": 100530 + }, + { + "epoch": 6.830751460796304, + "grad_norm": 0.8207949995994568, + "learning_rate": 0.00014615606740046202, + "loss": 3.4535, + "step": 100535 + }, + { + "epoch": 6.831091180866966, + "grad_norm": 0.9571805000305176, + "learning_rate": 0.0001461136023916293, + "loss": 3.5686, + "step": 100540 + }, + { + "epoch": 6.831430900937628, + "grad_norm": 1.2244988679885864, + "learning_rate": 0.00014607113738279658, + "loss": 3.4905, + "step": 100545 + }, + { + "epoch": 6.831770621008289, + "grad_norm": 0.7527838349342346, + "learning_rate": 0.00014602867237396386, + "loss": 3.5161, + "step": 100550 + }, + { + "epoch": 6.832110341078951, + "grad_norm": 0.9515064358711243, + "learning_rate": 0.00014598620736513114, + "loss": 3.4081, + "step": 100555 + }, + { + "epoch": 6.832450061149613, + "grad_norm": 0.8450833559036255, + "learning_rate": 0.00014594374235629842, + "loss": 3.2909, + "step": 100560 + }, + { + "epoch": 6.832789781220274, + "grad_norm": 0.8762406706809998, + "learning_rate": 0.0001459012773474657, + "loss": 3.5373, + "step": 100565 + }, + { + "epoch": 6.833129501290936, + "grad_norm": 0.8472840785980225, + "learning_rate": 0.00014585881233863295, + "loss": 3.3648, + "step": 100570 + }, + { + "epoch": 6.833469221361598, + "grad_norm": 0.837125301361084, + "learning_rate": 0.00014581634732980026, + "loss": 3.5402, + "step": 100575 + }, + { + "epoch": 6.83380894143226, + "grad_norm": 0.8889573812484741, + "learning_rate": 0.00014577388232096754, + "loss": 3.1978, + "step": 100580 + }, + { + "epoch": 6.834148661502922, + "grad_norm": 1.079724907875061, + "learning_rate": 0.0001457314173121348, + "loss": 3.4221, + "step": 100585 + }, + { + "epoch": 6.834488381573584, + "grad_norm": 0.7831591963768005, + "learning_rate": 0.0001456889523033021, + "loss": 3.6248, + "step": 100590 + }, + { + "epoch": 6.834828101644245, + "grad_norm": 0.8341687321662903, + "learning_rate": 0.00014564648729446935, + "loss": 3.5533, + "step": 100595 + }, + { + "epoch": 6.835167821714907, + "grad_norm": 0.8147615790367126, + "learning_rate": 0.00014560402228563663, + "loss": 3.38, + "step": 100600 + }, + { + "epoch": 6.835507541785569, + "grad_norm": 0.8650417923927307, + "learning_rate": 0.0001455615572768039, + "loss": 3.4579, + "step": 100605 + }, + { + "epoch": 6.83584726185623, + "grad_norm": 1.7366191148757935, + "learning_rate": 0.0001455190922679712, + "loss": 3.2616, + "step": 100610 + }, + { + "epoch": 6.836186981926892, + "grad_norm": 0.7795780897140503, + "learning_rate": 0.00014547662725913847, + "loss": 3.2107, + "step": 100615 + }, + { + "epoch": 6.836526701997554, + "grad_norm": 0.8736903071403503, + "learning_rate": 0.00014543416225030575, + "loss": 3.3444, + "step": 100620 + }, + { + "epoch": 6.836866422068216, + "grad_norm": 1.029105305671692, + "learning_rate": 0.00014539169724147303, + "loss": 3.1124, + "step": 100625 + }, + { + "epoch": 6.837206142138878, + "grad_norm": 0.9472302198410034, + "learning_rate": 0.00014534923223264028, + "loss": 3.2981, + "step": 100630 + }, + { + "epoch": 6.83754586220954, + "grad_norm": 0.9694245457649231, + "learning_rate": 0.0001453067672238076, + "loss": 3.3822, + "step": 100635 + }, + { + "epoch": 6.837885582280201, + "grad_norm": 0.8735710978507996, + "learning_rate": 0.00014526430221497487, + "loss": 3.4312, + "step": 100640 + }, + { + "epoch": 6.838225302350863, + "grad_norm": 1.0382264852523804, + "learning_rate": 0.00014522183720614215, + "loss": 3.0589, + "step": 100645 + }, + { + "epoch": 6.838565022421525, + "grad_norm": 0.9604737758636475, + "learning_rate": 0.00014517937219730943, + "loss": 3.4721, + "step": 100650 + }, + { + "epoch": 6.838904742492186, + "grad_norm": 0.9793075919151306, + "learning_rate": 0.00014513690718847668, + "loss": 3.3367, + "step": 100655 + }, + { + "epoch": 6.839244462562848, + "grad_norm": 0.7942293286323547, + "learning_rate": 0.000145094442179644, + "loss": 3.5251, + "step": 100660 + }, + { + "epoch": 6.83958418263351, + "grad_norm": 0.7261548638343811, + "learning_rate": 0.00014505197717081124, + "loss": 3.5219, + "step": 100665 + }, + { + "epoch": 6.839923902704172, + "grad_norm": 0.9981170892715454, + "learning_rate": 0.00014500951216197852, + "loss": 3.5413, + "step": 100670 + }, + { + "epoch": 6.840263622774834, + "grad_norm": 1.0394617319107056, + "learning_rate": 0.00014496704715314583, + "loss": 3.419, + "step": 100675 + }, + { + "epoch": 6.840603342845496, + "grad_norm": 0.7666508555412292, + "learning_rate": 0.00014492458214431308, + "loss": 3.3223, + "step": 100680 + }, + { + "epoch": 6.840943062916157, + "grad_norm": 1.1432828903198242, + "learning_rate": 0.00014488211713548036, + "loss": 3.4861, + "step": 100685 + }, + { + "epoch": 6.841282782986819, + "grad_norm": 0.9394174814224243, + "learning_rate": 0.00014483965212664765, + "loss": 3.6543, + "step": 100690 + }, + { + "epoch": 6.841622503057481, + "grad_norm": 1.1720502376556396, + "learning_rate": 0.00014479718711781493, + "loss": 3.4839, + "step": 100695 + }, + { + "epoch": 6.841962223128142, + "grad_norm": 0.8566363453865051, + "learning_rate": 0.00014475472210898218, + "loss": 3.4081, + "step": 100700 + }, + { + "epoch": 6.842301943198804, + "grad_norm": 0.9327955842018127, + "learning_rate": 0.00014471225710014949, + "loss": 3.5393, + "step": 100705 + }, + { + "epoch": 6.842641663269466, + "grad_norm": 0.9734935164451599, + "learning_rate": 0.00014466979209131677, + "loss": 3.369, + "step": 100710 + }, + { + "epoch": 6.842981383340128, + "grad_norm": 0.8572025299072266, + "learning_rate": 0.00014462732708248402, + "loss": 3.4087, + "step": 100715 + }, + { + "epoch": 6.84332110341079, + "grad_norm": 0.8700377345085144, + "learning_rate": 0.00014458486207365133, + "loss": 3.4553, + "step": 100720 + }, + { + "epoch": 6.843660823481452, + "grad_norm": 1.0506318807601929, + "learning_rate": 0.00014454239706481858, + "loss": 3.2546, + "step": 100725 + }, + { + "epoch": 6.844000543552113, + "grad_norm": 0.9769281148910522, + "learning_rate": 0.00014449993205598589, + "loss": 3.38, + "step": 100730 + }, + { + "epoch": 6.844340263622775, + "grad_norm": 1.1038548946380615, + "learning_rate": 0.00014445746704715314, + "loss": 3.2724, + "step": 100735 + }, + { + "epoch": 6.844679983693436, + "grad_norm": 0.8968616127967834, + "learning_rate": 0.00014441500203832042, + "loss": 3.25, + "step": 100740 + }, + { + "epoch": 6.845019703764098, + "grad_norm": 0.9997893571853638, + "learning_rate": 0.00014437253702948773, + "loss": 3.1677, + "step": 100745 + }, + { + "epoch": 6.84535942383476, + "grad_norm": 0.8686417937278748, + "learning_rate": 0.00014433007202065498, + "loss": 3.633, + "step": 100750 + }, + { + "epoch": 6.845699143905422, + "grad_norm": 0.8375125527381897, + "learning_rate": 0.00014428760701182226, + "loss": 3.4737, + "step": 100755 + }, + { + "epoch": 6.846038863976084, + "grad_norm": 0.817937970161438, + "learning_rate": 0.00014424514200298954, + "loss": 3.4163, + "step": 100760 + }, + { + "epoch": 6.846378584046746, + "grad_norm": 0.8897196054458618, + "learning_rate": 0.00014420267699415682, + "loss": 3.6355, + "step": 100765 + }, + { + "epoch": 6.846718304117407, + "grad_norm": 0.7632497549057007, + "learning_rate": 0.0001441602119853241, + "loss": 3.4569, + "step": 100770 + }, + { + "epoch": 6.847058024188069, + "grad_norm": 1.1038464307785034, + "learning_rate": 0.00014411774697649138, + "loss": 3.313, + "step": 100775 + }, + { + "epoch": 6.847397744258731, + "grad_norm": 0.9382354021072388, + "learning_rate": 0.00014407528196765866, + "loss": 3.4733, + "step": 100780 + }, + { + "epoch": 6.847737464329392, + "grad_norm": 0.8278224468231201, + "learning_rate": 0.0001440328169588259, + "loss": 3.6826, + "step": 100785 + }, + { + "epoch": 6.848077184400054, + "grad_norm": 0.941413938999176, + "learning_rate": 0.00014399035194999322, + "loss": 3.2974, + "step": 100790 + }, + { + "epoch": 6.848416904470716, + "grad_norm": 0.7823814749717712, + "learning_rate": 0.00014394788694116047, + "loss": 3.3791, + "step": 100795 + }, + { + "epoch": 6.848756624541378, + "grad_norm": 0.9404653310775757, + "learning_rate": 0.00014390542193232775, + "loss": 3.1892, + "step": 100800 + }, + { + "epoch": 6.84909634461204, + "grad_norm": 1.0439223051071167, + "learning_rate": 0.00014386295692349506, + "loss": 3.3695, + "step": 100805 + }, + { + "epoch": 6.849436064682702, + "grad_norm": 1.1628636121749878, + "learning_rate": 0.0001438204919146623, + "loss": 3.4865, + "step": 100810 + }, + { + "epoch": 6.849775784753363, + "grad_norm": 0.8943682909011841, + "learning_rate": 0.00014377802690582962, + "loss": 3.6905, + "step": 100815 + }, + { + "epoch": 6.850115504824025, + "grad_norm": 0.991104006767273, + "learning_rate": 0.00014373556189699687, + "loss": 3.3585, + "step": 100820 + }, + { + "epoch": 6.850455224894687, + "grad_norm": 1.0212881565093994, + "learning_rate": 0.00014369309688816415, + "loss": 3.3668, + "step": 100825 + }, + { + "epoch": 6.850794944965348, + "grad_norm": 0.8201507329940796, + "learning_rate": 0.00014365063187933143, + "loss": 3.3519, + "step": 100830 + }, + { + "epoch": 6.85113466503601, + "grad_norm": 0.898565948009491, + "learning_rate": 0.0001436081668704987, + "loss": 3.4623, + "step": 100835 + }, + { + "epoch": 6.851474385106672, + "grad_norm": 0.8707543611526489, + "learning_rate": 0.000143565701861666, + "loss": 3.5451, + "step": 100840 + }, + { + "epoch": 6.851814105177334, + "grad_norm": 0.7810743451118469, + "learning_rate": 0.00014352323685283327, + "loss": 3.3419, + "step": 100845 + }, + { + "epoch": 6.852153825247996, + "grad_norm": 0.7980993390083313, + "learning_rate": 0.00014348077184400055, + "loss": 3.3037, + "step": 100850 + }, + { + "epoch": 6.852493545318658, + "grad_norm": 0.9721657037734985, + "learning_rate": 0.0001434383068351678, + "loss": 3.5488, + "step": 100855 + }, + { + "epoch": 6.852833265389319, + "grad_norm": 0.8548057675361633, + "learning_rate": 0.0001433958418263351, + "loss": 3.5401, + "step": 100860 + }, + { + "epoch": 6.853172985459981, + "grad_norm": 0.804379940032959, + "learning_rate": 0.00014335337681750237, + "loss": 3.3104, + "step": 100865 + }, + { + "epoch": 6.853512705530643, + "grad_norm": 0.8433701992034912, + "learning_rate": 0.00014331091180866965, + "loss": 3.2795, + "step": 100870 + }, + { + "epoch": 6.853852425601304, + "grad_norm": 1.3477611541748047, + "learning_rate": 0.00014326844679983695, + "loss": 3.7712, + "step": 100875 + }, + { + "epoch": 6.854192145671966, + "grad_norm": 0.8915440440177917, + "learning_rate": 0.0001432259817910042, + "loss": 3.4046, + "step": 100880 + }, + { + "epoch": 6.854531865742628, + "grad_norm": 0.8018163442611694, + "learning_rate": 0.0001431835167821715, + "loss": 3.4949, + "step": 100885 + }, + { + "epoch": 6.85487158581329, + "grad_norm": 0.9001146554946899, + "learning_rate": 0.00014314105177333877, + "loss": 3.3791, + "step": 100890 + }, + { + "epoch": 6.855211305883952, + "grad_norm": 0.8978468775749207, + "learning_rate": 0.00014309858676450605, + "loss": 3.3004, + "step": 100895 + }, + { + "epoch": 6.855551025954614, + "grad_norm": 0.9628803133964539, + "learning_rate": 0.00014305612175567335, + "loss": 3.5417, + "step": 100900 + }, + { + "epoch": 6.855890746025275, + "grad_norm": 1.011817216873169, + "learning_rate": 0.0001430136567468406, + "loss": 3.1732, + "step": 100905 + }, + { + "epoch": 6.856230466095937, + "grad_norm": 1.0274728536605835, + "learning_rate": 0.0001429711917380079, + "loss": 3.3252, + "step": 100910 + }, + { + "epoch": 6.856570186166599, + "grad_norm": 0.986108124256134, + "learning_rate": 0.00014292872672917517, + "loss": 3.3218, + "step": 100915 + }, + { + "epoch": 6.85690990623726, + "grad_norm": 1.0948619842529297, + "learning_rate": 0.00014288626172034245, + "loss": 3.438, + "step": 100920 + }, + { + "epoch": 6.857249626307922, + "grad_norm": 0.7939807772636414, + "learning_rate": 0.0001428437967115097, + "loss": 3.6024, + "step": 100925 + }, + { + "epoch": 6.857589346378584, + "grad_norm": 0.7623628973960876, + "learning_rate": 0.000142801331702677, + "loss": 3.3814, + "step": 100930 + }, + { + "epoch": 6.857929066449246, + "grad_norm": 1.1625709533691406, + "learning_rate": 0.0001427588666938443, + "loss": 3.5444, + "step": 100935 + }, + { + "epoch": 6.858268786519908, + "grad_norm": 1.081380844116211, + "learning_rate": 0.00014271640168501154, + "loss": 3.2925, + "step": 100940 + }, + { + "epoch": 6.85860850659057, + "grad_norm": 0.8377594947814941, + "learning_rate": 0.00014267393667617885, + "loss": 3.5587, + "step": 100945 + }, + { + "epoch": 6.858948226661231, + "grad_norm": 0.8029831647872925, + "learning_rate": 0.0001426314716673461, + "loss": 3.2715, + "step": 100950 + }, + { + "epoch": 6.859287946731893, + "grad_norm": 0.7940414547920227, + "learning_rate": 0.00014258900665851338, + "loss": 3.1372, + "step": 100955 + }, + { + "epoch": 6.859627666802554, + "grad_norm": 0.8657073974609375, + "learning_rate": 0.00014254654164968066, + "loss": 3.2331, + "step": 100960 + }, + { + "epoch": 6.859967386873216, + "grad_norm": 0.907664954662323, + "learning_rate": 0.00014250407664084794, + "loss": 3.5149, + "step": 100965 + }, + { + "epoch": 6.860307106943878, + "grad_norm": 1.247492790222168, + "learning_rate": 0.00014246161163201522, + "loss": 3.2654, + "step": 100970 + }, + { + "epoch": 6.8606468270145395, + "grad_norm": 0.8293247818946838, + "learning_rate": 0.0001424191466231825, + "loss": 3.403, + "step": 100975 + }, + { + "epoch": 6.860986547085202, + "grad_norm": 0.8730805516242981, + "learning_rate": 0.00014237668161434978, + "loss": 3.4025, + "step": 100980 + }, + { + "epoch": 6.861326267155864, + "grad_norm": 1.352369785308838, + "learning_rate": 0.00014233421660551706, + "loss": 3.2625, + "step": 100985 + }, + { + "epoch": 6.861665987226525, + "grad_norm": 1.1155585050582886, + "learning_rate": 0.00014229175159668434, + "loss": 3.2941, + "step": 100990 + }, + { + "epoch": 6.862005707297187, + "grad_norm": 0.875763475894928, + "learning_rate": 0.0001422492865878516, + "loss": 3.6992, + "step": 100995 + }, + { + "epoch": 6.862345427367849, + "grad_norm": 0.8720851540565491, + "learning_rate": 0.0001422068215790189, + "loss": 3.5585, + "step": 101000 + }, + { + "epoch": 6.86268514743851, + "grad_norm": 0.9235142469406128, + "learning_rate": 0.00014216435657018618, + "loss": 3.4799, + "step": 101005 + }, + { + "epoch": 6.863024867509172, + "grad_norm": 0.7615699172019958, + "learning_rate": 0.00014212189156135343, + "loss": 3.2922, + "step": 101010 + }, + { + "epoch": 6.863364587579834, + "grad_norm": 1.1513049602508545, + "learning_rate": 0.00014207942655252074, + "loss": 3.2581, + "step": 101015 + }, + { + "epoch": 6.8637043076504956, + "grad_norm": 0.8185482621192932, + "learning_rate": 0.000142036961543688, + "loss": 3.6931, + "step": 101020 + }, + { + "epoch": 6.864044027721158, + "grad_norm": 1.5502245426177979, + "learning_rate": 0.00014199449653485527, + "loss": 3.3955, + "step": 101025 + }, + { + "epoch": 6.86438374779182, + "grad_norm": 0.7102898359298706, + "learning_rate": 0.00014195203152602255, + "loss": 3.4604, + "step": 101030 + }, + { + "epoch": 6.864723467862481, + "grad_norm": 0.7765747308731079, + "learning_rate": 0.00014190956651718983, + "loss": 3.6931, + "step": 101035 + }, + { + "epoch": 6.865063187933143, + "grad_norm": 1.1555076837539673, + "learning_rate": 0.00014186710150835711, + "loss": 3.3005, + "step": 101040 + }, + { + "epoch": 6.865402908003805, + "grad_norm": 0.9134440422058105, + "learning_rate": 0.0001418246364995244, + "loss": 3.4285, + "step": 101045 + }, + { + "epoch": 6.865742628074466, + "grad_norm": 1.017502784729004, + "learning_rate": 0.00014178217149069168, + "loss": 3.5214, + "step": 101050 + }, + { + "epoch": 6.866082348145128, + "grad_norm": 1.1551882028579712, + "learning_rate": 0.00014173970648185893, + "loss": 3.302, + "step": 101055 + }, + { + "epoch": 6.86642206821579, + "grad_norm": 0.7749224901199341, + "learning_rate": 0.00014169724147302624, + "loss": 3.5709, + "step": 101060 + }, + { + "epoch": 6.866761788286452, + "grad_norm": 0.764544665813446, + "learning_rate": 0.00014165477646419352, + "loss": 3.3714, + "step": 101065 + }, + { + "epoch": 6.867101508357114, + "grad_norm": 0.7951776385307312, + "learning_rate": 0.0001416123114553608, + "loss": 3.4816, + "step": 101070 + }, + { + "epoch": 6.867441228427776, + "grad_norm": 1.0528030395507812, + "learning_rate": 0.00014156984644652808, + "loss": 3.3216, + "step": 101075 + }, + { + "epoch": 6.867780948498437, + "grad_norm": 0.8078467845916748, + "learning_rate": 0.00014152738143769533, + "loss": 3.5055, + "step": 101080 + }, + { + "epoch": 6.868120668569099, + "grad_norm": 1.0797284841537476, + "learning_rate": 0.00014148491642886264, + "loss": 3.1357, + "step": 101085 + }, + { + "epoch": 6.868460388639761, + "grad_norm": 0.9443169236183167, + "learning_rate": 0.0001414424514200299, + "loss": 3.5055, + "step": 101090 + }, + { + "epoch": 6.868800108710422, + "grad_norm": 1.2047996520996094, + "learning_rate": 0.00014139998641119717, + "loss": 3.2355, + "step": 101095 + }, + { + "epoch": 6.869139828781084, + "grad_norm": 0.7169967293739319, + "learning_rate": 0.00014135752140236448, + "loss": 3.3825, + "step": 101100 + }, + { + "epoch": 6.869479548851746, + "grad_norm": 0.9865683913230896, + "learning_rate": 0.00014131505639353173, + "loss": 3.2421, + "step": 101105 + }, + { + "epoch": 6.869819268922408, + "grad_norm": 0.8149899244308472, + "learning_rate": 0.000141272591384699, + "loss": 3.7328, + "step": 101110 + }, + { + "epoch": 6.87015898899307, + "grad_norm": 0.7215856909751892, + "learning_rate": 0.0001412301263758663, + "loss": 3.2217, + "step": 101115 + }, + { + "epoch": 6.870498709063732, + "grad_norm": 0.9432133436203003, + "learning_rate": 0.00014118766136703357, + "loss": 3.4142, + "step": 101120 + }, + { + "epoch": 6.870838429134393, + "grad_norm": 0.9277271628379822, + "learning_rate": 0.00014114519635820082, + "loss": 3.4903, + "step": 101125 + }, + { + "epoch": 6.871178149205055, + "grad_norm": 1.4281291961669922, + "learning_rate": 0.00014110273134936813, + "loss": 3.3256, + "step": 101130 + }, + { + "epoch": 6.871517869275717, + "grad_norm": 1.0470941066741943, + "learning_rate": 0.0001410602663405354, + "loss": 3.3822, + "step": 101135 + }, + { + "epoch": 6.871857589346378, + "grad_norm": 1.1371681690216064, + "learning_rate": 0.00014101780133170266, + "loss": 3.6415, + "step": 101140 + }, + { + "epoch": 6.87219730941704, + "grad_norm": 0.8202137351036072, + "learning_rate": 0.00014097533632286997, + "loss": 3.6232, + "step": 101145 + }, + { + "epoch": 6.872537029487702, + "grad_norm": 0.9435778856277466, + "learning_rate": 0.00014093287131403722, + "loss": 3.6956, + "step": 101150 + }, + { + "epoch": 6.872876749558364, + "grad_norm": 0.7950998544692993, + "learning_rate": 0.00014089040630520453, + "loss": 3.571, + "step": 101155 + }, + { + "epoch": 6.873216469629026, + "grad_norm": 0.9063976407051086, + "learning_rate": 0.00014084794129637178, + "loss": 3.3168, + "step": 101160 + }, + { + "epoch": 6.873556189699688, + "grad_norm": 1.0573251247406006, + "learning_rate": 0.00014080547628753906, + "loss": 3.4272, + "step": 101165 + }, + { + "epoch": 6.873895909770349, + "grad_norm": 0.859754741191864, + "learning_rate": 0.00014076301127870637, + "loss": 3.55, + "step": 101170 + }, + { + "epoch": 6.874235629841011, + "grad_norm": 0.9990733861923218, + "learning_rate": 0.00014072054626987362, + "loss": 3.4242, + "step": 101175 + }, + { + "epoch": 6.874575349911673, + "grad_norm": 0.8812786340713501, + "learning_rate": 0.0001406780812610409, + "loss": 3.4928, + "step": 101180 + }, + { + "epoch": 6.874915069982334, + "grad_norm": 1.0609939098358154, + "learning_rate": 0.00014063561625220818, + "loss": 3.597, + "step": 101185 + }, + { + "epoch": 6.875254790052996, + "grad_norm": 0.9653779864311218, + "learning_rate": 0.00014059315124337546, + "loss": 3.2121, + "step": 101190 + }, + { + "epoch": 6.875594510123658, + "grad_norm": 0.8939583897590637, + "learning_rate": 0.00014055068623454272, + "loss": 3.2925, + "step": 101195 + }, + { + "epoch": 6.87593423019432, + "grad_norm": 0.8325219750404358, + "learning_rate": 0.00014050822122571002, + "loss": 3.2949, + "step": 101200 + }, + { + "epoch": 6.876273950264982, + "grad_norm": 0.9736220240592957, + "learning_rate": 0.0001404657562168773, + "loss": 3.3095, + "step": 101205 + }, + { + "epoch": 6.876613670335644, + "grad_norm": 0.7509075999259949, + "learning_rate": 0.00014042329120804456, + "loss": 3.2767, + "step": 101210 + }, + { + "epoch": 6.876953390406305, + "grad_norm": 1.1399192810058594, + "learning_rate": 0.00014038082619921186, + "loss": 3.6976, + "step": 101215 + }, + { + "epoch": 6.877293110476967, + "grad_norm": 1.0620923042297363, + "learning_rate": 0.00014033836119037912, + "loss": 3.1954, + "step": 101220 + }, + { + "epoch": 6.877632830547629, + "grad_norm": 0.9478781819343567, + "learning_rate": 0.0001402958961815464, + "loss": 3.5369, + "step": 101225 + }, + { + "epoch": 6.87797255061829, + "grad_norm": 0.9600107073783875, + "learning_rate": 0.0001402534311727137, + "loss": 3.6497, + "step": 101230 + }, + { + "epoch": 6.878312270688952, + "grad_norm": 1.019928216934204, + "learning_rate": 0.00014021096616388096, + "loss": 3.5396, + "step": 101235 + }, + { + "epoch": 6.878651990759614, + "grad_norm": 0.967535138130188, + "learning_rate": 0.00014016850115504826, + "loss": 3.3179, + "step": 101240 + }, + { + "epoch": 6.878991710830276, + "grad_norm": 0.9542655944824219, + "learning_rate": 0.00014012603614621552, + "loss": 3.4179, + "step": 101245 + }, + { + "epoch": 6.879331430900938, + "grad_norm": 0.9585425853729248, + "learning_rate": 0.0001400835711373828, + "loss": 3.3812, + "step": 101250 + }, + { + "epoch": 6.8796711509716, + "grad_norm": 0.7911840677261353, + "learning_rate": 0.00014004110612855008, + "loss": 3.5517, + "step": 101255 + }, + { + "epoch": 6.880010871042261, + "grad_norm": 0.724815309047699, + "learning_rate": 0.00013999864111971736, + "loss": 3.4475, + "step": 101260 + }, + { + "epoch": 6.880350591112923, + "grad_norm": 1.1129707098007202, + "learning_rate": 0.00013995617611088464, + "loss": 3.487, + "step": 101265 + }, + { + "epoch": 6.880690311183585, + "grad_norm": 0.8528203368186951, + "learning_rate": 0.00013991371110205192, + "loss": 3.2567, + "step": 101270 + }, + { + "epoch": 6.881030031254246, + "grad_norm": 0.9860325455665588, + "learning_rate": 0.0001398712460932192, + "loss": 3.7642, + "step": 101275 + }, + { + "epoch": 6.881369751324908, + "grad_norm": 1.0260109901428223, + "learning_rate": 0.00013982878108438645, + "loss": 3.3463, + "step": 101280 + }, + { + "epoch": 6.88170947139557, + "grad_norm": 0.8594717383384705, + "learning_rate": 0.00013978631607555376, + "loss": 3.3266, + "step": 101285 + }, + { + "epoch": 6.882049191466232, + "grad_norm": 0.7728205323219299, + "learning_rate": 0.000139743851066721, + "loss": 3.4005, + "step": 101290 + }, + { + "epoch": 6.882388911536894, + "grad_norm": 0.8960942029953003, + "learning_rate": 0.0001397013860578883, + "loss": 3.3692, + "step": 101295 + }, + { + "epoch": 6.882728631607556, + "grad_norm": 0.8587968945503235, + "learning_rate": 0.0001396589210490556, + "loss": 3.536, + "step": 101300 + }, + { + "epoch": 6.883068351678217, + "grad_norm": 1.0335484743118286, + "learning_rate": 0.00013961645604022285, + "loss": 3.316, + "step": 101305 + }, + { + "epoch": 6.883408071748879, + "grad_norm": 0.7530755996704102, + "learning_rate": 0.00013957399103139013, + "loss": 3.2945, + "step": 101310 + }, + { + "epoch": 6.883747791819541, + "grad_norm": 1.1303772926330566, + "learning_rate": 0.0001395315260225574, + "loss": 3.4204, + "step": 101315 + }, + { + "epoch": 6.884087511890202, + "grad_norm": 0.9682402610778809, + "learning_rate": 0.0001394890610137247, + "loss": 3.3024, + "step": 101320 + }, + { + "epoch": 6.884427231960864, + "grad_norm": 0.9693924188613892, + "learning_rate": 0.000139446596004892, + "loss": 3.541, + "step": 101325 + }, + { + "epoch": 6.884766952031526, + "grad_norm": 0.7893980741500854, + "learning_rate": 0.00013940413099605925, + "loss": 3.2648, + "step": 101330 + }, + { + "epoch": 6.885106672102188, + "grad_norm": 0.8071556687355042, + "learning_rate": 0.00013936166598722653, + "loss": 3.4559, + "step": 101335 + }, + { + "epoch": 6.88544639217285, + "grad_norm": 0.8542060256004333, + "learning_rate": 0.0001393192009783938, + "loss": 3.2506, + "step": 101340 + }, + { + "epoch": 6.885786112243512, + "grad_norm": 1.1016879081726074, + "learning_rate": 0.0001392767359695611, + "loss": 3.5506, + "step": 101345 + }, + { + "epoch": 6.886125832314173, + "grad_norm": 0.9285717606544495, + "learning_rate": 0.00013923427096072834, + "loss": 3.5815, + "step": 101350 + }, + { + "epoch": 6.886465552384835, + "grad_norm": 0.9052934050559998, + "learning_rate": 0.00013919180595189565, + "loss": 3.4356, + "step": 101355 + }, + { + "epoch": 6.886805272455497, + "grad_norm": 0.784863293170929, + "learning_rate": 0.00013914934094306293, + "loss": 3.3325, + "step": 101360 + }, + { + "epoch": 6.887144992526158, + "grad_norm": 0.7322821021080017, + "learning_rate": 0.00013910687593423018, + "loss": 3.5078, + "step": 101365 + }, + { + "epoch": 6.88748471259682, + "grad_norm": 0.8831473588943481, + "learning_rate": 0.0001390644109253975, + "loss": 3.3435, + "step": 101370 + }, + { + "epoch": 6.8878244326674825, + "grad_norm": 1.1633986234664917, + "learning_rate": 0.00013902194591656474, + "loss": 3.2625, + "step": 101375 + }, + { + "epoch": 6.888164152738144, + "grad_norm": 0.8711217641830444, + "learning_rate": 0.00013897948090773202, + "loss": 3.4136, + "step": 101380 + }, + { + "epoch": 6.888503872808806, + "grad_norm": 0.9397038221359253, + "learning_rate": 0.0001389370158988993, + "loss": 3.369, + "step": 101385 + }, + { + "epoch": 6.888843592879468, + "grad_norm": 0.9270291924476624, + "learning_rate": 0.00013889455089006658, + "loss": 3.4834, + "step": 101390 + }, + { + "epoch": 6.889183312950129, + "grad_norm": 1.0873943567276, + "learning_rate": 0.00013885208588123386, + "loss": 3.5854, + "step": 101395 + }, + { + "epoch": 6.889523033020791, + "grad_norm": 0.6862496137619019, + "learning_rate": 0.00013880962087240114, + "loss": 3.2661, + "step": 101400 + }, + { + "epoch": 6.889862753091453, + "grad_norm": 0.8782524466514587, + "learning_rate": 0.00013876715586356842, + "loss": 3.4167, + "step": 101405 + }, + { + "epoch": 6.890202473162114, + "grad_norm": 1.1416497230529785, + "learning_rate": 0.0001387246908547357, + "loss": 3.4708, + "step": 101410 + }, + { + "epoch": 6.890542193232776, + "grad_norm": 0.9186204075813293, + "learning_rate": 0.00013868222584590299, + "loss": 3.2242, + "step": 101415 + }, + { + "epoch": 6.8908819133034385, + "grad_norm": 1.111312985420227, + "learning_rate": 0.00013863976083707024, + "loss": 3.1752, + "step": 101420 + }, + { + "epoch": 6.8912216333741, + "grad_norm": 1.1170084476470947, + "learning_rate": 0.00013859729582823755, + "loss": 3.4916, + "step": 101425 + }, + { + "epoch": 6.891561353444762, + "grad_norm": 1.0672558546066284, + "learning_rate": 0.00013855483081940483, + "loss": 3.5215, + "step": 101430 + }, + { + "epoch": 6.891901073515423, + "grad_norm": 1.028028130531311, + "learning_rate": 0.00013851236581057208, + "loss": 3.3143, + "step": 101435 + }, + { + "epoch": 6.892240793586085, + "grad_norm": 1.2537147998809814, + "learning_rate": 0.00013846990080173939, + "loss": 3.7886, + "step": 101440 + }, + { + "epoch": 6.892580513656747, + "grad_norm": 0.6689179539680481, + "learning_rate": 0.00013842743579290664, + "loss": 3.3904, + "step": 101445 + }, + { + "epoch": 6.892920233727408, + "grad_norm": 0.876054048538208, + "learning_rate": 0.00013838497078407392, + "loss": 3.3871, + "step": 101450 + }, + { + "epoch": 6.89325995379807, + "grad_norm": 0.8758645057678223, + "learning_rate": 0.0001383425057752412, + "loss": 3.3715, + "step": 101455 + }, + { + "epoch": 6.893599673868732, + "grad_norm": 1.2365021705627441, + "learning_rate": 0.00013830004076640848, + "loss": 3.4602, + "step": 101460 + }, + { + "epoch": 6.893939393939394, + "grad_norm": 0.9217543005943298, + "learning_rate": 0.00013825757575757576, + "loss": 3.1677, + "step": 101465 + }, + { + "epoch": 6.894279114010056, + "grad_norm": 0.7891196608543396, + "learning_rate": 0.00013821511074874304, + "loss": 3.646, + "step": 101470 + }, + { + "epoch": 6.894618834080718, + "grad_norm": 1.3513967990875244, + "learning_rate": 0.00013817264573991032, + "loss": 3.4477, + "step": 101475 + }, + { + "epoch": 6.894958554151379, + "grad_norm": 0.8403365612030029, + "learning_rate": 0.00013813018073107757, + "loss": 3.1065, + "step": 101480 + }, + { + "epoch": 6.895298274222041, + "grad_norm": 0.8614407181739807, + "learning_rate": 0.00013808771572224488, + "loss": 3.2246, + "step": 101485 + }, + { + "epoch": 6.895637994292703, + "grad_norm": 0.854903519153595, + "learning_rate": 0.00013804525071341216, + "loss": 3.1885, + "step": 101490 + }, + { + "epoch": 6.895977714363364, + "grad_norm": 1.0327751636505127, + "learning_rate": 0.00013800278570457944, + "loss": 3.3503, + "step": 101495 + }, + { + "epoch": 6.896317434434026, + "grad_norm": 0.8288962841033936, + "learning_rate": 0.00013796032069574672, + "loss": 3.3173, + "step": 101500 + }, + { + "epoch": 6.896657154504688, + "grad_norm": 0.8965150117874146, + "learning_rate": 0.00013791785568691397, + "loss": 3.4314, + "step": 101505 + }, + { + "epoch": 6.89699687457535, + "grad_norm": 0.9726516604423523, + "learning_rate": 0.00013787539067808128, + "loss": 3.4469, + "step": 101510 + }, + { + "epoch": 6.897336594646012, + "grad_norm": 1.0832135677337646, + "learning_rate": 0.00013783292566924853, + "loss": 3.5436, + "step": 101515 + }, + { + "epoch": 6.897676314716674, + "grad_norm": 1.2080931663513184, + "learning_rate": 0.0001377904606604158, + "loss": 3.0385, + "step": 101520 + }, + { + "epoch": 6.898016034787335, + "grad_norm": 0.8961111903190613, + "learning_rate": 0.00013774799565158312, + "loss": 3.5095, + "step": 101525 + }, + { + "epoch": 6.898355754857997, + "grad_norm": 0.8710960149765015, + "learning_rate": 0.00013770553064275037, + "loss": 3.2362, + "step": 101530 + }, + { + "epoch": 6.898695474928659, + "grad_norm": 1.0490655899047852, + "learning_rate": 0.00013766306563391765, + "loss": 3.2673, + "step": 101535 + }, + { + "epoch": 6.89903519499932, + "grad_norm": 0.6962436437606812, + "learning_rate": 0.00013762060062508493, + "loss": 3.3939, + "step": 101540 + }, + { + "epoch": 6.899374915069982, + "grad_norm": 0.802471935749054, + "learning_rate": 0.0001375781356162522, + "loss": 3.2421, + "step": 101545 + }, + { + "epoch": 6.899714635140644, + "grad_norm": 0.8350061178207397, + "learning_rate": 0.00013753567060741947, + "loss": 3.5342, + "step": 101550 + }, + { + "epoch": 6.900054355211306, + "grad_norm": 0.9025750756263733, + "learning_rate": 0.00013749320559858677, + "loss": 3.4627, + "step": 101555 + }, + { + "epoch": 6.900394075281968, + "grad_norm": 0.9091207981109619, + "learning_rate": 0.00013745074058975405, + "loss": 3.5735, + "step": 101560 + }, + { + "epoch": 6.90073379535263, + "grad_norm": 0.7403337955474854, + "learning_rate": 0.0001374082755809213, + "loss": 3.4195, + "step": 101565 + }, + { + "epoch": 6.901073515423291, + "grad_norm": 2.100260019302368, + "learning_rate": 0.0001373658105720886, + "loss": 3.304, + "step": 101570 + }, + { + "epoch": 6.901413235493953, + "grad_norm": 0.8179365992546082, + "learning_rate": 0.00013732334556325587, + "loss": 3.3541, + "step": 101575 + }, + { + "epoch": 6.901752955564615, + "grad_norm": 1.1461533308029175, + "learning_rate": 0.00013728088055442317, + "loss": 3.1279, + "step": 101580 + }, + { + "epoch": 6.902092675635276, + "grad_norm": 0.9849970936775208, + "learning_rate": 0.00013723841554559043, + "loss": 3.4806, + "step": 101585 + }, + { + "epoch": 6.902432395705938, + "grad_norm": 0.8572201728820801, + "learning_rate": 0.0001371959505367577, + "loss": 3.2553, + "step": 101590 + }, + { + "epoch": 6.9027721157766, + "grad_norm": 0.8476952910423279, + "learning_rate": 0.00013715348552792501, + "loss": 3.2096, + "step": 101595 + }, + { + "epoch": 6.903111835847262, + "grad_norm": 0.802068829536438, + "learning_rate": 0.00013711102051909227, + "loss": 3.1891, + "step": 101600 + }, + { + "epoch": 6.903451555917924, + "grad_norm": 0.9268673062324524, + "learning_rate": 0.00013706855551025955, + "loss": 3.6145, + "step": 101605 + }, + { + "epoch": 6.903791275988586, + "grad_norm": 1.0828722715377808, + "learning_rate": 0.00013702609050142683, + "loss": 3.2491, + "step": 101610 + }, + { + "epoch": 6.904130996059247, + "grad_norm": 0.8985880017280579, + "learning_rate": 0.0001369836254925941, + "loss": 3.425, + "step": 101615 + }, + { + "epoch": 6.904470716129909, + "grad_norm": 0.8808119297027588, + "learning_rate": 0.00013694116048376136, + "loss": 3.7385, + "step": 101620 + }, + { + "epoch": 6.904810436200571, + "grad_norm": 0.881458580493927, + "learning_rate": 0.00013689869547492867, + "loss": 3.626, + "step": 101625 + }, + { + "epoch": 6.905150156271232, + "grad_norm": 1.1087405681610107, + "learning_rate": 0.00013685623046609595, + "loss": 3.4672, + "step": 101630 + }, + { + "epoch": 6.905489876341894, + "grad_norm": 1.0140860080718994, + "learning_rate": 0.0001368137654572632, + "loss": 3.4307, + "step": 101635 + }, + { + "epoch": 6.905829596412556, + "grad_norm": 0.9326143860816956, + "learning_rate": 0.0001367713004484305, + "loss": 3.5713, + "step": 101640 + }, + { + "epoch": 6.906169316483218, + "grad_norm": 0.9585304260253906, + "learning_rate": 0.00013672883543959776, + "loss": 3.1347, + "step": 101645 + }, + { + "epoch": 6.90650903655388, + "grad_norm": 0.7879514098167419, + "learning_rate": 0.00013668637043076504, + "loss": 3.7207, + "step": 101650 + }, + { + "epoch": 6.906848756624541, + "grad_norm": 1.6185957193374634, + "learning_rate": 0.00013664390542193235, + "loss": 3.5642, + "step": 101655 + }, + { + "epoch": 6.907188476695203, + "grad_norm": 0.8596517443656921, + "learning_rate": 0.0001366014404130996, + "loss": 3.5218, + "step": 101660 + }, + { + "epoch": 6.907528196765865, + "grad_norm": 1.162042260169983, + "learning_rate": 0.0001365589754042669, + "loss": 3.4671, + "step": 101665 + }, + { + "epoch": 6.907867916836526, + "grad_norm": 0.7792569994926453, + "learning_rate": 0.00013651651039543416, + "loss": 3.4768, + "step": 101670 + }, + { + "epoch": 6.908207636907188, + "grad_norm": 0.9232324361801147, + "learning_rate": 0.00013647404538660144, + "loss": 3.3371, + "step": 101675 + }, + { + "epoch": 6.90854735697785, + "grad_norm": 1.0003706216812134, + "learning_rate": 0.00013643158037776872, + "loss": 3.5705, + "step": 101680 + }, + { + "epoch": 6.908887077048512, + "grad_norm": 0.9934501051902771, + "learning_rate": 0.000136389115368936, + "loss": 3.4913, + "step": 101685 + }, + { + "epoch": 6.909226797119174, + "grad_norm": 0.8606526851654053, + "learning_rate": 0.00013634665036010328, + "loss": 3.2607, + "step": 101690 + }, + { + "epoch": 6.909566517189836, + "grad_norm": 0.8317620754241943, + "learning_rate": 0.00013630418535127056, + "loss": 3.1253, + "step": 101695 + }, + { + "epoch": 6.909906237260497, + "grad_norm": 1.0005552768707275, + "learning_rate": 0.00013626172034243784, + "loss": 3.3197, + "step": 101700 + }, + { + "epoch": 6.910245957331159, + "grad_norm": 0.8887527585029602, + "learning_rate": 0.0001362192553336051, + "loss": 3.2949, + "step": 101705 + }, + { + "epoch": 6.910585677401821, + "grad_norm": 0.9670987725257874, + "learning_rate": 0.0001361767903247724, + "loss": 3.592, + "step": 101710 + }, + { + "epoch": 6.910925397472482, + "grad_norm": 0.7432765364646912, + "learning_rate": 0.00013613432531593965, + "loss": 3.511, + "step": 101715 + }, + { + "epoch": 6.911265117543144, + "grad_norm": 0.8354858756065369, + "learning_rate": 0.00013609186030710693, + "loss": 3.3588, + "step": 101720 + }, + { + "epoch": 6.911604837613806, + "grad_norm": 0.9712925553321838, + "learning_rate": 0.00013604939529827424, + "loss": 3.5673, + "step": 101725 + }, + { + "epoch": 6.911944557684468, + "grad_norm": 0.9201673269271851, + "learning_rate": 0.0001360069302894415, + "loss": 3.4016, + "step": 101730 + }, + { + "epoch": 6.91228427775513, + "grad_norm": 0.7941307425498962, + "learning_rate": 0.00013596446528060877, + "loss": 3.6008, + "step": 101735 + }, + { + "epoch": 6.912623997825792, + "grad_norm": 1.0414470434188843, + "learning_rate": 0.00013592200027177605, + "loss": 3.3407, + "step": 101740 + }, + { + "epoch": 6.912963717896453, + "grad_norm": 0.8663234114646912, + "learning_rate": 0.00013587953526294333, + "loss": 3.4649, + "step": 101745 + }, + { + "epoch": 6.913303437967115, + "grad_norm": 1.0127794742584229, + "learning_rate": 0.00013583707025411064, + "loss": 3.2274, + "step": 101750 + }, + { + "epoch": 6.913643158037777, + "grad_norm": 0.9511242508888245, + "learning_rate": 0.0001357946052452779, + "loss": 3.5003, + "step": 101755 + }, + { + "epoch": 6.913982878108438, + "grad_norm": 1.1198405027389526, + "learning_rate": 0.00013575214023644517, + "loss": 3.3394, + "step": 101760 + }, + { + "epoch": 6.9143225981791, + "grad_norm": 0.8177787661552429, + "learning_rate": 0.00013570967522761246, + "loss": 3.4089, + "step": 101765 + }, + { + "epoch": 6.914662318249762, + "grad_norm": 0.7541701197624207, + "learning_rate": 0.00013566721021877974, + "loss": 3.3409, + "step": 101770 + }, + { + "epoch": 6.915002038320424, + "grad_norm": 1.0336447954177856, + "learning_rate": 0.000135624745209947, + "loss": 3.3172, + "step": 101775 + }, + { + "epoch": 6.915341758391086, + "grad_norm": 0.9538044333457947, + "learning_rate": 0.0001355822802011143, + "loss": 3.7397, + "step": 101780 + }, + { + "epoch": 6.915681478461748, + "grad_norm": 0.9641798138618469, + "learning_rate": 0.00013553981519228158, + "loss": 3.2033, + "step": 101785 + }, + { + "epoch": 6.916021198532409, + "grad_norm": 0.8513450622558594, + "learning_rate": 0.00013549735018344883, + "loss": 3.5112, + "step": 101790 + }, + { + "epoch": 6.916360918603071, + "grad_norm": 0.9310051798820496, + "learning_rate": 0.00013545488517461614, + "loss": 3.2491, + "step": 101795 + }, + { + "epoch": 6.916700638673733, + "grad_norm": 0.9847660660743713, + "learning_rate": 0.0001354124201657834, + "loss": 3.434, + "step": 101800 + }, + { + "epoch": 6.917040358744394, + "grad_norm": 1.3631609678268433, + "learning_rate": 0.00013536995515695067, + "loss": 3.2328, + "step": 101805 + }, + { + "epoch": 6.917380078815056, + "grad_norm": 1.0353211164474487, + "learning_rate": 0.00013532749014811795, + "loss": 3.6009, + "step": 101810 + }, + { + "epoch": 6.917719798885718, + "grad_norm": 0.874313235282898, + "learning_rate": 0.00013528502513928523, + "loss": 3.3019, + "step": 101815 + }, + { + "epoch": 6.91805951895638, + "grad_norm": 0.8505018949508667, + "learning_rate": 0.0001352425601304525, + "loss": 3.5431, + "step": 101820 + }, + { + "epoch": 6.918399239027042, + "grad_norm": 0.929074227809906, + "learning_rate": 0.0001352000951216198, + "loss": 3.4531, + "step": 101825 + }, + { + "epoch": 6.918738959097704, + "grad_norm": 0.7535291314125061, + "learning_rate": 0.00013515763011278707, + "loss": 3.4661, + "step": 101830 + }, + { + "epoch": 6.919078679168365, + "grad_norm": 1.0665334463119507, + "learning_rate": 0.00013511516510395435, + "loss": 3.3754, + "step": 101835 + }, + { + "epoch": 6.919418399239027, + "grad_norm": 0.7044690251350403, + "learning_rate": 0.00013507270009512163, + "loss": 3.4457, + "step": 101840 + }, + { + "epoch": 6.919758119309689, + "grad_norm": 1.198943018913269, + "learning_rate": 0.00013503023508628888, + "loss": 3.5305, + "step": 101845 + }, + { + "epoch": 6.92009783938035, + "grad_norm": 0.9499064683914185, + "learning_rate": 0.0001349877700774562, + "loss": 3.3021, + "step": 101850 + }, + { + "epoch": 6.920437559451012, + "grad_norm": 0.8623164296150208, + "learning_rate": 0.00013494530506862347, + "loss": 3.6557, + "step": 101855 + }, + { + "epoch": 6.920777279521674, + "grad_norm": 0.770013153553009, + "learning_rate": 0.00013490284005979072, + "loss": 3.3662, + "step": 101860 + }, + { + "epoch": 6.921116999592336, + "grad_norm": 1.0082358121871948, + "learning_rate": 0.00013486037505095803, + "loss": 3.3973, + "step": 101865 + }, + { + "epoch": 6.921456719662998, + "grad_norm": 0.7375884056091309, + "learning_rate": 0.00013481791004212528, + "loss": 3.3041, + "step": 101870 + }, + { + "epoch": 6.92179643973366, + "grad_norm": 1.02072012424469, + "learning_rate": 0.00013477544503329256, + "loss": 3.2051, + "step": 101875 + }, + { + "epoch": 6.922136159804321, + "grad_norm": 0.9061711430549622, + "learning_rate": 0.00013473298002445984, + "loss": 3.3249, + "step": 101880 + }, + { + "epoch": 6.922475879874983, + "grad_norm": 0.877646267414093, + "learning_rate": 0.00013469051501562712, + "loss": 3.2617, + "step": 101885 + }, + { + "epoch": 6.922815599945645, + "grad_norm": 0.9388894438743591, + "learning_rate": 0.0001346480500067944, + "loss": 3.371, + "step": 101890 + }, + { + "epoch": 6.923155320016306, + "grad_norm": 1.0341356992721558, + "learning_rate": 0.00013460558499796168, + "loss": 3.5744, + "step": 101895 + }, + { + "epoch": 6.923495040086968, + "grad_norm": 0.7376847267150879, + "learning_rate": 0.00013456311998912896, + "loss": 3.2475, + "step": 101900 + }, + { + "epoch": 6.92383476015763, + "grad_norm": 1.178245186805725, + "learning_rate": 0.00013452065498029622, + "loss": 3.3048, + "step": 101905 + }, + { + "epoch": 6.924174480228292, + "grad_norm": 0.7689367532730103, + "learning_rate": 0.00013447818997146352, + "loss": 3.2697, + "step": 101910 + }, + { + "epoch": 6.924514200298954, + "grad_norm": 1.1407854557037354, + "learning_rate": 0.0001344357249626308, + "loss": 3.2701, + "step": 101915 + }, + { + "epoch": 6.924853920369616, + "grad_norm": 1.1983026266098022, + "learning_rate": 0.00013439325995379808, + "loss": 3.4954, + "step": 101920 + }, + { + "epoch": 6.925193640440277, + "grad_norm": 0.9432186484336853, + "learning_rate": 0.00013435079494496536, + "loss": 3.4463, + "step": 101925 + }, + { + "epoch": 6.925533360510939, + "grad_norm": 0.883818507194519, + "learning_rate": 0.00013430832993613262, + "loss": 3.246, + "step": 101930 + }, + { + "epoch": 6.925873080581601, + "grad_norm": 1.0096933841705322, + "learning_rate": 0.00013426586492729992, + "loss": 3.2966, + "step": 101935 + }, + { + "epoch": 6.926212800652262, + "grad_norm": 1.0609618425369263, + "learning_rate": 0.00013422339991846718, + "loss": 3.3662, + "step": 101940 + }, + { + "epoch": 6.926552520722924, + "grad_norm": 0.9553772211074829, + "learning_rate": 0.00013418093490963446, + "loss": 3.3897, + "step": 101945 + }, + { + "epoch": 6.9268922407935865, + "grad_norm": 0.9351905584335327, + "learning_rate": 0.00013413846990080176, + "loss": 3.444, + "step": 101950 + }, + { + "epoch": 6.927231960864248, + "grad_norm": 0.9522320628166199, + "learning_rate": 0.00013409600489196902, + "loss": 3.2976, + "step": 101955 + }, + { + "epoch": 6.92757168093491, + "grad_norm": 0.782867431640625, + "learning_rate": 0.0001340535398831363, + "loss": 3.5784, + "step": 101960 + }, + { + "epoch": 6.927911401005572, + "grad_norm": 0.842818021774292, + "learning_rate": 0.00013401107487430358, + "loss": 3.5179, + "step": 101965 + }, + { + "epoch": 6.928251121076233, + "grad_norm": 0.8073701858520508, + "learning_rate": 0.00013396860986547086, + "loss": 3.7115, + "step": 101970 + }, + { + "epoch": 6.928590841146895, + "grad_norm": 1.0256638526916504, + "learning_rate": 0.0001339261448566381, + "loss": 3.5849, + "step": 101975 + }, + { + "epoch": 6.928930561217557, + "grad_norm": 0.8895252346992493, + "learning_rate": 0.00013388367984780542, + "loss": 3.4196, + "step": 101980 + }, + { + "epoch": 6.929270281288218, + "grad_norm": 1.03719162940979, + "learning_rate": 0.0001338412148389727, + "loss": 3.4293, + "step": 101985 + }, + { + "epoch": 6.92961000135888, + "grad_norm": 1.0631219148635864, + "learning_rate": 0.00013379874983013995, + "loss": 3.3011, + "step": 101990 + }, + { + "epoch": 6.9299497214295425, + "grad_norm": 0.9639312028884888, + "learning_rate": 0.00013375628482130726, + "loss": 3.3928, + "step": 101995 + }, + { + "epoch": 6.930289441500204, + "grad_norm": 0.9927453398704529, + "learning_rate": 0.0001337138198124745, + "loss": 3.6401, + "step": 102000 + }, + { + "epoch": 6.930629161570866, + "grad_norm": 1.0820658206939697, + "learning_rate": 0.00013367135480364182, + "loss": 3.5374, + "step": 102005 + }, + { + "epoch": 6.930968881641528, + "grad_norm": 0.996877133846283, + "learning_rate": 0.00013362888979480907, + "loss": 3.4703, + "step": 102010 + }, + { + "epoch": 6.931308601712189, + "grad_norm": 0.8691183924674988, + "learning_rate": 0.00013358642478597635, + "loss": 3.3888, + "step": 102015 + }, + { + "epoch": 6.931648321782851, + "grad_norm": 0.6742845177650452, + "learning_rate": 0.00013354395977714366, + "loss": 3.4689, + "step": 102020 + }, + { + "epoch": 6.931988041853513, + "grad_norm": 0.8561559319496155, + "learning_rate": 0.0001335014947683109, + "loss": 3.3927, + "step": 102025 + }, + { + "epoch": 6.932327761924174, + "grad_norm": 1.0555123090744019, + "learning_rate": 0.0001334590297594782, + "loss": 3.1907, + "step": 102030 + }, + { + "epoch": 6.932667481994836, + "grad_norm": 1.412047266960144, + "learning_rate": 0.00013341656475064547, + "loss": 3.5434, + "step": 102035 + }, + { + "epoch": 6.9330072020654985, + "grad_norm": 0.8208606243133545, + "learning_rate": 0.00013337409974181275, + "loss": 3.6342, + "step": 102040 + }, + { + "epoch": 6.93334692213616, + "grad_norm": 0.9457042813301086, + "learning_rate": 0.00013333163473298, + "loss": 3.3214, + "step": 102045 + }, + { + "epoch": 6.933686642206822, + "grad_norm": 1.1145840883255005, + "learning_rate": 0.0001332891697241473, + "loss": 3.4683, + "step": 102050 + }, + { + "epoch": 6.934026362277484, + "grad_norm": 1.0097225904464722, + "learning_rate": 0.0001332467047153146, + "loss": 3.4821, + "step": 102055 + }, + { + "epoch": 6.934366082348145, + "grad_norm": 1.1697115898132324, + "learning_rate": 0.00013320423970648184, + "loss": 3.368, + "step": 102060 + }, + { + "epoch": 6.934705802418807, + "grad_norm": 1.1186330318450928, + "learning_rate": 0.00013316177469764915, + "loss": 3.3783, + "step": 102065 + }, + { + "epoch": 6.935045522489469, + "grad_norm": 0.9686408042907715, + "learning_rate": 0.0001331193096888164, + "loss": 3.351, + "step": 102070 + }, + { + "epoch": 6.93538524256013, + "grad_norm": 0.9809467792510986, + "learning_rate": 0.00013307684467998368, + "loss": 3.3776, + "step": 102075 + }, + { + "epoch": 6.935724962630792, + "grad_norm": 1.0204620361328125, + "learning_rate": 0.000133034379671151, + "loss": 3.3031, + "step": 102080 + }, + { + "epoch": 6.9360646827014545, + "grad_norm": 1.0035405158996582, + "learning_rate": 0.00013299191466231824, + "loss": 3.4782, + "step": 102085 + }, + { + "epoch": 6.936404402772116, + "grad_norm": 1.050782561302185, + "learning_rate": 0.00013294944965348555, + "loss": 3.478, + "step": 102090 + }, + { + "epoch": 6.936744122842778, + "grad_norm": 1.098344326019287, + "learning_rate": 0.0001329069846446528, + "loss": 3.2669, + "step": 102095 + }, + { + "epoch": 6.93708384291344, + "grad_norm": 1.0050472021102905, + "learning_rate": 0.00013286451963582008, + "loss": 3.0791, + "step": 102100 + }, + { + "epoch": 6.937423562984101, + "grad_norm": 0.8295729756355286, + "learning_rate": 0.00013282205462698736, + "loss": 3.2415, + "step": 102105 + }, + { + "epoch": 6.937763283054763, + "grad_norm": 0.9085100293159485, + "learning_rate": 0.00013277958961815464, + "loss": 3.2109, + "step": 102110 + }, + { + "epoch": 6.938103003125424, + "grad_norm": 0.8997124433517456, + "learning_rate": 0.00013273712460932192, + "loss": 3.2249, + "step": 102115 + }, + { + "epoch": 6.938442723196086, + "grad_norm": 0.9896427392959595, + "learning_rate": 0.0001326946596004892, + "loss": 3.5464, + "step": 102120 + }, + { + "epoch": 6.938782443266748, + "grad_norm": 0.8199006915092468, + "learning_rate": 0.00013265219459165649, + "loss": 3.3721, + "step": 102125 + }, + { + "epoch": 6.93912216333741, + "grad_norm": 0.8422032594680786, + "learning_rate": 0.00013260972958282374, + "loss": 3.5763, + "step": 102130 + }, + { + "epoch": 6.939461883408072, + "grad_norm": 0.8876162171363831, + "learning_rate": 0.00013256726457399105, + "loss": 3.3871, + "step": 102135 + }, + { + "epoch": 6.939801603478734, + "grad_norm": 0.9517466425895691, + "learning_rate": 0.0001325247995651583, + "loss": 3.4835, + "step": 102140 + }, + { + "epoch": 6.940141323549395, + "grad_norm": 1.0903910398483276, + "learning_rate": 0.00013248233455632558, + "loss": 3.3902, + "step": 102145 + }, + { + "epoch": 6.940481043620057, + "grad_norm": 0.7423407435417175, + "learning_rate": 0.00013243986954749289, + "loss": 3.4727, + "step": 102150 + }, + { + "epoch": 6.940820763690719, + "grad_norm": 0.8958464860916138, + "learning_rate": 0.00013239740453866014, + "loss": 3.4627, + "step": 102155 + }, + { + "epoch": 6.94116048376138, + "grad_norm": 0.9520241618156433, + "learning_rate": 0.00013235493952982742, + "loss": 3.1967, + "step": 102160 + }, + { + "epoch": 6.941500203832042, + "grad_norm": 0.9037117958068848, + "learning_rate": 0.0001323124745209947, + "loss": 3.4332, + "step": 102165 + }, + { + "epoch": 6.941839923902704, + "grad_norm": 0.7253947854042053, + "learning_rate": 0.00013227000951216198, + "loss": 3.2662, + "step": 102170 + }, + { + "epoch": 6.942179643973366, + "grad_norm": 1.0575319528579712, + "learning_rate": 0.00013222754450332929, + "loss": 3.541, + "step": 102175 + }, + { + "epoch": 6.942519364044028, + "grad_norm": 0.768189549446106, + "learning_rate": 0.00013218507949449654, + "loss": 3.2466, + "step": 102180 + }, + { + "epoch": 6.94285908411469, + "grad_norm": 1.1129757165908813, + "learning_rate": 0.00013214261448566382, + "loss": 3.5956, + "step": 102185 + }, + { + "epoch": 6.943198804185351, + "grad_norm": 0.8753491640090942, + "learning_rate": 0.0001321001494768311, + "loss": 3.2772, + "step": 102190 + }, + { + "epoch": 6.943538524256013, + "grad_norm": 0.7835537791252136, + "learning_rate": 0.00013205768446799838, + "loss": 3.49, + "step": 102195 + }, + { + "epoch": 6.943878244326675, + "grad_norm": 1.0603389739990234, + "learning_rate": 0.00013201521945916563, + "loss": 3.4257, + "step": 102200 + }, + { + "epoch": 6.944217964397336, + "grad_norm": 0.8037093877792358, + "learning_rate": 0.00013197275445033294, + "loss": 3.5344, + "step": 102205 + }, + { + "epoch": 6.944557684467998, + "grad_norm": 1.007185697555542, + "learning_rate": 0.00013193028944150022, + "loss": 3.2158, + "step": 102210 + }, + { + "epoch": 6.9448974045386604, + "grad_norm": 1.0445743799209595, + "learning_rate": 0.00013188782443266747, + "loss": 3.4936, + "step": 102215 + }, + { + "epoch": 6.945237124609322, + "grad_norm": 0.7972914576530457, + "learning_rate": 0.00013184535942383478, + "loss": 3.3879, + "step": 102220 + }, + { + "epoch": 6.945576844679984, + "grad_norm": 0.8576540946960449, + "learning_rate": 0.00013180289441500203, + "loss": 3.4721, + "step": 102225 + }, + { + "epoch": 6.945916564750646, + "grad_norm": 0.8340522050857544, + "learning_rate": 0.0001317604294061693, + "loss": 3.5706, + "step": 102230 + }, + { + "epoch": 6.946256284821307, + "grad_norm": 0.8208369612693787, + "learning_rate": 0.0001317179643973366, + "loss": 3.4476, + "step": 102235 + }, + { + "epoch": 6.946596004891969, + "grad_norm": 0.980745792388916, + "learning_rate": 0.00013167549938850387, + "loss": 3.5657, + "step": 102240 + }, + { + "epoch": 6.946935724962631, + "grad_norm": 0.8236788511276245, + "learning_rate": 0.00013163303437967115, + "loss": 3.2576, + "step": 102245 + }, + { + "epoch": 6.947275445033292, + "grad_norm": 0.793920636177063, + "learning_rate": 0.00013159056937083843, + "loss": 3.6493, + "step": 102250 + }, + { + "epoch": 6.947615165103954, + "grad_norm": 0.8412822484970093, + "learning_rate": 0.0001315481043620057, + "loss": 3.5855, + "step": 102255 + }, + { + "epoch": 6.9479548851746165, + "grad_norm": 0.8216887712478638, + "learning_rate": 0.000131505639353173, + "loss": 3.431, + "step": 102260 + }, + { + "epoch": 6.948294605245278, + "grad_norm": 0.8421458601951599, + "learning_rate": 0.00013146317434434027, + "loss": 3.4566, + "step": 102265 + }, + { + "epoch": 6.94863432531594, + "grad_norm": 1.0956226587295532, + "learning_rate": 0.00013142070933550753, + "loss": 3.2404, + "step": 102270 + }, + { + "epoch": 6.948974045386602, + "grad_norm": 1.0827431678771973, + "learning_rate": 0.00013137824432667483, + "loss": 3.3203, + "step": 102275 + }, + { + "epoch": 6.949313765457263, + "grad_norm": 0.7419415712356567, + "learning_rate": 0.0001313357793178421, + "loss": 3.4353, + "step": 102280 + }, + { + "epoch": 6.949653485527925, + "grad_norm": 0.9260217547416687, + "learning_rate": 0.00013129331430900937, + "loss": 3.2728, + "step": 102285 + }, + { + "epoch": 6.949993205598587, + "grad_norm": 0.9497851133346558, + "learning_rate": 0.00013125084930017667, + "loss": 3.3722, + "step": 102290 + }, + { + "epoch": 6.950332925669248, + "grad_norm": 0.9470478892326355, + "learning_rate": 0.00013120838429134393, + "loss": 3.4107, + "step": 102295 + }, + { + "epoch": 6.95067264573991, + "grad_norm": 1.0524067878723145, + "learning_rate": 0.0001311659192825112, + "loss": 3.2654, + "step": 102300 + }, + { + "epoch": 6.9510123658105725, + "grad_norm": 0.8398501873016357, + "learning_rate": 0.0001311234542736785, + "loss": 3.5241, + "step": 102305 + }, + { + "epoch": 6.951352085881234, + "grad_norm": 1.9627203941345215, + "learning_rate": 0.00013108098926484577, + "loss": 3.524, + "step": 102310 + }, + { + "epoch": 6.951691805951896, + "grad_norm": 0.9312080144882202, + "learning_rate": 0.00013103852425601305, + "loss": 3.2839, + "step": 102315 + }, + { + "epoch": 6.952031526022557, + "grad_norm": 0.7939547896385193, + "learning_rate": 0.00013099605924718033, + "loss": 3.4548, + "step": 102320 + }, + { + "epoch": 6.952371246093219, + "grad_norm": 0.835465669631958, + "learning_rate": 0.0001309535942383476, + "loss": 3.3601, + "step": 102325 + }, + { + "epoch": 6.952710966163881, + "grad_norm": 0.9143009781837463, + "learning_rate": 0.00013091112922951486, + "loss": 3.5389, + "step": 102330 + }, + { + "epoch": 6.953050686234542, + "grad_norm": 0.8663129806518555, + "learning_rate": 0.00013086866422068217, + "loss": 3.3816, + "step": 102335 + }, + { + "epoch": 6.953390406305204, + "grad_norm": 1.064369797706604, + "learning_rate": 0.00013082619921184945, + "loss": 3.4104, + "step": 102340 + }, + { + "epoch": 6.953730126375866, + "grad_norm": 0.7299907803535461, + "learning_rate": 0.00013078373420301673, + "loss": 3.7263, + "step": 102345 + }, + { + "epoch": 6.954069846446528, + "grad_norm": 0.8585992455482483, + "learning_rate": 0.000130741269194184, + "loss": 3.69, + "step": 102350 + }, + { + "epoch": 6.95440956651719, + "grad_norm": 0.8451579213142395, + "learning_rate": 0.00013069880418535126, + "loss": 3.3807, + "step": 102355 + }, + { + "epoch": 6.954749286587852, + "grad_norm": 0.8157724142074585, + "learning_rate": 0.00013065633917651857, + "loss": 3.1407, + "step": 102360 + }, + { + "epoch": 6.955089006658513, + "grad_norm": 1.1549755334854126, + "learning_rate": 0.00013061387416768582, + "loss": 3.5116, + "step": 102365 + }, + { + "epoch": 6.955428726729175, + "grad_norm": 0.7800760865211487, + "learning_rate": 0.0001305714091588531, + "loss": 3.4991, + "step": 102370 + }, + { + "epoch": 6.955768446799837, + "grad_norm": 1.1157845258712769, + "learning_rate": 0.0001305289441500204, + "loss": 3.5242, + "step": 102375 + }, + { + "epoch": 6.956108166870498, + "grad_norm": 0.9433785080909729, + "learning_rate": 0.00013048647914118766, + "loss": 3.5078, + "step": 102380 + }, + { + "epoch": 6.95644788694116, + "grad_norm": 0.843789279460907, + "learning_rate": 0.00013044401413235494, + "loss": 3.4779, + "step": 102385 + }, + { + "epoch": 6.956787607011822, + "grad_norm": 0.9031049609184265, + "learning_rate": 0.00013040154912352222, + "loss": 3.4307, + "step": 102390 + }, + { + "epoch": 6.957127327082484, + "grad_norm": 0.8560187816619873, + "learning_rate": 0.0001303590841146895, + "loss": 3.0535, + "step": 102395 + }, + { + "epoch": 6.957467047153146, + "grad_norm": 1.0746128559112549, + "learning_rate": 0.00013031661910585675, + "loss": 3.5222, + "step": 102400 + }, + { + "epoch": 6.957806767223808, + "grad_norm": 0.816068708896637, + "learning_rate": 0.00013027415409702406, + "loss": 3.5894, + "step": 102405 + }, + { + "epoch": 6.958146487294469, + "grad_norm": 0.8742560148239136, + "learning_rate": 0.00013023168908819134, + "loss": 3.4061, + "step": 102410 + }, + { + "epoch": 6.958486207365131, + "grad_norm": 0.7056935429573059, + "learning_rate": 0.0001301892240793586, + "loss": 3.113, + "step": 102415 + }, + { + "epoch": 6.958825927435793, + "grad_norm": 0.7573054432868958, + "learning_rate": 0.0001301467590705259, + "loss": 3.3517, + "step": 102420 + }, + { + "epoch": 6.959165647506454, + "grad_norm": 0.8779467344284058, + "learning_rate": 0.00013010429406169315, + "loss": 3.5093, + "step": 102425 + }, + { + "epoch": 6.959505367577116, + "grad_norm": 0.943389892578125, + "learning_rate": 0.00013006182905286046, + "loss": 3.4692, + "step": 102430 + }, + { + "epoch": 6.959845087647778, + "grad_norm": 0.9890621304512024, + "learning_rate": 0.00013001936404402771, + "loss": 3.5019, + "step": 102435 + }, + { + "epoch": 6.96018480771844, + "grad_norm": 0.819722056388855, + "learning_rate": 0.000129976899035195, + "loss": 3.5175, + "step": 102440 + }, + { + "epoch": 6.960524527789102, + "grad_norm": 0.9415423274040222, + "learning_rate": 0.0001299344340263623, + "loss": 3.5077, + "step": 102445 + }, + { + "epoch": 6.960864247859764, + "grad_norm": 0.9733858704566956, + "learning_rate": 0.00012989196901752955, + "loss": 3.2583, + "step": 102450 + }, + { + "epoch": 6.961203967930425, + "grad_norm": 1.0280941724777222, + "learning_rate": 0.00012984950400869683, + "loss": 3.1945, + "step": 102455 + }, + { + "epoch": 6.961543688001087, + "grad_norm": 0.7908445000648499, + "learning_rate": 0.00012980703899986411, + "loss": 3.5773, + "step": 102460 + }, + { + "epoch": 6.961883408071749, + "grad_norm": 0.8520046472549438, + "learning_rate": 0.0001297645739910314, + "loss": 3.4523, + "step": 102465 + }, + { + "epoch": 6.96222312814241, + "grad_norm": 0.7173457145690918, + "learning_rate": 0.00012972210898219865, + "loss": 3.0592, + "step": 102470 + }, + { + "epoch": 6.962562848213072, + "grad_norm": 0.7747524380683899, + "learning_rate": 0.00012967964397336595, + "loss": 3.09, + "step": 102475 + }, + { + "epoch": 6.962902568283734, + "grad_norm": 1.1490086317062378, + "learning_rate": 0.00012963717896453324, + "loss": 3.2303, + "step": 102480 + }, + { + "epoch": 6.963242288354396, + "grad_norm": 0.7199576497077942, + "learning_rate": 0.0001295947139557005, + "loss": 3.5491, + "step": 102485 + }, + { + "epoch": 6.963582008425058, + "grad_norm": 1.0642887353897095, + "learning_rate": 0.0001295522489468678, + "loss": 3.5224, + "step": 102490 + }, + { + "epoch": 6.96392172849572, + "grad_norm": 0.9238139390945435, + "learning_rate": 0.00012950978393803505, + "loss": 3.6165, + "step": 102495 + }, + { + "epoch": 6.964261448566381, + "grad_norm": 1.009762167930603, + "learning_rate": 0.00012946731892920233, + "loss": 3.2707, + "step": 102500 + }, + { + "epoch": 6.964601168637043, + "grad_norm": 1.0261813402175903, + "learning_rate": 0.00012942485392036964, + "loss": 3.4891, + "step": 102505 + }, + { + "epoch": 6.964940888707705, + "grad_norm": 0.7334873676300049, + "learning_rate": 0.0001293823889115369, + "loss": 3.2688, + "step": 102510 + }, + { + "epoch": 6.965280608778366, + "grad_norm": 0.9204325675964355, + "learning_rate": 0.0001293399239027042, + "loss": 3.478, + "step": 102515 + }, + { + "epoch": 6.965620328849028, + "grad_norm": 0.7518738508224487, + "learning_rate": 0.00012929745889387145, + "loss": 3.3094, + "step": 102520 + }, + { + "epoch": 6.9659600489196905, + "grad_norm": 1.0627405643463135, + "learning_rate": 0.00012925499388503873, + "loss": 3.5566, + "step": 102525 + }, + { + "epoch": 6.966299768990352, + "grad_norm": 0.9394830465316772, + "learning_rate": 0.000129212528876206, + "loss": 3.3346, + "step": 102530 + }, + { + "epoch": 6.966639489061014, + "grad_norm": 1.0437250137329102, + "learning_rate": 0.0001291700638673733, + "loss": 3.5751, + "step": 102535 + }, + { + "epoch": 6.966979209131676, + "grad_norm": 0.9459506273269653, + "learning_rate": 0.00012912759885854057, + "loss": 3.7384, + "step": 102540 + }, + { + "epoch": 6.967318929202337, + "grad_norm": 1.0009866952896118, + "learning_rate": 0.00012908513384970785, + "loss": 3.4886, + "step": 102545 + }, + { + "epoch": 6.967658649272999, + "grad_norm": 1.0905600786209106, + "learning_rate": 0.00012904266884087513, + "loss": 3.4618, + "step": 102550 + }, + { + "epoch": 6.967998369343661, + "grad_norm": 0.8235769271850586, + "learning_rate": 0.00012900020383204238, + "loss": 3.4753, + "step": 102555 + }, + { + "epoch": 6.968338089414322, + "grad_norm": 1.0313431024551392, + "learning_rate": 0.0001289577388232097, + "loss": 3.5497, + "step": 102560 + }, + { + "epoch": 6.968677809484984, + "grad_norm": 0.9170039892196655, + "learning_rate": 0.00012891527381437694, + "loss": 3.5289, + "step": 102565 + }, + { + "epoch": 6.9690175295556465, + "grad_norm": 1.111119031906128, + "learning_rate": 0.00012887280880554422, + "loss": 3.3369, + "step": 102570 + }, + { + "epoch": 6.969357249626308, + "grad_norm": 0.9277728796005249, + "learning_rate": 0.00012883034379671153, + "loss": 3.4034, + "step": 102575 + }, + { + "epoch": 6.96969696969697, + "grad_norm": 0.9360367655754089, + "learning_rate": 0.00012878787878787878, + "loss": 3.5583, + "step": 102580 + }, + { + "epoch": 6.970036689767632, + "grad_norm": 1.0234925746917725, + "learning_rate": 0.00012874541377904606, + "loss": 3.5129, + "step": 102585 + }, + { + "epoch": 6.970376409838293, + "grad_norm": 0.9386270046234131, + "learning_rate": 0.00012870294877021334, + "loss": 3.1105, + "step": 102590 + }, + { + "epoch": 6.970716129908955, + "grad_norm": 0.8614767789840698, + "learning_rate": 0.00012866048376138062, + "loss": 3.1819, + "step": 102595 + }, + { + "epoch": 6.971055849979617, + "grad_norm": 1.0908740758895874, + "learning_rate": 0.0001286180187525479, + "loss": 3.3576, + "step": 102600 + }, + { + "epoch": 6.971395570050278, + "grad_norm": 1.2297250032424927, + "learning_rate": 0.00012857555374371518, + "loss": 3.5847, + "step": 102605 + }, + { + "epoch": 6.97173529012094, + "grad_norm": 1.092238187789917, + "learning_rate": 0.00012853308873488246, + "loss": 3.1208, + "step": 102610 + }, + { + "epoch": 6.9720750101916025, + "grad_norm": 0.893398642539978, + "learning_rate": 0.00012849062372604974, + "loss": 3.2105, + "step": 102615 + }, + { + "epoch": 6.972414730262264, + "grad_norm": 0.8836607933044434, + "learning_rate": 0.00012844815871721702, + "loss": 3.6456, + "step": 102620 + }, + { + "epoch": 6.972754450332926, + "grad_norm": 0.9234555959701538, + "learning_rate": 0.00012840569370838428, + "loss": 3.3672, + "step": 102625 + }, + { + "epoch": 6.973094170403588, + "grad_norm": 0.9487478733062744, + "learning_rate": 0.00012836322869955158, + "loss": 3.4791, + "step": 102630 + }, + { + "epoch": 6.973433890474249, + "grad_norm": 1.6744810342788696, + "learning_rate": 0.00012832076369071886, + "loss": 3.2897, + "step": 102635 + }, + { + "epoch": 6.973773610544911, + "grad_norm": 0.9064832329750061, + "learning_rate": 0.00012827829868188612, + "loss": 3.4921, + "step": 102640 + }, + { + "epoch": 6.974113330615573, + "grad_norm": 0.9632606506347656, + "learning_rate": 0.00012823583367305342, + "loss": 3.3943, + "step": 102645 + }, + { + "epoch": 6.974453050686234, + "grad_norm": 0.8962404131889343, + "learning_rate": 0.00012819336866422068, + "loss": 3.3029, + "step": 102650 + }, + { + "epoch": 6.974792770756896, + "grad_norm": 1.146951675415039, + "learning_rate": 0.00012815090365538796, + "loss": 3.4617, + "step": 102655 + }, + { + "epoch": 6.9751324908275585, + "grad_norm": 1.0537232160568237, + "learning_rate": 0.00012810843864655524, + "loss": 2.945, + "step": 102660 + }, + { + "epoch": 6.97547221089822, + "grad_norm": 0.8755468130111694, + "learning_rate": 0.00012806597363772252, + "loss": 3.3411, + "step": 102665 + }, + { + "epoch": 6.975811930968882, + "grad_norm": 0.8069353699684143, + "learning_rate": 0.0001280235086288898, + "loss": 3.429, + "step": 102670 + }, + { + "epoch": 6.976151651039544, + "grad_norm": 0.9910539388656616, + "learning_rate": 0.00012798104362005708, + "loss": 3.385, + "step": 102675 + }, + { + "epoch": 6.976491371110205, + "grad_norm": 0.8963002562522888, + "learning_rate": 0.00012793857861122436, + "loss": 3.3129, + "step": 102680 + }, + { + "epoch": 6.976831091180867, + "grad_norm": 0.8275030255317688, + "learning_rate": 0.00012789611360239164, + "loss": 3.4989, + "step": 102685 + }, + { + "epoch": 6.977170811251529, + "grad_norm": 0.9225301146507263, + "learning_rate": 0.00012785364859355892, + "loss": 3.2094, + "step": 102690 + }, + { + "epoch": 6.97751053132219, + "grad_norm": 0.860299825668335, + "learning_rate": 0.00012781118358472617, + "loss": 3.4534, + "step": 102695 + }, + { + "epoch": 6.977850251392852, + "grad_norm": 0.885494589805603, + "learning_rate": 0.00012776871857589348, + "loss": 3.3481, + "step": 102700 + }, + { + "epoch": 6.9781899714635145, + "grad_norm": 0.803236186504364, + "learning_rate": 0.00012772625356706076, + "loss": 3.4684, + "step": 102705 + }, + { + "epoch": 6.978529691534176, + "grad_norm": 0.8542935252189636, + "learning_rate": 0.000127683788558228, + "loss": 3.2632, + "step": 102710 + }, + { + "epoch": 6.978869411604838, + "grad_norm": 0.9718786478042603, + "learning_rate": 0.00012764132354939532, + "loss": 3.3318, + "step": 102715 + }, + { + "epoch": 6.9792091316755, + "grad_norm": 1.0639750957489014, + "learning_rate": 0.00012759885854056257, + "loss": 3.2774, + "step": 102720 + }, + { + "epoch": 6.979548851746161, + "grad_norm": 0.8505985736846924, + "learning_rate": 0.00012755639353172985, + "loss": 3.4278, + "step": 102725 + }, + { + "epoch": 6.979888571816823, + "grad_norm": 0.7428311109542847, + "learning_rate": 0.00012751392852289713, + "loss": 3.5076, + "step": 102730 + }, + { + "epoch": 6.980228291887485, + "grad_norm": 0.924034059047699, + "learning_rate": 0.0001274714635140644, + "loss": 3.3913, + "step": 102735 + }, + { + "epoch": 6.980568011958146, + "grad_norm": 1.003416895866394, + "learning_rate": 0.0001274289985052317, + "loss": 3.3488, + "step": 102740 + }, + { + "epoch": 6.980907732028808, + "grad_norm": 0.9170259833335876, + "learning_rate": 0.00012738653349639897, + "loss": 3.5946, + "step": 102745 + }, + { + "epoch": 6.9812474520994705, + "grad_norm": 0.9121410846710205, + "learning_rate": 0.00012734406848756625, + "loss": 3.4791, + "step": 102750 + }, + { + "epoch": 6.981587172170132, + "grad_norm": 1.0027610063552856, + "learning_rate": 0.0001273016034787335, + "loss": 3.604, + "step": 102755 + }, + { + "epoch": 6.981926892240794, + "grad_norm": 0.9977214932441711, + "learning_rate": 0.0001272591384699008, + "loss": 3.2275, + "step": 102760 + }, + { + "epoch": 6.982266612311456, + "grad_norm": 1.709061861038208, + "learning_rate": 0.0001272166734610681, + "loss": 3.4359, + "step": 102765 + }, + { + "epoch": 6.982606332382117, + "grad_norm": 0.800990879535675, + "learning_rate": 0.00012717420845223537, + "loss": 3.5482, + "step": 102770 + }, + { + "epoch": 6.982946052452779, + "grad_norm": 1.0715665817260742, + "learning_rate": 0.00012713174344340265, + "loss": 3.4258, + "step": 102775 + }, + { + "epoch": 6.983285772523441, + "grad_norm": 0.8524954319000244, + "learning_rate": 0.0001270892784345699, + "loss": 3.3739, + "step": 102780 + }, + { + "epoch": 6.983625492594102, + "grad_norm": 0.6482035517692566, + "learning_rate": 0.0001270468134257372, + "loss": 3.369, + "step": 102785 + }, + { + "epoch": 6.983965212664764, + "grad_norm": 1.0151177644729614, + "learning_rate": 0.00012700434841690446, + "loss": 3.2267, + "step": 102790 + }, + { + "epoch": 6.984304932735426, + "grad_norm": 0.941775381565094, + "learning_rate": 0.00012696188340807174, + "loss": 3.4762, + "step": 102795 + }, + { + "epoch": 6.984644652806088, + "grad_norm": 0.8015018105506897, + "learning_rate": 0.00012691941839923905, + "loss": 3.5169, + "step": 102800 + }, + { + "epoch": 6.98498437287675, + "grad_norm": 0.9029374718666077, + "learning_rate": 0.0001268769533904063, + "loss": 3.4682, + "step": 102805 + }, + { + "epoch": 6.985324092947411, + "grad_norm": 0.6554465293884277, + "learning_rate": 0.00012683448838157358, + "loss": 3.5906, + "step": 102810 + }, + { + "epoch": 6.985663813018073, + "grad_norm": 0.8859093189239502, + "learning_rate": 0.00012679202337274086, + "loss": 3.4186, + "step": 102815 + }, + { + "epoch": 6.986003533088735, + "grad_norm": 0.8486411571502686, + "learning_rate": 0.00012674955836390814, + "loss": 3.4465, + "step": 102820 + }, + { + "epoch": 6.986343253159396, + "grad_norm": 0.9382777810096741, + "learning_rate": 0.0001267070933550754, + "loss": 3.5192, + "step": 102825 + }, + { + "epoch": 6.986682973230058, + "grad_norm": 1.1505157947540283, + "learning_rate": 0.0001266646283462427, + "loss": 3.6039, + "step": 102830 + }, + { + "epoch": 6.9870226933007205, + "grad_norm": 1.16360604763031, + "learning_rate": 0.00012662216333740998, + "loss": 3.4998, + "step": 102835 + }, + { + "epoch": 6.987362413371382, + "grad_norm": 1.0355188846588135, + "learning_rate": 0.00012657969832857724, + "loss": 3.5213, + "step": 102840 + }, + { + "epoch": 6.987702133442044, + "grad_norm": 0.9478228092193604, + "learning_rate": 0.00012653723331974455, + "loss": 3.4591, + "step": 102845 + }, + { + "epoch": 6.988041853512706, + "grad_norm": 0.7454301118850708, + "learning_rate": 0.0001264947683109118, + "loss": 3.1834, + "step": 102850 + }, + { + "epoch": 6.988381573583367, + "grad_norm": 1.2105649709701538, + "learning_rate": 0.0001264523033020791, + "loss": 3.0606, + "step": 102855 + }, + { + "epoch": 6.988721293654029, + "grad_norm": 0.9867724180221558, + "learning_rate": 0.00012640983829324636, + "loss": 3.4605, + "step": 102860 + }, + { + "epoch": 6.989061013724691, + "grad_norm": 0.8359292149543762, + "learning_rate": 0.00012636737328441364, + "loss": 3.2267, + "step": 102865 + }, + { + "epoch": 6.989400733795352, + "grad_norm": 0.8524578809738159, + "learning_rate": 0.00012632490827558095, + "loss": 3.3104, + "step": 102870 + }, + { + "epoch": 6.989740453866014, + "grad_norm": 1.0059906244277954, + "learning_rate": 0.0001262824432667482, + "loss": 3.1913, + "step": 102875 + }, + { + "epoch": 6.9900801739366765, + "grad_norm": 0.619723916053772, + "learning_rate": 0.00012623997825791548, + "loss": 3.453, + "step": 102880 + }, + { + "epoch": 6.990419894007338, + "grad_norm": 0.9777804017066956, + "learning_rate": 0.00012619751324908276, + "loss": 3.4998, + "step": 102885 + }, + { + "epoch": 6.990759614078, + "grad_norm": 0.7802236080169678, + "learning_rate": 0.00012615504824025004, + "loss": 3.2834, + "step": 102890 + }, + { + "epoch": 6.991099334148662, + "grad_norm": 0.88691645860672, + "learning_rate": 0.0001261125832314173, + "loss": 3.1755, + "step": 102895 + }, + { + "epoch": 6.991439054219323, + "grad_norm": 0.9845026135444641, + "learning_rate": 0.0001260701182225846, + "loss": 3.4912, + "step": 102900 + }, + { + "epoch": 6.991778774289985, + "grad_norm": 1.0896276235580444, + "learning_rate": 0.00012602765321375188, + "loss": 3.3654, + "step": 102905 + }, + { + "epoch": 6.992118494360647, + "grad_norm": 0.9694443345069885, + "learning_rate": 0.00012598518820491913, + "loss": 3.7589, + "step": 102910 + }, + { + "epoch": 6.992458214431308, + "grad_norm": 1.1840616464614868, + "learning_rate": 0.00012594272319608644, + "loss": 3.6538, + "step": 102915 + }, + { + "epoch": 6.99279793450197, + "grad_norm": 0.9263381361961365, + "learning_rate": 0.0001259002581872537, + "loss": 3.4547, + "step": 102920 + }, + { + "epoch": 6.9931376545726325, + "grad_norm": 1.4067987203598022, + "learning_rate": 0.00012585779317842097, + "loss": 3.6237, + "step": 102925 + }, + { + "epoch": 6.993477374643294, + "grad_norm": 0.8740122318267822, + "learning_rate": 0.00012581532816958828, + "loss": 3.6823, + "step": 102930 + }, + { + "epoch": 6.993817094713956, + "grad_norm": 0.8420600891113281, + "learning_rate": 0.00012577286316075553, + "loss": 3.441, + "step": 102935 + }, + { + "epoch": 6.994156814784618, + "grad_norm": 0.8893803358078003, + "learning_rate": 0.00012573039815192284, + "loss": 3.3688, + "step": 102940 + }, + { + "epoch": 6.994496534855279, + "grad_norm": 0.9676928520202637, + "learning_rate": 0.0001256879331430901, + "loss": 3.6317, + "step": 102945 + }, + { + "epoch": 6.994836254925941, + "grad_norm": 1.0802574157714844, + "learning_rate": 0.00012564546813425737, + "loss": 3.3976, + "step": 102950 + }, + { + "epoch": 6.995175974996603, + "grad_norm": 0.7867299914360046, + "learning_rate": 0.00012560300312542465, + "loss": 3.3772, + "step": 102955 + }, + { + "epoch": 6.995515695067264, + "grad_norm": 0.8625629544258118, + "learning_rate": 0.00012556053811659193, + "loss": 3.523, + "step": 102960 + }, + { + "epoch": 6.995855415137926, + "grad_norm": 0.8038132190704346, + "learning_rate": 0.0001255180731077592, + "loss": 3.3026, + "step": 102965 + }, + { + "epoch": 6.9961951352085885, + "grad_norm": 0.8535280227661133, + "learning_rate": 0.0001254756080989265, + "loss": 3.5016, + "step": 102970 + }, + { + "epoch": 6.99653485527925, + "grad_norm": 0.8050394654273987, + "learning_rate": 0.00012543314309009377, + "loss": 3.5294, + "step": 102975 + }, + { + "epoch": 6.996874575349912, + "grad_norm": 0.7681966423988342, + "learning_rate": 0.00012539067808126103, + "loss": 3.5675, + "step": 102980 + }, + { + "epoch": 6.997214295420574, + "grad_norm": 0.9722830057144165, + "learning_rate": 0.00012534821307242833, + "loss": 3.6076, + "step": 102985 + }, + { + "epoch": 6.997554015491235, + "grad_norm": 0.7606921195983887, + "learning_rate": 0.00012530574806359559, + "loss": 3.4241, + "step": 102990 + }, + { + "epoch": 6.997893735561897, + "grad_norm": 2.495673894882202, + "learning_rate": 0.00012526328305476287, + "loss": 3.358, + "step": 102995 + }, + { + "epoch": 6.998233455632558, + "grad_norm": 1.3348685503005981, + "learning_rate": 0.00012522081804593017, + "loss": 3.51, + "step": 103000 + }, + { + "epoch": 6.99857317570322, + "grad_norm": 1.0126925706863403, + "learning_rate": 0.00012517835303709743, + "loss": 3.3994, + "step": 103005 + }, + { + "epoch": 6.998912895773882, + "grad_norm": 0.9949975609779358, + "learning_rate": 0.0001251358880282647, + "loss": 3.3387, + "step": 103010 + }, + { + "epoch": 6.999252615844544, + "grad_norm": 0.9600359201431274, + "learning_rate": 0.00012509342301943199, + "loss": 3.5088, + "step": 103015 + }, + { + "epoch": 6.999592335915206, + "grad_norm": 0.9191299676895142, + "learning_rate": 0.00012505095801059927, + "loss": 3.5832, + "step": 103020 + }, + { + "epoch": 6.999932055985868, + "grad_norm": 0.9539197087287903, + "learning_rate": 0.00012500849300176655, + "loss": 3.1945, + "step": 103025 + }, + { + "epoch": 7.0, + "eval_bertscore": { + "f1": 0.8445112908668603, + "precision": 0.8503523721506873, + "recall": 0.8394631683762909 + }, + "eval_bleu_4": 0.01492047203724886, + "eval_exact_match": 0.00019381723035177828, + "eval_loss": 3.390328884124756, + "eval_meteor": 0.08785069236010327, + "eval_rouge": { + "rouge1": 0.12909738199977155, + "rouge2": 0.018717240052289703, + "rougeL": 0.11122914498469796, + "rougeLsum": 0.11133224799394678 + }, + "eval_runtime": 2137.4271, + "eval_samples_per_second": 4.828, + "eval_steps_per_second": 0.604, + "step": 103026 + }, + { + "epoch": 7.00027177605653, + "grad_norm": 0.7556595802307129, + "learning_rate": 0.00012496602799293383, + "loss": 3.4095, + "step": 103030 + }, + { + "epoch": 7.000611496127191, + "grad_norm": 1.3806049823760986, + "learning_rate": 0.0001249235629841011, + "loss": 3.4094, + "step": 103035 + }, + { + "epoch": 7.000951216197853, + "grad_norm": 0.977888286113739, + "learning_rate": 0.0001248810979752684, + "loss": 3.4033, + "step": 103040 + }, + { + "epoch": 7.001290936268515, + "grad_norm": 1.075073480606079, + "learning_rate": 0.00012483863296643567, + "loss": 3.3974, + "step": 103045 + }, + { + "epoch": 7.001630656339176, + "grad_norm": 0.9881688356399536, + "learning_rate": 0.00012479616795760295, + "loss": 3.4537, + "step": 103050 + }, + { + "epoch": 7.001970376409838, + "grad_norm": 1.047606110572815, + "learning_rate": 0.0001247537029487702, + "loss": 3.4902, + "step": 103055 + }, + { + "epoch": 7.0023100964805, + "grad_norm": 0.9880174398422241, + "learning_rate": 0.0001247112379399375, + "loss": 3.3894, + "step": 103060 + }, + { + "epoch": 7.002649816551162, + "grad_norm": 1.0347806215286255, + "learning_rate": 0.0001246687729311048, + "loss": 3.5743, + "step": 103065 + }, + { + "epoch": 7.002989536621824, + "grad_norm": 1.0680855512619019, + "learning_rate": 0.00012462630792227204, + "loss": 3.1617, + "step": 103070 + }, + { + "epoch": 7.003329256692485, + "grad_norm": 1.0029562711715698, + "learning_rate": 0.00012458384291343932, + "loss": 3.3026, + "step": 103075 + }, + { + "epoch": 7.003668976763147, + "grad_norm": 0.9986989498138428, + "learning_rate": 0.0001245413779046066, + "loss": 3.0831, + "step": 103080 + }, + { + "epoch": 7.004008696833809, + "grad_norm": 1.0677562952041626, + "learning_rate": 0.00012449891289577388, + "loss": 3.5567, + "step": 103085 + }, + { + "epoch": 7.00434841690447, + "grad_norm": 1.201361060142517, + "learning_rate": 0.00012445644788694116, + "loss": 3.3887, + "step": 103090 + }, + { + "epoch": 7.004688136975132, + "grad_norm": 0.7751690149307251, + "learning_rate": 0.00012441398287810844, + "loss": 3.0715, + "step": 103095 + }, + { + "epoch": 7.0050278570457944, + "grad_norm": 0.9921829700469971, + "learning_rate": 0.00012437151786927572, + "loss": 3.4934, + "step": 103100 + }, + { + "epoch": 7.005367577116456, + "grad_norm": 0.8719357252120972, + "learning_rate": 0.000124329052860443, + "loss": 3.4044, + "step": 103105 + }, + { + "epoch": 7.005707297187118, + "grad_norm": 1.0200345516204834, + "learning_rate": 0.00012428658785161028, + "loss": 3.3336, + "step": 103110 + }, + { + "epoch": 7.00604701725778, + "grad_norm": 1.1511924266815186, + "learning_rate": 0.00012424412284277756, + "loss": 3.4612, + "step": 103115 + }, + { + "epoch": 7.006386737328441, + "grad_norm": 1.110174298286438, + "learning_rate": 0.00012420165783394481, + "loss": 3.3393, + "step": 103120 + }, + { + "epoch": 7.006726457399103, + "grad_norm": 0.8661626577377319, + "learning_rate": 0.00012415919282511212, + "loss": 3.2489, + "step": 103125 + }, + { + "epoch": 7.007066177469765, + "grad_norm": 1.117165446281433, + "learning_rate": 0.0001241167278162794, + "loss": 3.5899, + "step": 103130 + }, + { + "epoch": 7.007405897540426, + "grad_norm": 0.8223596811294556, + "learning_rate": 0.00012407426280744668, + "loss": 3.383, + "step": 103135 + }, + { + "epoch": 7.007745617611088, + "grad_norm": 0.9064943790435791, + "learning_rate": 0.00012403179779861393, + "loss": 3.33, + "step": 103140 + }, + { + "epoch": 7.0080853376817505, + "grad_norm": 1.2323074340820312, + "learning_rate": 0.00012398933278978121, + "loss": 3.4654, + "step": 103145 + }, + { + "epoch": 7.008425057752412, + "grad_norm": 0.9427741765975952, + "learning_rate": 0.0001239468677809485, + "loss": 3.3216, + "step": 103150 + }, + { + "epoch": 7.008764777823074, + "grad_norm": 0.824262261390686, + "learning_rate": 0.00012390440277211577, + "loss": 3.4009, + "step": 103155 + }, + { + "epoch": 7.009104497893736, + "grad_norm": 0.9519448280334473, + "learning_rate": 0.00012386193776328305, + "loss": 3.3466, + "step": 103160 + }, + { + "epoch": 7.009444217964397, + "grad_norm": 0.7327316403388977, + "learning_rate": 0.00012381947275445033, + "loss": 3.3756, + "step": 103165 + }, + { + "epoch": 7.009783938035059, + "grad_norm": 0.8212221264839172, + "learning_rate": 0.00012377700774561761, + "loss": 3.4383, + "step": 103170 + }, + { + "epoch": 7.010123658105721, + "grad_norm": 0.8465712070465088, + "learning_rate": 0.0001237345427367849, + "loss": 2.9339, + "step": 103175 + }, + { + "epoch": 7.010463378176382, + "grad_norm": 0.9735875725746155, + "learning_rate": 0.00012369207772795217, + "loss": 3.6412, + "step": 103180 + }, + { + "epoch": 7.010803098247044, + "grad_norm": 0.9648188948631287, + "learning_rate": 0.00012364961271911945, + "loss": 3.3042, + "step": 103185 + }, + { + "epoch": 7.0111428183177065, + "grad_norm": 1.0608359575271606, + "learning_rate": 0.00012360714771028673, + "loss": 3.4719, + "step": 103190 + }, + { + "epoch": 7.011482538388368, + "grad_norm": 0.7964227795600891, + "learning_rate": 0.00012356468270145402, + "loss": 3.3493, + "step": 103195 + }, + { + "epoch": 7.01182225845903, + "grad_norm": 0.851063072681427, + "learning_rate": 0.0001235222176926213, + "loss": 3.6049, + "step": 103200 + }, + { + "epoch": 7.012161978529692, + "grad_norm": 0.8426675796508789, + "learning_rate": 0.00012347975268378855, + "loss": 3.5, + "step": 103205 + }, + { + "epoch": 7.012501698600353, + "grad_norm": 0.7329621315002441, + "learning_rate": 0.00012343728767495583, + "loss": 3.3275, + "step": 103210 + }, + { + "epoch": 7.012841418671015, + "grad_norm": 0.8828045725822449, + "learning_rate": 0.0001233948226661231, + "loss": 3.3943, + "step": 103215 + }, + { + "epoch": 7.013181138741677, + "grad_norm": 0.861730694770813, + "learning_rate": 0.0001233523576572904, + "loss": 3.1515, + "step": 103220 + }, + { + "epoch": 7.013520858812338, + "grad_norm": 0.7020594477653503, + "learning_rate": 0.00012330989264845767, + "loss": 3.4758, + "step": 103225 + }, + { + "epoch": 7.013860578883, + "grad_norm": 1.0341037511825562, + "learning_rate": 0.00012326742763962495, + "loss": 3.35, + "step": 103230 + }, + { + "epoch": 7.0142002989536625, + "grad_norm": 0.8958227634429932, + "learning_rate": 0.00012322496263079223, + "loss": 3.4228, + "step": 103235 + }, + { + "epoch": 7.014540019024324, + "grad_norm": 0.8272492289543152, + "learning_rate": 0.0001231824976219595, + "loss": 3.3523, + "step": 103240 + }, + { + "epoch": 7.014879739094986, + "grad_norm": 1.0432785749435425, + "learning_rate": 0.0001231400326131268, + "loss": 3.5042, + "step": 103245 + }, + { + "epoch": 7.015219459165648, + "grad_norm": 0.8368958234786987, + "learning_rate": 0.00012309756760429407, + "loss": 3.3665, + "step": 103250 + }, + { + "epoch": 7.015559179236309, + "grad_norm": 0.9955891370773315, + "learning_rate": 0.00012305510259546135, + "loss": 3.4155, + "step": 103255 + }, + { + "epoch": 7.015898899306971, + "grad_norm": 0.9835084080696106, + "learning_rate": 0.00012301263758662863, + "loss": 3.6416, + "step": 103260 + }, + { + "epoch": 7.016238619377633, + "grad_norm": 0.8728025555610657, + "learning_rate": 0.0001229701725777959, + "loss": 3.4401, + "step": 103265 + }, + { + "epoch": 7.016578339448294, + "grad_norm": 0.8703916072845459, + "learning_rate": 0.0001229277075689632, + "loss": 3.3615, + "step": 103270 + }, + { + "epoch": 7.016918059518956, + "grad_norm": 0.9604029655456543, + "learning_rate": 0.00012288524256013044, + "loss": 3.3165, + "step": 103275 + }, + { + "epoch": 7.0172577795896185, + "grad_norm": 1.1020430326461792, + "learning_rate": 0.00012284277755129772, + "loss": 3.3729, + "step": 103280 + }, + { + "epoch": 7.01759749966028, + "grad_norm": 0.9676579236984253, + "learning_rate": 0.000122800312542465, + "loss": 3.265, + "step": 103285 + }, + { + "epoch": 7.017937219730942, + "grad_norm": 1.0049479007720947, + "learning_rate": 0.00012275784753363228, + "loss": 3.4752, + "step": 103290 + }, + { + "epoch": 7.018276939801604, + "grad_norm": 0.8005048036575317, + "learning_rate": 0.00012271538252479956, + "loss": 3.0751, + "step": 103295 + }, + { + "epoch": 7.018616659872265, + "grad_norm": 0.7989321947097778, + "learning_rate": 0.00012267291751596684, + "loss": 3.1419, + "step": 103300 + }, + { + "epoch": 7.018956379942927, + "grad_norm": 0.9743874669075012, + "learning_rate": 0.00012263045250713412, + "loss": 3.5505, + "step": 103305 + }, + { + "epoch": 7.019296100013589, + "grad_norm": 1.5062108039855957, + "learning_rate": 0.0001225879874983014, + "loss": 3.1706, + "step": 103310 + }, + { + "epoch": 7.01963582008425, + "grad_norm": 0.8398798108100891, + "learning_rate": 0.00012254552248946868, + "loss": 3.0668, + "step": 103315 + }, + { + "epoch": 7.019975540154912, + "grad_norm": 0.8939694166183472, + "learning_rate": 0.00012250305748063596, + "loss": 3.2286, + "step": 103320 + }, + { + "epoch": 7.0203152602255745, + "grad_norm": 1.0407171249389648, + "learning_rate": 0.00012246059247180324, + "loss": 3.6475, + "step": 103325 + }, + { + "epoch": 7.020654980296236, + "grad_norm": 1.0761795043945312, + "learning_rate": 0.00012241812746297052, + "loss": 3.6461, + "step": 103330 + }, + { + "epoch": 7.020994700366898, + "grad_norm": 1.1723111867904663, + "learning_rate": 0.0001223756624541378, + "loss": 3.4083, + "step": 103335 + }, + { + "epoch": 7.02133442043756, + "grad_norm": 0.9013713598251343, + "learning_rate": 0.00012233319744530508, + "loss": 3.4645, + "step": 103340 + }, + { + "epoch": 7.021674140508221, + "grad_norm": 0.883277952671051, + "learning_rate": 0.00012229073243647234, + "loss": 3.3477, + "step": 103345 + }, + { + "epoch": 7.022013860578883, + "grad_norm": 1.0849454402923584, + "learning_rate": 0.00012224826742763962, + "loss": 3.5066, + "step": 103350 + }, + { + "epoch": 7.022353580649545, + "grad_norm": 0.9670206308364868, + "learning_rate": 0.00012220580241880692, + "loss": 3.2926, + "step": 103355 + }, + { + "epoch": 7.022693300720206, + "grad_norm": 0.8435760140419006, + "learning_rate": 0.00012216333740997418, + "loss": 3.4668, + "step": 103360 + }, + { + "epoch": 7.023033020790868, + "grad_norm": 0.756401002407074, + "learning_rate": 0.00012212087240114146, + "loss": 3.2092, + "step": 103365 + }, + { + "epoch": 7.0233727408615305, + "grad_norm": 0.8724541664123535, + "learning_rate": 0.00012207840739230874, + "loss": 3.4899, + "step": 103370 + }, + { + "epoch": 7.023712460932192, + "grad_norm": 0.8192548751831055, + "learning_rate": 0.00012203594238347602, + "loss": 3.1674, + "step": 103375 + }, + { + "epoch": 7.024052181002854, + "grad_norm": 0.948924720287323, + "learning_rate": 0.0001219934773746433, + "loss": 3.3088, + "step": 103380 + }, + { + "epoch": 7.024391901073516, + "grad_norm": 1.0006436109542847, + "learning_rate": 0.00012195101236581058, + "loss": 3.5255, + "step": 103385 + }, + { + "epoch": 7.024731621144177, + "grad_norm": 0.780331552028656, + "learning_rate": 0.00012190854735697786, + "loss": 3.3321, + "step": 103390 + }, + { + "epoch": 7.025071341214839, + "grad_norm": 0.9306240081787109, + "learning_rate": 0.00012186608234814512, + "loss": 3.17, + "step": 103395 + }, + { + "epoch": 7.025411061285501, + "grad_norm": 0.9601434469223022, + "learning_rate": 0.0001218236173393124, + "loss": 3.3745, + "step": 103400 + }, + { + "epoch": 7.025750781356162, + "grad_norm": 0.7627377510070801, + "learning_rate": 0.0001217811523304797, + "loss": 3.4321, + "step": 103405 + }, + { + "epoch": 7.0260905014268245, + "grad_norm": 1.0485107898712158, + "learning_rate": 0.00012173868732164696, + "loss": 3.5169, + "step": 103410 + }, + { + "epoch": 7.026430221497486, + "grad_norm": 0.8586307764053345, + "learning_rate": 0.00012169622231281424, + "loss": 3.4532, + "step": 103415 + }, + { + "epoch": 7.026769941568148, + "grad_norm": 0.8529317378997803, + "learning_rate": 0.00012165375730398152, + "loss": 3.4455, + "step": 103420 + }, + { + "epoch": 7.02710966163881, + "grad_norm": 0.8717153072357178, + "learning_rate": 0.0001216112922951488, + "loss": 3.2829, + "step": 103425 + }, + { + "epoch": 7.027449381709471, + "grad_norm": 0.9729499816894531, + "learning_rate": 0.00012156882728631607, + "loss": 3.4763, + "step": 103430 + }, + { + "epoch": 7.027789101780133, + "grad_norm": 0.9755983948707581, + "learning_rate": 0.00012152636227748336, + "loss": 3.5203, + "step": 103435 + }, + { + "epoch": 7.028128821850795, + "grad_norm": 1.2274638414382935, + "learning_rate": 0.00012148389726865064, + "loss": 3.4449, + "step": 103440 + }, + { + "epoch": 7.028468541921456, + "grad_norm": 0.8213875889778137, + "learning_rate": 0.00012144143225981791, + "loss": 3.5323, + "step": 103445 + }, + { + "epoch": 7.028808261992118, + "grad_norm": 1.0100123882293701, + "learning_rate": 0.00012139896725098519, + "loss": 3.2406, + "step": 103450 + }, + { + "epoch": 7.0291479820627805, + "grad_norm": 1.0028756856918335, + "learning_rate": 0.00012135650224215247, + "loss": 3.2994, + "step": 103455 + }, + { + "epoch": 7.029487702133442, + "grad_norm": 1.0029258728027344, + "learning_rate": 0.00012131403723331974, + "loss": 3.5083, + "step": 103460 + }, + { + "epoch": 7.029827422204104, + "grad_norm": 1.0867985486984253, + "learning_rate": 0.00012127157222448702, + "loss": 3.5496, + "step": 103465 + }, + { + "epoch": 7.030167142274766, + "grad_norm": 0.8058934211730957, + "learning_rate": 0.00012122910721565431, + "loss": 3.4486, + "step": 103470 + }, + { + "epoch": 7.030506862345427, + "grad_norm": 0.7974064350128174, + "learning_rate": 0.00012118664220682159, + "loss": 3.2032, + "step": 103475 + }, + { + "epoch": 7.030846582416089, + "grad_norm": 1.3522186279296875, + "learning_rate": 0.00012114417719798886, + "loss": 3.3695, + "step": 103480 + }, + { + "epoch": 7.031186302486751, + "grad_norm": 1.0515469312667847, + "learning_rate": 0.00012110171218915614, + "loss": 3.3958, + "step": 103485 + }, + { + "epoch": 7.031526022557412, + "grad_norm": 1.367719054222107, + "learning_rate": 0.00012105924718032342, + "loss": 3.2879, + "step": 103490 + }, + { + "epoch": 7.031865742628074, + "grad_norm": 1.0373170375823975, + "learning_rate": 0.00012101678217149068, + "loss": 3.3068, + "step": 103495 + }, + { + "epoch": 7.0322054626987365, + "grad_norm": 0.9308684468269348, + "learning_rate": 0.00012097431716265796, + "loss": 3.4601, + "step": 103500 + }, + { + "epoch": 7.032545182769398, + "grad_norm": 0.7986583709716797, + "learning_rate": 0.00012093185215382526, + "loss": 3.2301, + "step": 103505 + }, + { + "epoch": 7.03288490284006, + "grad_norm": 0.800980269908905, + "learning_rate": 0.00012088938714499254, + "loss": 3.5215, + "step": 103510 + }, + { + "epoch": 7.033224622910722, + "grad_norm": 0.704522430896759, + "learning_rate": 0.0001208469221361598, + "loss": 3.4919, + "step": 103515 + }, + { + "epoch": 7.033564342981383, + "grad_norm": 1.0012117624282837, + "learning_rate": 0.00012080445712732708, + "loss": 3.254, + "step": 103520 + }, + { + "epoch": 7.033904063052045, + "grad_norm": 0.7963822484016418, + "learning_rate": 0.00012076199211849436, + "loss": 3.2772, + "step": 103525 + }, + { + "epoch": 7.034243783122707, + "grad_norm": 0.8562939167022705, + "learning_rate": 0.00012071952710966163, + "loss": 3.5954, + "step": 103530 + }, + { + "epoch": 7.034583503193368, + "grad_norm": 0.8998008370399475, + "learning_rate": 0.00012067706210082892, + "loss": 3.372, + "step": 103535 + }, + { + "epoch": 7.03492322326403, + "grad_norm": 0.8519073128700256, + "learning_rate": 0.0001206345970919962, + "loss": 3.6275, + "step": 103540 + }, + { + "epoch": 7.0352629433346925, + "grad_norm": 1.165730357170105, + "learning_rate": 0.00012059213208316347, + "loss": 3.3133, + "step": 103545 + }, + { + "epoch": 7.035602663405354, + "grad_norm": 1.018378496170044, + "learning_rate": 0.00012054966707433075, + "loss": 3.1918, + "step": 103550 + }, + { + "epoch": 7.035942383476016, + "grad_norm": 0.7490879893302917, + "learning_rate": 0.00012050720206549803, + "loss": 3.655, + "step": 103555 + }, + { + "epoch": 7.036282103546678, + "grad_norm": 0.8971711993217468, + "learning_rate": 0.00012046473705666531, + "loss": 3.7366, + "step": 103560 + }, + { + "epoch": 7.036621823617339, + "grad_norm": 0.8422214388847351, + "learning_rate": 0.00012042227204783258, + "loss": 3.5184, + "step": 103565 + }, + { + "epoch": 7.036961543688001, + "grad_norm": 0.8334212303161621, + "learning_rate": 0.00012037980703899987, + "loss": 3.3664, + "step": 103570 + }, + { + "epoch": 7.037301263758663, + "grad_norm": 0.7726700305938721, + "learning_rate": 0.00012033734203016715, + "loss": 3.4754, + "step": 103575 + }, + { + "epoch": 7.037640983829324, + "grad_norm": 0.9236964583396912, + "learning_rate": 0.00012029487702133442, + "loss": 3.2212, + "step": 103580 + }, + { + "epoch": 7.037980703899986, + "grad_norm": 0.8300935626029968, + "learning_rate": 0.0001202524120125017, + "loss": 3.4649, + "step": 103585 + }, + { + "epoch": 7.0383204239706485, + "grad_norm": 1.0133366584777832, + "learning_rate": 0.00012020994700366898, + "loss": 3.1674, + "step": 103590 + }, + { + "epoch": 7.03866014404131, + "grad_norm": 0.8820975422859192, + "learning_rate": 0.00012016748199483626, + "loss": 3.644, + "step": 103595 + }, + { + "epoch": 7.038999864111972, + "grad_norm": 0.9881974458694458, + "learning_rate": 0.00012012501698600354, + "loss": 3.3059, + "step": 103600 + }, + { + "epoch": 7.039339584182634, + "grad_norm": 0.7597233653068542, + "learning_rate": 0.00012008255197717082, + "loss": 3.5082, + "step": 103605 + }, + { + "epoch": 7.039679304253295, + "grad_norm": 0.940796434879303, + "learning_rate": 0.0001200400869683381, + "loss": 3.6471, + "step": 103610 + }, + { + "epoch": 7.040019024323957, + "grad_norm": 0.9195736050605774, + "learning_rate": 0.00011999762195950537, + "loss": 3.5053, + "step": 103615 + }, + { + "epoch": 7.040358744394619, + "grad_norm": 0.8696350455284119, + "learning_rate": 0.00011995515695067265, + "loss": 3.4775, + "step": 103620 + }, + { + "epoch": 7.04069846446528, + "grad_norm": 0.8776360154151917, + "learning_rate": 0.00011991269194183993, + "loss": 3.4651, + "step": 103625 + }, + { + "epoch": 7.041038184535942, + "grad_norm": 0.9520487785339355, + "learning_rate": 0.00011987022693300719, + "loss": 3.322, + "step": 103630 + }, + { + "epoch": 7.0413779046066045, + "grad_norm": 0.7562599182128906, + "learning_rate": 0.00011982776192417449, + "loss": 3.402, + "step": 103635 + }, + { + "epoch": 7.041717624677266, + "grad_norm": 0.9310691356658936, + "learning_rate": 0.00011978529691534177, + "loss": 3.292, + "step": 103640 + }, + { + "epoch": 7.042057344747928, + "grad_norm": 0.9521851539611816, + "learning_rate": 0.00011974283190650905, + "loss": 3.4752, + "step": 103645 + }, + { + "epoch": 7.04239706481859, + "grad_norm": 1.1520836353302002, + "learning_rate": 0.00011970036689767631, + "loss": 3.4968, + "step": 103650 + }, + { + "epoch": 7.042736784889251, + "grad_norm": 1.1612571477890015, + "learning_rate": 0.00011965790188884359, + "loss": 3.8366, + "step": 103655 + }, + { + "epoch": 7.043076504959913, + "grad_norm": 0.9156963229179382, + "learning_rate": 0.00011961543688001087, + "loss": 3.4113, + "step": 103660 + }, + { + "epoch": 7.043416225030575, + "grad_norm": 0.8288331627845764, + "learning_rate": 0.00011957297187117815, + "loss": 3.4381, + "step": 103665 + }, + { + "epoch": 7.043755945101236, + "grad_norm": 2.027019739151001, + "learning_rate": 0.00011953050686234543, + "loss": 3.5628, + "step": 103670 + }, + { + "epoch": 7.0440956651718984, + "grad_norm": 1.3186067342758179, + "learning_rate": 0.00011948804185351271, + "loss": 3.4894, + "step": 103675 + }, + { + "epoch": 7.0444353852425605, + "grad_norm": 1.2530373334884644, + "learning_rate": 0.00011944557684467999, + "loss": 3.4125, + "step": 103680 + }, + { + "epoch": 7.044775105313222, + "grad_norm": 0.9719299674034119, + "learning_rate": 0.00011940311183584726, + "loss": 3.5887, + "step": 103685 + }, + { + "epoch": 7.045114825383884, + "grad_norm": 0.9750646948814392, + "learning_rate": 0.00011936064682701454, + "loss": 3.1488, + "step": 103690 + }, + { + "epoch": 7.045454545454546, + "grad_norm": 0.8489910960197449, + "learning_rate": 0.00011931818181818182, + "loss": 3.2549, + "step": 103695 + }, + { + "epoch": 7.045794265525207, + "grad_norm": 0.9788417816162109, + "learning_rate": 0.0001192757168093491, + "loss": 3.3603, + "step": 103700 + }, + { + "epoch": 7.046133985595869, + "grad_norm": 1.0879371166229248, + "learning_rate": 0.00011923325180051638, + "loss": 3.5034, + "step": 103705 + }, + { + "epoch": 7.046473705666531, + "grad_norm": 1.024456262588501, + "learning_rate": 0.00011919078679168366, + "loss": 3.6326, + "step": 103710 + }, + { + "epoch": 7.046813425737192, + "grad_norm": 0.8746587634086609, + "learning_rate": 0.00011914832178285093, + "loss": 3.2745, + "step": 103715 + }, + { + "epoch": 7.0471531458078545, + "grad_norm": 1.0345426797866821, + "learning_rate": 0.0001191058567740182, + "loss": 3.3103, + "step": 103720 + }, + { + "epoch": 7.0474928658785165, + "grad_norm": 0.9651219844818115, + "learning_rate": 0.00011906339176518549, + "loss": 3.6789, + "step": 103725 + }, + { + "epoch": 7.047832585949178, + "grad_norm": 0.9432247877120972, + "learning_rate": 0.00011902092675635278, + "loss": 3.6148, + "step": 103730 + }, + { + "epoch": 7.04817230601984, + "grad_norm": 0.8724402189254761, + "learning_rate": 0.00011897846174752005, + "loss": 3.3433, + "step": 103735 + }, + { + "epoch": 7.048512026090501, + "grad_norm": 1.1888431310653687, + "learning_rate": 0.00011893599673868733, + "loss": 3.6637, + "step": 103740 + }, + { + "epoch": 7.048851746161163, + "grad_norm": 0.8431079983711243, + "learning_rate": 0.0001188935317298546, + "loss": 3.5936, + "step": 103745 + }, + { + "epoch": 7.049191466231825, + "grad_norm": 0.8296429514884949, + "learning_rate": 0.00011885106672102187, + "loss": 3.2708, + "step": 103750 + }, + { + "epoch": 7.049531186302486, + "grad_norm": 1.0992761850357056, + "learning_rate": 0.00011880860171218915, + "loss": 3.557, + "step": 103755 + }, + { + "epoch": 7.049870906373148, + "grad_norm": 1.1620677709579468, + "learning_rate": 0.00011876613670335643, + "loss": 3.1672, + "step": 103760 + }, + { + "epoch": 7.0502106264438105, + "grad_norm": 1.232216715812683, + "learning_rate": 0.00011872367169452373, + "loss": 3.0826, + "step": 103765 + }, + { + "epoch": 7.050550346514472, + "grad_norm": 0.9185592532157898, + "learning_rate": 0.000118681206685691, + "loss": 3.4651, + "step": 103770 + }, + { + "epoch": 7.050890066585134, + "grad_norm": 0.6896264553070068, + "learning_rate": 0.00011863874167685827, + "loss": 3.287, + "step": 103775 + }, + { + "epoch": 7.051229786655796, + "grad_norm": 1.0331223011016846, + "learning_rate": 0.00011859627666802555, + "loss": 3.3243, + "step": 103780 + }, + { + "epoch": 7.051569506726457, + "grad_norm": 0.8559095859527588, + "learning_rate": 0.00011855381165919282, + "loss": 3.6065, + "step": 103785 + }, + { + "epoch": 7.051909226797119, + "grad_norm": 0.933651864528656, + "learning_rate": 0.0001185113466503601, + "loss": 3.2909, + "step": 103790 + }, + { + "epoch": 7.052248946867781, + "grad_norm": 0.7275342345237732, + "learning_rate": 0.0001184688816415274, + "loss": 3.4243, + "step": 103795 + }, + { + "epoch": 7.052588666938442, + "grad_norm": 1.095494031906128, + "learning_rate": 0.00011842641663269466, + "loss": 3.4584, + "step": 103800 + }, + { + "epoch": 7.052928387009104, + "grad_norm": 0.7772966623306274, + "learning_rate": 0.00011838395162386194, + "loss": 3.258, + "step": 103805 + }, + { + "epoch": 7.0532681070797665, + "grad_norm": 0.85450679063797, + "learning_rate": 0.00011834148661502922, + "loss": 3.2563, + "step": 103810 + }, + { + "epoch": 7.053607827150428, + "grad_norm": 0.9391647577285767, + "learning_rate": 0.0001182990216061965, + "loss": 3.5654, + "step": 103815 + }, + { + "epoch": 7.05394754722109, + "grad_norm": 0.8327199816703796, + "learning_rate": 0.00011825655659736377, + "loss": 3.202, + "step": 103820 + }, + { + "epoch": 7.054287267291752, + "grad_norm": 1.1498095989227295, + "learning_rate": 0.00011821409158853105, + "loss": 3.2289, + "step": 103825 + }, + { + "epoch": 7.054626987362413, + "grad_norm": 0.9259196519851685, + "learning_rate": 0.00011817162657969834, + "loss": 3.369, + "step": 103830 + }, + { + "epoch": 7.054966707433075, + "grad_norm": 1.1602895259857178, + "learning_rate": 0.00011812916157086561, + "loss": 3.4256, + "step": 103835 + }, + { + "epoch": 7.055306427503737, + "grad_norm": 1.4370789527893066, + "learning_rate": 0.00011808669656203289, + "loss": 3.1684, + "step": 103840 + }, + { + "epoch": 7.055646147574398, + "grad_norm": 0.8487733006477356, + "learning_rate": 0.00011804423155320017, + "loss": 3.4128, + "step": 103845 + }, + { + "epoch": 7.05598586764506, + "grad_norm": 0.8723309636116028, + "learning_rate": 0.00011800176654436745, + "loss": 3.4696, + "step": 103850 + }, + { + "epoch": 7.0563255877157225, + "grad_norm": 0.8852866888046265, + "learning_rate": 0.00011795930153553471, + "loss": 3.5921, + "step": 103855 + }, + { + "epoch": 7.056665307786384, + "grad_norm": 1.0823743343353271, + "learning_rate": 0.00011791683652670201, + "loss": 3.5612, + "step": 103860 + }, + { + "epoch": 7.057005027857046, + "grad_norm": 0.7627637386322021, + "learning_rate": 0.00011787437151786929, + "loss": 3.3592, + "step": 103865 + }, + { + "epoch": 7.057344747927708, + "grad_norm": 1.1567111015319824, + "learning_rate": 0.00011783190650903655, + "loss": 3.5185, + "step": 103870 + }, + { + "epoch": 7.057684467998369, + "grad_norm": 1.1862696409225464, + "learning_rate": 0.00011778944150020383, + "loss": 3.4289, + "step": 103875 + }, + { + "epoch": 7.058024188069031, + "grad_norm": 1.107588768005371, + "learning_rate": 0.00011774697649137111, + "loss": 2.8041, + "step": 103880 + }, + { + "epoch": 7.058363908139693, + "grad_norm": 0.803798496723175, + "learning_rate": 0.00011770451148253838, + "loss": 3.38, + "step": 103885 + }, + { + "epoch": 7.058703628210354, + "grad_norm": 0.8542370200157166, + "learning_rate": 0.00011766204647370566, + "loss": 3.6477, + "step": 103890 + }, + { + "epoch": 7.059043348281016, + "grad_norm": 0.7950045466423035, + "learning_rate": 0.00011761958146487295, + "loss": 3.3765, + "step": 103895 + }, + { + "epoch": 7.0593830683516785, + "grad_norm": 0.9837733507156372, + "learning_rate": 0.00011757711645604023, + "loss": 3.4371, + "step": 103900 + }, + { + "epoch": 7.05972278842234, + "grad_norm": 0.9196634888648987, + "learning_rate": 0.0001175346514472075, + "loss": 3.5961, + "step": 103905 + }, + { + "epoch": 7.060062508493002, + "grad_norm": 0.7951719164848328, + "learning_rate": 0.00011749218643837478, + "loss": 3.5224, + "step": 103910 + }, + { + "epoch": 7.060402228563664, + "grad_norm": 0.7954001426696777, + "learning_rate": 0.00011744972142954206, + "loss": 3.0632, + "step": 103915 + }, + { + "epoch": 7.060741948634325, + "grad_norm": 1.2521532773971558, + "learning_rate": 0.00011740725642070933, + "loss": 3.2272, + "step": 103920 + }, + { + "epoch": 7.061081668704987, + "grad_norm": 0.8430784344673157, + "learning_rate": 0.00011736479141187661, + "loss": 3.435, + "step": 103925 + }, + { + "epoch": 7.061421388775649, + "grad_norm": 1.1518324613571167, + "learning_rate": 0.0001173223264030439, + "loss": 3.4359, + "step": 103930 + }, + { + "epoch": 7.06176110884631, + "grad_norm": 0.8483431935310364, + "learning_rate": 0.00011727986139421118, + "loss": 3.6172, + "step": 103935 + }, + { + "epoch": 7.062100828916972, + "grad_norm": 0.9614658951759338, + "learning_rate": 0.00011723739638537845, + "loss": 3.4237, + "step": 103940 + }, + { + "epoch": 7.0624405489876345, + "grad_norm": 0.7944939732551575, + "learning_rate": 0.00011719493137654573, + "loss": 3.4935, + "step": 103945 + }, + { + "epoch": 7.062780269058296, + "grad_norm": 1.1312716007232666, + "learning_rate": 0.00011715246636771301, + "loss": 3.252, + "step": 103950 + }, + { + "epoch": 7.063119989128958, + "grad_norm": 0.8617753386497498, + "learning_rate": 0.00011711000135888028, + "loss": 3.4909, + "step": 103955 + }, + { + "epoch": 7.06345970919962, + "grad_norm": 0.972758412361145, + "learning_rate": 0.00011706753635004757, + "loss": 3.3837, + "step": 103960 + }, + { + "epoch": 7.063799429270281, + "grad_norm": 0.8814369440078735, + "learning_rate": 0.00011702507134121485, + "loss": 3.5573, + "step": 103965 + }, + { + "epoch": 7.064139149340943, + "grad_norm": 1.041167974472046, + "learning_rate": 0.00011698260633238212, + "loss": 3.5357, + "step": 103970 + }, + { + "epoch": 7.064478869411605, + "grad_norm": 0.7475009560585022, + "learning_rate": 0.0001169401413235494, + "loss": 3.5246, + "step": 103975 + }, + { + "epoch": 7.064818589482266, + "grad_norm": 0.996152937412262, + "learning_rate": 0.00011689767631471668, + "loss": 3.3584, + "step": 103980 + }, + { + "epoch": 7.0651583095529285, + "grad_norm": 0.9559910893440247, + "learning_rate": 0.00011685521130588396, + "loss": 3.1757, + "step": 103985 + }, + { + "epoch": 7.0654980296235905, + "grad_norm": 0.8813297748565674, + "learning_rate": 0.00011681274629705122, + "loss": 3.1744, + "step": 103990 + }, + { + "epoch": 7.065837749694252, + "grad_norm": 0.9290105700492859, + "learning_rate": 0.00011677028128821852, + "loss": 3.3723, + "step": 103995 + }, + { + "epoch": 7.066177469764914, + "grad_norm": 1.083485722541809, + "learning_rate": 0.0001167278162793858, + "loss": 3.5393, + "step": 104000 + }, + { + "epoch": 7.066517189835576, + "grad_norm": 0.7771942019462585, + "learning_rate": 0.00011668535127055306, + "loss": 3.3225, + "step": 104005 + }, + { + "epoch": 7.066856909906237, + "grad_norm": 0.9171308279037476, + "learning_rate": 0.00011664288626172034, + "loss": 3.3008, + "step": 104010 + }, + { + "epoch": 7.067196629976899, + "grad_norm": 0.9718893766403198, + "learning_rate": 0.00011660042125288762, + "loss": 3.3102, + "step": 104015 + }, + { + "epoch": 7.067536350047561, + "grad_norm": 1.1542861461639404, + "learning_rate": 0.0001165579562440549, + "loss": 3.6627, + "step": 104020 + }, + { + "epoch": 7.067876070118222, + "grad_norm": 1.144759178161621, + "learning_rate": 0.00011651549123522218, + "loss": 3.2602, + "step": 104025 + }, + { + "epoch": 7.0682157901888845, + "grad_norm": 0.7834345698356628, + "learning_rate": 0.00011647302622638946, + "loss": 3.3549, + "step": 104030 + }, + { + "epoch": 7.0685555102595465, + "grad_norm": 0.7636922001838684, + "learning_rate": 0.00011643056121755674, + "loss": 3.4545, + "step": 104035 + }, + { + "epoch": 7.068895230330208, + "grad_norm": 0.8749014735221863, + "learning_rate": 0.00011638809620872401, + "loss": 3.5624, + "step": 104040 + }, + { + "epoch": 7.06923495040087, + "grad_norm": 0.8431821465492249, + "learning_rate": 0.00011634563119989129, + "loss": 3.5277, + "step": 104045 + }, + { + "epoch": 7.069574670471532, + "grad_norm": 0.7863978743553162, + "learning_rate": 0.00011630316619105857, + "loss": 3.4315, + "step": 104050 + }, + { + "epoch": 7.069914390542193, + "grad_norm": 1.0836535692214966, + "learning_rate": 0.00011626070118222584, + "loss": 3.2024, + "step": 104055 + }, + { + "epoch": 7.070254110612855, + "grad_norm": 0.7713770270347595, + "learning_rate": 0.00011621823617339313, + "loss": 3.6159, + "step": 104060 + }, + { + "epoch": 7.070593830683517, + "grad_norm": 0.8263673186302185, + "learning_rate": 0.00011617577116456041, + "loss": 3.1695, + "step": 104065 + }, + { + "epoch": 7.070933550754178, + "grad_norm": 0.9432210326194763, + "learning_rate": 0.00011613330615572769, + "loss": 3.4158, + "step": 104070 + }, + { + "epoch": 7.0712732708248405, + "grad_norm": 1.0932347774505615, + "learning_rate": 0.00011609084114689496, + "loss": 3.3052, + "step": 104075 + }, + { + "epoch": 7.0716129908955025, + "grad_norm": 1.0324294567108154, + "learning_rate": 0.00011604837613806224, + "loss": 3.2028, + "step": 104080 + }, + { + "epoch": 7.071952710966164, + "grad_norm": 0.9061249494552612, + "learning_rate": 0.00011600591112922952, + "loss": 3.4512, + "step": 104085 + }, + { + "epoch": 7.072292431036826, + "grad_norm": 1.0249773263931274, + "learning_rate": 0.0001159634461203968, + "loss": 3.3364, + "step": 104090 + }, + { + "epoch": 7.072632151107487, + "grad_norm": 0.8928452730178833, + "learning_rate": 0.00011592098111156408, + "loss": 3.3761, + "step": 104095 + }, + { + "epoch": 7.072971871178149, + "grad_norm": 1.2109380960464478, + "learning_rate": 0.00011587851610273136, + "loss": 3.2702, + "step": 104100 + }, + { + "epoch": 7.073311591248811, + "grad_norm": 0.8806576132774353, + "learning_rate": 0.00011583605109389864, + "loss": 3.4238, + "step": 104105 + }, + { + "epoch": 7.073651311319472, + "grad_norm": 0.899205207824707, + "learning_rate": 0.0001157935860850659, + "loss": 3.3622, + "step": 104110 + }, + { + "epoch": 7.073991031390134, + "grad_norm": 0.9000893831253052, + "learning_rate": 0.00011575112107623318, + "loss": 3.49, + "step": 104115 + }, + { + "epoch": 7.0743307514607965, + "grad_norm": 0.8828895688056946, + "learning_rate": 0.00011570865606740046, + "loss": 3.4338, + "step": 104120 + }, + { + "epoch": 7.074670471531458, + "grad_norm": 1.4540808200836182, + "learning_rate": 0.00011566619105856774, + "loss": 3.313, + "step": 104125 + }, + { + "epoch": 7.07501019160212, + "grad_norm": 0.8697982430458069, + "learning_rate": 0.00011562372604973502, + "loss": 3.4686, + "step": 104130 + }, + { + "epoch": 7.075349911672782, + "grad_norm": 0.8439257144927979, + "learning_rate": 0.0001155812610409023, + "loss": 3.4381, + "step": 104135 + }, + { + "epoch": 7.075689631743443, + "grad_norm": 0.7916607856750488, + "learning_rate": 0.00011553879603206957, + "loss": 3.2854, + "step": 104140 + }, + { + "epoch": 7.076029351814105, + "grad_norm": 0.6912140250205994, + "learning_rate": 0.00011549633102323685, + "loss": 3.4071, + "step": 104145 + }, + { + "epoch": 7.076369071884767, + "grad_norm": 0.8793339133262634, + "learning_rate": 0.00011545386601440413, + "loss": 3.496, + "step": 104150 + }, + { + "epoch": 7.076708791955428, + "grad_norm": 0.9650248289108276, + "learning_rate": 0.00011541140100557142, + "loss": 3.4218, + "step": 104155 + }, + { + "epoch": 7.07704851202609, + "grad_norm": 0.9607436656951904, + "learning_rate": 0.00011536893599673869, + "loss": 3.4216, + "step": 104160 + }, + { + "epoch": 7.0773882320967525, + "grad_norm": 1.2943716049194336, + "learning_rate": 0.00011532647098790597, + "loss": 3.2116, + "step": 104165 + }, + { + "epoch": 7.077727952167414, + "grad_norm": 1.0805705785751343, + "learning_rate": 0.00011528400597907325, + "loss": 3.1913, + "step": 104170 + }, + { + "epoch": 7.078067672238076, + "grad_norm": 1.2857097387313843, + "learning_rate": 0.00011524154097024052, + "loss": 3.5172, + "step": 104175 + }, + { + "epoch": 7.078407392308738, + "grad_norm": 0.9836971759796143, + "learning_rate": 0.0001151990759614078, + "loss": 3.329, + "step": 104180 + }, + { + "epoch": 7.078747112379399, + "grad_norm": 0.9362834095954895, + "learning_rate": 0.00011515661095257508, + "loss": 3.4744, + "step": 104185 + }, + { + "epoch": 7.079086832450061, + "grad_norm": 0.9694989323616028, + "learning_rate": 0.00011511414594374237, + "loss": 3.3325, + "step": 104190 + }, + { + "epoch": 7.079426552520723, + "grad_norm": 1.000629186630249, + "learning_rate": 0.00011507168093490964, + "loss": 3.5327, + "step": 104195 + }, + { + "epoch": 7.079766272591384, + "grad_norm": 0.8689146041870117, + "learning_rate": 0.00011502921592607692, + "loss": 3.3722, + "step": 104200 + }, + { + "epoch": 7.080105992662046, + "grad_norm": 0.8989437222480774, + "learning_rate": 0.0001149867509172442, + "loss": 3.199, + "step": 104205 + }, + { + "epoch": 7.0804457127327085, + "grad_norm": 1.1261626482009888, + "learning_rate": 0.00011494428590841146, + "loss": 3.5177, + "step": 104210 + }, + { + "epoch": 7.08078543280337, + "grad_norm": 0.9429791569709778, + "learning_rate": 0.00011490182089957874, + "loss": 3.4696, + "step": 104215 + }, + { + "epoch": 7.081125152874032, + "grad_norm": 0.8431165218353271, + "learning_rate": 0.00011485935589074604, + "loss": 3.4233, + "step": 104220 + }, + { + "epoch": 7.081464872944694, + "grad_norm": 0.7962741255760193, + "learning_rate": 0.0001148168908819133, + "loss": 3.229, + "step": 104225 + }, + { + "epoch": 7.081804593015355, + "grad_norm": 1.0444518327713013, + "learning_rate": 0.00011477442587308058, + "loss": 3.1655, + "step": 104230 + }, + { + "epoch": 7.082144313086017, + "grad_norm": 1.1500186920166016, + "learning_rate": 0.00011473196086424786, + "loss": 3.1653, + "step": 104235 + }, + { + "epoch": 7.082484033156679, + "grad_norm": 0.9361345171928406, + "learning_rate": 0.00011468949585541514, + "loss": 3.2209, + "step": 104240 + }, + { + "epoch": 7.08282375322734, + "grad_norm": 0.8623837828636169, + "learning_rate": 0.00011464703084658241, + "loss": 3.4358, + "step": 104245 + }, + { + "epoch": 7.083163473298002, + "grad_norm": 1.144843578338623, + "learning_rate": 0.00011460456583774969, + "loss": 3.4153, + "step": 104250 + }, + { + "epoch": 7.0835031933686645, + "grad_norm": 0.8630163073539734, + "learning_rate": 0.00011456210082891698, + "loss": 3.3388, + "step": 104255 + }, + { + "epoch": 7.083842913439326, + "grad_norm": 0.8040218949317932, + "learning_rate": 0.00011451963582008425, + "loss": 3.6193, + "step": 104260 + }, + { + "epoch": 7.084182633509988, + "grad_norm": 1.1537256240844727, + "learning_rate": 0.00011447717081125153, + "loss": 3.3024, + "step": 104265 + }, + { + "epoch": 7.08452235358065, + "grad_norm": 0.6432016491889954, + "learning_rate": 0.00011443470580241881, + "loss": 3.3416, + "step": 104270 + }, + { + "epoch": 7.084862073651311, + "grad_norm": 0.8754247426986694, + "learning_rate": 0.00011439224079358609, + "loss": 3.3943, + "step": 104275 + }, + { + "epoch": 7.085201793721973, + "grad_norm": 0.7359702587127686, + "learning_rate": 0.00011434977578475336, + "loss": 3.5522, + "step": 104280 + }, + { + "epoch": 7.085541513792635, + "grad_norm": 0.9946436882019043, + "learning_rate": 0.00011430731077592064, + "loss": 3.2961, + "step": 104285 + }, + { + "epoch": 7.085881233863296, + "grad_norm": 1.174688696861267, + "learning_rate": 0.00011426484576708793, + "loss": 3.6084, + "step": 104290 + }, + { + "epoch": 7.0862209539339585, + "grad_norm": 0.9201748967170715, + "learning_rate": 0.0001142223807582552, + "loss": 3.6033, + "step": 104295 + }, + { + "epoch": 7.0865606740046205, + "grad_norm": 0.8659526109695435, + "learning_rate": 0.00011417991574942248, + "loss": 3.4914, + "step": 104300 + }, + { + "epoch": 7.086900394075282, + "grad_norm": 0.8251466751098633, + "learning_rate": 0.00011413745074058976, + "loss": 3.4182, + "step": 104305 + }, + { + "epoch": 7.087240114145944, + "grad_norm": 1.0299113988876343, + "learning_rate": 0.00011409498573175702, + "loss": 3.2177, + "step": 104310 + }, + { + "epoch": 7.087579834216606, + "grad_norm": 1.2450931072235107, + "learning_rate": 0.0001140525207229243, + "loss": 3.3095, + "step": 104315 + }, + { + "epoch": 7.087919554287267, + "grad_norm": 0.9580865502357483, + "learning_rate": 0.0001140100557140916, + "loss": 3.3938, + "step": 104320 + }, + { + "epoch": 7.088259274357929, + "grad_norm": 0.9608656764030457, + "learning_rate": 0.00011396759070525888, + "loss": 3.722, + "step": 104325 + }, + { + "epoch": 7.088598994428591, + "grad_norm": 0.8754264116287231, + "learning_rate": 0.00011392512569642615, + "loss": 3.438, + "step": 104330 + }, + { + "epoch": 7.088938714499252, + "grad_norm": 0.9541337490081787, + "learning_rate": 0.00011388266068759343, + "loss": 3.4075, + "step": 104335 + }, + { + "epoch": 7.0892784345699145, + "grad_norm": 0.8053044080734253, + "learning_rate": 0.0001138401956787607, + "loss": 3.2775, + "step": 104340 + }, + { + "epoch": 7.0896181546405765, + "grad_norm": 0.9518930315971375, + "learning_rate": 0.00011379773066992797, + "loss": 3.3921, + "step": 104345 + }, + { + "epoch": 7.089957874711238, + "grad_norm": 0.9961082339286804, + "learning_rate": 0.00011375526566109525, + "loss": 3.1984, + "step": 104350 + }, + { + "epoch": 7.0902975947819, + "grad_norm": 0.7751923203468323, + "learning_rate": 0.00011371280065226255, + "loss": 3.4218, + "step": 104355 + }, + { + "epoch": 7.090637314852562, + "grad_norm": 1.0584887266159058, + "learning_rate": 0.00011367033564342983, + "loss": 3.4707, + "step": 104360 + }, + { + "epoch": 7.090977034923223, + "grad_norm": 0.7483612298965454, + "learning_rate": 0.00011362787063459709, + "loss": 3.4697, + "step": 104365 + }, + { + "epoch": 7.091316754993885, + "grad_norm": 0.7904231548309326, + "learning_rate": 0.00011358540562576437, + "loss": 3.4748, + "step": 104370 + }, + { + "epoch": 7.091656475064547, + "grad_norm": 0.7447105646133423, + "learning_rate": 0.00011354294061693165, + "loss": 3.5033, + "step": 104375 + }, + { + "epoch": 7.091996195135208, + "grad_norm": 0.9665389657020569, + "learning_rate": 0.00011350047560809892, + "loss": 3.2903, + "step": 104380 + }, + { + "epoch": 7.0923359152058705, + "grad_norm": 0.9210439920425415, + "learning_rate": 0.00011345801059926621, + "loss": 3.1729, + "step": 104385 + }, + { + "epoch": 7.0926756352765326, + "grad_norm": 0.9475433230400085, + "learning_rate": 0.00011341554559043349, + "loss": 3.4659, + "step": 104390 + }, + { + "epoch": 7.093015355347194, + "grad_norm": 0.9876008033752441, + "learning_rate": 0.00011337308058160076, + "loss": 3.5487, + "step": 104395 + }, + { + "epoch": 7.093355075417856, + "grad_norm": 0.8995936512947083, + "learning_rate": 0.00011333061557276804, + "loss": 3.3476, + "step": 104400 + }, + { + "epoch": 7.093694795488518, + "grad_norm": 0.8814787864685059, + "learning_rate": 0.00011328815056393532, + "loss": 3.2905, + "step": 104405 + }, + { + "epoch": 7.094034515559179, + "grad_norm": 1.0542019605636597, + "learning_rate": 0.0001132456855551026, + "loss": 3.3689, + "step": 104410 + }, + { + "epoch": 7.094374235629841, + "grad_norm": 0.8763954639434814, + "learning_rate": 0.00011320322054626987, + "loss": 3.3737, + "step": 104415 + }, + { + "epoch": 7.094713955700502, + "grad_norm": 0.9940475225448608, + "learning_rate": 0.00011316075553743716, + "loss": 3.2352, + "step": 104420 + }, + { + "epoch": 7.095053675771164, + "grad_norm": 0.8606086373329163, + "learning_rate": 0.00011311829052860444, + "loss": 3.2746, + "step": 104425 + }, + { + "epoch": 7.0953933958418265, + "grad_norm": 0.7688761949539185, + "learning_rate": 0.0001130758255197717, + "loss": 3.2147, + "step": 104430 + }, + { + "epoch": 7.095733115912488, + "grad_norm": 0.8183848857879639, + "learning_rate": 0.00011303336051093899, + "loss": 3.4895, + "step": 104435 + }, + { + "epoch": 7.09607283598315, + "grad_norm": 0.8531478047370911, + "learning_rate": 0.00011299089550210627, + "loss": 3.2974, + "step": 104440 + }, + { + "epoch": 7.096412556053812, + "grad_norm": 0.9086788296699524, + "learning_rate": 0.00011294843049327355, + "loss": 3.3471, + "step": 104445 + }, + { + "epoch": 7.096752276124473, + "grad_norm": 1.100573182106018, + "learning_rate": 0.00011290596548444083, + "loss": 3.2655, + "step": 104450 + }, + { + "epoch": 7.097091996195135, + "grad_norm": 1.0046926736831665, + "learning_rate": 0.0001128635004756081, + "loss": 3.2727, + "step": 104455 + }, + { + "epoch": 7.097431716265797, + "grad_norm": 0.9911916851997375, + "learning_rate": 0.00011282103546677539, + "loss": 3.4154, + "step": 104460 + }, + { + "epoch": 7.097771436336458, + "grad_norm": 0.9766772985458374, + "learning_rate": 0.00011277857045794265, + "loss": 3.4795, + "step": 104465 + }, + { + "epoch": 7.09811115640712, + "grad_norm": 0.8780450224876404, + "learning_rate": 0.00011273610544910993, + "loss": 3.0992, + "step": 104470 + }, + { + "epoch": 7.0984508764777825, + "grad_norm": 0.7890117168426514, + "learning_rate": 0.00011269364044027721, + "loss": 3.2542, + "step": 104475 + }, + { + "epoch": 7.098790596548444, + "grad_norm": 0.8777451515197754, + "learning_rate": 0.00011265117543144448, + "loss": 3.519, + "step": 104480 + }, + { + "epoch": 7.099130316619106, + "grad_norm": 0.9404387474060059, + "learning_rate": 0.00011260871042261177, + "loss": 3.1931, + "step": 104485 + }, + { + "epoch": 7.099470036689768, + "grad_norm": 0.7664154171943665, + "learning_rate": 0.00011256624541377905, + "loss": 3.5013, + "step": 104490 + }, + { + "epoch": 7.099809756760429, + "grad_norm": 1.0124828815460205, + "learning_rate": 0.00011252378040494633, + "loss": 3.0669, + "step": 104495 + }, + { + "epoch": 7.100149476831091, + "grad_norm": 0.9436324834823608, + "learning_rate": 0.0001124813153961136, + "loss": 3.5017, + "step": 104500 + }, + { + "epoch": 7.100489196901753, + "grad_norm": 0.9536551833152771, + "learning_rate": 0.00011243885038728088, + "loss": 3.2786, + "step": 104505 + }, + { + "epoch": 7.100828916972414, + "grad_norm": 0.8831111788749695, + "learning_rate": 0.00011239638537844816, + "loss": 3.4778, + "step": 104510 + }, + { + "epoch": 7.101168637043076, + "grad_norm": 0.9814079403877258, + "learning_rate": 0.00011235392036961544, + "loss": 3.5312, + "step": 104515 + }, + { + "epoch": 7.1015083571137385, + "grad_norm": 1.09671950340271, + "learning_rate": 0.00011231145536078272, + "loss": 3.2533, + "step": 104520 + }, + { + "epoch": 7.1018480771844, + "grad_norm": 1.040530800819397, + "learning_rate": 0.00011226899035195, + "loss": 3.163, + "step": 104525 + }, + { + "epoch": 7.102187797255062, + "grad_norm": 1.2957487106323242, + "learning_rate": 0.00011222652534311728, + "loss": 3.3449, + "step": 104530 + }, + { + "epoch": 7.102527517325724, + "grad_norm": 0.9211296439170837, + "learning_rate": 0.00011218406033428455, + "loss": 3.2084, + "step": 104535 + }, + { + "epoch": 7.102867237396385, + "grad_norm": 0.9784913659095764, + "learning_rate": 0.00011214159532545183, + "loss": 3.3464, + "step": 104540 + }, + { + "epoch": 7.103206957467047, + "grad_norm": 1.0219300985336304, + "learning_rate": 0.00011209913031661911, + "loss": 3.5206, + "step": 104545 + }, + { + "epoch": 7.103546677537709, + "grad_norm": 0.9067562818527222, + "learning_rate": 0.00011205666530778639, + "loss": 3.3906, + "step": 104550 + }, + { + "epoch": 7.10388639760837, + "grad_norm": 1.0711209774017334, + "learning_rate": 0.00011201420029895367, + "loss": 3.3198, + "step": 104555 + }, + { + "epoch": 7.1042261176790324, + "grad_norm": 0.9620311260223389, + "learning_rate": 0.00011197173529012095, + "loss": 3.2814, + "step": 104560 + }, + { + "epoch": 7.1045658377496945, + "grad_norm": 0.9803740382194519, + "learning_rate": 0.00011192927028128821, + "loss": 3.2009, + "step": 104565 + }, + { + "epoch": 7.104905557820356, + "grad_norm": 0.8016420006752014, + "learning_rate": 0.0001118868052724555, + "loss": 3.354, + "step": 104570 + }, + { + "epoch": 7.105245277891018, + "grad_norm": 1.1313728094100952, + "learning_rate": 0.00011184434026362277, + "loss": 3.3668, + "step": 104575 + }, + { + "epoch": 7.10558499796168, + "grad_norm": 1.2548948526382446, + "learning_rate": 0.00011180187525479007, + "loss": 3.3198, + "step": 104580 + }, + { + "epoch": 7.105924718032341, + "grad_norm": 1.3631517887115479, + "learning_rate": 0.00011175941024595733, + "loss": 3.5005, + "step": 104585 + }, + { + "epoch": 7.106264438103003, + "grad_norm": 0.930885910987854, + "learning_rate": 0.00011171694523712461, + "loss": 3.4713, + "step": 104590 + }, + { + "epoch": 7.106604158173665, + "grad_norm": 1.1657583713531494, + "learning_rate": 0.0001116744802282919, + "loss": 3.1651, + "step": 104595 + }, + { + "epoch": 7.106943878244326, + "grad_norm": 1.0181374549865723, + "learning_rate": 0.00011163201521945916, + "loss": 3.3506, + "step": 104600 + }, + { + "epoch": 7.1072835983149885, + "grad_norm": 1.1458723545074463, + "learning_rate": 0.00011158955021062644, + "loss": 3.6373, + "step": 104605 + }, + { + "epoch": 7.1076233183856505, + "grad_norm": 0.8473312854766846, + "learning_rate": 0.00011154708520179372, + "loss": 3.1067, + "step": 104610 + }, + { + "epoch": 7.107963038456312, + "grad_norm": 0.827388346195221, + "learning_rate": 0.00011150462019296101, + "loss": 3.3542, + "step": 104615 + }, + { + "epoch": 7.108302758526974, + "grad_norm": 1.1599130630493164, + "learning_rate": 0.00011146215518412828, + "loss": 3.4868, + "step": 104620 + }, + { + "epoch": 7.108642478597636, + "grad_norm": 0.8839167952537537, + "learning_rate": 0.00011141969017529556, + "loss": 3.2996, + "step": 104625 + }, + { + "epoch": 7.108982198668297, + "grad_norm": 0.8938928246498108, + "learning_rate": 0.00011137722516646284, + "loss": 3.4654, + "step": 104630 + }, + { + "epoch": 7.109321918738959, + "grad_norm": 1.1207026243209839, + "learning_rate": 0.00011133476015763011, + "loss": 3.3454, + "step": 104635 + }, + { + "epoch": 7.109661638809621, + "grad_norm": 0.9271222352981567, + "learning_rate": 0.00011129229514879739, + "loss": 3.4649, + "step": 104640 + }, + { + "epoch": 7.110001358880282, + "grad_norm": 0.9213207364082336, + "learning_rate": 0.00011124983013996468, + "loss": 3.2808, + "step": 104645 + }, + { + "epoch": 7.1103410789509445, + "grad_norm": 0.9870385527610779, + "learning_rate": 0.00011120736513113195, + "loss": 3.2436, + "step": 104650 + }, + { + "epoch": 7.1106807990216065, + "grad_norm": 1.0508153438568115, + "learning_rate": 0.00011116490012229923, + "loss": 3.2957, + "step": 104655 + }, + { + "epoch": 7.111020519092268, + "grad_norm": 0.9080775380134583, + "learning_rate": 0.00011112243511346651, + "loss": 3.531, + "step": 104660 + }, + { + "epoch": 7.11136023916293, + "grad_norm": 1.0877934694290161, + "learning_rate": 0.00011107997010463379, + "loss": 3.408, + "step": 104665 + }, + { + "epoch": 7.111699959233592, + "grad_norm": 0.8183296322822571, + "learning_rate": 0.00011103750509580106, + "loss": 3.5864, + "step": 104670 + }, + { + "epoch": 7.112039679304253, + "grad_norm": 1.3887633085250854, + "learning_rate": 0.00011099504008696834, + "loss": 3.6014, + "step": 104675 + }, + { + "epoch": 7.112379399374915, + "grad_norm": 0.9698776602745056, + "learning_rate": 0.00011095257507813563, + "loss": 3.4423, + "step": 104680 + }, + { + "epoch": 7.112719119445577, + "grad_norm": 0.9379838109016418, + "learning_rate": 0.0001109101100693029, + "loss": 3.5016, + "step": 104685 + }, + { + "epoch": 7.113058839516238, + "grad_norm": 0.9448671936988831, + "learning_rate": 0.00011086764506047018, + "loss": 3.3929, + "step": 104690 + }, + { + "epoch": 7.1133985595869005, + "grad_norm": 0.9891918897628784, + "learning_rate": 0.00011082518005163746, + "loss": 3.4392, + "step": 104695 + }, + { + "epoch": 7.113738279657563, + "grad_norm": 0.9893706440925598, + "learning_rate": 0.00011078271504280474, + "loss": 3.3825, + "step": 104700 + }, + { + "epoch": 7.114077999728224, + "grad_norm": 0.8882694840431213, + "learning_rate": 0.000110740250033972, + "loss": 3.5618, + "step": 104705 + }, + { + "epoch": 7.114417719798886, + "grad_norm": 1.1130719184875488, + "learning_rate": 0.00011069778502513928, + "loss": 3.3886, + "step": 104710 + }, + { + "epoch": 7.114757439869548, + "grad_norm": 0.823205828666687, + "learning_rate": 0.00011065532001630658, + "loss": 3.569, + "step": 104715 + }, + { + "epoch": 7.115097159940209, + "grad_norm": 0.8706493973731995, + "learning_rate": 0.00011061285500747384, + "loss": 3.2924, + "step": 104720 + }, + { + "epoch": 7.115436880010871, + "grad_norm": 0.9374810457229614, + "learning_rate": 0.00011057038999864112, + "loss": 3.106, + "step": 104725 + }, + { + "epoch": 7.115776600081533, + "grad_norm": 0.8858991861343384, + "learning_rate": 0.0001105279249898084, + "loss": 3.3948, + "step": 104730 + }, + { + "epoch": 7.116116320152194, + "grad_norm": 0.996624231338501, + "learning_rate": 0.00011048545998097567, + "loss": 3.2826, + "step": 104735 + }, + { + "epoch": 7.1164560402228565, + "grad_norm": 0.8946439623832703, + "learning_rate": 0.00011044299497214295, + "loss": 3.3864, + "step": 104740 + }, + { + "epoch": 7.116795760293519, + "grad_norm": 1.5515509843826294, + "learning_rate": 0.00011040052996331024, + "loss": 3.3677, + "step": 104745 + }, + { + "epoch": 7.11713548036418, + "grad_norm": 0.9665138125419617, + "learning_rate": 0.00011035806495447752, + "loss": 3.2952, + "step": 104750 + }, + { + "epoch": 7.117475200434842, + "grad_norm": 0.8526946902275085, + "learning_rate": 0.00011031559994564479, + "loss": 3.4143, + "step": 104755 + }, + { + "epoch": 7.117814920505504, + "grad_norm": 0.9987460374832153, + "learning_rate": 0.00011027313493681207, + "loss": 3.2476, + "step": 104760 + }, + { + "epoch": 7.118154640576165, + "grad_norm": 0.8227540850639343, + "learning_rate": 0.00011023066992797935, + "loss": 3.4039, + "step": 104765 + }, + { + "epoch": 7.118494360646827, + "grad_norm": 0.7653313875198364, + "learning_rate": 0.00011018820491914662, + "loss": 3.5478, + "step": 104770 + }, + { + "epoch": 7.118834080717488, + "grad_norm": 0.7579103112220764, + "learning_rate": 0.0001101457399103139, + "loss": 3.6004, + "step": 104775 + }, + { + "epoch": 7.11917380078815, + "grad_norm": 0.8750578761100769, + "learning_rate": 0.00011010327490148119, + "loss": 3.457, + "step": 104780 + }, + { + "epoch": 7.1195135208588125, + "grad_norm": 0.806609570980072, + "learning_rate": 0.00011006080989264847, + "loss": 3.4925, + "step": 104785 + }, + { + "epoch": 7.119853240929474, + "grad_norm": 0.823576033115387, + "learning_rate": 0.00011001834488381574, + "loss": 3.2609, + "step": 104790 + }, + { + "epoch": 7.120192961000136, + "grad_norm": 0.8302851915359497, + "learning_rate": 0.00010997587987498302, + "loss": 3.3943, + "step": 104795 + }, + { + "epoch": 7.120532681070798, + "grad_norm": 1.115756869316101, + "learning_rate": 0.0001099334148661503, + "loss": 3.4824, + "step": 104800 + }, + { + "epoch": 7.120872401141459, + "grad_norm": 0.755918562412262, + "learning_rate": 0.00010989094985731756, + "loss": 3.6285, + "step": 104805 + }, + { + "epoch": 7.121212121212121, + "grad_norm": 1.0858768224716187, + "learning_rate": 0.00010984848484848486, + "loss": 3.1491, + "step": 104810 + }, + { + "epoch": 7.121551841282783, + "grad_norm": 0.9142835736274719, + "learning_rate": 0.00010980601983965214, + "loss": 3.6279, + "step": 104815 + }, + { + "epoch": 7.121891561353444, + "grad_norm": 0.9842512607574463, + "learning_rate": 0.0001097635548308194, + "loss": 3.247, + "step": 104820 + }, + { + "epoch": 7.122231281424106, + "grad_norm": 0.8439558148384094, + "learning_rate": 0.00010972108982198668, + "loss": 3.2776, + "step": 104825 + }, + { + "epoch": 7.1225710014947685, + "grad_norm": 1.0940057039260864, + "learning_rate": 0.00010967862481315396, + "loss": 3.3462, + "step": 104830 + }, + { + "epoch": 7.12291072156543, + "grad_norm": 0.8542498350143433, + "learning_rate": 0.00010963615980432124, + "loss": 3.7535, + "step": 104835 + }, + { + "epoch": 7.123250441636092, + "grad_norm": 0.9584900736808777, + "learning_rate": 0.00010959369479548851, + "loss": 3.3254, + "step": 104840 + }, + { + "epoch": 7.123590161706754, + "grad_norm": 0.8507091403007507, + "learning_rate": 0.0001095512297866558, + "loss": 3.2815, + "step": 104845 + }, + { + "epoch": 7.123929881777415, + "grad_norm": 0.8372114896774292, + "learning_rate": 0.00010950876477782308, + "loss": 3.5363, + "step": 104850 + }, + { + "epoch": 7.124269601848077, + "grad_norm": 0.8371663093566895, + "learning_rate": 0.00010946629976899035, + "loss": 3.2965, + "step": 104855 + }, + { + "epoch": 7.124609321918739, + "grad_norm": 1.0005860328674316, + "learning_rate": 0.00010942383476015763, + "loss": 3.2428, + "step": 104860 + }, + { + "epoch": 7.1249490419894, + "grad_norm": 1.1143099069595337, + "learning_rate": 0.00010938136975132491, + "loss": 3.421, + "step": 104865 + }, + { + "epoch": 7.1252887620600625, + "grad_norm": 0.8251426815986633, + "learning_rate": 0.00010933890474249218, + "loss": 3.1957, + "step": 104870 + }, + { + "epoch": 7.1256284821307245, + "grad_norm": 1.026052474975586, + "learning_rate": 0.00010929643973365947, + "loss": 3.1803, + "step": 104875 + }, + { + "epoch": 7.125968202201386, + "grad_norm": 0.8286052346229553, + "learning_rate": 0.00010925397472482675, + "loss": 3.351, + "step": 104880 + }, + { + "epoch": 7.126307922272048, + "grad_norm": 0.9419455528259277, + "learning_rate": 0.00010921150971599403, + "loss": 3.5228, + "step": 104885 + }, + { + "epoch": 7.12664764234271, + "grad_norm": 0.8169238567352295, + "learning_rate": 0.0001091690447071613, + "loss": 3.6169, + "step": 104890 + }, + { + "epoch": 7.126987362413371, + "grad_norm": 0.6738440990447998, + "learning_rate": 0.00010912657969832858, + "loss": 3.1, + "step": 104895 + }, + { + "epoch": 7.127327082484033, + "grad_norm": 0.9724065065383911, + "learning_rate": 0.00010908411468949586, + "loss": 3.5274, + "step": 104900 + }, + { + "epoch": 7.127666802554695, + "grad_norm": 0.807737410068512, + "learning_rate": 0.00010904164968066312, + "loss": 3.618, + "step": 104905 + }, + { + "epoch": 7.128006522625356, + "grad_norm": 0.6784526705741882, + "learning_rate": 0.00010899918467183042, + "loss": 3.3454, + "step": 104910 + }, + { + "epoch": 7.1283462426960185, + "grad_norm": 1.134348750114441, + "learning_rate": 0.0001089567196629977, + "loss": 3.494, + "step": 104915 + }, + { + "epoch": 7.1286859627666805, + "grad_norm": 0.9215410351753235, + "learning_rate": 0.00010891425465416498, + "loss": 3.6111, + "step": 104920 + }, + { + "epoch": 7.129025682837342, + "grad_norm": 0.8379361033439636, + "learning_rate": 0.00010887178964533224, + "loss": 3.4924, + "step": 104925 + }, + { + "epoch": 7.129365402908004, + "grad_norm": 0.8353999257087708, + "learning_rate": 0.00010882932463649952, + "loss": 3.1546, + "step": 104930 + }, + { + "epoch": 7.129705122978666, + "grad_norm": 1.0775307416915894, + "learning_rate": 0.0001087868596276668, + "loss": 3.3997, + "step": 104935 + }, + { + "epoch": 7.130044843049327, + "grad_norm": 0.8716060519218445, + "learning_rate": 0.00010874439461883408, + "loss": 3.4409, + "step": 104940 + }, + { + "epoch": 7.130384563119989, + "grad_norm": 0.886695146560669, + "learning_rate": 0.00010870192961000136, + "loss": 3.491, + "step": 104945 + }, + { + "epoch": 7.130724283190651, + "grad_norm": 0.9418081641197205, + "learning_rate": 0.00010865946460116864, + "loss": 3.365, + "step": 104950 + }, + { + "epoch": 7.131064003261312, + "grad_norm": 1.0260639190673828, + "learning_rate": 0.00010861699959233591, + "loss": 3.5401, + "step": 104955 + }, + { + "epoch": 7.1314037233319745, + "grad_norm": 0.9025769233703613, + "learning_rate": 0.00010857453458350319, + "loss": 3.2549, + "step": 104960 + }, + { + "epoch": 7.1317434434026366, + "grad_norm": 0.9558852314949036, + "learning_rate": 0.00010853206957467047, + "loss": 3.3584, + "step": 104965 + }, + { + "epoch": 7.132083163473298, + "grad_norm": 0.8284828662872314, + "learning_rate": 0.00010848960456583775, + "loss": 3.446, + "step": 104970 + }, + { + "epoch": 7.13242288354396, + "grad_norm": 1.0207592248916626, + "learning_rate": 0.00010844713955700503, + "loss": 3.4668, + "step": 104975 + }, + { + "epoch": 7.132762603614622, + "grad_norm": 0.6995453238487244, + "learning_rate": 0.00010840467454817231, + "loss": 3.3827, + "step": 104980 + }, + { + "epoch": 7.133102323685283, + "grad_norm": 0.9572465419769287, + "learning_rate": 0.00010836220953933959, + "loss": 3.3614, + "step": 104985 + }, + { + "epoch": 7.133442043755945, + "grad_norm": 1.0266129970550537, + "learning_rate": 0.00010831974453050686, + "loss": 3.436, + "step": 104990 + }, + { + "epoch": 7.133781763826607, + "grad_norm": 0.7793362140655518, + "learning_rate": 0.00010827727952167414, + "loss": 3.2966, + "step": 104995 + }, + { + "epoch": 7.134121483897268, + "grad_norm": 1.287593960762024, + "learning_rate": 0.00010823481451284142, + "loss": 3.3833, + "step": 105000 + }, + { + "epoch": 7.1344612039679305, + "grad_norm": 1.0030570030212402, + "learning_rate": 0.00010819234950400871, + "loss": 3.3141, + "step": 105005 + }, + { + "epoch": 7.134800924038593, + "grad_norm": 1.4938615560531616, + "learning_rate": 0.00010814988449517598, + "loss": 3.4637, + "step": 105010 + }, + { + "epoch": 7.135140644109254, + "grad_norm": 0.6875920295715332, + "learning_rate": 0.00010810741948634326, + "loss": 3.3585, + "step": 105015 + }, + { + "epoch": 7.135480364179916, + "grad_norm": 0.9540504217147827, + "learning_rate": 0.00010806495447751054, + "loss": 3.2233, + "step": 105020 + }, + { + "epoch": 7.135820084250578, + "grad_norm": 0.9710580110549927, + "learning_rate": 0.0001080224894686778, + "loss": 3.3447, + "step": 105025 + }, + { + "epoch": 7.136159804321239, + "grad_norm": 1.286772608757019, + "learning_rate": 0.00010798002445984509, + "loss": 3.5892, + "step": 105030 + }, + { + "epoch": 7.136499524391901, + "grad_norm": 0.7953097820281982, + "learning_rate": 0.00010793755945101237, + "loss": 3.2817, + "step": 105035 + }, + { + "epoch": 7.136839244462563, + "grad_norm": 1.24703049659729, + "learning_rate": 0.00010789509444217965, + "loss": 3.38, + "step": 105040 + }, + { + "epoch": 7.137178964533224, + "grad_norm": 0.7659069895744324, + "learning_rate": 0.00010785262943334693, + "loss": 3.342, + "step": 105045 + }, + { + "epoch": 7.1375186846038865, + "grad_norm": 1.169379472732544, + "learning_rate": 0.0001078101644245142, + "loss": 3.6576, + "step": 105050 + }, + { + "epoch": 7.137858404674549, + "grad_norm": 1.168681263923645, + "learning_rate": 0.00010776769941568149, + "loss": 3.4254, + "step": 105055 + }, + { + "epoch": 7.13819812474521, + "grad_norm": 1.0372682809829712, + "learning_rate": 0.00010772523440684875, + "loss": 3.4579, + "step": 105060 + }, + { + "epoch": 7.138537844815872, + "grad_norm": 0.9886329770088196, + "learning_rate": 0.00010768276939801603, + "loss": 3.2702, + "step": 105065 + }, + { + "epoch": 7.138877564886534, + "grad_norm": 1.1953868865966797, + "learning_rate": 0.00010764030438918333, + "loss": 3.3236, + "step": 105070 + }, + { + "epoch": 7.139217284957195, + "grad_norm": 1.0586916208267212, + "learning_rate": 0.00010759783938035059, + "loss": 3.5445, + "step": 105075 + }, + { + "epoch": 7.139557005027857, + "grad_norm": 0.8925120830535889, + "learning_rate": 0.00010755537437151787, + "loss": 3.1074, + "step": 105080 + }, + { + "epoch": 7.139896725098519, + "grad_norm": 1.3355679512023926, + "learning_rate": 0.00010751290936268515, + "loss": 3.1222, + "step": 105085 + }, + { + "epoch": 7.14023644516918, + "grad_norm": 1.35747230052948, + "learning_rate": 0.00010747044435385243, + "loss": 3.4049, + "step": 105090 + }, + { + "epoch": 7.1405761652398425, + "grad_norm": 0.9668985605239868, + "learning_rate": 0.0001074279793450197, + "loss": 3.1466, + "step": 105095 + }, + { + "epoch": 7.140915885310504, + "grad_norm": 0.8496434688568115, + "learning_rate": 0.00010738551433618698, + "loss": 3.8114, + "step": 105100 + }, + { + "epoch": 7.141255605381166, + "grad_norm": 1.1547622680664062, + "learning_rate": 0.00010734304932735427, + "loss": 3.3867, + "step": 105105 + }, + { + "epoch": 7.141595325451828, + "grad_norm": 0.9061458110809326, + "learning_rate": 0.00010730058431852154, + "loss": 3.4054, + "step": 105110 + }, + { + "epoch": 7.141935045522489, + "grad_norm": 1.0058573484420776, + "learning_rate": 0.00010725811930968882, + "loss": 3.2273, + "step": 105115 + }, + { + "epoch": 7.142274765593151, + "grad_norm": 0.8792795538902283, + "learning_rate": 0.0001072156543008561, + "loss": 3.3067, + "step": 105120 + }, + { + "epoch": 7.142614485663813, + "grad_norm": 0.7967318892478943, + "learning_rate": 0.00010717318929202337, + "loss": 3.3808, + "step": 105125 + }, + { + "epoch": 7.142954205734474, + "grad_norm": 0.8754842877388, + "learning_rate": 0.00010713072428319065, + "loss": 3.3592, + "step": 105130 + }, + { + "epoch": 7.1432939258051364, + "grad_norm": 1.231602668762207, + "learning_rate": 0.00010708825927435793, + "loss": 3.3035, + "step": 105135 + }, + { + "epoch": 7.1436336458757985, + "grad_norm": 1.1309099197387695, + "learning_rate": 0.00010704579426552522, + "loss": 3.4353, + "step": 105140 + }, + { + "epoch": 7.14397336594646, + "grad_norm": 1.1117305755615234, + "learning_rate": 0.00010700332925669249, + "loss": 3.0555, + "step": 105145 + }, + { + "epoch": 7.144313086017122, + "grad_norm": 0.8909348249435425, + "learning_rate": 0.00010696086424785977, + "loss": 3.2119, + "step": 105150 + }, + { + "epoch": 7.144652806087784, + "grad_norm": 0.9070841073989868, + "learning_rate": 0.00010691839923902705, + "loss": 3.5853, + "step": 105155 + }, + { + "epoch": 7.144992526158445, + "grad_norm": 0.8125836849212646, + "learning_rate": 0.00010687593423019431, + "loss": 3.4184, + "step": 105160 + }, + { + "epoch": 7.145332246229107, + "grad_norm": 0.7945963144302368, + "learning_rate": 0.00010683346922136159, + "loss": 3.4367, + "step": 105165 + }, + { + "epoch": 7.145671966299769, + "grad_norm": 0.735683798789978, + "learning_rate": 0.00010679100421252889, + "loss": 3.6162, + "step": 105170 + }, + { + "epoch": 7.14601168637043, + "grad_norm": 0.9847286939620972, + "learning_rate": 0.00010674853920369617, + "loss": 3.3829, + "step": 105175 + }, + { + "epoch": 7.1463514064410925, + "grad_norm": 1.009589672088623, + "learning_rate": 0.00010670607419486343, + "loss": 3.3173, + "step": 105180 + }, + { + "epoch": 7.1466911265117545, + "grad_norm": 0.991890549659729, + "learning_rate": 0.00010666360918603071, + "loss": 3.3648, + "step": 105185 + }, + { + "epoch": 7.147030846582416, + "grad_norm": 0.828896164894104, + "learning_rate": 0.000106621144177198, + "loss": 3.5127, + "step": 105190 + }, + { + "epoch": 7.147370566653078, + "grad_norm": 1.257641077041626, + "learning_rate": 0.00010657867916836526, + "loss": 3.6097, + "step": 105195 + }, + { + "epoch": 7.14771028672374, + "grad_norm": 1.0393028259277344, + "learning_rate": 0.00010653621415953254, + "loss": 3.3493, + "step": 105200 + }, + { + "epoch": 7.148050006794401, + "grad_norm": 0.8000490069389343, + "learning_rate": 0.00010649374915069983, + "loss": 3.2549, + "step": 105205 + }, + { + "epoch": 7.148389726865063, + "grad_norm": 1.0089222192764282, + "learning_rate": 0.0001064512841418671, + "loss": 3.3653, + "step": 105210 + }, + { + "epoch": 7.148729446935725, + "grad_norm": 0.8408575057983398, + "learning_rate": 0.00010640881913303438, + "loss": 3.0937, + "step": 105215 + }, + { + "epoch": 7.149069167006386, + "grad_norm": 0.8415915966033936, + "learning_rate": 0.00010636635412420166, + "loss": 3.2231, + "step": 105220 + }, + { + "epoch": 7.1494088870770485, + "grad_norm": 0.9764803051948547, + "learning_rate": 0.00010632388911536894, + "loss": 3.5587, + "step": 105225 + }, + { + "epoch": 7.1497486071477105, + "grad_norm": 1.0565565824508667, + "learning_rate": 0.00010628142410653621, + "loss": 3.3567, + "step": 105230 + }, + { + "epoch": 7.150088327218372, + "grad_norm": 1.1094309091567993, + "learning_rate": 0.0001062389590977035, + "loss": 3.5132, + "step": 105235 + }, + { + "epoch": 7.150428047289034, + "grad_norm": 0.889115571975708, + "learning_rate": 0.00010619649408887078, + "loss": 3.2408, + "step": 105240 + }, + { + "epoch": 7.150767767359696, + "grad_norm": 0.8513706922531128, + "learning_rate": 0.00010615402908003805, + "loss": 3.5037, + "step": 105245 + }, + { + "epoch": 7.151107487430357, + "grad_norm": 0.7067617774009705, + "learning_rate": 0.00010611156407120533, + "loss": 3.5375, + "step": 105250 + }, + { + "epoch": 7.151447207501019, + "grad_norm": 0.9148487448692322, + "learning_rate": 0.00010606909906237261, + "loss": 3.5749, + "step": 105255 + }, + { + "epoch": 7.151786927571681, + "grad_norm": 1.18372642993927, + "learning_rate": 0.00010602663405353989, + "loss": 3.3871, + "step": 105260 + }, + { + "epoch": 7.152126647642342, + "grad_norm": 0.9945882558822632, + "learning_rate": 0.00010598416904470715, + "loss": 3.3549, + "step": 105265 + }, + { + "epoch": 7.1524663677130045, + "grad_norm": 0.73369300365448, + "learning_rate": 0.00010594170403587445, + "loss": 3.5699, + "step": 105270 + }, + { + "epoch": 7.1528060877836666, + "grad_norm": 0.9721579551696777, + "learning_rate": 0.00010589923902704173, + "loss": 3.4762, + "step": 105275 + }, + { + "epoch": 7.153145807854328, + "grad_norm": 0.9509632587432861, + "learning_rate": 0.000105856774018209, + "loss": 3.2965, + "step": 105280 + }, + { + "epoch": 7.15348552792499, + "grad_norm": 0.8079656958580017, + "learning_rate": 0.00010581430900937627, + "loss": 3.4285, + "step": 105285 + }, + { + "epoch": 7.153825247995652, + "grad_norm": 0.9730293154716492, + "learning_rate": 0.00010577184400054355, + "loss": 3.542, + "step": 105290 + }, + { + "epoch": 7.154164968066313, + "grad_norm": 1.0092912912368774, + "learning_rate": 0.00010572937899171082, + "loss": 3.4847, + "step": 105295 + }, + { + "epoch": 7.154504688136975, + "grad_norm": 0.7703951597213745, + "learning_rate": 0.00010568691398287811, + "loss": 3.6068, + "step": 105300 + }, + { + "epoch": 7.154844408207637, + "grad_norm": 0.8206380009651184, + "learning_rate": 0.0001056444489740454, + "loss": 3.3855, + "step": 105305 + }, + { + "epoch": 7.155184128278298, + "grad_norm": 0.9687917828559875, + "learning_rate": 0.00010560198396521267, + "loss": 3.6037, + "step": 105310 + }, + { + "epoch": 7.1555238483489605, + "grad_norm": 1.0661063194274902, + "learning_rate": 0.00010555951895637994, + "loss": 3.2723, + "step": 105315 + }, + { + "epoch": 7.155863568419623, + "grad_norm": 0.8910638093948364, + "learning_rate": 0.00010551705394754722, + "loss": 3.4851, + "step": 105320 + }, + { + "epoch": 7.156203288490284, + "grad_norm": 0.9043834805488586, + "learning_rate": 0.0001054745889387145, + "loss": 3.4169, + "step": 105325 + }, + { + "epoch": 7.156543008560946, + "grad_norm": 1.0247385501861572, + "learning_rate": 0.00010543212392988177, + "loss": 3.3444, + "step": 105330 + }, + { + "epoch": 7.156882728631608, + "grad_norm": 0.8655253648757935, + "learning_rate": 0.00010538965892104906, + "loss": 3.4105, + "step": 105335 + }, + { + "epoch": 7.157222448702269, + "grad_norm": 1.0728557109832764, + "learning_rate": 0.00010534719391221634, + "loss": 3.2814, + "step": 105340 + }, + { + "epoch": 7.157562168772931, + "grad_norm": 0.8208500146865845, + "learning_rate": 0.00010530472890338362, + "loss": 3.3092, + "step": 105345 + }, + { + "epoch": 7.157901888843593, + "grad_norm": 0.885455846786499, + "learning_rate": 0.00010526226389455089, + "loss": 3.5204, + "step": 105350 + }, + { + "epoch": 7.158241608914254, + "grad_norm": 1.1476936340332031, + "learning_rate": 0.00010521979888571817, + "loss": 3.385, + "step": 105355 + }, + { + "epoch": 7.1585813289849165, + "grad_norm": 0.9617382884025574, + "learning_rate": 0.00010517733387688545, + "loss": 3.3084, + "step": 105360 + }, + { + "epoch": 7.158921049055579, + "grad_norm": 0.8450812101364136, + "learning_rate": 0.00010513486886805273, + "loss": 3.4016, + "step": 105365 + }, + { + "epoch": 7.15926076912624, + "grad_norm": 0.9842977523803711, + "learning_rate": 0.00010509240385922001, + "loss": 3.3355, + "step": 105370 + }, + { + "epoch": 7.159600489196902, + "grad_norm": 0.9879416823387146, + "learning_rate": 0.00010504993885038729, + "loss": 3.5356, + "step": 105375 + }, + { + "epoch": 7.159940209267564, + "grad_norm": 0.851074755191803, + "learning_rate": 0.00010500747384155455, + "loss": 3.6506, + "step": 105380 + }, + { + "epoch": 7.160279929338225, + "grad_norm": 1.04993736743927, + "learning_rate": 0.00010496500883272184, + "loss": 3.5904, + "step": 105385 + }, + { + "epoch": 7.160619649408887, + "grad_norm": 0.8309604525566101, + "learning_rate": 0.00010492254382388912, + "loss": 3.493, + "step": 105390 + }, + { + "epoch": 7.160959369479549, + "grad_norm": 0.7210094928741455, + "learning_rate": 0.0001048800788150564, + "loss": 3.2946, + "step": 105395 + }, + { + "epoch": 7.16129908955021, + "grad_norm": 0.9488810300827026, + "learning_rate": 0.00010483761380622368, + "loss": 3.3975, + "step": 105400 + }, + { + "epoch": 7.1616388096208725, + "grad_norm": 0.7944637537002563, + "learning_rate": 0.00010479514879739096, + "loss": 3.1917, + "step": 105405 + }, + { + "epoch": 7.161978529691535, + "grad_norm": 1.0330379009246826, + "learning_rate": 0.00010475268378855824, + "loss": 3.423, + "step": 105410 + }, + { + "epoch": 7.162318249762196, + "grad_norm": 1.003043293952942, + "learning_rate": 0.0001047102187797255, + "loss": 3.2978, + "step": 105415 + }, + { + "epoch": 7.162657969832858, + "grad_norm": 0.9253548979759216, + "learning_rate": 0.00010466775377089278, + "loss": 3.3376, + "step": 105420 + }, + { + "epoch": 7.16299768990352, + "grad_norm": 1.206228494644165, + "learning_rate": 0.00010462528876206006, + "loss": 3.4082, + "step": 105425 + }, + { + "epoch": 7.163337409974181, + "grad_norm": 0.8775922060012817, + "learning_rate": 0.00010458282375322736, + "loss": 3.4785, + "step": 105430 + }, + { + "epoch": 7.163677130044843, + "grad_norm": 0.9363334774971008, + "learning_rate": 0.00010454035874439462, + "loss": 3.2104, + "step": 105435 + }, + { + "epoch": 7.164016850115505, + "grad_norm": 0.9033123254776001, + "learning_rate": 0.0001044978937355619, + "loss": 3.4611, + "step": 105440 + }, + { + "epoch": 7.1643565701861665, + "grad_norm": 0.9094349145889282, + "learning_rate": 0.00010445542872672918, + "loss": 3.4015, + "step": 105445 + }, + { + "epoch": 7.1646962902568285, + "grad_norm": 0.9718388319015503, + "learning_rate": 0.00010441296371789645, + "loss": 3.4342, + "step": 105450 + }, + { + "epoch": 7.165036010327491, + "grad_norm": 0.9820426106452942, + "learning_rate": 0.00010437049870906373, + "loss": 3.2946, + "step": 105455 + }, + { + "epoch": 7.165375730398152, + "grad_norm": 0.9477027058601379, + "learning_rate": 0.00010432803370023101, + "loss": 3.3686, + "step": 105460 + }, + { + "epoch": 7.165715450468814, + "grad_norm": 0.9578188061714172, + "learning_rate": 0.00010428556869139829, + "loss": 2.9678, + "step": 105465 + }, + { + "epoch": 7.166055170539475, + "grad_norm": 1.0504149198532104, + "learning_rate": 0.00010424310368256557, + "loss": 3.2768, + "step": 105470 + }, + { + "epoch": 7.166394890610137, + "grad_norm": 1.0200988054275513, + "learning_rate": 0.00010420063867373285, + "loss": 3.4482, + "step": 105475 + }, + { + "epoch": 7.166734610680799, + "grad_norm": 0.9812706708908081, + "learning_rate": 0.00010415817366490013, + "loss": 3.2025, + "step": 105480 + }, + { + "epoch": 7.16707433075146, + "grad_norm": 0.9808008074760437, + "learning_rate": 0.0001041157086560674, + "loss": 3.1228, + "step": 105485 + }, + { + "epoch": 7.1674140508221225, + "grad_norm": 0.841637134552002, + "learning_rate": 0.00010407324364723468, + "loss": 3.3186, + "step": 105490 + }, + { + "epoch": 7.1677537708927845, + "grad_norm": 0.8538141250610352, + "learning_rate": 0.00010403077863840196, + "loss": 3.4536, + "step": 105495 + }, + { + "epoch": 7.168093490963446, + "grad_norm": 0.9855509400367737, + "learning_rate": 0.00010398831362956924, + "loss": 3.5075, + "step": 105500 + }, + { + "epoch": 7.168433211034108, + "grad_norm": 0.8091641664505005, + "learning_rate": 0.00010394584862073652, + "loss": 3.4989, + "step": 105505 + }, + { + "epoch": 7.16877293110477, + "grad_norm": 1.2860428094863892, + "learning_rate": 0.0001039033836119038, + "loss": 3.1121, + "step": 105510 + }, + { + "epoch": 7.169112651175431, + "grad_norm": 1.2021417617797852, + "learning_rate": 0.00010386091860307108, + "loss": 3.3399, + "step": 105515 + }, + { + "epoch": 7.169452371246093, + "grad_norm": 0.9276304244995117, + "learning_rate": 0.00010381845359423834, + "loss": 3.3362, + "step": 105520 + }, + { + "epoch": 7.169792091316755, + "grad_norm": 0.896285891532898, + "learning_rate": 0.00010377598858540562, + "loss": 3.5677, + "step": 105525 + }, + { + "epoch": 7.170131811387416, + "grad_norm": 0.895337700843811, + "learning_rate": 0.00010373352357657292, + "loss": 3.1959, + "step": 105530 + }, + { + "epoch": 7.1704715314580785, + "grad_norm": 0.7732467651367188, + "learning_rate": 0.00010369105856774018, + "loss": 3.4009, + "step": 105535 + }, + { + "epoch": 7.1708112515287405, + "grad_norm": 1.119215726852417, + "learning_rate": 0.00010364859355890746, + "loss": 3.4642, + "step": 105540 + }, + { + "epoch": 7.171150971599402, + "grad_norm": 1.2033345699310303, + "learning_rate": 0.00010360612855007474, + "loss": 3.6296, + "step": 105545 + }, + { + "epoch": 7.171490691670064, + "grad_norm": 0.8673357367515564, + "learning_rate": 0.00010356366354124201, + "loss": 3.4841, + "step": 105550 + }, + { + "epoch": 7.171830411740726, + "grad_norm": 0.7636723518371582, + "learning_rate": 0.00010352119853240929, + "loss": 3.201, + "step": 105555 + }, + { + "epoch": 7.172170131811387, + "grad_norm": 1.3396176099777222, + "learning_rate": 0.00010347873352357657, + "loss": 3.4875, + "step": 105560 + }, + { + "epoch": 7.172509851882049, + "grad_norm": 0.9820488095283508, + "learning_rate": 0.00010343626851474386, + "loss": 3.5239, + "step": 105565 + }, + { + "epoch": 7.172849571952711, + "grad_norm": 1.6716663837432861, + "learning_rate": 0.00010339380350591113, + "loss": 3.5053, + "step": 105570 + }, + { + "epoch": 7.173189292023372, + "grad_norm": 0.8914536237716675, + "learning_rate": 0.00010335133849707841, + "loss": 3.2132, + "step": 105575 + }, + { + "epoch": 7.1735290120940345, + "grad_norm": 1.053883671760559, + "learning_rate": 0.00010330887348824569, + "loss": 3.3312, + "step": 105580 + }, + { + "epoch": 7.173868732164697, + "grad_norm": 1.0493875741958618, + "learning_rate": 0.00010326640847941296, + "loss": 3.5265, + "step": 105585 + }, + { + "epoch": 7.174208452235358, + "grad_norm": 0.7670869827270508, + "learning_rate": 0.00010322394347058024, + "loss": 3.3693, + "step": 105590 + }, + { + "epoch": 7.17454817230602, + "grad_norm": 0.8833093047142029, + "learning_rate": 0.00010318147846174753, + "loss": 3.5687, + "step": 105595 + }, + { + "epoch": 7.174887892376682, + "grad_norm": 0.8686231970787048, + "learning_rate": 0.00010313901345291481, + "loss": 3.4884, + "step": 105600 + }, + { + "epoch": 7.175227612447343, + "grad_norm": 1.0736154317855835, + "learning_rate": 0.00010309654844408208, + "loss": 3.7539, + "step": 105605 + }, + { + "epoch": 7.175567332518005, + "grad_norm": 1.0078892707824707, + "learning_rate": 0.00010305408343524936, + "loss": 3.5229, + "step": 105610 + }, + { + "epoch": 7.175907052588667, + "grad_norm": 1.0056298971176147, + "learning_rate": 0.00010301161842641664, + "loss": 3.4651, + "step": 105615 + }, + { + "epoch": 7.176246772659328, + "grad_norm": 0.8983667492866516, + "learning_rate": 0.0001029691534175839, + "loss": 3.2094, + "step": 105620 + }, + { + "epoch": 7.1765864927299905, + "grad_norm": 0.8735098242759705, + "learning_rate": 0.00010292668840875118, + "loss": 3.0979, + "step": 105625 + }, + { + "epoch": 7.176926212800653, + "grad_norm": 0.9093451499938965, + "learning_rate": 0.00010288422339991848, + "loss": 3.4822, + "step": 105630 + }, + { + "epoch": 7.177265932871314, + "grad_norm": 0.9553730487823486, + "learning_rate": 0.00010284175839108574, + "loss": 3.4281, + "step": 105635 + }, + { + "epoch": 7.177605652941976, + "grad_norm": 1.3950880765914917, + "learning_rate": 0.00010279929338225302, + "loss": 3.3303, + "step": 105640 + }, + { + "epoch": 7.177945373012638, + "grad_norm": 0.8582971692085266, + "learning_rate": 0.0001027568283734203, + "loss": 3.3405, + "step": 105645 + }, + { + "epoch": 7.178285093083299, + "grad_norm": 0.9516609311103821, + "learning_rate": 0.00010271436336458758, + "loss": 3.8566, + "step": 105650 + }, + { + "epoch": 7.178624813153961, + "grad_norm": 0.8130750060081482, + "learning_rate": 0.00010267189835575485, + "loss": 3.418, + "step": 105655 + }, + { + "epoch": 7.178964533224623, + "grad_norm": 0.8445088267326355, + "learning_rate": 0.00010262943334692214, + "loss": 3.3701, + "step": 105660 + }, + { + "epoch": 7.179304253295284, + "grad_norm": 1.3196977376937866, + "learning_rate": 0.00010258696833808942, + "loss": 3.4965, + "step": 105665 + }, + { + "epoch": 7.1796439733659465, + "grad_norm": 0.8116183280944824, + "learning_rate": 0.00010254450332925669, + "loss": 3.2593, + "step": 105670 + }, + { + "epoch": 7.179983693436609, + "grad_norm": 0.838330864906311, + "learning_rate": 0.00010250203832042397, + "loss": 3.4549, + "step": 105675 + }, + { + "epoch": 7.18032341350727, + "grad_norm": 0.6532452702522278, + "learning_rate": 0.00010245957331159125, + "loss": 3.4396, + "step": 105680 + }, + { + "epoch": 7.180663133577932, + "grad_norm": 0.7685763835906982, + "learning_rate": 0.00010241710830275853, + "loss": 3.523, + "step": 105685 + }, + { + "epoch": 7.181002853648594, + "grad_norm": 1.1238694190979004, + "learning_rate": 0.0001023746432939258, + "loss": 3.802, + "step": 105690 + }, + { + "epoch": 7.181342573719255, + "grad_norm": 0.9441054463386536, + "learning_rate": 0.00010233217828509309, + "loss": 3.3133, + "step": 105695 + }, + { + "epoch": 7.181682293789917, + "grad_norm": 0.8843408823013306, + "learning_rate": 0.00010228971327626037, + "loss": 3.5703, + "step": 105700 + }, + { + "epoch": 7.182022013860579, + "grad_norm": 1.0619723796844482, + "learning_rate": 0.00010224724826742764, + "loss": 3.1569, + "step": 105705 + }, + { + "epoch": 7.18236173393124, + "grad_norm": 1.0241400003433228, + "learning_rate": 0.00010220478325859492, + "loss": 3.3514, + "step": 105710 + }, + { + "epoch": 7.1827014540019025, + "grad_norm": 1.2680506706237793, + "learning_rate": 0.0001021623182497622, + "loss": 3.4547, + "step": 105715 + }, + { + "epoch": 7.183041174072565, + "grad_norm": 1.0607249736785889, + "learning_rate": 0.00010211985324092946, + "loss": 3.3873, + "step": 105720 + }, + { + "epoch": 7.183380894143226, + "grad_norm": 1.047842264175415, + "learning_rate": 0.00010207738823209676, + "loss": 3.2022, + "step": 105725 + }, + { + "epoch": 7.183720614213888, + "grad_norm": 0.9224300980567932, + "learning_rate": 0.00010203492322326404, + "loss": 3.3124, + "step": 105730 + }, + { + "epoch": 7.18406033428455, + "grad_norm": 1.0412085056304932, + "learning_rate": 0.00010199245821443132, + "loss": 3.1995, + "step": 105735 + }, + { + "epoch": 7.184400054355211, + "grad_norm": 0.8879486918449402, + "learning_rate": 0.00010194999320559858, + "loss": 3.4181, + "step": 105740 + }, + { + "epoch": 7.184739774425873, + "grad_norm": 1.7949129343032837, + "learning_rate": 0.00010190752819676587, + "loss": 3.097, + "step": 105745 + }, + { + "epoch": 7.185079494496535, + "grad_norm": 1.0383362770080566, + "learning_rate": 0.00010186506318793315, + "loss": 3.2474, + "step": 105750 + }, + { + "epoch": 7.1854192145671965, + "grad_norm": 0.8477770686149597, + "learning_rate": 0.00010182259817910041, + "loss": 3.5017, + "step": 105755 + }, + { + "epoch": 7.1857589346378585, + "grad_norm": 0.8353756070137024, + "learning_rate": 0.0001017801331702677, + "loss": 3.4038, + "step": 105760 + }, + { + "epoch": 7.186098654708521, + "grad_norm": 1.0113414525985718, + "learning_rate": 0.00010173766816143499, + "loss": 3.3444, + "step": 105765 + }, + { + "epoch": 7.186438374779182, + "grad_norm": 1.0477070808410645, + "learning_rate": 0.00010169520315260227, + "loss": 3.214, + "step": 105770 + }, + { + "epoch": 7.186778094849844, + "grad_norm": 0.7295812368392944, + "learning_rate": 0.00010165273814376953, + "loss": 3.2966, + "step": 105775 + }, + { + "epoch": 7.187117814920505, + "grad_norm": 1.0770543813705444, + "learning_rate": 0.00010161027313493681, + "loss": 3.3973, + "step": 105780 + }, + { + "epoch": 7.187457534991167, + "grad_norm": 0.8865898251533508, + "learning_rate": 0.00010156780812610409, + "loss": 3.4487, + "step": 105785 + }, + { + "epoch": 7.187797255061829, + "grad_norm": 0.775901198387146, + "learning_rate": 0.00010152534311727136, + "loss": 3.4726, + "step": 105790 + }, + { + "epoch": 7.18813697513249, + "grad_norm": 1.2249398231506348, + "learning_rate": 0.00010148287810843865, + "loss": 3.3754, + "step": 105795 + }, + { + "epoch": 7.1884766952031525, + "grad_norm": 0.9324944615364075, + "learning_rate": 0.00010144041309960593, + "loss": 3.1932, + "step": 105800 + }, + { + "epoch": 7.1888164152738145, + "grad_norm": 1.0449330806732178, + "learning_rate": 0.0001013979480907732, + "loss": 3.4025, + "step": 105805 + }, + { + "epoch": 7.189156135344476, + "grad_norm": 0.7634008526802063, + "learning_rate": 0.00010135548308194048, + "loss": 3.1747, + "step": 105810 + }, + { + "epoch": 7.189495855415138, + "grad_norm": 0.8755817413330078, + "learning_rate": 0.00010131301807310776, + "loss": 3.6406, + "step": 105815 + }, + { + "epoch": 7.1898355754858, + "grad_norm": 0.8032550811767578, + "learning_rate": 0.00010127055306427504, + "loss": 3.4333, + "step": 105820 + }, + { + "epoch": 7.190175295556461, + "grad_norm": 0.8995932936668396, + "learning_rate": 0.00010122808805544232, + "loss": 3.597, + "step": 105825 + }, + { + "epoch": 7.190515015627123, + "grad_norm": 0.9220735430717468, + "learning_rate": 0.0001011856230466096, + "loss": 3.5687, + "step": 105830 + }, + { + "epoch": 7.190854735697785, + "grad_norm": 1.0982773303985596, + "learning_rate": 0.00010114315803777688, + "loss": 3.2582, + "step": 105835 + }, + { + "epoch": 7.191194455768446, + "grad_norm": 0.9637282490730286, + "learning_rate": 0.00010110069302894415, + "loss": 3.3163, + "step": 105840 + }, + { + "epoch": 7.1915341758391085, + "grad_norm": 0.7943386435508728, + "learning_rate": 0.00010105822802011143, + "loss": 3.556, + "step": 105845 + }, + { + "epoch": 7.1918738959097706, + "grad_norm": 0.8871826529502869, + "learning_rate": 0.0001010157630112787, + "loss": 3.4607, + "step": 105850 + }, + { + "epoch": 7.192213615980432, + "grad_norm": 1.0352011919021606, + "learning_rate": 0.000100973298002446, + "loss": 3.5077, + "step": 105855 + }, + { + "epoch": 7.192553336051094, + "grad_norm": 0.9489971995353699, + "learning_rate": 0.00010093083299361327, + "loss": 3.2003, + "step": 105860 + }, + { + "epoch": 7.192893056121756, + "grad_norm": 1.0563486814498901, + "learning_rate": 0.00010088836798478055, + "loss": 3.4384, + "step": 105865 + }, + { + "epoch": 7.193232776192417, + "grad_norm": 0.773999035358429, + "learning_rate": 0.00010084590297594783, + "loss": 3.1967, + "step": 105870 + }, + { + "epoch": 7.193572496263079, + "grad_norm": 0.9139823317527771, + "learning_rate": 0.00010080343796711509, + "loss": 3.2587, + "step": 105875 + }, + { + "epoch": 7.193912216333741, + "grad_norm": 0.9232065677642822, + "learning_rate": 0.00010076097295828237, + "loss": 3.2331, + "step": 105880 + }, + { + "epoch": 7.194251936404402, + "grad_norm": 1.127182126045227, + "learning_rate": 0.00010071850794944965, + "loss": 3.3357, + "step": 105885 + }, + { + "epoch": 7.1945916564750645, + "grad_norm": 1.0326136350631714, + "learning_rate": 0.00010067604294061693, + "loss": 3.4213, + "step": 105890 + }, + { + "epoch": 7.194931376545727, + "grad_norm": 1.036206841468811, + "learning_rate": 0.00010063357793178421, + "loss": 3.5765, + "step": 105895 + }, + { + "epoch": 7.195271096616388, + "grad_norm": 0.8369729518890381, + "learning_rate": 0.00010059111292295149, + "loss": 3.4772, + "step": 105900 + }, + { + "epoch": 7.19561081668705, + "grad_norm": 0.94975745677948, + "learning_rate": 0.00010054864791411877, + "loss": 3.304, + "step": 105905 + }, + { + "epoch": 7.195950536757712, + "grad_norm": 0.9315518736839294, + "learning_rate": 0.00010050618290528604, + "loss": 3.3513, + "step": 105910 + }, + { + "epoch": 7.196290256828373, + "grad_norm": 0.8460417985916138, + "learning_rate": 0.00010046371789645332, + "loss": 3.434, + "step": 105915 + }, + { + "epoch": 7.196629976899035, + "grad_norm": 0.7528766393661499, + "learning_rate": 0.0001004212528876206, + "loss": 3.3003, + "step": 105920 + }, + { + "epoch": 7.196969696969697, + "grad_norm": 0.874121367931366, + "learning_rate": 0.00010037878787878788, + "loss": 3.3784, + "step": 105925 + }, + { + "epoch": 7.197309417040358, + "grad_norm": 1.1277674436569214, + "learning_rate": 0.00010033632286995516, + "loss": 3.1861, + "step": 105930 + }, + { + "epoch": 7.1976491371110205, + "grad_norm": 1.1928422451019287, + "learning_rate": 0.00010029385786112244, + "loss": 3.3464, + "step": 105935 + }, + { + "epoch": 7.197988857181683, + "grad_norm": 0.910466730594635, + "learning_rate": 0.00010025139285228972, + "loss": 3.703, + "step": 105940 + }, + { + "epoch": 7.198328577252344, + "grad_norm": 0.7978136539459229, + "learning_rate": 0.00010020892784345699, + "loss": 3.3941, + "step": 105945 + }, + { + "epoch": 7.198668297323006, + "grad_norm": 0.8590202927589417, + "learning_rate": 0.00010016646283462427, + "loss": 3.7392, + "step": 105950 + }, + { + "epoch": 7.199008017393668, + "grad_norm": 0.9960775971412659, + "learning_rate": 0.00010012399782579156, + "loss": 3.3078, + "step": 105955 + }, + { + "epoch": 7.199347737464329, + "grad_norm": 0.8982253074645996, + "learning_rate": 0.00010008153281695883, + "loss": 3.4699, + "step": 105960 + }, + { + "epoch": 7.199687457534991, + "grad_norm": 1.279239296913147, + "learning_rate": 0.00010003906780812611, + "loss": 3.2697, + "step": 105965 + }, + { + "epoch": 7.200027177605653, + "grad_norm": 1.0068720579147339, + "learning_rate": 9.999660279929339e-05, + "loss": 3.5486, + "step": 105970 + }, + { + "epoch": 7.200366897676314, + "grad_norm": 0.7631829977035522, + "learning_rate": 9.995413779046065e-05, + "loss": 3.4014, + "step": 105975 + }, + { + "epoch": 7.2007066177469765, + "grad_norm": 1.2071956396102905, + "learning_rate": 9.991167278162793e-05, + "loss": 3.4669, + "step": 105980 + }, + { + "epoch": 7.201046337817639, + "grad_norm": 0.8869341015815735, + "learning_rate": 9.986920777279521e-05, + "loss": 3.236, + "step": 105985 + }, + { + "epoch": 7.2013860578883, + "grad_norm": 0.9251505732536316, + "learning_rate": 9.982674276396251e-05, + "loss": 3.3596, + "step": 105990 + }, + { + "epoch": 7.201725777958962, + "grad_norm": 0.7432532906532288, + "learning_rate": 9.978427775512977e-05, + "loss": 3.3557, + "step": 105995 + }, + { + "epoch": 7.202065498029624, + "grad_norm": 1.1761373281478882, + "learning_rate": 9.974181274629705e-05, + "loss": 3.4384, + "step": 106000 + }, + { + "epoch": 7.202405218100285, + "grad_norm": 0.7667121887207031, + "learning_rate": 9.969934773746433e-05, + "loss": 3.4549, + "step": 106005 + }, + { + "epoch": 7.202744938170947, + "grad_norm": 1.0176739692687988, + "learning_rate": 9.96568827286316e-05, + "loss": 3.2587, + "step": 106010 + }, + { + "epoch": 7.203084658241609, + "grad_norm": 0.8901645541191101, + "learning_rate": 9.961441771979888e-05, + "loss": 3.548, + "step": 106015 + }, + { + "epoch": 7.2034243783122704, + "grad_norm": 1.0037531852722168, + "learning_rate": 9.957195271096617e-05, + "loss": 3.4532, + "step": 106020 + }, + { + "epoch": 7.2037640983829325, + "grad_norm": 0.8404086232185364, + "learning_rate": 9.952948770213345e-05, + "loss": 3.2965, + "step": 106025 + }, + { + "epoch": 7.204103818453595, + "grad_norm": 0.8060367107391357, + "learning_rate": 9.948702269330072e-05, + "loss": 3.4246, + "step": 106030 + }, + { + "epoch": 7.204443538524256, + "grad_norm": 1.0940654277801514, + "learning_rate": 9.9444557684468e-05, + "loss": 3.5824, + "step": 106035 + }, + { + "epoch": 7.204783258594918, + "grad_norm": 0.9696595668792725, + "learning_rate": 9.940209267563528e-05, + "loss": 3.0723, + "step": 106040 + }, + { + "epoch": 7.20512297866558, + "grad_norm": 0.9352137446403503, + "learning_rate": 9.935962766680255e-05, + "loss": 3.495, + "step": 106045 + }, + { + "epoch": 7.205462698736241, + "grad_norm": 0.9445629715919495, + "learning_rate": 9.931716265796983e-05, + "loss": 3.5773, + "step": 106050 + }, + { + "epoch": 7.205802418806903, + "grad_norm": 0.7802602052688599, + "learning_rate": 9.927469764913712e-05, + "loss": 3.4299, + "step": 106055 + }, + { + "epoch": 7.206142138877565, + "grad_norm": 0.8819277882575989, + "learning_rate": 9.923223264030439e-05, + "loss": 3.6914, + "step": 106060 + }, + { + "epoch": 7.2064818589482265, + "grad_norm": 0.9665668606758118, + "learning_rate": 9.918976763147167e-05, + "loss": 3.416, + "step": 106065 + }, + { + "epoch": 7.2068215790188885, + "grad_norm": 0.7506605386734009, + "learning_rate": 9.914730262263895e-05, + "loss": 3.5296, + "step": 106070 + }, + { + "epoch": 7.207161299089551, + "grad_norm": 0.8825567960739136, + "learning_rate": 9.910483761380623e-05, + "loss": 3.3014, + "step": 106075 + }, + { + "epoch": 7.207501019160212, + "grad_norm": 0.9029261469841003, + "learning_rate": 9.90623726049735e-05, + "loss": 3.3367, + "step": 106080 + }, + { + "epoch": 7.207840739230874, + "grad_norm": 0.9936944842338562, + "learning_rate": 9.901990759614079e-05, + "loss": 3.5106, + "step": 106085 + }, + { + "epoch": 7.208180459301536, + "grad_norm": 1.4346463680267334, + "learning_rate": 9.897744258730807e-05, + "loss": 3.2336, + "step": 106090 + }, + { + "epoch": 7.208520179372197, + "grad_norm": 1.3208298683166504, + "learning_rate": 9.893497757847533e-05, + "loss": 3.4608, + "step": 106095 + }, + { + "epoch": 7.208859899442859, + "grad_norm": 1.0012662410736084, + "learning_rate": 9.889251256964261e-05, + "loss": 3.6374, + "step": 106100 + }, + { + "epoch": 7.209199619513521, + "grad_norm": 0.9279007315635681, + "learning_rate": 9.88500475608099e-05, + "loss": 3.5553, + "step": 106105 + }, + { + "epoch": 7.2095393395841825, + "grad_norm": 0.7378093600273132, + "learning_rate": 9.880758255197718e-05, + "loss": 3.5479, + "step": 106110 + }, + { + "epoch": 7.2098790596548445, + "grad_norm": 0.9161143898963928, + "learning_rate": 9.876511754314444e-05, + "loss": 3.2313, + "step": 106115 + }, + { + "epoch": 7.210218779725507, + "grad_norm": 0.9537749290466309, + "learning_rate": 9.872265253431174e-05, + "loss": 3.5429, + "step": 106120 + }, + { + "epoch": 7.210558499796168, + "grad_norm": 0.9425577521324158, + "learning_rate": 9.868018752547902e-05, + "loss": 3.5929, + "step": 106125 + }, + { + "epoch": 7.21089821986683, + "grad_norm": 0.8596927523612976, + "learning_rate": 9.863772251664628e-05, + "loss": 3.3996, + "step": 106130 + }, + { + "epoch": 7.211237939937492, + "grad_norm": 0.9963798522949219, + "learning_rate": 9.859525750781356e-05, + "loss": 3.3819, + "step": 106135 + }, + { + "epoch": 7.211577660008153, + "grad_norm": 0.9810368418693542, + "learning_rate": 9.855279249898084e-05, + "loss": 3.3023, + "step": 106140 + }, + { + "epoch": 7.211917380078815, + "grad_norm": 0.8801648616790771, + "learning_rate": 9.851032749014811e-05, + "loss": 3.2993, + "step": 106145 + }, + { + "epoch": 7.212257100149476, + "grad_norm": 0.7867178916931152, + "learning_rate": 9.84678624813154e-05, + "loss": 3.3458, + "step": 106150 + }, + { + "epoch": 7.2125968202201385, + "grad_norm": 0.7964124083518982, + "learning_rate": 9.842539747248268e-05, + "loss": 3.4289, + "step": 106155 + }, + { + "epoch": 7.212936540290801, + "grad_norm": 1.0041835308074951, + "learning_rate": 9.838293246364996e-05, + "loss": 3.287, + "step": 106160 + }, + { + "epoch": 7.213276260361462, + "grad_norm": 0.9418566823005676, + "learning_rate": 9.834046745481723e-05, + "loss": 3.375, + "step": 106165 + }, + { + "epoch": 7.213615980432124, + "grad_norm": 0.7669152021408081, + "learning_rate": 9.829800244598451e-05, + "loss": 3.4298, + "step": 106170 + }, + { + "epoch": 7.213955700502786, + "grad_norm": 0.8807435035705566, + "learning_rate": 9.825553743715179e-05, + "loss": 3.2204, + "step": 106175 + }, + { + "epoch": 7.214295420573447, + "grad_norm": 1.0453490018844604, + "learning_rate": 9.821307242831906e-05, + "loss": 3.3506, + "step": 106180 + }, + { + "epoch": 7.214635140644109, + "grad_norm": 0.8652565479278564, + "learning_rate": 9.817060741948635e-05, + "loss": 3.4324, + "step": 106185 + }, + { + "epoch": 7.214974860714771, + "grad_norm": 0.8249659538269043, + "learning_rate": 9.812814241065363e-05, + "loss": 2.9769, + "step": 106190 + }, + { + "epoch": 7.215314580785432, + "grad_norm": 0.8028268814086914, + "learning_rate": 9.808567740182091e-05, + "loss": 3.3341, + "step": 106195 + }, + { + "epoch": 7.2156543008560945, + "grad_norm": 0.8075895309448242, + "learning_rate": 9.804321239298818e-05, + "loss": 3.2201, + "step": 106200 + }, + { + "epoch": 7.215994020926757, + "grad_norm": 0.7083885073661804, + "learning_rate": 9.800074738415546e-05, + "loss": 3.6813, + "step": 106205 + }, + { + "epoch": 7.216333740997418, + "grad_norm": 0.9383876323699951, + "learning_rate": 9.795828237532274e-05, + "loss": 3.34, + "step": 106210 + }, + { + "epoch": 7.21667346106808, + "grad_norm": 0.9488817453384399, + "learning_rate": 9.791581736649e-05, + "loss": 3.2611, + "step": 106215 + }, + { + "epoch": 7.217013181138742, + "grad_norm": 1.3868870735168457, + "learning_rate": 9.78733523576573e-05, + "loss": 3.2137, + "step": 106220 + }, + { + "epoch": 7.217352901209403, + "grad_norm": 1.0139656066894531, + "learning_rate": 9.783088734882458e-05, + "loss": 3.2422, + "step": 106225 + }, + { + "epoch": 7.217692621280065, + "grad_norm": 1.1859935522079468, + "learning_rate": 9.778842233999184e-05, + "loss": 3.642, + "step": 106230 + }, + { + "epoch": 7.218032341350727, + "grad_norm": 1.0873631238937378, + "learning_rate": 9.774595733115912e-05, + "loss": 3.5775, + "step": 106235 + }, + { + "epoch": 7.218372061421388, + "grad_norm": 0.9741585850715637, + "learning_rate": 9.77034923223264e-05, + "loss": 3.3283, + "step": 106240 + }, + { + "epoch": 7.2187117814920505, + "grad_norm": 1.202450156211853, + "learning_rate": 9.766102731349368e-05, + "loss": 3.0036, + "step": 106245 + }, + { + "epoch": 7.219051501562713, + "grad_norm": 1.305171012878418, + "learning_rate": 9.761856230466096e-05, + "loss": 3.3421, + "step": 106250 + }, + { + "epoch": 7.219391221633374, + "grad_norm": 0.9685468077659607, + "learning_rate": 9.757609729582824e-05, + "loss": 3.4182, + "step": 106255 + }, + { + "epoch": 7.219730941704036, + "grad_norm": 0.8622447848320007, + "learning_rate": 9.753363228699552e-05, + "loss": 3.5222, + "step": 106260 + }, + { + "epoch": 7.220070661774698, + "grad_norm": 1.420955777168274, + "learning_rate": 9.749116727816279e-05, + "loss": 3.5484, + "step": 106265 + }, + { + "epoch": 7.220410381845359, + "grad_norm": 0.9050021767616272, + "learning_rate": 9.744870226933007e-05, + "loss": 3.406, + "step": 106270 + }, + { + "epoch": 7.220750101916021, + "grad_norm": 1.005345106124878, + "learning_rate": 9.740623726049735e-05, + "loss": 3.4497, + "step": 106275 + }, + { + "epoch": 7.221089821986683, + "grad_norm": 1.1618106365203857, + "learning_rate": 9.736377225166464e-05, + "loss": 3.5338, + "step": 106280 + }, + { + "epoch": 7.221429542057344, + "grad_norm": 1.1702711582183838, + "learning_rate": 9.732130724283191e-05, + "loss": 3.5834, + "step": 106285 + }, + { + "epoch": 7.2217692621280065, + "grad_norm": 1.5039576292037964, + "learning_rate": 9.727884223399919e-05, + "loss": 3.2869, + "step": 106290 + }, + { + "epoch": 7.222108982198669, + "grad_norm": 0.9006040096282959, + "learning_rate": 9.723637722516647e-05, + "loss": 3.5822, + "step": 106295 + }, + { + "epoch": 7.22244870226933, + "grad_norm": 0.9088017344474792, + "learning_rate": 9.719391221633374e-05, + "loss": 3.2349, + "step": 106300 + }, + { + "epoch": 7.222788422339992, + "grad_norm": 0.9431678056716919, + "learning_rate": 9.715144720750102e-05, + "loss": 3.3184, + "step": 106305 + }, + { + "epoch": 7.223128142410654, + "grad_norm": 0.9114903211593628, + "learning_rate": 9.71089821986683e-05, + "loss": 3.6356, + "step": 106310 + }, + { + "epoch": 7.223467862481315, + "grad_norm": 1.3047395944595337, + "learning_rate": 9.706651718983558e-05, + "loss": 3.326, + "step": 106315 + }, + { + "epoch": 7.223807582551977, + "grad_norm": 1.2884082794189453, + "learning_rate": 9.702405218100286e-05, + "loss": 3.5703, + "step": 106320 + }, + { + "epoch": 7.224147302622639, + "grad_norm": 0.9177185893058777, + "learning_rate": 9.698158717217014e-05, + "loss": 3.4971, + "step": 106325 + }, + { + "epoch": 7.2244870226933005, + "grad_norm": 2.4811291694641113, + "learning_rate": 9.693912216333742e-05, + "loss": 3.1743, + "step": 106330 + }, + { + "epoch": 7.2248267427639625, + "grad_norm": 0.954371988773346, + "learning_rate": 9.689665715450468e-05, + "loss": 3.3148, + "step": 106335 + }, + { + "epoch": 7.225166462834625, + "grad_norm": 0.8355941772460938, + "learning_rate": 9.685419214567196e-05, + "loss": 3.5112, + "step": 106340 + }, + { + "epoch": 7.225506182905286, + "grad_norm": 0.9622858166694641, + "learning_rate": 9.681172713683924e-05, + "loss": 3.6044, + "step": 106345 + }, + { + "epoch": 7.225845902975948, + "grad_norm": 0.7657926678657532, + "learning_rate": 9.676926212800652e-05, + "loss": 3.4473, + "step": 106350 + }, + { + "epoch": 7.22618562304661, + "grad_norm": 1.1049054861068726, + "learning_rate": 9.67267971191738e-05, + "loss": 3.2212, + "step": 106355 + }, + { + "epoch": 7.226525343117271, + "grad_norm": 1.0420300960540771, + "learning_rate": 9.668433211034108e-05, + "loss": 3.3794, + "step": 106360 + }, + { + "epoch": 7.226865063187933, + "grad_norm": 1.0227950811386108, + "learning_rate": 9.664186710150836e-05, + "loss": 3.3709, + "step": 106365 + }, + { + "epoch": 7.227204783258595, + "grad_norm": 0.8876162171363831, + "learning_rate": 9.659940209267563e-05, + "loss": 3.7611, + "step": 106370 + }, + { + "epoch": 7.2275445033292565, + "grad_norm": 0.8694350719451904, + "learning_rate": 9.655693708384291e-05, + "loss": 3.2489, + "step": 106375 + }, + { + "epoch": 7.2278842233999185, + "grad_norm": 1.1915867328643799, + "learning_rate": 9.65144720750102e-05, + "loss": 3.6785, + "step": 106380 + }, + { + "epoch": 7.228223943470581, + "grad_norm": 0.9538058042526245, + "learning_rate": 9.647200706617747e-05, + "loss": 2.9739, + "step": 106385 + }, + { + "epoch": 7.228563663541242, + "grad_norm": 0.8760938048362732, + "learning_rate": 9.642954205734475e-05, + "loss": 3.3766, + "step": 106390 + }, + { + "epoch": 7.228903383611904, + "grad_norm": 1.1208409070968628, + "learning_rate": 9.638707704851203e-05, + "loss": 3.3962, + "step": 106395 + }, + { + "epoch": 7.229243103682566, + "grad_norm": 0.907090961933136, + "learning_rate": 9.63446120396793e-05, + "loss": 3.3709, + "step": 106400 + }, + { + "epoch": 7.229582823753227, + "grad_norm": 1.042445421218872, + "learning_rate": 9.630214703084658e-05, + "loss": 3.3612, + "step": 106405 + }, + { + "epoch": 7.229922543823889, + "grad_norm": 1.0175532102584839, + "learning_rate": 9.625968202201386e-05, + "loss": 3.3326, + "step": 106410 + }, + { + "epoch": 7.230262263894551, + "grad_norm": 0.8118071556091309, + "learning_rate": 9.621721701318115e-05, + "loss": 3.371, + "step": 106415 + }, + { + "epoch": 7.2306019839652125, + "grad_norm": 1.0454462766647339, + "learning_rate": 9.617475200434842e-05, + "loss": 3.4668, + "step": 106420 + }, + { + "epoch": 7.2309417040358746, + "grad_norm": 0.9100196361541748, + "learning_rate": 9.61322869955157e-05, + "loss": 3.3749, + "step": 106425 + }, + { + "epoch": 7.231281424106537, + "grad_norm": 0.9474098086357117, + "learning_rate": 9.608982198668298e-05, + "loss": 3.4178, + "step": 106430 + }, + { + "epoch": 7.231621144177198, + "grad_norm": 0.845032274723053, + "learning_rate": 9.604735697785024e-05, + "loss": 3.4, + "step": 106435 + }, + { + "epoch": 7.23196086424786, + "grad_norm": 0.914370059967041, + "learning_rate": 9.600489196901752e-05, + "loss": 3.6197, + "step": 106440 + }, + { + "epoch": 7.232300584318522, + "grad_norm": 0.7837552428245544, + "learning_rate": 9.596242696018482e-05, + "loss": 3.3617, + "step": 106445 + }, + { + "epoch": 7.232640304389183, + "grad_norm": 0.8302417993545532, + "learning_rate": 9.59199619513521e-05, + "loss": 3.5964, + "step": 106450 + }, + { + "epoch": 7.232980024459845, + "grad_norm": 0.941538393497467, + "learning_rate": 9.587749694251936e-05, + "loss": 3.167, + "step": 106455 + }, + { + "epoch": 7.233319744530506, + "grad_norm": 1.1152819395065308, + "learning_rate": 9.583503193368665e-05, + "loss": 3.2851, + "step": 106460 + }, + { + "epoch": 7.2336594646011685, + "grad_norm": 0.8581592440605164, + "learning_rate": 9.579256692485393e-05, + "loss": 3.4368, + "step": 106465 + }, + { + "epoch": 7.233999184671831, + "grad_norm": 1.0482348203659058, + "learning_rate": 9.575010191602119e-05, + "loss": 3.2968, + "step": 106470 + }, + { + "epoch": 7.234338904742492, + "grad_norm": 1.039038896560669, + "learning_rate": 9.570763690718847e-05, + "loss": 3.3529, + "step": 106475 + }, + { + "epoch": 7.234678624813154, + "grad_norm": 0.8165888786315918, + "learning_rate": 9.566517189835577e-05, + "loss": 3.2889, + "step": 106480 + }, + { + "epoch": 7.235018344883816, + "grad_norm": 0.8613719344139099, + "learning_rate": 9.562270688952303e-05, + "loss": 3.295, + "step": 106485 + }, + { + "epoch": 7.235358064954477, + "grad_norm": 1.0579370260238647, + "learning_rate": 9.558024188069031e-05, + "loss": 3.09, + "step": 106490 + }, + { + "epoch": 7.235697785025139, + "grad_norm": 0.9262233376502991, + "learning_rate": 9.553777687185759e-05, + "loss": 3.6548, + "step": 106495 + }, + { + "epoch": 7.236037505095801, + "grad_norm": 0.8270819783210754, + "learning_rate": 9.549531186302487e-05, + "loss": 3.3771, + "step": 106500 + }, + { + "epoch": 7.236377225166462, + "grad_norm": 1.293101191520691, + "learning_rate": 9.545284685419214e-05, + "loss": 3.2819, + "step": 106505 + }, + { + "epoch": 7.2367169452371245, + "grad_norm": 1.1277183294296265, + "learning_rate": 9.541038184535943e-05, + "loss": 3.3704, + "step": 106510 + }, + { + "epoch": 7.237056665307787, + "grad_norm": 1.0905203819274902, + "learning_rate": 9.536791683652671e-05, + "loss": 3.3035, + "step": 106515 + }, + { + "epoch": 7.237396385378448, + "grad_norm": 0.940873920917511, + "learning_rate": 9.532545182769398e-05, + "loss": 3.5163, + "step": 106520 + }, + { + "epoch": 7.23773610544911, + "grad_norm": 0.7385991215705872, + "learning_rate": 9.528298681886126e-05, + "loss": 3.0631, + "step": 106525 + }, + { + "epoch": 7.238075825519772, + "grad_norm": 0.9696252346038818, + "learning_rate": 9.524052181002854e-05, + "loss": 3.2609, + "step": 106530 + }, + { + "epoch": 7.238415545590433, + "grad_norm": 0.8513774871826172, + "learning_rate": 9.519805680119582e-05, + "loss": 3.3147, + "step": 106535 + }, + { + "epoch": 7.238755265661095, + "grad_norm": 0.8083130121231079, + "learning_rate": 9.515559179236309e-05, + "loss": 3.5808, + "step": 106540 + }, + { + "epoch": 7.239094985731757, + "grad_norm": 1.0888323783874512, + "learning_rate": 9.511312678353038e-05, + "loss": 3.566, + "step": 106545 + }, + { + "epoch": 7.239434705802418, + "grad_norm": 0.9836030006408691, + "learning_rate": 9.507066177469766e-05, + "loss": 3.1777, + "step": 106550 + }, + { + "epoch": 7.2397744258730805, + "grad_norm": 0.946246862411499, + "learning_rate": 9.502819676586493e-05, + "loss": 3.4177, + "step": 106555 + }, + { + "epoch": 7.240114145943743, + "grad_norm": 1.0546127557754517, + "learning_rate": 9.49857317570322e-05, + "loss": 3.3841, + "step": 106560 + }, + { + "epoch": 7.240453866014404, + "grad_norm": 0.951405942440033, + "learning_rate": 9.494326674819949e-05, + "loss": 3.2587, + "step": 106565 + }, + { + "epoch": 7.240793586085066, + "grad_norm": 0.8579548001289368, + "learning_rate": 9.490080173936675e-05, + "loss": 3.6087, + "step": 106570 + }, + { + "epoch": 7.241133306155728, + "grad_norm": 1.0019659996032715, + "learning_rate": 9.485833673053405e-05, + "loss": 3.5057, + "step": 106575 + }, + { + "epoch": 7.241473026226389, + "grad_norm": 0.8177088499069214, + "learning_rate": 9.481587172170133e-05, + "loss": 3.349, + "step": 106580 + }, + { + "epoch": 7.241812746297051, + "grad_norm": 1.1058872938156128, + "learning_rate": 9.47734067128686e-05, + "loss": 3.4896, + "step": 106585 + }, + { + "epoch": 7.242152466367713, + "grad_norm": 0.7770290970802307, + "learning_rate": 9.473094170403587e-05, + "loss": 3.4696, + "step": 106590 + }, + { + "epoch": 7.2424921864383744, + "grad_norm": 1.000299096107483, + "learning_rate": 9.468847669520315e-05, + "loss": 3.1953, + "step": 106595 + }, + { + "epoch": 7.2428319065090365, + "grad_norm": 0.918255627155304, + "learning_rate": 9.464601168637043e-05, + "loss": 3.573, + "step": 106600 + }, + { + "epoch": 7.243171626579699, + "grad_norm": 2.2672486305236816, + "learning_rate": 9.46035466775377e-05, + "loss": 3.3964, + "step": 106605 + }, + { + "epoch": 7.24351134665036, + "grad_norm": 0.8157080411911011, + "learning_rate": 9.456108166870499e-05, + "loss": 3.3527, + "step": 106610 + }, + { + "epoch": 7.243851066721022, + "grad_norm": 1.64271879196167, + "learning_rate": 9.451861665987227e-05, + "loss": 3.5048, + "step": 106615 + }, + { + "epoch": 7.244190786791684, + "grad_norm": 0.9110996127128601, + "learning_rate": 9.447615165103955e-05, + "loss": 3.1844, + "step": 106620 + }, + { + "epoch": 7.244530506862345, + "grad_norm": 0.7752372622489929, + "learning_rate": 9.443368664220682e-05, + "loss": 3.2265, + "step": 106625 + }, + { + "epoch": 7.244870226933007, + "grad_norm": 1.0056084394454956, + "learning_rate": 9.43912216333741e-05, + "loss": 3.4154, + "step": 106630 + }, + { + "epoch": 7.245209947003669, + "grad_norm": 0.8562089204788208, + "learning_rate": 9.434875662454138e-05, + "loss": 3.495, + "step": 106635 + }, + { + "epoch": 7.2455496670743305, + "grad_norm": 0.7597572207450867, + "learning_rate": 9.430629161570865e-05, + "loss": 3.334, + "step": 106640 + }, + { + "epoch": 7.2458893871449925, + "grad_norm": 0.9676788449287415, + "learning_rate": 9.426382660687594e-05, + "loss": 3.3377, + "step": 106645 + }, + { + "epoch": 7.246229107215655, + "grad_norm": 0.7256278991699219, + "learning_rate": 9.422136159804322e-05, + "loss": 3.5903, + "step": 106650 + }, + { + "epoch": 7.246568827286316, + "grad_norm": 1.0062614679336548, + "learning_rate": 9.417889658921049e-05, + "loss": 3.0629, + "step": 106655 + }, + { + "epoch": 7.246908547356978, + "grad_norm": 0.6900750994682312, + "learning_rate": 9.413643158037777e-05, + "loss": 3.2853, + "step": 106660 + }, + { + "epoch": 7.24724826742764, + "grad_norm": 0.776353120803833, + "learning_rate": 9.409396657154505e-05, + "loss": 3.4878, + "step": 106665 + }, + { + "epoch": 7.247587987498301, + "grad_norm": 0.9844589829444885, + "learning_rate": 9.405150156271233e-05, + "loss": 3.2636, + "step": 106670 + }, + { + "epoch": 7.247927707568963, + "grad_norm": 0.8031949400901794, + "learning_rate": 9.400903655387961e-05, + "loss": 3.391, + "step": 106675 + }, + { + "epoch": 7.248267427639625, + "grad_norm": 0.8952523469924927, + "learning_rate": 9.396657154504689e-05, + "loss": 3.3474, + "step": 106680 + }, + { + "epoch": 7.2486071477102865, + "grad_norm": 1.1714378595352173, + "learning_rate": 9.392410653621417e-05, + "loss": 3.5071, + "step": 106685 + }, + { + "epoch": 7.2489468677809485, + "grad_norm": 1.1700857877731323, + "learning_rate": 9.388164152738143e-05, + "loss": 3.2921, + "step": 106690 + }, + { + "epoch": 7.249286587851611, + "grad_norm": 0.8482294678688049, + "learning_rate": 9.383917651854871e-05, + "loss": 3.3805, + "step": 106695 + }, + { + "epoch": 7.249626307922272, + "grad_norm": 0.8425138592720032, + "learning_rate": 9.3796711509716e-05, + "loss": 3.3339, + "step": 106700 + }, + { + "epoch": 7.249966027992934, + "grad_norm": 0.8165473937988281, + "learning_rate": 9.375424650088327e-05, + "loss": 3.591, + "step": 106705 + }, + { + "epoch": 7.250305748063596, + "grad_norm": 0.7393984794616699, + "learning_rate": 9.371178149205055e-05, + "loss": 3.5354, + "step": 106710 + }, + { + "epoch": 7.250645468134257, + "grad_norm": 0.7846572399139404, + "learning_rate": 9.366931648321783e-05, + "loss": 3.4241, + "step": 106715 + }, + { + "epoch": 7.250985188204919, + "grad_norm": 1.0063045024871826, + "learning_rate": 9.362685147438511e-05, + "loss": 3.4309, + "step": 106720 + }, + { + "epoch": 7.251324908275581, + "grad_norm": 0.9192140698432922, + "learning_rate": 9.358438646555238e-05, + "loss": 3.4692, + "step": 106725 + }, + { + "epoch": 7.2516646283462425, + "grad_norm": 0.9471952319145203, + "learning_rate": 9.354192145671966e-05, + "loss": 3.3377, + "step": 106730 + }, + { + "epoch": 7.2520043484169046, + "grad_norm": 1.114580750465393, + "learning_rate": 9.349945644788694e-05, + "loss": 3.3222, + "step": 106735 + }, + { + "epoch": 7.252344068487567, + "grad_norm": 0.8823837041854858, + "learning_rate": 9.345699143905422e-05, + "loss": 3.3573, + "step": 106740 + }, + { + "epoch": 7.252683788558228, + "grad_norm": 0.9963676929473877, + "learning_rate": 9.34145264302215e-05, + "loss": 3.3804, + "step": 106745 + }, + { + "epoch": 7.25302350862889, + "grad_norm": 1.3270022869110107, + "learning_rate": 9.337206142138878e-05, + "loss": 3.2173, + "step": 106750 + }, + { + "epoch": 7.253363228699552, + "grad_norm": 1.1173151731491089, + "learning_rate": 9.332959641255606e-05, + "loss": 2.9138, + "step": 106755 + }, + { + "epoch": 7.253702948770213, + "grad_norm": 0.8571094274520874, + "learning_rate": 9.328713140372333e-05, + "loss": 3.4511, + "step": 106760 + }, + { + "epoch": 7.254042668840875, + "grad_norm": 1.3186668157577515, + "learning_rate": 9.324466639489061e-05, + "loss": 3.1313, + "step": 106765 + }, + { + "epoch": 7.254382388911537, + "grad_norm": 0.9716456532478333, + "learning_rate": 9.320220138605789e-05, + "loss": 3.2975, + "step": 106770 + }, + { + "epoch": 7.2547221089821985, + "grad_norm": 1.0269174575805664, + "learning_rate": 9.315973637722517e-05, + "loss": 3.4116, + "step": 106775 + }, + { + "epoch": 7.255061829052861, + "grad_norm": 0.9881914258003235, + "learning_rate": 9.311727136839245e-05, + "loss": 3.3755, + "step": 106780 + }, + { + "epoch": 7.255401549123523, + "grad_norm": 0.8333271145820618, + "learning_rate": 9.307480635955973e-05, + "loss": 3.2853, + "step": 106785 + }, + { + "epoch": 7.255741269194184, + "grad_norm": 0.90218186378479, + "learning_rate": 9.303234135072701e-05, + "loss": 3.3888, + "step": 106790 + }, + { + "epoch": 7.256080989264846, + "grad_norm": 0.895671546459198, + "learning_rate": 9.298987634189427e-05, + "loss": 3.3055, + "step": 106795 + }, + { + "epoch": 7.256420709335508, + "grad_norm": 0.7842230200767517, + "learning_rate": 9.294741133306155e-05, + "loss": 3.4232, + "step": 106800 + }, + { + "epoch": 7.256760429406169, + "grad_norm": 0.9530742168426514, + "learning_rate": 9.290494632422885e-05, + "loss": 3.2644, + "step": 106805 + }, + { + "epoch": 7.257100149476831, + "grad_norm": 1.1795860528945923, + "learning_rate": 9.286248131539611e-05, + "loss": 3.21, + "step": 106810 + }, + { + "epoch": 7.257439869547493, + "grad_norm": 1.3160592317581177, + "learning_rate": 9.28200163065634e-05, + "loss": 3.6, + "step": 106815 + }, + { + "epoch": 7.2577795896181545, + "grad_norm": 0.9464600682258606, + "learning_rate": 9.277755129773068e-05, + "loss": 3.5095, + "step": 106820 + }, + { + "epoch": 7.258119309688817, + "grad_norm": 1.620780348777771, + "learning_rate": 9.273508628889794e-05, + "loss": 3.4466, + "step": 106825 + }, + { + "epoch": 7.258459029759479, + "grad_norm": 0.7656317949295044, + "learning_rate": 9.269262128006522e-05, + "loss": 3.4939, + "step": 106830 + }, + { + "epoch": 7.25879874983014, + "grad_norm": 0.9947114586830139, + "learning_rate": 9.26501562712325e-05, + "loss": 3.6153, + "step": 106835 + }, + { + "epoch": 7.259138469900802, + "grad_norm": 1.2970656156539917, + "learning_rate": 9.26076912623998e-05, + "loss": 3.3452, + "step": 106840 + }, + { + "epoch": 7.259478189971463, + "grad_norm": 1.0291388034820557, + "learning_rate": 9.256522625356706e-05, + "loss": 3.5979, + "step": 106845 + }, + { + "epoch": 7.259817910042125, + "grad_norm": 0.7050414085388184, + "learning_rate": 9.252276124473434e-05, + "loss": 3.2076, + "step": 106850 + }, + { + "epoch": 7.260157630112787, + "grad_norm": 0.9933184385299683, + "learning_rate": 9.248029623590162e-05, + "loss": 3.3004, + "step": 106855 + }, + { + "epoch": 7.260497350183448, + "grad_norm": 0.9276649355888367, + "learning_rate": 9.243783122706889e-05, + "loss": 3.3571, + "step": 106860 + }, + { + "epoch": 7.2608370702541105, + "grad_norm": 1.1973949670791626, + "learning_rate": 9.239536621823617e-05, + "loss": 3.1992, + "step": 106865 + }, + { + "epoch": 7.261176790324773, + "grad_norm": 0.8103123903274536, + "learning_rate": 9.235290120940346e-05, + "loss": 3.3415, + "step": 106870 + }, + { + "epoch": 7.261516510395434, + "grad_norm": 0.666739821434021, + "learning_rate": 9.231043620057074e-05, + "loss": 3.2505, + "step": 106875 + }, + { + "epoch": 7.261856230466096, + "grad_norm": 0.96098792552948, + "learning_rate": 9.226797119173801e-05, + "loss": 3.4284, + "step": 106880 + }, + { + "epoch": 7.262195950536758, + "grad_norm": 0.8341097831726074, + "learning_rate": 9.222550618290529e-05, + "loss": 3.8441, + "step": 106885 + }, + { + "epoch": 7.262535670607419, + "grad_norm": 1.1385648250579834, + "learning_rate": 9.218304117407257e-05, + "loss": 3.1768, + "step": 106890 + }, + { + "epoch": 7.262875390678081, + "grad_norm": 0.823043704032898, + "learning_rate": 9.214057616523984e-05, + "loss": 3.3709, + "step": 106895 + }, + { + "epoch": 7.263215110748743, + "grad_norm": 1.0770483016967773, + "learning_rate": 9.209811115640712e-05, + "loss": 3.3513, + "step": 106900 + }, + { + "epoch": 7.2635548308194045, + "grad_norm": 1.1914007663726807, + "learning_rate": 9.205564614757441e-05, + "loss": 3.3502, + "step": 106905 + }, + { + "epoch": 7.2638945508900665, + "grad_norm": 0.9453372955322266, + "learning_rate": 9.201318113874168e-05, + "loss": 3.3294, + "step": 106910 + }, + { + "epoch": 7.264234270960729, + "grad_norm": 0.8473502397537231, + "learning_rate": 9.197071612990896e-05, + "loss": 3.4574, + "step": 106915 + }, + { + "epoch": 7.26457399103139, + "grad_norm": 0.966793417930603, + "learning_rate": 9.192825112107624e-05, + "loss": 3.5083, + "step": 106920 + }, + { + "epoch": 7.264913711102052, + "grad_norm": 1.0320755243301392, + "learning_rate": 9.188578611224352e-05, + "loss": 3.4288, + "step": 106925 + }, + { + "epoch": 7.265253431172714, + "grad_norm": 0.8788886666297913, + "learning_rate": 9.184332110341078e-05, + "loss": 3.2552, + "step": 106930 + }, + { + "epoch": 7.265593151243375, + "grad_norm": 1.2685620784759521, + "learning_rate": 9.180085609457808e-05, + "loss": 3.4224, + "step": 106935 + }, + { + "epoch": 7.265932871314037, + "grad_norm": 0.9594550132751465, + "learning_rate": 9.175839108574536e-05, + "loss": 3.3946, + "step": 106940 + }, + { + "epoch": 7.266272591384699, + "grad_norm": 0.8692086338996887, + "learning_rate": 9.171592607691262e-05, + "loss": 3.4721, + "step": 106945 + }, + { + "epoch": 7.2666123114553605, + "grad_norm": 1.11311674118042, + "learning_rate": 9.16734610680799e-05, + "loss": 3.5526, + "step": 106950 + }, + { + "epoch": 7.2669520315260225, + "grad_norm": 0.8636426329612732, + "learning_rate": 9.163099605924718e-05, + "loss": 3.4327, + "step": 106955 + }, + { + "epoch": 7.267291751596685, + "grad_norm": 0.8436318635940552, + "learning_rate": 9.158853105041446e-05, + "loss": 3.4264, + "step": 106960 + }, + { + "epoch": 7.267631471667346, + "grad_norm": 1.2998015880584717, + "learning_rate": 9.154606604158173e-05, + "loss": 3.4445, + "step": 106965 + }, + { + "epoch": 7.267971191738008, + "grad_norm": 0.9564030766487122, + "learning_rate": 9.150360103274902e-05, + "loss": 3.2172, + "step": 106970 + }, + { + "epoch": 7.26831091180867, + "grad_norm": 0.9448767900466919, + "learning_rate": 9.14611360239163e-05, + "loss": 3.4601, + "step": 106975 + }, + { + "epoch": 7.268650631879331, + "grad_norm": 0.9790486693382263, + "learning_rate": 9.141867101508357e-05, + "loss": 3.5017, + "step": 106980 + }, + { + "epoch": 7.268990351949993, + "grad_norm": 1.2402832508087158, + "learning_rate": 9.137620600625085e-05, + "loss": 3.3714, + "step": 106985 + }, + { + "epoch": 7.269330072020655, + "grad_norm": 0.8848008513450623, + "learning_rate": 9.133374099741813e-05, + "loss": 3.557, + "step": 106990 + }, + { + "epoch": 7.2696697920913165, + "grad_norm": 1.162688136100769, + "learning_rate": 9.12912759885854e-05, + "loss": 3.4596, + "step": 106995 + }, + { + "epoch": 7.2700095121619785, + "grad_norm": 0.8788232803344727, + "learning_rate": 9.124881097975269e-05, + "loss": 3.4341, + "step": 107000 + }, + { + "epoch": 7.270349232232641, + "grad_norm": 0.9277045726776123, + "learning_rate": 9.120634597091997e-05, + "loss": 3.4201, + "step": 107005 + }, + { + "epoch": 7.270688952303302, + "grad_norm": 0.9737425446510315, + "learning_rate": 9.116388096208725e-05, + "loss": 3.3658, + "step": 107010 + }, + { + "epoch": 7.271028672373964, + "grad_norm": 0.9126131534576416, + "learning_rate": 9.112141595325452e-05, + "loss": 3.4172, + "step": 107015 + }, + { + "epoch": 7.271368392444626, + "grad_norm": 0.923049807548523, + "learning_rate": 9.10789509444218e-05, + "loss": 3.3864, + "step": 107020 + }, + { + "epoch": 7.271708112515287, + "grad_norm": 1.052669882774353, + "learning_rate": 9.103648593558908e-05, + "loss": 3.5379, + "step": 107025 + }, + { + "epoch": 7.272047832585949, + "grad_norm": 0.9818555116653442, + "learning_rate": 9.099402092675634e-05, + "loss": 3.2325, + "step": 107030 + }, + { + "epoch": 7.272387552656611, + "grad_norm": 1.2338306903839111, + "learning_rate": 9.095155591792364e-05, + "loss": 3.2994, + "step": 107035 + }, + { + "epoch": 7.2727272727272725, + "grad_norm": 1.0855276584625244, + "learning_rate": 9.090909090909092e-05, + "loss": 3.3711, + "step": 107040 + }, + { + "epoch": 7.273066992797935, + "grad_norm": 0.9354802966117859, + "learning_rate": 9.08666259002582e-05, + "loss": 3.5215, + "step": 107045 + }, + { + "epoch": 7.273406712868597, + "grad_norm": 0.8256500959396362, + "learning_rate": 9.082416089142546e-05, + "loss": 3.1628, + "step": 107050 + }, + { + "epoch": 7.273746432939258, + "grad_norm": 0.8966386318206787, + "learning_rate": 9.078169588259274e-05, + "loss": 3.3606, + "step": 107055 + }, + { + "epoch": 7.27408615300992, + "grad_norm": 0.980743944644928, + "learning_rate": 9.073923087376002e-05, + "loss": 3.5194, + "step": 107060 + }, + { + "epoch": 7.274425873080582, + "grad_norm": 1.0098215341567993, + "learning_rate": 9.069676586492729e-05, + "loss": 3.3056, + "step": 107065 + }, + { + "epoch": 7.274765593151243, + "grad_norm": 0.8012374043464661, + "learning_rate": 9.065430085609458e-05, + "loss": 3.3531, + "step": 107070 + }, + { + "epoch": 7.275105313221905, + "grad_norm": 0.8106400966644287, + "learning_rate": 9.061183584726186e-05, + "loss": 3.3898, + "step": 107075 + }, + { + "epoch": 7.275445033292567, + "grad_norm": 0.9125608801841736, + "learning_rate": 9.056937083842913e-05, + "loss": 3.5024, + "step": 107080 + }, + { + "epoch": 7.2757847533632285, + "grad_norm": 0.7134227156639099, + "learning_rate": 9.052690582959641e-05, + "loss": 3.3964, + "step": 107085 + }, + { + "epoch": 7.276124473433891, + "grad_norm": 0.9514400362968445, + "learning_rate": 9.048444082076369e-05, + "loss": 3.3675, + "step": 107090 + }, + { + "epoch": 7.276464193504553, + "grad_norm": 1.1194278001785278, + "learning_rate": 9.044197581193097e-05, + "loss": 3.3864, + "step": 107095 + }, + { + "epoch": 7.276803913575214, + "grad_norm": 1.0230005979537964, + "learning_rate": 9.039951080309825e-05, + "loss": 3.3511, + "step": 107100 + }, + { + "epoch": 7.277143633645876, + "grad_norm": 1.2828495502471924, + "learning_rate": 9.035704579426553e-05, + "loss": 3.6634, + "step": 107105 + }, + { + "epoch": 7.277483353716538, + "grad_norm": 0.6929152011871338, + "learning_rate": 9.031458078543281e-05, + "loss": 3.4554, + "step": 107110 + }, + { + "epoch": 7.277823073787199, + "grad_norm": 0.8674505352973938, + "learning_rate": 9.027211577660008e-05, + "loss": 3.6957, + "step": 107115 + }, + { + "epoch": 7.278162793857861, + "grad_norm": 1.3988306522369385, + "learning_rate": 9.022965076776736e-05, + "loss": 3.3553, + "step": 107120 + }, + { + "epoch": 7.278502513928522, + "grad_norm": 0.8731162548065186, + "learning_rate": 9.018718575893464e-05, + "loss": 3.3902, + "step": 107125 + }, + { + "epoch": 7.2788422339991845, + "grad_norm": 0.910225510597229, + "learning_rate": 9.014472075010192e-05, + "loss": 3.4752, + "step": 107130 + }, + { + "epoch": 7.279181954069847, + "grad_norm": 0.9261617660522461, + "learning_rate": 9.01022557412692e-05, + "loss": 3.2025, + "step": 107135 + }, + { + "epoch": 7.279521674140508, + "grad_norm": 1.0815340280532837, + "learning_rate": 9.005979073243648e-05, + "loss": 3.1077, + "step": 107140 + }, + { + "epoch": 7.27986139421117, + "grad_norm": 0.9436856508255005, + "learning_rate": 9.001732572360376e-05, + "loss": 3.3943, + "step": 107145 + }, + { + "epoch": 7.280201114281832, + "grad_norm": 0.8332599401473999, + "learning_rate": 8.997486071477102e-05, + "loss": 3.3565, + "step": 107150 + }, + { + "epoch": 7.280540834352493, + "grad_norm": 0.7948506474494934, + "learning_rate": 8.99323957059383e-05, + "loss": 3.2373, + "step": 107155 + }, + { + "epoch": 7.280880554423155, + "grad_norm": 0.8848767280578613, + "learning_rate": 8.988993069710558e-05, + "loss": 3.3135, + "step": 107160 + }, + { + "epoch": 7.281220274493817, + "grad_norm": 0.9989780187606812, + "learning_rate": 8.984746568827286e-05, + "loss": 3.3632, + "step": 107165 + }, + { + "epoch": 7.2815599945644784, + "grad_norm": 1.09257972240448, + "learning_rate": 8.980500067944014e-05, + "loss": 3.2188, + "step": 107170 + }, + { + "epoch": 7.2818997146351405, + "grad_norm": 0.8660016059875488, + "learning_rate": 8.976253567060743e-05, + "loss": 3.4904, + "step": 107175 + }, + { + "epoch": 7.282239434705803, + "grad_norm": 0.970380961894989, + "learning_rate": 8.97200706617747e-05, + "loss": 3.4094, + "step": 107180 + }, + { + "epoch": 7.282579154776464, + "grad_norm": 0.8381536602973938, + "learning_rate": 8.967760565294197e-05, + "loss": 3.4481, + "step": 107185 + }, + { + "epoch": 7.282918874847126, + "grad_norm": 1.0733616352081299, + "learning_rate": 8.963514064410925e-05, + "loss": 3.346, + "step": 107190 + }, + { + "epoch": 7.283258594917788, + "grad_norm": 1.0107165575027466, + "learning_rate": 8.959267563527653e-05, + "loss": 3.2561, + "step": 107195 + }, + { + "epoch": 7.283598314988449, + "grad_norm": 0.971398651599884, + "learning_rate": 8.955021062644381e-05, + "loss": 3.4433, + "step": 107200 + }, + { + "epoch": 7.283938035059111, + "grad_norm": 0.9843540787696838, + "learning_rate": 8.950774561761109e-05, + "loss": 3.1803, + "step": 107205 + }, + { + "epoch": 7.284277755129773, + "grad_norm": 0.882333517074585, + "learning_rate": 8.946528060877837e-05, + "loss": 3.346, + "step": 107210 + }, + { + "epoch": 7.2846174752004345, + "grad_norm": 0.9383924603462219, + "learning_rate": 8.942281559994565e-05, + "loss": 3.143, + "step": 107215 + }, + { + "epoch": 7.2849571952710965, + "grad_norm": 1.336343765258789, + "learning_rate": 8.938035059111292e-05, + "loss": 3.2426, + "step": 107220 + }, + { + "epoch": 7.285296915341759, + "grad_norm": 0.9225680828094482, + "learning_rate": 8.93378855822802e-05, + "loss": 3.3826, + "step": 107225 + }, + { + "epoch": 7.28563663541242, + "grad_norm": 1.4366039037704468, + "learning_rate": 8.929542057344749e-05, + "loss": 3.7627, + "step": 107230 + }, + { + "epoch": 7.285976355483082, + "grad_norm": 1.2453701496124268, + "learning_rate": 8.925295556461476e-05, + "loss": 3.2445, + "step": 107235 + }, + { + "epoch": 7.286316075553744, + "grad_norm": 0.8516484498977661, + "learning_rate": 8.921049055578204e-05, + "loss": 3.2415, + "step": 107240 + }, + { + "epoch": 7.286655795624405, + "grad_norm": 1.17799711227417, + "learning_rate": 8.916802554694932e-05, + "loss": 3.2235, + "step": 107245 + }, + { + "epoch": 7.286995515695067, + "grad_norm": 0.9798224568367004, + "learning_rate": 8.912556053811659e-05, + "loss": 3.168, + "step": 107250 + }, + { + "epoch": 7.287335235765729, + "grad_norm": 1.0036633014678955, + "learning_rate": 8.908309552928387e-05, + "loss": 3.5133, + "step": 107255 + }, + { + "epoch": 7.2876749558363905, + "grad_norm": 0.799165666103363, + "learning_rate": 8.904063052045115e-05, + "loss": 3.5302, + "step": 107260 + }, + { + "epoch": 7.2880146759070525, + "grad_norm": 0.9476134777069092, + "learning_rate": 8.899816551161844e-05, + "loss": 3.2673, + "step": 107265 + }, + { + "epoch": 7.288354395977715, + "grad_norm": 0.802742600440979, + "learning_rate": 8.89557005027857e-05, + "loss": 3.3753, + "step": 107270 + }, + { + "epoch": 7.288694116048376, + "grad_norm": 0.9727922081947327, + "learning_rate": 8.891323549395299e-05, + "loss": 3.3171, + "step": 107275 + }, + { + "epoch": 7.289033836119038, + "grad_norm": 0.9356397986412048, + "learning_rate": 8.887077048512027e-05, + "loss": 3.0731, + "step": 107280 + }, + { + "epoch": 7.2893735561897, + "grad_norm": 0.9146900177001953, + "learning_rate": 8.882830547628753e-05, + "loss": 3.5595, + "step": 107285 + }, + { + "epoch": 7.289713276260361, + "grad_norm": 0.9211974740028381, + "learning_rate": 8.878584046745481e-05, + "loss": 3.4319, + "step": 107290 + }, + { + "epoch": 7.290052996331023, + "grad_norm": 1.039755940437317, + "learning_rate": 8.87433754586221e-05, + "loss": 3.4385, + "step": 107295 + }, + { + "epoch": 7.290392716401685, + "grad_norm": 0.857925534248352, + "learning_rate": 8.870091044978939e-05, + "loss": 3.727, + "step": 107300 + }, + { + "epoch": 7.2907324364723465, + "grad_norm": 0.956771194934845, + "learning_rate": 8.865844544095665e-05, + "loss": 3.3175, + "step": 107305 + }, + { + "epoch": 7.2910721565430086, + "grad_norm": 0.8232337832450867, + "learning_rate": 8.861598043212393e-05, + "loss": 3.369, + "step": 107310 + }, + { + "epoch": 7.291411876613671, + "grad_norm": 0.8263154029846191, + "learning_rate": 8.857351542329121e-05, + "loss": 3.4991, + "step": 107315 + }, + { + "epoch": 7.291751596684332, + "grad_norm": 0.9668517112731934, + "learning_rate": 8.853105041445848e-05, + "loss": 3.4338, + "step": 107320 + }, + { + "epoch": 7.292091316754994, + "grad_norm": 1.1281527280807495, + "learning_rate": 8.848858540562576e-05, + "loss": 3.5241, + "step": 107325 + }, + { + "epoch": 7.292431036825656, + "grad_norm": 0.8721659779548645, + "learning_rate": 8.844612039679305e-05, + "loss": 3.5407, + "step": 107330 + }, + { + "epoch": 7.292770756896317, + "grad_norm": 0.9380648732185364, + "learning_rate": 8.840365538796032e-05, + "loss": 3.2737, + "step": 107335 + }, + { + "epoch": 7.293110476966979, + "grad_norm": 0.7942481637001038, + "learning_rate": 8.83611903791276e-05, + "loss": 3.5498, + "step": 107340 + }, + { + "epoch": 7.293450197037641, + "grad_norm": 0.7314162850379944, + "learning_rate": 8.831872537029488e-05, + "loss": 3.7606, + "step": 107345 + }, + { + "epoch": 7.2937899171083025, + "grad_norm": 0.855686604976654, + "learning_rate": 8.827626036146216e-05, + "loss": 3.5797, + "step": 107350 + }, + { + "epoch": 7.294129637178965, + "grad_norm": 1.940801978111267, + "learning_rate": 8.823379535262943e-05, + "loss": 3.2045, + "step": 107355 + }, + { + "epoch": 7.294469357249627, + "grad_norm": 0.7938761115074158, + "learning_rate": 8.819133034379672e-05, + "loss": 3.3834, + "step": 107360 + }, + { + "epoch": 7.294809077320288, + "grad_norm": 0.8426679372787476, + "learning_rate": 8.8148865334964e-05, + "loss": 3.5452, + "step": 107365 + }, + { + "epoch": 7.29514879739095, + "grad_norm": 0.9958928823471069, + "learning_rate": 8.810640032613127e-05, + "loss": 3.4249, + "step": 107370 + }, + { + "epoch": 7.295488517461612, + "grad_norm": 0.862682580947876, + "learning_rate": 8.806393531729855e-05, + "loss": 3.3247, + "step": 107375 + }, + { + "epoch": 7.295828237532273, + "grad_norm": 1.0216355323791504, + "learning_rate": 8.802147030846583e-05, + "loss": 3.3511, + "step": 107380 + }, + { + "epoch": 7.296167957602935, + "grad_norm": 0.980289101600647, + "learning_rate": 8.797900529963311e-05, + "loss": 3.4056, + "step": 107385 + }, + { + "epoch": 7.296507677673597, + "grad_norm": 0.7868187427520752, + "learning_rate": 8.793654029080037e-05, + "loss": 3.5976, + "step": 107390 + }, + { + "epoch": 7.2968473977442585, + "grad_norm": 1.0141996145248413, + "learning_rate": 8.789407528196767e-05, + "loss": 3.412, + "step": 107395 + }, + { + "epoch": 7.297187117814921, + "grad_norm": 0.9061580896377563, + "learning_rate": 8.785161027313495e-05, + "loss": 3.5679, + "step": 107400 + }, + { + "epoch": 7.297526837885583, + "grad_norm": 0.8802809119224548, + "learning_rate": 8.780914526430221e-05, + "loss": 3.39, + "step": 107405 + }, + { + "epoch": 7.297866557956244, + "grad_norm": 1.0278078317642212, + "learning_rate": 8.77666802554695e-05, + "loss": 3.5225, + "step": 107410 + }, + { + "epoch": 7.298206278026906, + "grad_norm": 0.9690732955932617, + "learning_rate": 8.772421524663677e-05, + "loss": 3.443, + "step": 107415 + }, + { + "epoch": 7.298545998097568, + "grad_norm": 1.0511988401412964, + "learning_rate": 8.768175023780404e-05, + "loss": 3.4389, + "step": 107420 + }, + { + "epoch": 7.298885718168229, + "grad_norm": 0.8512231111526489, + "learning_rate": 8.763928522897132e-05, + "loss": 3.4266, + "step": 107425 + }, + { + "epoch": 7.299225438238891, + "grad_norm": 0.9343308806419373, + "learning_rate": 8.759682022013861e-05, + "loss": 3.4511, + "step": 107430 + }, + { + "epoch": 7.299565158309553, + "grad_norm": 1.022711157798767, + "learning_rate": 8.75543552113059e-05, + "loss": 3.2, + "step": 107435 + }, + { + "epoch": 7.2999048783802145, + "grad_norm": 0.7922827005386353, + "learning_rate": 8.751189020247316e-05, + "loss": 3.5134, + "step": 107440 + }, + { + "epoch": 7.300244598450877, + "grad_norm": 0.8957505822181702, + "learning_rate": 8.746942519364044e-05, + "loss": 3.1335, + "step": 107445 + }, + { + "epoch": 7.300584318521539, + "grad_norm": 0.9393259286880493, + "learning_rate": 8.742696018480772e-05, + "loss": 3.4615, + "step": 107450 + }, + { + "epoch": 7.3009240385922, + "grad_norm": 0.7673927545547485, + "learning_rate": 8.738449517597499e-05, + "loss": 3.2709, + "step": 107455 + }, + { + "epoch": 7.301263758662862, + "grad_norm": 0.900992214679718, + "learning_rate": 8.734203016714228e-05, + "loss": 3.5114, + "step": 107460 + }, + { + "epoch": 7.301603478733524, + "grad_norm": 0.8152751922607422, + "learning_rate": 8.729956515830956e-05, + "loss": 3.3306, + "step": 107465 + }, + { + "epoch": 7.301943198804185, + "grad_norm": 1.2068121433258057, + "learning_rate": 8.725710014947684e-05, + "loss": 3.5113, + "step": 107470 + }, + { + "epoch": 7.302282918874847, + "grad_norm": 0.9914538860321045, + "learning_rate": 8.721463514064411e-05, + "loss": 3.2835, + "step": 107475 + }, + { + "epoch": 7.302622638945509, + "grad_norm": 0.7942290902137756, + "learning_rate": 8.717217013181139e-05, + "loss": 3.4873, + "step": 107480 + }, + { + "epoch": 7.3029623590161705, + "grad_norm": 0.9575428366661072, + "learning_rate": 8.712970512297867e-05, + "loss": 3.3771, + "step": 107485 + }, + { + "epoch": 7.303302079086833, + "grad_norm": 0.8541513085365295, + "learning_rate": 8.708724011414593e-05, + "loss": 3.1897, + "step": 107490 + }, + { + "epoch": 7.303641799157495, + "grad_norm": 1.1054353713989258, + "learning_rate": 8.704477510531323e-05, + "loss": 3.1382, + "step": 107495 + }, + { + "epoch": 7.303981519228156, + "grad_norm": 1.1326476335525513, + "learning_rate": 8.700231009648051e-05, + "loss": 3.2063, + "step": 107500 + }, + { + "epoch": 7.304321239298818, + "grad_norm": 0.9287380576133728, + "learning_rate": 8.695984508764777e-05, + "loss": 3.587, + "step": 107505 + }, + { + "epoch": 7.30466095936948, + "grad_norm": 1.009476661682129, + "learning_rate": 8.691738007881505e-05, + "loss": 3.0109, + "step": 107510 + }, + { + "epoch": 7.305000679440141, + "grad_norm": 1.1685692071914673, + "learning_rate": 8.687491506998233e-05, + "loss": 3.3106, + "step": 107515 + }, + { + "epoch": 7.305340399510803, + "grad_norm": 0.9144592881202698, + "learning_rate": 8.683245006114961e-05, + "loss": 3.5155, + "step": 107520 + }, + { + "epoch": 7.3056801195814645, + "grad_norm": 0.9857518076896667, + "learning_rate": 8.67899850523169e-05, + "loss": 3.2508, + "step": 107525 + }, + { + "epoch": 7.3060198396521265, + "grad_norm": 0.9267323017120361, + "learning_rate": 8.674752004348417e-05, + "loss": 3.3765, + "step": 107530 + }, + { + "epoch": 7.306359559722789, + "grad_norm": 0.8758743405342102, + "learning_rate": 8.670505503465146e-05, + "loss": 3.6493, + "step": 107535 + }, + { + "epoch": 7.30669927979345, + "grad_norm": 1.1431121826171875, + "learning_rate": 8.666259002581872e-05, + "loss": 3.2143, + "step": 107540 + }, + { + "epoch": 7.307038999864112, + "grad_norm": 0.8707373738288879, + "learning_rate": 8.6620125016986e-05, + "loss": 3.1349, + "step": 107545 + }, + { + "epoch": 7.307378719934774, + "grad_norm": 1.0525777339935303, + "learning_rate": 8.657766000815328e-05, + "loss": 3.5077, + "step": 107550 + }, + { + "epoch": 7.307718440005435, + "grad_norm": 1.0003916025161743, + "learning_rate": 8.653519499932056e-05, + "loss": 3.7203, + "step": 107555 + }, + { + "epoch": 7.308058160076097, + "grad_norm": 0.9376387000083923, + "learning_rate": 8.649272999048784e-05, + "loss": 3.5558, + "step": 107560 + }, + { + "epoch": 7.308397880146759, + "grad_norm": 0.7915636301040649, + "learning_rate": 8.645026498165512e-05, + "loss": 3.3441, + "step": 107565 + }, + { + "epoch": 7.3087376002174205, + "grad_norm": 0.8931470513343811, + "learning_rate": 8.64077999728224e-05, + "loss": 3.7167, + "step": 107570 + }, + { + "epoch": 7.3090773202880825, + "grad_norm": 0.9761461615562439, + "learning_rate": 8.636533496398967e-05, + "loss": 3.2386, + "step": 107575 + }, + { + "epoch": 7.309417040358745, + "grad_norm": 0.9315457344055176, + "learning_rate": 8.632286995515695e-05, + "loss": 3.4492, + "step": 107580 + }, + { + "epoch": 7.309756760429406, + "grad_norm": 0.7896963357925415, + "learning_rate": 8.628040494632423e-05, + "loss": 3.5267, + "step": 107585 + }, + { + "epoch": 7.310096480500068, + "grad_norm": 0.8977712988853455, + "learning_rate": 8.623793993749151e-05, + "loss": 3.2949, + "step": 107590 + }, + { + "epoch": 7.31043620057073, + "grad_norm": 1.1018900871276855, + "learning_rate": 8.619547492865879e-05, + "loss": 3.4232, + "step": 107595 + }, + { + "epoch": 7.310775920641391, + "grad_norm": 1.0431997776031494, + "learning_rate": 8.615300991982607e-05, + "loss": 3.5361, + "step": 107600 + }, + { + "epoch": 7.311115640712053, + "grad_norm": 0.9323875308036804, + "learning_rate": 8.611054491099335e-05, + "loss": 3.3848, + "step": 107605 + }, + { + "epoch": 7.311455360782715, + "grad_norm": 0.9498881101608276, + "learning_rate": 8.606807990216062e-05, + "loss": 3.542, + "step": 107610 + }, + { + "epoch": 7.3117950808533765, + "grad_norm": 0.958308219909668, + "learning_rate": 8.60256148933279e-05, + "loss": 3.4491, + "step": 107615 + }, + { + "epoch": 7.312134800924039, + "grad_norm": 0.879248857498169, + "learning_rate": 8.598314988449518e-05, + "loss": 3.4146, + "step": 107620 + }, + { + "epoch": 7.312474520994701, + "grad_norm": 0.86686110496521, + "learning_rate": 8.594068487566246e-05, + "loss": 3.3151, + "step": 107625 + }, + { + "epoch": 7.312814241065362, + "grad_norm": 0.9246792793273926, + "learning_rate": 8.589821986682974e-05, + "loss": 3.5622, + "step": 107630 + }, + { + "epoch": 7.313153961136024, + "grad_norm": 0.7987502813339233, + "learning_rate": 8.585575485799702e-05, + "loss": 3.3266, + "step": 107635 + }, + { + "epoch": 7.313493681206686, + "grad_norm": 1.0450676679611206, + "learning_rate": 8.58132898491643e-05, + "loss": 3.2256, + "step": 107640 + }, + { + "epoch": 7.313833401277347, + "grad_norm": 0.7311716079711914, + "learning_rate": 8.577082484033156e-05, + "loss": 3.4172, + "step": 107645 + }, + { + "epoch": 7.314173121348009, + "grad_norm": 0.895795464515686, + "learning_rate": 8.572835983149884e-05, + "loss": 3.3012, + "step": 107650 + }, + { + "epoch": 7.314512841418671, + "grad_norm": 1.3504657745361328, + "learning_rate": 8.568589482266614e-05, + "loss": 3.2156, + "step": 107655 + }, + { + "epoch": 7.3148525614893325, + "grad_norm": 0.9048012495040894, + "learning_rate": 8.56434298138334e-05, + "loss": 3.1703, + "step": 107660 + }, + { + "epoch": 7.315192281559995, + "grad_norm": 1.081290602684021, + "learning_rate": 8.560096480500068e-05, + "loss": 3.4153, + "step": 107665 + }, + { + "epoch": 7.315532001630657, + "grad_norm": 0.7919420599937439, + "learning_rate": 8.555849979616796e-05, + "loss": 3.6144, + "step": 107670 + }, + { + "epoch": 7.315871721701318, + "grad_norm": 0.9178652167320251, + "learning_rate": 8.551603478733523e-05, + "loss": 3.6416, + "step": 107675 + }, + { + "epoch": 7.31621144177198, + "grad_norm": 0.847313642501831, + "learning_rate": 8.547356977850251e-05, + "loss": 3.3977, + "step": 107680 + }, + { + "epoch": 7.316551161842642, + "grad_norm": 0.731838047504425, + "learning_rate": 8.543110476966979e-05, + "loss": 3.2111, + "step": 107685 + }, + { + "epoch": 7.316890881913303, + "grad_norm": 0.9036993980407715, + "learning_rate": 8.538863976083708e-05, + "loss": 3.5582, + "step": 107690 + }, + { + "epoch": 7.317230601983965, + "grad_norm": 0.8204734325408936, + "learning_rate": 8.534617475200435e-05, + "loss": 3.2715, + "step": 107695 + }, + { + "epoch": 7.317570322054627, + "grad_norm": 0.8238948583602905, + "learning_rate": 8.530370974317163e-05, + "loss": 3.4013, + "step": 107700 + }, + { + "epoch": 7.3179100421252885, + "grad_norm": 1.4026081562042236, + "learning_rate": 8.526124473433891e-05, + "loss": 3.6283, + "step": 107705 + }, + { + "epoch": 7.318249762195951, + "grad_norm": 1.0963656902313232, + "learning_rate": 8.521877972550618e-05, + "loss": 3.3226, + "step": 107710 + }, + { + "epoch": 7.318589482266613, + "grad_norm": 0.988172709941864, + "learning_rate": 8.517631471667346e-05, + "loss": 3.5974, + "step": 107715 + }, + { + "epoch": 7.318929202337274, + "grad_norm": 0.893653154373169, + "learning_rate": 8.513384970784075e-05, + "loss": 3.7012, + "step": 107720 + }, + { + "epoch": 7.319268922407936, + "grad_norm": 1.0049911737442017, + "learning_rate": 8.509138469900803e-05, + "loss": 3.3531, + "step": 107725 + }, + { + "epoch": 7.319608642478598, + "grad_norm": 1.069619059562683, + "learning_rate": 8.50489196901753e-05, + "loss": 3.4676, + "step": 107730 + }, + { + "epoch": 7.319948362549259, + "grad_norm": 1.2385289669036865, + "learning_rate": 8.500645468134258e-05, + "loss": 3.3392, + "step": 107735 + }, + { + "epoch": 7.320288082619921, + "grad_norm": 1.0862877368927002, + "learning_rate": 8.496398967250986e-05, + "loss": 3.486, + "step": 107740 + }, + { + "epoch": 7.320627802690583, + "grad_norm": 0.842036783695221, + "learning_rate": 8.492152466367712e-05, + "loss": 3.3664, + "step": 107745 + }, + { + "epoch": 7.3209675227612445, + "grad_norm": 0.9630045294761658, + "learning_rate": 8.48790596548444e-05, + "loss": 3.4949, + "step": 107750 + }, + { + "epoch": 7.321307242831907, + "grad_norm": 1.0191376209259033, + "learning_rate": 8.48365946460117e-05, + "loss": 3.0274, + "step": 107755 + }, + { + "epoch": 7.321646962902569, + "grad_norm": 0.7934039235115051, + "learning_rate": 8.479412963717896e-05, + "loss": 3.387, + "step": 107760 + }, + { + "epoch": 7.32198668297323, + "grad_norm": 0.8726193904876709, + "learning_rate": 8.475166462834624e-05, + "loss": 3.2913, + "step": 107765 + }, + { + "epoch": 7.322326403043892, + "grad_norm": 1.0664807558059692, + "learning_rate": 8.470919961951352e-05, + "loss": 3.2865, + "step": 107770 + }, + { + "epoch": 7.322666123114554, + "grad_norm": 0.9074744582176208, + "learning_rate": 8.46667346106808e-05, + "loss": 3.2425, + "step": 107775 + }, + { + "epoch": 7.323005843185215, + "grad_norm": 0.872357964515686, + "learning_rate": 8.462426960184807e-05, + "loss": 3.5814, + "step": 107780 + }, + { + "epoch": 7.323345563255877, + "grad_norm": 0.9839628338813782, + "learning_rate": 8.458180459301536e-05, + "loss": 3.4004, + "step": 107785 + }, + { + "epoch": 7.323685283326539, + "grad_norm": 0.728886067867279, + "learning_rate": 8.453933958418264e-05, + "loss": 3.195, + "step": 107790 + }, + { + "epoch": 7.3240250033972005, + "grad_norm": 0.8242793679237366, + "learning_rate": 8.449687457534991e-05, + "loss": 3.4006, + "step": 107795 + }, + { + "epoch": 7.324364723467863, + "grad_norm": 1.072296380996704, + "learning_rate": 8.445440956651719e-05, + "loss": 3.4483, + "step": 107800 + }, + { + "epoch": 7.324704443538524, + "grad_norm": 1.0605820417404175, + "learning_rate": 8.441194455768447e-05, + "loss": 3.4133, + "step": 107805 + }, + { + "epoch": 7.325044163609186, + "grad_norm": 0.8600891828536987, + "learning_rate": 8.436947954885175e-05, + "loss": 3.5462, + "step": 107810 + }, + { + "epoch": 7.325383883679848, + "grad_norm": 0.8697220683097839, + "learning_rate": 8.432701454001902e-05, + "loss": 3.4229, + "step": 107815 + }, + { + "epoch": 7.325723603750509, + "grad_norm": 0.798488199710846, + "learning_rate": 8.428454953118631e-05, + "loss": 3.4019, + "step": 107820 + }, + { + "epoch": 7.326063323821171, + "grad_norm": 0.7795714735984802, + "learning_rate": 8.424208452235359e-05, + "loss": 3.2784, + "step": 107825 + }, + { + "epoch": 7.326403043891833, + "grad_norm": 0.9931150078773499, + "learning_rate": 8.419961951352086e-05, + "loss": 3.4317, + "step": 107830 + }, + { + "epoch": 7.3267427639624945, + "grad_norm": 0.8451365232467651, + "learning_rate": 8.415715450468814e-05, + "loss": 3.5254, + "step": 107835 + }, + { + "epoch": 7.3270824840331565, + "grad_norm": 0.962363064289093, + "learning_rate": 8.411468949585542e-05, + "loss": 3.3722, + "step": 107840 + }, + { + "epoch": 7.327422204103819, + "grad_norm": 0.9429982304573059, + "learning_rate": 8.407222448702268e-05, + "loss": 3.2821, + "step": 107845 + }, + { + "epoch": 7.32776192417448, + "grad_norm": 2.2665817737579346, + "learning_rate": 8.402975947818996e-05, + "loss": 3.5185, + "step": 107850 + }, + { + "epoch": 7.328101644245142, + "grad_norm": 1.0087672472000122, + "learning_rate": 8.398729446935726e-05, + "loss": 3.5055, + "step": 107855 + }, + { + "epoch": 7.328441364315804, + "grad_norm": 0.7280587553977966, + "learning_rate": 8.394482946052454e-05, + "loss": 3.4177, + "step": 107860 + }, + { + "epoch": 7.328781084386465, + "grad_norm": 0.8330195546150208, + "learning_rate": 8.39023644516918e-05, + "loss": 3.4727, + "step": 107865 + }, + { + "epoch": 7.329120804457127, + "grad_norm": 1.048551321029663, + "learning_rate": 8.385989944285908e-05, + "loss": 3.5467, + "step": 107870 + }, + { + "epoch": 7.329460524527789, + "grad_norm": 0.8761957287788391, + "learning_rate": 8.381743443402636e-05, + "loss": 3.4949, + "step": 107875 + }, + { + "epoch": 7.3298002445984505, + "grad_norm": 1.2490365505218506, + "learning_rate": 8.377496942519363e-05, + "loss": 3.4568, + "step": 107880 + }, + { + "epoch": 7.3301399646691126, + "grad_norm": 1.0214604139328003, + "learning_rate": 8.373250441636092e-05, + "loss": 3.5325, + "step": 107885 + }, + { + "epoch": 7.330479684739775, + "grad_norm": 1.0006330013275146, + "learning_rate": 8.36900394075282e-05, + "loss": 3.5759, + "step": 107890 + }, + { + "epoch": 7.330819404810436, + "grad_norm": 0.8977959752082825, + "learning_rate": 8.364757439869549e-05, + "loss": 3.3318, + "step": 107895 + }, + { + "epoch": 7.331159124881098, + "grad_norm": 0.8655332922935486, + "learning_rate": 8.360510938986275e-05, + "loss": 3.6581, + "step": 107900 + }, + { + "epoch": 7.33149884495176, + "grad_norm": 0.8481795787811279, + "learning_rate": 8.356264438103003e-05, + "loss": 3.3729, + "step": 107905 + }, + { + "epoch": 7.331838565022421, + "grad_norm": 0.8898553848266602, + "learning_rate": 8.352017937219731e-05, + "loss": 3.5049, + "step": 107910 + }, + { + "epoch": 7.332178285093083, + "grad_norm": 0.903596818447113, + "learning_rate": 8.347771436336458e-05, + "loss": 3.4754, + "step": 107915 + }, + { + "epoch": 7.332518005163745, + "grad_norm": 1.057130217552185, + "learning_rate": 8.343524935453187e-05, + "loss": 3.3978, + "step": 107920 + }, + { + "epoch": 7.3328577252344065, + "grad_norm": 0.8390891551971436, + "learning_rate": 8.339278434569915e-05, + "loss": 3.2319, + "step": 107925 + }, + { + "epoch": 7.333197445305069, + "grad_norm": 0.8714553713798523, + "learning_rate": 8.335031933686642e-05, + "loss": 3.2321, + "step": 107930 + }, + { + "epoch": 7.333537165375731, + "grad_norm": 0.9480482339859009, + "learning_rate": 8.33078543280337e-05, + "loss": 3.6039, + "step": 107935 + }, + { + "epoch": 7.333876885446392, + "grad_norm": 0.9670972228050232, + "learning_rate": 8.326538931920098e-05, + "loss": 3.5321, + "step": 107940 + }, + { + "epoch": 7.334216605517054, + "grad_norm": 0.9526634812355042, + "learning_rate": 8.322292431036826e-05, + "loss": 3.4828, + "step": 107945 + }, + { + "epoch": 7.334556325587716, + "grad_norm": 0.9669761061668396, + "learning_rate": 8.318045930153554e-05, + "loss": 3.3823, + "step": 107950 + }, + { + "epoch": 7.334896045658377, + "grad_norm": 0.7995757460594177, + "learning_rate": 8.313799429270282e-05, + "loss": 3.4967, + "step": 107955 + }, + { + "epoch": 7.335235765729039, + "grad_norm": 0.9108092784881592, + "learning_rate": 8.30955292838701e-05, + "loss": 3.2432, + "step": 107960 + }, + { + "epoch": 7.335575485799701, + "grad_norm": 0.8694682121276855, + "learning_rate": 8.305306427503737e-05, + "loss": 3.3208, + "step": 107965 + }, + { + "epoch": 7.3359152058703625, + "grad_norm": 1.5566470623016357, + "learning_rate": 8.301059926620465e-05, + "loss": 3.0164, + "step": 107970 + }, + { + "epoch": 7.336254925941025, + "grad_norm": 0.9309690594673157, + "learning_rate": 8.296813425737193e-05, + "loss": 3.1665, + "step": 107975 + }, + { + "epoch": 7.336594646011687, + "grad_norm": 1.0616323947906494, + "learning_rate": 8.29256692485392e-05, + "loss": 3.38, + "step": 107980 + }, + { + "epoch": 7.336934366082348, + "grad_norm": 0.8606554865837097, + "learning_rate": 8.288320423970649e-05, + "loss": 3.7303, + "step": 107985 + }, + { + "epoch": 7.33727408615301, + "grad_norm": 0.7815635204315186, + "learning_rate": 8.284073923087377e-05, + "loss": 3.2592, + "step": 107990 + }, + { + "epoch": 7.337613806223672, + "grad_norm": 1.0102550983428955, + "learning_rate": 8.279827422204105e-05, + "loss": 3.4743, + "step": 107995 + }, + { + "epoch": 7.337953526294333, + "grad_norm": 0.9203001260757446, + "learning_rate": 8.275580921320831e-05, + "loss": 3.3378, + "step": 108000 + }, + { + "epoch": 7.338293246364995, + "grad_norm": 1.0358006954193115, + "learning_rate": 8.271334420437559e-05, + "loss": 3.5542, + "step": 108005 + }, + { + "epoch": 7.338632966435657, + "grad_norm": 0.8995512127876282, + "learning_rate": 8.267087919554287e-05, + "loss": 3.4681, + "step": 108010 + }, + { + "epoch": 7.3389726865063185, + "grad_norm": 0.8615922331809998, + "learning_rate": 8.262841418671015e-05, + "loss": 3.5606, + "step": 108015 + }, + { + "epoch": 7.339312406576981, + "grad_norm": 0.949508547782898, + "learning_rate": 8.258594917787743e-05, + "loss": 3.3919, + "step": 108020 + }, + { + "epoch": 7.339652126647643, + "grad_norm": 0.8685413599014282, + "learning_rate": 8.254348416904471e-05, + "loss": 3.5122, + "step": 108025 + }, + { + "epoch": 7.339991846718304, + "grad_norm": 0.9927236437797546, + "learning_rate": 8.250101916021199e-05, + "loss": 3.6078, + "step": 108030 + }, + { + "epoch": 7.340331566788966, + "grad_norm": 0.9444804191589355, + "learning_rate": 8.245855415137926e-05, + "loss": 3.3532, + "step": 108035 + }, + { + "epoch": 7.340671286859628, + "grad_norm": 0.8974268436431885, + "learning_rate": 8.241608914254654e-05, + "loss": 3.4356, + "step": 108040 + }, + { + "epoch": 7.341011006930289, + "grad_norm": 1.2377915382385254, + "learning_rate": 8.237362413371382e-05, + "loss": 3.6819, + "step": 108045 + }, + { + "epoch": 7.341350727000951, + "grad_norm": 1.1674448251724243, + "learning_rate": 8.23311591248811e-05, + "loss": 3.3233, + "step": 108050 + }, + { + "epoch": 7.341690447071613, + "grad_norm": 0.9541130661964417, + "learning_rate": 8.228869411604838e-05, + "loss": 3.4151, + "step": 108055 + }, + { + "epoch": 7.3420301671422745, + "grad_norm": 0.8057922720909119, + "learning_rate": 8.224622910721566e-05, + "loss": 3.5368, + "step": 108060 + }, + { + "epoch": 7.342369887212937, + "grad_norm": 0.7784551382064819, + "learning_rate": 8.220376409838294e-05, + "loss": 3.4019, + "step": 108065 + }, + { + "epoch": 7.342709607283599, + "grad_norm": 0.7299333810806274, + "learning_rate": 8.21612990895502e-05, + "loss": 3.56, + "step": 108070 + }, + { + "epoch": 7.34304932735426, + "grad_norm": 0.7801570296287537, + "learning_rate": 8.211883408071749e-05, + "loss": 3.299, + "step": 108075 + }, + { + "epoch": 7.343389047424922, + "grad_norm": 0.7803922295570374, + "learning_rate": 8.207636907188478e-05, + "loss": 3.4279, + "step": 108080 + }, + { + "epoch": 7.343728767495584, + "grad_norm": 0.7911625504493713, + "learning_rate": 8.203390406305205e-05, + "loss": 3.462, + "step": 108085 + }, + { + "epoch": 7.344068487566245, + "grad_norm": 1.0405676364898682, + "learning_rate": 8.199143905421933e-05, + "loss": 3.3255, + "step": 108090 + }, + { + "epoch": 7.344408207636907, + "grad_norm": 0.9080454707145691, + "learning_rate": 8.194897404538661e-05, + "loss": 3.3057, + "step": 108095 + }, + { + "epoch": 7.344747927707569, + "grad_norm": 0.7944557070732117, + "learning_rate": 8.190650903655387e-05, + "loss": 3.1201, + "step": 108100 + }, + { + "epoch": 7.3450876477782305, + "grad_norm": 0.7965056300163269, + "learning_rate": 8.186404402772115e-05, + "loss": 3.4372, + "step": 108105 + }, + { + "epoch": 7.345427367848893, + "grad_norm": 0.7723160982131958, + "learning_rate": 8.182157901888843e-05, + "loss": 3.8188, + "step": 108110 + }, + { + "epoch": 7.345767087919555, + "grad_norm": 0.6610286831855774, + "learning_rate": 8.177911401005573e-05, + "loss": 3.4049, + "step": 108115 + }, + { + "epoch": 7.346106807990216, + "grad_norm": 0.9732816815376282, + "learning_rate": 8.1736649001223e-05, + "loss": 3.0379, + "step": 108120 + }, + { + "epoch": 7.346446528060878, + "grad_norm": 1.4723745584487915, + "learning_rate": 8.169418399239027e-05, + "loss": 3.5451, + "step": 108125 + }, + { + "epoch": 7.34678624813154, + "grad_norm": 1.0435247421264648, + "learning_rate": 8.165171898355755e-05, + "loss": 3.3065, + "step": 108130 + }, + { + "epoch": 7.347125968202201, + "grad_norm": 0.9035091996192932, + "learning_rate": 8.160925397472482e-05, + "loss": 3.1052, + "step": 108135 + }, + { + "epoch": 7.347465688272863, + "grad_norm": 0.8666511178016663, + "learning_rate": 8.15667889658921e-05, + "loss": 3.2337, + "step": 108140 + }, + { + "epoch": 7.347805408343525, + "grad_norm": 1.0023013353347778, + "learning_rate": 8.15243239570594e-05, + "loss": 3.1479, + "step": 108145 + }, + { + "epoch": 7.3481451284141865, + "grad_norm": 1.0774047374725342, + "learning_rate": 8.148185894822667e-05, + "loss": 3.5066, + "step": 108150 + }, + { + "epoch": 7.348484848484849, + "grad_norm": 1.1266034841537476, + "learning_rate": 8.143939393939394e-05, + "loss": 3.355, + "step": 108155 + }, + { + "epoch": 7.348824568555511, + "grad_norm": 0.873637318611145, + "learning_rate": 8.139692893056122e-05, + "loss": 3.609, + "step": 108160 + }, + { + "epoch": 7.349164288626172, + "grad_norm": 0.9227760434150696, + "learning_rate": 8.13544639217285e-05, + "loss": 3.4881, + "step": 108165 + }, + { + "epoch": 7.349504008696834, + "grad_norm": 1.1071733236312866, + "learning_rate": 8.131199891289577e-05, + "loss": 3.3653, + "step": 108170 + }, + { + "epoch": 7.349843728767496, + "grad_norm": 0.9310802221298218, + "learning_rate": 8.126953390406305e-05, + "loss": 3.4962, + "step": 108175 + }, + { + "epoch": 7.350183448838157, + "grad_norm": 0.8424489498138428, + "learning_rate": 8.122706889523034e-05, + "loss": 3.6485, + "step": 108180 + }, + { + "epoch": 7.350523168908819, + "grad_norm": 0.8273205161094666, + "learning_rate": 8.118460388639761e-05, + "loss": 3.2509, + "step": 108185 + }, + { + "epoch": 7.350862888979481, + "grad_norm": 0.8774218559265137, + "learning_rate": 8.114213887756489e-05, + "loss": 3.4807, + "step": 108190 + }, + { + "epoch": 7.3512026090501426, + "grad_norm": 0.9253801703453064, + "learning_rate": 8.109967386873217e-05, + "loss": 3.461, + "step": 108195 + }, + { + "epoch": 7.351542329120805, + "grad_norm": 1.0547844171524048, + "learning_rate": 8.105720885989945e-05, + "loss": 3.0784, + "step": 108200 + }, + { + "epoch": 7.351882049191466, + "grad_norm": 1.129296064376831, + "learning_rate": 8.101474385106671e-05, + "loss": 3.2888, + "step": 108205 + }, + { + "epoch": 7.352221769262128, + "grad_norm": 0.7377391457557678, + "learning_rate": 8.097227884223401e-05, + "loss": 3.4174, + "step": 108210 + }, + { + "epoch": 7.35256148933279, + "grad_norm": 0.8187551498413086, + "learning_rate": 8.092981383340129e-05, + "loss": 3.6157, + "step": 108215 + }, + { + "epoch": 7.352901209403451, + "grad_norm": 0.8447620272636414, + "learning_rate": 8.088734882456855e-05, + "loss": 3.3799, + "step": 108220 + }, + { + "epoch": 7.353240929474113, + "grad_norm": 1.030419945716858, + "learning_rate": 8.084488381573583e-05, + "loss": 3.4766, + "step": 108225 + }, + { + "epoch": 7.353580649544775, + "grad_norm": 0.884154736995697, + "learning_rate": 8.080241880690311e-05, + "loss": 3.4768, + "step": 108230 + }, + { + "epoch": 7.3539203696154365, + "grad_norm": 1.1728287935256958, + "learning_rate": 8.07599537980704e-05, + "loss": 3.2911, + "step": 108235 + }, + { + "epoch": 7.354260089686099, + "grad_norm": 0.6798505783081055, + "learning_rate": 8.071748878923766e-05, + "loss": 3.4404, + "step": 108240 + }, + { + "epoch": 7.354599809756761, + "grad_norm": 0.8838697075843811, + "learning_rate": 8.067502378040495e-05, + "loss": 3.7085, + "step": 108245 + }, + { + "epoch": 7.354939529827422, + "grad_norm": 0.7868896126747131, + "learning_rate": 8.063255877157224e-05, + "loss": 3.4759, + "step": 108250 + }, + { + "epoch": 7.355279249898084, + "grad_norm": 1.115308165550232, + "learning_rate": 8.05900937627395e-05, + "loss": 3.4094, + "step": 108255 + }, + { + "epoch": 7.355618969968746, + "grad_norm": 1.1845316886901855, + "learning_rate": 8.054762875390678e-05, + "loss": 3.5418, + "step": 108260 + }, + { + "epoch": 7.355958690039407, + "grad_norm": 0.8508946895599365, + "learning_rate": 8.050516374507406e-05, + "loss": 3.4961, + "step": 108265 + }, + { + "epoch": 7.356298410110069, + "grad_norm": 0.8119141459465027, + "learning_rate": 8.046269873624133e-05, + "loss": 3.4016, + "step": 108270 + }, + { + "epoch": 7.356638130180731, + "grad_norm": 0.9017042517662048, + "learning_rate": 8.042023372740861e-05, + "loss": 3.5362, + "step": 108275 + }, + { + "epoch": 7.3569778502513925, + "grad_norm": 0.775424599647522, + "learning_rate": 8.03777687185759e-05, + "loss": 3.5469, + "step": 108280 + }, + { + "epoch": 7.357317570322055, + "grad_norm": 0.9882447123527527, + "learning_rate": 8.033530370974318e-05, + "loss": 3.5374, + "step": 108285 + }, + { + "epoch": 7.357657290392717, + "grad_norm": 0.9443387389183044, + "learning_rate": 8.029283870091045e-05, + "loss": 3.2298, + "step": 108290 + }, + { + "epoch": 7.357997010463378, + "grad_norm": 0.9321837425231934, + "learning_rate": 8.025037369207773e-05, + "loss": 3.5566, + "step": 108295 + }, + { + "epoch": 7.35833673053404, + "grad_norm": 0.8999015688896179, + "learning_rate": 8.020790868324501e-05, + "loss": 3.2951, + "step": 108300 + }, + { + "epoch": 7.358676450604702, + "grad_norm": 1.0706946849822998, + "learning_rate": 8.016544367441228e-05, + "loss": 3.5397, + "step": 108305 + }, + { + "epoch": 7.359016170675363, + "grad_norm": 0.9312024712562561, + "learning_rate": 8.012297866557957e-05, + "loss": 3.4407, + "step": 108310 + }, + { + "epoch": 7.359355890746025, + "grad_norm": 0.7699271440505981, + "learning_rate": 8.008051365674685e-05, + "loss": 3.384, + "step": 108315 + }, + { + "epoch": 7.359695610816687, + "grad_norm": 0.9022674560546875, + "learning_rate": 8.003804864791413e-05, + "loss": 3.344, + "step": 108320 + }, + { + "epoch": 7.3600353308873485, + "grad_norm": 1.0527420043945312, + "learning_rate": 7.99955836390814e-05, + "loss": 3.4335, + "step": 108325 + }, + { + "epoch": 7.360375050958011, + "grad_norm": 0.7516013383865356, + "learning_rate": 7.995311863024868e-05, + "loss": 3.2959, + "step": 108330 + }, + { + "epoch": 7.360714771028673, + "grad_norm": 0.7266879081726074, + "learning_rate": 7.991065362141596e-05, + "loss": 3.672, + "step": 108335 + }, + { + "epoch": 7.361054491099334, + "grad_norm": 1.1054710149765015, + "learning_rate": 7.986818861258322e-05, + "loss": 3.5137, + "step": 108340 + }, + { + "epoch": 7.361394211169996, + "grad_norm": 0.890555739402771, + "learning_rate": 7.982572360375052e-05, + "loss": 3.2393, + "step": 108345 + }, + { + "epoch": 7.361733931240658, + "grad_norm": 1.1597766876220703, + "learning_rate": 7.97832585949178e-05, + "loss": 3.0959, + "step": 108350 + }, + { + "epoch": 7.362073651311319, + "grad_norm": 1.1957168579101562, + "learning_rate": 7.974079358608506e-05, + "loss": 3.2867, + "step": 108355 + }, + { + "epoch": 7.362413371381981, + "grad_norm": 0.9507166743278503, + "learning_rate": 7.969832857725234e-05, + "loss": 3.4213, + "step": 108360 + }, + { + "epoch": 7.362753091452643, + "grad_norm": 0.7997114062309265, + "learning_rate": 7.965586356841962e-05, + "loss": 3.4923, + "step": 108365 + }, + { + "epoch": 7.3630928115233045, + "grad_norm": 1.0644311904907227, + "learning_rate": 7.96133985595869e-05, + "loss": 3.423, + "step": 108370 + }, + { + "epoch": 7.363432531593967, + "grad_norm": 1.05824613571167, + "learning_rate": 7.957093355075418e-05, + "loss": 3.3916, + "step": 108375 + }, + { + "epoch": 7.363772251664629, + "grad_norm": 0.8700168132781982, + "learning_rate": 7.952846854192146e-05, + "loss": 3.6112, + "step": 108380 + }, + { + "epoch": 7.36411197173529, + "grad_norm": 1.0517839193344116, + "learning_rate": 7.948600353308874e-05, + "loss": 3.544, + "step": 108385 + }, + { + "epoch": 7.364451691805952, + "grad_norm": 1.1109884977340698, + "learning_rate": 7.944353852425601e-05, + "loss": 2.955, + "step": 108390 + }, + { + "epoch": 7.364791411876614, + "grad_norm": 1.0801697969436646, + "learning_rate": 7.940107351542329e-05, + "loss": 3.382, + "step": 108395 + }, + { + "epoch": 7.365131131947275, + "grad_norm": 0.9609904885292053, + "learning_rate": 7.935860850659057e-05, + "loss": 3.2012, + "step": 108400 + }, + { + "epoch": 7.365470852017937, + "grad_norm": 0.8115131258964539, + "learning_rate": 7.931614349775785e-05, + "loss": 3.2604, + "step": 108405 + }, + { + "epoch": 7.365810572088599, + "grad_norm": 1.0339816808700562, + "learning_rate": 7.927367848892513e-05, + "loss": 3.1233, + "step": 108410 + }, + { + "epoch": 7.3661502921592605, + "grad_norm": 0.8646503686904907, + "learning_rate": 7.923121348009241e-05, + "loss": 3.2328, + "step": 108415 + }, + { + "epoch": 7.366490012229923, + "grad_norm": 1.0038965940475464, + "learning_rate": 7.918874847125969e-05, + "loss": 3.2664, + "step": 108420 + }, + { + "epoch": 7.366829732300585, + "grad_norm": 0.9860743284225464, + "learning_rate": 7.914628346242696e-05, + "loss": 3.5601, + "step": 108425 + }, + { + "epoch": 7.367169452371246, + "grad_norm": 3.7118241786956787, + "learning_rate": 7.910381845359424e-05, + "loss": 3.1668, + "step": 108430 + }, + { + "epoch": 7.367509172441908, + "grad_norm": 0.9937576651573181, + "learning_rate": 7.906135344476152e-05, + "loss": 3.6225, + "step": 108435 + }, + { + "epoch": 7.36784889251257, + "grad_norm": 0.8687572479248047, + "learning_rate": 7.90188884359288e-05, + "loss": 3.5621, + "step": 108440 + }, + { + "epoch": 7.368188612583231, + "grad_norm": 1.1762768030166626, + "learning_rate": 7.897642342709608e-05, + "loss": 3.2681, + "step": 108445 + }, + { + "epoch": 7.368528332653893, + "grad_norm": 0.9401125311851501, + "learning_rate": 7.893395841826336e-05, + "loss": 3.5379, + "step": 108450 + }, + { + "epoch": 7.368868052724555, + "grad_norm": 0.8845216631889343, + "learning_rate": 7.889149340943064e-05, + "loss": 3.4379, + "step": 108455 + }, + { + "epoch": 7.3692077727952165, + "grad_norm": 0.9081280827522278, + "learning_rate": 7.88490284005979e-05, + "loss": 3.4007, + "step": 108460 + }, + { + "epoch": 7.369547492865879, + "grad_norm": 0.8586682081222534, + "learning_rate": 7.880656339176518e-05, + "loss": 3.2975, + "step": 108465 + }, + { + "epoch": 7.369887212936541, + "grad_norm": 0.8872316479682922, + "learning_rate": 7.876409838293246e-05, + "loss": 3.3245, + "step": 108470 + }, + { + "epoch": 7.370226933007202, + "grad_norm": 0.7201442718505859, + "learning_rate": 7.872163337409974e-05, + "loss": 3.3255, + "step": 108475 + }, + { + "epoch": 7.370566653077864, + "grad_norm": 0.784247875213623, + "learning_rate": 7.867916836526702e-05, + "loss": 3.4149, + "step": 108480 + }, + { + "epoch": 7.370906373148525, + "grad_norm": 0.9139862060546875, + "learning_rate": 7.86367033564343e-05, + "loss": 3.5281, + "step": 108485 + }, + { + "epoch": 7.371246093219187, + "grad_norm": 1.018827199935913, + "learning_rate": 7.859423834760158e-05, + "loss": 3.3271, + "step": 108490 + }, + { + "epoch": 7.371585813289849, + "grad_norm": 1.4374910593032837, + "learning_rate": 7.855177333876885e-05, + "loss": 3.2927, + "step": 108495 + }, + { + "epoch": 7.3719255333605105, + "grad_norm": 0.8623530268669128, + "learning_rate": 7.850930832993613e-05, + "loss": 3.361, + "step": 108500 + }, + { + "epoch": 7.372265253431173, + "grad_norm": 0.7925111055374146, + "learning_rate": 7.846684332110342e-05, + "loss": 3.4554, + "step": 108505 + }, + { + "epoch": 7.372604973501835, + "grad_norm": 0.8856112957000732, + "learning_rate": 7.842437831227069e-05, + "loss": 3.4435, + "step": 108510 + }, + { + "epoch": 7.372944693572496, + "grad_norm": 0.8187741637229919, + "learning_rate": 7.838191330343797e-05, + "loss": 3.6619, + "step": 108515 + }, + { + "epoch": 7.373284413643158, + "grad_norm": 0.7865045070648193, + "learning_rate": 7.833944829460525e-05, + "loss": 3.6965, + "step": 108520 + }, + { + "epoch": 7.37362413371382, + "grad_norm": 1.07732355594635, + "learning_rate": 7.829698328577252e-05, + "loss": 3.3292, + "step": 108525 + }, + { + "epoch": 7.373963853784481, + "grad_norm": 1.1698788404464722, + "learning_rate": 7.82545182769398e-05, + "loss": 3.2386, + "step": 108530 + }, + { + "epoch": 7.374303573855143, + "grad_norm": 0.9270304441452026, + "learning_rate": 7.821205326810708e-05, + "loss": 3.4056, + "step": 108535 + }, + { + "epoch": 7.374643293925805, + "grad_norm": 0.8269317150115967, + "learning_rate": 7.816958825927437e-05, + "loss": 3.3837, + "step": 108540 + }, + { + "epoch": 7.3749830139964665, + "grad_norm": 0.796012818813324, + "learning_rate": 7.812712325044164e-05, + "loss": 3.4102, + "step": 108545 + }, + { + "epoch": 7.375322734067129, + "grad_norm": 0.8778824210166931, + "learning_rate": 7.808465824160892e-05, + "loss": 3.2212, + "step": 108550 + }, + { + "epoch": 7.375662454137791, + "grad_norm": 0.7691001892089844, + "learning_rate": 7.80421932327762e-05, + "loss": 3.6094, + "step": 108555 + }, + { + "epoch": 7.376002174208452, + "grad_norm": 1.2202004194259644, + "learning_rate": 7.799972822394346e-05, + "loss": 3.5188, + "step": 108560 + }, + { + "epoch": 7.376341894279114, + "grad_norm": 0.9468342065811157, + "learning_rate": 7.795726321511074e-05, + "loss": 3.3915, + "step": 108565 + }, + { + "epoch": 7.376681614349776, + "grad_norm": 0.889205276966095, + "learning_rate": 7.791479820627804e-05, + "loss": 3.5099, + "step": 108570 + }, + { + "epoch": 7.377021334420437, + "grad_norm": 1.050233244895935, + "learning_rate": 7.78723331974453e-05, + "loss": 3.5637, + "step": 108575 + }, + { + "epoch": 7.377361054491099, + "grad_norm": 1.0188441276550293, + "learning_rate": 7.782986818861258e-05, + "loss": 3.4345, + "step": 108580 + }, + { + "epoch": 7.377700774561761, + "grad_norm": 0.9444283843040466, + "learning_rate": 7.778740317977986e-05, + "loss": 3.4514, + "step": 108585 + }, + { + "epoch": 7.3780404946324225, + "grad_norm": 1.0428990125656128, + "learning_rate": 7.774493817094714e-05, + "loss": 3.5051, + "step": 108590 + }, + { + "epoch": 7.378380214703085, + "grad_norm": 0.9623711705207825, + "learning_rate": 7.770247316211441e-05, + "loss": 3.6014, + "step": 108595 + }, + { + "epoch": 7.378719934773747, + "grad_norm": 0.770611584186554, + "learning_rate": 7.766000815328169e-05, + "loss": 3.29, + "step": 108600 + }, + { + "epoch": 7.379059654844408, + "grad_norm": 0.9828957319259644, + "learning_rate": 7.761754314444898e-05, + "loss": 3.3056, + "step": 108605 + }, + { + "epoch": 7.37939937491507, + "grad_norm": 0.8040779829025269, + "learning_rate": 7.757507813561625e-05, + "loss": 3.2686, + "step": 108610 + }, + { + "epoch": 7.379739094985732, + "grad_norm": 0.8790755867958069, + "learning_rate": 7.753261312678353e-05, + "loss": 3.6983, + "step": 108615 + }, + { + "epoch": 7.380078815056393, + "grad_norm": 0.9021618366241455, + "learning_rate": 7.749014811795081e-05, + "loss": 3.4091, + "step": 108620 + }, + { + "epoch": 7.380418535127055, + "grad_norm": 1.1559405326843262, + "learning_rate": 7.744768310911809e-05, + "loss": 3.5123, + "step": 108625 + }, + { + "epoch": 7.380758255197717, + "grad_norm": 0.8557308316230774, + "learning_rate": 7.740521810028536e-05, + "loss": 3.3387, + "step": 108630 + }, + { + "epoch": 7.3810979752683785, + "grad_norm": 0.8494070172309875, + "learning_rate": 7.736275309145264e-05, + "loss": 3.3037, + "step": 108635 + }, + { + "epoch": 7.381437695339041, + "grad_norm": 0.9849132299423218, + "learning_rate": 7.732028808261993e-05, + "loss": 3.5677, + "step": 108640 + }, + { + "epoch": 7.381777415409703, + "grad_norm": 0.9838067293167114, + "learning_rate": 7.72778230737872e-05, + "loss": 3.5172, + "step": 108645 + }, + { + "epoch": 7.382117135480364, + "grad_norm": 0.8861747980117798, + "learning_rate": 7.723535806495448e-05, + "loss": 3.3951, + "step": 108650 + }, + { + "epoch": 7.382456855551026, + "grad_norm": 1.1223593950271606, + "learning_rate": 7.719289305612176e-05, + "loss": 3.3503, + "step": 108655 + }, + { + "epoch": 7.382796575621688, + "grad_norm": 0.8330774903297424, + "learning_rate": 7.715042804728903e-05, + "loss": 3.449, + "step": 108660 + }, + { + "epoch": 7.383136295692349, + "grad_norm": 1.0974870920181274, + "learning_rate": 7.71079630384563e-05, + "loss": 3.0533, + "step": 108665 + }, + { + "epoch": 7.383476015763011, + "grad_norm": 0.947439968585968, + "learning_rate": 7.70654980296236e-05, + "loss": 3.2871, + "step": 108670 + }, + { + "epoch": 7.383815735833673, + "grad_norm": 0.8520537614822388, + "learning_rate": 7.702303302079088e-05, + "loss": 3.3696, + "step": 108675 + }, + { + "epoch": 7.3841554559043345, + "grad_norm": 0.9360328912734985, + "learning_rate": 7.698056801195815e-05, + "loss": 3.3586, + "step": 108680 + }, + { + "epoch": 7.384495175974997, + "grad_norm": 0.8290603756904602, + "learning_rate": 7.693810300312543e-05, + "loss": 3.726, + "step": 108685 + }, + { + "epoch": 7.384834896045659, + "grad_norm": 0.8845439553260803, + "learning_rate": 7.68956379942927e-05, + "loss": 3.379, + "step": 108690 + }, + { + "epoch": 7.38517461611632, + "grad_norm": 1.0937515497207642, + "learning_rate": 7.685317298545997e-05, + "loss": 3.3014, + "step": 108695 + }, + { + "epoch": 7.385514336186982, + "grad_norm": 1.256439208984375, + "learning_rate": 7.681070797662725e-05, + "loss": 3.254, + "step": 108700 + }, + { + "epoch": 7.385854056257644, + "grad_norm": 1.2918915748596191, + "learning_rate": 7.676824296779455e-05, + "loss": 3.4275, + "step": 108705 + }, + { + "epoch": 7.386193776328305, + "grad_norm": 0.7958886623382568, + "learning_rate": 7.672577795896183e-05, + "loss": 3.4371, + "step": 108710 + }, + { + "epoch": 7.386533496398967, + "grad_norm": 0.8987610340118408, + "learning_rate": 7.668331295012909e-05, + "loss": 3.2994, + "step": 108715 + }, + { + "epoch": 7.386873216469629, + "grad_norm": 0.9478507041931152, + "learning_rate": 7.664084794129637e-05, + "loss": 3.5705, + "step": 108720 + }, + { + "epoch": 7.3872129365402905, + "grad_norm": 0.9301944375038147, + "learning_rate": 7.659838293246365e-05, + "loss": 3.51, + "step": 108725 + }, + { + "epoch": 7.387552656610953, + "grad_norm": 1.069493055343628, + "learning_rate": 7.655591792363092e-05, + "loss": 3.3749, + "step": 108730 + }, + { + "epoch": 7.387892376681615, + "grad_norm": 1.0526881217956543, + "learning_rate": 7.651345291479821e-05, + "loss": 3.5893, + "step": 108735 + }, + { + "epoch": 7.388232096752276, + "grad_norm": 1.0788267850875854, + "learning_rate": 7.647098790596549e-05, + "loss": 3.4895, + "step": 108740 + }, + { + "epoch": 7.388571816822938, + "grad_norm": 0.8771752715110779, + "learning_rate": 7.642852289713276e-05, + "loss": 3.4834, + "step": 108745 + }, + { + "epoch": 7.3889115368936, + "grad_norm": 0.8939110040664673, + "learning_rate": 7.638605788830004e-05, + "loss": 3.3556, + "step": 108750 + }, + { + "epoch": 7.389251256964261, + "grad_norm": 0.8363263607025146, + "learning_rate": 7.634359287946732e-05, + "loss": 3.7008, + "step": 108755 + }, + { + "epoch": 7.389590977034923, + "grad_norm": 0.987054705619812, + "learning_rate": 7.63011278706346e-05, + "loss": 3.3091, + "step": 108760 + }, + { + "epoch": 7.389930697105585, + "grad_norm": 1.3022857904434204, + "learning_rate": 7.625866286180187e-05, + "loss": 3.3306, + "step": 108765 + }, + { + "epoch": 7.3902704171762466, + "grad_norm": 0.9912842512130737, + "learning_rate": 7.621619785296916e-05, + "loss": 3.4346, + "step": 108770 + }, + { + "epoch": 7.390610137246909, + "grad_norm": 0.833530604839325, + "learning_rate": 7.617373284413644e-05, + "loss": 3.4538, + "step": 108775 + }, + { + "epoch": 7.390949857317571, + "grad_norm": 1.0182992219924927, + "learning_rate": 7.61312678353037e-05, + "loss": 3.7227, + "step": 108780 + }, + { + "epoch": 7.391289577388232, + "grad_norm": 0.9638214111328125, + "learning_rate": 7.608880282647099e-05, + "loss": 3.5231, + "step": 108785 + }, + { + "epoch": 7.391629297458894, + "grad_norm": 0.9906840324401855, + "learning_rate": 7.604633781763827e-05, + "loss": 3.4886, + "step": 108790 + }, + { + "epoch": 7.391969017529556, + "grad_norm": 1.0783493518829346, + "learning_rate": 7.600387280880555e-05, + "loss": 3.3802, + "step": 108795 + }, + { + "epoch": 7.392308737600217, + "grad_norm": 0.9286695122718811, + "learning_rate": 7.596140779997283e-05, + "loss": 3.3918, + "step": 108800 + }, + { + "epoch": 7.392648457670879, + "grad_norm": 1.0722134113311768, + "learning_rate": 7.59189427911401e-05, + "loss": 3.6304, + "step": 108805 + }, + { + "epoch": 7.392988177741541, + "grad_norm": 1.048500418663025, + "learning_rate": 7.587647778230739e-05, + "loss": 3.4126, + "step": 108810 + }, + { + "epoch": 7.393327897812203, + "grad_norm": 0.8750117421150208, + "learning_rate": 7.583401277347465e-05, + "loss": 3.3311, + "step": 108815 + }, + { + "epoch": 7.393667617882865, + "grad_norm": 0.9940992593765259, + "learning_rate": 7.579154776464193e-05, + "loss": 3.6763, + "step": 108820 + }, + { + "epoch": 7.394007337953527, + "grad_norm": 0.8729440569877625, + "learning_rate": 7.574908275580921e-05, + "loss": 3.4946, + "step": 108825 + }, + { + "epoch": 7.394347058024188, + "grad_norm": 0.7479689717292786, + "learning_rate": 7.570661774697648e-05, + "loss": 3.664, + "step": 108830 + }, + { + "epoch": 7.39468677809485, + "grad_norm": 0.8776648044586182, + "learning_rate": 7.566415273814377e-05, + "loss": 3.2808, + "step": 108835 + }, + { + "epoch": 7.395026498165512, + "grad_norm": 0.7777795195579529, + "learning_rate": 7.562168772931105e-05, + "loss": 3.3919, + "step": 108840 + }, + { + "epoch": 7.395366218236173, + "grad_norm": 0.9615136384963989, + "learning_rate": 7.557922272047833e-05, + "loss": 3.5661, + "step": 108845 + }, + { + "epoch": 7.395705938306835, + "grad_norm": 0.8547780513763428, + "learning_rate": 7.55367577116456e-05, + "loss": 3.2303, + "step": 108850 + }, + { + "epoch": 7.396045658377497, + "grad_norm": 0.8625694513320923, + "learning_rate": 7.549429270281288e-05, + "loss": 3.3904, + "step": 108855 + }, + { + "epoch": 7.396385378448159, + "grad_norm": 0.8853688836097717, + "learning_rate": 7.545182769398016e-05, + "loss": 3.3837, + "step": 108860 + }, + { + "epoch": 7.396725098518821, + "grad_norm": 1.1897940635681152, + "learning_rate": 7.540936268514744e-05, + "loss": 3.2404, + "step": 108865 + }, + { + "epoch": 7.397064818589483, + "grad_norm": 0.8775194883346558, + "learning_rate": 7.536689767631472e-05, + "loss": 3.3294, + "step": 108870 + }, + { + "epoch": 7.397404538660144, + "grad_norm": 2.016775369644165, + "learning_rate": 7.5324432667482e-05, + "loss": 3.5714, + "step": 108875 + }, + { + "epoch": 7.397744258730806, + "grad_norm": 0.8640131950378418, + "learning_rate": 7.528196765864928e-05, + "loss": 3.5967, + "step": 108880 + }, + { + "epoch": 7.398083978801467, + "grad_norm": 1.1205579042434692, + "learning_rate": 7.523950264981655e-05, + "loss": 3.4264, + "step": 108885 + }, + { + "epoch": 7.398423698872129, + "grad_norm": 0.8593012094497681, + "learning_rate": 7.519703764098383e-05, + "loss": 3.4121, + "step": 108890 + }, + { + "epoch": 7.398763418942791, + "grad_norm": 0.8305783867835999, + "learning_rate": 7.515457263215111e-05, + "loss": 3.1685, + "step": 108895 + }, + { + "epoch": 7.3991031390134525, + "grad_norm": 1.059627890586853, + "learning_rate": 7.511210762331839e-05, + "loss": 3.2316, + "step": 108900 + }, + { + "epoch": 7.399442859084115, + "grad_norm": 0.8926761746406555, + "learning_rate": 7.506964261448567e-05, + "loss": 3.41, + "step": 108905 + }, + { + "epoch": 7.399782579154777, + "grad_norm": 0.7658839821815491, + "learning_rate": 7.502717760565295e-05, + "loss": 3.3431, + "step": 108910 + }, + { + "epoch": 7.400122299225438, + "grad_norm": 1.2107939720153809, + "learning_rate": 7.498471259682021e-05, + "loss": 3.3389, + "step": 108915 + }, + { + "epoch": 7.4004620192961, + "grad_norm": 0.8407041430473328, + "learning_rate": 7.49422475879875e-05, + "loss": 3.2322, + "step": 108920 + }, + { + "epoch": 7.400801739366762, + "grad_norm": 0.9181031584739685, + "learning_rate": 7.489978257915477e-05, + "loss": 3.4538, + "step": 108925 + }, + { + "epoch": 7.401141459437423, + "grad_norm": 1.0137367248535156, + "learning_rate": 7.485731757032207e-05, + "loss": 3.4811, + "step": 108930 + }, + { + "epoch": 7.401481179508085, + "grad_norm": 0.9607706069946289, + "learning_rate": 7.481485256148933e-05, + "loss": 3.4217, + "step": 108935 + }, + { + "epoch": 7.401820899578747, + "grad_norm": 0.9311674237251282, + "learning_rate": 7.477238755265661e-05, + "loss": 3.4537, + "step": 108940 + }, + { + "epoch": 7.4021606196494085, + "grad_norm": 1.0805103778839111, + "learning_rate": 7.47299225438239e-05, + "loss": 3.1805, + "step": 108945 + }, + { + "epoch": 7.402500339720071, + "grad_norm": 0.8960440158843994, + "learning_rate": 7.468745753499116e-05, + "loss": 3.3954, + "step": 108950 + }, + { + "epoch": 7.402840059790733, + "grad_norm": 0.9124037027359009, + "learning_rate": 7.464499252615844e-05, + "loss": 3.3928, + "step": 108955 + }, + { + "epoch": 7.403179779861394, + "grad_norm": 0.8860113620758057, + "learning_rate": 7.460252751732572e-05, + "loss": 3.5762, + "step": 108960 + }, + { + "epoch": 7.403519499932056, + "grad_norm": 0.9793386459350586, + "learning_rate": 7.456006250849302e-05, + "loss": 3.4473, + "step": 108965 + }, + { + "epoch": 7.403859220002718, + "grad_norm": 1.0126973390579224, + "learning_rate": 7.451759749966028e-05, + "loss": 3.324, + "step": 108970 + }, + { + "epoch": 7.404198940073379, + "grad_norm": 1.0484081506729126, + "learning_rate": 7.447513249082756e-05, + "loss": 3.4212, + "step": 108975 + }, + { + "epoch": 7.404538660144041, + "grad_norm": 1.0509047508239746, + "learning_rate": 7.443266748199484e-05, + "loss": 3.2109, + "step": 108980 + }, + { + "epoch": 7.404878380214703, + "grad_norm": 1.0090023279190063, + "learning_rate": 7.439020247316211e-05, + "loss": 3.2653, + "step": 108985 + }, + { + "epoch": 7.4052181002853645, + "grad_norm": 0.9249692559242249, + "learning_rate": 7.434773746432939e-05, + "loss": 3.4443, + "step": 108990 + }, + { + "epoch": 7.405557820356027, + "grad_norm": 2.793853759765625, + "learning_rate": 7.430527245549668e-05, + "loss": 3.3124, + "step": 108995 + }, + { + "epoch": 7.405897540426689, + "grad_norm": 0.8026891350746155, + "learning_rate": 7.426280744666395e-05, + "loss": 3.3444, + "step": 109000 + }, + { + "epoch": 7.40623726049735, + "grad_norm": 1.0014301538467407, + "learning_rate": 7.422034243783123e-05, + "loss": 3.3705, + "step": 109005 + }, + { + "epoch": 7.406576980568012, + "grad_norm": 9.743577003479004, + "learning_rate": 7.417787742899851e-05, + "loss": 3.3388, + "step": 109010 + }, + { + "epoch": 7.406916700638674, + "grad_norm": 1.0356522798538208, + "learning_rate": 7.413541242016579e-05, + "loss": 3.3262, + "step": 109015 + }, + { + "epoch": 7.407256420709335, + "grad_norm": 0.993087887763977, + "learning_rate": 7.409294741133306e-05, + "loss": 3.2299, + "step": 109020 + }, + { + "epoch": 7.407596140779997, + "grad_norm": 1.5235745906829834, + "learning_rate": 7.405048240250034e-05, + "loss": 3.38, + "step": 109025 + }, + { + "epoch": 7.407935860850659, + "grad_norm": 1.2068055868148804, + "learning_rate": 7.400801739366763e-05, + "loss": 3.5387, + "step": 109030 + }, + { + "epoch": 7.4082755809213205, + "grad_norm": 0.8238664865493774, + "learning_rate": 7.39655523848349e-05, + "loss": 3.262, + "step": 109035 + }, + { + "epoch": 7.408615300991983, + "grad_norm": 0.9751163125038147, + "learning_rate": 7.392308737600218e-05, + "loss": 3.1025, + "step": 109040 + }, + { + "epoch": 7.408955021062645, + "grad_norm": 0.8880235552787781, + "learning_rate": 7.388062236716946e-05, + "loss": 3.3305, + "step": 109045 + }, + { + "epoch": 7.409294741133306, + "grad_norm": 0.9892449378967285, + "learning_rate": 7.383815735833674e-05, + "loss": 3.1956, + "step": 109050 + }, + { + "epoch": 7.409634461203968, + "grad_norm": 0.9087174534797668, + "learning_rate": 7.3795692349504e-05, + "loss": 3.2998, + "step": 109055 + }, + { + "epoch": 7.40997418127463, + "grad_norm": 1.097226858139038, + "learning_rate": 7.375322734067128e-05, + "loss": 3.1743, + "step": 109060 + }, + { + "epoch": 7.410313901345291, + "grad_norm": 1.0096023082733154, + "learning_rate": 7.371076233183858e-05, + "loss": 3.4655, + "step": 109065 + }, + { + "epoch": 7.410653621415953, + "grad_norm": 0.9846969246864319, + "learning_rate": 7.366829732300584e-05, + "loss": 3.2782, + "step": 109070 + }, + { + "epoch": 7.410993341486615, + "grad_norm": 1.2822694778442383, + "learning_rate": 7.362583231417312e-05, + "loss": 3.2555, + "step": 109075 + }, + { + "epoch": 7.411333061557277, + "grad_norm": 0.8766348958015442, + "learning_rate": 7.35833673053404e-05, + "loss": 3.5165, + "step": 109080 + }, + { + "epoch": 7.411672781627939, + "grad_norm": 1.0228028297424316, + "learning_rate": 7.354090229650767e-05, + "loss": 3.2676, + "step": 109085 + }, + { + "epoch": 7.412012501698601, + "grad_norm": 1.0322678089141846, + "learning_rate": 7.349843728767495e-05, + "loss": 3.2237, + "step": 109090 + }, + { + "epoch": 7.412352221769262, + "grad_norm": 0.9516665935516357, + "learning_rate": 7.345597227884224e-05, + "loss": 3.3254, + "step": 109095 + }, + { + "epoch": 7.412691941839924, + "grad_norm": 0.8405289649963379, + "learning_rate": 7.341350727000952e-05, + "loss": 3.2449, + "step": 109100 + }, + { + "epoch": 7.413031661910586, + "grad_norm": 1.3262921571731567, + "learning_rate": 7.337104226117679e-05, + "loss": 3.4065, + "step": 109105 + }, + { + "epoch": 7.413371381981247, + "grad_norm": 1.021611213684082, + "learning_rate": 7.332857725234407e-05, + "loss": 3.3436, + "step": 109110 + }, + { + "epoch": 7.413711102051909, + "grad_norm": 0.9913219213485718, + "learning_rate": 7.328611224351135e-05, + "loss": 3.2277, + "step": 109115 + }, + { + "epoch": 7.414050822122571, + "grad_norm": 0.896756649017334, + "learning_rate": 7.324364723467862e-05, + "loss": 3.5591, + "step": 109120 + }, + { + "epoch": 7.414390542193233, + "grad_norm": 0.9391492009162903, + "learning_rate": 7.32011822258459e-05, + "loss": 3.5823, + "step": 109125 + }, + { + "epoch": 7.414730262263895, + "grad_norm": 0.7776646614074707, + "learning_rate": 7.315871721701319e-05, + "loss": 3.5321, + "step": 109130 + }, + { + "epoch": 7.415069982334557, + "grad_norm": 0.873918354511261, + "learning_rate": 7.311625220818047e-05, + "loss": 3.5976, + "step": 109135 + }, + { + "epoch": 7.415409702405218, + "grad_norm": 0.6959112882614136, + "learning_rate": 7.307378719934774e-05, + "loss": 3.3658, + "step": 109140 + }, + { + "epoch": 7.41574942247588, + "grad_norm": 1.1407252550125122, + "learning_rate": 7.303132219051502e-05, + "loss": 3.3689, + "step": 109145 + }, + { + "epoch": 7.416089142546542, + "grad_norm": 1.0464898347854614, + "learning_rate": 7.29888571816823e-05, + "loss": 2.919, + "step": 109150 + }, + { + "epoch": 7.416428862617203, + "grad_norm": 0.7282983064651489, + "learning_rate": 7.294639217284956e-05, + "loss": 3.3706, + "step": 109155 + }, + { + "epoch": 7.416768582687865, + "grad_norm": 0.7946872711181641, + "learning_rate": 7.290392716401686e-05, + "loss": 3.218, + "step": 109160 + }, + { + "epoch": 7.4171083027585265, + "grad_norm": 0.9517340660095215, + "learning_rate": 7.286146215518414e-05, + "loss": 3.32, + "step": 109165 + }, + { + "epoch": 7.417448022829189, + "grad_norm": 0.8556929230690002, + "learning_rate": 7.28189971463514e-05, + "loss": 3.3439, + "step": 109170 + }, + { + "epoch": 7.417787742899851, + "grad_norm": 0.7873017191886902, + "learning_rate": 7.277653213751868e-05, + "loss": 3.5728, + "step": 109175 + }, + { + "epoch": 7.418127462970512, + "grad_norm": 0.9487882852554321, + "learning_rate": 7.273406712868596e-05, + "loss": 3.3549, + "step": 109180 + }, + { + "epoch": 7.418467183041174, + "grad_norm": 0.8733442425727844, + "learning_rate": 7.269160211985324e-05, + "loss": 3.373, + "step": 109185 + }, + { + "epoch": 7.418806903111836, + "grad_norm": 0.8508400321006775, + "learning_rate": 7.264913711102051e-05, + "loss": 3.4433, + "step": 109190 + }, + { + "epoch": 7.419146623182497, + "grad_norm": 1.095154881477356, + "learning_rate": 7.26066721021878e-05, + "loss": 3.4764, + "step": 109195 + }, + { + "epoch": 7.419486343253159, + "grad_norm": 1.101104974746704, + "learning_rate": 7.256420709335508e-05, + "loss": 3.3057, + "step": 109200 + }, + { + "epoch": 7.419826063323821, + "grad_norm": 0.9127247929573059, + "learning_rate": 7.252174208452235e-05, + "loss": 3.3728, + "step": 109205 + }, + { + "epoch": 7.4201657833944825, + "grad_norm": 1.091130256652832, + "learning_rate": 7.247927707568963e-05, + "loss": 3.127, + "step": 109210 + }, + { + "epoch": 7.420505503465145, + "grad_norm": 0.8310306668281555, + "learning_rate": 7.243681206685691e-05, + "loss": 3.2889, + "step": 109215 + }, + { + "epoch": 7.420845223535807, + "grad_norm": 0.7557002902030945, + "learning_rate": 7.239434705802419e-05, + "loss": 3.2906, + "step": 109220 + }, + { + "epoch": 7.421184943606468, + "grad_norm": 0.9455674290657043, + "learning_rate": 7.235188204919147e-05, + "loss": 3.6486, + "step": 109225 + }, + { + "epoch": 7.42152466367713, + "grad_norm": 1.038344383239746, + "learning_rate": 7.230941704035875e-05, + "loss": 3.3907, + "step": 109230 + }, + { + "epoch": 7.421864383747792, + "grad_norm": 0.7879281044006348, + "learning_rate": 7.226695203152603e-05, + "loss": 3.4652, + "step": 109235 + }, + { + "epoch": 7.422204103818453, + "grad_norm": 0.9279722571372986, + "learning_rate": 7.22244870226933e-05, + "loss": 3.5836, + "step": 109240 + }, + { + "epoch": 7.422543823889115, + "grad_norm": 0.9697195291519165, + "learning_rate": 7.218202201386058e-05, + "loss": 3.4864, + "step": 109245 + }, + { + "epoch": 7.422883543959777, + "grad_norm": 1.0397840738296509, + "learning_rate": 7.213955700502786e-05, + "loss": 3.3327, + "step": 109250 + }, + { + "epoch": 7.4232232640304385, + "grad_norm": 0.8250311613082886, + "learning_rate": 7.209709199619512e-05, + "loss": 3.328, + "step": 109255 + }, + { + "epoch": 7.423562984101101, + "grad_norm": 1.0358904600143433, + "learning_rate": 7.205462698736242e-05, + "loss": 3.2039, + "step": 109260 + }, + { + "epoch": 7.423902704171763, + "grad_norm": 0.8110485672950745, + "learning_rate": 7.20121619785297e-05, + "loss": 3.5292, + "step": 109265 + }, + { + "epoch": 7.424242424242424, + "grad_norm": 0.7948475480079651, + "learning_rate": 7.196969696969698e-05, + "loss": 3.6926, + "step": 109270 + }, + { + "epoch": 7.424582144313086, + "grad_norm": 1.7271445989608765, + "learning_rate": 7.192723196086424e-05, + "loss": 3.5266, + "step": 109275 + }, + { + "epoch": 7.424921864383748, + "grad_norm": 0.9946459531784058, + "learning_rate": 7.188476695203152e-05, + "loss": 3.4679, + "step": 109280 + }, + { + "epoch": 7.425261584454409, + "grad_norm": 0.8152720928192139, + "learning_rate": 7.18423019431988e-05, + "loss": 3.589, + "step": 109285 + }, + { + "epoch": 7.425601304525071, + "grad_norm": 0.9734251499176025, + "learning_rate": 7.179983693436608e-05, + "loss": 3.4673, + "step": 109290 + }, + { + "epoch": 7.425941024595733, + "grad_norm": 1.0814473628997803, + "learning_rate": 7.175737192553336e-05, + "loss": 3.0955, + "step": 109295 + }, + { + "epoch": 7.4262807446663945, + "grad_norm": 0.8925777077674866, + "learning_rate": 7.171490691670064e-05, + "loss": 3.5171, + "step": 109300 + }, + { + "epoch": 7.426620464737057, + "grad_norm": 1.0907917022705078, + "learning_rate": 7.167244190786792e-05, + "loss": 3.1838, + "step": 109305 + }, + { + "epoch": 7.426960184807719, + "grad_norm": 1.172927975654602, + "learning_rate": 7.162997689903519e-05, + "loss": 3.254, + "step": 109310 + }, + { + "epoch": 7.42729990487838, + "grad_norm": 0.866870641708374, + "learning_rate": 7.158751189020247e-05, + "loss": 3.4389, + "step": 109315 + }, + { + "epoch": 7.427639624949042, + "grad_norm": 0.9416316747665405, + "learning_rate": 7.154504688136975e-05, + "loss": 3.5011, + "step": 109320 + }, + { + "epoch": 7.427979345019704, + "grad_norm": 0.9312907457351685, + "learning_rate": 7.150258187253703e-05, + "loss": 3.4294, + "step": 109325 + }, + { + "epoch": 7.428319065090365, + "grad_norm": 0.7198126912117004, + "learning_rate": 7.146011686370431e-05, + "loss": 3.3865, + "step": 109330 + }, + { + "epoch": 7.428658785161027, + "grad_norm": 0.9124802947044373, + "learning_rate": 7.141765185487159e-05, + "loss": 3.2169, + "step": 109335 + }, + { + "epoch": 7.428998505231689, + "grad_norm": 0.8203264474868774, + "learning_rate": 7.137518684603886e-05, + "loss": 3.5869, + "step": 109340 + }, + { + "epoch": 7.4293382253023506, + "grad_norm": 0.9073609113693237, + "learning_rate": 7.133272183720614e-05, + "loss": 3.3305, + "step": 109345 + }, + { + "epoch": 7.429677945373013, + "grad_norm": 0.944314181804657, + "learning_rate": 7.129025682837342e-05, + "loss": 3.5182, + "step": 109350 + }, + { + "epoch": 7.430017665443675, + "grad_norm": 0.9842053055763245, + "learning_rate": 7.124779181954071e-05, + "loss": 3.284, + "step": 109355 + }, + { + "epoch": 7.430357385514336, + "grad_norm": 0.918709397315979, + "learning_rate": 7.120532681070798e-05, + "loss": 3.4144, + "step": 109360 + }, + { + "epoch": 7.430697105584998, + "grad_norm": 1.278515100479126, + "learning_rate": 7.116286180187526e-05, + "loss": 3.6131, + "step": 109365 + }, + { + "epoch": 7.43103682565566, + "grad_norm": 0.9255961179733276, + "learning_rate": 7.112039679304254e-05, + "loss": 3.5364, + "step": 109370 + }, + { + "epoch": 7.431376545726321, + "grad_norm": 0.9755925536155701, + "learning_rate": 7.10779317842098e-05, + "loss": 3.433, + "step": 109375 + }, + { + "epoch": 7.431716265796983, + "grad_norm": 0.9686803817749023, + "learning_rate": 7.103546677537709e-05, + "loss": 3.5441, + "step": 109380 + }, + { + "epoch": 7.432055985867645, + "grad_norm": 1.0482949018478394, + "learning_rate": 7.099300176654437e-05, + "loss": 3.3705, + "step": 109385 + }, + { + "epoch": 7.432395705938307, + "grad_norm": 0.876022219657898, + "learning_rate": 7.095053675771166e-05, + "loss": 3.4434, + "step": 109390 + }, + { + "epoch": 7.432735426008969, + "grad_norm": 1.2994740009307861, + "learning_rate": 7.090807174887893e-05, + "loss": 3.3367, + "step": 109395 + }, + { + "epoch": 7.433075146079631, + "grad_norm": 1.1505136489868164, + "learning_rate": 7.08656067400462e-05, + "loss": 3.4587, + "step": 109400 + }, + { + "epoch": 7.433414866150292, + "grad_norm": 0.9328205585479736, + "learning_rate": 7.082314173121349e-05, + "loss": 3.4286, + "step": 109405 + }, + { + "epoch": 7.433754586220954, + "grad_norm": 0.8587508797645569, + "learning_rate": 7.078067672238075e-05, + "loss": 3.4289, + "step": 109410 + }, + { + "epoch": 7.434094306291616, + "grad_norm": 0.756179690361023, + "learning_rate": 7.073821171354803e-05, + "loss": 3.386, + "step": 109415 + }, + { + "epoch": 7.434434026362277, + "grad_norm": 1.032590627670288, + "learning_rate": 7.069574670471533e-05, + "loss": 3.5387, + "step": 109420 + }, + { + "epoch": 7.434773746432939, + "grad_norm": 1.1768051385879517, + "learning_rate": 7.065328169588259e-05, + "loss": 3.5423, + "step": 109425 + }, + { + "epoch": 7.435113466503601, + "grad_norm": 1.0286988019943237, + "learning_rate": 7.061081668704987e-05, + "loss": 3.2038, + "step": 109430 + }, + { + "epoch": 7.435453186574263, + "grad_norm": 0.8535590767860413, + "learning_rate": 7.056835167821715e-05, + "loss": 3.1513, + "step": 109435 + }, + { + "epoch": 7.435792906644925, + "grad_norm": 0.725544273853302, + "learning_rate": 7.052588666938443e-05, + "loss": 3.0781, + "step": 109440 + }, + { + "epoch": 7.436132626715587, + "grad_norm": 0.86250239610672, + "learning_rate": 7.04834216605517e-05, + "loss": 3.4086, + "step": 109445 + }, + { + "epoch": 7.436472346786248, + "grad_norm": 0.7648467421531677, + "learning_rate": 7.044095665171898e-05, + "loss": 3.5776, + "step": 109450 + }, + { + "epoch": 7.43681206685691, + "grad_norm": 0.9063030481338501, + "learning_rate": 7.039849164288627e-05, + "loss": 3.3784, + "step": 109455 + }, + { + "epoch": 7.437151786927572, + "grad_norm": 1.0014652013778687, + "learning_rate": 7.035602663405354e-05, + "loss": 3.1754, + "step": 109460 + }, + { + "epoch": 7.437491506998233, + "grad_norm": 1.0241363048553467, + "learning_rate": 7.031356162522082e-05, + "loss": 3.1623, + "step": 109465 + }, + { + "epoch": 7.437831227068895, + "grad_norm": 0.8997647166252136, + "learning_rate": 7.02710966163881e-05, + "loss": 3.3834, + "step": 109470 + }, + { + "epoch": 7.438170947139557, + "grad_norm": 0.9125089049339294, + "learning_rate": 7.022863160755538e-05, + "loss": 3.4018, + "step": 109475 + }, + { + "epoch": 7.438510667210219, + "grad_norm": 1.7133721113204956, + "learning_rate": 7.018616659872265e-05, + "loss": 3.6005, + "step": 109480 + }, + { + "epoch": 7.438850387280881, + "grad_norm": 0.8551002740859985, + "learning_rate": 7.014370158988993e-05, + "loss": 3.4097, + "step": 109485 + }, + { + "epoch": 7.439190107351543, + "grad_norm": 0.9649596810340881, + "learning_rate": 7.010123658105722e-05, + "loss": 3.4248, + "step": 109490 + }, + { + "epoch": 7.439529827422204, + "grad_norm": 0.9358288049697876, + "learning_rate": 7.005877157222449e-05, + "loss": 3.5513, + "step": 109495 + }, + { + "epoch": 7.439869547492866, + "grad_norm": 1.1833605766296387, + "learning_rate": 7.001630656339177e-05, + "loss": 3.2971, + "step": 109500 + }, + { + "epoch": 7.440209267563528, + "grad_norm": 1.0178806781768799, + "learning_rate": 6.997384155455905e-05, + "loss": 3.2322, + "step": 109505 + }, + { + "epoch": 7.440548987634189, + "grad_norm": 0.8572435975074768, + "learning_rate": 6.993137654572631e-05, + "loss": 3.2185, + "step": 109510 + }, + { + "epoch": 7.440888707704851, + "grad_norm": 1.0251275300979614, + "learning_rate": 6.988891153689359e-05, + "loss": 3.3901, + "step": 109515 + }, + { + "epoch": 7.441228427775513, + "grad_norm": 0.8845797181129456, + "learning_rate": 6.984644652806089e-05, + "loss": 3.2952, + "step": 109520 + }, + { + "epoch": 7.441568147846175, + "grad_norm": 0.8172441124916077, + "learning_rate": 6.980398151922817e-05, + "loss": 3.3019, + "step": 109525 + }, + { + "epoch": 7.441907867916837, + "grad_norm": 0.9353725910186768, + "learning_rate": 6.976151651039543e-05, + "loss": 3.4391, + "step": 109530 + }, + { + "epoch": 7.442247587987499, + "grad_norm": 0.8781663775444031, + "learning_rate": 6.971905150156271e-05, + "loss": 3.3819, + "step": 109535 + }, + { + "epoch": 7.44258730805816, + "grad_norm": 1.0336495637893677, + "learning_rate": 6.967658649273e-05, + "loss": 3.3218, + "step": 109540 + }, + { + "epoch": 7.442927028128822, + "grad_norm": 1.1334969997406006, + "learning_rate": 6.963412148389726e-05, + "loss": 3.2498, + "step": 109545 + }, + { + "epoch": 7.443266748199484, + "grad_norm": 1.2293639183044434, + "learning_rate": 6.959165647506454e-05, + "loss": 3.643, + "step": 109550 + }, + { + "epoch": 7.443606468270145, + "grad_norm": 0.8461442589759827, + "learning_rate": 6.954919146623183e-05, + "loss": 3.3863, + "step": 109555 + }, + { + "epoch": 7.443946188340807, + "grad_norm": 1.2041168212890625, + "learning_rate": 6.950672645739911e-05, + "loss": 3.3953, + "step": 109560 + }, + { + "epoch": 7.444285908411469, + "grad_norm": 0.8950439095497131, + "learning_rate": 6.946426144856638e-05, + "loss": 3.28, + "step": 109565 + }, + { + "epoch": 7.444625628482131, + "grad_norm": 0.9156578779220581, + "learning_rate": 6.942179643973366e-05, + "loss": 3.5885, + "step": 109570 + }, + { + "epoch": 7.444965348552793, + "grad_norm": 0.8092699646949768, + "learning_rate": 6.937933143090094e-05, + "loss": 3.3307, + "step": 109575 + }, + { + "epoch": 7.445305068623454, + "grad_norm": 1.0184001922607422, + "learning_rate": 6.933686642206821e-05, + "loss": 3.5924, + "step": 109580 + }, + { + "epoch": 7.445644788694116, + "grad_norm": 0.9650581479072571, + "learning_rate": 6.92944014132355e-05, + "loss": 3.5832, + "step": 109585 + }, + { + "epoch": 7.445984508764778, + "grad_norm": 0.8706843256950378, + "learning_rate": 6.925193640440278e-05, + "loss": 3.3137, + "step": 109590 + }, + { + "epoch": 7.446324228835439, + "grad_norm": 0.9196293354034424, + "learning_rate": 6.920947139557005e-05, + "loss": 3.2726, + "step": 109595 + }, + { + "epoch": 7.446663948906101, + "grad_norm": 1.427722692489624, + "learning_rate": 6.916700638673733e-05, + "loss": 3.3183, + "step": 109600 + }, + { + "epoch": 7.447003668976763, + "grad_norm": 1.0083094835281372, + "learning_rate": 6.912454137790461e-05, + "loss": 3.3831, + "step": 109605 + }, + { + "epoch": 7.4473433890474245, + "grad_norm": 1.0163116455078125, + "learning_rate": 6.908207636907189e-05, + "loss": 3.3127, + "step": 109610 + }, + { + "epoch": 7.447683109118087, + "grad_norm": 1.0467088222503662, + "learning_rate": 6.903961136023915e-05, + "loss": 3.3333, + "step": 109615 + }, + { + "epoch": 7.448022829188749, + "grad_norm": 0.980232834815979, + "learning_rate": 6.899714635140645e-05, + "loss": 3.582, + "step": 109620 + }, + { + "epoch": 7.44836254925941, + "grad_norm": 1.1514155864715576, + "learning_rate": 6.895468134257373e-05, + "loss": 3.3008, + "step": 109625 + }, + { + "epoch": 7.448702269330072, + "grad_norm": 0.9963608384132385, + "learning_rate": 6.8912216333741e-05, + "loss": 3.479, + "step": 109630 + }, + { + "epoch": 7.449041989400734, + "grad_norm": 1.2647864818572998, + "learning_rate": 6.886975132490827e-05, + "loss": 3.4977, + "step": 109635 + }, + { + "epoch": 7.449381709471395, + "grad_norm": 0.8244101405143738, + "learning_rate": 6.882728631607555e-05, + "loss": 3.4747, + "step": 109640 + }, + { + "epoch": 7.449721429542057, + "grad_norm": 1.155725359916687, + "learning_rate": 6.878482130724283e-05, + "loss": 3.4733, + "step": 109645 + }, + { + "epoch": 7.450061149612719, + "grad_norm": 0.9018498659133911, + "learning_rate": 6.874235629841011e-05, + "loss": 3.3252, + "step": 109650 + }, + { + "epoch": 7.4504008696833806, + "grad_norm": 1.0593016147613525, + "learning_rate": 6.86998912895774e-05, + "loss": 3.0404, + "step": 109655 + }, + { + "epoch": 7.450740589754043, + "grad_norm": 0.7719897627830505, + "learning_rate": 6.865742628074467e-05, + "loss": 3.3786, + "step": 109660 + }, + { + "epoch": 7.451080309824705, + "grad_norm": 0.8079191446304321, + "learning_rate": 6.861496127191194e-05, + "loss": 3.2604, + "step": 109665 + }, + { + "epoch": 7.451420029895366, + "grad_norm": 1.2059556245803833, + "learning_rate": 6.857249626307922e-05, + "loss": 3.4505, + "step": 109670 + }, + { + "epoch": 7.451759749966028, + "grad_norm": 0.7341389656066895, + "learning_rate": 6.85300312542465e-05, + "loss": 3.4584, + "step": 109675 + }, + { + "epoch": 7.45209947003669, + "grad_norm": 0.8052537441253662, + "learning_rate": 6.848756624541377e-05, + "loss": 3.2597, + "step": 109680 + }, + { + "epoch": 7.452439190107351, + "grad_norm": 1.1313880681991577, + "learning_rate": 6.844510123658106e-05, + "loss": 3.2831, + "step": 109685 + }, + { + "epoch": 7.452778910178013, + "grad_norm": 1.3197563886642456, + "learning_rate": 6.840263622774834e-05, + "loss": 3.4352, + "step": 109690 + }, + { + "epoch": 7.453118630248675, + "grad_norm": 0.7725565433502197, + "learning_rate": 6.836017121891562e-05, + "loss": 3.3574, + "step": 109695 + }, + { + "epoch": 7.453458350319337, + "grad_norm": 1.2184996604919434, + "learning_rate": 6.831770621008289e-05, + "loss": 3.4323, + "step": 109700 + }, + { + "epoch": 7.453798070389999, + "grad_norm": 0.9786549806594849, + "learning_rate": 6.827524120125017e-05, + "loss": 3.2893, + "step": 109705 + }, + { + "epoch": 7.454137790460661, + "grad_norm": 0.8651266694068909, + "learning_rate": 6.823277619241745e-05, + "loss": 3.2716, + "step": 109710 + }, + { + "epoch": 7.454477510531322, + "grad_norm": 1.109071969985962, + "learning_rate": 6.819031118358473e-05, + "loss": 3.2961, + "step": 109715 + }, + { + "epoch": 7.454817230601984, + "grad_norm": 0.859356701374054, + "learning_rate": 6.814784617475201e-05, + "loss": 3.342, + "step": 109720 + }, + { + "epoch": 7.455156950672646, + "grad_norm": 0.9217710494995117, + "learning_rate": 6.810538116591929e-05, + "loss": 3.4405, + "step": 109725 + }, + { + "epoch": 7.455496670743307, + "grad_norm": 0.9754170775413513, + "learning_rate": 6.806291615708657e-05, + "loss": 3.0987, + "step": 109730 + }, + { + "epoch": 7.455836390813969, + "grad_norm": 0.8167212605476379, + "learning_rate": 6.802045114825384e-05, + "loss": 3.4727, + "step": 109735 + }, + { + "epoch": 7.456176110884631, + "grad_norm": 0.9721459150314331, + "learning_rate": 6.797798613942112e-05, + "loss": 3.374, + "step": 109740 + }, + { + "epoch": 7.456515830955293, + "grad_norm": 0.8616061210632324, + "learning_rate": 6.79355211305884e-05, + "loss": 3.3096, + "step": 109745 + }, + { + "epoch": 7.456855551025955, + "grad_norm": 0.7160918712615967, + "learning_rate": 6.789305612175568e-05, + "loss": 3.3844, + "step": 109750 + }, + { + "epoch": 7.457195271096617, + "grad_norm": 1.029139757156372, + "learning_rate": 6.785059111292296e-05, + "loss": 3.4676, + "step": 109755 + }, + { + "epoch": 7.457534991167278, + "grad_norm": 0.8555437326431274, + "learning_rate": 6.780812610409024e-05, + "loss": 3.3715, + "step": 109760 + }, + { + "epoch": 7.45787471123794, + "grad_norm": 0.9523018598556519, + "learning_rate": 6.77656610952575e-05, + "loss": 3.4401, + "step": 109765 + }, + { + "epoch": 7.458214431308602, + "grad_norm": 0.966242790222168, + "learning_rate": 6.772319608642478e-05, + "loss": 3.3293, + "step": 109770 + }, + { + "epoch": 7.458554151379263, + "grad_norm": 1.0818426609039307, + "learning_rate": 6.768073107759206e-05, + "loss": 3.4204, + "step": 109775 + }, + { + "epoch": 7.458893871449925, + "grad_norm": 0.9440369009971619, + "learning_rate": 6.763826606875936e-05, + "loss": 3.1671, + "step": 109780 + }, + { + "epoch": 7.459233591520587, + "grad_norm": 1.3470290899276733, + "learning_rate": 6.759580105992662e-05, + "loss": 3.3855, + "step": 109785 + }, + { + "epoch": 7.459573311591249, + "grad_norm": 1.259413480758667, + "learning_rate": 6.75533360510939e-05, + "loss": 3.2829, + "step": 109790 + }, + { + "epoch": 7.459913031661911, + "grad_norm": 1.4947993755340576, + "learning_rate": 6.751087104226118e-05, + "loss": 3.6058, + "step": 109795 + }, + { + "epoch": 7.460252751732573, + "grad_norm": 1.0873783826828003, + "learning_rate": 6.746840603342845e-05, + "loss": 3.2982, + "step": 109800 + }, + { + "epoch": 7.460592471803234, + "grad_norm": 0.8453444242477417, + "learning_rate": 6.742594102459573e-05, + "loss": 3.5045, + "step": 109805 + }, + { + "epoch": 7.460932191873896, + "grad_norm": 0.8120226263999939, + "learning_rate": 6.738347601576301e-05, + "loss": 3.464, + "step": 109810 + }, + { + "epoch": 7.461271911944558, + "grad_norm": 1.0531060695648193, + "learning_rate": 6.73410110069303e-05, + "loss": 3.6566, + "step": 109815 + }, + { + "epoch": 7.461611632015219, + "grad_norm": 1.2124356031417847, + "learning_rate": 6.729854599809757e-05, + "loss": 3.4443, + "step": 109820 + }, + { + "epoch": 7.461951352085881, + "grad_norm": 0.8837572336196899, + "learning_rate": 6.725608098926485e-05, + "loss": 3.32, + "step": 109825 + }, + { + "epoch": 7.462291072156543, + "grad_norm": 0.8994693756103516, + "learning_rate": 6.721361598043213e-05, + "loss": 3.7608, + "step": 109830 + }, + { + "epoch": 7.462630792227205, + "grad_norm": 0.7993336319923401, + "learning_rate": 6.71711509715994e-05, + "loss": 3.3227, + "step": 109835 + }, + { + "epoch": 7.462970512297867, + "grad_norm": 1.0240929126739502, + "learning_rate": 6.712868596276668e-05, + "loss": 3.3113, + "step": 109840 + }, + { + "epoch": 7.463310232368528, + "grad_norm": 0.8717549443244934, + "learning_rate": 6.708622095393396e-05, + "loss": 3.3376, + "step": 109845 + }, + { + "epoch": 7.46364995243919, + "grad_norm": 1.8585435152053833, + "learning_rate": 6.704375594510124e-05, + "loss": 3.1827, + "step": 109850 + }, + { + "epoch": 7.463989672509852, + "grad_norm": 0.9209631681442261, + "learning_rate": 6.700129093626852e-05, + "loss": 3.1437, + "step": 109855 + }, + { + "epoch": 7.464329392580513, + "grad_norm": 1.0209424495697021, + "learning_rate": 6.69588259274358e-05, + "loss": 3.513, + "step": 109860 + }, + { + "epoch": 7.464669112651175, + "grad_norm": 0.8306548595428467, + "learning_rate": 6.691636091860308e-05, + "loss": 3.2772, + "step": 109865 + }, + { + "epoch": 7.465008832721837, + "grad_norm": 1.1270288228988647, + "learning_rate": 6.687389590977034e-05, + "loss": 2.9915, + "step": 109870 + }, + { + "epoch": 7.4653485527924985, + "grad_norm": 0.8859511017799377, + "learning_rate": 6.683143090093762e-05, + "loss": 3.6846, + "step": 109875 + }, + { + "epoch": 7.465688272863161, + "grad_norm": 1.0498641729354858, + "learning_rate": 6.678896589210492e-05, + "loss": 3.2946, + "step": 109880 + }, + { + "epoch": 7.466027992933823, + "grad_norm": 1.0783065557479858, + "learning_rate": 6.674650088327218e-05, + "loss": 3.4213, + "step": 109885 + }, + { + "epoch": 7.466367713004484, + "grad_norm": 1.1296368837356567, + "learning_rate": 6.670403587443946e-05, + "loss": 3.3604, + "step": 109890 + }, + { + "epoch": 7.466707433075146, + "grad_norm": 1.0610769987106323, + "learning_rate": 6.666157086560674e-05, + "loss": 3.6073, + "step": 109895 + }, + { + "epoch": 7.467047153145808, + "grad_norm": 0.863092303276062, + "learning_rate": 6.661910585677402e-05, + "loss": 3.493, + "step": 109900 + }, + { + "epoch": 7.467386873216469, + "grad_norm": 1.038997769355774, + "learning_rate": 6.657664084794129e-05, + "loss": 3.4336, + "step": 109905 + }, + { + "epoch": 7.467726593287131, + "grad_norm": 0.7702967524528503, + "learning_rate": 6.653417583910857e-05, + "loss": 3.1778, + "step": 109910 + }, + { + "epoch": 7.468066313357793, + "grad_norm": 0.9763985276222229, + "learning_rate": 6.649171083027586e-05, + "loss": 3.4871, + "step": 109915 + }, + { + "epoch": 7.4684060334284545, + "grad_norm": 0.9700543880462646, + "learning_rate": 6.644924582144313e-05, + "loss": 3.3641, + "step": 109920 + }, + { + "epoch": 7.468745753499117, + "grad_norm": 0.9121419191360474, + "learning_rate": 6.640678081261041e-05, + "loss": 3.291, + "step": 109925 + }, + { + "epoch": 7.469085473569779, + "grad_norm": 0.9845113754272461, + "learning_rate": 6.636431580377769e-05, + "loss": 3.3431, + "step": 109930 + }, + { + "epoch": 7.46942519364044, + "grad_norm": 1.0607548952102661, + "learning_rate": 6.632185079494496e-05, + "loss": 3.3117, + "step": 109935 + }, + { + "epoch": 7.469764913711102, + "grad_norm": 0.8916406631469727, + "learning_rate": 6.627938578611224e-05, + "loss": 3.2717, + "step": 109940 + }, + { + "epoch": 7.470104633781764, + "grad_norm": 0.978239893913269, + "learning_rate": 6.623692077727953e-05, + "loss": 3.1803, + "step": 109945 + }, + { + "epoch": 7.470444353852425, + "grad_norm": 1.0926940441131592, + "learning_rate": 6.619445576844681e-05, + "loss": 3.3716, + "step": 109950 + }, + { + "epoch": 7.470784073923087, + "grad_norm": 1.2490861415863037, + "learning_rate": 6.615199075961408e-05, + "loss": 3.1489, + "step": 109955 + }, + { + "epoch": 7.471123793993749, + "grad_norm": 1.123196005821228, + "learning_rate": 6.610952575078136e-05, + "loss": 3.5709, + "step": 109960 + }, + { + "epoch": 7.471463514064411, + "grad_norm": 1.1652201414108276, + "learning_rate": 6.606706074194864e-05, + "loss": 3.3638, + "step": 109965 + }, + { + "epoch": 7.471803234135073, + "grad_norm": 0.9513775110244751, + "learning_rate": 6.60245957331159e-05, + "loss": 3.3575, + "step": 109970 + }, + { + "epoch": 7.472142954205735, + "grad_norm": 1.206178069114685, + "learning_rate": 6.598213072428318e-05, + "loss": 3.3889, + "step": 109975 + }, + { + "epoch": 7.472482674276396, + "grad_norm": 1.2734203338623047, + "learning_rate": 6.593966571545048e-05, + "loss": 3.6237, + "step": 109980 + }, + { + "epoch": 7.472822394347058, + "grad_norm": 0.8745378851890564, + "learning_rate": 6.589720070661776e-05, + "loss": 3.6265, + "step": 109985 + }, + { + "epoch": 7.47316211441772, + "grad_norm": 1.004935622215271, + "learning_rate": 6.585473569778502e-05, + "loss": 3.1999, + "step": 109990 + }, + { + "epoch": 7.473501834488381, + "grad_norm": 0.7792134881019592, + "learning_rate": 6.58122706889523e-05, + "loss": 3.3795, + "step": 109995 + }, + { + "epoch": 7.473841554559043, + "grad_norm": 1.0985711812973022, + "learning_rate": 6.576980568011958e-05, + "loss": 3.2151, + "step": 110000 + }, + { + "epoch": 7.474181274629705, + "grad_norm": 0.8279844522476196, + "learning_rate": 6.572734067128685e-05, + "loss": 3.5036, + "step": 110005 + }, + { + "epoch": 7.474520994700367, + "grad_norm": 0.9400378465652466, + "learning_rate": 6.568487566245414e-05, + "loss": 3.4129, + "step": 110010 + }, + { + "epoch": 7.474860714771029, + "grad_norm": 0.9410541653633118, + "learning_rate": 6.564241065362142e-05, + "loss": 3.322, + "step": 110015 + }, + { + "epoch": 7.475200434841691, + "grad_norm": 1.2322354316711426, + "learning_rate": 6.559994564478869e-05, + "loss": 3.0401, + "step": 110020 + }, + { + "epoch": 7.475540154912352, + "grad_norm": 1.120863437652588, + "learning_rate": 6.555748063595597e-05, + "loss": 3.3288, + "step": 110025 + }, + { + "epoch": 7.475879874983014, + "grad_norm": 0.8781113624572754, + "learning_rate": 6.551501562712325e-05, + "loss": 3.3982, + "step": 110030 + }, + { + "epoch": 7.476219595053676, + "grad_norm": 0.8783310055732727, + "learning_rate": 6.547255061829053e-05, + "loss": 3.2538, + "step": 110035 + }, + { + "epoch": 7.476559315124337, + "grad_norm": 1.0434095859527588, + "learning_rate": 6.54300856094578e-05, + "loss": 3.1123, + "step": 110040 + }, + { + "epoch": 7.476899035194999, + "grad_norm": 1.0029726028442383, + "learning_rate": 6.538762060062509e-05, + "loss": 3.0504, + "step": 110045 + }, + { + "epoch": 7.477238755265661, + "grad_norm": 0.7957786321640015, + "learning_rate": 6.534515559179237e-05, + "loss": 3.4034, + "step": 110050 + }, + { + "epoch": 7.477578475336323, + "grad_norm": 0.8482705354690552, + "learning_rate": 6.530269058295964e-05, + "loss": 3.3606, + "step": 110055 + }, + { + "epoch": 7.477918195406985, + "grad_norm": 0.9182798862457275, + "learning_rate": 6.526022557412692e-05, + "loss": 3.1333, + "step": 110060 + }, + { + "epoch": 7.478257915477647, + "grad_norm": 0.769528329372406, + "learning_rate": 6.52177605652942e-05, + "loss": 3.6173, + "step": 110065 + }, + { + "epoch": 7.478597635548308, + "grad_norm": 1.0616117715835571, + "learning_rate": 6.517529555646148e-05, + "loss": 3.2168, + "step": 110070 + }, + { + "epoch": 7.47893735561897, + "grad_norm": 0.8723009824752808, + "learning_rate": 6.513283054762876e-05, + "loss": 3.4909, + "step": 110075 + }, + { + "epoch": 7.479277075689632, + "grad_norm": 1.0169991254806519, + "learning_rate": 6.509036553879604e-05, + "loss": 3.3889, + "step": 110080 + }, + { + "epoch": 7.479616795760293, + "grad_norm": 0.9382302761077881, + "learning_rate": 6.504790052996332e-05, + "loss": 3.2369, + "step": 110085 + }, + { + "epoch": 7.479956515830955, + "grad_norm": 0.8897489309310913, + "learning_rate": 6.500543552113059e-05, + "loss": 3.2164, + "step": 110090 + }, + { + "epoch": 7.480296235901617, + "grad_norm": 0.9164875745773315, + "learning_rate": 6.496297051229787e-05, + "loss": 3.5752, + "step": 110095 + }, + { + "epoch": 7.480635955972279, + "grad_norm": 1.2019174098968506, + "learning_rate": 6.492050550346515e-05, + "loss": 3.2788, + "step": 110100 + }, + { + "epoch": 7.480975676042941, + "grad_norm": 0.9698094129562378, + "learning_rate": 6.487804049463241e-05, + "loss": 3.2482, + "step": 110105 + }, + { + "epoch": 7.481315396113603, + "grad_norm": 0.9467402100563049, + "learning_rate": 6.48355754857997e-05, + "loss": 3.4055, + "step": 110110 + }, + { + "epoch": 7.481655116184264, + "grad_norm": 0.8264342546463013, + "learning_rate": 6.479311047696699e-05, + "loss": 3.4822, + "step": 110115 + }, + { + "epoch": 7.481994836254926, + "grad_norm": 1.130481481552124, + "learning_rate": 6.475064546813427e-05, + "loss": 3.4897, + "step": 110120 + }, + { + "epoch": 7.482334556325588, + "grad_norm": 0.8664780259132385, + "learning_rate": 6.470818045930153e-05, + "loss": 3.2281, + "step": 110125 + }, + { + "epoch": 7.482674276396249, + "grad_norm": 1.0685182809829712, + "learning_rate": 6.466571545046881e-05, + "loss": 3.3116, + "step": 110130 + }, + { + "epoch": 7.483013996466911, + "grad_norm": 0.8888327479362488, + "learning_rate": 6.462325044163609e-05, + "loss": 3.3106, + "step": 110135 + }, + { + "epoch": 7.483353716537573, + "grad_norm": 0.9667490720748901, + "learning_rate": 6.458078543280336e-05, + "loss": 3.379, + "step": 110140 + }, + { + "epoch": 7.483693436608235, + "grad_norm": 0.7528200149536133, + "learning_rate": 6.453832042397065e-05, + "loss": 3.3795, + "step": 110145 + }, + { + "epoch": 7.484033156678897, + "grad_norm": 0.7744521498680115, + "learning_rate": 6.449585541513793e-05, + "loss": 3.4366, + "step": 110150 + }, + { + "epoch": 7.484372876749559, + "grad_norm": 0.8301141262054443, + "learning_rate": 6.445339040630521e-05, + "loss": 3.4008, + "step": 110155 + }, + { + "epoch": 7.48471259682022, + "grad_norm": 0.8483744859695435, + "learning_rate": 6.441092539747248e-05, + "loss": 3.2094, + "step": 110160 + }, + { + "epoch": 7.485052316890882, + "grad_norm": 1.0173803567886353, + "learning_rate": 6.436846038863976e-05, + "loss": 3.3398, + "step": 110165 + }, + { + "epoch": 7.485392036961544, + "grad_norm": 1.0134838819503784, + "learning_rate": 6.432599537980704e-05, + "loss": 3.3892, + "step": 110170 + }, + { + "epoch": 7.485731757032205, + "grad_norm": 0.9155483245849609, + "learning_rate": 6.428353037097432e-05, + "loss": 3.5328, + "step": 110175 + }, + { + "epoch": 7.486071477102867, + "grad_norm": 1.059468150138855, + "learning_rate": 6.42410653621416e-05, + "loss": 3.3379, + "step": 110180 + }, + { + "epoch": 7.486411197173529, + "grad_norm": 0.8377454280853271, + "learning_rate": 6.419860035330888e-05, + "loss": 3.3353, + "step": 110185 + }, + { + "epoch": 7.486750917244191, + "grad_norm": 0.9867743253707886, + "learning_rate": 6.415613534447615e-05, + "loss": 3.4663, + "step": 110190 + }, + { + "epoch": 7.487090637314853, + "grad_norm": 0.7696905732154846, + "learning_rate": 6.411367033564343e-05, + "loss": 3.3867, + "step": 110195 + }, + { + "epoch": 7.487430357385515, + "grad_norm": 1.213222861289978, + "learning_rate": 6.40712053268107e-05, + "loss": 3.0941, + "step": 110200 + }, + { + "epoch": 7.487770077456176, + "grad_norm": 0.9977549314498901, + "learning_rate": 6.4028740317978e-05, + "loss": 3.4353, + "step": 110205 + }, + { + "epoch": 7.488109797526838, + "grad_norm": 1.2338696718215942, + "learning_rate": 6.398627530914527e-05, + "loss": 3.5673, + "step": 110210 + }, + { + "epoch": 7.4884495175975, + "grad_norm": 0.8429946303367615, + "learning_rate": 6.394381030031255e-05, + "loss": 3.6174, + "step": 110215 + }, + { + "epoch": 7.488789237668161, + "grad_norm": 1.0333470106124878, + "learning_rate": 6.390134529147983e-05, + "loss": 3.0664, + "step": 110220 + }, + { + "epoch": 7.489128957738823, + "grad_norm": 1.0017236471176147, + "learning_rate": 6.385888028264709e-05, + "loss": 3.4479, + "step": 110225 + }, + { + "epoch": 7.489468677809485, + "grad_norm": 0.937288224697113, + "learning_rate": 6.381641527381437e-05, + "loss": 3.4217, + "step": 110230 + }, + { + "epoch": 7.489808397880147, + "grad_norm": 1.70301353931427, + "learning_rate": 6.377395026498165e-05, + "loss": 3.4344, + "step": 110235 + }, + { + "epoch": 7.490148117950809, + "grad_norm": 0.928008496761322, + "learning_rate": 6.373148525614895e-05, + "loss": 3.4273, + "step": 110240 + }, + { + "epoch": 7.490487838021471, + "grad_norm": 1.0675901174545288, + "learning_rate": 6.368902024731621e-05, + "loss": 3.239, + "step": 110245 + }, + { + "epoch": 7.490827558092132, + "grad_norm": 0.9522601366043091, + "learning_rate": 6.36465552384835e-05, + "loss": 3.3622, + "step": 110250 + }, + { + "epoch": 7.491167278162794, + "grad_norm": 1.8949909210205078, + "learning_rate": 6.360409022965077e-05, + "loss": 3.3874, + "step": 110255 + }, + { + "epoch": 7.491506998233455, + "grad_norm": 0.8863766193389893, + "learning_rate": 6.356162522081804e-05, + "loss": 3.4353, + "step": 110260 + }, + { + "epoch": 7.491846718304117, + "grad_norm": 1.0836173295974731, + "learning_rate": 6.351916021198532e-05, + "loss": 3.5273, + "step": 110265 + }, + { + "epoch": 7.492186438374779, + "grad_norm": 0.9811696410179138, + "learning_rate": 6.34766952031526e-05, + "loss": 3.4572, + "step": 110270 + }, + { + "epoch": 7.492526158445441, + "grad_norm": 1.3892436027526855, + "learning_rate": 6.343423019431988e-05, + "loss": 3.4045, + "step": 110275 + }, + { + "epoch": 7.492865878516103, + "grad_norm": 1.1974519491195679, + "learning_rate": 6.339176518548716e-05, + "loss": 3.268, + "step": 110280 + }, + { + "epoch": 7.493205598586765, + "grad_norm": 0.9281141757965088, + "learning_rate": 6.334930017665444e-05, + "loss": 3.4522, + "step": 110285 + }, + { + "epoch": 7.493545318657426, + "grad_norm": 0.92299485206604, + "learning_rate": 6.330683516782172e-05, + "loss": 3.1524, + "step": 110290 + }, + { + "epoch": 7.493885038728088, + "grad_norm": 0.8222960233688354, + "learning_rate": 6.326437015898899e-05, + "loss": 3.3939, + "step": 110295 + }, + { + "epoch": 7.49422475879875, + "grad_norm": 1.0637671947479248, + "learning_rate": 6.322190515015627e-05, + "loss": 3.3778, + "step": 110300 + }, + { + "epoch": 7.494564478869411, + "grad_norm": 0.8994595408439636, + "learning_rate": 6.317944014132356e-05, + "loss": 3.2872, + "step": 110305 + }, + { + "epoch": 7.494904198940073, + "grad_norm": 0.8412138223648071, + "learning_rate": 6.313697513249083e-05, + "loss": 3.2762, + "step": 110310 + }, + { + "epoch": 7.495243919010735, + "grad_norm": 1.230734944343567, + "learning_rate": 6.309451012365811e-05, + "loss": 3.685, + "step": 110315 + }, + { + "epoch": 7.495583639081397, + "grad_norm": 0.8197237849235535, + "learning_rate": 6.305204511482539e-05, + "loss": 3.4371, + "step": 110320 + }, + { + "epoch": 7.495923359152059, + "grad_norm": 0.7397778630256653, + "learning_rate": 6.300958010599267e-05, + "loss": 3.2114, + "step": 110325 + }, + { + "epoch": 7.496263079222721, + "grad_norm": 0.9007188081741333, + "learning_rate": 6.296711509715993e-05, + "loss": 3.2351, + "step": 110330 + }, + { + "epoch": 7.496602799293382, + "grad_norm": 0.9673104286193848, + "learning_rate": 6.292465008832721e-05, + "loss": 3.4435, + "step": 110335 + }, + { + "epoch": 7.496942519364044, + "grad_norm": 0.9388792514801025, + "learning_rate": 6.288218507949451e-05, + "loss": 3.311, + "step": 110340 + }, + { + "epoch": 7.497282239434706, + "grad_norm": 1.1008421182632446, + "learning_rate": 6.283972007066177e-05, + "loss": 3.4385, + "step": 110345 + }, + { + "epoch": 7.497621959505367, + "grad_norm": 0.740371584892273, + "learning_rate": 6.279725506182905e-05, + "loss": 3.5186, + "step": 110350 + }, + { + "epoch": 7.497961679576029, + "grad_norm": 0.9865629076957703, + "learning_rate": 6.275479005299633e-05, + "loss": 3.5576, + "step": 110355 + }, + { + "epoch": 7.498301399646691, + "grad_norm": 0.90883868932724, + "learning_rate": 6.27123250441636e-05, + "loss": 3.5518, + "step": 110360 + }, + { + "epoch": 7.498641119717353, + "grad_norm": 0.8720169067382812, + "learning_rate": 6.266986003533088e-05, + "loss": 3.3595, + "step": 110365 + }, + { + "epoch": 7.498980839788015, + "grad_norm": 0.9422600865364075, + "learning_rate": 6.262739502649817e-05, + "loss": 3.4448, + "step": 110370 + }, + { + "epoch": 7.499320559858677, + "grad_norm": 1.0052266120910645, + "learning_rate": 6.258493001766545e-05, + "loss": 3.33, + "step": 110375 + }, + { + "epoch": 7.499660279929338, + "grad_norm": 0.7994112372398376, + "learning_rate": 6.254246500883272e-05, + "loss": 3.5647, + "step": 110380 + }, + { + "epoch": 7.5, + "grad_norm": 1.0158820152282715, + "learning_rate": 6.25e-05, + "loss": 3.7862, + "step": 110385 + }, + { + "epoch": 7.500339720070662, + "grad_norm": 0.7447687983512878, + "learning_rate": 6.245753499116728e-05, + "loss": 3.4067, + "step": 110390 + }, + { + "epoch": 7.500679440141323, + "grad_norm": 0.8959037661552429, + "learning_rate": 6.241506998233456e-05, + "loss": 3.4112, + "step": 110395 + }, + { + "epoch": 7.501019160211985, + "grad_norm": 0.9309865236282349, + "learning_rate": 6.237260497350183e-05, + "loss": 3.506, + "step": 110400 + }, + { + "epoch": 7.501358880282647, + "grad_norm": 1.0196192264556885, + "learning_rate": 6.233013996466912e-05, + "loss": 3.2239, + "step": 110405 + }, + { + "epoch": 7.501698600353309, + "grad_norm": 0.8816798329353333, + "learning_rate": 6.228767495583639e-05, + "loss": 3.0331, + "step": 110410 + }, + { + "epoch": 7.502038320423971, + "grad_norm": 1.0327301025390625, + "learning_rate": 6.224520994700367e-05, + "loss": 3.1314, + "step": 110415 + }, + { + "epoch": 7.502378040494633, + "grad_norm": 0.9996874332427979, + "learning_rate": 6.220274493817095e-05, + "loss": 3.516, + "step": 110420 + }, + { + "epoch": 7.502717760565294, + "grad_norm": 0.8745957612991333, + "learning_rate": 6.216027992933823e-05, + "loss": 3.4623, + "step": 110425 + }, + { + "epoch": 7.503057480635956, + "grad_norm": 0.8808220624923706, + "learning_rate": 6.211781492050551e-05, + "loss": 3.0748, + "step": 110430 + }, + { + "epoch": 7.503397200706618, + "grad_norm": 0.9667528867721558, + "learning_rate": 6.207534991167279e-05, + "loss": 3.694, + "step": 110435 + }, + { + "epoch": 7.503736920777279, + "grad_norm": 0.8834923505783081, + "learning_rate": 6.203288490284007e-05, + "loss": 3.1377, + "step": 110440 + }, + { + "epoch": 7.504076640847941, + "grad_norm": 0.804199755191803, + "learning_rate": 6.199041989400734e-05, + "loss": 3.5141, + "step": 110445 + }, + { + "epoch": 7.504416360918603, + "grad_norm": 0.9727156758308411, + "learning_rate": 6.194795488517462e-05, + "loss": 3.4728, + "step": 110450 + }, + { + "epoch": 7.504756080989265, + "grad_norm": 0.7573269009590149, + "learning_rate": 6.19054898763419e-05, + "loss": 3.394, + "step": 110455 + }, + { + "epoch": 7.505095801059927, + "grad_norm": 0.9327318668365479, + "learning_rate": 6.186302486750918e-05, + "loss": 2.9099, + "step": 110460 + }, + { + "epoch": 7.505435521130589, + "grad_norm": 1.1397061347961426, + "learning_rate": 6.182055985867646e-05, + "loss": 3.3328, + "step": 110465 + }, + { + "epoch": 7.50577524120125, + "grad_norm": 0.7765927314758301, + "learning_rate": 6.177809484984374e-05, + "loss": 3.1524, + "step": 110470 + }, + { + "epoch": 7.506114961271912, + "grad_norm": 1.0021628141403198, + "learning_rate": 6.1735629841011e-05, + "loss": 3.5066, + "step": 110475 + }, + { + "epoch": 7.506454681342574, + "grad_norm": 0.8079969882965088, + "learning_rate": 6.169316483217828e-05, + "loss": 3.4588, + "step": 110480 + }, + { + "epoch": 7.506794401413235, + "grad_norm": 1.05609130859375, + "learning_rate": 6.165069982334556e-05, + "loss": 3.4456, + "step": 110485 + }, + { + "epoch": 7.507134121483897, + "grad_norm": 0.9366332292556763, + "learning_rate": 6.160823481451284e-05, + "loss": 3.2353, + "step": 110490 + }, + { + "epoch": 7.5074738415545585, + "grad_norm": 0.7777292728424072, + "learning_rate": 6.156576980568012e-05, + "loss": 3.1774, + "step": 110495 + }, + { + "epoch": 7.507813561625221, + "grad_norm": 0.8372637629508972, + "learning_rate": 6.15233047968474e-05, + "loss": 3.2536, + "step": 110500 + }, + { + "epoch": 7.508153281695883, + "grad_norm": 1.0268827676773071, + "learning_rate": 6.148083978801468e-05, + "loss": 3.5246, + "step": 110505 + }, + { + "epoch": 7.508493001766544, + "grad_norm": 0.8031108379364014, + "learning_rate": 6.143837477918195e-05, + "loss": 3.4491, + "step": 110510 + }, + { + "epoch": 7.508832721837206, + "grad_norm": 1.009747862815857, + "learning_rate": 6.139590977034923e-05, + "loss": 3.3208, + "step": 110515 + }, + { + "epoch": 7.509172441907868, + "grad_norm": 1.0752285718917847, + "learning_rate": 6.135344476151651e-05, + "loss": 3.204, + "step": 110520 + }, + { + "epoch": 7.509512161978529, + "grad_norm": 0.9887376427650452, + "learning_rate": 6.131097975268379e-05, + "loss": 3.2481, + "step": 110525 + }, + { + "epoch": 7.509851882049191, + "grad_norm": 1.0036227703094482, + "learning_rate": 6.126851474385107e-05, + "loss": 3.4157, + "step": 110530 + }, + { + "epoch": 7.510191602119853, + "grad_norm": 0.8518768548965454, + "learning_rate": 6.122604973501835e-05, + "loss": 3.2571, + "step": 110535 + }, + { + "epoch": 7.510531322190515, + "grad_norm": 0.7758731842041016, + "learning_rate": 6.118358472618563e-05, + "loss": 3.5678, + "step": 110540 + }, + { + "epoch": 7.510871042261177, + "grad_norm": 1.1809183359146118, + "learning_rate": 6.11411197173529e-05, + "loss": 3.3915, + "step": 110545 + }, + { + "epoch": 7.511210762331839, + "grad_norm": 1.0335195064544678, + "learning_rate": 6.109865470852019e-05, + "loss": 3.4576, + "step": 110550 + }, + { + "epoch": 7.5115504824025, + "grad_norm": 0.9356069564819336, + "learning_rate": 6.105618969968746e-05, + "loss": 3.2295, + "step": 110555 + }, + { + "epoch": 7.511890202473162, + "grad_norm": 0.8270926475524902, + "learning_rate": 6.1013724690854736e-05, + "loss": 3.5674, + "step": 110560 + }, + { + "epoch": 7.512229922543824, + "grad_norm": 1.1222186088562012, + "learning_rate": 6.0971259682022016e-05, + "loss": 3.5977, + "step": 110565 + }, + { + "epoch": 7.512569642614485, + "grad_norm": 0.8027486205101013, + "learning_rate": 6.092879467318929e-05, + "loss": 3.2549, + "step": 110570 + }, + { + "epoch": 7.512909362685147, + "grad_norm": 1.1956684589385986, + "learning_rate": 6.0886329664356576e-05, + "loss": 3.5008, + "step": 110575 + }, + { + "epoch": 7.513249082755809, + "grad_norm": 0.7072540521621704, + "learning_rate": 6.084386465552385e-05, + "loss": 3.7228, + "step": 110580 + }, + { + "epoch": 7.513588802826471, + "grad_norm": 1.0129390954971313, + "learning_rate": 6.080139964669112e-05, + "loss": 3.4964, + "step": 110585 + }, + { + "epoch": 7.513928522897133, + "grad_norm": 1.0159703493118286, + "learning_rate": 6.075893463785841e-05, + "loss": 3.4283, + "step": 110590 + }, + { + "epoch": 7.514268242967795, + "grad_norm": 0.9266612529754639, + "learning_rate": 6.071646962902568e-05, + "loss": 3.1992, + "step": 110595 + }, + { + "epoch": 7.514607963038456, + "grad_norm": 0.7949339747428894, + "learning_rate": 6.067400462019296e-05, + "loss": 3.4798, + "step": 110600 + }, + { + "epoch": 7.514947683109118, + "grad_norm": 0.9583854675292969, + "learning_rate": 6.0631539611360237e-05, + "loss": 3.306, + "step": 110605 + }, + { + "epoch": 7.51528740317978, + "grad_norm": 1.1017957925796509, + "learning_rate": 6.0589074602527523e-05, + "loss": 3.4751, + "step": 110610 + }, + { + "epoch": 7.515627123250441, + "grad_norm": 0.9266512393951416, + "learning_rate": 6.05466095936948e-05, + "loss": 3.267, + "step": 110615 + }, + { + "epoch": 7.515966843321103, + "grad_norm": 0.899669885635376, + "learning_rate": 6.050414458486207e-05, + "loss": 3.4402, + "step": 110620 + }, + { + "epoch": 7.516306563391765, + "grad_norm": 0.7971221804618835, + "learning_rate": 6.046167957602936e-05, + "loss": 3.2269, + "step": 110625 + }, + { + "epoch": 7.516646283462427, + "grad_norm": 0.782314121723175, + "learning_rate": 6.041921456719663e-05, + "loss": 3.3896, + "step": 110630 + }, + { + "epoch": 7.516986003533089, + "grad_norm": 0.8437812328338623, + "learning_rate": 6.037674955836391e-05, + "loss": 3.2425, + "step": 110635 + }, + { + "epoch": 7.517325723603751, + "grad_norm": 0.9737811088562012, + "learning_rate": 6.033428454953119e-05, + "loss": 3.2822, + "step": 110640 + }, + { + "epoch": 7.517665443674412, + "grad_norm": 1.18149995803833, + "learning_rate": 6.0291819540698464e-05, + "loss": 3.3138, + "step": 110645 + }, + { + "epoch": 7.518005163745074, + "grad_norm": 1.1644518375396729, + "learning_rate": 6.0249354531865744e-05, + "loss": 3.4322, + "step": 110650 + }, + { + "epoch": 7.518344883815736, + "grad_norm": 1.5118253231048584, + "learning_rate": 6.0206889523033024e-05, + "loss": 3.4106, + "step": 110655 + }, + { + "epoch": 7.518684603886397, + "grad_norm": 0.7145471572875977, + "learning_rate": 6.0164424514200304e-05, + "loss": 3.246, + "step": 110660 + }, + { + "epoch": 7.519024323957059, + "grad_norm": 0.9976124167442322, + "learning_rate": 6.012195950536758e-05, + "loss": 3.4608, + "step": 110665 + }, + { + "epoch": 7.519364044027721, + "grad_norm": 0.8302885890007019, + "learning_rate": 6.007949449653485e-05, + "loss": 3.4203, + "step": 110670 + }, + { + "epoch": 7.519703764098383, + "grad_norm": 0.8916659355163574, + "learning_rate": 6.003702948770214e-05, + "loss": 3.1223, + "step": 110675 + }, + { + "epoch": 7.520043484169045, + "grad_norm": 0.8793354034423828, + "learning_rate": 5.999456447886941e-05, + "loss": 3.5354, + "step": 110680 + }, + { + "epoch": 7.520383204239707, + "grad_norm": 0.8647778034210205, + "learning_rate": 5.995209947003669e-05, + "loss": 3.3961, + "step": 110685 + }, + { + "epoch": 7.520722924310368, + "grad_norm": 0.8541690111160278, + "learning_rate": 5.990963446120397e-05, + "loss": 3.3699, + "step": 110690 + }, + { + "epoch": 7.52106264438103, + "grad_norm": 0.9605346918106079, + "learning_rate": 5.986716945237125e-05, + "loss": 3.3941, + "step": 110695 + }, + { + "epoch": 7.521402364451692, + "grad_norm": 0.8266472816467285, + "learning_rate": 5.9824704443538524e-05, + "loss": 3.3233, + "step": 110700 + }, + { + "epoch": 7.521742084522353, + "grad_norm": 0.9988973140716553, + "learning_rate": 5.9782239434705804e-05, + "loss": 3.2044, + "step": 110705 + }, + { + "epoch": 7.522081804593015, + "grad_norm": 0.8522038459777832, + "learning_rate": 5.9739774425873084e-05, + "loss": 3.8336, + "step": 110710 + }, + { + "epoch": 7.522421524663677, + "grad_norm": 1.4595845937728882, + "learning_rate": 5.969730941704036e-05, + "loss": 3.681, + "step": 110715 + }, + { + "epoch": 7.522761244734339, + "grad_norm": 0.927713930606842, + "learning_rate": 5.965484440820764e-05, + "loss": 3.3351, + "step": 110720 + }, + { + "epoch": 7.523100964805001, + "grad_norm": 1.1836175918579102, + "learning_rate": 5.961237939937492e-05, + "loss": 3.4249, + "step": 110725 + }, + { + "epoch": 7.523440684875663, + "grad_norm": 1.0297906398773193, + "learning_rate": 5.956991439054219e-05, + "loss": 3.6606, + "step": 110730 + }, + { + "epoch": 7.523780404946324, + "grad_norm": 0.9714031219482422, + "learning_rate": 5.952744938170947e-05, + "loss": 3.583, + "step": 110735 + }, + { + "epoch": 7.524120125016986, + "grad_norm": 1.016584038734436, + "learning_rate": 5.948498437287675e-05, + "loss": 3.3547, + "step": 110740 + }, + { + "epoch": 7.524459845087648, + "grad_norm": 0.8429428339004517, + "learning_rate": 5.944251936404403e-05, + "loss": 3.2729, + "step": 110745 + }, + { + "epoch": 7.524799565158309, + "grad_norm": 1.0250431299209595, + "learning_rate": 5.9400054355211305e-05, + "loss": 3.2669, + "step": 110750 + }, + { + "epoch": 7.525139285228971, + "grad_norm": 0.9043570756912231, + "learning_rate": 5.9357589346378585e-05, + "loss": 3.4406, + "step": 110755 + }, + { + "epoch": 7.525479005299633, + "grad_norm": 1.0241528749465942, + "learning_rate": 5.9315124337545865e-05, + "loss": 3.2787, + "step": 110760 + }, + { + "epoch": 7.525818725370295, + "grad_norm": 0.7843253016471863, + "learning_rate": 5.927265932871314e-05, + "loss": 3.4287, + "step": 110765 + }, + { + "epoch": 7.526158445440957, + "grad_norm": 0.9194327592849731, + "learning_rate": 5.9230194319880425e-05, + "loss": 3.3078, + "step": 110770 + }, + { + "epoch": 7.526498165511619, + "grad_norm": 0.901494562625885, + "learning_rate": 5.91877293110477e-05, + "loss": 3.5899, + "step": 110775 + }, + { + "epoch": 7.52683788558228, + "grad_norm": 1.326223611831665, + "learning_rate": 5.914526430221498e-05, + "loss": 3.409, + "step": 110780 + }, + { + "epoch": 7.527177605652942, + "grad_norm": 1.0278558731079102, + "learning_rate": 5.910279929338225e-05, + "loss": 3.3788, + "step": 110785 + }, + { + "epoch": 7.527517325723604, + "grad_norm": 1.0182487964630127, + "learning_rate": 5.906033428454953e-05, + "loss": 3.2658, + "step": 110790 + }, + { + "epoch": 7.527857045794265, + "grad_norm": 0.7487565875053406, + "learning_rate": 5.901786927571681e-05, + "loss": 3.5164, + "step": 110795 + }, + { + "epoch": 7.528196765864927, + "grad_norm": 0.7943599820137024, + "learning_rate": 5.8975404266884085e-05, + "loss": 3.5736, + "step": 110800 + }, + { + "epoch": 7.528536485935589, + "grad_norm": 0.8003950119018555, + "learning_rate": 5.893293925805137e-05, + "loss": 3.578, + "step": 110805 + }, + { + "epoch": 7.528876206006251, + "grad_norm": 0.7713513970375061, + "learning_rate": 5.8890474249218645e-05, + "loss": 3.2671, + "step": 110810 + }, + { + "epoch": 7.529215926076913, + "grad_norm": 1.1556978225708008, + "learning_rate": 5.884800924038592e-05, + "loss": 3.2835, + "step": 110815 + }, + { + "epoch": 7.529555646147575, + "grad_norm": 1.0342425107955933, + "learning_rate": 5.8805544231553205e-05, + "loss": 3.3264, + "step": 110820 + }, + { + "epoch": 7.529895366218236, + "grad_norm": 1.0844454765319824, + "learning_rate": 5.876307922272048e-05, + "loss": 3.2014, + "step": 110825 + }, + { + "epoch": 7.530235086288898, + "grad_norm": 0.9166903495788574, + "learning_rate": 5.872061421388776e-05, + "loss": 3.2511, + "step": 110830 + }, + { + "epoch": 7.53057480635956, + "grad_norm": 0.8426485061645508, + "learning_rate": 5.867814920505504e-05, + "loss": 3.3816, + "step": 110835 + }, + { + "epoch": 7.530914526430221, + "grad_norm": 0.8101131319999695, + "learning_rate": 5.863568419622231e-05, + "loss": 3.1572, + "step": 110840 + }, + { + "epoch": 7.531254246500883, + "grad_norm": 1.164159893989563, + "learning_rate": 5.859321918738959e-05, + "loss": 3.3974, + "step": 110845 + }, + { + "epoch": 7.5315939665715455, + "grad_norm": 1.116913080215454, + "learning_rate": 5.8550754178556866e-05, + "loss": 3.5965, + "step": 110850 + }, + { + "epoch": 7.531933686642207, + "grad_norm": 0.8997089266777039, + "learning_rate": 5.850828916972415e-05, + "loss": 3.2523, + "step": 110855 + }, + { + "epoch": 7.532273406712869, + "grad_norm": 1.3026572465896606, + "learning_rate": 5.8465824160891426e-05, + "loss": 3.287, + "step": 110860 + }, + { + "epoch": 7.532613126783531, + "grad_norm": 0.8404808044433594, + "learning_rate": 5.8423359152058706e-05, + "loss": 3.3439, + "step": 110865 + }, + { + "epoch": 7.532952846854192, + "grad_norm": 1.0361379384994507, + "learning_rate": 5.8380894143225986e-05, + "loss": 3.2802, + "step": 110870 + }, + { + "epoch": 7.533292566924854, + "grad_norm": 0.7402912378311157, + "learning_rate": 5.833842913439326e-05, + "loss": 3.431, + "step": 110875 + }, + { + "epoch": 7.533632286995516, + "grad_norm": 0.8056209087371826, + "learning_rate": 5.829596412556054e-05, + "loss": 3.3012, + "step": 110880 + }, + { + "epoch": 7.533972007066177, + "grad_norm": 0.9108080863952637, + "learning_rate": 5.825349911672782e-05, + "loss": 3.4773, + "step": 110885 + }, + { + "epoch": 7.534311727136839, + "grad_norm": 0.8450350165367126, + "learning_rate": 5.82110341078951e-05, + "loss": 3.5516, + "step": 110890 + }, + { + "epoch": 7.5346514472075015, + "grad_norm": 0.9609054923057556, + "learning_rate": 5.816856909906237e-05, + "loss": 3.7756, + "step": 110895 + }, + { + "epoch": 7.534991167278163, + "grad_norm": 0.8301565647125244, + "learning_rate": 5.8126104090229646e-05, + "loss": 3.1732, + "step": 110900 + }, + { + "epoch": 7.535330887348825, + "grad_norm": 0.8348388671875, + "learning_rate": 5.808363908139693e-05, + "loss": 3.3399, + "step": 110905 + }, + { + "epoch": 7.535670607419487, + "grad_norm": 1.1150299310684204, + "learning_rate": 5.8041174072564206e-05, + "loss": 3.5783, + "step": 110910 + }, + { + "epoch": 7.536010327490148, + "grad_norm": 0.8724519610404968, + "learning_rate": 5.7998709063731486e-05, + "loss": 3.5743, + "step": 110915 + }, + { + "epoch": 7.53635004756081, + "grad_norm": 0.9027546644210815, + "learning_rate": 5.7956244054898766e-05, + "loss": 3.3876, + "step": 110920 + }, + { + "epoch": 7.536689767631472, + "grad_norm": 0.9550749063491821, + "learning_rate": 5.791377904606604e-05, + "loss": 3.1839, + "step": 110925 + }, + { + "epoch": 7.537029487702133, + "grad_norm": 0.9734463691711426, + "learning_rate": 5.787131403723332e-05, + "loss": 3.3843, + "step": 110930 + }, + { + "epoch": 7.537369207772795, + "grad_norm": 0.7345141172409058, + "learning_rate": 5.78288490284006e-05, + "loss": 3.5637, + "step": 110935 + }, + { + "epoch": 7.5377089278434575, + "grad_norm": 0.8129519820213318, + "learning_rate": 5.778638401956788e-05, + "loss": 3.7011, + "step": 110940 + }, + { + "epoch": 7.538048647914119, + "grad_norm": 0.9844523668289185, + "learning_rate": 5.774391901073515e-05, + "loss": 3.641, + "step": 110945 + }, + { + "epoch": 7.538388367984781, + "grad_norm": 0.9990353584289551, + "learning_rate": 5.770145400190244e-05, + "loss": 3.3164, + "step": 110950 + }, + { + "epoch": 7.538728088055443, + "grad_norm": 1.2633637189865112, + "learning_rate": 5.765898899306971e-05, + "loss": 3.5162, + "step": 110955 + }, + { + "epoch": 7.539067808126104, + "grad_norm": 0.8242394924163818, + "learning_rate": 5.7616523984236987e-05, + "loss": 3.2188, + "step": 110960 + }, + { + "epoch": 7.539407528196766, + "grad_norm": 0.8457367420196533, + "learning_rate": 5.757405897540427e-05, + "loss": 3.3692, + "step": 110965 + }, + { + "epoch": 7.539747248267427, + "grad_norm": 0.8837199807167053, + "learning_rate": 5.753159396657155e-05, + "loss": 3.0778, + "step": 110970 + }, + { + "epoch": 7.540086968338089, + "grad_norm": 0.8514642119407654, + "learning_rate": 5.748912895773883e-05, + "loss": 3.4046, + "step": 110975 + }, + { + "epoch": 7.540426688408751, + "grad_norm": 1.052740454673767, + "learning_rate": 5.74466639489061e-05, + "loss": 3.1915, + "step": 110980 + }, + { + "epoch": 7.540766408479413, + "grad_norm": 0.7095946073532104, + "learning_rate": 5.740419894007338e-05, + "loss": 3.3612, + "step": 110985 + }, + { + "epoch": 7.541106128550075, + "grad_norm": 0.8209061622619629, + "learning_rate": 5.736173393124066e-05, + "loss": 3.4611, + "step": 110990 + }, + { + "epoch": 7.541445848620737, + "grad_norm": 0.9028964638710022, + "learning_rate": 5.7319268922407934e-05, + "loss": 3.2264, + "step": 110995 + }, + { + "epoch": 7.541785568691398, + "grad_norm": 0.990241527557373, + "learning_rate": 5.727680391357522e-05, + "loss": 3.4277, + "step": 111000 + }, + { + "epoch": 7.54212528876206, + "grad_norm": 1.0544756650924683, + "learning_rate": 5.7234338904742494e-05, + "loss": 3.3067, + "step": 111005 + }, + { + "epoch": 7.542465008832722, + "grad_norm": 0.9079288840293884, + "learning_rate": 5.719187389590977e-05, + "loss": 3.5406, + "step": 111010 + }, + { + "epoch": 7.542804728903383, + "grad_norm": 0.8898398280143738, + "learning_rate": 5.7149408887077054e-05, + "loss": 3.3121, + "step": 111015 + }, + { + "epoch": 7.543144448974045, + "grad_norm": 0.7700204849243164, + "learning_rate": 5.710694387824433e-05, + "loss": 3.429, + "step": 111020 + }, + { + "epoch": 7.543484169044707, + "grad_norm": 0.9840584397315979, + "learning_rate": 5.706447886941161e-05, + "loss": 3.3142, + "step": 111025 + }, + { + "epoch": 7.543823889115369, + "grad_norm": 1.0792350769042969, + "learning_rate": 5.702201386057888e-05, + "loss": 3.2678, + "step": 111030 + }, + { + "epoch": 7.544163609186031, + "grad_norm": 0.9383200407028198, + "learning_rate": 5.697954885174617e-05, + "loss": 3.3106, + "step": 111035 + }, + { + "epoch": 7.544503329256693, + "grad_norm": 1.0594160556793213, + "learning_rate": 5.693708384291344e-05, + "loss": 3.4251, + "step": 111040 + }, + { + "epoch": 7.544843049327354, + "grad_norm": 0.6557683348655701, + "learning_rate": 5.6894618834080714e-05, + "loss": 3.205, + "step": 111045 + }, + { + "epoch": 7.545182769398016, + "grad_norm": 0.7983291149139404, + "learning_rate": 5.6852153825248e-05, + "loss": 3.567, + "step": 111050 + }, + { + "epoch": 7.545522489468678, + "grad_norm": 0.9805039167404175, + "learning_rate": 5.6809688816415274e-05, + "loss": 3.5541, + "step": 111055 + }, + { + "epoch": 7.545862209539339, + "grad_norm": 0.9153196215629578, + "learning_rate": 5.6767223807582554e-05, + "loss": 3.2351, + "step": 111060 + }, + { + "epoch": 7.546201929610001, + "grad_norm": 1.0798025131225586, + "learning_rate": 5.6724758798749834e-05, + "loss": 3.2609, + "step": 111065 + }, + { + "epoch": 7.546541649680663, + "grad_norm": 0.8197891116142273, + "learning_rate": 5.668229378991711e-05, + "loss": 3.3271, + "step": 111070 + }, + { + "epoch": 7.546881369751325, + "grad_norm": 0.950478732585907, + "learning_rate": 5.663982878108439e-05, + "loss": 3.5432, + "step": 111075 + }, + { + "epoch": 7.547221089821987, + "grad_norm": 0.8308697938919067, + "learning_rate": 5.659736377225166e-05, + "loss": 3.2764, + "step": 111080 + }, + { + "epoch": 7.547560809892649, + "grad_norm": 0.9928393959999084, + "learning_rate": 5.655489876341895e-05, + "loss": 3.3101, + "step": 111085 + }, + { + "epoch": 7.54790052996331, + "grad_norm": 0.8602257966995239, + "learning_rate": 5.651243375458622e-05, + "loss": 3.5033, + "step": 111090 + }, + { + "epoch": 7.548240250033972, + "grad_norm": 1.0435714721679688, + "learning_rate": 5.6469968745753494e-05, + "loss": 3.5323, + "step": 111095 + }, + { + "epoch": 7.548579970104634, + "grad_norm": 0.8535974025726318, + "learning_rate": 5.642750373692078e-05, + "loss": 3.6639, + "step": 111100 + }, + { + "epoch": 7.548919690175295, + "grad_norm": 1.2821152210235596, + "learning_rate": 5.6385038728088055e-05, + "loss": 3.4679, + "step": 111105 + }, + { + "epoch": 7.549259410245957, + "grad_norm": 0.9056418538093567, + "learning_rate": 5.6342573719255335e-05, + "loss": 3.2576, + "step": 111110 + }, + { + "epoch": 7.5495991303166194, + "grad_norm": 0.996463418006897, + "learning_rate": 5.6300108710422615e-05, + "loss": 3.1544, + "step": 111115 + }, + { + "epoch": 7.549938850387281, + "grad_norm": 0.9648915529251099, + "learning_rate": 5.6257643701589895e-05, + "loss": 3.6583, + "step": 111120 + }, + { + "epoch": 7.550278570457943, + "grad_norm": 0.9583186507225037, + "learning_rate": 5.621517869275717e-05, + "loss": 3.3492, + "step": 111125 + }, + { + "epoch": 7.550618290528605, + "grad_norm": 0.8273849487304688, + "learning_rate": 5.617271368392445e-05, + "loss": 3.3091, + "step": 111130 + }, + { + "epoch": 7.550958010599266, + "grad_norm": 1.073859453201294, + "learning_rate": 5.613024867509173e-05, + "loss": 3.155, + "step": 111135 + }, + { + "epoch": 7.551297730669928, + "grad_norm": 1.0131057500839233, + "learning_rate": 5.6087783666259e-05, + "loss": 3.2966, + "step": 111140 + }, + { + "epoch": 7.55163745074059, + "grad_norm": 0.8649951815605164, + "learning_rate": 5.604531865742628e-05, + "loss": 3.2025, + "step": 111145 + }, + { + "epoch": 7.551977170811251, + "grad_norm": 0.793768584728241, + "learning_rate": 5.600285364859356e-05, + "loss": 3.4503, + "step": 111150 + }, + { + "epoch": 7.552316890881913, + "grad_norm": 0.9928789138793945, + "learning_rate": 5.5960388639760835e-05, + "loss": 3.4188, + "step": 111155 + }, + { + "epoch": 7.5526566109525755, + "grad_norm": 0.8680434823036194, + "learning_rate": 5.5917923630928115e-05, + "loss": 3.3565, + "step": 111160 + }, + { + "epoch": 7.552996331023237, + "grad_norm": 0.79287189245224, + "learning_rate": 5.5875458622095395e-05, + "loss": 3.3951, + "step": 111165 + }, + { + "epoch": 7.553336051093899, + "grad_norm": 1.0913664102554321, + "learning_rate": 5.5832993613262675e-05, + "loss": 3.4343, + "step": 111170 + }, + { + "epoch": 7.55367577116456, + "grad_norm": 0.8757702708244324, + "learning_rate": 5.579052860442995e-05, + "loss": 3.267, + "step": 111175 + }, + { + "epoch": 7.554015491235222, + "grad_norm": 0.947044312953949, + "learning_rate": 5.574806359559723e-05, + "loss": 3.3175, + "step": 111180 + }, + { + "epoch": 7.554355211305884, + "grad_norm": 0.8836825489997864, + "learning_rate": 5.570559858676451e-05, + "loss": 3.3906, + "step": 111185 + }, + { + "epoch": 7.554694931376545, + "grad_norm": 0.8767936825752258, + "learning_rate": 5.566313357793178e-05, + "loss": 3.4949, + "step": 111190 + }, + { + "epoch": 7.555034651447207, + "grad_norm": 1.2824370861053467, + "learning_rate": 5.562066856909907e-05, + "loss": 3.2157, + "step": 111195 + }, + { + "epoch": 7.555374371517869, + "grad_norm": 0.9593544006347656, + "learning_rate": 5.557820356026634e-05, + "loss": 3.5528, + "step": 111200 + }, + { + "epoch": 7.555714091588531, + "grad_norm": 1.1223008632659912, + "learning_rate": 5.553573855143362e-05, + "loss": 3.4589, + "step": 111205 + }, + { + "epoch": 7.556053811659193, + "grad_norm": 0.8613304495811462, + "learning_rate": 5.5493273542600896e-05, + "loss": 3.4855, + "step": 111210 + }, + { + "epoch": 7.556393531729855, + "grad_norm": 1.0302340984344482, + "learning_rate": 5.5450808533768176e-05, + "loss": 3.1251, + "step": 111215 + }, + { + "epoch": 7.556733251800516, + "grad_norm": 0.8375474214553833, + "learning_rate": 5.5408343524935456e-05, + "loss": 3.2986, + "step": 111220 + }, + { + "epoch": 7.557072971871178, + "grad_norm": 0.8688676357269287, + "learning_rate": 5.536587851610273e-05, + "loss": 3.3115, + "step": 111225 + }, + { + "epoch": 7.55741269194184, + "grad_norm": 1.0311559438705444, + "learning_rate": 5.5323413507270016e-05, + "loss": 3.3202, + "step": 111230 + }, + { + "epoch": 7.557752412012501, + "grad_norm": 1.0510016679763794, + "learning_rate": 5.528094849843729e-05, + "loss": 3.1988, + "step": 111235 + }, + { + "epoch": 7.558092132083163, + "grad_norm": 0.8699060082435608, + "learning_rate": 5.523848348960456e-05, + "loss": 3.4952, + "step": 111240 + }, + { + "epoch": 7.558431852153825, + "grad_norm": 1.679100513458252, + "learning_rate": 5.519601848077185e-05, + "loss": 3.4295, + "step": 111245 + }, + { + "epoch": 7.558771572224487, + "grad_norm": 0.8721213936805725, + "learning_rate": 5.515355347193912e-05, + "loss": 3.6323, + "step": 111250 + }, + { + "epoch": 7.559111292295149, + "grad_norm": 0.9326490163803101, + "learning_rate": 5.51110884631064e-05, + "loss": 3.4991, + "step": 111255 + }, + { + "epoch": 7.559451012365811, + "grad_norm": 0.8105872273445129, + "learning_rate": 5.506862345427368e-05, + "loss": 3.464, + "step": 111260 + }, + { + "epoch": 7.559790732436472, + "grad_norm": 0.8136166334152222, + "learning_rate": 5.5026158445440956e-05, + "loss": 3.4634, + "step": 111265 + }, + { + "epoch": 7.560130452507134, + "grad_norm": 0.9995914101600647, + "learning_rate": 5.4983693436608236e-05, + "loss": 3.1918, + "step": 111270 + }, + { + "epoch": 7.560470172577796, + "grad_norm": 0.8489272594451904, + "learning_rate": 5.494122842777551e-05, + "loss": 3.4538, + "step": 111275 + }, + { + "epoch": 7.560809892648457, + "grad_norm": 0.7710446715354919, + "learning_rate": 5.4898763418942796e-05, + "loss": 3.5303, + "step": 111280 + }, + { + "epoch": 7.561149612719119, + "grad_norm": 0.9068520069122314, + "learning_rate": 5.485629841011007e-05, + "loss": 3.3076, + "step": 111285 + }, + { + "epoch": 7.561489332789781, + "grad_norm": 1.1166956424713135, + "learning_rate": 5.481383340127735e-05, + "loss": 3.535, + "step": 111290 + }, + { + "epoch": 7.561829052860443, + "grad_norm": 0.8714536428451538, + "learning_rate": 5.477136839244463e-05, + "loss": 3.3533, + "step": 111295 + }, + { + "epoch": 7.562168772931105, + "grad_norm": 0.8512662053108215, + "learning_rate": 5.47289033836119e-05, + "loss": 3.7226, + "step": 111300 + }, + { + "epoch": 7.562508493001767, + "grad_norm": 0.8756119608879089, + "learning_rate": 5.468643837477918e-05, + "loss": 3.3734, + "step": 111305 + }, + { + "epoch": 7.562848213072428, + "grad_norm": 0.8924235105514526, + "learning_rate": 5.464397336594646e-05, + "loss": 3.6224, + "step": 111310 + }, + { + "epoch": 7.56318793314309, + "grad_norm": 1.0415834188461304, + "learning_rate": 5.460150835711374e-05, + "loss": 3.4665, + "step": 111315 + }, + { + "epoch": 7.563527653213752, + "grad_norm": 0.9248286485671997, + "learning_rate": 5.4559043348281017e-05, + "loss": 3.3935, + "step": 111320 + }, + { + "epoch": 7.563867373284413, + "grad_norm": 1.2285680770874023, + "learning_rate": 5.451657833944829e-05, + "loss": 3.3506, + "step": 111325 + }, + { + "epoch": 7.564207093355075, + "grad_norm": 0.9439234137535095, + "learning_rate": 5.447411333061558e-05, + "loss": 3.3653, + "step": 111330 + }, + { + "epoch": 7.564546813425737, + "grad_norm": 1.0479625463485718, + "learning_rate": 5.443164832178285e-05, + "loss": 3.548, + "step": 111335 + }, + { + "epoch": 7.564886533496399, + "grad_norm": 0.9372516870498657, + "learning_rate": 5.438918331295013e-05, + "loss": 3.4224, + "step": 111340 + }, + { + "epoch": 7.565226253567061, + "grad_norm": 0.8609635829925537, + "learning_rate": 5.434671830411741e-05, + "loss": 3.5155, + "step": 111345 + }, + { + "epoch": 7.565565973637723, + "grad_norm": 0.9773989319801331, + "learning_rate": 5.4304253295284684e-05, + "loss": 3.5107, + "step": 111350 + }, + { + "epoch": 7.565905693708384, + "grad_norm": 0.9072912931442261, + "learning_rate": 5.4261788286451964e-05, + "loss": 3.2729, + "step": 111355 + }, + { + "epoch": 7.566245413779046, + "grad_norm": 0.9252513647079468, + "learning_rate": 5.4219323277619244e-05, + "loss": 3.3696, + "step": 111360 + }, + { + "epoch": 7.566585133849708, + "grad_norm": 0.8352742791175842, + "learning_rate": 5.4176858268786524e-05, + "loss": 3.3752, + "step": 111365 + }, + { + "epoch": 7.566924853920369, + "grad_norm": 1.008560299873352, + "learning_rate": 5.41343932599538e-05, + "loss": 3.5391, + "step": 111370 + }, + { + "epoch": 7.567264573991031, + "grad_norm": 0.921235203742981, + "learning_rate": 5.409192825112108e-05, + "loss": 3.4074, + "step": 111375 + }, + { + "epoch": 7.567604294061693, + "grad_norm": 1.1577335596084595, + "learning_rate": 5.404946324228836e-05, + "loss": 3.4149, + "step": 111380 + }, + { + "epoch": 7.567944014132355, + "grad_norm": 0.858457088470459, + "learning_rate": 5.400699823345563e-05, + "loss": 3.4334, + "step": 111385 + }, + { + "epoch": 7.568283734203017, + "grad_norm": 0.8581017851829529, + "learning_rate": 5.396453322462291e-05, + "loss": 3.501, + "step": 111390 + }, + { + "epoch": 7.568623454273679, + "grad_norm": 0.8719479441642761, + "learning_rate": 5.392206821579019e-05, + "loss": 3.503, + "step": 111395 + }, + { + "epoch": 7.56896317434434, + "grad_norm": 0.9647455215454102, + "learning_rate": 5.387960320695747e-05, + "loss": 3.5857, + "step": 111400 + }, + { + "epoch": 7.569302894415002, + "grad_norm": 0.9365659952163696, + "learning_rate": 5.3837138198124744e-05, + "loss": 3.4399, + "step": 111405 + }, + { + "epoch": 7.569642614485664, + "grad_norm": 0.7658882737159729, + "learning_rate": 5.3794673189292024e-05, + "loss": 3.3854, + "step": 111410 + }, + { + "epoch": 7.569982334556325, + "grad_norm": 0.9190279245376587, + "learning_rate": 5.3752208180459304e-05, + "loss": 2.9707, + "step": 111415 + }, + { + "epoch": 7.570322054626987, + "grad_norm": 0.836206316947937, + "learning_rate": 5.370974317162658e-05, + "loss": 3.296, + "step": 111420 + }, + { + "epoch": 7.5706617746976494, + "grad_norm": 0.8941786885261536, + "learning_rate": 5.3667278162793864e-05, + "loss": 3.5768, + "step": 111425 + }, + { + "epoch": 7.571001494768311, + "grad_norm": 0.7884923815727234, + "learning_rate": 5.362481315396114e-05, + "loss": 3.4935, + "step": 111430 + }, + { + "epoch": 7.571341214838973, + "grad_norm": 1.1731880903244019, + "learning_rate": 5.358234814512841e-05, + "loss": 3.2385, + "step": 111435 + }, + { + "epoch": 7.571680934909635, + "grad_norm": 0.7744308114051819, + "learning_rate": 5.35398831362957e-05, + "loss": 3.4657, + "step": 111440 + }, + { + "epoch": 7.572020654980296, + "grad_norm": 1.080836534500122, + "learning_rate": 5.349741812746297e-05, + "loss": 3.6728, + "step": 111445 + }, + { + "epoch": 7.572360375050958, + "grad_norm": 1.0004469156265259, + "learning_rate": 5.345495311863025e-05, + "loss": 3.1852, + "step": 111450 + }, + { + "epoch": 7.57270009512162, + "grad_norm": 0.7801219820976257, + "learning_rate": 5.3412488109797524e-05, + "loss": 3.5011, + "step": 111455 + }, + { + "epoch": 7.573039815192281, + "grad_norm": 0.9347019791603088, + "learning_rate": 5.3370023100964805e-05, + "loss": 3.2902, + "step": 111460 + }, + { + "epoch": 7.573379535262943, + "grad_norm": 1.0607267618179321, + "learning_rate": 5.3327558092132085e-05, + "loss": 3.3505, + "step": 111465 + }, + { + "epoch": 7.5737192553336055, + "grad_norm": 0.9375846982002258, + "learning_rate": 5.328509308329936e-05, + "loss": 3.4122, + "step": 111470 + }, + { + "epoch": 7.574058975404267, + "grad_norm": 0.9699857234954834, + "learning_rate": 5.3242628074466645e-05, + "loss": 3.2719, + "step": 111475 + }, + { + "epoch": 7.574398695474929, + "grad_norm": 0.9520700573921204, + "learning_rate": 5.320016306563392e-05, + "loss": 3.644, + "step": 111480 + }, + { + "epoch": 7.574738415545591, + "grad_norm": 0.7177578806877136, + "learning_rate": 5.31576980568012e-05, + "loss": 3.1985, + "step": 111485 + }, + { + "epoch": 7.575078135616252, + "grad_norm": 0.8251407146453857, + "learning_rate": 5.311523304796848e-05, + "loss": 3.4786, + "step": 111490 + }, + { + "epoch": 7.575417855686914, + "grad_norm": 1.1382341384887695, + "learning_rate": 5.307276803913575e-05, + "loss": 3.5278, + "step": 111495 + }, + { + "epoch": 7.575757575757576, + "grad_norm": 0.8811636567115784, + "learning_rate": 5.303030303030303e-05, + "loss": 3.182, + "step": 111500 + }, + { + "epoch": 7.576097295828237, + "grad_norm": 0.8976104855537415, + "learning_rate": 5.2987838021470305e-05, + "loss": 3.4394, + "step": 111505 + }, + { + "epoch": 7.576437015898899, + "grad_norm": 1.2122604846954346, + "learning_rate": 5.294537301263759e-05, + "loss": 3.4827, + "step": 111510 + }, + { + "epoch": 7.5767767359695615, + "grad_norm": 0.8663927912712097, + "learning_rate": 5.2902908003804865e-05, + "loss": 3.2921, + "step": 111515 + }, + { + "epoch": 7.577116456040223, + "grad_norm": 1.1392680406570435, + "learning_rate": 5.286044299497214e-05, + "loss": 3.5626, + "step": 111520 + }, + { + "epoch": 7.577456176110885, + "grad_norm": 0.9306991696357727, + "learning_rate": 5.2817977986139425e-05, + "loss": 3.3624, + "step": 111525 + }, + { + "epoch": 7.577795896181547, + "grad_norm": 0.8655908703804016, + "learning_rate": 5.27755129773067e-05, + "loss": 3.4799, + "step": 111530 + }, + { + "epoch": 7.578135616252208, + "grad_norm": 0.9743436574935913, + "learning_rate": 5.273304796847398e-05, + "loss": 3.4766, + "step": 111535 + }, + { + "epoch": 7.57847533632287, + "grad_norm": 0.9043291211128235, + "learning_rate": 5.269058295964126e-05, + "loss": 3.1709, + "step": 111540 + }, + { + "epoch": 7.578815056393532, + "grad_norm": 0.8965036273002625, + "learning_rate": 5.264811795080853e-05, + "loss": 3.2828, + "step": 111545 + }, + { + "epoch": 7.579154776464193, + "grad_norm": 0.7237759828567505, + "learning_rate": 5.260565294197581e-05, + "loss": 3.2862, + "step": 111550 + }, + { + "epoch": 7.579494496534855, + "grad_norm": 1.0033278465270996, + "learning_rate": 5.256318793314309e-05, + "loss": 3.3481, + "step": 111555 + }, + { + "epoch": 7.5798342166055175, + "grad_norm": 0.8908172249794006, + "learning_rate": 5.252072292431037e-05, + "loss": 3.2198, + "step": 111560 + }, + { + "epoch": 7.580173936676179, + "grad_norm": 1.0971717834472656, + "learning_rate": 5.2478257915477646e-05, + "loss": 3.3697, + "step": 111565 + }, + { + "epoch": 7.580513656746841, + "grad_norm": 0.9116031527519226, + "learning_rate": 5.2435792906644926e-05, + "loss": 2.9649, + "step": 111570 + }, + { + "epoch": 7.580853376817503, + "grad_norm": 1.0805922746658325, + "learning_rate": 5.2393327897812206e-05, + "loss": 3.3218, + "step": 111575 + }, + { + "epoch": 7.581193096888164, + "grad_norm": 0.8234644532203674, + "learning_rate": 5.235086288897948e-05, + "loss": 3.1931, + "step": 111580 + }, + { + "epoch": 7.581532816958826, + "grad_norm": 0.7996472716331482, + "learning_rate": 5.230839788014676e-05, + "loss": 3.2902, + "step": 111585 + }, + { + "epoch": 7.581872537029488, + "grad_norm": 0.8287175893783569, + "learning_rate": 5.226593287131404e-05, + "loss": 3.4333, + "step": 111590 + }, + { + "epoch": 7.582212257100149, + "grad_norm": 0.8609719276428223, + "learning_rate": 5.222346786248132e-05, + "loss": 3.5679, + "step": 111595 + }, + { + "epoch": 7.582551977170811, + "grad_norm": 0.9544534087181091, + "learning_rate": 5.218100285364859e-05, + "loss": 3.4397, + "step": 111600 + }, + { + "epoch": 7.5828916972414735, + "grad_norm": 0.7875822186470032, + "learning_rate": 5.213853784481587e-05, + "loss": 3.5595, + "step": 111605 + }, + { + "epoch": 7.583231417312135, + "grad_norm": 0.8415759205818176, + "learning_rate": 5.209607283598315e-05, + "loss": 3.3663, + "step": 111610 + }, + { + "epoch": 7.583571137382797, + "grad_norm": 1.1494702100753784, + "learning_rate": 5.2053607827150426e-05, + "loss": 3.5064, + "step": 111615 + }, + { + "epoch": 7.583910857453459, + "grad_norm": 0.9578989744186401, + "learning_rate": 5.201114281831771e-05, + "loss": 3.3762, + "step": 111620 + }, + { + "epoch": 7.58425057752412, + "grad_norm": 0.9339717626571655, + "learning_rate": 5.1968677809484986e-05, + "loss": 3.3761, + "step": 111625 + }, + { + "epoch": 7.584590297594782, + "grad_norm": 0.8762722015380859, + "learning_rate": 5.192621280065226e-05, + "loss": 3.6402, + "step": 111630 + }, + { + "epoch": 7.584930017665444, + "grad_norm": 0.9513866901397705, + "learning_rate": 5.188374779181954e-05, + "loss": 3.4276, + "step": 111635 + }, + { + "epoch": 7.585269737736105, + "grad_norm": 1.098723292350769, + "learning_rate": 5.184128278298682e-05, + "loss": 3.2605, + "step": 111640 + }, + { + "epoch": 7.585609457806767, + "grad_norm": 0.858253002166748, + "learning_rate": 5.17988177741541e-05, + "loss": 3.3653, + "step": 111645 + }, + { + "epoch": 7.585949177877429, + "grad_norm": 0.9648464322090149, + "learning_rate": 5.175635276532137e-05, + "loss": 3.2448, + "step": 111650 + }, + { + "epoch": 7.586288897948091, + "grad_norm": 1.3925187587738037, + "learning_rate": 5.171388775648866e-05, + "loss": 3.4189, + "step": 111655 + }, + { + "epoch": 7.586628618018753, + "grad_norm": 1.016889214515686, + "learning_rate": 5.167142274765593e-05, + "loss": 3.527, + "step": 111660 + }, + { + "epoch": 7.586968338089414, + "grad_norm": 1.0235481262207031, + "learning_rate": 5.1628957738823206e-05, + "loss": 3.4228, + "step": 111665 + }, + { + "epoch": 7.587308058160076, + "grad_norm": 1.007304310798645, + "learning_rate": 5.158649272999049e-05, + "loss": 3.0922, + "step": 111670 + }, + { + "epoch": 7.587647778230738, + "grad_norm": 0.8527780771255493, + "learning_rate": 5.1544027721157767e-05, + "loss": 3.4633, + "step": 111675 + }, + { + "epoch": 7.587987498301399, + "grad_norm": 0.7841089367866516, + "learning_rate": 5.150156271232505e-05, + "loss": 3.3193, + "step": 111680 + }, + { + "epoch": 7.588327218372061, + "grad_norm": 0.9207803010940552, + "learning_rate": 5.145909770349232e-05, + "loss": 3.2593, + "step": 111685 + }, + { + "epoch": 7.588666938442723, + "grad_norm": 1.1558310985565186, + "learning_rate": 5.14166326946596e-05, + "loss": 3.3829, + "step": 111690 + }, + { + "epoch": 7.589006658513385, + "grad_norm": 1.0471961498260498, + "learning_rate": 5.137416768582688e-05, + "loss": 3.6023, + "step": 111695 + }, + { + "epoch": 7.589346378584047, + "grad_norm": 1.0591930150985718, + "learning_rate": 5.133170267699415e-05, + "loss": 3.2067, + "step": 111700 + }, + { + "epoch": 7.589686098654709, + "grad_norm": 0.813962459564209, + "learning_rate": 5.128923766816144e-05, + "loss": 3.4467, + "step": 111705 + }, + { + "epoch": 7.59002581872537, + "grad_norm": 0.8207576274871826, + "learning_rate": 5.1246772659328714e-05, + "loss": 3.3299, + "step": 111710 + }, + { + "epoch": 7.590365538796032, + "grad_norm": 1.1710807085037231, + "learning_rate": 5.120430765049599e-05, + "loss": 3.333, + "step": 111715 + }, + { + "epoch": 7.590705258866694, + "grad_norm": 0.7814620733261108, + "learning_rate": 5.1161842641663274e-05, + "loss": 3.5887, + "step": 111720 + }, + { + "epoch": 7.591044978937355, + "grad_norm": 0.8091585636138916, + "learning_rate": 5.111937763283055e-05, + "loss": 3.2846, + "step": 111725 + }, + { + "epoch": 7.591384699008017, + "grad_norm": 0.9756494164466858, + "learning_rate": 5.107691262399783e-05, + "loss": 3.1297, + "step": 111730 + }, + { + "epoch": 7.5917244190786795, + "grad_norm": 0.8965809941291809, + "learning_rate": 5.103444761516511e-05, + "loss": 3.6046, + "step": 111735 + }, + { + "epoch": 7.592064139149341, + "grad_norm": 0.824178159236908, + "learning_rate": 5.099198260633239e-05, + "loss": 3.4889, + "step": 111740 + }, + { + "epoch": 7.592403859220003, + "grad_norm": 1.0022507905960083, + "learning_rate": 5.094951759749966e-05, + "loss": 3.3245, + "step": 111745 + }, + { + "epoch": 7.592743579290665, + "grad_norm": 1.1455185413360596, + "learning_rate": 5.0907052588666934e-05, + "loss": 3.2896, + "step": 111750 + }, + { + "epoch": 7.593083299361326, + "grad_norm": 0.9631938338279724, + "learning_rate": 5.086458757983422e-05, + "loss": 3.3978, + "step": 111755 + }, + { + "epoch": 7.593423019431988, + "grad_norm": 0.9272576570510864, + "learning_rate": 5.0822122571001494e-05, + "loss": 3.3197, + "step": 111760 + }, + { + "epoch": 7.59376273950265, + "grad_norm": 0.8973345756530762, + "learning_rate": 5.0779657562168774e-05, + "loss": 3.3131, + "step": 111765 + }, + { + "epoch": 7.594102459573311, + "grad_norm": 0.8350628614425659, + "learning_rate": 5.0737192553336054e-05, + "loss": 3.3707, + "step": 111770 + }, + { + "epoch": 7.594442179643973, + "grad_norm": 0.9025888442993164, + "learning_rate": 5.069472754450333e-05, + "loss": 3.442, + "step": 111775 + }, + { + "epoch": 7.5947818997146355, + "grad_norm": 0.8326866626739502, + "learning_rate": 5.065226253567061e-05, + "loss": 3.3572, + "step": 111780 + }, + { + "epoch": 7.595121619785297, + "grad_norm": 1.0786900520324707, + "learning_rate": 5.060979752683789e-05, + "loss": 3.4053, + "step": 111785 + }, + { + "epoch": 7.595461339855959, + "grad_norm": 1.2602208852767944, + "learning_rate": 5.056733251800517e-05, + "loss": 3.2716, + "step": 111790 + }, + { + "epoch": 7.595801059926621, + "grad_norm": 0.915758490562439, + "learning_rate": 5.052486750917244e-05, + "loss": 3.4231, + "step": 111795 + }, + { + "epoch": 7.596140779997282, + "grad_norm": 0.8831523060798645, + "learning_rate": 5.048240250033972e-05, + "loss": 3.5992, + "step": 111800 + }, + { + "epoch": 7.596480500067944, + "grad_norm": 1.0511351823806763, + "learning_rate": 5.0439937491507e-05, + "loss": 3.3471, + "step": 111805 + }, + { + "epoch": 7.596820220138606, + "grad_norm": 1.0544785261154175, + "learning_rate": 5.0397472482674274e-05, + "loss": 3.4881, + "step": 111810 + }, + { + "epoch": 7.597159940209267, + "grad_norm": 0.8400909304618835, + "learning_rate": 5.0355007473841555e-05, + "loss": 3.1808, + "step": 111815 + }, + { + "epoch": 7.597499660279929, + "grad_norm": 0.9156176447868347, + "learning_rate": 5.0312542465008835e-05, + "loss": 3.2506, + "step": 111820 + }, + { + "epoch": 7.5978393803505915, + "grad_norm": 0.8244140148162842, + "learning_rate": 5.0270077456176115e-05, + "loss": 3.3569, + "step": 111825 + }, + { + "epoch": 7.598179100421253, + "grad_norm": 1.1398950815200806, + "learning_rate": 5.022761244734339e-05, + "loss": 3.5533, + "step": 111830 + }, + { + "epoch": 7.598518820491915, + "grad_norm": 0.7431821823120117, + "learning_rate": 5.018514743851067e-05, + "loss": 3.5162, + "step": 111835 + }, + { + "epoch": 7.598858540562577, + "grad_norm": 0.8085132241249084, + "learning_rate": 5.014268242967795e-05, + "loss": 3.403, + "step": 111840 + }, + { + "epoch": 7.599198260633238, + "grad_norm": 0.8091210126876831, + "learning_rate": 5.010021742084522e-05, + "loss": 3.2847, + "step": 111845 + }, + { + "epoch": 7.5995379807039, + "grad_norm": 1.2004276514053345, + "learning_rate": 5.005775241201251e-05, + "loss": 3.088, + "step": 111850 + }, + { + "epoch": 7.599877700774561, + "grad_norm": 1.0050560235977173, + "learning_rate": 5.001528740317978e-05, + "loss": 3.3809, + "step": 111855 + }, + { + "epoch": 7.600217420845223, + "grad_norm": 0.9806979894638062, + "learning_rate": 4.9972822394347055e-05, + "loss": 3.4645, + "step": 111860 + }, + { + "epoch": 7.600557140915885, + "grad_norm": 0.9567335247993469, + "learning_rate": 4.993035738551434e-05, + "loss": 3.3414, + "step": 111865 + }, + { + "epoch": 7.600896860986547, + "grad_norm": 1.2434544563293457, + "learning_rate": 4.9887892376681615e-05, + "loss": 3.3734, + "step": 111870 + }, + { + "epoch": 7.601236581057209, + "grad_norm": 0.7304790616035461, + "learning_rate": 4.9845427367848895e-05, + "loss": 3.1336, + "step": 111875 + }, + { + "epoch": 7.601576301127871, + "grad_norm": 0.8604336380958557, + "learning_rate": 4.980296235901617e-05, + "loss": 3.5269, + "step": 111880 + }, + { + "epoch": 7.601916021198532, + "grad_norm": 0.8884493708610535, + "learning_rate": 4.976049735018345e-05, + "loss": 3.335, + "step": 111885 + }, + { + "epoch": 7.602255741269194, + "grad_norm": 1.1965054273605347, + "learning_rate": 4.971803234135073e-05, + "loss": 3.3182, + "step": 111890 + }, + { + "epoch": 7.602595461339856, + "grad_norm": 0.837051510810852, + "learning_rate": 4.9675567332518e-05, + "loss": 3.4079, + "step": 111895 + }, + { + "epoch": 7.602935181410517, + "grad_norm": 0.8242466449737549, + "learning_rate": 4.963310232368529e-05, + "loss": 3.2895, + "step": 111900 + }, + { + "epoch": 7.603274901481179, + "grad_norm": 0.8733306527137756, + "learning_rate": 4.959063731485256e-05, + "loss": 3.2425, + "step": 111905 + }, + { + "epoch": 7.603614621551841, + "grad_norm": 0.8786354064941406, + "learning_rate": 4.954817230601984e-05, + "loss": 3.4336, + "step": 111910 + }, + { + "epoch": 7.603954341622503, + "grad_norm": 0.9797596335411072, + "learning_rate": 4.950570729718712e-05, + "loss": 3.393, + "step": 111915 + }, + { + "epoch": 7.604294061693165, + "grad_norm": 1.133903980255127, + "learning_rate": 4.9463242288354395e-05, + "loss": 3.4264, + "step": 111920 + }, + { + "epoch": 7.604633781763827, + "grad_norm": 0.9727842211723328, + "learning_rate": 4.9420777279521676e-05, + "loss": 3.2625, + "step": 111925 + }, + { + "epoch": 7.604973501834488, + "grad_norm": 0.9241257309913635, + "learning_rate": 4.937831227068895e-05, + "loss": 3.1284, + "step": 111930 + }, + { + "epoch": 7.60531322190515, + "grad_norm": 0.924166738986969, + "learning_rate": 4.9335847261856236e-05, + "loss": 3.3276, + "step": 111935 + }, + { + "epoch": 7.605652941975812, + "grad_norm": 0.9583518505096436, + "learning_rate": 4.929338225302351e-05, + "loss": 3.615, + "step": 111940 + }, + { + "epoch": 7.605992662046473, + "grad_norm": 0.9016560316085815, + "learning_rate": 4.925091724419078e-05, + "loss": 3.5444, + "step": 111945 + }, + { + "epoch": 7.606332382117135, + "grad_norm": 0.7895097732543945, + "learning_rate": 4.920845223535807e-05, + "loss": 3.5564, + "step": 111950 + }, + { + "epoch": 7.606672102187797, + "grad_norm": 0.9264260530471802, + "learning_rate": 4.916598722652534e-05, + "loss": 3.4691, + "step": 111955 + }, + { + "epoch": 7.607011822258459, + "grad_norm": 0.7818512916564941, + "learning_rate": 4.912352221769262e-05, + "loss": 3.2046, + "step": 111960 + }, + { + "epoch": 7.607351542329121, + "grad_norm": 0.8162993788719177, + "learning_rate": 4.90810572088599e-05, + "loss": 3.3246, + "step": 111965 + }, + { + "epoch": 7.607691262399783, + "grad_norm": 0.9048073291778564, + "learning_rate": 4.9038592200027176e-05, + "loss": 3.4005, + "step": 111970 + }, + { + "epoch": 7.608030982470444, + "grad_norm": 1.3690139055252075, + "learning_rate": 4.8996127191194456e-05, + "loss": 3.5115, + "step": 111975 + }, + { + "epoch": 7.608370702541106, + "grad_norm": 0.8214221000671387, + "learning_rate": 4.8953662182361736e-05, + "loss": 3.2958, + "step": 111980 + }, + { + "epoch": 7.608710422611768, + "grad_norm": 1.1427544355392456, + "learning_rate": 4.8911197173529016e-05, + "loss": 3.1946, + "step": 111985 + }, + { + "epoch": 7.609050142682429, + "grad_norm": 0.7745341658592224, + "learning_rate": 4.886873216469629e-05, + "loss": 3.4174, + "step": 111990 + }, + { + "epoch": 7.609389862753091, + "grad_norm": 0.7793834805488586, + "learning_rate": 4.882626715586357e-05, + "loss": 3.5307, + "step": 111995 + }, + { + "epoch": 7.6097295828237534, + "grad_norm": 0.9361981153488159, + "learning_rate": 4.878380214703085e-05, + "loss": 3.5153, + "step": 112000 + }, + { + "epoch": 7.610069302894415, + "grad_norm": 0.8734991550445557, + "learning_rate": 4.874133713819812e-05, + "loss": 3.3194, + "step": 112005 + }, + { + "epoch": 7.610409022965077, + "grad_norm": 0.8654208779335022, + "learning_rate": 4.86988721293654e-05, + "loss": 3.5245, + "step": 112010 + }, + { + "epoch": 7.610748743035739, + "grad_norm": 0.9341074824333191, + "learning_rate": 4.865640712053268e-05, + "loss": 2.858, + "step": 112015 + }, + { + "epoch": 7.6110884631064, + "grad_norm": 0.9080499410629272, + "learning_rate": 4.861394211169996e-05, + "loss": 3.4265, + "step": 112020 + }, + { + "epoch": 7.611428183177062, + "grad_norm": 0.9377245903015137, + "learning_rate": 4.8571477102867236e-05, + "loss": 3.4497, + "step": 112025 + }, + { + "epoch": 7.611767903247724, + "grad_norm": 0.839765191078186, + "learning_rate": 4.8529012094034517e-05, + "loss": 3.4943, + "step": 112030 + }, + { + "epoch": 7.612107623318385, + "grad_norm": 0.9068822264671326, + "learning_rate": 4.8486547085201797e-05, + "loss": 3.5063, + "step": 112035 + }, + { + "epoch": 7.612447343389047, + "grad_norm": 1.0940858125686646, + "learning_rate": 4.844408207636907e-05, + "loss": 3.1556, + "step": 112040 + }, + { + "epoch": 7.6127870634597095, + "grad_norm": 0.9192482233047485, + "learning_rate": 4.840161706753636e-05, + "loss": 3.1811, + "step": 112045 + }, + { + "epoch": 7.613126783530371, + "grad_norm": 0.8146933317184448, + "learning_rate": 4.835915205870363e-05, + "loss": 3.3829, + "step": 112050 + }, + { + "epoch": 7.613466503601033, + "grad_norm": 1.079905390739441, + "learning_rate": 4.83166870498709e-05, + "loss": 3.4877, + "step": 112055 + }, + { + "epoch": 7.613806223671695, + "grad_norm": 1.1164724826812744, + "learning_rate": 4.8274222041038183e-05, + "loss": 3.1894, + "step": 112060 + }, + { + "epoch": 7.614145943742356, + "grad_norm": 0.990015983581543, + "learning_rate": 4.8231757032205463e-05, + "loss": 3.189, + "step": 112065 + }, + { + "epoch": 7.614485663813018, + "grad_norm": 1.0278226137161255, + "learning_rate": 4.8189292023372744e-05, + "loss": 3.2766, + "step": 112070 + }, + { + "epoch": 7.61482538388368, + "grad_norm": 1.0016294717788696, + "learning_rate": 4.814682701454002e-05, + "loss": 3.4699, + "step": 112075 + }, + { + "epoch": 7.615165103954341, + "grad_norm": 1.0545889139175415, + "learning_rate": 4.8104362005707304e-05, + "loss": 3.3877, + "step": 112080 + }, + { + "epoch": 7.615504824025003, + "grad_norm": 1.142643690109253, + "learning_rate": 4.806189699687458e-05, + "loss": 3.3883, + "step": 112085 + }, + { + "epoch": 7.6158445440956655, + "grad_norm": 1.0644557476043701, + "learning_rate": 4.801943198804185e-05, + "loss": 3.4035, + "step": 112090 + }, + { + "epoch": 7.616184264166327, + "grad_norm": 0.8540008068084717, + "learning_rate": 4.797696697920914e-05, + "loss": 3.5173, + "step": 112095 + }, + { + "epoch": 7.616523984236989, + "grad_norm": 0.9194421768188477, + "learning_rate": 4.793450197037641e-05, + "loss": 3.4647, + "step": 112100 + }, + { + "epoch": 7.616863704307651, + "grad_norm": 0.9604576230049133, + "learning_rate": 4.789203696154369e-05, + "loss": 3.529, + "step": 112105 + }, + { + "epoch": 7.617203424378312, + "grad_norm": 0.9075765013694763, + "learning_rate": 4.7849571952710964e-05, + "loss": 3.3764, + "step": 112110 + }, + { + "epoch": 7.617543144448974, + "grad_norm": 0.8346485495567322, + "learning_rate": 4.7807106943878244e-05, + "loss": 3.4051, + "step": 112115 + }, + { + "epoch": 7.617882864519636, + "grad_norm": 0.7817507982254028, + "learning_rate": 4.7764641935045524e-05, + "loss": 3.4449, + "step": 112120 + }, + { + "epoch": 7.618222584590297, + "grad_norm": 0.7323435544967651, + "learning_rate": 4.77221769262128e-05, + "loss": 3.6375, + "step": 112125 + }, + { + "epoch": 7.618562304660959, + "grad_norm": 1.0636945962905884, + "learning_rate": 4.7679711917380084e-05, + "loss": 3.2779, + "step": 112130 + }, + { + "epoch": 7.6189020247316215, + "grad_norm": 3.713388442993164, + "learning_rate": 4.763724690854736e-05, + "loss": 3.6197, + "step": 112135 + }, + { + "epoch": 7.619241744802283, + "grad_norm": 1.050538182258606, + "learning_rate": 4.759478189971463e-05, + "loss": 3.2127, + "step": 112140 + }, + { + "epoch": 7.619581464872945, + "grad_norm": 0.9820727109909058, + "learning_rate": 4.755231689088192e-05, + "loss": 3.4757, + "step": 112145 + }, + { + "epoch": 7.619921184943607, + "grad_norm": 1.0723868608474731, + "learning_rate": 4.750985188204919e-05, + "loss": 3.1791, + "step": 112150 + }, + { + "epoch": 7.620260905014268, + "grad_norm": 1.0317169427871704, + "learning_rate": 4.746738687321647e-05, + "loss": 3.4365, + "step": 112155 + }, + { + "epoch": 7.62060062508493, + "grad_norm": 0.858337938785553, + "learning_rate": 4.742492186438375e-05, + "loss": 3.5017, + "step": 112160 + }, + { + "epoch": 7.620940345155592, + "grad_norm": 1.2008229494094849, + "learning_rate": 4.738245685555103e-05, + "loss": 3.4689, + "step": 112165 + }, + { + "epoch": 7.621280065226253, + "grad_norm": 0.9559288024902344, + "learning_rate": 4.7339991846718304e-05, + "loss": 3.3222, + "step": 112170 + }, + { + "epoch": 7.621619785296915, + "grad_norm": 1.2915655374526978, + "learning_rate": 4.729752683788558e-05, + "loss": 3.4272, + "step": 112175 + }, + { + "epoch": 7.6219595053675775, + "grad_norm": 1.3384760618209839, + "learning_rate": 4.7255061829052865e-05, + "loss": 3.226, + "step": 112180 + }, + { + "epoch": 7.622299225438239, + "grad_norm": 0.8469418883323669, + "learning_rate": 4.721259682022014e-05, + "loss": 3.507, + "step": 112185 + }, + { + "epoch": 7.622638945508901, + "grad_norm": 1.1013795137405396, + "learning_rate": 4.717013181138742e-05, + "loss": 3.5279, + "step": 112190 + }, + { + "epoch": 7.622978665579563, + "grad_norm": 0.9401018619537354, + "learning_rate": 4.71276668025547e-05, + "loss": 3.2521, + "step": 112195 + }, + { + "epoch": 7.623318385650224, + "grad_norm": 0.7541585564613342, + "learning_rate": 4.708520179372197e-05, + "loss": 3.3648, + "step": 112200 + }, + { + "epoch": 7.623658105720886, + "grad_norm": 0.947014331817627, + "learning_rate": 4.704273678488925e-05, + "loss": 3.2813, + "step": 112205 + }, + { + "epoch": 7.623997825791548, + "grad_norm": 0.8956754803657532, + "learning_rate": 4.700027177605653e-05, + "loss": 3.1928, + "step": 112210 + }, + { + "epoch": 7.624337545862209, + "grad_norm": 1.0171830654144287, + "learning_rate": 4.695780676722381e-05, + "loss": 3.4644, + "step": 112215 + }, + { + "epoch": 7.624677265932871, + "grad_norm": 1.0344239473342896, + "learning_rate": 4.6915341758391085e-05, + "loss": 3.3253, + "step": 112220 + }, + { + "epoch": 7.6250169860035335, + "grad_norm": 1.0483454465866089, + "learning_rate": 4.6872876749558365e-05, + "loss": 3.3824, + "step": 112225 + }, + { + "epoch": 7.625356706074195, + "grad_norm": 0.9976480007171631, + "learning_rate": 4.6830411740725645e-05, + "loss": 3.6758, + "step": 112230 + }, + { + "epoch": 7.625696426144857, + "grad_norm": 0.7807690501213074, + "learning_rate": 4.678794673189292e-05, + "loss": 3.4928, + "step": 112235 + }, + { + "epoch": 7.626036146215519, + "grad_norm": 0.9342118501663208, + "learning_rate": 4.67454817230602e-05, + "loss": 3.4237, + "step": 112240 + }, + { + "epoch": 7.62637586628618, + "grad_norm": 0.8268876075744629, + "learning_rate": 4.670301671422748e-05, + "loss": 3.5006, + "step": 112245 + }, + { + "epoch": 7.626715586356842, + "grad_norm": 0.906776487827301, + "learning_rate": 4.666055170539476e-05, + "loss": 3.5164, + "step": 112250 + }, + { + "epoch": 7.627055306427504, + "grad_norm": 0.9543066024780273, + "learning_rate": 4.661808669656203e-05, + "loss": 3.1111, + "step": 112255 + }, + { + "epoch": 7.627395026498165, + "grad_norm": 0.8475157618522644, + "learning_rate": 4.657562168772931e-05, + "loss": 3.4357, + "step": 112260 + }, + { + "epoch": 7.627734746568827, + "grad_norm": 1.1420366764068604, + "learning_rate": 4.653315667889659e-05, + "loss": 3.5245, + "step": 112265 + }, + { + "epoch": 7.6280744666394895, + "grad_norm": 0.9010941982269287, + "learning_rate": 4.6490691670063865e-05, + "loss": 3.3425, + "step": 112270 + }, + { + "epoch": 7.628414186710151, + "grad_norm": 1.088998794555664, + "learning_rate": 4.644822666123115e-05, + "loss": 3.2931, + "step": 112275 + }, + { + "epoch": 7.628753906780813, + "grad_norm": 0.9250776171684265, + "learning_rate": 4.6405761652398426e-05, + "loss": 3.4557, + "step": 112280 + }, + { + "epoch": 7.629093626851475, + "grad_norm": 0.9425126910209656, + "learning_rate": 4.63632966435657e-05, + "loss": 3.5754, + "step": 112285 + }, + { + "epoch": 7.629433346922136, + "grad_norm": 0.9662245512008667, + "learning_rate": 4.632083163473298e-05, + "loss": 3.2027, + "step": 112290 + }, + { + "epoch": 7.629773066992798, + "grad_norm": 0.9992597699165344, + "learning_rate": 4.627836662590026e-05, + "loss": 3.3521, + "step": 112295 + }, + { + "epoch": 7.63011278706346, + "grad_norm": 1.381895661354065, + "learning_rate": 4.623590161706754e-05, + "loss": 3.4987, + "step": 112300 + }, + { + "epoch": 7.630452507134121, + "grad_norm": 0.9785231947898865, + "learning_rate": 4.619343660823481e-05, + "loss": 3.289, + "step": 112305 + }, + { + "epoch": 7.6307922272047835, + "grad_norm": 1.3509185314178467, + "learning_rate": 4.615097159940209e-05, + "loss": 3.4052, + "step": 112310 + }, + { + "epoch": 7.6311319472754455, + "grad_norm": 1.4965996742248535, + "learning_rate": 4.610850659056937e-05, + "loss": 3.3932, + "step": 112315 + }, + { + "epoch": 7.631471667346107, + "grad_norm": 0.9859437346458435, + "learning_rate": 4.6066041581736646e-05, + "loss": 3.5009, + "step": 112320 + }, + { + "epoch": 7.631811387416769, + "grad_norm": 0.8212019801139832, + "learning_rate": 4.602357657290393e-05, + "loss": 3.3645, + "step": 112325 + }, + { + "epoch": 7.63215110748743, + "grad_norm": 1.9555498361587524, + "learning_rate": 4.5981111564071206e-05, + "loss": 3.6064, + "step": 112330 + }, + { + "epoch": 7.632490827558092, + "grad_norm": 0.9153217077255249, + "learning_rate": 4.5938646555238486e-05, + "loss": 3.4385, + "step": 112335 + }, + { + "epoch": 7.632830547628754, + "grad_norm": 0.8776251077651978, + "learning_rate": 4.5896181546405766e-05, + "loss": 3.4141, + "step": 112340 + }, + { + "epoch": 7.633170267699415, + "grad_norm": 1.0604723691940308, + "learning_rate": 4.585371653757304e-05, + "loss": 3.4364, + "step": 112345 + }, + { + "epoch": 7.633509987770077, + "grad_norm": 0.9204199314117432, + "learning_rate": 4.581125152874032e-05, + "loss": 3.4274, + "step": 112350 + }, + { + "epoch": 7.6338497078407395, + "grad_norm": 0.9260985255241394, + "learning_rate": 4.576878651990759e-05, + "loss": 3.3995, + "step": 112355 + }, + { + "epoch": 7.634189427911401, + "grad_norm": 1.137837290763855, + "learning_rate": 4.572632151107488e-05, + "loss": 3.5071, + "step": 112360 + }, + { + "epoch": 7.634529147982063, + "grad_norm": 0.873797595500946, + "learning_rate": 4.568385650224215e-05, + "loss": 3.2141, + "step": 112365 + }, + { + "epoch": 7.634868868052725, + "grad_norm": 0.8722084760665894, + "learning_rate": 4.5641391493409426e-05, + "loss": 3.4455, + "step": 112370 + }, + { + "epoch": 7.635208588123386, + "grad_norm": 0.8186711668968201, + "learning_rate": 4.559892648457671e-05, + "loss": 3.4403, + "step": 112375 + }, + { + "epoch": 7.635548308194048, + "grad_norm": 0.9097861647605896, + "learning_rate": 4.5556461475743986e-05, + "loss": 3.4898, + "step": 112380 + }, + { + "epoch": 7.63588802826471, + "grad_norm": 1.4737153053283691, + "learning_rate": 4.5513996466911266e-05, + "loss": 3.3171, + "step": 112385 + }, + { + "epoch": 7.636227748335371, + "grad_norm": 0.9207075834274292, + "learning_rate": 4.5471531458078547e-05, + "loss": 3.1552, + "step": 112390 + }, + { + "epoch": 7.636567468406033, + "grad_norm": 0.9077394604682922, + "learning_rate": 4.542906644924582e-05, + "loss": 3.4447, + "step": 112395 + }, + { + "epoch": 7.6369071884766955, + "grad_norm": 0.9876229166984558, + "learning_rate": 4.53866014404131e-05, + "loss": 3.1586, + "step": 112400 + }, + { + "epoch": 7.637246908547357, + "grad_norm": 1.0863884687423706, + "learning_rate": 4.534413643158038e-05, + "loss": 3.6068, + "step": 112405 + }, + { + "epoch": 7.637586628618019, + "grad_norm": 0.9499063491821289, + "learning_rate": 4.530167142274766e-05, + "loss": 3.388, + "step": 112410 + }, + { + "epoch": 7.637926348688681, + "grad_norm": 0.9988776445388794, + "learning_rate": 4.525920641391493e-05, + "loss": 3.2331, + "step": 112415 + }, + { + "epoch": 7.638266068759342, + "grad_norm": 0.9412990808486938, + "learning_rate": 4.5216741405082213e-05, + "loss": 3.4073, + "step": 112420 + }, + { + "epoch": 7.638605788830004, + "grad_norm": 2.2250349521636963, + "learning_rate": 4.5174276396249494e-05, + "loss": 3.429, + "step": 112425 + }, + { + "epoch": 7.638945508900666, + "grad_norm": 1.0390174388885498, + "learning_rate": 4.513181138741677e-05, + "loss": 3.4601, + "step": 112430 + }, + { + "epoch": 7.639285228971327, + "grad_norm": 0.8819515109062195, + "learning_rate": 4.508934637858405e-05, + "loss": 3.1811, + "step": 112435 + }, + { + "epoch": 7.639624949041989, + "grad_norm": 1.0449726581573486, + "learning_rate": 4.504688136975133e-05, + "loss": 3.492, + "step": 112440 + }, + { + "epoch": 7.6399646691126515, + "grad_norm": 0.9633982181549072, + "learning_rate": 4.500441636091861e-05, + "loss": 3.5389, + "step": 112445 + }, + { + "epoch": 7.640304389183313, + "grad_norm": 1.4608879089355469, + "learning_rate": 4.496195135208588e-05, + "loss": 3.4427, + "step": 112450 + }, + { + "epoch": 7.640644109253975, + "grad_norm": 1.3589016199111938, + "learning_rate": 4.491948634325316e-05, + "loss": 3.3485, + "step": 112455 + }, + { + "epoch": 7.640983829324637, + "grad_norm": 1.045647382736206, + "learning_rate": 4.487702133442044e-05, + "loss": 3.2072, + "step": 112460 + }, + { + "epoch": 7.641323549395298, + "grad_norm": 0.9246186017990112, + "learning_rate": 4.4834556325587714e-05, + "loss": 3.3442, + "step": 112465 + }, + { + "epoch": 7.64166326946596, + "grad_norm": 1.0020135641098022, + "learning_rate": 4.4792091316755e-05, + "loss": 3.3602, + "step": 112470 + }, + { + "epoch": 7.642002989536622, + "grad_norm": 0.8674826622009277, + "learning_rate": 4.4749626307922274e-05, + "loss": 3.4681, + "step": 112475 + }, + { + "epoch": 7.642342709607283, + "grad_norm": 0.9258780479431152, + "learning_rate": 4.470716129908955e-05, + "loss": 3.4779, + "step": 112480 + }, + { + "epoch": 7.642682429677945, + "grad_norm": 1.165784478187561, + "learning_rate": 4.466469629025683e-05, + "loss": 3.4729, + "step": 112485 + }, + { + "epoch": 7.6430221497486075, + "grad_norm": 0.9795543551445007, + "learning_rate": 4.462223128142411e-05, + "loss": 2.959, + "step": 112490 + }, + { + "epoch": 7.643361869819269, + "grad_norm": 1.0003658533096313, + "learning_rate": 4.457976627259139e-05, + "loss": 3.3433, + "step": 112495 + }, + { + "epoch": 7.643701589889931, + "grad_norm": 0.7739466428756714, + "learning_rate": 4.453730126375866e-05, + "loss": 3.5583, + "step": 112500 + }, + { + "epoch": 7.644041309960593, + "grad_norm": 0.8650500774383545, + "learning_rate": 4.449483625492595e-05, + "loss": 3.4858, + "step": 112505 + }, + { + "epoch": 7.644381030031254, + "grad_norm": 0.701994001865387, + "learning_rate": 4.445237124609322e-05, + "loss": 3.4541, + "step": 112510 + }, + { + "epoch": 7.644720750101916, + "grad_norm": 0.7601056694984436, + "learning_rate": 4.4409906237260494e-05, + "loss": 3.4546, + "step": 112515 + }, + { + "epoch": 7.645060470172578, + "grad_norm": 0.9422383308410645, + "learning_rate": 4.436744122842778e-05, + "loss": 3.3052, + "step": 112520 + }, + { + "epoch": 7.645400190243239, + "grad_norm": 1.045641303062439, + "learning_rate": 4.4324976219595054e-05, + "loss": 3.536, + "step": 112525 + }, + { + "epoch": 7.645739910313901, + "grad_norm": 1.1622087955474854, + "learning_rate": 4.4282511210762334e-05, + "loss": 3.3761, + "step": 112530 + }, + { + "epoch": 7.6460796303845635, + "grad_norm": 0.7555384039878845, + "learning_rate": 4.424004620192961e-05, + "loss": 3.5168, + "step": 112535 + }, + { + "epoch": 7.646419350455225, + "grad_norm": 0.9282976984977722, + "learning_rate": 4.419758119309689e-05, + "loss": 3.1276, + "step": 112540 + }, + { + "epoch": 7.646759070525887, + "grad_norm": 1.043043851852417, + "learning_rate": 4.415511618426417e-05, + "loss": 3.5077, + "step": 112545 + }, + { + "epoch": 7.647098790596548, + "grad_norm": 0.9584177136421204, + "learning_rate": 4.411265117543144e-05, + "loss": 3.6907, + "step": 112550 + }, + { + "epoch": 7.64743851066721, + "grad_norm": 0.9730686545372009, + "learning_rate": 4.407018616659873e-05, + "loss": 3.1828, + "step": 112555 + }, + { + "epoch": 7.647778230737872, + "grad_norm": 0.7485954761505127, + "learning_rate": 4.4027721157766e-05, + "loss": 3.4326, + "step": 112560 + }, + { + "epoch": 7.648117950808533, + "grad_norm": 1.2227376699447632, + "learning_rate": 4.3985256148933275e-05, + "loss": 3.1249, + "step": 112565 + }, + { + "epoch": 7.648457670879195, + "grad_norm": 0.8316325545310974, + "learning_rate": 4.394279114010056e-05, + "loss": 3.3478, + "step": 112570 + }, + { + "epoch": 7.6487973909498574, + "grad_norm": 0.9437422752380371, + "learning_rate": 4.3900326131267835e-05, + "loss": 3.3848, + "step": 112575 + }, + { + "epoch": 7.649137111020519, + "grad_norm": 1.0993919372558594, + "learning_rate": 4.3857861122435115e-05, + "loss": 3.4643, + "step": 112580 + }, + { + "epoch": 7.649476831091181, + "grad_norm": 0.831368088722229, + "learning_rate": 4.3815396113602395e-05, + "loss": 3.3183, + "step": 112585 + }, + { + "epoch": 7.649816551161843, + "grad_norm": 0.9772201180458069, + "learning_rate": 4.3772931104769675e-05, + "loss": 3.5778, + "step": 112590 + }, + { + "epoch": 7.650156271232504, + "grad_norm": 0.7355365753173828, + "learning_rate": 4.373046609593695e-05, + "loss": 3.6079, + "step": 112595 + }, + { + "epoch": 7.650495991303166, + "grad_norm": 0.9603703618049622, + "learning_rate": 4.368800108710422e-05, + "loss": 3.2912, + "step": 112600 + }, + { + "epoch": 7.650835711373828, + "grad_norm": 1.1698410511016846, + "learning_rate": 4.364553607827151e-05, + "loss": 3.4516, + "step": 112605 + }, + { + "epoch": 7.651175431444489, + "grad_norm": 1.5403573513031006, + "learning_rate": 4.360307106943878e-05, + "loss": 3.1489, + "step": 112610 + }, + { + "epoch": 7.651515151515151, + "grad_norm": 0.9207063317298889, + "learning_rate": 4.356060606060606e-05, + "loss": 3.541, + "step": 112615 + }, + { + "epoch": 7.6518548715858135, + "grad_norm": 0.8532674908638, + "learning_rate": 4.351814105177334e-05, + "loss": 3.0516, + "step": 112620 + }, + { + "epoch": 7.652194591656475, + "grad_norm": 1.1170414686203003, + "learning_rate": 4.3475676042940615e-05, + "loss": 3.4088, + "step": 112625 + }, + { + "epoch": 7.652534311727137, + "grad_norm": 0.9707779288291931, + "learning_rate": 4.3433211034107895e-05, + "loss": 3.4437, + "step": 112630 + }, + { + "epoch": 7.652874031797799, + "grad_norm": 1.0121650695800781, + "learning_rate": 4.3390746025275175e-05, + "loss": 3.2012, + "step": 112635 + }, + { + "epoch": 7.65321375186846, + "grad_norm": 0.863158643245697, + "learning_rate": 4.3348281016442456e-05, + "loss": 3.5915, + "step": 112640 + }, + { + "epoch": 7.653553471939122, + "grad_norm": 1.148604393005371, + "learning_rate": 4.330581600760973e-05, + "loss": 3.3969, + "step": 112645 + }, + { + "epoch": 7.653893192009784, + "grad_norm": 0.8766696453094482, + "learning_rate": 4.3263350998777e-05, + "loss": 3.5022, + "step": 112650 + }, + { + "epoch": 7.654232912080445, + "grad_norm": 0.9564868211746216, + "learning_rate": 4.322088598994429e-05, + "loss": 3.4201, + "step": 112655 + }, + { + "epoch": 7.654572632151107, + "grad_norm": 1.0905125141143799, + "learning_rate": 4.317842098111156e-05, + "loss": 3.3234, + "step": 112660 + }, + { + "epoch": 7.6549123522217695, + "grad_norm": 1.1032371520996094, + "learning_rate": 4.313595597227884e-05, + "loss": 3.2015, + "step": 112665 + }, + { + "epoch": 7.655252072292431, + "grad_norm": 0.9453433752059937, + "learning_rate": 4.309349096344612e-05, + "loss": 3.0931, + "step": 112670 + }, + { + "epoch": 7.655591792363093, + "grad_norm": 1.0070751905441284, + "learning_rate": 4.30510259546134e-05, + "loss": 3.6004, + "step": 112675 + }, + { + "epoch": 7.655931512433755, + "grad_norm": 1.0812243223190308, + "learning_rate": 4.3008560945780676e-05, + "loss": 3.403, + "step": 112680 + }, + { + "epoch": 7.656271232504416, + "grad_norm": 1.069840431213379, + "learning_rate": 4.2966095936947956e-05, + "loss": 3.255, + "step": 112685 + }, + { + "epoch": 7.656610952575078, + "grad_norm": 0.9023341536521912, + "learning_rate": 4.2923630928115236e-05, + "loss": 3.3823, + "step": 112690 + }, + { + "epoch": 7.65695067264574, + "grad_norm": 1.076233983039856, + "learning_rate": 4.288116591928251e-05, + "loss": 3.3689, + "step": 112695 + }, + { + "epoch": 7.657290392716401, + "grad_norm": 0.8512434363365173, + "learning_rate": 4.2838700910449796e-05, + "loss": 3.2427, + "step": 112700 + }, + { + "epoch": 7.657630112787063, + "grad_norm": 1.0255650281906128, + "learning_rate": 4.279623590161707e-05, + "loss": 3.288, + "step": 112705 + }, + { + "epoch": 7.6579698328577255, + "grad_norm": 0.9518836736679077, + "learning_rate": 4.275377089278434e-05, + "loss": 3.3735, + "step": 112710 + }, + { + "epoch": 7.658309552928387, + "grad_norm": 1.060813307762146, + "learning_rate": 4.271130588395162e-05, + "loss": 3.5745, + "step": 112715 + }, + { + "epoch": 7.658649272999049, + "grad_norm": 1.2778875827789307, + "learning_rate": 4.26688408751189e-05, + "loss": 3.7225, + "step": 112720 + }, + { + "epoch": 7.658988993069711, + "grad_norm": 1.035233736038208, + "learning_rate": 4.262637586628618e-05, + "loss": 3.177, + "step": 112725 + }, + { + "epoch": 7.659328713140372, + "grad_norm": 1.0202096700668335, + "learning_rate": 4.2583910857453456e-05, + "loss": 3.2512, + "step": 112730 + }, + { + "epoch": 7.659668433211034, + "grad_norm": 0.9746663570404053, + "learning_rate": 4.2541445848620736e-05, + "loss": 3.4056, + "step": 112735 + }, + { + "epoch": 7.660008153281696, + "grad_norm": 0.8413648009300232, + "learning_rate": 4.2498980839788016e-05, + "loss": 3.6096, + "step": 112740 + }, + { + "epoch": 7.660347873352357, + "grad_norm": 0.9626955389976501, + "learning_rate": 4.245651583095529e-05, + "loss": 3.6162, + "step": 112745 + }, + { + "epoch": 7.660687593423019, + "grad_norm": 0.9585137367248535, + "learning_rate": 4.2414050822122577e-05, + "loss": 3.3415, + "step": 112750 + }, + { + "epoch": 7.6610273134936815, + "grad_norm": 0.9695428609848022, + "learning_rate": 4.237158581328985e-05, + "loss": 3.5522, + "step": 112755 + }, + { + "epoch": 7.661367033564343, + "grad_norm": 0.9291563630104065, + "learning_rate": 4.232912080445713e-05, + "loss": 3.2421, + "step": 112760 + }, + { + "epoch": 7.661706753635005, + "grad_norm": 0.8623147010803223, + "learning_rate": 4.228665579562441e-05, + "loss": 3.3069, + "step": 112765 + }, + { + "epoch": 7.662046473705667, + "grad_norm": 0.7963269948959351, + "learning_rate": 4.224419078679168e-05, + "loss": 3.4123, + "step": 112770 + }, + { + "epoch": 7.662386193776328, + "grad_norm": 0.7782647609710693, + "learning_rate": 4.2201725777958963e-05, + "loss": 3.5804, + "step": 112775 + }, + { + "epoch": 7.66272591384699, + "grad_norm": 0.9194507002830505, + "learning_rate": 4.215926076912624e-05, + "loss": 3.2847, + "step": 112780 + }, + { + "epoch": 7.663065633917652, + "grad_norm": 0.8478006720542908, + "learning_rate": 4.2116795760293524e-05, + "loss": 3.2472, + "step": 112785 + }, + { + "epoch": 7.663405353988313, + "grad_norm": 0.8304161429405212, + "learning_rate": 4.20743307514608e-05, + "loss": 3.3509, + "step": 112790 + }, + { + "epoch": 7.663745074058975, + "grad_norm": 0.9024617075920105, + "learning_rate": 4.203186574262807e-05, + "loss": 3.4375, + "step": 112795 + }, + { + "epoch": 7.6640847941296375, + "grad_norm": 0.831393837928772, + "learning_rate": 4.198940073379536e-05, + "loss": 3.4697, + "step": 112800 + }, + { + "epoch": 7.664424514200299, + "grad_norm": 0.909951388835907, + "learning_rate": 4.194693572496263e-05, + "loss": 3.5808, + "step": 112805 + }, + { + "epoch": 7.664764234270961, + "grad_norm": 1.0104529857635498, + "learning_rate": 4.190447071612991e-05, + "loss": 3.1735, + "step": 112810 + }, + { + "epoch": 7.665103954341623, + "grad_norm": 1.2135754823684692, + "learning_rate": 4.186200570729719e-05, + "loss": 3.4233, + "step": 112815 + }, + { + "epoch": 7.665443674412284, + "grad_norm": 0.9478224515914917, + "learning_rate": 4.1819540698464464e-05, + "loss": 3.3623, + "step": 112820 + }, + { + "epoch": 7.665783394482946, + "grad_norm": 0.8355873227119446, + "learning_rate": 4.1777075689631744e-05, + "loss": 3.3108, + "step": 112825 + }, + { + "epoch": 7.666123114553608, + "grad_norm": 0.8958069086074829, + "learning_rate": 4.1734610680799024e-05, + "loss": 3.062, + "step": 112830 + }, + { + "epoch": 7.666462834624269, + "grad_norm": 1.049147129058838, + "learning_rate": 4.1692145671966304e-05, + "loss": 3.5068, + "step": 112835 + }, + { + "epoch": 7.666802554694931, + "grad_norm": 0.9330539703369141, + "learning_rate": 4.164968066313358e-05, + "loss": 3.4726, + "step": 112840 + }, + { + "epoch": 7.6671422747655935, + "grad_norm": 0.9201052784919739, + "learning_rate": 4.160721565430086e-05, + "loss": 3.3784, + "step": 112845 + }, + { + "epoch": 7.667481994836255, + "grad_norm": 0.9034565687179565, + "learning_rate": 4.156475064546814e-05, + "loss": 3.255, + "step": 112850 + }, + { + "epoch": 7.667821714906917, + "grad_norm": 0.897429347038269, + "learning_rate": 4.152228563663541e-05, + "loss": 3.0533, + "step": 112855 + }, + { + "epoch": 7.668161434977579, + "grad_norm": 1.1917376518249512, + "learning_rate": 4.147982062780269e-05, + "loss": 3.42, + "step": 112860 + }, + { + "epoch": 7.66850115504824, + "grad_norm": 1.137567400932312, + "learning_rate": 4.143735561896997e-05, + "loss": 3.4709, + "step": 112865 + }, + { + "epoch": 7.668840875118902, + "grad_norm": 1.2040715217590332, + "learning_rate": 4.139489061013725e-05, + "loss": 3.269, + "step": 112870 + }, + { + "epoch": 7.669180595189564, + "grad_norm": 0.975372314453125, + "learning_rate": 4.1352425601304524e-05, + "loss": 3.4667, + "step": 112875 + }, + { + "epoch": 7.669520315260225, + "grad_norm": 1.0705811977386475, + "learning_rate": 4.1309960592471804e-05, + "loss": 3.328, + "step": 112880 + }, + { + "epoch": 7.6698600353308874, + "grad_norm": 1.375104308128357, + "learning_rate": 4.1267495583639084e-05, + "loss": 3.4625, + "step": 112885 + }, + { + "epoch": 7.6701997554015495, + "grad_norm": 0.8962164521217346, + "learning_rate": 4.122503057480636e-05, + "loss": 3.5149, + "step": 112890 + }, + { + "epoch": 7.670539475472211, + "grad_norm": 1.028198003768921, + "learning_rate": 4.118256556597364e-05, + "loss": 3.5048, + "step": 112895 + }, + { + "epoch": 7.670879195542873, + "grad_norm": 0.8083811402320862, + "learning_rate": 4.114010055714092e-05, + "loss": 3.3865, + "step": 112900 + }, + { + "epoch": 7.671218915613535, + "grad_norm": 0.8612976670265198, + "learning_rate": 4.109763554830819e-05, + "loss": 3.0398, + "step": 112905 + }, + { + "epoch": 7.671558635684196, + "grad_norm": 0.9592646360397339, + "learning_rate": 4.105517053947547e-05, + "loss": 3.2481, + "step": 112910 + }, + { + "epoch": 7.671898355754858, + "grad_norm": 0.9222086071968079, + "learning_rate": 4.101270553064275e-05, + "loss": 3.5538, + "step": 112915 + }, + { + "epoch": 7.67223807582552, + "grad_norm": 0.9122011065483093, + "learning_rate": 4.097024052181003e-05, + "loss": 3.5829, + "step": 112920 + }, + { + "epoch": 7.672577795896181, + "grad_norm": 0.96656733751297, + "learning_rate": 4.0927775512977305e-05, + "loss": 3.4555, + "step": 112925 + }, + { + "epoch": 7.6729175159668435, + "grad_norm": 0.9290505051612854, + "learning_rate": 4.088531050414459e-05, + "loss": 3.3994, + "step": 112930 + }, + { + "epoch": 7.6732572360375055, + "grad_norm": 1.0312902927398682, + "learning_rate": 4.0842845495311865e-05, + "loss": 3.2223, + "step": 112935 + }, + { + "epoch": 7.673596956108167, + "grad_norm": 3.9769961833953857, + "learning_rate": 4.080038048647914e-05, + "loss": 3.2783, + "step": 112940 + }, + { + "epoch": 7.673936676178829, + "grad_norm": 0.865813136100769, + "learning_rate": 4.0757915477646425e-05, + "loss": 3.0834, + "step": 112945 + }, + { + "epoch": 7.674276396249491, + "grad_norm": 0.8614926338195801, + "learning_rate": 4.07154504688137e-05, + "loss": 3.5268, + "step": 112950 + }, + { + "epoch": 7.674616116320152, + "grad_norm": 1.27047860622406, + "learning_rate": 4.067298545998098e-05, + "loss": 3.4623, + "step": 112955 + }, + { + "epoch": 7.674955836390814, + "grad_norm": 1.2663475275039673, + "learning_rate": 4.063052045114825e-05, + "loss": 3.3181, + "step": 112960 + }, + { + "epoch": 7.675295556461476, + "grad_norm": 0.9340328574180603, + "learning_rate": 4.058805544231553e-05, + "loss": 3.2778, + "step": 112965 + }, + { + "epoch": 7.675635276532137, + "grad_norm": 1.139778733253479, + "learning_rate": 4.054559043348281e-05, + "loss": 3.2994, + "step": 112970 + }, + { + "epoch": 7.6759749966027995, + "grad_norm": 0.9346972703933716, + "learning_rate": 4.0503125424650085e-05, + "loss": 3.184, + "step": 112975 + }, + { + "epoch": 7.6763147166734615, + "grad_norm": 1.0922173261642456, + "learning_rate": 4.046066041581737e-05, + "loss": 3.4643, + "step": 112980 + }, + { + "epoch": 7.676654436744123, + "grad_norm": 1.0221103429794312, + "learning_rate": 4.0418195406984645e-05, + "loss": 3.2509, + "step": 112985 + }, + { + "epoch": 7.676994156814785, + "grad_norm": 1.007397174835205, + "learning_rate": 4.037573039815192e-05, + "loss": 3.309, + "step": 112990 + }, + { + "epoch": 7.677333876885447, + "grad_norm": 0.9656966328620911, + "learning_rate": 4.0333265389319205e-05, + "loss": 3.4763, + "step": 112995 + }, + { + "epoch": 7.677673596956108, + "grad_norm": 0.8479506969451904, + "learning_rate": 4.029080038048648e-05, + "loss": 3.2221, + "step": 113000 + }, + { + "epoch": 7.67801331702677, + "grad_norm": 0.940873384475708, + "learning_rate": 4.024833537165376e-05, + "loss": 3.477, + "step": 113005 + }, + { + "epoch": 7.678353037097431, + "grad_norm": 2.517364978790283, + "learning_rate": 4.020587036282104e-05, + "loss": 3.2836, + "step": 113010 + }, + { + "epoch": 7.678692757168093, + "grad_norm": 0.8576843738555908, + "learning_rate": 4.016340535398832e-05, + "loss": 3.5285, + "step": 113015 + }, + { + "epoch": 7.6790324772387555, + "grad_norm": 0.934738039970398, + "learning_rate": 4.012094034515559e-05, + "loss": 3.3235, + "step": 113020 + }, + { + "epoch": 7.679372197309417, + "grad_norm": 0.9214470982551575, + "learning_rate": 4.0078475336322866e-05, + "loss": 3.2819, + "step": 113025 + }, + { + "epoch": 7.679711917380079, + "grad_norm": 0.9106442332267761, + "learning_rate": 4.003601032749015e-05, + "loss": 3.2932, + "step": 113030 + }, + { + "epoch": 7.680051637450741, + "grad_norm": 1.2295221090316772, + "learning_rate": 3.9993545318657426e-05, + "loss": 3.4559, + "step": 113035 + }, + { + "epoch": 7.680391357521402, + "grad_norm": 0.9699468612670898, + "learning_rate": 3.9951080309824706e-05, + "loss": 3.464, + "step": 113040 + }, + { + "epoch": 7.680731077592064, + "grad_norm": 0.9912946820259094, + "learning_rate": 3.9908615300991986e-05, + "loss": 3.273, + "step": 113045 + }, + { + "epoch": 7.681070797662726, + "grad_norm": 0.9219380617141724, + "learning_rate": 3.986615029215926e-05, + "loss": 3.3508, + "step": 113050 + }, + { + "epoch": 7.681410517733387, + "grad_norm": 0.8667637705802917, + "learning_rate": 3.982368528332654e-05, + "loss": 3.4197, + "step": 113055 + }, + { + "epoch": 7.681750237804049, + "grad_norm": 1.0289479494094849, + "learning_rate": 3.978122027449382e-05, + "loss": 3.494, + "step": 113060 + }, + { + "epoch": 7.6820899578747115, + "grad_norm": 0.8048513531684875, + "learning_rate": 3.97387552656611e-05, + "loss": 3.4229, + "step": 113065 + }, + { + "epoch": 7.682429677945373, + "grad_norm": 0.973792314529419, + "learning_rate": 3.969629025682837e-05, + "loss": 3.6065, + "step": 113070 + }, + { + "epoch": 7.682769398016035, + "grad_norm": 0.8827299475669861, + "learning_rate": 3.9653825247995646e-05, + "loss": 3.3814, + "step": 113075 + }, + { + "epoch": 7.683109118086697, + "grad_norm": 0.9804157018661499, + "learning_rate": 3.961136023916293e-05, + "loss": 3.3267, + "step": 113080 + }, + { + "epoch": 7.683448838157358, + "grad_norm": 0.8970969915390015, + "learning_rate": 3.9568895230330206e-05, + "loss": 3.6, + "step": 113085 + }, + { + "epoch": 7.68378855822802, + "grad_norm": 0.9736630320549011, + "learning_rate": 3.9526430221497486e-05, + "loss": 3.2885, + "step": 113090 + }, + { + "epoch": 7.684128278298682, + "grad_norm": 0.9099250435829163, + "learning_rate": 3.9483965212664766e-05, + "loss": 3.427, + "step": 113095 + }, + { + "epoch": 7.684467998369343, + "grad_norm": 1.0332337617874146, + "learning_rate": 3.9441500203832046e-05, + "loss": 3.1831, + "step": 113100 + }, + { + "epoch": 7.684807718440005, + "grad_norm": 0.8867408633232117, + "learning_rate": 3.939903519499932e-05, + "loss": 3.3367, + "step": 113105 + }, + { + "epoch": 7.6851474385106675, + "grad_norm": 1.0113941431045532, + "learning_rate": 3.93565701861666e-05, + "loss": 3.2707, + "step": 113110 + }, + { + "epoch": 7.685487158581329, + "grad_norm": 1.162502408027649, + "learning_rate": 3.931410517733388e-05, + "loss": 3.3804, + "step": 113115 + }, + { + "epoch": 7.685826878651991, + "grad_norm": 0.9091602563858032, + "learning_rate": 3.927164016850115e-05, + "loss": 3.2829, + "step": 113120 + }, + { + "epoch": 7.686166598722653, + "grad_norm": 1.1228057146072388, + "learning_rate": 3.922917515966844e-05, + "loss": 3.3278, + "step": 113125 + }, + { + "epoch": 7.686506318793314, + "grad_norm": 0.9915242791175842, + "learning_rate": 3.918671015083571e-05, + "loss": 3.2563, + "step": 113130 + }, + { + "epoch": 7.686846038863976, + "grad_norm": 1.2443594932556152, + "learning_rate": 3.914424514200299e-05, + "loss": 3.0283, + "step": 113135 + }, + { + "epoch": 7.687185758934638, + "grad_norm": 0.8993411064147949, + "learning_rate": 3.910178013317027e-05, + "loss": 3.4551, + "step": 113140 + }, + { + "epoch": 7.687525479005299, + "grad_norm": 0.7880287766456604, + "learning_rate": 3.905931512433755e-05, + "loss": 3.2647, + "step": 113145 + }, + { + "epoch": 7.687865199075961, + "grad_norm": 1.233643889427185, + "learning_rate": 3.901685011550483e-05, + "loss": 3.5276, + "step": 113150 + }, + { + "epoch": 7.6882049191466235, + "grad_norm": 0.8512731790542603, + "learning_rate": 3.89743851066721e-05, + "loss": 3.5765, + "step": 113155 + }, + { + "epoch": 7.688544639217285, + "grad_norm": 1.081068992614746, + "learning_rate": 3.893192009783938e-05, + "loss": 3.3151, + "step": 113160 + }, + { + "epoch": 7.688884359287947, + "grad_norm": 0.9587036371231079, + "learning_rate": 3.888945508900666e-05, + "loss": 3.2992, + "step": 113165 + }, + { + "epoch": 7.689224079358609, + "grad_norm": 1.0871073007583618, + "learning_rate": 3.8846990080173934e-05, + "loss": 3.3335, + "step": 113170 + }, + { + "epoch": 7.68956379942927, + "grad_norm": 0.8617308735847473, + "learning_rate": 3.880452507134122e-05, + "loss": 3.5581, + "step": 113175 + }, + { + "epoch": 7.689903519499932, + "grad_norm": 0.8178585171699524, + "learning_rate": 3.8762060062508494e-05, + "loss": 3.6009, + "step": 113180 + }, + { + "epoch": 7.690243239570594, + "grad_norm": 0.9507669806480408, + "learning_rate": 3.871959505367577e-05, + "loss": 3.3052, + "step": 113185 + }, + { + "epoch": 7.690582959641255, + "grad_norm": 0.8862653970718384, + "learning_rate": 3.8677130044843054e-05, + "loss": 3.4436, + "step": 113190 + }, + { + "epoch": 7.6909226797119175, + "grad_norm": 0.9875726699829102, + "learning_rate": 3.863466503601033e-05, + "loss": 3.2466, + "step": 113195 + }, + { + "epoch": 7.6912623997825795, + "grad_norm": 0.7971572875976562, + "learning_rate": 3.859220002717761e-05, + "loss": 3.0694, + "step": 113200 + }, + { + "epoch": 7.691602119853241, + "grad_norm": 0.9528030157089233, + "learning_rate": 3.854973501834488e-05, + "loss": 3.2317, + "step": 113205 + }, + { + "epoch": 7.691941839923903, + "grad_norm": 1.0599384307861328, + "learning_rate": 3.850727000951217e-05, + "loss": 3.3592, + "step": 113210 + }, + { + "epoch": 7.692281559994565, + "grad_norm": 0.799289345741272, + "learning_rate": 3.846480500067944e-05, + "loss": 3.3407, + "step": 113215 + }, + { + "epoch": 7.692621280065226, + "grad_norm": 0.9994924068450928, + "learning_rate": 3.8422339991846714e-05, + "loss": 3.569, + "step": 113220 + }, + { + "epoch": 7.692961000135888, + "grad_norm": 1.02475106716156, + "learning_rate": 3.8379874983014e-05, + "loss": 3.4141, + "step": 113225 + }, + { + "epoch": 7.693300720206549, + "grad_norm": 0.8844285607337952, + "learning_rate": 3.8337409974181274e-05, + "loss": 3.2993, + "step": 113230 + }, + { + "epoch": 7.693640440277211, + "grad_norm": 0.9268152713775635, + "learning_rate": 3.8294944965348554e-05, + "loss": 3.2696, + "step": 113235 + }, + { + "epoch": 7.6939801603478735, + "grad_norm": 0.9977164268493652, + "learning_rate": 3.8252479956515834e-05, + "loss": 3.6025, + "step": 113240 + }, + { + "epoch": 7.694319880418535, + "grad_norm": 1.1638150215148926, + "learning_rate": 3.821001494768311e-05, + "loss": 3.5264, + "step": 113245 + }, + { + "epoch": 7.694659600489197, + "grad_norm": 0.9683764576911926, + "learning_rate": 3.816754993885039e-05, + "loss": 3.2012, + "step": 113250 + }, + { + "epoch": 7.694999320559859, + "grad_norm": 0.9531885385513306, + "learning_rate": 3.812508493001766e-05, + "loss": 3.2437, + "step": 113255 + }, + { + "epoch": 7.69533904063052, + "grad_norm": 0.9002404808998108, + "learning_rate": 3.808261992118495e-05, + "loss": 3.2928, + "step": 113260 + }, + { + "epoch": 7.695678760701182, + "grad_norm": 1.0933682918548584, + "learning_rate": 3.804015491235222e-05, + "loss": 3.2359, + "step": 113265 + }, + { + "epoch": 7.696018480771844, + "grad_norm": 1.1104542016983032, + "learning_rate": 3.7997689903519495e-05, + "loss": 3.384, + "step": 113270 + }, + { + "epoch": 7.696358200842505, + "grad_norm": 0.9055066108703613, + "learning_rate": 3.795522489468678e-05, + "loss": 3.5815, + "step": 113275 + }, + { + "epoch": 7.696697920913167, + "grad_norm": 1.0558894872665405, + "learning_rate": 3.7912759885854055e-05, + "loss": 3.5125, + "step": 113280 + }, + { + "epoch": 7.6970376409838295, + "grad_norm": 0.9281837940216064, + "learning_rate": 3.7870294877021335e-05, + "loss": 3.3563, + "step": 113285 + }, + { + "epoch": 7.697377361054491, + "grad_norm": 0.9928522109985352, + "learning_rate": 3.7827829868188615e-05, + "loss": 3.3488, + "step": 113290 + }, + { + "epoch": 7.697717081125153, + "grad_norm": 0.9897965788841248, + "learning_rate": 3.7785364859355895e-05, + "loss": 3.3082, + "step": 113295 + }, + { + "epoch": 7.698056801195815, + "grad_norm": 0.841359555721283, + "learning_rate": 3.774289985052317e-05, + "loss": 3.2055, + "step": 113300 + }, + { + "epoch": 7.698396521266476, + "grad_norm": 0.9711959958076477, + "learning_rate": 3.770043484169045e-05, + "loss": 3.4149, + "step": 113305 + }, + { + "epoch": 7.698736241337138, + "grad_norm": 1.1004890203475952, + "learning_rate": 3.765796983285773e-05, + "loss": 3.3858, + "step": 113310 + }, + { + "epoch": 7.6990759614078, + "grad_norm": 0.9776637554168701, + "learning_rate": 3.7615504824025e-05, + "loss": 3.4077, + "step": 113315 + }, + { + "epoch": 7.699415681478461, + "grad_norm": 1.1359803676605225, + "learning_rate": 3.757303981519228e-05, + "loss": 3.3488, + "step": 113320 + }, + { + "epoch": 7.699755401549123, + "grad_norm": 0.8339403867721558, + "learning_rate": 3.753057480635956e-05, + "loss": 3.4668, + "step": 113325 + }, + { + "epoch": 7.7000951216197855, + "grad_norm": 1.0698423385620117, + "learning_rate": 3.7488109797526835e-05, + "loss": 3.351, + "step": 113330 + }, + { + "epoch": 7.700434841690447, + "grad_norm": 0.972842812538147, + "learning_rate": 3.7445644788694115e-05, + "loss": 3.4135, + "step": 113335 + }, + { + "epoch": 7.700774561761109, + "grad_norm": 0.9648342132568359, + "learning_rate": 3.7403179779861395e-05, + "loss": 3.2978, + "step": 113340 + }, + { + "epoch": 7.701114281831771, + "grad_norm": 1.4392061233520508, + "learning_rate": 3.7360714771028675e-05, + "loss": 3.3899, + "step": 113345 + }, + { + "epoch": 7.701454001902432, + "grad_norm": 1.353940725326538, + "learning_rate": 3.731824976219595e-05, + "loss": 3.5316, + "step": 113350 + }, + { + "epoch": 7.701793721973094, + "grad_norm": 0.9321033954620361, + "learning_rate": 3.727578475336323e-05, + "loss": 3.5741, + "step": 113355 + }, + { + "epoch": 7.702133442043756, + "grad_norm": 0.8545920848846436, + "learning_rate": 3.723331974453051e-05, + "loss": 3.3234, + "step": 113360 + }, + { + "epoch": 7.702473162114417, + "grad_norm": 1.0665922164916992, + "learning_rate": 3.719085473569778e-05, + "loss": 2.9656, + "step": 113365 + }, + { + "epoch": 7.702812882185079, + "grad_norm": 1.1007592678070068, + "learning_rate": 3.714838972686507e-05, + "loss": 3.5569, + "step": 113370 + }, + { + "epoch": 7.7031526022557415, + "grad_norm": 0.9060133099555969, + "learning_rate": 3.710592471803234e-05, + "loss": 3.3604, + "step": 113375 + }, + { + "epoch": 7.703492322326403, + "grad_norm": 0.8918707966804504, + "learning_rate": 3.706345970919962e-05, + "loss": 3.4635, + "step": 113380 + }, + { + "epoch": 7.703832042397065, + "grad_norm": 0.8818008303642273, + "learning_rate": 3.7020994700366896e-05, + "loss": 3.6852, + "step": 113385 + }, + { + "epoch": 7.704171762467727, + "grad_norm": 0.907250165939331, + "learning_rate": 3.6978529691534176e-05, + "loss": 3.6949, + "step": 113390 + }, + { + "epoch": 7.704511482538388, + "grad_norm": 0.9445988535881042, + "learning_rate": 3.6936064682701456e-05, + "loss": 3.2876, + "step": 113395 + }, + { + "epoch": 7.70485120260905, + "grad_norm": 0.9742817282676697, + "learning_rate": 3.689359967386873e-05, + "loss": 3.361, + "step": 113400 + }, + { + "epoch": 7.705190922679712, + "grad_norm": 0.8917263746261597, + "learning_rate": 3.6851134665036016e-05, + "loss": 3.5154, + "step": 113405 + }, + { + "epoch": 7.705530642750373, + "grad_norm": 1.0964200496673584, + "learning_rate": 3.680866965620329e-05, + "loss": 3.5509, + "step": 113410 + }, + { + "epoch": 7.705870362821035, + "grad_norm": 0.9949285387992859, + "learning_rate": 3.676620464737056e-05, + "loss": 3.2082, + "step": 113415 + }, + { + "epoch": 7.7062100828916975, + "grad_norm": 0.740581750869751, + "learning_rate": 3.672373963853785e-05, + "loss": 3.265, + "step": 113420 + }, + { + "epoch": 7.706549802962359, + "grad_norm": 1.098183512687683, + "learning_rate": 3.668127462970512e-05, + "loss": 3.4009, + "step": 113425 + }, + { + "epoch": 7.706889523033021, + "grad_norm": 0.9968330264091492, + "learning_rate": 3.66388096208724e-05, + "loss": 3.1923, + "step": 113430 + }, + { + "epoch": 7.707229243103683, + "grad_norm": 0.8729807138442993, + "learning_rate": 3.659634461203968e-05, + "loss": 3.2898, + "step": 113435 + }, + { + "epoch": 7.707568963174344, + "grad_norm": 1.2368438243865967, + "learning_rate": 3.6553879603206956e-05, + "loss": 3.1457, + "step": 113440 + }, + { + "epoch": 7.707908683245006, + "grad_norm": 1.048829436302185, + "learning_rate": 3.6511414594374236e-05, + "loss": 3.3566, + "step": 113445 + }, + { + "epoch": 7.708248403315668, + "grad_norm": 0.7585592865943909, + "learning_rate": 3.646894958554151e-05, + "loss": 3.1953, + "step": 113450 + }, + { + "epoch": 7.708588123386329, + "grad_norm": 0.8734019994735718, + "learning_rate": 3.6426484576708796e-05, + "loss": 3.2249, + "step": 113455 + }, + { + "epoch": 7.7089278434569914, + "grad_norm": 0.975897490978241, + "learning_rate": 3.638401956787607e-05, + "loss": 3.3908, + "step": 113460 + }, + { + "epoch": 7.7092675635276535, + "grad_norm": 0.9720220565795898, + "learning_rate": 3.634155455904335e-05, + "loss": 3.322, + "step": 113465 + }, + { + "epoch": 7.709607283598315, + "grad_norm": 1.1024682521820068, + "learning_rate": 3.629908955021063e-05, + "loss": 3.4784, + "step": 113470 + }, + { + "epoch": 7.709947003668977, + "grad_norm": 1.422379732131958, + "learning_rate": 3.62566245413779e-05, + "loss": 3.4087, + "step": 113475 + }, + { + "epoch": 7.710286723739639, + "grad_norm": 0.9648181796073914, + "learning_rate": 3.621415953254518e-05, + "loss": 3.3466, + "step": 113480 + }, + { + "epoch": 7.7106264438103, + "grad_norm": 0.949038565158844, + "learning_rate": 3.617169452371246e-05, + "loss": 3.3757, + "step": 113485 + }, + { + "epoch": 7.710966163880962, + "grad_norm": 0.8850806355476379, + "learning_rate": 3.6129229514879743e-05, + "loss": 3.3518, + "step": 113490 + }, + { + "epoch": 7.711305883951624, + "grad_norm": 0.7472370266914368, + "learning_rate": 3.608676450604702e-05, + "loss": 3.4433, + "step": 113495 + }, + { + "epoch": 7.711645604022285, + "grad_norm": 1.8258973360061646, + "learning_rate": 3.604429949721429e-05, + "loss": 3.4244, + "step": 113500 + }, + { + "epoch": 7.7119853240929475, + "grad_norm": 0.8307035565376282, + "learning_rate": 3.600183448838158e-05, + "loss": 3.6467, + "step": 113505 + }, + { + "epoch": 7.7123250441636095, + "grad_norm": 0.8768848776817322, + "learning_rate": 3.595936947954885e-05, + "loss": 3.4304, + "step": 113510 + }, + { + "epoch": 7.712664764234271, + "grad_norm": 0.8644377589225769, + "learning_rate": 3.591690447071613e-05, + "loss": 3.3561, + "step": 113515 + }, + { + "epoch": 7.713004484304933, + "grad_norm": 1.017575740814209, + "learning_rate": 3.587443946188341e-05, + "loss": 3.2337, + "step": 113520 + }, + { + "epoch": 7.713344204375595, + "grad_norm": 0.9515251517295837, + "learning_rate": 3.5831974453050684e-05, + "loss": 3.3394, + "step": 113525 + }, + { + "epoch": 7.713683924446256, + "grad_norm": 0.943169891834259, + "learning_rate": 3.5789509444217964e-05, + "loss": 3.4024, + "step": 113530 + }, + { + "epoch": 7.714023644516918, + "grad_norm": 0.9988165497779846, + "learning_rate": 3.5747044435385244e-05, + "loss": 3.2431, + "step": 113535 + }, + { + "epoch": 7.71436336458758, + "grad_norm": 0.9356712698936462, + "learning_rate": 3.5704579426552524e-05, + "loss": 3.1875, + "step": 113540 + }, + { + "epoch": 7.714703084658241, + "grad_norm": 1.0942528247833252, + "learning_rate": 3.56621144177198e-05, + "loss": 3.0644, + "step": 113545 + }, + { + "epoch": 7.7150428047289035, + "grad_norm": 1.0685149431228638, + "learning_rate": 3.5619649408887084e-05, + "loss": 3.3855, + "step": 113550 + }, + { + "epoch": 7.7153825247995655, + "grad_norm": 1.0907410383224487, + "learning_rate": 3.557718440005436e-05, + "loss": 3.2517, + "step": 113555 + }, + { + "epoch": 7.715722244870227, + "grad_norm": 0.9323757886886597, + "learning_rate": 3.553471939122163e-05, + "loss": 3.3046, + "step": 113560 + }, + { + "epoch": 7.716061964940889, + "grad_norm": 0.784040093421936, + "learning_rate": 3.549225438238891e-05, + "loss": 3.1813, + "step": 113565 + }, + { + "epoch": 7.716401685011551, + "grad_norm": 0.9919095039367676, + "learning_rate": 3.544978937355619e-05, + "loss": 3.2207, + "step": 113570 + }, + { + "epoch": 7.716741405082212, + "grad_norm": 0.8731863498687744, + "learning_rate": 3.540732436472347e-05, + "loss": 3.1849, + "step": 113575 + }, + { + "epoch": 7.717081125152874, + "grad_norm": 1.1570897102355957, + "learning_rate": 3.5364859355890744e-05, + "loss": 3.3035, + "step": 113580 + }, + { + "epoch": 7.717420845223536, + "grad_norm": 1.2216670513153076, + "learning_rate": 3.5322394347058024e-05, + "loss": 3.342, + "step": 113585 + }, + { + "epoch": 7.717760565294197, + "grad_norm": 1.1029223203659058, + "learning_rate": 3.5279929338225304e-05, + "loss": 3.2201, + "step": 113590 + }, + { + "epoch": 7.7181002853648595, + "grad_norm": 1.1149717569351196, + "learning_rate": 3.523746432939258e-05, + "loss": 3.166, + "step": 113595 + }, + { + "epoch": 7.7184400054355216, + "grad_norm": 1.0737558603286743, + "learning_rate": 3.5194999320559864e-05, + "loss": 2.9622, + "step": 113600 + }, + { + "epoch": 7.718779725506183, + "grad_norm": 0.8682258725166321, + "learning_rate": 3.515253431172714e-05, + "loss": 3.2916, + "step": 113605 + }, + { + "epoch": 7.719119445576845, + "grad_norm": 1.1763073205947876, + "learning_rate": 3.511006930289441e-05, + "loss": 3.3409, + "step": 113610 + }, + { + "epoch": 7.719459165647507, + "grad_norm": 0.8123204112052917, + "learning_rate": 3.50676042940617e-05, + "loss": 3.1804, + "step": 113615 + }, + { + "epoch": 7.719798885718168, + "grad_norm": 0.906078577041626, + "learning_rate": 3.502513928522897e-05, + "loss": 3.4301, + "step": 113620 + }, + { + "epoch": 7.72013860578883, + "grad_norm": 0.9765132069587708, + "learning_rate": 3.498267427639625e-05, + "loss": 3.3074, + "step": 113625 + }, + { + "epoch": 7.720478325859492, + "grad_norm": 0.7404305338859558, + "learning_rate": 3.4940209267563525e-05, + "loss": 3.4971, + "step": 113630 + }, + { + "epoch": 7.720818045930153, + "grad_norm": 1.0825858116149902, + "learning_rate": 3.489774425873081e-05, + "loss": 3.3856, + "step": 113635 + }, + { + "epoch": 7.7211577660008155, + "grad_norm": 0.8596729636192322, + "learning_rate": 3.4855279249898085e-05, + "loss": 3.2872, + "step": 113640 + }, + { + "epoch": 7.721497486071478, + "grad_norm": 1.1652276515960693, + "learning_rate": 3.481281424106536e-05, + "loss": 3.3576, + "step": 113645 + }, + { + "epoch": 7.721837206142139, + "grad_norm": 1.0913307666778564, + "learning_rate": 3.4770349232232645e-05, + "loss": 3.3585, + "step": 113650 + }, + { + "epoch": 7.722176926212801, + "grad_norm": 0.8490902781486511, + "learning_rate": 3.472788422339992e-05, + "loss": 3.3367, + "step": 113655 + }, + { + "epoch": 7.722516646283463, + "grad_norm": 1.0220253467559814, + "learning_rate": 3.46854192145672e-05, + "loss": 3.4987, + "step": 113660 + }, + { + "epoch": 7.722856366354124, + "grad_norm": 1.2331782579421997, + "learning_rate": 3.464295420573448e-05, + "loss": 3.3085, + "step": 113665 + }, + { + "epoch": 7.723196086424786, + "grad_norm": 0.9492802023887634, + "learning_rate": 3.460048919690175e-05, + "loss": 3.2797, + "step": 113670 + }, + { + "epoch": 7.723535806495448, + "grad_norm": 1.0869511365890503, + "learning_rate": 3.455802418806903e-05, + "loss": 3.6133, + "step": 113675 + }, + { + "epoch": 7.723875526566109, + "grad_norm": 1.0648388862609863, + "learning_rate": 3.4515559179236305e-05, + "loss": 3.5645, + "step": 113680 + }, + { + "epoch": 7.7242152466367715, + "grad_norm": 0.9240715503692627, + "learning_rate": 3.447309417040359e-05, + "loss": 3.5719, + "step": 113685 + }, + { + "epoch": 7.724554966707433, + "grad_norm": 1.0042150020599365, + "learning_rate": 3.4430629161570865e-05, + "loss": 3.3822, + "step": 113690 + }, + { + "epoch": 7.724894686778095, + "grad_norm": 1.0342767238616943, + "learning_rate": 3.438816415273814e-05, + "loss": 3.4746, + "step": 113695 + }, + { + "epoch": 7.725234406848757, + "grad_norm": 0.8841719031333923, + "learning_rate": 3.4345699143905425e-05, + "loss": 3.302, + "step": 113700 + }, + { + "epoch": 7.725574126919418, + "grad_norm": 0.9574698805809021, + "learning_rate": 3.43032341350727e-05, + "loss": 3.5742, + "step": 113705 + }, + { + "epoch": 7.72591384699008, + "grad_norm": 1.1598056554794312, + "learning_rate": 3.426076912623998e-05, + "loss": 3.2764, + "step": 113710 + }, + { + "epoch": 7.726253567060742, + "grad_norm": 0.8311142325401306, + "learning_rate": 3.421830411740726e-05, + "loss": 3.4455, + "step": 113715 + }, + { + "epoch": 7.726593287131403, + "grad_norm": 0.8907772302627563, + "learning_rate": 3.417583910857454e-05, + "loss": 3.3839, + "step": 113720 + }, + { + "epoch": 7.726933007202065, + "grad_norm": 1.0370668172836304, + "learning_rate": 3.413337409974181e-05, + "loss": 3.2644, + "step": 113725 + }, + { + "epoch": 7.7272727272727275, + "grad_norm": 1.1003183126449585, + "learning_rate": 3.409090909090909e-05, + "loss": 3.4496, + "step": 113730 + }, + { + "epoch": 7.727612447343389, + "grad_norm": 1.0406689643859863, + "learning_rate": 3.404844408207637e-05, + "loss": 3.496, + "step": 113735 + }, + { + "epoch": 7.727952167414051, + "grad_norm": 0.7385467886924744, + "learning_rate": 3.4005979073243646e-05, + "loss": 3.5698, + "step": 113740 + }, + { + "epoch": 7.728291887484713, + "grad_norm": 0.9130847454071045, + "learning_rate": 3.3963514064410926e-05, + "loss": 3.5051, + "step": 113745 + }, + { + "epoch": 7.728631607555374, + "grad_norm": 1.162708044052124, + "learning_rate": 3.3921049055578206e-05, + "loss": 3.3036, + "step": 113750 + }, + { + "epoch": 7.728971327626036, + "grad_norm": 0.9672364592552185, + "learning_rate": 3.387858404674548e-05, + "loss": 3.3128, + "step": 113755 + }, + { + "epoch": 7.729311047696698, + "grad_norm": 1.1194792985916138, + "learning_rate": 3.383611903791276e-05, + "loss": 3.333, + "step": 113760 + }, + { + "epoch": 7.729650767767359, + "grad_norm": 0.9060764908790588, + "learning_rate": 3.379365402908004e-05, + "loss": 3.5796, + "step": 113765 + }, + { + "epoch": 7.7299904878380215, + "grad_norm": 1.0373371839523315, + "learning_rate": 3.375118902024732e-05, + "loss": 3.3717, + "step": 113770 + }, + { + "epoch": 7.7303302079086835, + "grad_norm": 1.0068814754486084, + "learning_rate": 3.370872401141459e-05, + "loss": 3.3177, + "step": 113775 + }, + { + "epoch": 7.730669927979345, + "grad_norm": 0.8220153450965881, + "learning_rate": 3.366625900258187e-05, + "loss": 3.394, + "step": 113780 + }, + { + "epoch": 7.731009648050007, + "grad_norm": 0.96247398853302, + "learning_rate": 3.362379399374915e-05, + "loss": 3.3104, + "step": 113785 + }, + { + "epoch": 7.731349368120669, + "grad_norm": 0.7842761874198914, + "learning_rate": 3.3581328984916426e-05, + "loss": 3.3254, + "step": 113790 + }, + { + "epoch": 7.73168908819133, + "grad_norm": 0.922290027141571, + "learning_rate": 3.353886397608371e-05, + "loss": 3.2703, + "step": 113795 + }, + { + "epoch": 7.732028808261992, + "grad_norm": 0.9200266599655151, + "learning_rate": 3.3496398967250986e-05, + "loss": 3.4782, + "step": 113800 + }, + { + "epoch": 7.732368528332654, + "grad_norm": 0.9380212426185608, + "learning_rate": 3.3453933958418266e-05, + "loss": 3.4808, + "step": 113805 + }, + { + "epoch": 7.732708248403315, + "grad_norm": 0.9445865750312805, + "learning_rate": 3.341146894958554e-05, + "loss": 2.9813, + "step": 113810 + }, + { + "epoch": 7.7330479684739775, + "grad_norm": 0.832582950592041, + "learning_rate": 3.336900394075282e-05, + "loss": 3.4513, + "step": 113815 + }, + { + "epoch": 7.7333876885446395, + "grad_norm": 1.3271418809890747, + "learning_rate": 3.33265389319201e-05, + "loss": 3.3803, + "step": 113820 + }, + { + "epoch": 7.733727408615301, + "grad_norm": 0.7975448369979858, + "learning_rate": 3.328407392308737e-05, + "loss": 3.484, + "step": 113825 + }, + { + "epoch": 7.734067128685963, + "grad_norm": 0.8881714344024658, + "learning_rate": 3.324160891425466e-05, + "loss": 3.2764, + "step": 113830 + }, + { + "epoch": 7.734406848756625, + "grad_norm": 1.2146074771881104, + "learning_rate": 3.319914390542193e-05, + "loss": 3.8335, + "step": 113835 + }, + { + "epoch": 7.734746568827286, + "grad_norm": 0.7625477313995361, + "learning_rate": 3.3156678896589206e-05, + "loss": 3.547, + "step": 113840 + }, + { + "epoch": 7.735086288897948, + "grad_norm": 0.8455623388290405, + "learning_rate": 3.311421388775649e-05, + "loss": 3.3827, + "step": 113845 + }, + { + "epoch": 7.73542600896861, + "grad_norm": 1.0966302156448364, + "learning_rate": 3.307174887892377e-05, + "loss": 3.3407, + "step": 113850 + }, + { + "epoch": 7.735765729039271, + "grad_norm": 1.156745433807373, + "learning_rate": 3.302928387009105e-05, + "loss": 3.38, + "step": 113855 + }, + { + "epoch": 7.7361054491099335, + "grad_norm": 0.8351176381111145, + "learning_rate": 3.298681886125832e-05, + "loss": 3.1059, + "step": 113860 + }, + { + "epoch": 7.7364451691805955, + "grad_norm": 1.0259850025177002, + "learning_rate": 3.29443538524256e-05, + "loss": 3.6076, + "step": 113865 + }, + { + "epoch": 7.736784889251257, + "grad_norm": 0.7871024012565613, + "learning_rate": 3.290188884359288e-05, + "loss": 3.3744, + "step": 113870 + }, + { + "epoch": 7.737124609321919, + "grad_norm": 0.7106296420097351, + "learning_rate": 3.2859423834760153e-05, + "loss": 3.4921, + "step": 113875 + }, + { + "epoch": 7.737464329392581, + "grad_norm": 0.8866576552391052, + "learning_rate": 3.281695882592744e-05, + "loss": 3.5679, + "step": 113880 + }, + { + "epoch": 7.737804049463242, + "grad_norm": 1.3110193014144897, + "learning_rate": 3.2774493817094714e-05, + "loss": 3.2442, + "step": 113885 + }, + { + "epoch": 7.738143769533904, + "grad_norm": 1.0011311769485474, + "learning_rate": 3.2732028808261994e-05, + "loss": 3.336, + "step": 113890 + }, + { + "epoch": 7.738483489604566, + "grad_norm": 0.9322420954704285, + "learning_rate": 3.2689563799429274e-05, + "loss": 3.462, + "step": 113895 + }, + { + "epoch": 7.738823209675227, + "grad_norm": 1.5511772632598877, + "learning_rate": 3.264709879059655e-05, + "loss": 3.3741, + "step": 113900 + }, + { + "epoch": 7.7391629297458895, + "grad_norm": 0.9550550580024719, + "learning_rate": 3.260463378176383e-05, + "loss": 3.4015, + "step": 113905 + }, + { + "epoch": 7.739502649816551, + "grad_norm": 0.8009645938873291, + "learning_rate": 3.256216877293111e-05, + "loss": 3.3525, + "step": 113910 + }, + { + "epoch": 7.739842369887213, + "grad_norm": 1.4414423704147339, + "learning_rate": 3.251970376409839e-05, + "loss": 3.4241, + "step": 113915 + }, + { + "epoch": 7.740182089957875, + "grad_norm": 0.8328583836555481, + "learning_rate": 3.247723875526566e-05, + "loss": 3.3802, + "step": 113920 + }, + { + "epoch": 7.740521810028536, + "grad_norm": 1.174438238143921, + "learning_rate": 3.2434773746432934e-05, + "loss": 3.6131, + "step": 113925 + }, + { + "epoch": 7.740861530099198, + "grad_norm": 0.9082760214805603, + "learning_rate": 3.239230873760022e-05, + "loss": 3.6432, + "step": 113930 + }, + { + "epoch": 7.74120125016986, + "grad_norm": 0.8694512844085693, + "learning_rate": 3.2349843728767494e-05, + "loss": 3.2741, + "step": 113935 + }, + { + "epoch": 7.741540970240521, + "grad_norm": 1.0000150203704834, + "learning_rate": 3.2307378719934774e-05, + "loss": 3.487, + "step": 113940 + }, + { + "epoch": 7.741880690311183, + "grad_norm": 0.9299226999282837, + "learning_rate": 3.2264913711102054e-05, + "loss": 3.3987, + "step": 113945 + }, + { + "epoch": 7.7422204103818455, + "grad_norm": 1.0082813501358032, + "learning_rate": 3.222244870226933e-05, + "loss": 3.3459, + "step": 113950 + }, + { + "epoch": 7.742560130452507, + "grad_norm": 0.8642444610595703, + "learning_rate": 3.217998369343661e-05, + "loss": 3.3634, + "step": 113955 + }, + { + "epoch": 7.742899850523169, + "grad_norm": 1.0015373229980469, + "learning_rate": 3.213751868460389e-05, + "loss": 3.5062, + "step": 113960 + }, + { + "epoch": 7.743239570593831, + "grad_norm": 1.120120644569397, + "learning_rate": 3.209505367577117e-05, + "loss": 3.3435, + "step": 113965 + }, + { + "epoch": 7.743579290664492, + "grad_norm": 0.999608039855957, + "learning_rate": 3.205258866693844e-05, + "loss": 3.437, + "step": 113970 + }, + { + "epoch": 7.743919010735154, + "grad_norm": 0.8387864232063293, + "learning_rate": 3.201012365810573e-05, + "loss": 3.4182, + "step": 113975 + }, + { + "epoch": 7.744258730805816, + "grad_norm": 0.8603915572166443, + "learning_rate": 3.1967658649273e-05, + "loss": 3.5137, + "step": 113980 + }, + { + "epoch": 7.744598450876477, + "grad_norm": 1.0373486280441284, + "learning_rate": 3.1925193640440275e-05, + "loss": 3.2552, + "step": 113985 + }, + { + "epoch": 7.744938170947139, + "grad_norm": 0.7786288857460022, + "learning_rate": 3.1882728631607555e-05, + "loss": 3.1748, + "step": 113990 + }, + { + "epoch": 7.7452778910178015, + "grad_norm": 0.822038471698761, + "learning_rate": 3.1840263622774835e-05, + "loss": 3.3457, + "step": 113995 + }, + { + "epoch": 7.745617611088463, + "grad_norm": 0.9334049224853516, + "learning_rate": 3.1797798613942115e-05, + "loss": 3.6696, + "step": 114000 + }, + { + "epoch": 7.745957331159125, + "grad_norm": 0.7708850502967834, + "learning_rate": 3.175533360510939e-05, + "loss": 3.2378, + "step": 114005 + }, + { + "epoch": 7.746297051229787, + "grad_norm": 0.8382863998413086, + "learning_rate": 3.171286859627667e-05, + "loss": 3.3822, + "step": 114010 + }, + { + "epoch": 7.746636771300448, + "grad_norm": 0.988573431968689, + "learning_rate": 3.167040358744395e-05, + "loss": 3.4609, + "step": 114015 + }, + { + "epoch": 7.74697649137111, + "grad_norm": 0.9047374725341797, + "learning_rate": 3.162793857861122e-05, + "loss": 3.2243, + "step": 114020 + }, + { + "epoch": 7.747316211441772, + "grad_norm": 0.9409074187278748, + "learning_rate": 3.158547356977851e-05, + "loss": 3.4034, + "step": 114025 + }, + { + "epoch": 7.747655931512433, + "grad_norm": 0.9250293970108032, + "learning_rate": 3.154300856094578e-05, + "loss": 3.4198, + "step": 114030 + }, + { + "epoch": 7.7479956515830954, + "grad_norm": 0.7691288590431213, + "learning_rate": 3.1500543552113055e-05, + "loss": 3.3257, + "step": 114035 + }, + { + "epoch": 7.7483353716537575, + "grad_norm": 0.8379738330841064, + "learning_rate": 3.145807854328034e-05, + "loss": 3.5631, + "step": 114040 + }, + { + "epoch": 7.748675091724419, + "grad_norm": 0.672504723072052, + "learning_rate": 3.1415613534447615e-05, + "loss": 3.3422, + "step": 114045 + }, + { + "epoch": 7.749014811795081, + "grad_norm": 1.0758037567138672, + "learning_rate": 3.1373148525614895e-05, + "loss": 3.3585, + "step": 114050 + }, + { + "epoch": 7.749354531865743, + "grad_norm": 1.0037755966186523, + "learning_rate": 3.133068351678217e-05, + "loss": 3.373, + "step": 114055 + }, + { + "epoch": 7.749694251936404, + "grad_norm": 0.9887966513633728, + "learning_rate": 3.1288218507949455e-05, + "loss": 3.2481, + "step": 114060 + }, + { + "epoch": 7.750033972007066, + "grad_norm": 1.0573867559432983, + "learning_rate": 3.124575349911673e-05, + "loss": 3.3936, + "step": 114065 + }, + { + "epoch": 7.750373692077728, + "grad_norm": 0.9229081273078918, + "learning_rate": 3.120328849028401e-05, + "loss": 3.4013, + "step": 114070 + }, + { + "epoch": 7.750713412148389, + "grad_norm": 2.166881561279297, + "learning_rate": 3.116082348145129e-05, + "loss": 3.2851, + "step": 114075 + }, + { + "epoch": 7.7510531322190515, + "grad_norm": 0.9379826188087463, + "learning_rate": 3.111835847261856e-05, + "loss": 3.3473, + "step": 114080 + }, + { + "epoch": 7.7513928522897135, + "grad_norm": 0.9816350340843201, + "learning_rate": 3.107589346378584e-05, + "loss": 3.3369, + "step": 114085 + }, + { + "epoch": 7.751732572360375, + "grad_norm": 0.7438561916351318, + "learning_rate": 3.103342845495312e-05, + "loss": 3.4441, + "step": 114090 + }, + { + "epoch": 7.752072292431037, + "grad_norm": 0.9561501741409302, + "learning_rate": 3.0990963446120396e-05, + "loss": 3.5673, + "step": 114095 + }, + { + "epoch": 7.752412012501699, + "grad_norm": 0.7099414467811584, + "learning_rate": 3.0948498437287676e-05, + "loss": 3.3384, + "step": 114100 + }, + { + "epoch": 7.75275173257236, + "grad_norm": 1.4832388162612915, + "learning_rate": 3.090603342845495e-05, + "loss": 3.2917, + "step": 114105 + }, + { + "epoch": 7.753091452643022, + "grad_norm": 0.8157397508621216, + "learning_rate": 3.086356841962223e-05, + "loss": 3.3562, + "step": 114110 + }, + { + "epoch": 7.753431172713684, + "grad_norm": 1.1557585000991821, + "learning_rate": 3.082110341078951e-05, + "loss": 3.121, + "step": 114115 + }, + { + "epoch": 7.753770892784345, + "grad_norm": 1.0858123302459717, + "learning_rate": 3.077863840195679e-05, + "loss": 3.7526, + "step": 114120 + }, + { + "epoch": 7.7541106128550075, + "grad_norm": 0.8906903266906738, + "learning_rate": 3.073617339312407e-05, + "loss": 3.4147, + "step": 114125 + }, + { + "epoch": 7.7544503329256695, + "grad_norm": 0.9888713359832764, + "learning_rate": 3.069370838429135e-05, + "loss": 3.3775, + "step": 114130 + }, + { + "epoch": 7.754790052996331, + "grad_norm": 1.0424038171768188, + "learning_rate": 3.065124337545862e-05, + "loss": 3.3077, + "step": 114135 + }, + { + "epoch": 7.755129773066993, + "grad_norm": 1.27492356300354, + "learning_rate": 3.06087783666259e-05, + "loss": 3.431, + "step": 114140 + }, + { + "epoch": 7.755469493137655, + "grad_norm": 0.7953582406044006, + "learning_rate": 3.0566313357793176e-05, + "loss": 3.354, + "step": 114145 + }, + { + "epoch": 7.755809213208316, + "grad_norm": 0.8894882798194885, + "learning_rate": 3.0523848348960456e-05, + "loss": 3.3904, + "step": 114150 + }, + { + "epoch": 7.756148933278978, + "grad_norm": 0.7544663548469543, + "learning_rate": 3.0481383340127736e-05, + "loss": 3.3427, + "step": 114155 + }, + { + "epoch": 7.75648865334964, + "grad_norm": 0.9547638297080994, + "learning_rate": 3.0438918331295016e-05, + "loss": 3.5469, + "step": 114160 + }, + { + "epoch": 7.756828373420301, + "grad_norm": 0.9028273820877075, + "learning_rate": 3.039645332246229e-05, + "loss": 3.0767, + "step": 114165 + }, + { + "epoch": 7.7571680934909635, + "grad_norm": 1.0528265237808228, + "learning_rate": 3.035398831362957e-05, + "loss": 3.3497, + "step": 114170 + }, + { + "epoch": 7.7575078135616256, + "grad_norm": 0.9833452105522156, + "learning_rate": 3.031152330479685e-05, + "loss": 3.6516, + "step": 114175 + }, + { + "epoch": 7.757847533632287, + "grad_norm": 0.9690506458282471, + "learning_rate": 3.0269058295964126e-05, + "loss": 3.4683, + "step": 114180 + }, + { + "epoch": 7.758187253702949, + "grad_norm": 1.1274112462997437, + "learning_rate": 3.0226593287131406e-05, + "loss": 3.6698, + "step": 114185 + }, + { + "epoch": 7.758526973773611, + "grad_norm": 0.849075973033905, + "learning_rate": 3.018412827829868e-05, + "loss": 3.4295, + "step": 114190 + }, + { + "epoch": 7.758866693844272, + "grad_norm": 0.791931688785553, + "learning_rate": 3.014166326946596e-05, + "loss": 3.3358, + "step": 114195 + }, + { + "epoch": 7.759206413914934, + "grad_norm": 1.158375859260559, + "learning_rate": 3.009919826063324e-05, + "loss": 3.4053, + "step": 114200 + }, + { + "epoch": 7.759546133985596, + "grad_norm": 1.044782280921936, + "learning_rate": 3.0056733251800517e-05, + "loss": 3.3761, + "step": 114205 + }, + { + "epoch": 7.759885854056257, + "grad_norm": 1.0813876390457153, + "learning_rate": 3.0014268242967797e-05, + "loss": 3.4088, + "step": 114210 + }, + { + "epoch": 7.7602255741269195, + "grad_norm": 0.787753164768219, + "learning_rate": 2.9971803234135073e-05, + "loss": 3.3962, + "step": 114215 + }, + { + "epoch": 7.760565294197582, + "grad_norm": 1.0374958515167236, + "learning_rate": 2.992933822530235e-05, + "loss": 3.2249, + "step": 114220 + }, + { + "epoch": 7.760905014268243, + "grad_norm": 1.136501669883728, + "learning_rate": 2.988687321646963e-05, + "loss": 3.423, + "step": 114225 + }, + { + "epoch": 7.761244734338905, + "grad_norm": 0.7475631237030029, + "learning_rate": 2.9844408207636907e-05, + "loss": 3.6824, + "step": 114230 + }, + { + "epoch": 7.761584454409567, + "grad_norm": 1.2390520572662354, + "learning_rate": 2.9801943198804187e-05, + "loss": 3.1779, + "step": 114235 + }, + { + "epoch": 7.761924174480228, + "grad_norm": 1.3004416227340698, + "learning_rate": 2.9759478189971464e-05, + "loss": 3.3234, + "step": 114240 + }, + { + "epoch": 7.76226389455089, + "grad_norm": 0.976012110710144, + "learning_rate": 2.9717013181138744e-05, + "loss": 3.2305, + "step": 114245 + }, + { + "epoch": 7.762603614621552, + "grad_norm": 0.7753512263298035, + "learning_rate": 2.967454817230602e-05, + "loss": 3.1898, + "step": 114250 + }, + { + "epoch": 7.762943334692213, + "grad_norm": 0.9792818427085876, + "learning_rate": 2.9632083163473297e-05, + "loss": 3.5001, + "step": 114255 + }, + { + "epoch": 7.7632830547628755, + "grad_norm": 0.9479405879974365, + "learning_rate": 2.9589618154640577e-05, + "loss": 3.6559, + "step": 114260 + }, + { + "epoch": 7.763622774833538, + "grad_norm": 2.598273754119873, + "learning_rate": 2.9547153145807857e-05, + "loss": 3.4733, + "step": 114265 + }, + { + "epoch": 7.763962494904199, + "grad_norm": 0.9582160711288452, + "learning_rate": 2.9504688136975134e-05, + "loss": 3.1489, + "step": 114270 + }, + { + "epoch": 7.764302214974861, + "grad_norm": 0.9806935787200928, + "learning_rate": 2.946222312814241e-05, + "loss": 3.2845, + "step": 114275 + }, + { + "epoch": 7.764641935045523, + "grad_norm": 0.9592546224594116, + "learning_rate": 2.9419758119309687e-05, + "loss": 3.3568, + "step": 114280 + }, + { + "epoch": 7.764981655116184, + "grad_norm": 0.8980819582939148, + "learning_rate": 2.9377293110476967e-05, + "loss": 3.4697, + "step": 114285 + }, + { + "epoch": 7.765321375186846, + "grad_norm": 0.9821886420249939, + "learning_rate": 2.9334828101644247e-05, + "loss": 3.5693, + "step": 114290 + }, + { + "epoch": 7.765661095257508, + "grad_norm": 0.8852136135101318, + "learning_rate": 2.9292363092811524e-05, + "loss": 3.3675, + "step": 114295 + }, + { + "epoch": 7.766000815328169, + "grad_norm": 1.0026886463165283, + "learning_rate": 2.9249898083978804e-05, + "loss": 3.5161, + "step": 114300 + }, + { + "epoch": 7.7663405353988315, + "grad_norm": 0.9857033491134644, + "learning_rate": 2.9207433075146077e-05, + "loss": 3.4412, + "step": 114305 + }, + { + "epoch": 7.766680255469494, + "grad_norm": 0.8843522071838379, + "learning_rate": 2.9164968066313358e-05, + "loss": 3.3603, + "step": 114310 + }, + { + "epoch": 7.767019975540155, + "grad_norm": 0.890105128288269, + "learning_rate": 2.9122503057480638e-05, + "loss": 3.2317, + "step": 114315 + }, + { + "epoch": 7.767359695610817, + "grad_norm": 0.8671925663948059, + "learning_rate": 2.9080038048647914e-05, + "loss": 3.5512, + "step": 114320 + }, + { + "epoch": 7.767699415681479, + "grad_norm": 0.8924959301948547, + "learning_rate": 2.9037573039815194e-05, + "loss": 3.5532, + "step": 114325 + }, + { + "epoch": 7.76803913575214, + "grad_norm": 0.8594526648521423, + "learning_rate": 2.899510803098247e-05, + "loss": 3.2612, + "step": 114330 + }, + { + "epoch": 7.768378855822802, + "grad_norm": 0.9476993083953857, + "learning_rate": 2.8952643022149748e-05, + "loss": 3.2115, + "step": 114335 + }, + { + "epoch": 7.768718575893464, + "grad_norm": 0.7819574475288391, + "learning_rate": 2.8910178013317028e-05, + "loss": 3.4154, + "step": 114340 + }, + { + "epoch": 7.7690582959641254, + "grad_norm": 0.8789820075035095, + "learning_rate": 2.8867713004484305e-05, + "loss": 3.3467, + "step": 114345 + }, + { + "epoch": 7.7693980160347875, + "grad_norm": 1.0064070224761963, + "learning_rate": 2.8825247995651585e-05, + "loss": 3.46, + "step": 114350 + }, + { + "epoch": 7.76973773610545, + "grad_norm": 0.8601114153862, + "learning_rate": 2.8782782986818865e-05, + "loss": 3.2521, + "step": 114355 + }, + { + "epoch": 7.770077456176111, + "grad_norm": 1.005618691444397, + "learning_rate": 2.8740317977986138e-05, + "loss": 3.3567, + "step": 114360 + }, + { + "epoch": 7.770417176246773, + "grad_norm": 1.0151828527450562, + "learning_rate": 2.8697852969153418e-05, + "loss": 3.4396, + "step": 114365 + }, + { + "epoch": 7.770756896317434, + "grad_norm": 1.0841989517211914, + "learning_rate": 2.8655387960320695e-05, + "loss": 3.3213, + "step": 114370 + }, + { + "epoch": 7.771096616388096, + "grad_norm": 0.8799169063568115, + "learning_rate": 2.8612922951487975e-05, + "loss": 3.4149, + "step": 114375 + }, + { + "epoch": 7.771436336458758, + "grad_norm": 1.0212934017181396, + "learning_rate": 2.8570457942655255e-05, + "loss": 3.3497, + "step": 114380 + }, + { + "epoch": 7.771776056529419, + "grad_norm": 0.9085917472839355, + "learning_rate": 2.852799293382253e-05, + "loss": 3.4761, + "step": 114385 + }, + { + "epoch": 7.7721157766000815, + "grad_norm": 0.8188848495483398, + "learning_rate": 2.848552792498981e-05, + "loss": 3.2682, + "step": 114390 + }, + { + "epoch": 7.7724554966707435, + "grad_norm": 0.9037489295005798, + "learning_rate": 2.8443062916157085e-05, + "loss": 3.4099, + "step": 114395 + }, + { + "epoch": 7.772795216741405, + "grad_norm": 0.8570337295532227, + "learning_rate": 2.8400597907324365e-05, + "loss": 3.2268, + "step": 114400 + }, + { + "epoch": 7.773134936812067, + "grad_norm": 0.9288970232009888, + "learning_rate": 2.8358132898491645e-05, + "loss": 3.3232, + "step": 114405 + }, + { + "epoch": 7.773474656882729, + "grad_norm": 0.7623081207275391, + "learning_rate": 2.8315667889658922e-05, + "loss": 3.1853, + "step": 114410 + }, + { + "epoch": 7.77381437695339, + "grad_norm": 0.966468870639801, + "learning_rate": 2.8273202880826202e-05, + "loss": 3.494, + "step": 114415 + }, + { + "epoch": 7.774154097024052, + "grad_norm": 0.7410699725151062, + "learning_rate": 2.8230737871993475e-05, + "loss": 3.4778, + "step": 114420 + }, + { + "epoch": 7.774493817094714, + "grad_norm": 1.0408437252044678, + "learning_rate": 2.8188272863160755e-05, + "loss": 3.2265, + "step": 114425 + }, + { + "epoch": 7.774833537165375, + "grad_norm": 1.0460132360458374, + "learning_rate": 2.8145807854328035e-05, + "loss": 3.291, + "step": 114430 + }, + { + "epoch": 7.7751732572360375, + "grad_norm": 0.8750938773155212, + "learning_rate": 2.8103342845495312e-05, + "loss": 3.4979, + "step": 114435 + }, + { + "epoch": 7.7755129773066995, + "grad_norm": 0.913745641708374, + "learning_rate": 2.8060877836662592e-05, + "loss": 3.6756, + "step": 114440 + }, + { + "epoch": 7.775852697377361, + "grad_norm": 0.8830406665802002, + "learning_rate": 2.801841282782987e-05, + "loss": 3.2006, + "step": 114445 + }, + { + "epoch": 7.776192417448023, + "grad_norm": 0.953270435333252, + "learning_rate": 2.7975947818997146e-05, + "loss": 3.5368, + "step": 114450 + }, + { + "epoch": 7.776532137518685, + "grad_norm": 0.8127518892288208, + "learning_rate": 2.7933482810164426e-05, + "loss": 3.1846, + "step": 114455 + }, + { + "epoch": 7.776871857589346, + "grad_norm": 0.910478949546814, + "learning_rate": 2.7891017801331702e-05, + "loss": 3.2827, + "step": 114460 + }, + { + "epoch": 7.777211577660008, + "grad_norm": 0.9566697478294373, + "learning_rate": 2.7848552792498982e-05, + "loss": 3.3286, + "step": 114465 + }, + { + "epoch": 7.77755129773067, + "grad_norm": 1.330091118812561, + "learning_rate": 2.7806087783666262e-05, + "loss": 3.438, + "step": 114470 + }, + { + "epoch": 7.777891017801331, + "grad_norm": 1.0843656063079834, + "learning_rate": 2.7763622774833536e-05, + "loss": 3.2229, + "step": 114475 + }, + { + "epoch": 7.7782307378719935, + "grad_norm": 1.303063988685608, + "learning_rate": 2.7721157766000816e-05, + "loss": 3.219, + "step": 114480 + }, + { + "epoch": 7.778570457942656, + "grad_norm": 0.8026272058486938, + "learning_rate": 2.7678692757168093e-05, + "loss": 3.5766, + "step": 114485 + }, + { + "epoch": 7.778910178013317, + "grad_norm": 1.2248594760894775, + "learning_rate": 2.7636227748335373e-05, + "loss": 3.7205, + "step": 114490 + }, + { + "epoch": 7.779249898083979, + "grad_norm": 0.7396495342254639, + "learning_rate": 2.7593762739502653e-05, + "loss": 3.6514, + "step": 114495 + }, + { + "epoch": 7.779589618154641, + "grad_norm": 0.9443865418434143, + "learning_rate": 2.755129773066993e-05, + "loss": 2.9924, + "step": 114500 + }, + { + "epoch": 7.779929338225302, + "grad_norm": 0.8681981563568115, + "learning_rate": 2.7508832721837206e-05, + "loss": 3.0812, + "step": 114505 + }, + { + "epoch": 7.780269058295964, + "grad_norm": 0.9357820153236389, + "learning_rate": 2.7466367713004483e-05, + "loss": 3.6228, + "step": 114510 + }, + { + "epoch": 7.780608778366626, + "grad_norm": 0.7911099791526794, + "learning_rate": 2.7423902704171763e-05, + "loss": 3.3708, + "step": 114515 + }, + { + "epoch": 7.780948498437287, + "grad_norm": 0.9692290425300598, + "learning_rate": 2.7381437695339043e-05, + "loss": 3.3418, + "step": 114520 + }, + { + "epoch": 7.7812882185079495, + "grad_norm": 1.0767818689346313, + "learning_rate": 2.733897268650632e-05, + "loss": 3.4383, + "step": 114525 + }, + { + "epoch": 7.781627938578612, + "grad_norm": 0.866648256778717, + "learning_rate": 2.7296507677673596e-05, + "loss": 3.3952, + "step": 114530 + }, + { + "epoch": 7.781967658649273, + "grad_norm": 1.1640971899032593, + "learning_rate": 2.7254042668840876e-05, + "loss": 3.4438, + "step": 114535 + }, + { + "epoch": 7.782307378719935, + "grad_norm": 0.8714810013771057, + "learning_rate": 2.7211577660008153e-05, + "loss": 3.2008, + "step": 114540 + }, + { + "epoch": 7.782647098790597, + "grad_norm": 0.8949848413467407, + "learning_rate": 2.7169112651175433e-05, + "loss": 3.488, + "step": 114545 + }, + { + "epoch": 7.782986818861258, + "grad_norm": 1.5420397520065308, + "learning_rate": 2.712664764234271e-05, + "loss": 3.5261, + "step": 114550 + }, + { + "epoch": 7.78332653893192, + "grad_norm": 0.8699105381965637, + "learning_rate": 2.708418263350999e-05, + "loss": 3.2225, + "step": 114555 + }, + { + "epoch": 7.783666259002582, + "grad_norm": 0.6761823892593384, + "learning_rate": 2.7041717624677267e-05, + "loss": 3.4038, + "step": 114560 + }, + { + "epoch": 7.784005979073243, + "grad_norm": 0.8767006993293762, + "learning_rate": 2.6999252615844543e-05, + "loss": 3.5756, + "step": 114565 + }, + { + "epoch": 7.7843456991439055, + "grad_norm": 1.0600621700286865, + "learning_rate": 2.6956787607011823e-05, + "loss": 3.1279, + "step": 114570 + }, + { + "epoch": 7.784685419214568, + "grad_norm": 0.871412456035614, + "learning_rate": 2.69143225981791e-05, + "loss": 3.3503, + "step": 114575 + }, + { + "epoch": 7.785025139285229, + "grad_norm": 0.9404487013816833, + "learning_rate": 2.687185758934638e-05, + "loss": 3.4979, + "step": 114580 + }, + { + "epoch": 7.785364859355891, + "grad_norm": 0.8498484492301941, + "learning_rate": 2.6829392580513657e-05, + "loss": 3.1921, + "step": 114585 + }, + { + "epoch": 7.785704579426552, + "grad_norm": 0.8837407231330872, + "learning_rate": 2.6786927571680933e-05, + "loss": 3.2811, + "step": 114590 + }, + { + "epoch": 7.786044299497214, + "grad_norm": 0.8760497570037842, + "learning_rate": 2.6744462562848214e-05, + "loss": 3.347, + "step": 114595 + }, + { + "epoch": 7.786384019567876, + "grad_norm": 0.9425506591796875, + "learning_rate": 2.670199755401549e-05, + "loss": 3.4214, + "step": 114600 + }, + { + "epoch": 7.786723739638537, + "grad_norm": 0.8882789611816406, + "learning_rate": 2.665953254518277e-05, + "loss": 3.4522, + "step": 114605 + }, + { + "epoch": 7.787063459709199, + "grad_norm": 0.9491703510284424, + "learning_rate": 2.661706753635005e-05, + "loss": 3.4284, + "step": 114610 + }, + { + "epoch": 7.7874031797798615, + "grad_norm": 0.8808254599571228, + "learning_rate": 2.6574602527517324e-05, + "loss": 3.4238, + "step": 114615 + }, + { + "epoch": 7.787742899850523, + "grad_norm": 0.728868842124939, + "learning_rate": 2.6532137518684604e-05, + "loss": 3.4336, + "step": 114620 + }, + { + "epoch": 7.788082619921185, + "grad_norm": 1.028918981552124, + "learning_rate": 2.6489672509851884e-05, + "loss": 3.3523, + "step": 114625 + }, + { + "epoch": 7.788422339991847, + "grad_norm": 0.9009609818458557, + "learning_rate": 2.644720750101916e-05, + "loss": 3.2874, + "step": 114630 + }, + { + "epoch": 7.788762060062508, + "grad_norm": 1.554229497909546, + "learning_rate": 2.640474249218644e-05, + "loss": 3.4524, + "step": 114635 + }, + { + "epoch": 7.78910178013317, + "grad_norm": 0.7338384985923767, + "learning_rate": 2.6362277483353717e-05, + "loss": 3.4901, + "step": 114640 + }, + { + "epoch": 7.789441500203832, + "grad_norm": 0.8928853869438171, + "learning_rate": 2.6319812474520994e-05, + "loss": 3.1303, + "step": 114645 + }, + { + "epoch": 7.789781220274493, + "grad_norm": 0.9731579422950745, + "learning_rate": 2.6277347465688274e-05, + "loss": 3.3878, + "step": 114650 + }, + { + "epoch": 7.7901209403451555, + "grad_norm": 0.9692927598953247, + "learning_rate": 2.623488245685555e-05, + "loss": 3.3208, + "step": 114655 + }, + { + "epoch": 7.7904606604158175, + "grad_norm": 1.1727588176727295, + "learning_rate": 2.619241744802283e-05, + "loss": 3.3699, + "step": 114660 + }, + { + "epoch": 7.790800380486479, + "grad_norm": 1.1700143814086914, + "learning_rate": 2.6149952439190108e-05, + "loss": 3.4973, + "step": 114665 + }, + { + "epoch": 7.791140100557141, + "grad_norm": 0.8853758573532104, + "learning_rate": 2.6107487430357384e-05, + "loss": 3.2868, + "step": 114670 + }, + { + "epoch": 7.791479820627803, + "grad_norm": 0.9971444010734558, + "learning_rate": 2.6065022421524664e-05, + "loss": 3.4429, + "step": 114675 + }, + { + "epoch": 7.791819540698464, + "grad_norm": 0.8227654099464417, + "learning_rate": 2.602255741269194e-05, + "loss": 3.3518, + "step": 114680 + }, + { + "epoch": 7.792159260769126, + "grad_norm": 0.9407891035079956, + "learning_rate": 2.598009240385922e-05, + "loss": 3.1857, + "step": 114685 + }, + { + "epoch": 7.792498980839788, + "grad_norm": 0.9739158153533936, + "learning_rate": 2.59376273950265e-05, + "loss": 3.2681, + "step": 114690 + }, + { + "epoch": 7.792838700910449, + "grad_norm": 0.8474039435386658, + "learning_rate": 2.5895162386193778e-05, + "loss": 3.1404, + "step": 114695 + }, + { + "epoch": 7.7931784209811115, + "grad_norm": 1.071753740310669, + "learning_rate": 2.5852697377361055e-05, + "loss": 3.5149, + "step": 114700 + }, + { + "epoch": 7.7935181410517735, + "grad_norm": 0.8578097820281982, + "learning_rate": 2.581023236852833e-05, + "loss": 3.1574, + "step": 114705 + }, + { + "epoch": 7.793857861122435, + "grad_norm": 0.8999024033546448, + "learning_rate": 2.576776735969561e-05, + "loss": 2.9981, + "step": 114710 + }, + { + "epoch": 7.794197581193097, + "grad_norm": 1.0198650360107422, + "learning_rate": 2.572530235086289e-05, + "loss": 3.2125, + "step": 114715 + }, + { + "epoch": 7.794537301263759, + "grad_norm": 1.075173020362854, + "learning_rate": 2.5682837342030168e-05, + "loss": 3.3118, + "step": 114720 + }, + { + "epoch": 7.79487702133442, + "grad_norm": 0.9192383885383606, + "learning_rate": 2.5640372333197448e-05, + "loss": 3.3439, + "step": 114725 + }, + { + "epoch": 7.795216741405082, + "grad_norm": 0.9810198545455933, + "learning_rate": 2.559790732436472e-05, + "loss": 3.2944, + "step": 114730 + }, + { + "epoch": 7.795556461475744, + "grad_norm": 1.017330527305603, + "learning_rate": 2.5555442315532e-05, + "loss": 3.057, + "step": 114735 + }, + { + "epoch": 7.795896181546405, + "grad_norm": 0.9980810284614563, + "learning_rate": 2.551297730669928e-05, + "loss": 3.3922, + "step": 114740 + }, + { + "epoch": 7.7962359016170675, + "grad_norm": 0.9151728749275208, + "learning_rate": 2.5470512297866558e-05, + "loss": 3.6797, + "step": 114745 + }, + { + "epoch": 7.7965756216877296, + "grad_norm": 1.0833868980407715, + "learning_rate": 2.542804728903384e-05, + "loss": 3.5391, + "step": 114750 + }, + { + "epoch": 7.796915341758391, + "grad_norm": 1.0976717472076416, + "learning_rate": 2.538558228020111e-05, + "loss": 3.164, + "step": 114755 + }, + { + "epoch": 7.797255061829053, + "grad_norm": 0.9158300757408142, + "learning_rate": 2.5343117271368392e-05, + "loss": 3.3446, + "step": 114760 + }, + { + "epoch": 7.797594781899715, + "grad_norm": 0.7317060828208923, + "learning_rate": 2.5300652262535672e-05, + "loss": 3.53, + "step": 114765 + }, + { + "epoch": 7.797934501970376, + "grad_norm": 1.1109212636947632, + "learning_rate": 2.525818725370295e-05, + "loss": 3.3984, + "step": 114770 + }, + { + "epoch": 7.798274222041038, + "grad_norm": 0.7330110669136047, + "learning_rate": 2.521572224487023e-05, + "loss": 3.3434, + "step": 114775 + }, + { + "epoch": 7.7986139421117, + "grad_norm": 0.9277752637863159, + "learning_rate": 2.517325723603751e-05, + "loss": 3.6163, + "step": 114780 + }, + { + "epoch": 7.798953662182361, + "grad_norm": 0.853769838809967, + "learning_rate": 2.5130792227204782e-05, + "loss": 3.2192, + "step": 114785 + }, + { + "epoch": 7.7992933822530235, + "grad_norm": 0.9021204113960266, + "learning_rate": 2.5088327218372062e-05, + "loss": 3.2833, + "step": 114790 + }, + { + "epoch": 7.799633102323686, + "grad_norm": 0.8701243996620178, + "learning_rate": 2.504586220953934e-05, + "loss": 3.5386, + "step": 114795 + }, + { + "epoch": 7.799972822394347, + "grad_norm": 0.8565212488174438, + "learning_rate": 2.500339720070662e-05, + "loss": 3.3747, + "step": 114800 + }, + { + "epoch": 7.800312542465009, + "grad_norm": 0.8150926232337952, + "learning_rate": 2.49609321918739e-05, + "loss": 3.6512, + "step": 114805 + }, + { + "epoch": 7.800652262535671, + "grad_norm": 0.8237428665161133, + "learning_rate": 2.4918467183041176e-05, + "loss": 3.5052, + "step": 114810 + }, + { + "epoch": 7.800991982606332, + "grad_norm": 0.9455736875534058, + "learning_rate": 2.4876002174208452e-05, + "loss": 3.3149, + "step": 114815 + }, + { + "epoch": 7.801331702676994, + "grad_norm": 1.2897381782531738, + "learning_rate": 2.483353716537573e-05, + "loss": 3.311, + "step": 114820 + }, + { + "epoch": 7.801671422747656, + "grad_norm": 0.9023627638816833, + "learning_rate": 2.479107215654301e-05, + "loss": 3.3873, + "step": 114825 + }, + { + "epoch": 7.802011142818317, + "grad_norm": 1.1589488983154297, + "learning_rate": 2.474860714771029e-05, + "loss": 3.5299, + "step": 114830 + }, + { + "epoch": 7.8023508628889795, + "grad_norm": 0.7763519287109375, + "learning_rate": 2.4706142138877566e-05, + "loss": 3.4136, + "step": 114835 + }, + { + "epoch": 7.802690582959642, + "grad_norm": 0.8556327223777771, + "learning_rate": 2.4663677130044842e-05, + "loss": 3.2883, + "step": 114840 + }, + { + "epoch": 7.803030303030303, + "grad_norm": 0.8624034523963928, + "learning_rate": 2.462121212121212e-05, + "loss": 3.4647, + "step": 114845 + }, + { + "epoch": 7.803370023100965, + "grad_norm": 0.8620632290840149, + "learning_rate": 2.45787471123794e-05, + "loss": 3.2735, + "step": 114850 + }, + { + "epoch": 7.803709743171627, + "grad_norm": 0.9168095588684082, + "learning_rate": 2.453628210354668e-05, + "loss": 3.4397, + "step": 114855 + }, + { + "epoch": 7.804049463242288, + "grad_norm": 0.7841411232948303, + "learning_rate": 2.4493817094713956e-05, + "loss": 3.1564, + "step": 114860 + }, + { + "epoch": 7.80438918331295, + "grad_norm": 1.1918741464614868, + "learning_rate": 2.4451352085881236e-05, + "loss": 3.6456, + "step": 114865 + }, + { + "epoch": 7.804728903383612, + "grad_norm": 1.0526901483535767, + "learning_rate": 2.4408887077048513e-05, + "loss": 3.2246, + "step": 114870 + }, + { + "epoch": 7.805068623454273, + "grad_norm": 0.909204363822937, + "learning_rate": 2.436642206821579e-05, + "loss": 3.2226, + "step": 114875 + }, + { + "epoch": 7.8054083435249355, + "grad_norm": 0.9034633040428162, + "learning_rate": 2.432395705938307e-05, + "loss": 3.2116, + "step": 114880 + }, + { + "epoch": 7.805748063595598, + "grad_norm": 0.9939238429069519, + "learning_rate": 2.4281492050550346e-05, + "loss": 3.4899, + "step": 114885 + }, + { + "epoch": 7.806087783666259, + "grad_norm": 0.9055208563804626, + "learning_rate": 2.4239027041717626e-05, + "loss": 3.1281, + "step": 114890 + }, + { + "epoch": 7.806427503736921, + "grad_norm": 0.8697766661643982, + "learning_rate": 2.4196562032884906e-05, + "loss": 3.456, + "step": 114895 + }, + { + "epoch": 7.806767223807583, + "grad_norm": 0.9880050420761108, + "learning_rate": 2.415409702405218e-05, + "loss": 3.4218, + "step": 114900 + }, + { + "epoch": 7.807106943878244, + "grad_norm": 0.8860488533973694, + "learning_rate": 2.411163201521946e-05, + "loss": 2.9476, + "step": 114905 + }, + { + "epoch": 7.807446663948906, + "grad_norm": 0.8634893894195557, + "learning_rate": 2.4069167006386736e-05, + "loss": 3.3474, + "step": 114910 + }, + { + "epoch": 7.807786384019568, + "grad_norm": 1.1205270290374756, + "learning_rate": 2.4026701997554017e-05, + "loss": 3.4568, + "step": 114915 + }, + { + "epoch": 7.8081261040902294, + "grad_norm": 0.9975250959396362, + "learning_rate": 2.3984236988721297e-05, + "loss": 3.2078, + "step": 114920 + }, + { + "epoch": 7.8084658241608915, + "grad_norm": 1.0028150081634521, + "learning_rate": 2.394177197988857e-05, + "loss": 3.1584, + "step": 114925 + }, + { + "epoch": 7.808805544231554, + "grad_norm": 1.0083413124084473, + "learning_rate": 2.389930697105585e-05, + "loss": 3.5501, + "step": 114930 + }, + { + "epoch": 7.809145264302215, + "grad_norm": 1.0009464025497437, + "learning_rate": 2.3856841962223127e-05, + "loss": 3.3586, + "step": 114935 + }, + { + "epoch": 7.809484984372877, + "grad_norm": 0.966263473033905, + "learning_rate": 2.3814376953390407e-05, + "loss": 3.432, + "step": 114940 + }, + { + "epoch": 7.809824704443539, + "grad_norm": 0.8706415295600891, + "learning_rate": 2.3771911944557687e-05, + "loss": 3.3229, + "step": 114945 + }, + { + "epoch": 7.8101644245142, + "grad_norm": 0.8541893362998962, + "learning_rate": 2.3729446935724964e-05, + "loss": 3.3803, + "step": 114950 + }, + { + "epoch": 7.810504144584862, + "grad_norm": 1.062772274017334, + "learning_rate": 2.368698192689224e-05, + "loss": 3.478, + "step": 114955 + }, + { + "epoch": 7.810843864655524, + "grad_norm": 1.0575497150421143, + "learning_rate": 2.364451691805952e-05, + "loss": 3.5381, + "step": 114960 + }, + { + "epoch": 7.8111835847261855, + "grad_norm": 1.0644818544387817, + "learning_rate": 2.3602051909226797e-05, + "loss": 3.3801, + "step": 114965 + }, + { + "epoch": 7.8115233047968475, + "grad_norm": 0.9183666706085205, + "learning_rate": 2.3559586900394077e-05, + "loss": 3.4671, + "step": 114970 + }, + { + "epoch": 7.81186302486751, + "grad_norm": 0.9890817999839783, + "learning_rate": 2.3517121891561354e-05, + "loss": 3.3245, + "step": 114975 + }, + { + "epoch": 7.812202744938171, + "grad_norm": 1.1990106105804443, + "learning_rate": 2.3474656882728634e-05, + "loss": 3.0016, + "step": 114980 + }, + { + "epoch": 7.812542465008833, + "grad_norm": 1.0460859537124634, + "learning_rate": 2.343219187389591e-05, + "loss": 3.3156, + "step": 114985 + }, + { + "epoch": 7.812882185079495, + "grad_norm": 0.9644384384155273, + "learning_rate": 2.3389726865063187e-05, + "loss": 3.4547, + "step": 114990 + }, + { + "epoch": 7.813221905150156, + "grad_norm": 0.8196182250976562, + "learning_rate": 2.3347261856230467e-05, + "loss": 3.4869, + "step": 114995 + }, + { + "epoch": 7.813561625220818, + "grad_norm": 1.2234023809432983, + "learning_rate": 2.3304796847397744e-05, + "loss": 3.4382, + "step": 115000 + }, + { + "epoch": 7.81390134529148, + "grad_norm": 0.7400530576705933, + "learning_rate": 2.3262331838565024e-05, + "loss": 3.3007, + "step": 115005 + }, + { + "epoch": 7.8142410653621415, + "grad_norm": 0.9713556170463562, + "learning_rate": 2.32198668297323e-05, + "loss": 3.4274, + "step": 115010 + }, + { + "epoch": 7.8145807854328035, + "grad_norm": 0.7948873043060303, + "learning_rate": 2.3177401820899577e-05, + "loss": 3.3371, + "step": 115015 + }, + { + "epoch": 7.814920505503466, + "grad_norm": 0.8976809978485107, + "learning_rate": 2.3134936812066857e-05, + "loss": 3.4772, + "step": 115020 + }, + { + "epoch": 7.815260225574127, + "grad_norm": 0.9539018869400024, + "learning_rate": 2.3092471803234134e-05, + "loss": 3.1566, + "step": 115025 + }, + { + "epoch": 7.815599945644789, + "grad_norm": 0.7990810871124268, + "learning_rate": 2.3050006794401414e-05, + "loss": 3.1235, + "step": 115030 + }, + { + "epoch": 7.815939665715451, + "grad_norm": 0.9958177208900452, + "learning_rate": 2.3007541785568694e-05, + "loss": 3.5067, + "step": 115035 + }, + { + "epoch": 7.816279385786112, + "grad_norm": 0.8618928790092468, + "learning_rate": 2.2965076776735968e-05, + "loss": 3.1871, + "step": 115040 + }, + { + "epoch": 7.816619105856774, + "grad_norm": 1.0150038003921509, + "learning_rate": 2.2922611767903248e-05, + "loss": 3.383, + "step": 115045 + }, + { + "epoch": 7.816958825927435, + "grad_norm": 1.1347754001617432, + "learning_rate": 2.2880146759070528e-05, + "loss": 3.0971, + "step": 115050 + }, + { + "epoch": 7.8172985459980975, + "grad_norm": 1.2435821294784546, + "learning_rate": 2.2837681750237804e-05, + "loss": 3.5735, + "step": 115055 + }, + { + "epoch": 7.81763826606876, + "grad_norm": 0.9344014525413513, + "learning_rate": 2.2795216741405085e-05, + "loss": 3.7198, + "step": 115060 + }, + { + "epoch": 7.817977986139421, + "grad_norm": 0.7925999760627747, + "learning_rate": 2.275275173257236e-05, + "loss": 3.5238, + "step": 115065 + }, + { + "epoch": 7.818317706210083, + "grad_norm": 0.8728433847427368, + "learning_rate": 2.2710286723739638e-05, + "loss": 3.5542, + "step": 115070 + }, + { + "epoch": 7.818657426280745, + "grad_norm": 0.916090726852417, + "learning_rate": 2.2667821714906918e-05, + "loss": 3.0709, + "step": 115075 + }, + { + "epoch": 7.818997146351406, + "grad_norm": 0.9672450423240662, + "learning_rate": 2.2625356706074195e-05, + "loss": 3.391, + "step": 115080 + }, + { + "epoch": 7.819336866422068, + "grad_norm": 0.9503450393676758, + "learning_rate": 2.2582891697241475e-05, + "loss": 3.2224, + "step": 115085 + }, + { + "epoch": 7.81967658649273, + "grad_norm": 0.8778010010719299, + "learning_rate": 2.254042668840875e-05, + "loss": 3.3837, + "step": 115090 + }, + { + "epoch": 7.820016306563391, + "grad_norm": 1.0164954662322998, + "learning_rate": 2.2497961679576028e-05, + "loss": 3.6466, + "step": 115095 + }, + { + "epoch": 7.8203560266340535, + "grad_norm": 0.7432065010070801, + "learning_rate": 2.2455496670743308e-05, + "loss": 3.2694, + "step": 115100 + }, + { + "epoch": 7.820695746704716, + "grad_norm": 0.948091447353363, + "learning_rate": 2.2413031661910585e-05, + "loss": 3.2755, + "step": 115105 + }, + { + "epoch": 7.821035466775377, + "grad_norm": 0.914864182472229, + "learning_rate": 2.2370566653077865e-05, + "loss": 3.5278, + "step": 115110 + }, + { + "epoch": 7.821375186846039, + "grad_norm": 0.9578237533569336, + "learning_rate": 2.232810164424514e-05, + "loss": 3.5982, + "step": 115115 + }, + { + "epoch": 7.821714906916701, + "grad_norm": 0.7923611402511597, + "learning_rate": 2.2285636635412422e-05, + "loss": 3.3967, + "step": 115120 + }, + { + "epoch": 7.822054626987362, + "grad_norm": 1.063409447669983, + "learning_rate": 2.22431716265797e-05, + "loss": 3.1461, + "step": 115125 + }, + { + "epoch": 7.822394347058024, + "grad_norm": 1.1569054126739502, + "learning_rate": 2.2200706617746975e-05, + "loss": 3.3187, + "step": 115130 + }, + { + "epoch": 7.822734067128686, + "grad_norm": 0.9701105356216431, + "learning_rate": 2.2158241608914255e-05, + "loss": 3.3888, + "step": 115135 + }, + { + "epoch": 7.823073787199347, + "grad_norm": 0.8298749327659607, + "learning_rate": 2.2115776600081535e-05, + "loss": 3.2633, + "step": 115140 + }, + { + "epoch": 7.8234135072700095, + "grad_norm": 0.9460019469261169, + "learning_rate": 2.2073311591248812e-05, + "loss": 3.2652, + "step": 115145 + }, + { + "epoch": 7.823753227340672, + "grad_norm": 0.8895694613456726, + "learning_rate": 2.2030846582416092e-05, + "loss": 3.3738, + "step": 115150 + }, + { + "epoch": 7.824092947411333, + "grad_norm": 0.9868344664573669, + "learning_rate": 2.1988381573583365e-05, + "loss": 3.2977, + "step": 115155 + }, + { + "epoch": 7.824432667481995, + "grad_norm": 1.5767133235931396, + "learning_rate": 2.1945916564750645e-05, + "loss": 3.261, + "step": 115160 + }, + { + "epoch": 7.824772387552657, + "grad_norm": 0.7985988855361938, + "learning_rate": 2.1903451555917926e-05, + "loss": 3.2836, + "step": 115165 + }, + { + "epoch": 7.825112107623318, + "grad_norm": 0.9089409708976746, + "learning_rate": 2.1860986547085202e-05, + "loss": 3.334, + "step": 115170 + }, + { + "epoch": 7.82545182769398, + "grad_norm": 0.8227800130844116, + "learning_rate": 2.1818521538252482e-05, + "loss": 3.2743, + "step": 115175 + }, + { + "epoch": 7.825791547764642, + "grad_norm": 0.8672536015510559, + "learning_rate": 2.1776056529419756e-05, + "loss": 3.5063, + "step": 115180 + }, + { + "epoch": 7.826131267835303, + "grad_norm": 1.2156225442886353, + "learning_rate": 2.1733591520587036e-05, + "loss": 3.3321, + "step": 115185 + }, + { + "epoch": 7.8264709879059655, + "grad_norm": 0.9809209704399109, + "learning_rate": 2.1691126511754316e-05, + "loss": 3.4753, + "step": 115190 + }, + { + "epoch": 7.826810707976628, + "grad_norm": 0.8580672144889832, + "learning_rate": 2.1648661502921592e-05, + "loss": 3.5361, + "step": 115195 + }, + { + "epoch": 7.827150428047289, + "grad_norm": 0.8548184037208557, + "learning_rate": 2.1606196494088873e-05, + "loss": 3.4635, + "step": 115200 + }, + { + "epoch": 7.827490148117951, + "grad_norm": 1.1245819330215454, + "learning_rate": 2.156373148525615e-05, + "loss": 3.3361, + "step": 115205 + }, + { + "epoch": 7.827829868188613, + "grad_norm": 1.125425934791565, + "learning_rate": 2.1521266476423426e-05, + "loss": 3.3965, + "step": 115210 + }, + { + "epoch": 7.828169588259274, + "grad_norm": 1.0023362636566162, + "learning_rate": 2.1478801467590706e-05, + "loss": 3.4291, + "step": 115215 + }, + { + "epoch": 7.828509308329936, + "grad_norm": 1.1055757999420166, + "learning_rate": 2.1436336458757983e-05, + "loss": 3.3867, + "step": 115220 + }, + { + "epoch": 7.828849028400598, + "grad_norm": 1.0182008743286133, + "learning_rate": 2.1393871449925263e-05, + "loss": 3.4234, + "step": 115225 + }, + { + "epoch": 7.8291887484712595, + "grad_norm": 0.897865891456604, + "learning_rate": 2.1351406441092543e-05, + "loss": 3.3121, + "step": 115230 + }, + { + "epoch": 7.8295284685419215, + "grad_norm": 0.9163982272148132, + "learning_rate": 2.130894143225982e-05, + "loss": 3.2934, + "step": 115235 + }, + { + "epoch": 7.829868188612584, + "grad_norm": 1.1169025897979736, + "learning_rate": 2.1266476423427096e-05, + "loss": 3.4869, + "step": 115240 + }, + { + "epoch": 7.830207908683245, + "grad_norm": 1.164669156074524, + "learning_rate": 2.1224011414594373e-05, + "loss": 3.5116, + "step": 115245 + }, + { + "epoch": 7.830547628753907, + "grad_norm": 0.9497455954551697, + "learning_rate": 2.1181546405761653e-05, + "loss": 3.3894, + "step": 115250 + }, + { + "epoch": 7.830887348824569, + "grad_norm": 1.1535075902938843, + "learning_rate": 2.1139081396928933e-05, + "loss": 3.2919, + "step": 115255 + }, + { + "epoch": 7.83122706889523, + "grad_norm": 0.8567995429039001, + "learning_rate": 2.109661638809621e-05, + "loss": 3.3377, + "step": 115260 + }, + { + "epoch": 7.831566788965892, + "grad_norm": 0.8599972128868103, + "learning_rate": 2.1054151379263486e-05, + "loss": 3.2718, + "step": 115265 + }, + { + "epoch": 7.831906509036553, + "grad_norm": 0.9488184452056885, + "learning_rate": 2.1011686370430763e-05, + "loss": 3.7362, + "step": 115270 + }, + { + "epoch": 7.8322462291072155, + "grad_norm": 0.8111642599105835, + "learning_rate": 2.0969221361598043e-05, + "loss": 3.3419, + "step": 115275 + }, + { + "epoch": 7.8325859491778775, + "grad_norm": 0.9364269375801086, + "learning_rate": 2.0926756352765323e-05, + "loss": 3.4639, + "step": 115280 + }, + { + "epoch": 7.832925669248539, + "grad_norm": 1.669872760772705, + "learning_rate": 2.08842913439326e-05, + "loss": 3.4333, + "step": 115285 + }, + { + "epoch": 7.833265389319201, + "grad_norm": 0.766767680644989, + "learning_rate": 2.084182633509988e-05, + "loss": 3.3141, + "step": 115290 + }, + { + "epoch": 7.833605109389863, + "grad_norm": 0.9940740466117859, + "learning_rate": 2.0799361326267153e-05, + "loss": 3.3317, + "step": 115295 + }, + { + "epoch": 7.833944829460524, + "grad_norm": 0.9935349822044373, + "learning_rate": 2.0756896317434433e-05, + "loss": 3.1255, + "step": 115300 + }, + { + "epoch": 7.834284549531186, + "grad_norm": 0.766322910785675, + "learning_rate": 2.0714431308601713e-05, + "loss": 3.2624, + "step": 115305 + }, + { + "epoch": 7.834624269601848, + "grad_norm": 0.8190920352935791, + "learning_rate": 2.067196629976899e-05, + "loss": 3.5322, + "step": 115310 + }, + { + "epoch": 7.834963989672509, + "grad_norm": 0.9058427810668945, + "learning_rate": 2.062950129093627e-05, + "loss": 3.216, + "step": 115315 + }, + { + "epoch": 7.8353037097431715, + "grad_norm": 1.0502796173095703, + "learning_rate": 2.058703628210355e-05, + "loss": 3.5487, + "step": 115320 + }, + { + "epoch": 7.8356434298138335, + "grad_norm": 0.8332571983337402, + "learning_rate": 2.0544571273270824e-05, + "loss": 3.4709, + "step": 115325 + }, + { + "epoch": 7.835983149884495, + "grad_norm": 0.872771143913269, + "learning_rate": 2.0502106264438104e-05, + "loss": 3.4435, + "step": 115330 + }, + { + "epoch": 7.836322869955157, + "grad_norm": 0.829639196395874, + "learning_rate": 2.045964125560538e-05, + "loss": 3.4226, + "step": 115335 + }, + { + "epoch": 7.836662590025819, + "grad_norm": 0.8424528241157532, + "learning_rate": 2.041717624677266e-05, + "loss": 3.3118, + "step": 115340 + }, + { + "epoch": 7.83700231009648, + "grad_norm": 0.8807824850082397, + "learning_rate": 2.037471123793994e-05, + "loss": 3.4897, + "step": 115345 + }, + { + "epoch": 7.837342030167142, + "grad_norm": 0.9277797937393188, + "learning_rate": 2.0332246229107214e-05, + "loss": 3.2767, + "step": 115350 + }, + { + "epoch": 7.837681750237804, + "grad_norm": 0.8324109315872192, + "learning_rate": 2.0289781220274494e-05, + "loss": 3.4191, + "step": 115355 + }, + { + "epoch": 7.838021470308465, + "grad_norm": 0.9083201289176941, + "learning_rate": 2.024731621144177e-05, + "loss": 3.4304, + "step": 115360 + }, + { + "epoch": 7.8383611903791275, + "grad_norm": 0.9980582594871521, + "learning_rate": 2.020485120260905e-05, + "loss": 3.5723, + "step": 115365 + }, + { + "epoch": 7.83870091044979, + "grad_norm": 1.0795093774795532, + "learning_rate": 2.016238619377633e-05, + "loss": 3.306, + "step": 115370 + }, + { + "epoch": 7.839040630520451, + "grad_norm": 1.0115547180175781, + "learning_rate": 2.0119921184943607e-05, + "loss": 3.2006, + "step": 115375 + }, + { + "epoch": 7.839380350591113, + "grad_norm": 0.9588574767112732, + "learning_rate": 2.0077456176110884e-05, + "loss": 3.614, + "step": 115380 + }, + { + "epoch": 7.839720070661775, + "grad_norm": 0.9923123717308044, + "learning_rate": 2.003499116727816e-05, + "loss": 3.4326, + "step": 115385 + }, + { + "epoch": 7.840059790732436, + "grad_norm": 1.0727978944778442, + "learning_rate": 1.999252615844544e-05, + "loss": 3.3574, + "step": 115390 + }, + { + "epoch": 7.840399510803098, + "grad_norm": 0.8572938442230225, + "learning_rate": 1.995006114961272e-05, + "loss": 3.3588, + "step": 115395 + }, + { + "epoch": 7.84073923087376, + "grad_norm": 0.9549247622489929, + "learning_rate": 1.9907596140779998e-05, + "loss": 3.3397, + "step": 115400 + }, + { + "epoch": 7.841078950944421, + "grad_norm": 1.0399845838546753, + "learning_rate": 1.9865131131947278e-05, + "loss": 3.1832, + "step": 115405 + }, + { + "epoch": 7.8414186710150835, + "grad_norm": 1.0191320180892944, + "learning_rate": 1.9822666123114554e-05, + "loss": 3.2286, + "step": 115410 + }, + { + "epoch": 7.841758391085746, + "grad_norm": 0.9057300090789795, + "learning_rate": 1.978020111428183e-05, + "loss": 3.3002, + "step": 115415 + }, + { + "epoch": 7.842098111156407, + "grad_norm": 0.9151212573051453, + "learning_rate": 1.973773610544911e-05, + "loss": 3.3469, + "step": 115420 + }, + { + "epoch": 7.842437831227069, + "grad_norm": 0.9801045656204224, + "learning_rate": 1.9695271096616388e-05, + "loss": 3.5093, + "step": 115425 + }, + { + "epoch": 7.842777551297731, + "grad_norm": 0.8040044903755188, + "learning_rate": 1.9652806087783668e-05, + "loss": 3.3357, + "step": 115430 + }, + { + "epoch": 7.843117271368392, + "grad_norm": 0.8929601907730103, + "learning_rate": 1.9610341078950945e-05, + "loss": 3.5006, + "step": 115435 + }, + { + "epoch": 7.843456991439054, + "grad_norm": 2.238642930984497, + "learning_rate": 1.956787607011822e-05, + "loss": 3.3249, + "step": 115440 + }, + { + "epoch": 7.843796711509716, + "grad_norm": 1.0795100927352905, + "learning_rate": 1.95254110612855e-05, + "loss": 3.4716, + "step": 115445 + }, + { + "epoch": 7.844136431580377, + "grad_norm": 0.8675016164779663, + "learning_rate": 1.9482946052452778e-05, + "loss": 3.4947, + "step": 115450 + }, + { + "epoch": 7.8444761516510395, + "grad_norm": 0.6946950554847717, + "learning_rate": 1.9440481043620058e-05, + "loss": 3.1129, + "step": 115455 + }, + { + "epoch": 7.844815871721702, + "grad_norm": 0.8245530724525452, + "learning_rate": 1.9398016034787338e-05, + "loss": 3.247, + "step": 115460 + }, + { + "epoch": 7.845155591792363, + "grad_norm": 0.8503345847129822, + "learning_rate": 1.935555102595461e-05, + "loss": 3.3864, + "step": 115465 + }, + { + "epoch": 7.845495311863025, + "grad_norm": 0.9380565881729126, + "learning_rate": 1.931308601712189e-05, + "loss": 3.6117, + "step": 115470 + }, + { + "epoch": 7.845835031933687, + "grad_norm": 1.0350505113601685, + "learning_rate": 1.9270621008289172e-05, + "loss": 3.1813, + "step": 115475 + }, + { + "epoch": 7.846174752004348, + "grad_norm": 1.0414255857467651, + "learning_rate": 1.922815599945645e-05, + "loss": 3.3194, + "step": 115480 + }, + { + "epoch": 7.84651447207501, + "grad_norm": 0.8497803807258606, + "learning_rate": 1.918569099062373e-05, + "loss": 3.4169, + "step": 115485 + }, + { + "epoch": 7.846854192145672, + "grad_norm": 1.05471932888031, + "learning_rate": 1.9143225981791002e-05, + "loss": 3.3687, + "step": 115490 + }, + { + "epoch": 7.8471939122163334, + "grad_norm": 1.11383056640625, + "learning_rate": 1.9100760972958282e-05, + "loss": 3.5065, + "step": 115495 + }, + { + "epoch": 7.8475336322869955, + "grad_norm": 0.8325208425521851, + "learning_rate": 1.9058295964125562e-05, + "loss": 3.7348, + "step": 115500 + }, + { + "epoch": 7.847873352357658, + "grad_norm": 1.2329552173614502, + "learning_rate": 1.901583095529284e-05, + "loss": 3.3799, + "step": 115505 + }, + { + "epoch": 7.848213072428319, + "grad_norm": 1.192344069480896, + "learning_rate": 1.897336594646012e-05, + "loss": 3.2465, + "step": 115510 + }, + { + "epoch": 7.848552792498981, + "grad_norm": 0.9235686659812927, + "learning_rate": 1.8930900937627395e-05, + "loss": 3.0488, + "step": 115515 + }, + { + "epoch": 7.848892512569643, + "grad_norm": 1.491414189338684, + "learning_rate": 1.8888435928794672e-05, + "loss": 3.5116, + "step": 115520 + }, + { + "epoch": 7.849232232640304, + "grad_norm": 0.8831323385238647, + "learning_rate": 1.8845970919961952e-05, + "loss": 3.1097, + "step": 115525 + }, + { + "epoch": 7.849571952710966, + "grad_norm": 1.028315782546997, + "learning_rate": 1.880350591112923e-05, + "loss": 3.0957, + "step": 115530 + }, + { + "epoch": 7.849911672781628, + "grad_norm": 0.8374600410461426, + "learning_rate": 1.876104090229651e-05, + "loss": 3.3363, + "step": 115535 + }, + { + "epoch": 7.8502513928522895, + "grad_norm": 0.9059107899665833, + "learning_rate": 1.8718575893463786e-05, + "loss": 3.5201, + "step": 115540 + }, + { + "epoch": 7.8505911129229515, + "grad_norm": 1.0360203981399536, + "learning_rate": 1.8676110884631066e-05, + "loss": 3.5926, + "step": 115545 + }, + { + "epoch": 7.850930832993614, + "grad_norm": 0.8973623514175415, + "learning_rate": 1.8633645875798342e-05, + "loss": 3.488, + "step": 115550 + }, + { + "epoch": 7.851270553064275, + "grad_norm": 1.004514455795288, + "learning_rate": 1.859118086696562e-05, + "loss": 3.2363, + "step": 115555 + }, + { + "epoch": 7.851610273134937, + "grad_norm": 0.7481683492660522, + "learning_rate": 1.85487158581329e-05, + "loss": 3.1773, + "step": 115560 + }, + { + "epoch": 7.851949993205599, + "grad_norm": 0.7638279795646667, + "learning_rate": 1.850625084930018e-05, + "loss": 3.3742, + "step": 115565 + }, + { + "epoch": 7.85228971327626, + "grad_norm": 0.87955242395401, + "learning_rate": 1.8463785840467456e-05, + "loss": 3.2667, + "step": 115570 + }, + { + "epoch": 7.852629433346922, + "grad_norm": 0.8077199459075928, + "learning_rate": 1.8421320831634733e-05, + "loss": 3.6291, + "step": 115575 + }, + { + "epoch": 7.852969153417584, + "grad_norm": 0.7823739051818848, + "learning_rate": 1.837885582280201e-05, + "loss": 3.363, + "step": 115580 + }, + { + "epoch": 7.8533088734882455, + "grad_norm": 0.7929497957229614, + "learning_rate": 1.833639081396929e-05, + "loss": 3.2181, + "step": 115585 + }, + { + "epoch": 7.8536485935589075, + "grad_norm": 1.136122465133667, + "learning_rate": 1.829392580513657e-05, + "loss": 3.2883, + "step": 115590 + }, + { + "epoch": 7.85398831362957, + "grad_norm": 0.8703322410583496, + "learning_rate": 1.8251460796303846e-05, + "loss": 3.3356, + "step": 115595 + }, + { + "epoch": 7.854328033700231, + "grad_norm": 0.8284199833869934, + "learning_rate": 1.8208995787471126e-05, + "loss": 3.3301, + "step": 115600 + }, + { + "epoch": 7.854667753770893, + "grad_norm": 1.1532025337219238, + "learning_rate": 1.81665307786384e-05, + "loss": 3.3421, + "step": 115605 + }, + { + "epoch": 7.855007473841555, + "grad_norm": 0.9762461185455322, + "learning_rate": 1.812406576980568e-05, + "loss": 3.253, + "step": 115610 + }, + { + "epoch": 7.855347193912216, + "grad_norm": 0.8755745887756348, + "learning_rate": 1.808160076097296e-05, + "loss": 3.3775, + "step": 115615 + }, + { + "epoch": 7.855686913982878, + "grad_norm": 0.9227385520935059, + "learning_rate": 1.8039135752140236e-05, + "loss": 3.113, + "step": 115620 + }, + { + "epoch": 7.85602663405354, + "grad_norm": 0.7725024223327637, + "learning_rate": 1.7996670743307516e-05, + "loss": 3.5269, + "step": 115625 + }, + { + "epoch": 7.8563663541242015, + "grad_norm": 0.8470501899719238, + "learning_rate": 1.7954205734474793e-05, + "loss": 3.2524, + "step": 115630 + }, + { + "epoch": 7.8567060741948636, + "grad_norm": 0.8843463659286499, + "learning_rate": 1.791174072564207e-05, + "loss": 3.4602, + "step": 115635 + }, + { + "epoch": 7.857045794265526, + "grad_norm": 0.804688036441803, + "learning_rate": 1.786927571680935e-05, + "loss": 3.4103, + "step": 115640 + }, + { + "epoch": 7.857385514336187, + "grad_norm": 0.9151496887207031, + "learning_rate": 1.7826810707976627e-05, + "loss": 3.1336, + "step": 115645 + }, + { + "epoch": 7.857725234406849, + "grad_norm": 0.9277817010879517, + "learning_rate": 1.7784345699143907e-05, + "loss": 3.3846, + "step": 115650 + }, + { + "epoch": 7.858064954477511, + "grad_norm": 0.8769875764846802, + "learning_rate": 1.7741880690311187e-05, + "loss": 3.3161, + "step": 115655 + }, + { + "epoch": 7.858404674548172, + "grad_norm": 1.1779841184616089, + "learning_rate": 1.769941568147846e-05, + "loss": 3.2253, + "step": 115660 + }, + { + "epoch": 7.858744394618834, + "grad_norm": 0.8765807151794434, + "learning_rate": 1.765695067264574e-05, + "loss": 3.2674, + "step": 115665 + }, + { + "epoch": 7.859084114689496, + "grad_norm": 0.9855905175209045, + "learning_rate": 1.7614485663813017e-05, + "loss": 3.4436, + "step": 115670 + }, + { + "epoch": 7.8594238347601575, + "grad_norm": 0.9194306135177612, + "learning_rate": 1.7572020654980297e-05, + "loss": 3.327, + "step": 115675 + }, + { + "epoch": 7.85976355483082, + "grad_norm": 0.894448459148407, + "learning_rate": 1.7529555646147577e-05, + "loss": 3.3961, + "step": 115680 + }, + { + "epoch": 7.860103274901482, + "grad_norm": 0.8753439784049988, + "learning_rate": 1.7487090637314854e-05, + "loss": 3.5525, + "step": 115685 + }, + { + "epoch": 7.860442994972143, + "grad_norm": 0.8302602171897888, + "learning_rate": 1.744462562848213e-05, + "loss": 3.4649, + "step": 115690 + }, + { + "epoch": 7.860782715042805, + "grad_norm": 0.9713963270187378, + "learning_rate": 1.7402160619649407e-05, + "loss": 3.2336, + "step": 115695 + }, + { + "epoch": 7.861122435113467, + "grad_norm": 0.8491024971008301, + "learning_rate": 1.7359695610816687e-05, + "loss": 3.397, + "step": 115700 + }, + { + "epoch": 7.861462155184128, + "grad_norm": 0.9135463237762451, + "learning_rate": 1.7317230601983967e-05, + "loss": 3.2601, + "step": 115705 + }, + { + "epoch": 7.86180187525479, + "grad_norm": 0.8147472739219666, + "learning_rate": 1.7274765593151244e-05, + "loss": 3.3101, + "step": 115710 + }, + { + "epoch": 7.862141595325452, + "grad_norm": 0.9264710545539856, + "learning_rate": 1.7232300584318524e-05, + "loss": 3.2193, + "step": 115715 + }, + { + "epoch": 7.8624813153961135, + "grad_norm": 0.972954511642456, + "learning_rate": 1.7189835575485797e-05, + "loss": 3.2803, + "step": 115720 + }, + { + "epoch": 7.862821035466776, + "grad_norm": 1.0205347537994385, + "learning_rate": 1.7147370566653077e-05, + "loss": 3.2672, + "step": 115725 + }, + { + "epoch": 7.863160755537437, + "grad_norm": 0.8236509561538696, + "learning_rate": 1.7104905557820357e-05, + "loss": 3.3405, + "step": 115730 + }, + { + "epoch": 7.863500475608099, + "grad_norm": 0.8200693130493164, + "learning_rate": 1.7062440548987634e-05, + "loss": 3.2295, + "step": 115735 + }, + { + "epoch": 7.863840195678761, + "grad_norm": 0.8511335253715515, + "learning_rate": 1.7019975540154914e-05, + "loss": 3.3714, + "step": 115740 + }, + { + "epoch": 7.864179915749422, + "grad_norm": 0.8673534989356995, + "learning_rate": 1.697751053132219e-05, + "loss": 3.4285, + "step": 115745 + }, + { + "epoch": 7.864519635820084, + "grad_norm": 0.9394187331199646, + "learning_rate": 1.6935045522489468e-05, + "loss": 3.2564, + "step": 115750 + }, + { + "epoch": 7.864859355890746, + "grad_norm": 1.0448726415634155, + "learning_rate": 1.6892580513656748e-05, + "loss": 3.5284, + "step": 115755 + }, + { + "epoch": 7.865199075961407, + "grad_norm": 0.8527834415435791, + "learning_rate": 1.6850115504824024e-05, + "loss": 3.4796, + "step": 115760 + }, + { + "epoch": 7.8655387960320695, + "grad_norm": 0.9138684868812561, + "learning_rate": 1.6807650495991304e-05, + "loss": 3.5063, + "step": 115765 + }, + { + "epoch": 7.865878516102732, + "grad_norm": 0.8502715229988098, + "learning_rate": 1.6765185487158584e-05, + "loss": 3.483, + "step": 115770 + }, + { + "epoch": 7.866218236173393, + "grad_norm": 0.8813223838806152, + "learning_rate": 1.6722720478325858e-05, + "loss": 3.3439, + "step": 115775 + }, + { + "epoch": 7.866557956244055, + "grad_norm": 0.8714665174484253, + "learning_rate": 1.6680255469493138e-05, + "loss": 3.4891, + "step": 115780 + }, + { + "epoch": 7.866897676314717, + "grad_norm": 0.8910839557647705, + "learning_rate": 1.6637790460660415e-05, + "loss": 3.4159, + "step": 115785 + }, + { + "epoch": 7.867237396385378, + "grad_norm": 0.9133066534996033, + "learning_rate": 1.6595325451827695e-05, + "loss": 3.5824, + "step": 115790 + }, + { + "epoch": 7.86757711645604, + "grad_norm": 1.0064222812652588, + "learning_rate": 1.6552860442994975e-05, + "loss": 3.5941, + "step": 115795 + }, + { + "epoch": 7.867916836526702, + "grad_norm": 0.7977838516235352, + "learning_rate": 1.651039543416225e-05, + "loss": 3.3486, + "step": 115800 + }, + { + "epoch": 7.8682565565973634, + "grad_norm": 0.978630006313324, + "learning_rate": 1.6467930425329528e-05, + "loss": 3.4512, + "step": 115805 + }, + { + "epoch": 7.8685962766680255, + "grad_norm": 1.1104168891906738, + "learning_rate": 1.6425465416496805e-05, + "loss": 3.4998, + "step": 115810 + }, + { + "epoch": 7.868935996738688, + "grad_norm": 0.8991864323616028, + "learning_rate": 1.6383000407664085e-05, + "loss": 3.4697, + "step": 115815 + }, + { + "epoch": 7.869275716809349, + "grad_norm": 0.7949345707893372, + "learning_rate": 1.6340535398831365e-05, + "loss": 3.4257, + "step": 115820 + }, + { + "epoch": 7.869615436880011, + "grad_norm": 0.9320369362831116, + "learning_rate": 1.629807038999864e-05, + "loss": 3.4998, + "step": 115825 + }, + { + "epoch": 7.869955156950673, + "grad_norm": 0.8939493298530579, + "learning_rate": 1.6255605381165918e-05, + "loss": 3.6005, + "step": 115830 + }, + { + "epoch": 7.870294877021334, + "grad_norm": 1.0706661939620972, + "learning_rate": 1.62131403723332e-05, + "loss": 3.1706, + "step": 115835 + }, + { + "epoch": 7.870634597091996, + "grad_norm": 0.7849816083908081, + "learning_rate": 1.6170675363500475e-05, + "loss": 3.2958, + "step": 115840 + }, + { + "epoch": 7.870974317162658, + "grad_norm": 0.9772716164588928, + "learning_rate": 1.6128210354667755e-05, + "loss": 3.3693, + "step": 115845 + }, + { + "epoch": 7.8713140372333195, + "grad_norm": 1.0856086015701294, + "learning_rate": 1.6085745345835032e-05, + "loss": 3.2889, + "step": 115850 + }, + { + "epoch": 7.8716537573039815, + "grad_norm": 0.9893754124641418, + "learning_rate": 1.6043280337002312e-05, + "loss": 3.2238, + "step": 115855 + }, + { + "epoch": 7.871993477374644, + "grad_norm": 0.9256649613380432, + "learning_rate": 1.600081532816959e-05, + "loss": 3.4883, + "step": 115860 + }, + { + "epoch": 7.872333197445305, + "grad_norm": 0.8375445008277893, + "learning_rate": 1.5958350319336865e-05, + "loss": 3.5227, + "step": 115865 + }, + { + "epoch": 7.872672917515967, + "grad_norm": 0.973482072353363, + "learning_rate": 1.5915885310504145e-05, + "loss": 3.4382, + "step": 115870 + }, + { + "epoch": 7.873012637586629, + "grad_norm": 0.8639630079269409, + "learning_rate": 1.5873420301671422e-05, + "loss": 3.5828, + "step": 115875 + }, + { + "epoch": 7.87335235765729, + "grad_norm": 0.6852543354034424, + "learning_rate": 1.5830955292838702e-05, + "loss": 3.3958, + "step": 115880 + }, + { + "epoch": 7.873692077727952, + "grad_norm": 1.041929006576538, + "learning_rate": 1.5788490284005982e-05, + "loss": 3.5917, + "step": 115885 + }, + { + "epoch": 7.874031797798614, + "grad_norm": 0.9538796544075012, + "learning_rate": 1.5746025275173255e-05, + "loss": 3.3286, + "step": 115890 + }, + { + "epoch": 7.8743715178692755, + "grad_norm": 0.9735327363014221, + "learning_rate": 1.5703560266340536e-05, + "loss": 3.3204, + "step": 115895 + }, + { + "epoch": 7.8747112379399375, + "grad_norm": 0.953525960445404, + "learning_rate": 1.5661095257507812e-05, + "loss": 3.2771, + "step": 115900 + }, + { + "epoch": 7.8750509580106, + "grad_norm": 0.7903362512588501, + "learning_rate": 1.5618630248675092e-05, + "loss": 3.2709, + "step": 115905 + }, + { + "epoch": 7.875390678081261, + "grad_norm": 0.855486273765564, + "learning_rate": 1.557616523984237e-05, + "loss": 3.5122, + "step": 115910 + }, + { + "epoch": 7.875730398151923, + "grad_norm": 0.9481089115142822, + "learning_rate": 1.553370023100965e-05, + "loss": 3.4177, + "step": 115915 + }, + { + "epoch": 7.876070118222585, + "grad_norm": 0.8967079520225525, + "learning_rate": 1.5491235222176926e-05, + "loss": 3.1976, + "step": 115920 + }, + { + "epoch": 7.876409838293246, + "grad_norm": 0.8644145727157593, + "learning_rate": 1.5448770213344206e-05, + "loss": 3.5683, + "step": 115925 + }, + { + "epoch": 7.876749558363908, + "grad_norm": 0.9408068060874939, + "learning_rate": 1.5406305204511483e-05, + "loss": 3.4466, + "step": 115930 + }, + { + "epoch": 7.87708927843457, + "grad_norm": 1.1728311777114868, + "learning_rate": 1.5363840195678763e-05, + "loss": 3.0541, + "step": 115935 + }, + { + "epoch": 7.8774289985052315, + "grad_norm": 0.968498170375824, + "learning_rate": 1.532137518684604e-05, + "loss": 3.1369, + "step": 115940 + }, + { + "epoch": 7.877768718575894, + "grad_norm": 0.7429043054580688, + "learning_rate": 1.5278910178013316e-05, + "loss": 3.4767, + "step": 115945 + }, + { + "epoch": 7.878108438646555, + "grad_norm": 0.8333841562271118, + "learning_rate": 1.5236445169180594e-05, + "loss": 3.371, + "step": 115950 + }, + { + "epoch": 7.878448158717217, + "grad_norm": 1.1009293794631958, + "learning_rate": 1.5193980160347873e-05, + "loss": 3.2606, + "step": 115955 + }, + { + "epoch": 7.878787878787879, + "grad_norm": 0.9611184000968933, + "learning_rate": 1.5151515151515153e-05, + "loss": 3.4462, + "step": 115960 + }, + { + "epoch": 7.87912759885854, + "grad_norm": 0.7863991856575012, + "learning_rate": 1.510905014268243e-05, + "loss": 3.4159, + "step": 115965 + }, + { + "epoch": 7.879467318929202, + "grad_norm": 0.9138818979263306, + "learning_rate": 1.5066585133849708e-05, + "loss": 3.3111, + "step": 115970 + }, + { + "epoch": 7.879807038999864, + "grad_norm": 1.1962251663208008, + "learning_rate": 1.5024120125016986e-05, + "loss": 3.5322, + "step": 115975 + }, + { + "epoch": 7.880146759070525, + "grad_norm": 0.9184339046478271, + "learning_rate": 1.4981655116184265e-05, + "loss": 3.4904, + "step": 115980 + }, + { + "epoch": 7.8804864791411875, + "grad_norm": 0.8110573291778564, + "learning_rate": 1.4939190107351543e-05, + "loss": 3.2881, + "step": 115985 + }, + { + "epoch": 7.88082619921185, + "grad_norm": 1.4057265520095825, + "learning_rate": 1.4896725098518821e-05, + "loss": 3.4327, + "step": 115990 + }, + { + "epoch": 7.881165919282511, + "grad_norm": 0.9920687079429626, + "learning_rate": 1.4854260089686098e-05, + "loss": 3.3708, + "step": 115995 + }, + { + "epoch": 7.881505639353173, + "grad_norm": 1.6504878997802734, + "learning_rate": 1.4811795080853377e-05, + "loss": 3.351, + "step": 116000 + }, + { + "epoch": 7.881845359423835, + "grad_norm": 1.000727653503418, + "learning_rate": 1.4769330072020657e-05, + "loss": 3.2407, + "step": 116005 + }, + { + "epoch": 7.882185079494496, + "grad_norm": 0.7792304158210754, + "learning_rate": 1.4726865063187933e-05, + "loss": 3.3918, + "step": 116010 + }, + { + "epoch": 7.882524799565158, + "grad_norm": 1.1380507946014404, + "learning_rate": 1.4684400054355212e-05, + "loss": 3.4704, + "step": 116015 + }, + { + "epoch": 7.88286451963582, + "grad_norm": 1.029637098312378, + "learning_rate": 1.464193504552249e-05, + "loss": 3.5473, + "step": 116020 + }, + { + "epoch": 7.883204239706481, + "grad_norm": 0.9730058312416077, + "learning_rate": 1.4599470036689768e-05, + "loss": 3.5002, + "step": 116025 + }, + { + "epoch": 7.8835439597771435, + "grad_norm": 0.886907696723938, + "learning_rate": 1.4557005027857047e-05, + "loss": 3.4963, + "step": 116030 + }, + { + "epoch": 7.883883679847806, + "grad_norm": 0.7390263676643372, + "learning_rate": 1.4514540019024324e-05, + "loss": 3.444, + "step": 116035 + }, + { + "epoch": 7.884223399918467, + "grad_norm": 0.8614093065261841, + "learning_rate": 1.4472075010191602e-05, + "loss": 3.5679, + "step": 116040 + }, + { + "epoch": 7.884563119989129, + "grad_norm": 0.9878278970718384, + "learning_rate": 1.442961000135888e-05, + "loss": 3.0877, + "step": 116045 + }, + { + "epoch": 7.884902840059791, + "grad_norm": 1.2395403385162354, + "learning_rate": 1.4387144992526159e-05, + "loss": 3.565, + "step": 116050 + }, + { + "epoch": 7.885242560130452, + "grad_norm": 0.7843355536460876, + "learning_rate": 1.4344679983693437e-05, + "loss": 3.4503, + "step": 116055 + }, + { + "epoch": 7.885582280201114, + "grad_norm": 0.968789279460907, + "learning_rate": 1.4302214974860715e-05, + "loss": 3.4427, + "step": 116060 + }, + { + "epoch": 7.885922000271776, + "grad_norm": 1.0255775451660156, + "learning_rate": 1.4259749966027992e-05, + "loss": 3.4881, + "step": 116065 + }, + { + "epoch": 7.886261720342437, + "grad_norm": 0.8173617124557495, + "learning_rate": 1.4217284957195272e-05, + "loss": 3.4947, + "step": 116070 + }, + { + "epoch": 7.8866014404130995, + "grad_norm": 0.9437533020973206, + "learning_rate": 1.417481994836255e-05, + "loss": 3.2355, + "step": 116075 + }, + { + "epoch": 7.886941160483762, + "grad_norm": 1.139997959136963, + "learning_rate": 1.4132354939529827e-05, + "loss": 3.3406, + "step": 116080 + }, + { + "epoch": 7.887280880554423, + "grad_norm": 1.093639850616455, + "learning_rate": 1.4089889930697106e-05, + "loss": 3.3522, + "step": 116085 + }, + { + "epoch": 7.887620600625085, + "grad_norm": 1.0997406244277954, + "learning_rate": 1.4047424921864384e-05, + "loss": 3.3398, + "step": 116090 + }, + { + "epoch": 7.887960320695747, + "grad_norm": 0.9577206373214722, + "learning_rate": 1.4004959913031662e-05, + "loss": 3.3103, + "step": 116095 + }, + { + "epoch": 7.888300040766408, + "grad_norm": 1.0121787786483765, + "learning_rate": 1.396249490419894e-05, + "loss": 3.4641, + "step": 116100 + }, + { + "epoch": 7.88863976083707, + "grad_norm": 0.8642535209655762, + "learning_rate": 1.392002989536622e-05, + "loss": 3.4237, + "step": 116105 + }, + { + "epoch": 7.888979480907732, + "grad_norm": 0.8501123785972595, + "learning_rate": 1.3877564886533496e-05, + "loss": 3.0231, + "step": 116110 + }, + { + "epoch": 7.8893192009783935, + "grad_norm": 1.0056288242340088, + "learning_rate": 1.3835099877700776e-05, + "loss": 3.4425, + "step": 116115 + }, + { + "epoch": 7.8896589210490555, + "grad_norm": 0.895125150680542, + "learning_rate": 1.3792634868868053e-05, + "loss": 3.22, + "step": 116120 + }, + { + "epoch": 7.889998641119718, + "grad_norm": 0.8771841526031494, + "learning_rate": 1.3750169860035331e-05, + "loss": 3.2737, + "step": 116125 + }, + { + "epoch": 7.890338361190379, + "grad_norm": 0.8405683636665344, + "learning_rate": 1.370770485120261e-05, + "loss": 3.5593, + "step": 116130 + }, + { + "epoch": 7.890678081261041, + "grad_norm": 1.0541144609451294, + "learning_rate": 1.3665239842369886e-05, + "loss": 3.5166, + "step": 116135 + }, + { + "epoch": 7.891017801331703, + "grad_norm": 1.037105679512024, + "learning_rate": 1.3622774833537166e-05, + "loss": 3.1661, + "step": 116140 + }, + { + "epoch": 7.891357521402364, + "grad_norm": 0.9399018287658691, + "learning_rate": 1.3580309824704445e-05, + "loss": 3.1803, + "step": 116145 + }, + { + "epoch": 7.891697241473026, + "grad_norm": 1.1674349308013916, + "learning_rate": 1.3537844815871721e-05, + "loss": 3.3106, + "step": 116150 + }, + { + "epoch": 7.892036961543688, + "grad_norm": 0.7831346392631531, + "learning_rate": 1.3495379807039e-05, + "loss": 3.2909, + "step": 116155 + }, + { + "epoch": 7.8923766816143495, + "grad_norm": 1.100667119026184, + "learning_rate": 1.345291479820628e-05, + "loss": 3.374, + "step": 116160 + }, + { + "epoch": 7.8927164016850115, + "grad_norm": 0.9304864406585693, + "learning_rate": 1.3410449789373556e-05, + "loss": 3.4739, + "step": 116165 + }, + { + "epoch": 7.893056121755674, + "grad_norm": 0.754474401473999, + "learning_rate": 1.3367984780540835e-05, + "loss": 3.4389, + "step": 116170 + }, + { + "epoch": 7.893395841826335, + "grad_norm": 0.8526752591133118, + "learning_rate": 1.3325519771708113e-05, + "loss": 3.5185, + "step": 116175 + }, + { + "epoch": 7.893735561896997, + "grad_norm": 0.8931655883789062, + "learning_rate": 1.328305476287539e-05, + "loss": 3.4248, + "step": 116180 + }, + { + "epoch": 7.894075281967659, + "grad_norm": 0.8842072486877441, + "learning_rate": 1.324058975404267e-05, + "loss": 3.5429, + "step": 116185 + }, + { + "epoch": 7.89441500203832, + "grad_norm": 0.8182488083839417, + "learning_rate": 1.3198124745209947e-05, + "loss": 3.419, + "step": 116190 + }, + { + "epoch": 7.894754722108982, + "grad_norm": 1.0888721942901611, + "learning_rate": 1.3155659736377225e-05, + "loss": 3.2304, + "step": 116195 + }, + { + "epoch": 7.895094442179644, + "grad_norm": 1.0264707803726196, + "learning_rate": 1.3113194727544503e-05, + "loss": 3.3243, + "step": 116200 + }, + { + "epoch": 7.8954341622503055, + "grad_norm": 0.9690226912498474, + "learning_rate": 1.3070729718711782e-05, + "loss": 3.5109, + "step": 116205 + }, + { + "epoch": 7.8957738823209676, + "grad_norm": 0.8066064119338989, + "learning_rate": 1.302826470987906e-05, + "loss": 3.1403, + "step": 116210 + }, + { + "epoch": 7.89611360239163, + "grad_norm": 0.8047357201576233, + "learning_rate": 1.2985799701046339e-05, + "loss": 3.4805, + "step": 116215 + }, + { + "epoch": 7.896453322462291, + "grad_norm": 0.8546112775802612, + "learning_rate": 1.2943334692213615e-05, + "loss": 3.2297, + "step": 116220 + }, + { + "epoch": 7.896793042532953, + "grad_norm": 0.8248113393783569, + "learning_rate": 1.2900869683380894e-05, + "loss": 3.3832, + "step": 116225 + }, + { + "epoch": 7.897132762603615, + "grad_norm": 1.1342116594314575, + "learning_rate": 1.2858404674548174e-05, + "loss": 3.2245, + "step": 116230 + }, + { + "epoch": 7.897472482674276, + "grad_norm": 0.970888614654541, + "learning_rate": 1.281593966571545e-05, + "loss": 3.2168, + "step": 116235 + }, + { + "epoch": 7.897812202744938, + "grad_norm": 0.969092845916748, + "learning_rate": 1.2773474656882729e-05, + "loss": 3.1847, + "step": 116240 + }, + { + "epoch": 7.8981519228156, + "grad_norm": 0.9926217794418335, + "learning_rate": 1.2731009648050007e-05, + "loss": 3.3868, + "step": 116245 + }, + { + "epoch": 7.8984916428862615, + "grad_norm": 0.8798980712890625, + "learning_rate": 1.2688544639217286e-05, + "loss": 3.3333, + "step": 116250 + }, + { + "epoch": 7.898831362956924, + "grad_norm": 0.9403606057167053, + "learning_rate": 1.2646079630384564e-05, + "loss": 3.2514, + "step": 116255 + }, + { + "epoch": 7.899171083027586, + "grad_norm": 0.902208149433136, + "learning_rate": 1.2603614621551842e-05, + "loss": 3.0811, + "step": 116260 + }, + { + "epoch": 7.899510803098247, + "grad_norm": 0.7908784747123718, + "learning_rate": 1.2561149612719119e-05, + "loss": 3.5496, + "step": 116265 + }, + { + "epoch": 7.899850523168909, + "grad_norm": 1.02617347240448, + "learning_rate": 1.2518684603886397e-05, + "loss": 3.6049, + "step": 116270 + }, + { + "epoch": 7.900190243239571, + "grad_norm": 1.312915563583374, + "learning_rate": 1.2476219595053676e-05, + "loss": 3.3431, + "step": 116275 + }, + { + "epoch": 7.900529963310232, + "grad_norm": 0.9801099300384521, + "learning_rate": 1.2433754586220954e-05, + "loss": 3.397, + "step": 116280 + }, + { + "epoch": 7.900869683380894, + "grad_norm": 1.157461166381836, + "learning_rate": 1.2391289577388233e-05, + "loss": 3.4393, + "step": 116285 + }, + { + "epoch": 7.901209403451556, + "grad_norm": 0.8412001132965088, + "learning_rate": 1.234882456855551e-05, + "loss": 3.4782, + "step": 116290 + }, + { + "epoch": 7.9015491235222175, + "grad_norm": 0.9981868863105774, + "learning_rate": 1.230635955972279e-05, + "loss": 3.3793, + "step": 116295 + }, + { + "epoch": 7.90188884359288, + "grad_norm": 1.2528963088989258, + "learning_rate": 1.2263894550890068e-05, + "loss": 3.3021, + "step": 116300 + }, + { + "epoch": 7.902228563663542, + "grad_norm": 0.82230544090271, + "learning_rate": 1.2221429542057344e-05, + "loss": 3.1515, + "step": 116305 + }, + { + "epoch": 7.902568283734203, + "grad_norm": 1.1834254264831543, + "learning_rate": 1.2178964533224623e-05, + "loss": 3.3279, + "step": 116310 + }, + { + "epoch": 7.902908003804865, + "grad_norm": 0.8819334506988525, + "learning_rate": 1.2136499524391901e-05, + "loss": 3.3466, + "step": 116315 + }, + { + "epoch": 7.903247723875527, + "grad_norm": 0.9079425930976868, + "learning_rate": 1.209403451555918e-05, + "loss": 3.5111, + "step": 116320 + }, + { + "epoch": 7.903587443946188, + "grad_norm": 0.8851118087768555, + "learning_rate": 1.2051569506726458e-05, + "loss": 3.5183, + "step": 116325 + }, + { + "epoch": 7.90392716401685, + "grad_norm": 0.7290522456169128, + "learning_rate": 1.2009104497893736e-05, + "loss": 2.9273, + "step": 116330 + }, + { + "epoch": 7.904266884087512, + "grad_norm": 0.8187059164047241, + "learning_rate": 1.1966639489061013e-05, + "loss": 3.4412, + "step": 116335 + }, + { + "epoch": 7.9046066041581735, + "grad_norm": 0.9912301301956177, + "learning_rate": 1.1924174480228293e-05, + "loss": 3.3999, + "step": 116340 + }, + { + "epoch": 7.904946324228836, + "grad_norm": 0.9212816953659058, + "learning_rate": 1.1881709471395571e-05, + "loss": 3.1428, + "step": 116345 + }, + { + "epoch": 7.905286044299498, + "grad_norm": 1.0085595846176147, + "learning_rate": 1.1839244462562848e-05, + "loss": 3.5854, + "step": 116350 + }, + { + "epoch": 7.905625764370159, + "grad_norm": 0.9472259879112244, + "learning_rate": 1.1796779453730126e-05, + "loss": 3.4241, + "step": 116355 + }, + { + "epoch": 7.905965484440821, + "grad_norm": 0.832780659198761, + "learning_rate": 1.1754314444897403e-05, + "loss": 3.3403, + "step": 116360 + }, + { + "epoch": 7.906305204511483, + "grad_norm": 1.0243120193481445, + "learning_rate": 1.1711849436064683e-05, + "loss": 3.5587, + "step": 116365 + }, + { + "epoch": 7.906644924582144, + "grad_norm": 0.8870436549186707, + "learning_rate": 1.1669384427231962e-05, + "loss": 3.4784, + "step": 116370 + }, + { + "epoch": 7.906984644652806, + "grad_norm": 0.8798761367797852, + "learning_rate": 1.1626919418399238e-05, + "loss": 3.1403, + "step": 116375 + }, + { + "epoch": 7.907324364723468, + "grad_norm": 0.7565405368804932, + "learning_rate": 1.1584454409566517e-05, + "loss": 3.1942, + "step": 116380 + }, + { + "epoch": 7.9076640847941295, + "grad_norm": 1.0169252157211304, + "learning_rate": 1.1541989400733797e-05, + "loss": 3.4277, + "step": 116385 + }, + { + "epoch": 7.908003804864792, + "grad_norm": 0.9257920384407043, + "learning_rate": 1.1499524391901073e-05, + "loss": 3.3046, + "step": 116390 + }, + { + "epoch": 7.908343524935454, + "grad_norm": 0.7543999552726746, + "learning_rate": 1.1457059383068352e-05, + "loss": 3.4168, + "step": 116395 + }, + { + "epoch": 7.908683245006115, + "grad_norm": 0.9582787156105042, + "learning_rate": 1.141459437423563e-05, + "loss": 3.312, + "step": 116400 + }, + { + "epoch": 7.909022965076777, + "grad_norm": 0.8955578207969666, + "learning_rate": 1.1372129365402907e-05, + "loss": 3.3775, + "step": 116405 + }, + { + "epoch": 7.909362685147439, + "grad_norm": 1.2059531211853027, + "learning_rate": 1.1329664356570187e-05, + "loss": 3.3689, + "step": 116410 + }, + { + "epoch": 7.9097024052181, + "grad_norm": 0.7103554010391235, + "learning_rate": 1.1287199347737465e-05, + "loss": 3.272, + "step": 116415 + }, + { + "epoch": 7.910042125288762, + "grad_norm": 1.0887857675552368, + "learning_rate": 1.1244734338904742e-05, + "loss": 3.3495, + "step": 116420 + }, + { + "epoch": 7.9103818453594235, + "grad_norm": 1.0334372520446777, + "learning_rate": 1.120226933007202e-05, + "loss": 3.4262, + "step": 116425 + }, + { + "epoch": 7.9107215654300855, + "grad_norm": 1.0910909175872803, + "learning_rate": 1.11598043212393e-05, + "loss": 3.6177, + "step": 116430 + }, + { + "epoch": 7.911061285500748, + "grad_norm": 0.9447363018989563, + "learning_rate": 1.1117339312406577e-05, + "loss": 3.4448, + "step": 116435 + }, + { + "epoch": 7.911401005571409, + "grad_norm": 0.7630108594894409, + "learning_rate": 1.1074874303573856e-05, + "loss": 3.4714, + "step": 116440 + }, + { + "epoch": 7.911740725642071, + "grad_norm": 0.8881083130836487, + "learning_rate": 1.1032409294741132e-05, + "loss": 3.2898, + "step": 116445 + }, + { + "epoch": 7.912080445712733, + "grad_norm": 0.9479274153709412, + "learning_rate": 1.098994428590841e-05, + "loss": 3.3803, + "step": 116450 + }, + { + "epoch": 7.912420165783394, + "grad_norm": 0.9609630107879639, + "learning_rate": 1.094747927707569e-05, + "loss": 3.5971, + "step": 116455 + }, + { + "epoch": 7.912759885854056, + "grad_norm": 0.8990363478660583, + "learning_rate": 1.0905014268242967e-05, + "loss": 3.5049, + "step": 116460 + }, + { + "epoch": 7.913099605924718, + "grad_norm": 1.0314533710479736, + "learning_rate": 1.0862549259410246e-05, + "loss": 3.2326, + "step": 116465 + }, + { + "epoch": 7.9134393259953795, + "grad_norm": 1.1155558824539185, + "learning_rate": 1.0820084250577524e-05, + "loss": 3.3462, + "step": 116470 + }, + { + "epoch": 7.9137790460660415, + "grad_norm": 0.8047475218772888, + "learning_rate": 1.0777619241744803e-05, + "loss": 3.6743, + "step": 116475 + }, + { + "epoch": 7.914118766136704, + "grad_norm": 0.8101286888122559, + "learning_rate": 1.0735154232912081e-05, + "loss": 3.2304, + "step": 116480 + }, + { + "epoch": 7.914458486207365, + "grad_norm": 0.9542962312698364, + "learning_rate": 1.069268922407936e-05, + "loss": 3.5306, + "step": 116485 + }, + { + "epoch": 7.914798206278027, + "grad_norm": 1.2630078792572021, + "learning_rate": 1.0650224215246636e-05, + "loss": 3.4686, + "step": 116490 + }, + { + "epoch": 7.915137926348689, + "grad_norm": 0.8135185241699219, + "learning_rate": 1.0607759206413914e-05, + "loss": 3.1497, + "step": 116495 + }, + { + "epoch": 7.91547764641935, + "grad_norm": 1.165574073791504, + "learning_rate": 1.0565294197581195e-05, + "loss": 3.3609, + "step": 116500 + }, + { + "epoch": 7.915817366490012, + "grad_norm": 1.0405018329620361, + "learning_rate": 1.0522829188748471e-05, + "loss": 3.4351, + "step": 116505 + }, + { + "epoch": 7.916157086560674, + "grad_norm": 0.7651498317718506, + "learning_rate": 1.048036417991575e-05, + "loss": 3.2155, + "step": 116510 + }, + { + "epoch": 7.9164968066313355, + "grad_norm": 0.8433520197868347, + "learning_rate": 1.0437899171083028e-05, + "loss": 3.4223, + "step": 116515 + }, + { + "epoch": 7.916836526701998, + "grad_norm": 0.9173451066017151, + "learning_rate": 1.0395434162250306e-05, + "loss": 3.4539, + "step": 116520 + }, + { + "epoch": 7.91717624677266, + "grad_norm": 1.1709953546524048, + "learning_rate": 1.0352969153417585e-05, + "loss": 3.4682, + "step": 116525 + }, + { + "epoch": 7.917515966843321, + "grad_norm": 0.9710039496421814, + "learning_rate": 1.0310504144584861e-05, + "loss": 3.346, + "step": 116530 + }, + { + "epoch": 7.917855686913983, + "grad_norm": 1.053895354270935, + "learning_rate": 1.026803913575214e-05, + "loss": 3.6771, + "step": 116535 + }, + { + "epoch": 7.918195406984645, + "grad_norm": 0.9125542640686035, + "learning_rate": 1.022557412691942e-05, + "loss": 3.524, + "step": 116540 + }, + { + "epoch": 7.918535127055306, + "grad_norm": 0.7748126983642578, + "learning_rate": 1.0183109118086697e-05, + "loss": 3.592, + "step": 116545 + }, + { + "epoch": 7.918874847125968, + "grad_norm": 0.757634162902832, + "learning_rate": 1.0140644109253975e-05, + "loss": 3.2934, + "step": 116550 + }, + { + "epoch": 7.91921456719663, + "grad_norm": 0.7898083329200745, + "learning_rate": 1.0098179100421253e-05, + "loss": 3.3093, + "step": 116555 + }, + { + "epoch": 7.9195542872672915, + "grad_norm": 1.049175500869751, + "learning_rate": 1.005571409158853e-05, + "loss": 3.3408, + "step": 116560 + }, + { + "epoch": 7.919894007337954, + "grad_norm": 1.1482396125793457, + "learning_rate": 1.001324908275581e-05, + "loss": 3.3459, + "step": 116565 + }, + { + "epoch": 7.920233727408616, + "grad_norm": 1.0461783409118652, + "learning_rate": 9.970784073923088e-06, + "loss": 3.2872, + "step": 116570 + }, + { + "epoch": 7.920573447479277, + "grad_norm": 0.8091289401054382, + "learning_rate": 9.928319065090365e-06, + "loss": 3.5216, + "step": 116575 + }, + { + "epoch": 7.920913167549939, + "grad_norm": 0.9465789198875427, + "learning_rate": 9.885854056257644e-06, + "loss": 3.1783, + "step": 116580 + }, + { + "epoch": 7.921252887620601, + "grad_norm": 1.041419506072998, + "learning_rate": 9.843389047424924e-06, + "loss": 3.4625, + "step": 116585 + }, + { + "epoch": 7.921592607691262, + "grad_norm": 1.1957157850265503, + "learning_rate": 9.8009240385922e-06, + "loss": 3.2522, + "step": 116590 + }, + { + "epoch": 7.921932327761924, + "grad_norm": 0.8935316205024719, + "learning_rate": 9.758459029759479e-06, + "loss": 3.4103, + "step": 116595 + }, + { + "epoch": 7.922272047832586, + "grad_norm": 0.7957669496536255, + "learning_rate": 9.715994020926755e-06, + "loss": 3.3366, + "step": 116600 + }, + { + "epoch": 7.9226117679032475, + "grad_norm": 0.7131280899047852, + "learning_rate": 9.673529012094034e-06, + "loss": 3.4133, + "step": 116605 + }, + { + "epoch": 7.92295148797391, + "grad_norm": 1.1158721446990967, + "learning_rate": 9.631064003261314e-06, + "loss": 3.1345, + "step": 116610 + }, + { + "epoch": 7.923291208044572, + "grad_norm": 0.9798750877380371, + "learning_rate": 9.58859899442859e-06, + "loss": 3.4314, + "step": 116615 + }, + { + "epoch": 7.923630928115233, + "grad_norm": 0.9280117154121399, + "learning_rate": 9.546133985595869e-06, + "loss": 3.6138, + "step": 116620 + }, + { + "epoch": 7.923970648185895, + "grad_norm": 1.2367814779281616, + "learning_rate": 9.503668976763147e-06, + "loss": 3.286, + "step": 116625 + }, + { + "epoch": 7.924310368256556, + "grad_norm": 0.8620408773422241, + "learning_rate": 9.461203967930426e-06, + "loss": 3.1177, + "step": 116630 + }, + { + "epoch": 7.924650088327218, + "grad_norm": 0.9883774518966675, + "learning_rate": 9.418738959097704e-06, + "loss": 3.3993, + "step": 116635 + }, + { + "epoch": 7.92498980839788, + "grad_norm": 0.9168908596038818, + "learning_rate": 9.376273950264982e-06, + "loss": 3.633, + "step": 116640 + }, + { + "epoch": 7.925329528468541, + "grad_norm": 0.9126530885696411, + "learning_rate": 9.33380894143226e-06, + "loss": 3.5473, + "step": 116645 + }, + { + "epoch": 7.9256692485392035, + "grad_norm": 0.9786132574081421, + "learning_rate": 9.291343932599538e-06, + "loss": 3.4364, + "step": 116650 + }, + { + "epoch": 7.926008968609866, + "grad_norm": 1.1496821641921997, + "learning_rate": 9.248878923766818e-06, + "loss": 3.5594, + "step": 116655 + }, + { + "epoch": 7.926348688680527, + "grad_norm": 0.9915406107902527, + "learning_rate": 9.206413914934094e-06, + "loss": 3.2977, + "step": 116660 + }, + { + "epoch": 7.926688408751189, + "grad_norm": 0.8908513784408569, + "learning_rate": 9.163948906101373e-06, + "loss": 3.5035, + "step": 116665 + }, + { + "epoch": 7.927028128821851, + "grad_norm": 0.9005915522575378, + "learning_rate": 9.121483897268651e-06, + "loss": 3.4625, + "step": 116670 + }, + { + "epoch": 7.927367848892512, + "grad_norm": 0.9213376045227051, + "learning_rate": 9.07901888843593e-06, + "loss": 3.5001, + "step": 116675 + }, + { + "epoch": 7.927707568963174, + "grad_norm": 1.1502889394760132, + "learning_rate": 9.036553879603208e-06, + "loss": 3.325, + "step": 116680 + }, + { + "epoch": 7.928047289033836, + "grad_norm": 0.9220241904258728, + "learning_rate": 8.994088870770485e-06, + "loss": 3.3364, + "step": 116685 + }, + { + "epoch": 7.9283870091044975, + "grad_norm": 1.0447853803634644, + "learning_rate": 8.951623861937763e-06, + "loss": 3.5196, + "step": 116690 + }, + { + "epoch": 7.9287267291751595, + "grad_norm": 0.9900646209716797, + "learning_rate": 8.909158853105041e-06, + "loss": 3.1454, + "step": 116695 + }, + { + "epoch": 7.929066449245822, + "grad_norm": 1.1344099044799805, + "learning_rate": 8.86669384427232e-06, + "loss": 3.4105, + "step": 116700 + }, + { + "epoch": 7.929406169316483, + "grad_norm": 1.1317803859710693, + "learning_rate": 8.824228835439598e-06, + "loss": 3.3264, + "step": 116705 + }, + { + "epoch": 7.929745889387145, + "grad_norm": 0.919607937335968, + "learning_rate": 8.781763826606876e-06, + "loss": 3.4154, + "step": 116710 + }, + { + "epoch": 7.930085609457807, + "grad_norm": 0.980049192905426, + "learning_rate": 8.739298817774153e-06, + "loss": 3.6081, + "step": 116715 + }, + { + "epoch": 7.930425329528468, + "grad_norm": 0.9300215840339661, + "learning_rate": 8.696833808941433e-06, + "loss": 3.1969, + "step": 116720 + }, + { + "epoch": 7.93076504959913, + "grad_norm": 1.0515718460083008, + "learning_rate": 8.654368800108712e-06, + "loss": 3.5782, + "step": 116725 + }, + { + "epoch": 7.931104769669792, + "grad_norm": 0.899366021156311, + "learning_rate": 8.611903791275988e-06, + "loss": 3.3287, + "step": 116730 + }, + { + "epoch": 7.9314444897404535, + "grad_norm": 1.0204436779022217, + "learning_rate": 8.569438782443267e-06, + "loss": 3.1524, + "step": 116735 + }, + { + "epoch": 7.9317842098111155, + "grad_norm": 0.8489725589752197, + "learning_rate": 8.526973773610545e-06, + "loss": 3.2551, + "step": 116740 + }, + { + "epoch": 7.932123929881778, + "grad_norm": 0.9045321941375732, + "learning_rate": 8.484508764777823e-06, + "loss": 3.4167, + "step": 116745 + }, + { + "epoch": 7.932463649952439, + "grad_norm": 1.044059157371521, + "learning_rate": 8.442043755945102e-06, + "loss": 3.1268, + "step": 116750 + }, + { + "epoch": 7.932803370023101, + "grad_norm": 1.116809368133545, + "learning_rate": 8.39957874711238e-06, + "loss": 3.3831, + "step": 116755 + }, + { + "epoch": 7.933143090093763, + "grad_norm": 0.8635019659996033, + "learning_rate": 8.357113738279657e-06, + "loss": 3.4304, + "step": 116760 + }, + { + "epoch": 7.933482810164424, + "grad_norm": 0.7758601903915405, + "learning_rate": 8.314648729446937e-06, + "loss": 3.4489, + "step": 116765 + }, + { + "epoch": 7.933822530235086, + "grad_norm": 1.2761722803115845, + "learning_rate": 8.272183720614214e-06, + "loss": 3.6635, + "step": 116770 + }, + { + "epoch": 7.934162250305748, + "grad_norm": 0.8660451769828796, + "learning_rate": 8.229718711781492e-06, + "loss": 3.4136, + "step": 116775 + }, + { + "epoch": 7.9345019703764095, + "grad_norm": 1.0811628103256226, + "learning_rate": 8.18725370294877e-06, + "loss": 3.0163, + "step": 116780 + }, + { + "epoch": 7.9348416904470715, + "grad_norm": 0.9553582072257996, + "learning_rate": 8.144788694116047e-06, + "loss": 3.5903, + "step": 116785 + }, + { + "epoch": 7.935181410517734, + "grad_norm": 0.9756330251693726, + "learning_rate": 8.102323685283327e-06, + "loss": 3.3173, + "step": 116790 + }, + { + "epoch": 7.935521130588395, + "grad_norm": 0.9023902416229248, + "learning_rate": 8.059858676450606e-06, + "loss": 3.4293, + "step": 116795 + }, + { + "epoch": 7.935860850659057, + "grad_norm": 0.9373165965080261, + "learning_rate": 8.017393667617882e-06, + "loss": 3.2695, + "step": 116800 + }, + { + "epoch": 7.936200570729719, + "grad_norm": 1.1906187534332275, + "learning_rate": 7.97492865878516e-06, + "loss": 3.5076, + "step": 116805 + }, + { + "epoch": 7.93654029080038, + "grad_norm": 0.7174252867698669, + "learning_rate": 7.93246364995244e-06, + "loss": 3.3858, + "step": 116810 + }, + { + "epoch": 7.936880010871042, + "grad_norm": 0.8776425123214722, + "learning_rate": 7.889998641119717e-06, + "loss": 3.4447, + "step": 116815 + }, + { + "epoch": 7.937219730941704, + "grad_norm": 0.8012998700141907, + "learning_rate": 7.847533632286996e-06, + "loss": 3.2479, + "step": 116820 + }, + { + "epoch": 7.9375594510123655, + "grad_norm": 0.9881122708320618, + "learning_rate": 7.805068623454274e-06, + "loss": 3.3318, + "step": 116825 + }, + { + "epoch": 7.937899171083028, + "grad_norm": 0.7949721217155457, + "learning_rate": 7.762603614621551e-06, + "loss": 3.2715, + "step": 116830 + }, + { + "epoch": 7.93823889115369, + "grad_norm": 1.0404679775238037, + "learning_rate": 7.720138605788831e-06, + "loss": 3.6821, + "step": 116835 + }, + { + "epoch": 7.938578611224351, + "grad_norm": 0.8906493186950684, + "learning_rate": 7.677673596956108e-06, + "loss": 3.141, + "step": 116840 + }, + { + "epoch": 7.938918331295013, + "grad_norm": 0.8593282699584961, + "learning_rate": 7.635208588123386e-06, + "loss": 3.4438, + "step": 116845 + }, + { + "epoch": 7.939258051365675, + "grad_norm": 0.9846568703651428, + "learning_rate": 7.592743579290665e-06, + "loss": 3.4544, + "step": 116850 + }, + { + "epoch": 7.939597771436336, + "grad_norm": 0.9080834984779358, + "learning_rate": 7.550278570457943e-06, + "loss": 3.1109, + "step": 116855 + }, + { + "epoch": 7.939937491506998, + "grad_norm": 0.857833981513977, + "learning_rate": 7.507813561625221e-06, + "loss": 3.5373, + "step": 116860 + }, + { + "epoch": 7.94027721157766, + "grad_norm": 1.2052174806594849, + "learning_rate": 7.4653485527924995e-06, + "loss": 3.4322, + "step": 116865 + }, + { + "epoch": 7.9406169316483215, + "grad_norm": 0.9748688340187073, + "learning_rate": 7.422883543959777e-06, + "loss": 3.3742, + "step": 116870 + }, + { + "epoch": 7.940956651718984, + "grad_norm": 1.0850085020065308, + "learning_rate": 7.3804185351270555e-06, + "loss": 3.5159, + "step": 116875 + }, + { + "epoch": 7.941296371789646, + "grad_norm": 0.8634036779403687, + "learning_rate": 7.337953526294333e-06, + "loss": 3.5795, + "step": 116880 + }, + { + "epoch": 7.941636091860307, + "grad_norm": 0.9484114646911621, + "learning_rate": 7.295488517461612e-06, + "loss": 3.4473, + "step": 116885 + }, + { + "epoch": 7.941975811930969, + "grad_norm": 0.8810225129127502, + "learning_rate": 7.25302350862889e-06, + "loss": 3.6015, + "step": 116890 + }, + { + "epoch": 7.942315532001631, + "grad_norm": 1.1251698732376099, + "learning_rate": 7.210558499796168e-06, + "loss": 3.2867, + "step": 116895 + }, + { + "epoch": 7.942655252072292, + "grad_norm": 0.9934365153312683, + "learning_rate": 7.1680934909634465e-06, + "loss": 3.4122, + "step": 116900 + }, + { + "epoch": 7.942994972142954, + "grad_norm": 0.7642443776130676, + "learning_rate": 7.125628482130725e-06, + "loss": 3.4922, + "step": 116905 + }, + { + "epoch": 7.943334692213616, + "grad_norm": 0.7167701721191406, + "learning_rate": 7.0831634732980025e-06, + "loss": 3.3455, + "step": 116910 + }, + { + "epoch": 7.9436744122842775, + "grad_norm": 0.9419924020767212, + "learning_rate": 7.04069846446528e-06, + "loss": 3.4923, + "step": 116915 + }, + { + "epoch": 7.94401413235494, + "grad_norm": 0.9468636512756348, + "learning_rate": 6.998233455632559e-06, + "loss": 3.3494, + "step": 116920 + }, + { + "epoch": 7.944353852425602, + "grad_norm": 0.9170337319374084, + "learning_rate": 6.955768446799837e-06, + "loss": 3.459, + "step": 116925 + }, + { + "epoch": 7.944693572496263, + "grad_norm": 0.964834988117218, + "learning_rate": 6.913303437967115e-06, + "loss": 3.2329, + "step": 116930 + }, + { + "epoch": 7.945033292566925, + "grad_norm": 0.8253658413887024, + "learning_rate": 6.8708384291343935e-06, + "loss": 3.6426, + "step": 116935 + }, + { + "epoch": 7.945373012637587, + "grad_norm": 0.8415079116821289, + "learning_rate": 6.828373420301672e-06, + "loss": 3.238, + "step": 116940 + }, + { + "epoch": 7.945712732708248, + "grad_norm": 0.8107269406318665, + "learning_rate": 6.7859084114689494e-06, + "loss": 3.4998, + "step": 116945 + }, + { + "epoch": 7.94605245277891, + "grad_norm": 1.0693302154541016, + "learning_rate": 6.743443402636228e-06, + "loss": 3.4555, + "step": 116950 + }, + { + "epoch": 7.946392172849572, + "grad_norm": 1.0729098320007324, + "learning_rate": 6.700978393803506e-06, + "loss": 3.4828, + "step": 116955 + }, + { + "epoch": 7.9467318929202335, + "grad_norm": 0.9820566773414612, + "learning_rate": 6.658513384970784e-06, + "loss": 3.4702, + "step": 116960 + }, + { + "epoch": 7.947071612990896, + "grad_norm": 0.852262556552887, + "learning_rate": 6.616048376138062e-06, + "loss": 3.3637, + "step": 116965 + }, + { + "epoch": 7.947411333061558, + "grad_norm": 0.7867757678031921, + "learning_rate": 6.5735833673053405e-06, + "loss": 3.2947, + "step": 116970 + }, + { + "epoch": 7.947751053132219, + "grad_norm": 0.9071044325828552, + "learning_rate": 6.531118358472619e-06, + "loss": 3.1471, + "step": 116975 + }, + { + "epoch": 7.948090773202881, + "grad_norm": 1.0466712713241577, + "learning_rate": 6.4886533496398964e-06, + "loss": 3.5548, + "step": 116980 + }, + { + "epoch": 7.948430493273543, + "grad_norm": 0.9598124027252197, + "learning_rate": 6.446188340807176e-06, + "loss": 3.4106, + "step": 116985 + }, + { + "epoch": 7.948770213344204, + "grad_norm": 1.2580794095993042, + "learning_rate": 6.403723331974453e-06, + "loss": 3.5428, + "step": 116990 + }, + { + "epoch": 7.949109933414866, + "grad_norm": 1.1790755987167358, + "learning_rate": 6.361258323141732e-06, + "loss": 3.375, + "step": 116995 + }, + { + "epoch": 7.949449653485528, + "grad_norm": 0.8624858260154724, + "learning_rate": 6.318793314309009e-06, + "loss": 3.4033, + "step": 117000 + }, + { + "epoch": 7.9497893735561895, + "grad_norm": 0.8241692185401917, + "learning_rate": 6.2763283054762875e-06, + "loss": 3.3585, + "step": 117005 + }, + { + "epoch": 7.950129093626852, + "grad_norm": 1.231656789779663, + "learning_rate": 6.233863296643566e-06, + "loss": 3.2411, + "step": 117010 + }, + { + "epoch": 7.950468813697514, + "grad_norm": 1.0513490438461304, + "learning_rate": 6.191398287810843e-06, + "loss": 3.3112, + "step": 117015 + }, + { + "epoch": 7.950808533768175, + "grad_norm": 0.6685965657234192, + "learning_rate": 6.148933278978123e-06, + "loss": 3.1061, + "step": 117020 + }, + { + "epoch": 7.951148253838837, + "grad_norm": 1.0533660650253296, + "learning_rate": 6.1064682701454e-06, + "loss": 2.8891, + "step": 117025 + }, + { + "epoch": 7.951487973909499, + "grad_norm": 0.9563286900520325, + "learning_rate": 6.0640032613126786e-06, + "loss": 3.3718, + "step": 117030 + }, + { + "epoch": 7.95182769398016, + "grad_norm": 1.2360289096832275, + "learning_rate": 6.021538252479956e-06, + "loss": 3.4311, + "step": 117035 + }, + { + "epoch": 7.952167414050822, + "grad_norm": 0.923394501209259, + "learning_rate": 5.979073243647235e-06, + "loss": 3.3402, + "step": 117040 + }, + { + "epoch": 7.952507134121484, + "grad_norm": 0.7249405384063721, + "learning_rate": 5.936608234814513e-06, + "loss": 3.5187, + "step": 117045 + }, + { + "epoch": 7.9528468541921455, + "grad_norm": 1.041886806488037, + "learning_rate": 5.89414322598179e-06, + "loss": 3.3704, + "step": 117050 + }, + { + "epoch": 7.953186574262808, + "grad_norm": 0.9308632016181946, + "learning_rate": 5.85167821714907e-06, + "loss": 3.334, + "step": 117055 + }, + { + "epoch": 7.95352629433347, + "grad_norm": 0.9283150434494019, + "learning_rate": 5.809213208316347e-06, + "loss": 3.3775, + "step": 117060 + }, + { + "epoch": 7.953866014404131, + "grad_norm": 0.7827629446983337, + "learning_rate": 5.7667481994836256e-06, + "loss": 3.1737, + "step": 117065 + }, + { + "epoch": 7.954205734474793, + "grad_norm": 0.880403459072113, + "learning_rate": 5.724283190650904e-06, + "loss": 3.3896, + "step": 117070 + }, + { + "epoch": 7.954545454545455, + "grad_norm": 0.8952151536941528, + "learning_rate": 5.681818181818182e-06, + "loss": 3.2054, + "step": 117075 + }, + { + "epoch": 7.954885174616116, + "grad_norm": 1.1582629680633545, + "learning_rate": 5.63935317298546e-06, + "loss": 3.3695, + "step": 117080 + }, + { + "epoch": 7.955224894686778, + "grad_norm": 1.2200863361358643, + "learning_rate": 5.596888164152738e-06, + "loss": 3.2734, + "step": 117085 + }, + { + "epoch": 7.95556461475744, + "grad_norm": 1.1265164613723755, + "learning_rate": 5.554423155320017e-06, + "loss": 3.2546, + "step": 117090 + }, + { + "epoch": 7.9559043348281016, + "grad_norm": 0.7346317768096924, + "learning_rate": 5.511958146487295e-06, + "loss": 3.516, + "step": 117095 + }, + { + "epoch": 7.956244054898764, + "grad_norm": 0.7067522406578064, + "learning_rate": 5.4694931376545725e-06, + "loss": 3.459, + "step": 117100 + }, + { + "epoch": 7.956583774969425, + "grad_norm": 0.7995157837867737, + "learning_rate": 5.427028128821851e-06, + "loss": 3.3017, + "step": 117105 + }, + { + "epoch": 7.956923495040087, + "grad_norm": 0.934941291809082, + "learning_rate": 5.384563119989129e-06, + "loss": 3.4066, + "step": 117110 + }, + { + "epoch": 7.957263215110749, + "grad_norm": 1.0386303663253784, + "learning_rate": 5.342098111156407e-06, + "loss": 3.3611, + "step": 117115 + }, + { + "epoch": 7.95760293518141, + "grad_norm": 0.9446911215782166, + "learning_rate": 5.299633102323685e-06, + "loss": 3.2839, + "step": 117120 + }, + { + "epoch": 7.957942655252072, + "grad_norm": 1.1061009168624878, + "learning_rate": 5.257168093490964e-06, + "loss": 3.2804, + "step": 117125 + }, + { + "epoch": 7.958282375322734, + "grad_norm": 1.6965149641036987, + "learning_rate": 5.214703084658242e-06, + "loss": 3.1062, + "step": 117130 + }, + { + "epoch": 7.9586220953933955, + "grad_norm": 0.707755982875824, + "learning_rate": 5.1722380758255195e-06, + "loss": 3.3531, + "step": 117135 + }, + { + "epoch": 7.958961815464058, + "grad_norm": 0.8550378680229187, + "learning_rate": 5.129773066992799e-06, + "loss": 3.5847, + "step": 117140 + }, + { + "epoch": 7.95930153553472, + "grad_norm": 1.0414197444915771, + "learning_rate": 5.087308058160076e-06, + "loss": 3.467, + "step": 117145 + }, + { + "epoch": 7.959641255605381, + "grad_norm": 1.161255955696106, + "learning_rate": 5.044843049327354e-06, + "loss": 3.3593, + "step": 117150 + }, + { + "epoch": 7.959980975676043, + "grad_norm": 1.0965960025787354, + "learning_rate": 5.002378040494633e-06, + "loss": 3.4398, + "step": 117155 + }, + { + "epoch": 7.960320695746705, + "grad_norm": 1.241086721420288, + "learning_rate": 4.959913031661911e-06, + "loss": 3.6161, + "step": 117160 + }, + { + "epoch": 7.960660415817366, + "grad_norm": 0.7522113919258118, + "learning_rate": 4.917448022829189e-06, + "loss": 3.4378, + "step": 117165 + }, + { + "epoch": 7.961000135888028, + "grad_norm": 1.0866358280181885, + "learning_rate": 4.8749830139964665e-06, + "loss": 3.3327, + "step": 117170 + }, + { + "epoch": 7.96133985595869, + "grad_norm": 1.1129308938980103, + "learning_rate": 4.832518005163746e-06, + "loss": 3.3706, + "step": 117175 + }, + { + "epoch": 7.9616795760293515, + "grad_norm": 0.8184992074966431, + "learning_rate": 4.790052996331023e-06, + "loss": 3.4187, + "step": 117180 + }, + { + "epoch": 7.962019296100014, + "grad_norm": 0.8610861897468567, + "learning_rate": 4.747587987498302e-06, + "loss": 3.2782, + "step": 117185 + }, + { + "epoch": 7.962359016170676, + "grad_norm": 0.9382168650627136, + "learning_rate": 4.70512297866558e-06, + "loss": 3.2928, + "step": 117190 + }, + { + "epoch": 7.962698736241337, + "grad_norm": 1.2094587087631226, + "learning_rate": 4.662657969832858e-06, + "loss": 3.1973, + "step": 117195 + }, + { + "epoch": 7.963038456311999, + "grad_norm": 0.795108437538147, + "learning_rate": 4.620192961000136e-06, + "loss": 3.5026, + "step": 117200 + }, + { + "epoch": 7.963378176382661, + "grad_norm": 0.8571358919143677, + "learning_rate": 4.5777279521674135e-06, + "loss": 3.3499, + "step": 117205 + }, + { + "epoch": 7.963717896453322, + "grad_norm": 1.0056043863296509, + "learning_rate": 4.535262943334693e-06, + "loss": 3.564, + "step": 117210 + }, + { + "epoch": 7.964057616523984, + "grad_norm": 0.95423823595047, + "learning_rate": 4.49279793450197e-06, + "loss": 3.5568, + "step": 117215 + }, + { + "epoch": 7.964397336594646, + "grad_norm": 2.031769275665283, + "learning_rate": 4.450332925669249e-06, + "loss": 3.3581, + "step": 117220 + }, + { + "epoch": 7.9647370566653075, + "grad_norm": 0.9947285056114197, + "learning_rate": 4.407867916836527e-06, + "loss": 3.4185, + "step": 117225 + }, + { + "epoch": 7.96507677673597, + "grad_norm": 0.754567563533783, + "learning_rate": 4.365402908003805e-06, + "loss": 3.4265, + "step": 117230 + }, + { + "epoch": 7.965416496806632, + "grad_norm": 0.8990226984024048, + "learning_rate": 4.322937899171083e-06, + "loss": 3.2396, + "step": 117235 + }, + { + "epoch": 7.965756216877293, + "grad_norm": 0.8965315222740173, + "learning_rate": 4.2804728903383605e-06, + "loss": 3.504, + "step": 117240 + }, + { + "epoch": 7.966095936947955, + "grad_norm": 0.9321632981300354, + "learning_rate": 4.23800788150564e-06, + "loss": 3.5851, + "step": 117245 + }, + { + "epoch": 7.966435657018617, + "grad_norm": 0.6918849349021912, + "learning_rate": 4.195542872672917e-06, + "loss": 3.5045, + "step": 117250 + }, + { + "epoch": 7.966775377089278, + "grad_norm": 0.8357621431350708, + "learning_rate": 4.153077863840196e-06, + "loss": 3.3252, + "step": 117255 + }, + { + "epoch": 7.96711509715994, + "grad_norm": 0.8325751423835754, + "learning_rate": 4.110612855007474e-06, + "loss": 3.0731, + "step": 117260 + }, + { + "epoch": 7.967454817230602, + "grad_norm": 0.9251959919929504, + "learning_rate": 4.068147846174752e-06, + "loss": 3.3126, + "step": 117265 + }, + { + "epoch": 7.9677945373012635, + "grad_norm": 0.8767516613006592, + "learning_rate": 4.02568283734203e-06, + "loss": 3.4261, + "step": 117270 + }, + { + "epoch": 7.968134257371926, + "grad_norm": 0.9840179681777954, + "learning_rate": 3.983217828509309e-06, + "loss": 3.3895, + "step": 117275 + }, + { + "epoch": 7.968473977442588, + "grad_norm": 0.8881438374519348, + "learning_rate": 3.940752819676587e-06, + "loss": 3.2782, + "step": 117280 + }, + { + "epoch": 7.968813697513249, + "grad_norm": 0.9147993922233582, + "learning_rate": 3.898287810843864e-06, + "loss": 3.3152, + "step": 117285 + }, + { + "epoch": 7.969153417583911, + "grad_norm": 0.7856869101524353, + "learning_rate": 3.855822802011143e-06, + "loss": 3.6197, + "step": 117290 + }, + { + "epoch": 7.969493137654573, + "grad_norm": 0.7834372520446777, + "learning_rate": 3.813357793178421e-06, + "loss": 3.5415, + "step": 117295 + }, + { + "epoch": 7.969832857725234, + "grad_norm": 1.0984309911727905, + "learning_rate": 3.7708927843456994e-06, + "loss": 3.3333, + "step": 117300 + }, + { + "epoch": 7.970172577795896, + "grad_norm": 1.0919263362884521, + "learning_rate": 3.7284277755129774e-06, + "loss": 3.5566, + "step": 117305 + }, + { + "epoch": 7.9705122978665575, + "grad_norm": 0.8635883331298828, + "learning_rate": 3.6859627666802553e-06, + "loss": 3.2925, + "step": 117310 + }, + { + "epoch": 7.9708520179372195, + "grad_norm": 0.890950620174408, + "learning_rate": 3.6434977578475337e-06, + "loss": 3.1284, + "step": 117315 + }, + { + "epoch": 7.971191738007882, + "grad_norm": 1.0664921998977661, + "learning_rate": 3.6010327490148117e-06, + "loss": 3.3645, + "step": 117320 + }, + { + "epoch": 7.971531458078543, + "grad_norm": 1.0163023471832275, + "learning_rate": 3.55856774018209e-06, + "loss": 3.3146, + "step": 117325 + }, + { + "epoch": 7.971871178149205, + "grad_norm": 1.0596181154251099, + "learning_rate": 3.5161027313493684e-06, + "loss": 3.3545, + "step": 117330 + }, + { + "epoch": 7.972210898219867, + "grad_norm": 0.8811275362968445, + "learning_rate": 3.4736377225166464e-06, + "loss": 3.5082, + "step": 117335 + }, + { + "epoch": 7.972550618290528, + "grad_norm": 0.9626274704933167, + "learning_rate": 3.4311727136839248e-06, + "loss": 3.5404, + "step": 117340 + }, + { + "epoch": 7.97289033836119, + "grad_norm": 0.801450252532959, + "learning_rate": 3.3887077048512027e-06, + "loss": 3.3834, + "step": 117345 + }, + { + "epoch": 7.973230058431852, + "grad_norm": 0.9591950178146362, + "learning_rate": 3.346242696018481e-06, + "loss": 3.0878, + "step": 117350 + }, + { + "epoch": 7.9735697785025135, + "grad_norm": 0.9341144561767578, + "learning_rate": 3.3037776871857586e-06, + "loss": 3.4549, + "step": 117355 + }, + { + "epoch": 7.9739094985731755, + "grad_norm": 0.9842540621757507, + "learning_rate": 3.261312678353037e-06, + "loss": 3.4394, + "step": 117360 + }, + { + "epoch": 7.974249218643838, + "grad_norm": 0.9357299208641052, + "learning_rate": 3.2188476695203154e-06, + "loss": 3.3428, + "step": 117365 + }, + { + "epoch": 7.974588938714499, + "grad_norm": 1.4678248167037964, + "learning_rate": 3.1763826606875934e-06, + "loss": 3.4715, + "step": 117370 + }, + { + "epoch": 7.974928658785161, + "grad_norm": 1.0750056505203247, + "learning_rate": 3.1339176518548718e-06, + "loss": 3.536, + "step": 117375 + }, + { + "epoch": 7.975268378855823, + "grad_norm": 1.0269533395767212, + "learning_rate": 3.0914526430221497e-06, + "loss": 3.3948, + "step": 117380 + }, + { + "epoch": 7.975608098926484, + "grad_norm": 0.7205071449279785, + "learning_rate": 3.048987634189428e-06, + "loss": 3.3953, + "step": 117385 + }, + { + "epoch": 7.975947818997146, + "grad_norm": 0.7449992299079895, + "learning_rate": 3.0065226253567065e-06, + "loss": 3.5169, + "step": 117390 + }, + { + "epoch": 7.976287539067808, + "grad_norm": 0.9857437610626221, + "learning_rate": 2.9640576165239844e-06, + "loss": 3.4438, + "step": 117395 + }, + { + "epoch": 7.9766272591384695, + "grad_norm": 0.9466061592102051, + "learning_rate": 2.9215926076912624e-06, + "loss": 3.2442, + "step": 117400 + }, + { + "epoch": 7.976966979209132, + "grad_norm": 0.9160231351852417, + "learning_rate": 2.8791275988585404e-06, + "loss": 3.41, + "step": 117405 + }, + { + "epoch": 7.977306699279794, + "grad_norm": 0.9187971949577332, + "learning_rate": 2.8366625900258187e-06, + "loss": 3.3849, + "step": 117410 + }, + { + "epoch": 7.977646419350455, + "grad_norm": 0.8400639891624451, + "learning_rate": 2.7941975811930967e-06, + "loss": 3.2882, + "step": 117415 + }, + { + "epoch": 7.977986139421117, + "grad_norm": 0.8512306213378906, + "learning_rate": 2.751732572360375e-06, + "loss": 3.3574, + "step": 117420 + }, + { + "epoch": 7.978325859491779, + "grad_norm": 0.7712929844856262, + "learning_rate": 2.7092675635276535e-06, + "loss": 3.7121, + "step": 117425 + }, + { + "epoch": 7.97866557956244, + "grad_norm": 1.1179834604263306, + "learning_rate": 2.6668025546949314e-06, + "loss": 3.3696, + "step": 117430 + }, + { + "epoch": 7.979005299633102, + "grad_norm": 1.2161744832992554, + "learning_rate": 2.62433754586221e-06, + "loss": 3.6275, + "step": 117435 + }, + { + "epoch": 7.979345019703764, + "grad_norm": 1.0528584718704224, + "learning_rate": 2.5818725370294878e-06, + "loss": 3.5379, + "step": 117440 + }, + { + "epoch": 7.9796847397744255, + "grad_norm": 0.9144910573959351, + "learning_rate": 2.5394075281967657e-06, + "loss": 3.6865, + "step": 117445 + }, + { + "epoch": 7.980024459845088, + "grad_norm": 1.1014667749404907, + "learning_rate": 2.496942519364044e-06, + "loss": 3.2717, + "step": 117450 + }, + { + "epoch": 7.98036417991575, + "grad_norm": 0.8425962328910828, + "learning_rate": 2.454477510531322e-06, + "loss": 3.2017, + "step": 117455 + }, + { + "epoch": 7.980703899986411, + "grad_norm": 0.9816661477088928, + "learning_rate": 2.4120125016986005e-06, + "loss": 3.5618, + "step": 117460 + }, + { + "epoch": 7.981043620057073, + "grad_norm": 0.8144105076789856, + "learning_rate": 2.3695474928658784e-06, + "loss": 3.364, + "step": 117465 + }, + { + "epoch": 7.981383340127735, + "grad_norm": 0.9697515964508057, + "learning_rate": 2.327082484033157e-06, + "loss": 3.3009, + "step": 117470 + }, + { + "epoch": 7.981723060198396, + "grad_norm": 0.9875323176383972, + "learning_rate": 2.2846174752004348e-06, + "loss": 3.4902, + "step": 117475 + }, + { + "epoch": 7.982062780269058, + "grad_norm": 0.8714432120323181, + "learning_rate": 2.242152466367713e-06, + "loss": 3.474, + "step": 117480 + }, + { + "epoch": 7.98240250033972, + "grad_norm": 0.8043555617332458, + "learning_rate": 2.1996874575349915e-06, + "loss": 3.374, + "step": 117485 + }, + { + "epoch": 7.9827422204103815, + "grad_norm": 0.976438581943512, + "learning_rate": 2.157222448702269e-06, + "loss": 3.6436, + "step": 117490 + }, + { + "epoch": 7.983081940481044, + "grad_norm": 1.240046501159668, + "learning_rate": 2.1147574398695474e-06, + "loss": 3.11, + "step": 117495 + }, + { + "epoch": 7.983421660551706, + "grad_norm": 0.8442407250404358, + "learning_rate": 2.0722924310368254e-06, + "loss": 3.4401, + "step": 117500 + }, + { + "epoch": 7.983761380622367, + "grad_norm": 0.8077302575111389, + "learning_rate": 2.029827422204104e-06, + "loss": 3.3518, + "step": 117505 + }, + { + "epoch": 7.984101100693029, + "grad_norm": 1.0180342197418213, + "learning_rate": 1.987362413371382e-06, + "loss": 3.7232, + "step": 117510 + }, + { + "epoch": 7.984440820763691, + "grad_norm": 0.9209889769554138, + "learning_rate": 1.94489740453866e-06, + "loss": 3.395, + "step": 117515 + }, + { + "epoch": 7.984780540834352, + "grad_norm": 0.8372121453285217, + "learning_rate": 1.9024323957059385e-06, + "loss": 3.5169, + "step": 117520 + }, + { + "epoch": 7.985120260905014, + "grad_norm": 0.7748185992240906, + "learning_rate": 1.8599673868732165e-06, + "loss": 3.3279, + "step": 117525 + }, + { + "epoch": 7.985459980975676, + "grad_norm": 0.8915907144546509, + "learning_rate": 1.8175023780404946e-06, + "loss": 2.9991, + "step": 117530 + }, + { + "epoch": 7.9857997010463375, + "grad_norm": 0.8457152247428894, + "learning_rate": 1.7750373692077728e-06, + "loss": 3.4123, + "step": 117535 + }, + { + "epoch": 7.986139421117, + "grad_norm": 0.9299027919769287, + "learning_rate": 1.732572360375051e-06, + "loss": 3.3462, + "step": 117540 + }, + { + "epoch": 7.986479141187662, + "grad_norm": 0.8148669004440308, + "learning_rate": 1.6901073515423292e-06, + "loss": 3.3617, + "step": 117545 + }, + { + "epoch": 7.986818861258323, + "grad_norm": 1.7525157928466797, + "learning_rate": 1.6476423427096073e-06, + "loss": 3.2856, + "step": 117550 + }, + { + "epoch": 7.987158581328985, + "grad_norm": 0.9342226982116699, + "learning_rate": 1.6051773338768855e-06, + "loss": 3.5598, + "step": 117555 + }, + { + "epoch": 7.987498301399647, + "grad_norm": 0.8415552973747253, + "learning_rate": 1.5627123250441637e-06, + "loss": 3.5028, + "step": 117560 + }, + { + "epoch": 7.987838021470308, + "grad_norm": 0.9067577123641968, + "learning_rate": 1.5202473162114418e-06, + "loss": 3.5499, + "step": 117565 + }, + { + "epoch": 7.98817774154097, + "grad_norm": 0.8860252499580383, + "learning_rate": 1.47778230737872e-06, + "loss": 3.1826, + "step": 117570 + }, + { + "epoch": 7.988517461611632, + "grad_norm": 1.0610146522521973, + "learning_rate": 1.435317298545998e-06, + "loss": 3.2648, + "step": 117575 + }, + { + "epoch": 7.9888571816822935, + "grad_norm": 0.9219157099723816, + "learning_rate": 1.3928522897132764e-06, + "loss": 3.2216, + "step": 117580 + }, + { + "epoch": 7.989196901752956, + "grad_norm": 1.0076751708984375, + "learning_rate": 1.3503872808805545e-06, + "loss": 3.1932, + "step": 117585 + }, + { + "epoch": 7.989536621823618, + "grad_norm": 1.2312184572219849, + "learning_rate": 1.3079222720478327e-06, + "loss": 3.3436, + "step": 117590 + }, + { + "epoch": 7.989876341894279, + "grad_norm": 0.8389595746994019, + "learning_rate": 1.2654572632151107e-06, + "loss": 3.3788, + "step": 117595 + }, + { + "epoch": 7.990216061964941, + "grad_norm": 0.9500887393951416, + "learning_rate": 1.2229922543823888e-06, + "loss": 3.5214, + "step": 117600 + }, + { + "epoch": 7.990555782035603, + "grad_norm": 0.9042538404464722, + "learning_rate": 1.180527245549667e-06, + "loss": 3.3474, + "step": 117605 + }, + { + "epoch": 7.990895502106264, + "grad_norm": 0.9139134287834167, + "learning_rate": 1.1380622367169454e-06, + "loss": 3.3077, + "step": 117610 + }, + { + "epoch": 7.991235222176926, + "grad_norm": 0.9732170104980469, + "learning_rate": 1.0955972278842236e-06, + "loss": 3.2478, + "step": 117615 + }, + { + "epoch": 7.991574942247588, + "grad_norm": 0.9352133274078369, + "learning_rate": 1.0531322190515015e-06, + "loss": 3.3771, + "step": 117620 + }, + { + "epoch": 7.9919146623182495, + "grad_norm": 1.0262112617492676, + "learning_rate": 1.0106672102187797e-06, + "loss": 3.5818, + "step": 117625 + }, + { + "epoch": 7.992254382388912, + "grad_norm": 0.7651551961898804, + "learning_rate": 9.682022013860579e-07, + "loss": 3.4427, + "step": 117630 + }, + { + "epoch": 7.992594102459574, + "grad_norm": 1.0805021524429321, + "learning_rate": 9.25737192553336e-07, + "loss": 3.3505, + "step": 117635 + }, + { + "epoch": 7.992933822530235, + "grad_norm": 0.9265249371528625, + "learning_rate": 8.832721837206142e-07, + "loss": 3.5632, + "step": 117640 + }, + { + "epoch": 7.993273542600897, + "grad_norm": 0.9067485928535461, + "learning_rate": 8.408071748878925e-07, + "loss": 3.1899, + "step": 117645 + }, + { + "epoch": 7.993613262671559, + "grad_norm": 1.059639573097229, + "learning_rate": 7.983421660551705e-07, + "loss": 3.2527, + "step": 117650 + }, + { + "epoch": 7.99395298274222, + "grad_norm": 0.8762375712394714, + "learning_rate": 7.558771572224487e-07, + "loss": 3.3769, + "step": 117655 + }, + { + "epoch": 7.994292702812882, + "grad_norm": 0.8246251940727234, + "learning_rate": 7.134121483897269e-07, + "loss": 3.2435, + "step": 117660 + }, + { + "epoch": 7.994632422883544, + "grad_norm": 1.3981503248214722, + "learning_rate": 6.709471395570051e-07, + "loss": 3.2458, + "step": 117665 + }, + { + "epoch": 7.9949721429542056, + "grad_norm": 1.0905605554580688, + "learning_rate": 6.284821307242831e-07, + "loss": 3.2317, + "step": 117670 + }, + { + "epoch": 7.995311863024868, + "grad_norm": 0.7600978016853333, + "learning_rate": 5.860171218915614e-07, + "loss": 3.2222, + "step": 117675 + }, + { + "epoch": 7.99565158309553, + "grad_norm": 0.8162209391593933, + "learning_rate": 5.435521130588396e-07, + "loss": 3.2217, + "step": 117680 + }, + { + "epoch": 7.995991303166191, + "grad_norm": 0.8791228532791138, + "learning_rate": 5.010871042261176e-07, + "loss": 3.4581, + "step": 117685 + }, + { + "epoch": 7.996331023236853, + "grad_norm": 1.0229488611221313, + "learning_rate": 4.586220953933958e-07, + "loss": 3.4328, + "step": 117690 + }, + { + "epoch": 7.996670743307515, + "grad_norm": 0.7252069711685181, + "learning_rate": 4.16157086560674e-07, + "loss": 3.4484, + "step": 117695 + }, + { + "epoch": 7.997010463378176, + "grad_norm": 0.956166684627533, + "learning_rate": 3.736920777279522e-07, + "loss": 3.4358, + "step": 117700 + }, + { + "epoch": 7.997350183448838, + "grad_norm": 0.6945189833641052, + "learning_rate": 3.3122706889523033e-07, + "loss": 3.2222, + "step": 117705 + }, + { + "epoch": 7.9976899035195, + "grad_norm": 0.9757670164108276, + "learning_rate": 2.887620600625085e-07, + "loss": 3.2734, + "step": 117710 + }, + { + "epoch": 7.998029623590162, + "grad_norm": 0.9869692325592041, + "learning_rate": 2.4629705122978667e-07, + "loss": 3.332, + "step": 117715 + }, + { + "epoch": 7.998369343660824, + "grad_norm": 1.0143767595291138, + "learning_rate": 2.0383204239706481e-07, + "loss": 3.3149, + "step": 117720 + }, + { + "epoch": 7.998709063731486, + "grad_norm": 0.9417544603347778, + "learning_rate": 1.6136703356434299e-07, + "loss": 3.4784, + "step": 117725 + }, + { + "epoch": 7.999048783802147, + "grad_norm": 1.039652943611145, + "learning_rate": 1.1890202473162116e-07, + "loss": 3.5618, + "step": 117730 + }, + { + "epoch": 7.999388503872809, + "grad_norm": 1.0154411792755127, + "learning_rate": 7.64370158988993e-08, + "loss": 3.2547, + "step": 117735 + }, + { + "epoch": 7.999728223943471, + "grad_norm": 0.9872223138809204, + "learning_rate": 3.3972007066177473e-08, + "loss": 3.5329, + "step": 117740 + }, + { + "epoch": 8.0, + "eval_bertscore": { + "f1": 0.8416534208814723, + "precision": 0.8448592209980024, + "recall": 0.8392182110538662 + }, + "eval_bleu_4": 0.018096643140168338, + "eval_exact_match": 0.000678360306231224, + "eval_loss": 3.3866779804229736, + "eval_meteor": 0.08833788525644948, + "eval_rouge": { + "rouge1": 0.12877786154270793, + "rouge2": 0.02000782267977022, + "rougeL": 0.11199251991643619, + "rougeLsum": 0.11215112315766845 + }, + "eval_runtime": 2186.5532, + "eval_samples_per_second": 4.719, + "eval_steps_per_second": 0.59, + "step": 117744 + } + ], + "logging_steps": 5, + "max_steps": 117744, + "num_input_tokens_seen": 0, + "num_train_epochs": 8, + "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": 5.787048068020961e+17, + "train_batch_size": 8, + "trial_name": null, + "trial_params": null +}