{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.375, "eval_steps": 500, "global_step": 2250, "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 } ], "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 }