| { |
| "best_global_step": 2000, |
| "best_metric": 1.5431796145591616, |
| "best_model_checkpoint": "./SALAMA_NEW8/checkpoint-2000", |
| "epoch": 0.798881565807869, |
| "eval_steps": 2000, |
| "global_step": 2000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.003994407829039345, |
| "grad_norm": 2.2177462577819824, |
| "learning_rate": 1.8e-07, |
| "loss": 0.022, |
| "step": 10 |
| }, |
| { |
| "epoch": 0.00798881565807869, |
| "grad_norm": 2.724672794342041, |
| "learning_rate": 3.8e-07, |
| "loss": 0.0249, |
| "step": 20 |
| }, |
| { |
| "epoch": 0.011983223487118035, |
| "grad_norm": 2.5243964195251465, |
| "learning_rate": 5.800000000000001e-07, |
| "loss": 0.0215, |
| "step": 30 |
| }, |
| { |
| "epoch": 0.01597763131615738, |
| "grad_norm": 2.4517173767089844, |
| "learning_rate": 7.8e-07, |
| "loss": 0.0229, |
| "step": 40 |
| }, |
| { |
| "epoch": 0.019972039145196723, |
| "grad_norm": 1.1104662418365479, |
| "learning_rate": 9.800000000000001e-07, |
| "loss": 0.0212, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.02396644697423607, |
| "grad_norm": 3.2835190296173096, |
| "learning_rate": 1.1800000000000001e-06, |
| "loss": 0.0177, |
| "step": 60 |
| }, |
| { |
| "epoch": 0.027960854803275415, |
| "grad_norm": 2.933971405029297, |
| "learning_rate": 1.3800000000000001e-06, |
| "loss": 0.0181, |
| "step": 70 |
| }, |
| { |
| "epoch": 0.03195526263231476, |
| "grad_norm": 2.667775869369507, |
| "learning_rate": 1.5800000000000001e-06, |
| "loss": 0.0243, |
| "step": 80 |
| }, |
| { |
| "epoch": 0.035949670461354104, |
| "grad_norm": 2.8503150939941406, |
| "learning_rate": 1.7800000000000001e-06, |
| "loss": 0.0158, |
| "step": 90 |
| }, |
| { |
| "epoch": 0.039944078290393446, |
| "grad_norm": 2.6173582077026367, |
| "learning_rate": 1.98e-06, |
| "loss": 0.0217, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.043938486119432796, |
| "grad_norm": 2.148167371749878, |
| "learning_rate": 2.1800000000000003e-06, |
| "loss": 0.0216, |
| "step": 110 |
| }, |
| { |
| "epoch": 0.04793289394847214, |
| "grad_norm": 1.6248950958251953, |
| "learning_rate": 2.38e-06, |
| "loss": 0.0211, |
| "step": 120 |
| }, |
| { |
| "epoch": 0.05192730177751148, |
| "grad_norm": 2.5831820964813232, |
| "learning_rate": 2.5800000000000003e-06, |
| "loss": 0.0282, |
| "step": 130 |
| }, |
| { |
| "epoch": 0.05592170960655083, |
| "grad_norm": 2.1337149143218994, |
| "learning_rate": 2.7800000000000005e-06, |
| "loss": 0.0215, |
| "step": 140 |
| }, |
| { |
| "epoch": 0.05991611743559017, |
| "grad_norm": 2.95695161819458, |
| "learning_rate": 2.9800000000000003e-06, |
| "loss": 0.0177, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.06391052526462952, |
| "grad_norm": 2.277665853500366, |
| "learning_rate": 3.1800000000000005e-06, |
| "loss": 0.0217, |
| "step": 160 |
| }, |
| { |
| "epoch": 0.06790493309366886, |
| "grad_norm": 2.531909704208374, |
| "learning_rate": 3.3800000000000007e-06, |
| "loss": 0.0375, |
| "step": 170 |
| }, |
| { |
| "epoch": 0.07189934092270821, |
| "grad_norm": 2.5161631107330322, |
| "learning_rate": 3.58e-06, |
| "loss": 0.0262, |
| "step": 180 |
| }, |
| { |
| "epoch": 0.07589374875174755, |
| "grad_norm": 2.9737093448638916, |
| "learning_rate": 3.7800000000000002e-06, |
| "loss": 0.0237, |
| "step": 190 |
| }, |
| { |
| "epoch": 0.07988815658078689, |
| "grad_norm": 2.2957046031951904, |
| "learning_rate": 3.980000000000001e-06, |
| "loss": 0.0211, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.08388256440982625, |
| "grad_norm": 1.9400784969329834, |
| "learning_rate": 4.18e-06, |
| "loss": 0.0241, |
| "step": 210 |
| }, |
| { |
| "epoch": 0.08787697223886559, |
| "grad_norm": 3.7293314933776855, |
| "learning_rate": 4.38e-06, |
| "loss": 0.0263, |
| "step": 220 |
| }, |
| { |
| "epoch": 0.09187138006790493, |
| "grad_norm": 4.266306400299072, |
| "learning_rate": 4.58e-06, |
| "loss": 0.0387, |
| "step": 230 |
| }, |
| { |
| "epoch": 0.09586578789694428, |
| "grad_norm": 2.510282278060913, |
| "learning_rate": 4.78e-06, |
| "loss": 0.029, |
| "step": 240 |
| }, |
| { |
| "epoch": 0.09986019572598362, |
| "grad_norm": 1.4938842058181763, |
| "learning_rate": 4.980000000000001e-06, |
| "loss": 0.0154, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.10385460355502296, |
| "grad_norm": 2.0076608657836914, |
| "learning_rate": 5.18e-06, |
| "loss": 0.0263, |
| "step": 260 |
| }, |
| { |
| "epoch": 0.10784901138406232, |
| "grad_norm": 2.252916097640991, |
| "learning_rate": 5.380000000000001e-06, |
| "loss": 0.0273, |
| "step": 270 |
| }, |
| { |
| "epoch": 0.11184341921310166, |
| "grad_norm": 0.9760869145393372, |
| "learning_rate": 5.580000000000001e-06, |
| "loss": 0.0227, |
| "step": 280 |
| }, |
| { |
| "epoch": 0.115837827042141, |
| "grad_norm": 1.7485207319259644, |
| "learning_rate": 5.78e-06, |
| "loss": 0.0234, |
| "step": 290 |
| }, |
| { |
| "epoch": 0.11983223487118035, |
| "grad_norm": 2.540903329849243, |
| "learning_rate": 5.98e-06, |
| "loss": 0.0216, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.12382664270021969, |
| "grad_norm": 2.970639228820801, |
| "learning_rate": 6.18e-06, |
| "loss": 0.0244, |
| "step": 310 |
| }, |
| { |
| "epoch": 0.12782105052925904, |
| "grad_norm": 4.563722610473633, |
| "learning_rate": 6.380000000000001e-06, |
| "loss": 0.0332, |
| "step": 320 |
| }, |
| { |
| "epoch": 0.1318154583582984, |
| "grad_norm": 3.670563220977783, |
| "learning_rate": 6.5800000000000005e-06, |
| "loss": 0.0259, |
| "step": 330 |
| }, |
| { |
| "epoch": 0.13580986618733773, |
| "grad_norm": 3.6714515686035156, |
| "learning_rate": 6.780000000000001e-06, |
| "loss": 0.032, |
| "step": 340 |
| }, |
| { |
| "epoch": 0.13980427401637707, |
| "grad_norm": 2.029430389404297, |
| "learning_rate": 6.98e-06, |
| "loss": 0.0264, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.14379868184541642, |
| "grad_norm": 2.609748363494873, |
| "learning_rate": 7.180000000000001e-06, |
| "loss": 0.0369, |
| "step": 360 |
| }, |
| { |
| "epoch": 0.14779308967445576, |
| "grad_norm": 1.438359260559082, |
| "learning_rate": 7.3800000000000005e-06, |
| "loss": 0.0242, |
| "step": 370 |
| }, |
| { |
| "epoch": 0.1517874975034951, |
| "grad_norm": 2.1897497177124023, |
| "learning_rate": 7.58e-06, |
| "loss": 0.02, |
| "step": 380 |
| }, |
| { |
| "epoch": 0.15578190533253444, |
| "grad_norm": 1.9179186820983887, |
| "learning_rate": 7.78e-06, |
| "loss": 0.0366, |
| "step": 390 |
| }, |
| { |
| "epoch": 0.15977631316157379, |
| "grad_norm": 2.6540606021881104, |
| "learning_rate": 7.980000000000002e-06, |
| "loss": 0.0253, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.16377072099061313, |
| "grad_norm": 2.5292742252349854, |
| "learning_rate": 8.18e-06, |
| "loss": 0.0356, |
| "step": 410 |
| }, |
| { |
| "epoch": 0.1677651288196525, |
| "grad_norm": 2.4729959964752197, |
| "learning_rate": 8.380000000000001e-06, |
| "loss": 0.0327, |
| "step": 420 |
| }, |
| { |
| "epoch": 0.17175953664869184, |
| "grad_norm": 3.639272689819336, |
| "learning_rate": 8.580000000000001e-06, |
| "loss": 0.032, |
| "step": 430 |
| }, |
| { |
| "epoch": 0.17575394447773118, |
| "grad_norm": 3.2385575771331787, |
| "learning_rate": 8.78e-06, |
| "loss": 0.0359, |
| "step": 440 |
| }, |
| { |
| "epoch": 0.17974835230677053, |
| "grad_norm": 3.476755142211914, |
| "learning_rate": 8.98e-06, |
| "loss": 0.0286, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.18374276013580987, |
| "grad_norm": 2.612086772918701, |
| "learning_rate": 9.180000000000002e-06, |
| "loss": 0.0228, |
| "step": 460 |
| }, |
| { |
| "epoch": 0.1877371679648492, |
| "grad_norm": 3.5403027534484863, |
| "learning_rate": 9.38e-06, |
| "loss": 0.0243, |
| "step": 470 |
| }, |
| { |
| "epoch": 0.19173157579388855, |
| "grad_norm": 2.7765207290649414, |
| "learning_rate": 9.58e-06, |
| "loss": 0.0348, |
| "step": 480 |
| }, |
| { |
| "epoch": 0.1957259836229279, |
| "grad_norm": 3.589587450027466, |
| "learning_rate": 9.780000000000001e-06, |
| "loss": 0.0323, |
| "step": 490 |
| }, |
| { |
| "epoch": 0.19972039145196724, |
| "grad_norm": 1.8719650506973267, |
| "learning_rate": 9.980000000000001e-06, |
| "loss": 0.0275, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.20371479928100658, |
| "grad_norm": 2.750931739807129, |
| "learning_rate": 9.983636363636364e-06, |
| "loss": 0.0348, |
| "step": 510 |
| }, |
| { |
| "epoch": 0.20770920711004592, |
| "grad_norm": 1.9383965730667114, |
| "learning_rate": 9.965454545454546e-06, |
| "loss": 0.0282, |
| "step": 520 |
| }, |
| { |
| "epoch": 0.2117036149390853, |
| "grad_norm": 3.6504414081573486, |
| "learning_rate": 9.947272727272728e-06, |
| "loss": 0.0316, |
| "step": 530 |
| }, |
| { |
| "epoch": 0.21569802276812464, |
| "grad_norm": 3.26383376121521, |
| "learning_rate": 9.92909090909091e-06, |
| "loss": 0.0273, |
| "step": 540 |
| }, |
| { |
| "epoch": 0.21969243059716398, |
| "grad_norm": 4.769904613494873, |
| "learning_rate": 9.910909090909092e-06, |
| "loss": 0.0295, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.22368683842620332, |
| "grad_norm": 1.143069863319397, |
| "learning_rate": 9.892727272727273e-06, |
| "loss": 0.027, |
| "step": 560 |
| }, |
| { |
| "epoch": 0.22768124625524266, |
| "grad_norm": 2.988460063934326, |
| "learning_rate": 9.874545454545455e-06, |
| "loss": 0.0349, |
| "step": 570 |
| }, |
| { |
| "epoch": 0.231675654084282, |
| "grad_norm": 2.6448323726654053, |
| "learning_rate": 9.856363636363637e-06, |
| "loss": 0.036, |
| "step": 580 |
| }, |
| { |
| "epoch": 0.23567006191332135, |
| "grad_norm": 3.077092409133911, |
| "learning_rate": 9.838181818181819e-06, |
| "loss": 0.0491, |
| "step": 590 |
| }, |
| { |
| "epoch": 0.2396644697423607, |
| "grad_norm": 6.214341163635254, |
| "learning_rate": 9.820000000000001e-06, |
| "loss": 0.0313, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.24365887757140003, |
| "grad_norm": 3.1780426502227783, |
| "learning_rate": 9.801818181818183e-06, |
| "loss": 0.0251, |
| "step": 610 |
| }, |
| { |
| "epoch": 0.24765328540043938, |
| "grad_norm": 1.7672063112258911, |
| "learning_rate": 9.783636363636365e-06, |
| "loss": 0.0259, |
| "step": 620 |
| }, |
| { |
| "epoch": 0.2516476932294787, |
| "grad_norm": 3.9317915439605713, |
| "learning_rate": 9.765454545454546e-06, |
| "loss": 0.0263, |
| "step": 630 |
| }, |
| { |
| "epoch": 0.2556421010585181, |
| "grad_norm": 2.23710298538208, |
| "learning_rate": 9.747272727272728e-06, |
| "loss": 0.0335, |
| "step": 640 |
| }, |
| { |
| "epoch": 0.2596365088875574, |
| "grad_norm": 3.24867844581604, |
| "learning_rate": 9.72909090909091e-06, |
| "loss": 0.0383, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.2636309167165968, |
| "grad_norm": 3.0995874404907227, |
| "learning_rate": 9.710909090909092e-06, |
| "loss": 0.0296, |
| "step": 660 |
| }, |
| { |
| "epoch": 0.2676253245456361, |
| "grad_norm": 1.4767210483551025, |
| "learning_rate": 9.692727272727274e-06, |
| "loss": 0.04, |
| "step": 670 |
| }, |
| { |
| "epoch": 0.27161973237467546, |
| "grad_norm": 1.7182132005691528, |
| "learning_rate": 9.674545454545456e-06, |
| "loss": 0.0294, |
| "step": 680 |
| }, |
| { |
| "epoch": 0.2756141402037148, |
| "grad_norm": 3.0770034790039062, |
| "learning_rate": 9.656363636363637e-06, |
| "loss": 0.0393, |
| "step": 690 |
| }, |
| { |
| "epoch": 0.27960854803275415, |
| "grad_norm": 2.452723503112793, |
| "learning_rate": 9.63818181818182e-06, |
| "loss": 0.0321, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.2836029558617935, |
| "grad_norm": 2.9560225009918213, |
| "learning_rate": 9.620000000000001e-06, |
| "loss": 0.0292, |
| "step": 710 |
| }, |
| { |
| "epoch": 0.28759736369083283, |
| "grad_norm": 2.8095853328704834, |
| "learning_rate": 9.601818181818183e-06, |
| "loss": 0.028, |
| "step": 720 |
| }, |
| { |
| "epoch": 0.2915917715198722, |
| "grad_norm": 3.4851746559143066, |
| "learning_rate": 9.583636363636365e-06, |
| "loss": 0.0296, |
| "step": 730 |
| }, |
| { |
| "epoch": 0.2955861793489115, |
| "grad_norm": 4.2011027336120605, |
| "learning_rate": 9.565454545454547e-06, |
| "loss": 0.0353, |
| "step": 740 |
| }, |
| { |
| "epoch": 0.2995805871779509, |
| "grad_norm": 5.23146390914917, |
| "learning_rate": 9.547272727272728e-06, |
| "loss": 0.0393, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.3035749950069902, |
| "grad_norm": 2.8136441707611084, |
| "learning_rate": 9.52909090909091e-06, |
| "loss": 0.0339, |
| "step": 760 |
| }, |
| { |
| "epoch": 0.30756940283602957, |
| "grad_norm": 3.0712337493896484, |
| "learning_rate": 9.510909090909092e-06, |
| "loss": 0.0414, |
| "step": 770 |
| }, |
| { |
| "epoch": 0.3115638106650689, |
| "grad_norm": 2.1798360347747803, |
| "learning_rate": 9.492727272727274e-06, |
| "loss": 0.03, |
| "step": 780 |
| }, |
| { |
| "epoch": 0.31555821849410826, |
| "grad_norm": 3.7893567085266113, |
| "learning_rate": 9.474545454545456e-06, |
| "loss": 0.0325, |
| "step": 790 |
| }, |
| { |
| "epoch": 0.31955262632314757, |
| "grad_norm": 2.0599608421325684, |
| "learning_rate": 9.456363636363638e-06, |
| "loss": 0.0389, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.32354703415218694, |
| "grad_norm": 2.7613537311553955, |
| "learning_rate": 9.438181818181818e-06, |
| "loss": 0.0284, |
| "step": 810 |
| }, |
| { |
| "epoch": 0.32754144198122626, |
| "grad_norm": 2.630587577819824, |
| "learning_rate": 9.42e-06, |
| "loss": 0.0308, |
| "step": 820 |
| }, |
| { |
| "epoch": 0.3315358498102656, |
| "grad_norm": 2.8372304439544678, |
| "learning_rate": 9.401818181818183e-06, |
| "loss": 0.029, |
| "step": 830 |
| }, |
| { |
| "epoch": 0.335530257639305, |
| "grad_norm": 2.21663236618042, |
| "learning_rate": 9.383636363636365e-06, |
| "loss": 0.0367, |
| "step": 840 |
| }, |
| { |
| "epoch": 0.3395246654683443, |
| "grad_norm": 3.325446367263794, |
| "learning_rate": 9.365454545454547e-06, |
| "loss": 0.0256, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.3435190732973837, |
| "grad_norm": 2.7013704776763916, |
| "learning_rate": 9.347272727272729e-06, |
| "loss": 0.0279, |
| "step": 860 |
| }, |
| { |
| "epoch": 0.347513481126423, |
| "grad_norm": 1.6712983846664429, |
| "learning_rate": 9.32909090909091e-06, |
| "loss": 0.0256, |
| "step": 870 |
| }, |
| { |
| "epoch": 0.35150788895546237, |
| "grad_norm": 3.0498671531677246, |
| "learning_rate": 9.310909090909092e-06, |
| "loss": 0.0351, |
| "step": 880 |
| }, |
| { |
| "epoch": 0.3555022967845017, |
| "grad_norm": 2.060898542404175, |
| "learning_rate": 9.292727272727272e-06, |
| "loss": 0.0349, |
| "step": 890 |
| }, |
| { |
| "epoch": 0.35949670461354105, |
| "grad_norm": 2.191037178039551, |
| "learning_rate": 9.274545454545454e-06, |
| "loss": 0.0236, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.36349111244258037, |
| "grad_norm": 2.7442283630371094, |
| "learning_rate": 9.256363636363636e-06, |
| "loss": 0.0339, |
| "step": 910 |
| }, |
| { |
| "epoch": 0.36748552027161974, |
| "grad_norm": 2.151813268661499, |
| "learning_rate": 9.23818181818182e-06, |
| "loss": 0.0272, |
| "step": 920 |
| }, |
| { |
| "epoch": 0.37147992810065905, |
| "grad_norm": 3.0903069972991943, |
| "learning_rate": 9.220000000000002e-06, |
| "loss": 0.0343, |
| "step": 930 |
| }, |
| { |
| "epoch": 0.3754743359296984, |
| "grad_norm": 3.2857322692871094, |
| "learning_rate": 9.201818181818183e-06, |
| "loss": 0.028, |
| "step": 940 |
| }, |
| { |
| "epoch": 0.3794687437587378, |
| "grad_norm": 2.7144224643707275, |
| "learning_rate": 9.183636363636365e-06, |
| "loss": 0.0224, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.3834631515877771, |
| "grad_norm": 2.985253095626831, |
| "learning_rate": 9.165454545454547e-06, |
| "loss": 0.0297, |
| "step": 960 |
| }, |
| { |
| "epoch": 0.3874575594168165, |
| "grad_norm": 2.6801540851593018, |
| "learning_rate": 9.147272727272727e-06, |
| "loss": 0.0306, |
| "step": 970 |
| }, |
| { |
| "epoch": 0.3914519672458558, |
| "grad_norm": 2.944761276245117, |
| "learning_rate": 9.129090909090909e-06, |
| "loss": 0.0324, |
| "step": 980 |
| }, |
| { |
| "epoch": 0.39544637507489516, |
| "grad_norm": 2.6776340007781982, |
| "learning_rate": 9.11090909090909e-06, |
| "loss": 0.04, |
| "step": 990 |
| }, |
| { |
| "epoch": 0.3994407829039345, |
| "grad_norm": 4.007734775543213, |
| "learning_rate": 9.092727272727273e-06, |
| "loss": 0.0286, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.40343519073297385, |
| "grad_norm": 3.652127265930176, |
| "learning_rate": 9.074545454545455e-06, |
| "loss": 0.0373, |
| "step": 1010 |
| }, |
| { |
| "epoch": 0.40742959856201316, |
| "grad_norm": 1.8009936809539795, |
| "learning_rate": 9.056363636363638e-06, |
| "loss": 0.024, |
| "step": 1020 |
| }, |
| { |
| "epoch": 0.41142400639105253, |
| "grad_norm": 2.0933914184570312, |
| "learning_rate": 9.03818181818182e-06, |
| "loss": 0.0341, |
| "step": 1030 |
| }, |
| { |
| "epoch": 0.41541841422009185, |
| "grad_norm": 2.683140277862549, |
| "learning_rate": 9.020000000000002e-06, |
| "loss": 0.0473, |
| "step": 1040 |
| }, |
| { |
| "epoch": 0.4194128220491312, |
| "grad_norm": 2.7582523822784424, |
| "learning_rate": 9.001818181818182e-06, |
| "loss": 0.0322, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.4234072298781706, |
| "grad_norm": 3.6831445693969727, |
| "learning_rate": 8.983636363636364e-06, |
| "loss": 0.0382, |
| "step": 1060 |
| }, |
| { |
| "epoch": 0.4274016377072099, |
| "grad_norm": 2.053619623184204, |
| "learning_rate": 8.965454545454546e-06, |
| "loss": 0.0304, |
| "step": 1070 |
| }, |
| { |
| "epoch": 0.4313960455362493, |
| "grad_norm": 2.5573840141296387, |
| "learning_rate": 8.947272727272727e-06, |
| "loss": 0.0567, |
| "step": 1080 |
| }, |
| { |
| "epoch": 0.4353904533652886, |
| "grad_norm": 3.5455336570739746, |
| "learning_rate": 8.92909090909091e-06, |
| "loss": 0.0338, |
| "step": 1090 |
| }, |
| { |
| "epoch": 0.43938486119432796, |
| "grad_norm": 2.8842337131500244, |
| "learning_rate": 8.910909090909091e-06, |
| "loss": 0.0314, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.4433792690233673, |
| "grad_norm": 2.779266834259033, |
| "learning_rate": 8.892727272727275e-06, |
| "loss": 0.0486, |
| "step": 1110 |
| }, |
| { |
| "epoch": 0.44737367685240664, |
| "grad_norm": 3.959299325942993, |
| "learning_rate": 8.874545454545456e-06, |
| "loss": 0.0396, |
| "step": 1120 |
| }, |
| { |
| "epoch": 0.45136808468144596, |
| "grad_norm": 2.053091526031494, |
| "learning_rate": 8.856363636363637e-06, |
| "loss": 0.0329, |
| "step": 1130 |
| }, |
| { |
| "epoch": 0.45536249251048533, |
| "grad_norm": 1.1432011127471924, |
| "learning_rate": 8.838181818181818e-06, |
| "loss": 0.0268, |
| "step": 1140 |
| }, |
| { |
| "epoch": 0.45935690033952464, |
| "grad_norm": 2.738510847091675, |
| "learning_rate": 8.82e-06, |
| "loss": 0.0388, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.463351308168564, |
| "grad_norm": 1.9435960054397583, |
| "learning_rate": 8.801818181818182e-06, |
| "loss": 0.0321, |
| "step": 1160 |
| }, |
| { |
| "epoch": 0.46734571599760333, |
| "grad_norm": 2.462301015853882, |
| "learning_rate": 8.783636363636364e-06, |
| "loss": 0.0344, |
| "step": 1170 |
| }, |
| { |
| "epoch": 0.4713401238266427, |
| "grad_norm": 2.8463451862335205, |
| "learning_rate": 8.765454545454546e-06, |
| "loss": 0.0372, |
| "step": 1180 |
| }, |
| { |
| "epoch": 0.47533453165568207, |
| "grad_norm": 4.407367706298828, |
| "learning_rate": 8.747272727272728e-06, |
| "loss": 0.0399, |
| "step": 1190 |
| }, |
| { |
| "epoch": 0.4793289394847214, |
| "grad_norm": 2.5179121494293213, |
| "learning_rate": 8.72909090909091e-06, |
| "loss": 0.0331, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.48332334731376075, |
| "grad_norm": 3.2438509464263916, |
| "learning_rate": 8.710909090909091e-06, |
| "loss": 0.0348, |
| "step": 1210 |
| }, |
| { |
| "epoch": 0.48731775514280007, |
| "grad_norm": 2.54004168510437, |
| "learning_rate": 8.692727272727273e-06, |
| "loss": 0.0317, |
| "step": 1220 |
| }, |
| { |
| "epoch": 0.49131216297183944, |
| "grad_norm": 2.8079185485839844, |
| "learning_rate": 8.674545454545455e-06, |
| "loss": 0.0386, |
| "step": 1230 |
| }, |
| { |
| "epoch": 0.49530657080087875, |
| "grad_norm": 4.827033519744873, |
| "learning_rate": 8.656363636363637e-06, |
| "loss": 0.0338, |
| "step": 1240 |
| }, |
| { |
| "epoch": 0.4993009786299181, |
| "grad_norm": 2.919968843460083, |
| "learning_rate": 8.638181818181819e-06, |
| "loss": 0.0265, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.5032953864589574, |
| "grad_norm": 3.6935982704162598, |
| "learning_rate": 8.62e-06, |
| "loss": 0.0327, |
| "step": 1260 |
| }, |
| { |
| "epoch": 0.5072897942879968, |
| "grad_norm": 4.243749141693115, |
| "learning_rate": 8.601818181818182e-06, |
| "loss": 0.0282, |
| "step": 1270 |
| }, |
| { |
| "epoch": 0.5112842021170362, |
| "grad_norm": 3.979485273361206, |
| "learning_rate": 8.583636363636364e-06, |
| "loss": 0.0242, |
| "step": 1280 |
| }, |
| { |
| "epoch": 0.5152786099460755, |
| "grad_norm": 3.32183575630188, |
| "learning_rate": 8.565454545454546e-06, |
| "loss": 0.0356, |
| "step": 1290 |
| }, |
| { |
| "epoch": 0.5192730177751148, |
| "grad_norm": 2.474187135696411, |
| "learning_rate": 8.547272727272728e-06, |
| "loss": 0.0201, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.5232674256041542, |
| "grad_norm": 3.150151252746582, |
| "learning_rate": 8.52909090909091e-06, |
| "loss": 0.0304, |
| "step": 1310 |
| }, |
| { |
| "epoch": 0.5272618334331935, |
| "grad_norm": 2.039504289627075, |
| "learning_rate": 8.510909090909092e-06, |
| "loss": 0.045, |
| "step": 1320 |
| }, |
| { |
| "epoch": 0.5312562412622329, |
| "grad_norm": 2.7388851642608643, |
| "learning_rate": 8.492727272727273e-06, |
| "loss": 0.0268, |
| "step": 1330 |
| }, |
| { |
| "epoch": 0.5352506490912722, |
| "grad_norm": 2.9932122230529785, |
| "learning_rate": 8.474545454545455e-06, |
| "loss": 0.0331, |
| "step": 1340 |
| }, |
| { |
| "epoch": 0.5392450569203115, |
| "grad_norm": 3.3818037509918213, |
| "learning_rate": 8.456363636363637e-06, |
| "loss": 0.0365, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.5432394647493509, |
| "grad_norm": 3.251274585723877, |
| "learning_rate": 8.438181818181819e-06, |
| "loss": 0.0395, |
| "step": 1360 |
| }, |
| { |
| "epoch": 0.5472338725783903, |
| "grad_norm": 3.46567964553833, |
| "learning_rate": 8.42e-06, |
| "loss": 0.0287, |
| "step": 1370 |
| }, |
| { |
| "epoch": 0.5512282804074295, |
| "grad_norm": 2.459820032119751, |
| "learning_rate": 8.401818181818183e-06, |
| "loss": 0.0329, |
| "step": 1380 |
| }, |
| { |
| "epoch": 0.5552226882364689, |
| "grad_norm": 1.735729694366455, |
| "learning_rate": 8.383636363636364e-06, |
| "loss": 0.0313, |
| "step": 1390 |
| }, |
| { |
| "epoch": 0.5592170960655083, |
| "grad_norm": 2.2887370586395264, |
| "learning_rate": 8.365454545454546e-06, |
| "loss": 0.0215, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.5632115038945477, |
| "grad_norm": 3.110576868057251, |
| "learning_rate": 8.347272727272728e-06, |
| "loss": 0.0213, |
| "step": 1410 |
| }, |
| { |
| "epoch": 0.567205911723587, |
| "grad_norm": 3.1144895553588867, |
| "learning_rate": 8.32909090909091e-06, |
| "loss": 0.0312, |
| "step": 1420 |
| }, |
| { |
| "epoch": 0.5712003195526263, |
| "grad_norm": 2.9777989387512207, |
| "learning_rate": 8.310909090909092e-06, |
| "loss": 0.0322, |
| "step": 1430 |
| }, |
| { |
| "epoch": 0.5751947273816657, |
| "grad_norm": 2.2770936489105225, |
| "learning_rate": 8.292727272727274e-06, |
| "loss": 0.0304, |
| "step": 1440 |
| }, |
| { |
| "epoch": 0.579189135210705, |
| "grad_norm": 3.5356180667877197, |
| "learning_rate": 8.274545454545455e-06, |
| "loss": 0.0347, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.5831835430397444, |
| "grad_norm": 3.6384565830230713, |
| "learning_rate": 8.256363636363637e-06, |
| "loss": 0.0289, |
| "step": 1460 |
| }, |
| { |
| "epoch": 0.5871779508687837, |
| "grad_norm": 3.053424119949341, |
| "learning_rate": 8.238181818181819e-06, |
| "loss": 0.0356, |
| "step": 1470 |
| }, |
| { |
| "epoch": 0.591172358697823, |
| "grad_norm": 4.9105963706970215, |
| "learning_rate": 8.220000000000001e-06, |
| "loss": 0.0315, |
| "step": 1480 |
| }, |
| { |
| "epoch": 0.5951667665268624, |
| "grad_norm": 3.0485212802886963, |
| "learning_rate": 8.201818181818183e-06, |
| "loss": 0.0298, |
| "step": 1490 |
| }, |
| { |
| "epoch": 0.5991611743559018, |
| "grad_norm": 3.3632636070251465, |
| "learning_rate": 8.183636363636365e-06, |
| "loss": 0.0268, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.603155582184941, |
| "grad_norm": 2.593235969543457, |
| "learning_rate": 8.165454545454546e-06, |
| "loss": 0.0286, |
| "step": 1510 |
| }, |
| { |
| "epoch": 0.6071499900139804, |
| "grad_norm": 2.542865753173828, |
| "learning_rate": 8.147272727272728e-06, |
| "loss": 0.0317, |
| "step": 1520 |
| }, |
| { |
| "epoch": 0.6111443978430198, |
| "grad_norm": 3.246321201324463, |
| "learning_rate": 8.12909090909091e-06, |
| "loss": 0.0358, |
| "step": 1530 |
| }, |
| { |
| "epoch": 0.6151388056720591, |
| "grad_norm": 4.592155456542969, |
| "learning_rate": 8.110909090909092e-06, |
| "loss": 0.0295, |
| "step": 1540 |
| }, |
| { |
| "epoch": 0.6191332135010985, |
| "grad_norm": 2.1040351390838623, |
| "learning_rate": 8.092727272727274e-06, |
| "loss": 0.044, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.6231276213301378, |
| "grad_norm": 2.832470417022705, |
| "learning_rate": 8.074545454545456e-06, |
| "loss": 0.0335, |
| "step": 1560 |
| }, |
| { |
| "epoch": 0.6271220291591771, |
| "grad_norm": 3.7737035751342773, |
| "learning_rate": 8.056363636363636e-06, |
| "loss": 0.0289, |
| "step": 1570 |
| }, |
| { |
| "epoch": 0.6311164369882165, |
| "grad_norm": 2.9322657585144043, |
| "learning_rate": 8.038181818181818e-06, |
| "loss": 0.0209, |
| "step": 1580 |
| }, |
| { |
| "epoch": 0.6351108448172559, |
| "grad_norm": 3.0506591796875, |
| "learning_rate": 8.020000000000001e-06, |
| "loss": 0.0277, |
| "step": 1590 |
| }, |
| { |
| "epoch": 0.6391052526462951, |
| "grad_norm": 1.679126501083374, |
| "learning_rate": 8.001818181818183e-06, |
| "loss": 0.0313, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.6430996604753345, |
| "grad_norm": 1.4025191068649292, |
| "learning_rate": 7.983636363636365e-06, |
| "loss": 0.0342, |
| "step": 1610 |
| }, |
| { |
| "epoch": 0.6470940683043739, |
| "grad_norm": 3.2698376178741455, |
| "learning_rate": 7.965454545454547e-06, |
| "loss": 0.0334, |
| "step": 1620 |
| }, |
| { |
| "epoch": 0.6510884761334133, |
| "grad_norm": 2.007560968399048, |
| "learning_rate": 7.947272727272728e-06, |
| "loss": 0.0361, |
| "step": 1630 |
| }, |
| { |
| "epoch": 0.6550828839624525, |
| "grad_norm": 3.021299123764038, |
| "learning_rate": 7.92909090909091e-06, |
| "loss": 0.0316, |
| "step": 1640 |
| }, |
| { |
| "epoch": 0.6590772917914919, |
| "grad_norm": 3.4559459686279297, |
| "learning_rate": 7.91090909090909e-06, |
| "loss": 0.0336, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.6630716996205313, |
| "grad_norm": 3.786959648132324, |
| "learning_rate": 7.892727272727272e-06, |
| "loss": 0.0337, |
| "step": 1660 |
| }, |
| { |
| "epoch": 0.6670661074495706, |
| "grad_norm": 2.5222368240356445, |
| "learning_rate": 7.874545454545454e-06, |
| "loss": 0.026, |
| "step": 1670 |
| }, |
| { |
| "epoch": 0.67106051527861, |
| "grad_norm": 2.880535364151001, |
| "learning_rate": 7.856363636363638e-06, |
| "loss": 0.023, |
| "step": 1680 |
| }, |
| { |
| "epoch": 0.6750549231076493, |
| "grad_norm": 3.375427007675171, |
| "learning_rate": 7.83818181818182e-06, |
| "loss": 0.025, |
| "step": 1690 |
| }, |
| { |
| "epoch": 0.6790493309366886, |
| "grad_norm": 1.8976999521255493, |
| "learning_rate": 7.820000000000001e-06, |
| "loss": 0.0267, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.683043738765728, |
| "grad_norm": 3.0969624519348145, |
| "learning_rate": 7.801818181818183e-06, |
| "loss": 0.0334, |
| "step": 1710 |
| }, |
| { |
| "epoch": 0.6870381465947674, |
| "grad_norm": 2.949564218521118, |
| "learning_rate": 7.783636363636365e-06, |
| "loss": 0.0317, |
| "step": 1720 |
| }, |
| { |
| "epoch": 0.6910325544238066, |
| "grad_norm": 3.1288681030273438, |
| "learning_rate": 7.765454545454545e-06, |
| "loss": 0.036, |
| "step": 1730 |
| }, |
| { |
| "epoch": 0.695026962252846, |
| "grad_norm": 2.337353229522705, |
| "learning_rate": 7.747272727272727e-06, |
| "loss": 0.0233, |
| "step": 1740 |
| }, |
| { |
| "epoch": 0.6990213700818854, |
| "grad_norm": 2.593578577041626, |
| "learning_rate": 7.729090909090909e-06, |
| "loss": 0.0338, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.7030157779109247, |
| "grad_norm": 2.1465232372283936, |
| "learning_rate": 7.71090909090909e-06, |
| "loss": 0.0289, |
| "step": 1760 |
| }, |
| { |
| "epoch": 0.7070101857399641, |
| "grad_norm": 2.932419538497925, |
| "learning_rate": 7.692727272727273e-06, |
| "loss": 0.0263, |
| "step": 1770 |
| }, |
| { |
| "epoch": 0.7110045935690034, |
| "grad_norm": 4.221918106079102, |
| "learning_rate": 7.674545454545456e-06, |
| "loss": 0.0262, |
| "step": 1780 |
| }, |
| { |
| "epoch": 0.7149990013980427, |
| "grad_norm": 3.3574588298797607, |
| "learning_rate": 7.656363636363638e-06, |
| "loss": 0.0298, |
| "step": 1790 |
| }, |
| { |
| "epoch": 0.7189934092270821, |
| "grad_norm": 3.61956787109375, |
| "learning_rate": 7.63818181818182e-06, |
| "loss": 0.0361, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.7229878170561215, |
| "grad_norm": 3.8640527725219727, |
| "learning_rate": 7.620000000000001e-06, |
| "loss": 0.0379, |
| "step": 1810 |
| }, |
| { |
| "epoch": 0.7269822248851607, |
| "grad_norm": 3.021359920501709, |
| "learning_rate": 7.6018181818181826e-06, |
| "loss": 0.0256, |
| "step": 1820 |
| }, |
| { |
| "epoch": 0.7309766327142001, |
| "grad_norm": 3.4771692752838135, |
| "learning_rate": 7.583636363636364e-06, |
| "loss": 0.0317, |
| "step": 1830 |
| }, |
| { |
| "epoch": 0.7349710405432395, |
| "grad_norm": 1.799800992012024, |
| "learning_rate": 7.565454545454546e-06, |
| "loss": 0.0318, |
| "step": 1840 |
| }, |
| { |
| "epoch": 0.7389654483722788, |
| "grad_norm": 2.8619134426116943, |
| "learning_rate": 7.547272727272727e-06, |
| "loss": 0.0351, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.7429598562013181, |
| "grad_norm": 1.9078686237335205, |
| "learning_rate": 7.529090909090909e-06, |
| "loss": 0.0223, |
| "step": 1860 |
| }, |
| { |
| "epoch": 0.7469542640303575, |
| "grad_norm": 3.3508458137512207, |
| "learning_rate": 7.510909090909092e-06, |
| "loss": 0.0219, |
| "step": 1870 |
| }, |
| { |
| "epoch": 0.7509486718593968, |
| "grad_norm": 2.5900681018829346, |
| "learning_rate": 7.492727272727274e-06, |
| "loss": 0.0231, |
| "step": 1880 |
| }, |
| { |
| "epoch": 0.7549430796884362, |
| "grad_norm": 2.878042697906494, |
| "learning_rate": 7.4745454545454554e-06, |
| "loss": 0.0269, |
| "step": 1890 |
| }, |
| { |
| "epoch": 0.7589374875174756, |
| "grad_norm": 2.814326524734497, |
| "learning_rate": 7.456363636363637e-06, |
| "loss": 0.027, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.7629318953465148, |
| "grad_norm": 3.0135231018066406, |
| "learning_rate": 7.438181818181819e-06, |
| "loss": 0.0363, |
| "step": 1910 |
| }, |
| { |
| "epoch": 0.7669263031755542, |
| "grad_norm": 4.630500793457031, |
| "learning_rate": 7.420000000000001e-06, |
| "loss": 0.0318, |
| "step": 1920 |
| }, |
| { |
| "epoch": 0.7709207110045936, |
| "grad_norm": 2.1792924404144287, |
| "learning_rate": 7.401818181818182e-06, |
| "loss": 0.0299, |
| "step": 1930 |
| }, |
| { |
| "epoch": 0.774915118833633, |
| "grad_norm": 2.716294527053833, |
| "learning_rate": 7.383636363636364e-06, |
| "loss": 0.028, |
| "step": 1940 |
| }, |
| { |
| "epoch": 0.7789095266626722, |
| "grad_norm": 2.591440439224243, |
| "learning_rate": 7.365454545454546e-06, |
| "loss": 0.0311, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.7829039344917116, |
| "grad_norm": 1.3216569423675537, |
| "learning_rate": 7.3472727272727275e-06, |
| "loss": 0.0213, |
| "step": 1960 |
| }, |
| { |
| "epoch": 0.786898342320751, |
| "grad_norm": 1.4774867296218872, |
| "learning_rate": 7.32909090909091e-06, |
| "loss": 0.0218, |
| "step": 1970 |
| }, |
| { |
| "epoch": 0.7908927501497903, |
| "grad_norm": 2.8911545276641846, |
| "learning_rate": 7.310909090909092e-06, |
| "loss": 0.0251, |
| "step": 1980 |
| }, |
| { |
| "epoch": 0.7948871579788296, |
| "grad_norm": 4.558359146118164, |
| "learning_rate": 7.292727272727274e-06, |
| "loss": 0.023, |
| "step": 1990 |
| }, |
| { |
| "epoch": 0.798881565807869, |
| "grad_norm": 2.78124737739563, |
| "learning_rate": 7.274545454545456e-06, |
| "loss": 0.0361, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.798881565807869, |
| "eval_loss": 0.017742320895195007, |
| "eval_runtime": 7487.3562, |
| "eval_samples_per_second": 2.675, |
| "eval_steps_per_second": 0.334, |
| "eval_wer": 1.5431796145591616, |
| "step": 2000 |
| } |
| ], |
| "logging_steps": 10, |
| "max_steps": 6000, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 3, |
| "save_steps": 2000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 1.846946562048e+19, |
| "train_batch_size": 16, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
