| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.272984441301273, | |
| "eval_steps": 500, | |
| "global_step": 900, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.014144271570014143, | |
| "grad_norm": 85.7519302368164, | |
| "learning_rate": 4.976426214049977e-05, | |
| "loss": 9.3898, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.028288543140028287, | |
| "grad_norm": 33.235904693603516, | |
| "learning_rate": 4.952852428099953e-05, | |
| "loss": 5.3413, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.042432814710042434, | |
| "grad_norm": 5.113602161407471, | |
| "learning_rate": 4.9292786421499294e-05, | |
| "loss": 4.854, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.056577086280056574, | |
| "grad_norm": 36.069705963134766, | |
| "learning_rate": 4.9057048561999055e-05, | |
| "loss": 5.0425, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07072135785007072, | |
| "grad_norm": 33.71372604370117, | |
| "learning_rate": 4.882131070249882e-05, | |
| "loss": 5.04, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08486562942008487, | |
| "grad_norm": 10.95967960357666, | |
| "learning_rate": 4.858557284299859e-05, | |
| "loss": 4.5402, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09900990099009901, | |
| "grad_norm": 57.9084358215332, | |
| "learning_rate": 4.834983498349835e-05, | |
| "loss": 4.8505, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.11315417256011315, | |
| "grad_norm": 34.007869720458984, | |
| "learning_rate": 4.8114097123998114e-05, | |
| "loss": 4.3305, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.1272984441301273, | |
| "grad_norm": 15.245634078979492, | |
| "learning_rate": 4.787835926449788e-05, | |
| "loss": 4.4707, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.14144271570014144, | |
| "grad_norm": 24.47016143798828, | |
| "learning_rate": 4.7642621404997644e-05, | |
| "loss": 4.5379, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15558698727015557, | |
| "grad_norm": 14.591358184814453, | |
| "learning_rate": 4.740688354549741e-05, | |
| "loss": 4.5205, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.16973125884016974, | |
| "grad_norm": 18.776493072509766, | |
| "learning_rate": 4.7171145685997174e-05, | |
| "loss": 4.6578, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.18387553041018387, | |
| "grad_norm": 4.018069267272949, | |
| "learning_rate": 4.6935407826496935e-05, | |
| "loss": 4.1864, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.19801980198019803, | |
| "grad_norm": 8.038480758666992, | |
| "learning_rate": 4.6699669966996704e-05, | |
| "loss": 4.5046, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.21216407355021216, | |
| "grad_norm": 2.9573426246643066, | |
| "learning_rate": 4.6463932107496465e-05, | |
| "loss": 4.2317, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.2263083451202263, | |
| "grad_norm": 3.0037927627563477, | |
| "learning_rate": 4.622819424799623e-05, | |
| "loss": 4.0951, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.24045261669024046, | |
| "grad_norm": 9.886232376098633, | |
| "learning_rate": 4.5992456388495995e-05, | |
| "loss": 4.0168, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2545968882602546, | |
| "grad_norm": 2.463179588317871, | |
| "learning_rate": 4.5756718528995756e-05, | |
| "loss": 4.0372, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.26874115983026875, | |
| "grad_norm": 13.546555519104004, | |
| "learning_rate": 4.5520980669495525e-05, | |
| "loss": 4.1929, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2828854314002829, | |
| "grad_norm": 2.4467103481292725, | |
| "learning_rate": 4.5285242809995286e-05, | |
| "loss": 4.2146, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.297029702970297, | |
| "grad_norm": 5.936313152313232, | |
| "learning_rate": 4.5049504950495054e-05, | |
| "loss": 4.6069, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.31117397454031115, | |
| "grad_norm": 6.5272536277771, | |
| "learning_rate": 4.4813767090994816e-05, | |
| "loss": 4.3263, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.32531824611032534, | |
| "grad_norm": 4.2881598472595215, | |
| "learning_rate": 4.457802923149458e-05, | |
| "loss": 3.8218, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.33946251768033947, | |
| "grad_norm": 3.9945058822631836, | |
| "learning_rate": 4.4342291371994345e-05, | |
| "loss": 4.1532, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3536067892503536, | |
| "grad_norm": 4.577730655670166, | |
| "learning_rate": 4.410655351249411e-05, | |
| "loss": 4.0727, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.36775106082036774, | |
| "grad_norm": 2.6052353382110596, | |
| "learning_rate": 4.3870815652993875e-05, | |
| "loss": 3.9074, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.38189533239038187, | |
| "grad_norm": 15.787618637084961, | |
| "learning_rate": 4.363507779349364e-05, | |
| "loss": 4.2402, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.39603960396039606, | |
| "grad_norm": 23.970670700073242, | |
| "learning_rate": 4.33993399339934e-05, | |
| "loss": 4.3765, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4101838755304102, | |
| "grad_norm": 2.5313973426818848, | |
| "learning_rate": 4.3163602074493166e-05, | |
| "loss": 3.9863, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4243281471004243, | |
| "grad_norm": 21.020267486572266, | |
| "learning_rate": 4.292786421499293e-05, | |
| "loss": 4.0341, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.43847241867043846, | |
| "grad_norm": 9.731268882751465, | |
| "learning_rate": 4.2692126355492696e-05, | |
| "loss": 4.1089, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.4526166902404526, | |
| "grad_norm": 4.240326881408691, | |
| "learning_rate": 4.245638849599246e-05, | |
| "loss": 3.9737, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4667609618104668, | |
| "grad_norm": 15.72867202758789, | |
| "learning_rate": 4.222065063649222e-05, | |
| "loss": 4.454, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4809052333804809, | |
| "grad_norm": 10.669405937194824, | |
| "learning_rate": 4.198491277699199e-05, | |
| "loss": 4.1596, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.49504950495049505, | |
| "grad_norm": 11.927492141723633, | |
| "learning_rate": 4.174917491749175e-05, | |
| "loss": 4.0485, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5091937765205092, | |
| "grad_norm": 7.629958629608154, | |
| "learning_rate": 4.151343705799152e-05, | |
| "loss": 4.1567, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5233380480905233, | |
| "grad_norm": 32.22209930419922, | |
| "learning_rate": 4.1277699198491285e-05, | |
| "loss": 4.9187, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5374823196605375, | |
| "grad_norm": 7.841526985168457, | |
| "learning_rate": 4.104196133899104e-05, | |
| "loss": 4.0989, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5516265912305516, | |
| "grad_norm": 3.8099868297576904, | |
| "learning_rate": 4.080622347949081e-05, | |
| "loss": 4.4586, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5657708628005658, | |
| "grad_norm": 11.720135688781738, | |
| "learning_rate": 4.057048561999057e-05, | |
| "loss": 4.1915, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.57991513437058, | |
| "grad_norm": 5.8960280418396, | |
| "learning_rate": 4.033474776049034e-05, | |
| "loss": 4.3458, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.594059405940594, | |
| "grad_norm": 3.532780885696411, | |
| "learning_rate": 4.0099009900990106e-05, | |
| "loss": 4.2094, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6082036775106082, | |
| "grad_norm": 6.77498722076416, | |
| "learning_rate": 3.986327204148986e-05, | |
| "loss": 4.3139, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6223479490806223, | |
| "grad_norm": 23.035005569458008, | |
| "learning_rate": 3.962753418198963e-05, | |
| "loss": 4.0213, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6364922206506365, | |
| "grad_norm": 3.033621311187744, | |
| "learning_rate": 3.939179632248939e-05, | |
| "loss": 4.0612, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6506364922206507, | |
| "grad_norm": 32.6967887878418, | |
| "learning_rate": 3.915605846298916e-05, | |
| "loss": 4.1847, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6647807637906648, | |
| "grad_norm": 9.779464721679688, | |
| "learning_rate": 3.892032060348893e-05, | |
| "loss": 4.1741, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.6789250353606789, | |
| "grad_norm": 14.904414176940918, | |
| "learning_rate": 3.868458274398868e-05, | |
| "loss": 4.1909, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.693069306930693, | |
| "grad_norm": 34.94367218017578, | |
| "learning_rate": 3.844884488448845e-05, | |
| "loss": 4.9592, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7072135785007072, | |
| "grad_norm": 6.339000701904297, | |
| "learning_rate": 3.821310702498822e-05, | |
| "loss": 3.7726, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7213578500707214, | |
| "grad_norm": 17.672000885009766, | |
| "learning_rate": 3.797736916548798e-05, | |
| "loss": 3.9599, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.7355021216407355, | |
| "grad_norm": 13.348356246948242, | |
| "learning_rate": 3.774163130598775e-05, | |
| "loss": 3.8342, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7496463932107497, | |
| "grad_norm": 1.8930085897445679, | |
| "learning_rate": 3.75058934464875e-05, | |
| "loss": 4.0049, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.7637906647807637, | |
| "grad_norm": 35.62409210205078, | |
| "learning_rate": 3.727015558698727e-05, | |
| "loss": 3.9412, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.7779349363507779, | |
| "grad_norm": 2.246541738510132, | |
| "learning_rate": 3.703441772748704e-05, | |
| "loss": 3.893, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.7920792079207921, | |
| "grad_norm": 57.89748001098633, | |
| "learning_rate": 3.67986798679868e-05, | |
| "loss": 4.2004, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8062234794908062, | |
| "grad_norm": 13.958605766296387, | |
| "learning_rate": 3.656294200848657e-05, | |
| "loss": 4.5987, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.8203677510608204, | |
| "grad_norm": 7.963130950927734, | |
| "learning_rate": 3.6327204148986324e-05, | |
| "loss": 3.8688, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8345120226308345, | |
| "grad_norm": 12.124194145202637, | |
| "learning_rate": 3.609146628948609e-05, | |
| "loss": 3.8787, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.8486562942008486, | |
| "grad_norm": 19.39701271057129, | |
| "learning_rate": 3.585572842998586e-05, | |
| "loss": 3.8274, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.8628005657708628, | |
| "grad_norm": 7.561882495880127, | |
| "learning_rate": 3.561999057048562e-05, | |
| "loss": 3.917, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.8769448373408769, | |
| "grad_norm": 8.699311256408691, | |
| "learning_rate": 3.538425271098539e-05, | |
| "loss": 3.8819, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.8910891089108911, | |
| "grad_norm": 10.60632038116455, | |
| "learning_rate": 3.514851485148515e-05, | |
| "loss": 4.3288, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9052333804809052, | |
| "grad_norm": 5.851240634918213, | |
| "learning_rate": 3.491277699198491e-05, | |
| "loss": 3.7157, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.9193776520509194, | |
| "grad_norm": 12.624049186706543, | |
| "learning_rate": 3.467703913248468e-05, | |
| "loss": 4.0151, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9335219236209336, | |
| "grad_norm": 10.379075050354004, | |
| "learning_rate": 3.444130127298444e-05, | |
| "loss": 4.1419, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9476661951909476, | |
| "grad_norm": 11.247940063476562, | |
| "learning_rate": 3.420556341348421e-05, | |
| "loss": 3.5958, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.9618104667609618, | |
| "grad_norm": 10.014704704284668, | |
| "learning_rate": 3.396982555398397e-05, | |
| "loss": 3.7822, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.9759547383309759, | |
| "grad_norm": 8.791955947875977, | |
| "learning_rate": 3.3734087694483734e-05, | |
| "loss": 3.9809, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.9900990099009901, | |
| "grad_norm": 30.620357513427734, | |
| "learning_rate": 3.34983498349835e-05, | |
| "loss": 3.9702, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 1.0042432814710043, | |
| "grad_norm": 23.29230499267578, | |
| "learning_rate": 3.326261197548326e-05, | |
| "loss": 3.8758, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 1.0183875530410185, | |
| "grad_norm": 6.364682674407959, | |
| "learning_rate": 3.302687411598303e-05, | |
| "loss": 3.6476, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 1.0325318246110324, | |
| "grad_norm": 22.594091415405273, | |
| "learning_rate": 3.279113625648279e-05, | |
| "loss": 4.2427, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 1.0466760961810466, | |
| "grad_norm": 31.865617752075195, | |
| "learning_rate": 3.2555398396982555e-05, | |
| "loss": 3.8927, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 1.0608203677510608, | |
| "grad_norm": 2.553858757019043, | |
| "learning_rate": 3.231966053748232e-05, | |
| "loss": 3.6406, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 1.074964639321075, | |
| "grad_norm": 24.558555603027344, | |
| "learning_rate": 3.2083922677982084e-05, | |
| "loss": 3.9391, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 1.0891089108910892, | |
| "grad_norm": 3.9098362922668457, | |
| "learning_rate": 3.184818481848185e-05, | |
| "loss": 4.2241, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 1.1032531824611032, | |
| "grad_norm": 14.435652732849121, | |
| "learning_rate": 3.1612446958981614e-05, | |
| "loss": 3.5212, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 1.1173974540311173, | |
| "grad_norm": 4.891509056091309, | |
| "learning_rate": 3.1376709099481375e-05, | |
| "loss": 3.3049, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 1.1315417256011315, | |
| "grad_norm": 10.893304824829102, | |
| "learning_rate": 3.1140971239981144e-05, | |
| "loss": 3.7428, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 1.1456859971711457, | |
| "grad_norm": 3.631542921066284, | |
| "learning_rate": 3.0905233380480905e-05, | |
| "loss": 3.7109, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 1.15983026874116, | |
| "grad_norm": 17.292734146118164, | |
| "learning_rate": 3.0669495520980673e-05, | |
| "loss": 3.748, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 1.1739745403111739, | |
| "grad_norm": 12.438305854797363, | |
| "learning_rate": 3.043375766148043e-05, | |
| "loss": 4.5263, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 1.188118811881188, | |
| "grad_norm": 7.694697380065918, | |
| "learning_rate": 3.01980198019802e-05, | |
| "loss": 3.9575, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 1.2022630834512023, | |
| "grad_norm": 9.88021469116211, | |
| "learning_rate": 2.9962281942479965e-05, | |
| "loss": 3.5793, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 1.2164073550212164, | |
| "grad_norm": 18.16057586669922, | |
| "learning_rate": 2.972654408297973e-05, | |
| "loss": 3.7951, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 1.2305516265912306, | |
| "grad_norm": 3.5214946269989014, | |
| "learning_rate": 2.9490806223479494e-05, | |
| "loss": 4.0263, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 1.2446958981612446, | |
| "grad_norm": 20.135046005249023, | |
| "learning_rate": 2.9255068363979256e-05, | |
| "loss": 3.678, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 1.2588401697312588, | |
| "grad_norm": 35.220733642578125, | |
| "learning_rate": 2.901933050447902e-05, | |
| "loss": 4.1981, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 1.272984441301273, | |
| "grad_norm": 39.2838134765625, | |
| "learning_rate": 2.8783592644978786e-05, | |
| "loss": 3.9173, | |
| "step": 900 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 2121, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.3469133437927424e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |