{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.9166666666666666, "eval_steps": 500, "global_step": 5500, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 1.1998956203460693, "learning_rate": 1.5e-06, "loss": 0.0948, "step": 10 }, { "grad_norm": 1.149483561515808, "learning_rate": 3.1666666666666667e-06, "loss": 0.1133, "step": 20 }, { "grad_norm": 0.5785359144210815, "learning_rate": 4.833333333333333e-06, "loss": 0.0921, "step": 30 }, { "grad_norm": 0.5978452563285828, "learning_rate": 6.5000000000000004e-06, "loss": 0.0424, "step": 40 }, { "grad_norm": 0.5148229598999023, "learning_rate": 8.166666666666666e-06, "loss": 0.0314, "step": 50 }, { "grad_norm": 0.456398606300354, "learning_rate": 9.833333333333333e-06, "loss": 0.0311, "step": 60 }, { "grad_norm": 0.2702748775482178, "learning_rate": 1.1500000000000002e-05, "loss": 0.0249, "step": 70 }, { "grad_norm": 0.2341022789478302, "learning_rate": 1.3166666666666667e-05, "loss": 0.0208, "step": 80 }, { "grad_norm": 0.24386702477931976, "learning_rate": 1.4833333333333334e-05, "loss": 0.0168, "step": 90 }, { "grad_norm": 0.15167136490345, "learning_rate": 1.65e-05, "loss": 0.0114, "step": 100 }, { "grad_norm": 0.12867538630962372, "learning_rate": 1.8166666666666667e-05, "loss": 0.015, "step": 110 }, { "grad_norm": 0.11734911054372787, "learning_rate": 1.9833333333333332e-05, "loss": 0.0108, "step": 120 }, { "grad_norm": 0.10739504545927048, "learning_rate": 2.15e-05, "loss": 0.0085, "step": 130 }, { "grad_norm": 0.11796178668737411, "learning_rate": 2.316666666666667e-05, "loss": 0.0108, "step": 140 }, { "grad_norm": 0.1394447386264801, "learning_rate": 2.483333333333333e-05, "loss": 0.0111, "step": 150 }, { "grad_norm": 0.1239742562174797, "learning_rate": 2.65e-05, "loss": 0.01, "step": 160 }, { "grad_norm": 0.15108101069927216, "learning_rate": 2.8166666666666666e-05, "loss": 0.0088, "step": 170 }, { "grad_norm": 0.10625599324703217, "learning_rate": 2.9833333333333335e-05, "loss": 0.0073, "step": 180 }, { "grad_norm": 0.18963077664375305, "learning_rate": 2.9999822989428527e-05, "loss": 0.0086, "step": 190 }, { "grad_norm": 0.13748835027217865, "learning_rate": 2.9999211106396838e-05, "loss": 0.008, "step": 200 }, { "grad_norm": 0.11299511790275574, "learning_rate": 2.9998162183413746e-05, "loss": 0.0084, "step": 210 }, { "grad_norm": 0.09556759148836136, "learning_rate": 2.9996676251042288e-05, "loss": 0.0075, "step": 220 }, { "grad_norm": 0.13729563355445862, "learning_rate": 2.9994753352578854e-05, "loss": 0.0078, "step": 230 }, { "grad_norm": 0.13592101633548737, "learning_rate": 2.999239354405195e-05, "loss": 0.0087, "step": 240 }, { "grad_norm": 0.13999280333518982, "learning_rate": 2.9989596894220574e-05, "loss": 0.0072, "step": 250 }, { "grad_norm": 0.20841121673583984, "learning_rate": 2.998636348457219e-05, "loss": 0.0078, "step": 260 }, { "grad_norm": 0.1364365965127945, "learning_rate": 2.998269340932036e-05, "loss": 0.0086, "step": 270 }, { "grad_norm": 0.1565951108932495, "learning_rate": 2.9978586775402017e-05, "loss": 0.0066, "step": 280 }, { "grad_norm": 0.1358197182416916, "learning_rate": 2.9974043702474322e-05, "loss": 0.0057, "step": 290 }, { "grad_norm": 0.11862654238939285, "learning_rate": 2.9969064322911198e-05, "loss": 0.0074, "step": 300 }, { "grad_norm": 0.13869383931159973, "learning_rate": 2.9963648781799457e-05, "loss": 0.0073, "step": 310 }, { "grad_norm": 0.10812313109636307, "learning_rate": 2.9957797236934595e-05, "loss": 0.0055, "step": 320 }, { "grad_norm": 0.10657292604446411, "learning_rate": 2.995150985881616e-05, "loss": 0.0066, "step": 330 }, { "grad_norm": 0.1428467333316803, "learning_rate": 2.9944786830642822e-05, "loss": 0.0075, "step": 340 }, { "grad_norm": 0.2037973701953888, "learning_rate": 2.9937628348307015e-05, "loss": 0.0068, "step": 350 }, { "grad_norm": 0.17431488633155823, "learning_rate": 2.9930034620389217e-05, "loss": 0.0085, "step": 360 }, { "grad_norm": 0.11183494329452515, "learning_rate": 2.9922005868151907e-05, "loss": 0.0075, "step": 370 }, { "grad_norm": 0.15082184970378876, "learning_rate": 2.991354232553308e-05, "loss": 0.0062, "step": 380 }, { "grad_norm": 0.18662261962890625, "learning_rate": 2.9904644239139453e-05, "loss": 0.0065, "step": 390 }, { "grad_norm": 0.08828644454479218, "learning_rate": 2.9895311868239293e-05, "loss": 0.0063, "step": 400 }, { "grad_norm": 0.09585582464933395, "learning_rate": 2.988554548475481e-05, "loss": 0.0085, "step": 410 }, { "grad_norm": 0.1154177337884903, "learning_rate": 2.9875345373254302e-05, "loss": 0.0062, "step": 420 }, { "grad_norm": 0.13848039507865906, "learning_rate": 2.986471183094382e-05, "loss": 0.0072, "step": 430 }, { "grad_norm": 0.11135748773813248, "learning_rate": 2.9853645167658494e-05, "loss": 0.0052, "step": 440 }, { "grad_norm": 0.15800616145133972, "learning_rate": 2.9842145705853565e-05, "loss": 0.0057, "step": 450 }, { "grad_norm": 0.11519774794578552, "learning_rate": 2.9830213780594925e-05, "loss": 0.0041, "step": 460 }, { "grad_norm": 0.06121053546667099, "learning_rate": 2.9817849739549408e-05, "loss": 0.0045, "step": 470 }, { "grad_norm": 0.11012984067201614, "learning_rate": 2.980505394297461e-05, "loss": 0.0047, "step": 480 }, { "grad_norm": 0.12504155933856964, "learning_rate": 2.9791826763708432e-05, "loss": 0.0043, "step": 490 }, { "grad_norm": 0.10626716166734695, "learning_rate": 2.9778168587158196e-05, "loss": 0.005, "step": 500 }, { "grad_norm": 0.100404754281044, "learning_rate": 2.976407981128942e-05, "loss": 0.0037, "step": 510 }, { "grad_norm": 0.11580510437488556, "learning_rate": 2.9749560846614216e-05, "loss": 0.0042, "step": 520 }, { "grad_norm": 0.1404998004436493, "learning_rate": 2.973461211617935e-05, "loss": 0.0055, "step": 530 }, { "grad_norm": 0.1066383644938469, "learning_rate": 2.9719234055553884e-05, "loss": 0.0055, "step": 540 }, { "grad_norm": 0.16468606889247894, "learning_rate": 2.970342711281651e-05, "loss": 0.0038, "step": 550 }, { "grad_norm": 0.1270781308412552, "learning_rate": 2.9687191748542485e-05, "loss": 0.0056, "step": 560 }, { "grad_norm": 0.13871441781520844, "learning_rate": 2.967052843579021e-05, "loss": 0.0079, "step": 570 }, { "grad_norm": 0.13760028779506683, "learning_rate": 2.965343766008745e-05, "loss": 0.0047, "step": 580 }, { "grad_norm": 0.23738646507263184, "learning_rate": 2.9635919919417175e-05, "loss": 0.0058, "step": 590 }, { "grad_norm": 0.12334813177585602, "learning_rate": 2.9617975724203065e-05, "loss": 0.006, "step": 600 }, { "grad_norm": 0.11787013709545135, "learning_rate": 2.9599605597294635e-05, "loss": 0.0042, "step": 610 }, { "grad_norm": 0.10075175762176514, "learning_rate": 2.958081007395199e-05, "loss": 0.0051, "step": 620 }, { "grad_norm": 0.12161751091480255, "learning_rate": 2.956158970183024e-05, "loss": 0.006, "step": 630 }, { "grad_norm": 0.1339486688375473, "learning_rate": 2.954194504096355e-05, "loss": 0.0053, "step": 640 }, { "grad_norm": 0.09474559128284454, "learning_rate": 2.9521876663748787e-05, "loss": 0.0069, "step": 650 }, { "grad_norm": 0.10630160570144653, "learning_rate": 2.950138515492888e-05, "loss": 0.0053, "step": 660 }, { "grad_norm": 0.1059340387582779, "learning_rate": 2.9480471111575768e-05, "loss": 0.0047, "step": 670 }, { "grad_norm": 0.13061285018920898, "learning_rate": 2.9459135143073e-05, "loss": 0.0044, "step": 680 }, { "grad_norm": 0.11301890015602112, "learning_rate": 2.9437377871097974e-05, "loss": 0.004, "step": 690 }, { "grad_norm": 0.11071354895830154, "learning_rate": 2.9415199929603843e-05, "loss": 0.0048, "step": 700 }, { "grad_norm": 0.10549204051494598, "learning_rate": 2.939260196480103e-05, "loss": 0.0053, "step": 710 }, { "grad_norm": 0.13164091110229492, "learning_rate": 2.9369584635138386e-05, "loss": 0.0051, "step": 720 }, { "grad_norm": 0.1299622356891632, "learning_rate": 2.9346148611284025e-05, "loss": 0.0057, "step": 730 }, { "grad_norm": 0.12443673610687256, "learning_rate": 2.9322294576105785e-05, "loss": 0.0054, "step": 740 }, { "grad_norm": 0.2119351178407669, "learning_rate": 2.92980232246513e-05, "loss": 0.006, "step": 750 }, { "grad_norm": 0.17323265969753265, "learning_rate": 2.927333526412779e-05, "loss": 0.0051, "step": 760 }, { "grad_norm": 0.14540384709835052, "learning_rate": 2.9248231413881423e-05, "loss": 0.0051, "step": 770 }, { "grad_norm": 0.1042582094669342, "learning_rate": 2.9222712405376372e-05, "loss": 0.0058, "step": 780 }, { "grad_norm": 0.10313008725643158, "learning_rate": 2.919677898217348e-05, "loss": 0.0047, "step": 790 }, { "grad_norm": 0.09979265928268433, "learning_rate": 2.917043189990864e-05, "loss": 0.0046, "step": 800 }, { "grad_norm": 0.08516388386487961, "learning_rate": 2.914367192627072e-05, "loss": 0.0041, "step": 810 }, { "grad_norm": 0.12445282191038132, "learning_rate": 2.9116499840979232e-05, "loss": 0.0042, "step": 820 }, { "grad_norm": 0.07022149115800858, "learning_rate": 2.9088916435761604e-05, "loss": 0.0043, "step": 830 }, { "grad_norm": 0.12865330278873444, "learning_rate": 2.9060922514330117e-05, "loss": 0.0049, "step": 840 }, { "grad_norm": 0.16881509125232697, "learning_rate": 2.903251889235847e-05, "loss": 0.0054, "step": 850 }, { "grad_norm": 0.12935705482959747, "learning_rate": 2.9003706397458025e-05, "loss": 0.0053, "step": 860 }, { "grad_norm": 0.11892224848270416, "learning_rate": 2.8974485869153698e-05, "loss": 0.0039, "step": 870 }, { "grad_norm": 0.12481220811605453, "learning_rate": 2.8944858158859484e-05, "loss": 0.0059, "step": 880 }, { "grad_norm": 0.09241459518671036, "learning_rate": 2.8914824129853653e-05, "loss": 0.0051, "step": 890 }, { "grad_norm": 0.08305121958255768, "learning_rate": 2.8884384657253603e-05, "loss": 0.0057, "step": 900 }, { "grad_norm": 0.12699495255947113, "learning_rate": 2.8853540627990356e-05, "loss": 0.005, "step": 910 }, { "grad_norm": 0.11409413814544678, "learning_rate": 2.8822292940782718e-05, "loss": 0.0042, "step": 920 }, { "grad_norm": 0.1391240954399109, "learning_rate": 2.8790642506111083e-05, "loss": 0.0042, "step": 930 }, { "grad_norm": 0.12300487607717514, "learning_rate": 2.8758590246190914e-05, "loss": 0.005, "step": 940 }, { "grad_norm": 0.1318342536687851, "learning_rate": 2.8726137094945875e-05, "loss": 0.0058, "step": 950 }, { "grad_norm": 0.1381806582212448, "learning_rate": 2.8693283997980604e-05, "loss": 0.0049, "step": 960 }, { "grad_norm": 0.11659003049135208, "learning_rate": 2.8660031912553177e-05, "loss": 0.0059, "step": 970 }, { "grad_norm": 0.09772439301013947, "learning_rate": 2.8626381807547197e-05, "loss": 0.004, "step": 980 }, { "grad_norm": 0.11248010396957397, "learning_rate": 2.859233466344358e-05, "loss": 0.0062, "step": 990 }, { "grad_norm": 0.08497095108032227, "learning_rate": 2.8557891472291994e-05, "loss": 0.0051, "step": 1000 }, { "grad_norm": 0.10097184032201767, "learning_rate": 2.852305323768191e-05, "loss": 0.0048, "step": 1010 }, { "grad_norm": 0.12525945901870728, "learning_rate": 2.8487820974713416e-05, "loss": 0.0067, "step": 1020 }, { "grad_norm": 0.12628182768821716, "learning_rate": 2.8452195709967595e-05, "loss": 0.0051, "step": 1030 }, { "grad_norm": 0.09084980189800262, "learning_rate": 2.8416178481476652e-05, "loss": 0.0044, "step": 1040 }, { "grad_norm": 0.09692411869764328, "learning_rate": 2.837977033869362e-05, "loss": 0.0043, "step": 1050 }, { "grad_norm": 0.11363832652568817, "learning_rate": 2.8342972342461827e-05, "loss": 0.0046, "step": 1060 }, { "grad_norm": 0.11651736497879028, "learning_rate": 2.8305785564983964e-05, "loss": 0.0049, "step": 1070 }, { "grad_norm": 0.11450709402561188, "learning_rate": 2.8268211089790837e-05, "loss": 0.006, "step": 1080 }, { "grad_norm": 0.13134150207042694, "learning_rate": 2.8230250011709824e-05, "loss": 0.0051, "step": 1090 }, { "grad_norm": 0.06729510426521301, "learning_rate": 2.819190343683294e-05, "loss": 0.0046, "step": 1100 }, { "grad_norm": 0.07857424765825272, "learning_rate": 2.8153172482484634e-05, "loss": 0.0041, "step": 1110 }, { "grad_norm": 0.12144170701503754, "learning_rate": 2.811405827718921e-05, "loss": 0.0042, "step": 1120 }, { "grad_norm": 0.1513696163892746, "learning_rate": 2.8074561960637976e-05, "loss": 0.0058, "step": 1130 }, { "grad_norm": 0.10109832137823105, "learning_rate": 2.8034684683656003e-05, "loss": 0.0045, "step": 1140 }, { "grad_norm": 0.07752282917499542, "learning_rate": 2.7994427608168617e-05, "loss": 0.0044, "step": 1150 }, { "grad_norm": 0.0888523980975151, "learning_rate": 2.7953791907167525e-05, "loss": 0.0055, "step": 1160 }, { "grad_norm": 0.10832788050174713, "learning_rate": 2.791277876467665e-05, "loss": 0.0061, "step": 1170 }, { "grad_norm": 0.08513602614402771, "learning_rate": 2.7871389375717637e-05, "loss": 0.0058, "step": 1180 }, { "grad_norm": 0.09278848767280579, "learning_rate": 2.782962494627501e-05, "loss": 0.0043, "step": 1190 }, { "grad_norm": 0.09228791296482086, "learning_rate": 2.7787486693261047e-05, "loss": 0.0053, "step": 1200 }, { "grad_norm": 0.11885028332471848, "learning_rate": 2.7744975844480326e-05, "loss": 0.0043, "step": 1210 }, { "grad_norm": 0.08647848665714264, "learning_rate": 2.7702093638593952e-05, "loss": 0.0043, "step": 1220 }, { "grad_norm": 0.1418542116880417, "learning_rate": 2.7658841325083445e-05, "loss": 0.0048, "step": 1230 }, { "grad_norm": 0.1654549092054367, "learning_rate": 2.761522016421435e-05, "loss": 0.0047, "step": 1240 }, { "grad_norm": 0.13846398890018463, "learning_rate": 2.757123142699952e-05, "loss": 0.0044, "step": 1250 }, { "grad_norm": 0.11066365987062454, "learning_rate": 2.7526876395162063e-05, "loss": 0.0047, "step": 1260 }, { "grad_norm": 0.1423611342906952, "learning_rate": 2.748215636109802e-05, "loss": 0.0049, "step": 1270 }, { "grad_norm": 0.1386536806821823, "learning_rate": 2.7437072627838685e-05, "loss": 0.0056, "step": 1280 }, { "grad_norm": 0.1230684295296669, "learning_rate": 2.7391626509012657e-05, "loss": 0.0051, "step": 1290 }, { "grad_norm": 0.1601068675518036, "learning_rate": 2.7345819328807545e-05, "loss": 0.0046, "step": 1300 }, { "grad_norm": 0.11355318874120712, "learning_rate": 2.7299652421931398e-05, "loss": 0.0045, "step": 1310 }, { "grad_norm": 0.09870899468660355, "learning_rate": 2.7253127133573818e-05, "loss": 0.004, "step": 1320 }, { "grad_norm": 0.13512560725212097, "learning_rate": 2.7206244819366748e-05, "loss": 0.0072, "step": 1330 }, { "grad_norm": 0.12584540247917175, "learning_rate": 2.7159006845344985e-05, "loss": 0.0059, "step": 1340 }, { "grad_norm": 0.1445896178483963, "learning_rate": 2.7111414587906383e-05, "loss": 0.0045, "step": 1350 }, { "grad_norm": 0.08940842747688293, "learning_rate": 2.7063469433771726e-05, "loss": 0.0049, "step": 1360 }, { "grad_norm": 0.07216592133045197, "learning_rate": 2.701517277994434e-05, "loss": 0.0041, "step": 1370 }, { "grad_norm": 0.0641094371676445, "learning_rate": 2.6966526033669398e-05, "loss": 0.0037, "step": 1380 }, { "grad_norm": 0.1202094703912735, "learning_rate": 2.6917530612392874e-05, "loss": 0.004, "step": 1390 }, { "grad_norm": 0.10160712152719498, "learning_rate": 2.6868187943720294e-05, "loss": 0.0039, "step": 1400 }, { "grad_norm": 0.0838804617524147, "learning_rate": 2.681849946537511e-05, "loss": 0.0044, "step": 1410 }, { "grad_norm": 0.07244281470775604, "learning_rate": 2.6768466625156797e-05, "loss": 0.0042, "step": 1420 }, { "grad_norm": 0.09038420021533966, "learning_rate": 2.671809088089871e-05, "loss": 0.0043, "step": 1430 }, { "grad_norm": 0.1908736675977707, "learning_rate": 2.666737370042556e-05, "loss": 0.0037, "step": 1440 }, { "grad_norm": 0.07205282896757126, "learning_rate": 2.661631656151068e-05, "loss": 0.0039, "step": 1450 }, { "grad_norm": 0.14247483015060425, "learning_rate": 2.6564920951832937e-05, "loss": 0.0053, "step": 1460 }, { "grad_norm": 0.09355881810188293, "learning_rate": 2.6513188368933416e-05, "loss": 0.0035, "step": 1470 }, { "grad_norm": 0.1190556213259697, "learning_rate": 2.6461120320171757e-05, "loss": 0.0049, "step": 1480 }, { "grad_norm": 0.1366739571094513, "learning_rate": 2.6408718322682246e-05, "loss": 0.0047, "step": 1490 }, { "grad_norm": 0.12044065445661545, "learning_rate": 2.6355983903329627e-05, "loss": 0.0036, "step": 1500 }, { "grad_norm": 0.11228612065315247, "learning_rate": 2.630291859866457e-05, "loss": 0.005, "step": 1510 }, { "grad_norm": 0.09218034148216248, "learning_rate": 2.624952395487895e-05, "loss": 0.0033, "step": 1520 }, { "grad_norm": 0.07770651578903198, "learning_rate": 2.619580152776075e-05, "loss": 0.0047, "step": 1530 }, { "grad_norm": 0.15504707396030426, "learning_rate": 2.6141752882648758e-05, "loss": 0.0041, "step": 1540 }, { "grad_norm": 0.16510525345802307, "learning_rate": 2.6087379594386956e-05, "loss": 0.0041, "step": 1550 }, { "grad_norm": 0.09923221915960312, "learning_rate": 2.6032683247278613e-05, "loss": 0.0046, "step": 1560 }, { "grad_norm": 0.10440414398908615, "learning_rate": 2.5977665435040135e-05, "loss": 0.0045, "step": 1570 }, { "grad_norm": 0.11448866128921509, "learning_rate": 2.592232776075464e-05, "loss": 0.0038, "step": 1580 }, { "grad_norm": 0.10592176020145416, "learning_rate": 2.5866671836825226e-05, "loss": 0.0041, "step": 1590 }, { "grad_norm": 0.10503924638032913, "learning_rate": 2.5810699284928e-05, "loss": 0.0048, "step": 1600 }, { "grad_norm": 0.08060912042856216, "learning_rate": 2.575441173596482e-05, "loss": 0.0042, "step": 1610 }, { "grad_norm": 0.08904369175434113, "learning_rate": 2.569781083001579e-05, "loss": 0.0035, "step": 1620 }, { "grad_norm": 0.10122060775756836, "learning_rate": 2.5640898216291467e-05, "loss": 0.005, "step": 1630 }, { "grad_norm": 0.07322278618812561, "learning_rate": 2.558367555308479e-05, "loss": 0.0036, "step": 1640 }, { "grad_norm": 0.10172346234321594, "learning_rate": 2.5526144507722777e-05, "loss": 0.0044, "step": 1650 }, { "grad_norm": 0.1091361865401268, "learning_rate": 2.546830675651794e-05, "loss": 0.0035, "step": 1660 }, { "grad_norm": 0.09636865556240082, "learning_rate": 2.5410163984719442e-05, "loss": 0.0037, "step": 1670 }, { "grad_norm": 0.11563992500305176, "learning_rate": 2.5351717886464e-05, "loss": 0.0048, "step": 1680 }, { "grad_norm": 0.0999450832605362, "learning_rate": 2.5292970164726503e-05, "loss": 0.0058, "step": 1690 }, { "grad_norm": 0.10548657923936844, "learning_rate": 2.5233922531270407e-05, "loss": 0.0057, "step": 1700 }, { "grad_norm": 0.11799237877130508, "learning_rate": 2.517457670659786e-05, "loss": 0.005, "step": 1710 }, { "grad_norm": 0.16543568670749664, "learning_rate": 2.5114934419899557e-05, "loss": 0.0047, "step": 1720 }, { "grad_norm": 0.13398373126983643, "learning_rate": 2.5054997409004368e-05, "loss": 0.0048, "step": 1730 }, { "grad_norm": 0.12377601861953735, "learning_rate": 2.4994767420328692e-05, "loss": 0.0042, "step": 1740 }, { "grad_norm": 0.1447695940732956, "learning_rate": 2.4934246208825594e-05, "loss": 0.0053, "step": 1750 }, { "grad_norm": 0.09021581709384918, "learning_rate": 2.4873435537933627e-05, "loss": 0.0041, "step": 1760 }, { "grad_norm": 0.2032671719789505, "learning_rate": 2.48123371795255e-05, "loss": 0.0059, "step": 1770 }, { "grad_norm": 0.12990808486938477, "learning_rate": 2.475095291385641e-05, "loss": 0.0045, "step": 1780 }, { "grad_norm": 0.1114439144730568, "learning_rate": 2.468928452951219e-05, "loss": 0.0061, "step": 1790 }, { "grad_norm": 0.13164660334587097, "learning_rate": 2.4627333823357185e-05, "loss": 0.0046, "step": 1800 }, { "grad_norm": 0.099468894302845, "learning_rate": 2.456510260048191e-05, "loss": 0.0038, "step": 1810 }, { "grad_norm": 0.13928678631782532, "learning_rate": 2.4502592674150436e-05, "loss": 0.005, "step": 1820 }, { "grad_norm": 0.08072284609079361, "learning_rate": 2.4439805865747564e-05, "loss": 0.0043, "step": 1830 }, { "grad_norm": 0.11410820484161377, "learning_rate": 2.4376744004725756e-05, "loss": 0.0043, "step": 1840 }, { "grad_norm": 0.11230126023292542, "learning_rate": 2.4313408928551834e-05, "loss": 0.0041, "step": 1850 }, { "grad_norm": 0.10740366578102112, "learning_rate": 2.424980248265342e-05, "loss": 0.0058, "step": 1860 }, { "grad_norm": 0.05973527953028679, "learning_rate": 2.418592652036519e-05, "loss": 0.0051, "step": 1870 }, { "grad_norm": 0.07472732663154602, "learning_rate": 2.4121782902874866e-05, "loss": 0.0043, "step": 1880 }, { "grad_norm": 0.09412290155887604, "learning_rate": 2.4057373499168968e-05, "loss": 0.0052, "step": 1890 }, { "grad_norm": 0.10615745931863785, "learning_rate": 2.399270018597838e-05, "loss": 0.0047, "step": 1900 }, { "grad_norm": 0.10771957784891129, "learning_rate": 2.392776484772365e-05, "loss": 0.0047, "step": 1910 }, { "grad_norm": 0.17450952529907227, "learning_rate": 2.3862569376460096e-05, "loss": 0.0046, "step": 1920 }, { "grad_norm": 0.14669440686702728, "learning_rate": 2.3797115671822665e-05, "loss": 0.0052, "step": 1930 }, { "grad_norm": 0.080330990254879, "learning_rate": 2.3731405640970578e-05, "loss": 0.0042, "step": 1940 }, { "grad_norm": 0.13788354396820068, "learning_rate": 2.366544119853178e-05, "loss": 0.0043, "step": 1950 }, { "grad_norm": 0.12239138782024384, "learning_rate": 2.3599224266547145e-05, "loss": 0.0039, "step": 1960 }, { "grad_norm": 0.12057246267795563, "learning_rate": 2.3532756774414454e-05, "loss": 0.0047, "step": 1970 }, { "grad_norm": 0.08931958675384521, "learning_rate": 2.3466040658832202e-05, "loss": 0.0043, "step": 1980 }, { "grad_norm": 0.11142227053642273, "learning_rate": 2.3399077863743154e-05, "loss": 0.0043, "step": 1990 }, { "grad_norm": 0.11959561705589294, "learning_rate": 2.3331870340277703e-05, "loss": 0.0043, "step": 2000 }, { "grad_norm": 0.07246878743171692, "learning_rate": 2.3264420046697033e-05, "loss": 0.0038, "step": 2010 }, { "grad_norm": 0.10344090312719345, "learning_rate": 2.3196728948336035e-05, "loss": 0.0051, "step": 2020 }, { "grad_norm": 0.11588168889284134, "learning_rate": 2.3128799017546066e-05, "loss": 0.0054, "step": 2030 }, { "grad_norm": 0.08791792392730713, "learning_rate": 2.306063223363747e-05, "loss": 0.0039, "step": 2040 }, { "grad_norm": 0.11206529289484024, "learning_rate": 2.299223058282189e-05, "loss": 0.0044, "step": 2050 }, { "grad_norm": 0.10127924382686615, "learning_rate": 2.2923596058154437e-05, "loss": 0.0049, "step": 2060 }, { "grad_norm": 0.09655720740556717, "learning_rate": 2.2854730659475575e-05, "loss": 0.004, "step": 2070 }, { "grad_norm": 0.0919470265507698, "learning_rate": 2.278563639335286e-05, "loss": 0.0037, "step": 2080 }, { "grad_norm": 0.1400183141231537, "learning_rate": 2.27163152730225e-05, "loss": 0.0061, "step": 2090 }, { "grad_norm": 0.09806305915117264, "learning_rate": 2.264676931833065e-05, "loss": 0.0044, "step": 2100 }, { "grad_norm": 0.07071761786937714, "learning_rate": 2.25770005556746e-05, "loss": 0.0027, "step": 2110 }, { "grad_norm": 0.08000872284173965, "learning_rate": 2.250701101794371e-05, "loss": 0.0039, "step": 2120 }, { "grad_norm": 0.09656523168087006, "learning_rate": 2.243680274446017e-05, "loss": 0.0057, "step": 2130 }, { "grad_norm": 0.09651938825845718, "learning_rate": 2.236637778091961e-05, "loss": 0.0034, "step": 2140 }, { "grad_norm": 0.09979426115751266, "learning_rate": 2.2295738179331452e-05, "loss": 0.004, "step": 2150 }, { "grad_norm": 0.05974041670560837, "learning_rate": 2.2224885997959155e-05, "loss": 0.0035, "step": 2160 }, { "grad_norm": 0.15712539851665497, "learning_rate": 2.215382330126021e-05, "loss": 0.0031, "step": 2170 }, { "grad_norm": 0.09726399183273315, "learning_rate": 2.2082552159826037e-05, "loss": 0.0031, "step": 2180 }, { "grad_norm": 0.06862549483776093, "learning_rate": 2.201107465032158e-05, "loss": 0.0037, "step": 2190 }, { "grad_norm": 0.18924099206924438, "learning_rate": 2.1939392855424865e-05, "loss": 0.0051, "step": 2200 }, { "grad_norm": 0.10739069432020187, "learning_rate": 2.186750886376629e-05, "loss": 0.0043, "step": 2210 }, { "grad_norm": 0.12378190457820892, "learning_rate": 2.1795424769867748e-05, "loss": 0.0065, "step": 2220 }, { "grad_norm": 0.09460093080997467, "learning_rate": 2.172314267408163e-05, "loss": 0.0033, "step": 2230 }, { "grad_norm": 0.0978519544005394, "learning_rate": 2.1650664682529626e-05, "loss": 0.003, "step": 2240 }, { "grad_norm": 0.09626927971839905, "learning_rate": 2.157799290704132e-05, "loss": 0.0046, "step": 2250 }, { "grad_norm": 0.1985124796628952, "learning_rate": 2.150512946509269e-05, "loss": 0.0044, "step": 2260 }, { "grad_norm": 0.17612029612064362, "learning_rate": 2.1432076479744403e-05, "loss": 0.0036, "step": 2270 }, { "grad_norm": 0.13746587932109833, "learning_rate": 2.1358836079579953e-05, "loss": 0.0036, "step": 2280 }, { "grad_norm": 0.1004326120018959, "learning_rate": 2.128541039864364e-05, "loss": 0.0044, "step": 2290 }, { "grad_norm": 0.126906156539917, "learning_rate": 2.1211801576378384e-05, "loss": 0.003, "step": 2300 }, { "grad_norm": 0.1311343014240265, "learning_rate": 2.1138011757563388e-05, "loss": 0.0048, "step": 2310 }, { "grad_norm": 0.13708554208278656, "learning_rate": 2.1064043092251654e-05, "loss": 0.0046, "step": 2320 }, { "grad_norm": 0.14371265470981598, "learning_rate": 2.098989773570733e-05, "loss": 0.0065, "step": 2330 }, { "grad_norm": 0.09575030952692032, "learning_rate": 2.0915577848342905e-05, "loss": 0.0032, "step": 2340 }, { "grad_norm": 0.09642798453569412, "learning_rate": 2.084108559565626e-05, "loss": 0.0058, "step": 2350 }, { "grad_norm": 0.09199727326631546, "learning_rate": 2.0766423148167592e-05, "loss": 0.0036, "step": 2360 }, { "grad_norm": 0.10715307295322418, "learning_rate": 2.0691592681356134e-05, "loss": 0.0038, "step": 2370 }, { "grad_norm": 0.13233180344104767, "learning_rate": 2.0616596375596805e-05, "loss": 0.0045, "step": 2380 }, { "grad_norm": 0.07768990844488144, "learning_rate": 2.054143641609665e-05, "loss": 0.0042, "step": 2390 }, { "grad_norm": 0.13608784973621368, "learning_rate": 2.0466114992831175e-05, "loss": 0.005, "step": 2400 }, { "grad_norm": 0.10314002633094788, "learning_rate": 2.039063430048056e-05, "loss": 0.0036, "step": 2410 }, { "grad_norm": 0.15518108010292053, "learning_rate": 2.031499653836568e-05, "loss": 0.0077, "step": 2420 }, { "grad_norm": 0.15236151218414307, "learning_rate": 2.0239203910384035e-05, "loss": 0.0055, "step": 2430 }, { "grad_norm": 0.11873604357242584, "learning_rate": 2.0163258624945532e-05, "loss": 0.0047, "step": 2440 }, { "grad_norm": 0.10288285464048386, "learning_rate": 2.008716289490814e-05, "loss": 0.0038, "step": 2450 }, { "grad_norm": 0.09240736067295074, "learning_rate": 2.0010918937513418e-05, "loss": 0.0032, "step": 2460 }, { "grad_norm": 0.10177815705537796, "learning_rate": 1.9934528974321904e-05, "loss": 0.0037, "step": 2470 }, { "grad_norm": 0.10068890452384949, "learning_rate": 1.9857995231148374e-05, "loss": 0.0034, "step": 2480 }, { "grad_norm": 0.09852269291877747, "learning_rate": 1.9781319937997003e-05, "loss": 0.0057, "step": 2490 }, { "grad_norm": 0.10190477967262268, "learning_rate": 1.9704505328996387e-05, "loss": 0.0034, "step": 2500 }, { "grad_norm": 0.11787329614162445, "learning_rate": 1.9627553642334435e-05, "loss": 0.0059, "step": 2510 }, { "grad_norm": 0.12991037964820862, "learning_rate": 1.9550467120193167e-05, "loss": 0.004, "step": 2520 }, { "grad_norm": 0.12583398818969727, "learning_rate": 1.947324800868337e-05, "loss": 0.0042, "step": 2530 }, { "grad_norm": 0.1513422727584839, "learning_rate": 1.9395898557779154e-05, "loss": 0.0039, "step": 2540 }, { "grad_norm": 0.10720251500606537, "learning_rate": 1.9318421021252408e-05, "loss": 0.0059, "step": 2550 }, { "grad_norm": 0.108665332198143, "learning_rate": 1.9240817656607114e-05, "loss": 0.0056, "step": 2560 }, { "grad_norm": 0.18657821416854858, "learning_rate": 1.9163090725013578e-05, "loss": 0.0044, "step": 2570 }, { "grad_norm": 0.12283720821142197, "learning_rate": 1.9085242491242524e-05, "loss": 0.0046, "step": 2580 }, { "grad_norm": 0.10580199956893921, "learning_rate": 1.900727522359915e-05, "loss": 0.0035, "step": 2590 }, { "grad_norm": 0.1246466338634491, "learning_rate": 1.8929191193856985e-05, "loss": 0.0059, "step": 2600 }, { "grad_norm": 0.11310068517923355, "learning_rate": 1.8850992677191726e-05, "loss": 0.0042, "step": 2610 }, { "grad_norm": 0.10423216968774796, "learning_rate": 1.877268195211494e-05, "loss": 0.0044, "step": 2620 }, { "grad_norm": 0.09140152484178543, "learning_rate": 1.8694261300407653e-05, "loss": 0.004, "step": 2630 }, { "grad_norm": 0.15965282917022705, "learning_rate": 1.861573300705391e-05, "loss": 0.0044, "step": 2640 }, { "grad_norm": 0.11887385696172714, "learning_rate": 1.8537099360174137e-05, "loss": 0.0042, "step": 2650 }, { "grad_norm": 0.09578254073858261, "learning_rate": 1.845836265095852e-05, "loss": 0.0038, "step": 2660 }, { "grad_norm": 0.09519214928150177, "learning_rate": 1.837952517360022e-05, "loss": 0.0041, "step": 2670 }, { "grad_norm": 0.12270296365022659, "learning_rate": 1.8300589225228533e-05, "loss": 0.0044, "step": 2680 }, { "grad_norm": 0.11333918571472168, "learning_rate": 1.8221557105841958e-05, "loss": 0.0047, "step": 2690 }, { "grad_norm": 0.08164770901203156, "learning_rate": 1.814243111824119e-05, "loss": 0.0032, "step": 2700 }, { "grad_norm": 0.11921174079179764, "learning_rate": 1.8063213567961994e-05, "loss": 0.0037, "step": 2710 }, { "grad_norm": 0.09655310958623886, "learning_rate": 1.7983906763208055e-05, "loss": 0.0041, "step": 2720 }, { "grad_norm": 0.0791718065738678, "learning_rate": 1.7904513014783706e-05, "loss": 0.0042, "step": 2730 }, { "grad_norm": 0.11868743598461151, "learning_rate": 1.782503463602662e-05, "loss": 0.0038, "step": 2740 }, { "grad_norm": 0.08265439420938492, "learning_rate": 1.7745473942740373e-05, "loss": 0.0035, "step": 2750 }, { "grad_norm": 0.10162845999002457, "learning_rate": 1.7665833253126987e-05, "loss": 0.0033, "step": 2760 }, { "grad_norm": 0.10096478462219238, "learning_rate": 1.7586114887719378e-05, "loss": 0.0062, "step": 2770 }, { "grad_norm": 0.11772424727678299, "learning_rate": 1.750632116931374e-05, "loss": 0.0032, "step": 2780 }, { "grad_norm": 0.10858657211065292, "learning_rate": 1.7426454422901888e-05, "loss": 0.0035, "step": 2790 }, { "grad_norm": 0.11779221147298813, "learning_rate": 1.7346516975603466e-05, "loss": 0.0042, "step": 2800 }, { "grad_norm": 0.09140808880329132, "learning_rate": 1.7266511156598176e-05, "loss": 0.0043, "step": 2810 }, { "grad_norm": 0.08126810193061829, "learning_rate": 1.7186439297057906e-05, "loss": 0.0042, "step": 2820 }, { "grad_norm": 0.09498797357082367, "learning_rate": 1.7106303730078806e-05, "loss": 0.0037, "step": 2830 }, { "grad_norm": 0.08488073945045471, "learning_rate": 1.7026106790613295e-05, "loss": 0.0034, "step": 2840 }, { "grad_norm": 0.07819471508264542, "learning_rate": 1.6945850815402033e-05, "loss": 0.0043, "step": 2850 }, { "grad_norm": 0.13870303332805634, "learning_rate": 1.6865538142905846e-05, "loss": 0.0047, "step": 2860 }, { "grad_norm": 0.0990184098482132, "learning_rate": 1.6785171113237562e-05, "loss": 0.0048, "step": 2870 }, { "grad_norm": 0.05612372234463692, "learning_rate": 1.670475206809387e-05, "loss": 0.0045, "step": 2880 }, { "grad_norm": 0.11454035341739655, "learning_rate": 1.662428335068704e-05, "loss": 0.0041, "step": 2890 }, { "grad_norm": 0.09158965945243835, "learning_rate": 1.6543767305676663e-05, "loss": 0.0039, "step": 2900 }, { "grad_norm": 0.09644990414381027, "learning_rate": 1.6463206279101363e-05, "loss": 0.004, "step": 2910 }, { "grad_norm": 0.09871806204319, "learning_rate": 1.6382602618310396e-05, "loss": 0.0046, "step": 2920 }, { "grad_norm": 0.1133292093873024, "learning_rate": 1.6301958671895288e-05, "loss": 0.0057, "step": 2930 }, { "grad_norm": 0.15669667720794678, "learning_rate": 1.6221276789621377e-05, "loss": 0.0056, "step": 2940 }, { "grad_norm": 0.07493694126605988, "learning_rate": 1.6140559322359358e-05, "loss": 0.0047, "step": 2950 }, { "grad_norm": 0.14167147874832153, "learning_rate": 1.605980862201679e-05, "loss": 0.004, "step": 2960 }, { "grad_norm": 0.12322021275758743, "learning_rate": 1.5979027041469576e-05, "loss": 0.0038, "step": 2970 }, { "grad_norm": 0.1101364865899086, "learning_rate": 1.5898216934493365e-05, "loss": 0.0035, "step": 2980 }, { "grad_norm": 0.1453101634979248, "learning_rate": 1.5817380655695005e-05, "loss": 0.0034, "step": 2990 }, { "grad_norm": 0.09624999016523361, "learning_rate": 1.5736520560443933e-05, "loss": 0.0039, "step": 3000 }, { "grad_norm": 0.10005682706832886, "learning_rate": 1.5655639004803537e-05, "loss": 0.0036, "step": 3010 }, { "grad_norm": 0.11586767435073853, "learning_rate": 1.55747383454625e-05, "loss": 0.0038, "step": 3020 }, { "grad_norm": 0.06066455319523811, "learning_rate": 1.5493820939666146e-05, "loss": 0.0037, "step": 3030 }, { "grad_norm": 0.06761951744556427, "learning_rate": 1.541288914514775e-05, "loss": 0.0045, "step": 3040 }, { "grad_norm": 0.08163075149059296, "learning_rate": 1.533194532005983e-05, "loss": 0.0037, "step": 3050 }, { "grad_norm": 0.1107611283659935, "learning_rate": 1.5250991822905468e-05, "loss": 0.0036, "step": 3060 }, { "grad_norm": 0.12838762998580933, "learning_rate": 1.5170031012469538e-05, "loss": 0.004, "step": 3070 }, { "grad_norm": 0.10671990364789963, "learning_rate": 1.508906524775002e-05, "loss": 0.0044, "step": 3080 }, { "grad_norm": 0.1331837922334671, "learning_rate": 1.5008096887889245e-05, "loss": 0.0048, "step": 3090 }, { "grad_norm": 0.1242435947060585, "learning_rate": 1.492712829210516e-05, "loss": 0.0042, "step": 3100 }, { "grad_norm": 0.08521351218223572, "learning_rate": 1.4846161819622594e-05, "loss": 0.0038, "step": 3110 }, { "grad_norm": 0.07415805011987686, "learning_rate": 1.4765199829604498e-05, "loss": 0.0032, "step": 3120 }, { "grad_norm": 0.06467261910438538, "learning_rate": 1.468424468108322e-05, "loss": 0.0061, "step": 3130 }, { "grad_norm": 0.09693094342947006, "learning_rate": 1.4603298732891762e-05, "loss": 0.0038, "step": 3140 }, { "grad_norm": 0.05952545627951622, "learning_rate": 1.4522364343595054e-05, "loss": 0.0044, "step": 3150 }, { "grad_norm": 0.0716821476817131, "learning_rate": 1.4441443871421228e-05, "loss": 0.0034, "step": 3160 }, { "grad_norm": 0.08609385043382645, "learning_rate": 1.4360539674192899e-05, "loss": 0.0033, "step": 3170 }, { "grad_norm": 0.08336781710386276, "learning_rate": 1.4279654109258472e-05, "loss": 0.003, "step": 3180 }, { "grad_norm": 0.12197946757078171, "learning_rate": 1.4198789533423466e-05, "loss": 0.0039, "step": 3190 }, { "grad_norm": 0.09763512015342712, "learning_rate": 1.4117948302881811e-05, "loss": 0.0038, "step": 3200 }, { "grad_norm": 0.07868342101573944, "learning_rate": 1.403713277314723e-05, "loss": 0.0038, "step": 3210 }, { "grad_norm": 0.051395829766988754, "learning_rate": 1.3956345298984574e-05, "loss": 0.0028, "step": 3220 }, { "grad_norm": 0.11672254651784897, "learning_rate": 1.3875588234341233e-05, "loss": 0.0028, "step": 3230 }, { "grad_norm": 0.11190174520015717, "learning_rate": 1.3794863932278527e-05, "loss": 0.0041, "step": 3240 }, { "grad_norm": 0.09065545350313187, "learning_rate": 1.3714174744903172e-05, "loss": 0.0032, "step": 3250 }, { "grad_norm": 0.11228013038635254, "learning_rate": 1.3633523023298717e-05, "loss": 0.0046, "step": 3260 }, { "grad_norm": 0.06375761330127716, "learning_rate": 1.3552911117457055e-05, "loss": 0.0044, "step": 3270 }, { "grad_norm": 0.07220669835805893, "learning_rate": 1.347234137620994e-05, "loss": 0.0032, "step": 3280 }, { "grad_norm": 0.09625758230686188, "learning_rate": 1.3391816147160568e-05, "loss": 0.0056, "step": 3290 }, { "grad_norm": 0.08487574756145477, "learning_rate": 1.3311337776615146e-05, "loss": 0.0037, "step": 3300 }, { "grad_norm": 0.07956377416849136, "learning_rate": 1.3230908609514544e-05, "loss": 0.0031, "step": 3310 }, { "grad_norm": 0.11078336089849472, "learning_rate": 1.3150530989365966e-05, "loss": 0.0044, "step": 3320 }, { "grad_norm": 0.08106371760368347, "learning_rate": 1.307020725817467e-05, "loss": 0.0033, "step": 3330 }, { "grad_norm": 0.05347970128059387, "learning_rate": 1.2989939756375711e-05, "loss": 0.004, "step": 3340 }, { "grad_norm": 0.08384165912866592, "learning_rate": 1.290973082276577e-05, "loss": 0.003, "step": 3350 }, { "grad_norm": 0.09720702469348907, "learning_rate": 1.2829582794434987e-05, "loss": 0.0031, "step": 3360 }, { "grad_norm": 0.10105139762163162, "learning_rate": 1.2749498006698879e-05, "loss": 0.0043, "step": 3370 }, { "grad_norm": 0.08893254399299622, "learning_rate": 1.2669478793030287e-05, "loss": 0.0029, "step": 3380 }, { "grad_norm": 0.09673310816287994, "learning_rate": 1.2589527484991387e-05, "loss": 0.0025, "step": 3390 }, { "grad_norm": 0.12181838601827621, "learning_rate": 1.2509646412165748e-05, "loss": 0.0039, "step": 3400 }, { "grad_norm": 0.11474212259054184, "learning_rate": 1.2429837902090462e-05, "loss": 0.0027, "step": 3410 }, { "grad_norm": 0.0729747787117958, "learning_rate": 1.2350104280188316e-05, "loss": 0.0028, "step": 3420 }, { "grad_norm": 0.09193936735391617, "learning_rate": 1.2270447869700051e-05, "loss": 0.0035, "step": 3430 }, { "grad_norm": 0.07126328349113464, "learning_rate": 1.219087099161665e-05, "loss": 0.0043, "step": 3440 }, { "grad_norm": 0.09029316902160645, "learning_rate": 1.2111375964611726e-05, "loss": 0.0032, "step": 3450 }, { "grad_norm": 0.10429364442825317, "learning_rate": 1.2031965104973935e-05, "loss": 0.0046, "step": 3460 }, { "grad_norm": 0.08040885627269745, "learning_rate": 1.1952640726539533e-05, "loss": 0.0025, "step": 3470 }, { "grad_norm": 0.11755494773387909, "learning_rate": 1.1873405140624904e-05, "loss": 0.0039, "step": 3480 }, { "grad_norm": 0.12123305350542068, "learning_rate": 1.1794260655959247e-05, "loss": 0.0044, "step": 3490 }, { "grad_norm": 0.10822794586420059, "learning_rate": 1.1715209578617294e-05, "loss": 0.0037, "step": 3500 }, { "grad_norm": 0.065816231071949, "learning_rate": 1.1636254211952112e-05, "loss": 0.0028, "step": 3510 }, { "grad_norm": 0.06965353339910507, "learning_rate": 1.1557396856528013e-05, "loss": 0.0033, "step": 3520 }, { "grad_norm": 0.07491380721330643, "learning_rate": 1.1478639810053487e-05, "loss": 0.0044, "step": 3530 }, { "grad_norm": 0.08226829022169113, "learning_rate": 1.1399985367314274e-05, "loss": 0.0032, "step": 3540 }, { "grad_norm": 0.05845038592815399, "learning_rate": 1.1321435820106487e-05, "loss": 0.0026, "step": 3550 }, { "grad_norm": 0.11162237823009491, "learning_rate": 1.1242993457169867e-05, "loss": 0.003, "step": 3560 }, { "grad_norm": 0.13599853217601776, "learning_rate": 1.116466056412104e-05, "loss": 0.0033, "step": 3570 }, { "grad_norm": 0.09740614145994186, "learning_rate": 1.1086439423386972e-05, "loss": 0.0043, "step": 3580 }, { "grad_norm": 0.08986973017454147, "learning_rate": 1.1008332314138433e-05, "loss": 0.0026, "step": 3590 }, { "grad_norm": 0.09561492502689362, "learning_rate": 1.0930341512223596e-05, "loss": 0.0038, "step": 3600 }, { "grad_norm": 0.10061262547969818, "learning_rate": 1.0852469290101724e-05, "loss": 0.0034, "step": 3610 }, { "grad_norm": 0.07297629117965698, "learning_rate": 1.077471791677697e-05, "loss": 0.0025, "step": 3620 }, { "grad_norm": 0.10886093229055405, "learning_rate": 1.069708965773224e-05, "loss": 0.0026, "step": 3630 }, { "grad_norm": 0.102382592856884, "learning_rate": 1.0619586774863198e-05, "loss": 0.0031, "step": 3640 }, { "grad_norm": 0.08297941088676453, "learning_rate": 1.0542211526412351e-05, "loss": 0.0029, "step": 3650 }, { "grad_norm": 0.0679573193192482, "learning_rate": 1.0464966166903266e-05, "loss": 0.0035, "step": 3660 }, { "grad_norm": 0.0714852511882782, "learning_rate": 1.0387852947074858e-05, "loss": 0.0036, "step": 3670 }, { "grad_norm": 0.07032801955938339, "learning_rate": 1.0310874113815824e-05, "loss": 0.0026, "step": 3680 }, { "grad_norm": 0.0894472673535347, "learning_rate": 1.0234031910099153e-05, "loss": 0.0028, "step": 3690 }, { "grad_norm": 0.08256567269563675, "learning_rate": 1.0157328574916811e-05, "loss": 0.0047, "step": 3700 }, { "grad_norm": 0.08830638229846954, "learning_rate": 1.0080766343214462e-05, "loss": 0.0029, "step": 3710 }, { "grad_norm": 0.07382957637310028, "learning_rate": 1.0004347445826368e-05, "loss": 0.0034, "step": 3720 }, { "grad_norm": 0.0672360509634018, "learning_rate": 9.928074109410371e-06, "loss": 0.0036, "step": 3730 }, { "grad_norm": 0.062298908829689026, "learning_rate": 9.851948556383055e-06, "loss": 0.0034, "step": 3740 }, { "grad_norm": 0.054670773446559906, "learning_rate": 9.77597300485493e-06, "loss": 0.0031, "step": 3750 }, { "grad_norm": 0.0936974510550499, "learning_rate": 9.700149668565853e-06, "loss": 0.0037, "step": 3760 }, { "grad_norm": 0.06597235053777695, "learning_rate": 9.624480756820495e-06, "loss": 0.0031, "step": 3770 }, { "grad_norm": 0.06676555424928665, "learning_rate": 9.548968474423984e-06, "loss": 0.0023, "step": 3780 }, { "grad_norm": 0.08981256186962128, "learning_rate": 9.473615021617646e-06, "loss": 0.004, "step": 3790 }, { "grad_norm": 0.07909717410802841, "learning_rate": 9.398422594014917e-06, "loss": 0.004, "step": 3800 }, { "grad_norm": 0.05415402725338936, "learning_rate": 9.32339338253735e-06, "loss": 0.0022, "step": 3810 }, { "grad_norm": 0.05480325594544411, "learning_rate": 9.248529573350782e-06, "loss": 0.0027, "step": 3820 }, { "grad_norm": 0.058878351002931595, "learning_rate": 9.173833347801639e-06, "loss": 0.0037, "step": 3830 }, { "grad_norm": 0.06265062093734741, "learning_rate": 9.099306882353377e-06, "loss": 0.0027, "step": 3840 }, { "grad_norm": 0.0681389793753624, "learning_rate": 9.024952348523053e-06, "loss": 0.0026, "step": 3850 }, { "grad_norm": 0.08739670366048813, "learning_rate": 8.95077191281807e-06, "loss": 0.0032, "step": 3860 }, { "grad_norm": 0.07649361342191696, "learning_rate": 8.876767736673037e-06, "loss": 0.003, "step": 3870 }, { "grad_norm": 0.08005782216787338, "learning_rate": 8.802941976386793e-06, "loss": 0.0023, "step": 3880 }, { "grad_norm": 0.0897689238190651, "learning_rate": 8.72929678305959e-06, "loss": 0.0026, "step": 3890 }, { "grad_norm": 0.07536846399307251, "learning_rate": 8.655834302530391e-06, "loss": 0.0029, "step": 3900 }, { "grad_norm": 0.07567661255598068, "learning_rate": 8.58255667531437e-06, "loss": 0.0035, "step": 3910 }, { "grad_norm": 0.12472916394472122, "learning_rate": 8.509466036540531e-06, "loss": 0.005, "step": 3920 }, { "grad_norm": 0.06811542809009552, "learning_rate": 8.436564515889497e-06, "loss": 0.0025, "step": 3930 }, { "grad_norm": 0.049239810556173325, "learning_rate": 8.363854237531442e-06, "loss": 0.0025, "step": 3940 }, { "grad_norm": 0.050166867673397064, "learning_rate": 8.291337320064241e-06, "loss": 0.0024, "step": 3950 }, { "grad_norm": 0.05276251584291458, "learning_rate": 8.219015876451675e-06, "loss": 0.0033, "step": 3960 }, { "grad_norm": 0.08960229158401489, "learning_rate": 8.146892013961928e-06, "loss": 0.0028, "step": 3970 }, { "grad_norm": 0.110183484852314, "learning_rate": 8.074967834106153e-06, "loss": 0.006, "step": 3980 }, { "grad_norm": 0.11177948862314224, "learning_rate": 8.003245432577229e-06, "loss": 0.0056, "step": 3990 }, { "grad_norm": 0.05508233234286308, "learning_rate": 7.931726899188727e-06, "loss": 0.0027, "step": 4000 }, { "grad_norm": 0.052221231162548065, "learning_rate": 7.860414317814007e-06, "loss": 0.0029, "step": 4010 }, { "grad_norm": 0.06484582275152206, "learning_rate": 7.789309766325483e-06, "loss": 0.003, "step": 4020 }, { "grad_norm": 0.0864526629447937, "learning_rate": 7.718415316534107e-06, "loss": 0.0026, "step": 4030 }, { "grad_norm": 0.05630047991871834, "learning_rate": 7.64773303412897e-06, "loss": 0.0025, "step": 4040 }, { "grad_norm": 0.07152245938777924, "learning_rate": 7.5772649786171415e-06, "loss": 0.0026, "step": 4050 }, { "grad_norm": 0.07800120860338211, "learning_rate": 7.507013203263653e-06, "loss": 0.0026, "step": 4060 }, { "grad_norm": 0.062226876616477966, "learning_rate": 7.436979755031649e-06, "loss": 0.0024, "step": 4070 }, { "grad_norm": 0.08144303411245346, "learning_rate": 7.367166674522775e-06, "loss": 0.0026, "step": 4080 }, { "grad_norm": 0.049730513244867325, "learning_rate": 7.297575995917708e-06, "loss": 0.003, "step": 4090 }, { "grad_norm": 0.07160552591085434, "learning_rate": 7.228209746916862e-06, "loss": 0.0026, "step": 4100 }, { "grad_norm": 0.06243562698364258, "learning_rate": 7.159069948681363e-06, "loss": 0.0029, "step": 4110 }, { "grad_norm": 0.0679006278514862, "learning_rate": 7.090158615774082e-06, "loss": 0.0028, "step": 4120 }, { "grad_norm": 0.08142750710248947, "learning_rate": 7.021477756101007e-06, "loss": 0.0024, "step": 4130 }, { "grad_norm": 0.05774086341261864, "learning_rate": 6.953029370852677e-06, "loss": 0.0036, "step": 4140 }, { "grad_norm": 0.0558801144361496, "learning_rate": 6.884815454445923e-06, "loss": 0.0027, "step": 4150 }, { "grad_norm": 0.061070751398801804, "learning_rate": 6.816837994465731e-06, "loss": 0.0045, "step": 4160 }, { "grad_norm": 0.06962040066719055, "learning_rate": 6.749098971607311e-06, "loss": 0.0023, "step": 4170 }, { "grad_norm": 0.07268668711185455, "learning_rate": 6.681600359618428e-06, "loss": 0.0026, "step": 4180 }, { "grad_norm": 0.05980467423796654, "learning_rate": 6.614344125241862e-06, "loss": 0.0024, "step": 4190 }, { "grad_norm": 0.09890970587730408, "learning_rate": 6.547332228158094e-06, "loss": 0.0026, "step": 4200 }, { "grad_norm": 0.047991473227739334, "learning_rate": 6.480566620928245e-06, "loss": 0.0027, "step": 4210 }, { "grad_norm": 0.09556986391544342, "learning_rate": 6.41404924893713e-06, "loss": 0.0024, "step": 4220 }, { "grad_norm": 0.08843955397605896, "learning_rate": 6.347782050336625e-06, "loss": 0.0022, "step": 4230 }, { "grad_norm": 0.06960665434598923, "learning_rate": 6.281766955989165e-06, "loss": 0.0048, "step": 4240 }, { "grad_norm": 0.08095689862966537, "learning_rate": 6.2160058894114835e-06, "loss": 0.0028, "step": 4250 }, { "grad_norm": 0.06031321734189987, "learning_rate": 6.150500766718582e-06, "loss": 0.0027, "step": 4260 }, { "grad_norm": 0.06310360133647919, "learning_rate": 6.085253496567878e-06, "loss": 0.0019, "step": 4270 }, { "grad_norm": 0.06618604063987732, "learning_rate": 6.020265980103609e-06, "loss": 0.0027, "step": 4280 }, { "grad_norm": 0.058316055685281754, "learning_rate": 5.955540110901447e-06, "loss": 0.0028, "step": 4290 }, { "grad_norm": 0.05983656644821167, "learning_rate": 5.891077774913282e-06, "loss": 0.0025, "step": 4300 }, { "grad_norm": 0.05791456997394562, "learning_rate": 5.8268808504123185e-06, "loss": 0.003, "step": 4310 }, { "grad_norm": 0.04390927404165268, "learning_rate": 5.762951207938309e-06, "loss": 0.0026, "step": 4320 }, { "grad_norm": 0.05586939677596092, "learning_rate": 5.699290710243082e-06, "loss": 0.003, "step": 4330 }, { "grad_norm": 0.10256128013134003, "learning_rate": 5.635901212236251e-06, "loss": 0.0032, "step": 4340 }, { "grad_norm": 0.04283665493130684, "learning_rate": 5.5727845609311545e-06, "loss": 0.0025, "step": 4350 }, { "grad_norm": 0.05426718667149544, "learning_rate": 5.509942595391073e-06, "loss": 0.0021, "step": 4360 }, { "grad_norm": 0.08707068860530853, "learning_rate": 5.447377146675602e-06, "loss": 0.0029, "step": 4370 }, { "grad_norm": 0.10243280977010727, "learning_rate": 5.385090037787334e-06, "loss": 0.0029, "step": 4380 }, { "grad_norm": 0.05103703960776329, "learning_rate": 5.323083083618727e-06, "loss": 0.0033, "step": 4390 }, { "grad_norm": 0.060124531388282776, "learning_rate": 5.261358090899206e-06, "loss": 0.0033, "step": 4400 }, { "grad_norm": 0.10158316045999527, "learning_rate": 5.199916858142556e-06, "loss": 0.0033, "step": 4410 }, { "grad_norm": 0.07515929639339447, "learning_rate": 5.138761175594495e-06, "loss": 0.0025, "step": 4420 }, { "grad_norm": 0.07259644567966461, "learning_rate": 5.077892825180497e-06, "loss": 0.0029, "step": 4430 }, { "grad_norm": 0.07611123472452164, "learning_rate": 5.017313580453905e-06, "loss": 0.0025, "step": 4440 }, { "grad_norm": 0.08785836398601532, "learning_rate": 4.9570252065442274e-06, "loss": 0.0034, "step": 4450 }, { "grad_norm": 0.039934225380420685, "learning_rate": 4.89702946010572e-06, "loss": 0.002, "step": 4460 }, { "grad_norm": 0.07307283580303192, "learning_rate": 4.837328089266195e-06, "loss": 0.0024, "step": 4470 }, { "grad_norm": 0.0927189365029335, "learning_rate": 4.777922833576088e-06, "loss": 0.0034, "step": 4480 }, { "grad_norm": 0.04589124023914337, "learning_rate": 4.718815423957778e-06, "loss": 0.0039, "step": 4490 }, { "grad_norm": 0.0727127268910408, "learning_rate": 4.660007582655125e-06, "loss": 0.0031, "step": 4500 }, { "grad_norm": 0.07311692833900452, "learning_rate": 4.6015010231833275e-06, "loss": 0.0025, "step": 4510 }, { "grad_norm": 0.06421023607254028, "learning_rate": 4.5432974502789754e-06, "loss": 0.0025, "step": 4520 }, { "grad_norm": 0.04857269302010536, "learning_rate": 4.4853985598503625e-06, "loss": 0.0031, "step": 4530 }, { "grad_norm": 0.059340883046388626, "learning_rate": 4.427806038928109e-06, "loss": 0.0032, "step": 4540 }, { "grad_norm": 0.08296361565589905, "learning_rate": 4.370521565615962e-06, "loss": 0.0029, "step": 4550 }, { "grad_norm": 0.07902832329273224, "learning_rate": 4.313546809041942e-06, "loss": 0.0027, "step": 4560 }, { "grad_norm": 0.1092028021812439, "learning_rate": 4.256883429309685e-06, "loss": 0.0022, "step": 4570 }, { "grad_norm": 0.052551236003637314, "learning_rate": 4.2005330774500645e-06, "loss": 0.0035, "step": 4580 }, { "grad_norm": 0.05813279747962952, "learning_rate": 4.144497395373106e-06, "loss": 0.0031, "step": 4590 }, { "grad_norm": 0.0455247201025486, "learning_rate": 4.088778015820135e-06, "loss": 0.0031, "step": 4600 }, { "grad_norm": 0.0658392682671547, "learning_rate": 4.033376562316191e-06, "loss": 0.0033, "step": 4610 }, { "grad_norm": 0.06841374188661575, "learning_rate": 3.978294649122752e-06, "loss": 0.0026, "step": 4620 }, { "grad_norm": 0.06853021681308746, "learning_rate": 3.923533881190662e-06, "loss": 0.003, "step": 4630 }, { "grad_norm": 0.07420346140861511, "learning_rate": 3.8690958541134e-06, "loss": 0.003, "step": 4640 }, { "grad_norm": 0.06575186550617218, "learning_rate": 3.814982154080569e-06, "loss": 0.0034, "step": 4650 }, { "grad_norm": 0.07205885648727417, "learning_rate": 3.7611943578316824e-06, "loss": 0.0027, "step": 4660 }, { "grad_norm": 0.04615858942270279, "learning_rate": 3.707734032610228e-06, "loss": 0.003, "step": 4670 }, { "grad_norm": 0.07501405477523804, "learning_rate": 3.654602736117982e-06, "loss": 0.0035, "step": 4680 }, { "grad_norm": 0.09194950014352798, "learning_rate": 3.601802016469653e-06, "loss": 0.0042, "step": 4690 }, { "grad_norm": 0.04328564181923866, "learning_rate": 3.5493334121477544e-06, "loss": 0.0028, "step": 4700 }, { "grad_norm": 0.05529750511050224, "learning_rate": 3.49719845195777e-06, "loss": 0.003, "step": 4710 }, { "grad_norm": 0.10001294314861298, "learning_rate": 3.445398654983628e-06, "loss": 0.0034, "step": 4720 }, { "grad_norm": 0.06471127271652222, "learning_rate": 3.393935530543418e-06, "loss": 0.0027, "step": 4730 }, { "grad_norm": 0.05960933491587639, "learning_rate": 3.3428105781454364e-06, "loss": 0.0025, "step": 4740 }, { "grad_norm": 0.08707603067159653, "learning_rate": 3.2920252874444788e-06, "loss": 0.0043, "step": 4750 }, { "grad_norm": 0.08540203422307968, "learning_rate": 3.2415811381984264e-06, "loss": 0.0037, "step": 4760 }, { "grad_norm": 0.0713149905204773, "learning_rate": 3.191479600225159e-06, "loss": 0.0033, "step": 4770 }, { "grad_norm": 0.06586753576993942, "learning_rate": 3.1417221333596918e-06, "loss": 0.0032, "step": 4780 }, { "grad_norm": 0.07737500220537186, "learning_rate": 3.0923101874116756e-06, "loss": 0.0031, "step": 4790 }, { "grad_norm": 0.07109493762254715, "learning_rate": 3.0432452021231287e-06, "loss": 0.0022, "step": 4800 }, { "grad_norm": 0.07665184885263443, "learning_rate": 2.9945286071264845e-06, "loss": 0.0038, "step": 4810 }, { "grad_norm": 0.05713554471731186, "learning_rate": 2.9461618219029563e-06, "loss": 0.0028, "step": 4820 }, { "grad_norm": 0.06393369287252426, "learning_rate": 2.89814625574116e-06, "loss": 0.0029, "step": 4830 }, { "grad_norm": 0.06203538551926613, "learning_rate": 2.8504833076960567e-06, "loss": 0.0027, "step": 4840 }, { "grad_norm": 0.04352898150682449, "learning_rate": 2.803174366548188e-06, "loss": 0.0021, "step": 4850 }, { "grad_norm": 0.07385428994894028, "learning_rate": 2.756220810763202e-06, "loss": 0.0033, "step": 4860 }, { "grad_norm": 0.06604447215795517, "learning_rate": 2.709624008451707e-06, "loss": 0.0028, "step": 4870 }, { "grad_norm": 0.05847585201263428, "learning_rate": 2.6633853173293834e-06, "loss": 0.002, "step": 4880 }, { "grad_norm": 0.0321679562330246, "learning_rate": 2.6175060846774485e-06, "loss": 0.0024, "step": 4890 }, { "grad_norm": 0.035010624676942825, "learning_rate": 2.5719876473033825e-06, "loss": 0.0025, "step": 4900 }, { "grad_norm": 0.06074056029319763, "learning_rate": 2.5268313315019796e-06, "loss": 0.0024, "step": 4910 }, { "grad_norm": 0.04323519021272659, "learning_rate": 2.482038453016712e-06, "loss": 0.0038, "step": 4920 }, { "grad_norm": 0.06676605343818665, "learning_rate": 2.437610317001387e-06, "loss": 0.0044, "step": 4930 }, { "grad_norm": 0.05648193508386612, "learning_rate": 2.3935482179821067e-06, "loss": 0.0028, "step": 4940 }, { "grad_norm": 0.05948924645781517, "learning_rate": 2.349853439819576e-06, "loss": 0.0032, "step": 4950 }, { "grad_norm": 0.05742521211504936, "learning_rate": 2.3065272556716604e-06, "loss": 0.0043, "step": 4960 }, { "grad_norm": 0.07226058095693588, "learning_rate": 2.2635709279563204e-06, "loss": 0.0027, "step": 4970 }, { "grad_norm": 0.04860719293355942, "learning_rate": 2.2209857083148098e-06, "loss": 0.0033, "step": 4980 }, { "grad_norm": 0.06960118561983109, "learning_rate": 2.1787728375752042e-06, "loss": 0.0024, "step": 4990 }, { "grad_norm": 0.0939183309674263, "learning_rate": 2.136933545716257e-06, "loss": 0.0037, "step": 5000 }, { "grad_norm": 0.0371067188680172, "learning_rate": 2.095469051831561e-06, "loss": 0.0029, "step": 5010 }, { "grad_norm": 0.10542552173137665, "learning_rate": 2.0543805640940117e-06, "loss": 0.0034, "step": 5020 }, { "grad_norm": 0.08033856749534607, "learning_rate": 2.0136692797206276e-06, "loss": 0.0029, "step": 5030 }, { "grad_norm": 0.048041995614767075, "learning_rate": 1.973336384937634e-06, "loss": 0.0028, "step": 5040 }, { "grad_norm": 0.07258376479148865, "learning_rate": 1.9333830549459415e-06, "loss": 0.0021, "step": 5050 }, { "grad_norm": 0.07235071808099747, "learning_rate": 1.8938104538868595e-06, "loss": 0.0025, "step": 5060 }, { "grad_norm": 0.04660137742757797, "learning_rate": 1.854619734808209e-06, "loss": 0.002, "step": 5070 }, { "grad_norm": 0.06338319182395935, "learning_rate": 1.815812039630716e-06, "loss": 0.0022, "step": 5080 }, { "grad_norm": 0.06109411269426346, "learning_rate": 1.7773884991147243e-06, "loss": 0.0028, "step": 5090 }, { "grad_norm": 0.04851362481713295, "learning_rate": 1.7393502328272699e-06, "loss": 0.0022, "step": 5100 }, { "grad_norm": 0.08846496045589447, "learning_rate": 1.7016983491094513e-06, "loss": 0.0024, "step": 5110 }, { "grad_norm": 0.04170979931950569, "learning_rate": 1.6644339450441214e-06, "loss": 0.0024, "step": 5120 }, { "grad_norm": 0.04883277043700218, "learning_rate": 1.6275581064239497e-06, "loss": 0.0027, "step": 5130 }, { "grad_norm": 0.09936371445655823, "learning_rate": 1.5910719077197516e-06, "loss": 0.0032, "step": 5140 }, { "grad_norm": 0.028741255402565002, "learning_rate": 1.554976412049211e-06, "loss": 0.0025, "step": 5150 }, { "grad_norm": 0.03936922922730446, "learning_rate": 1.5192726711458899e-06, "loss": 0.0032, "step": 5160 }, { "grad_norm": 0.03335015848278999, "learning_rate": 1.4839617253285758e-06, "loss": 0.0019, "step": 5170 }, { "grad_norm": 0.03196994587779045, "learning_rate": 1.449044603470987e-06, "loss": 0.0018, "step": 5180 }, { "grad_norm": 0.0717940479516983, "learning_rate": 1.414522322971772e-06, "loss": 0.0026, "step": 5190 }, { "grad_norm": 0.025321830064058304, "learning_rate": 1.3803958897249003e-06, "loss": 0.0025, "step": 5200 }, { "grad_norm": 0.054642051458358765, "learning_rate": 1.3466662980903065e-06, "loss": 0.0026, "step": 5210 }, { "grad_norm": 0.03861148655414581, "learning_rate": 1.313334530864953e-06, "loss": 0.0027, "step": 5220 }, { "grad_norm": 0.07443683594465256, "learning_rate": 1.2804015592541856e-06, "loss": 0.0019, "step": 5230 }, { "grad_norm": 0.04590648040175438, "learning_rate": 1.247868342843419e-06, "loss": 0.0028, "step": 5240 }, { "grad_norm": 0.02897215634584427, "learning_rate": 1.2157358295701981e-06, "loss": 0.0026, "step": 5250 }, { "grad_norm": 0.043594203889369965, "learning_rate": 1.1840049556965693e-06, "loss": 0.0043, "step": 5260 }, { "grad_norm": 0.0517614483833313, "learning_rate": 1.1526766457817916e-06, "loss": 0.0025, "step": 5270 }, { "grad_norm": 0.07484445720911026, "learning_rate": 1.1217518126554156e-06, "loss": 0.0022, "step": 5280 }, { "grad_norm": 0.03570960462093353, "learning_rate": 1.0912313573906617e-06, "loss": 0.0019, "step": 5290 }, { "grad_norm": 0.05601223558187485, "learning_rate": 1.0611161692781885e-06, "loss": 0.0023, "step": 5300 }, { "grad_norm": 0.021517347544431686, "learning_rate": 1.0314071258001722e-06, "loss": 0.0021, "step": 5310 }, { "grad_norm": 0.03928108140826225, "learning_rate": 1.0021050926047315e-06, "loss": 0.002, "step": 5320 }, { "grad_norm": 0.032030925154685974, "learning_rate": 9.732109234807124e-07, "loss": 0.0025, "step": 5330 }, { "grad_norm": 0.050179291516542435, "learning_rate": 9.447254603328142e-07, "loss": 0.0029, "step": 5340 }, { "grad_norm": 0.02549734152853489, "learning_rate": 9.166495331570491e-07, "loss": 0.0018, "step": 5350 }, { "grad_norm": 0.04321710392832756, "learning_rate": 8.889839600165656e-07, "loss": 0.0025, "step": 5360 }, { "grad_norm": 0.06186482310295105, "learning_rate": 8.617295470178039e-07, "loss": 0.0023, "step": 5370 }, { "grad_norm": 0.04183085635304451, "learning_rate": 8.348870882870269e-07, "loss": 0.002, "step": 5380 }, { "grad_norm": 0.019996939226984978, "learning_rate": 8.084573659471512e-07, "loss": 0.0019, "step": 5390 }, { "grad_norm": 0.06442452222108841, "learning_rate": 7.82441150094988e-07, "loss": 0.0037, "step": 5400 }, { "grad_norm": 0.077437624335289, "learning_rate": 7.568391987787849e-07, "loss": 0.0023, "step": 5410 }, { "grad_norm": 0.027246888726949692, "learning_rate": 7.316522579761431e-07, "loss": 0.0021, "step": 5420 }, { "grad_norm": 0.0590815469622612, "learning_rate": 7.068810615722887e-07, "loss": 0.0024, "step": 5430 }, { "grad_norm": 0.03910129517316818, "learning_rate": 6.825263313386793e-07, "loss": 0.0018, "step": 5440 }, { "grad_norm": 0.04055808112025261, "learning_rate": 6.585887769119758e-07, "loss": 0.0022, "step": 5450 }, { "grad_norm": 0.032348066568374634, "learning_rate": 6.350690957733712e-07, "loss": 0.0017, "step": 5460 }, { "grad_norm": 0.028897378593683243, "learning_rate": 6.119679732282596e-07, "loss": 0.0022, "step": 5470 }, { "grad_norm": 0.03753242269158363, "learning_rate": 5.89286082386274e-07, "loss": 0.002, "step": 5480 }, { "grad_norm": 0.037839531898498535, "learning_rate": 5.670240841416718e-07, "loss": 0.0022, "step": 5490 }, { "grad_norm": 0.05074967443943024, "learning_rate": 5.451826271540744e-07, "loss": 0.0032, "step": 5500 } ], "logging_steps": 10, "max_steps": 6000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 250, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }