| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 19.908045977011493, | |
| "eval_steps": 500, | |
| "global_step": 1732, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.11494252873563218, | |
| "grad_norm": 5.833221435546875, | |
| "learning_rate": 2.2988505747126437e-05, | |
| "loss": 1.1691, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.22988505747126436, | |
| "grad_norm": 3.3890936374664307, | |
| "learning_rate": 4.597701149425287e-05, | |
| "loss": 0.4962, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.3448275862068966, | |
| "grad_norm": 1.4635858535766602, | |
| "learning_rate": 6.896551724137931e-05, | |
| "loss": 0.2852, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.45977011494252873, | |
| "grad_norm": 1.3263287544250488, | |
| "learning_rate": 9.195402298850575e-05, | |
| "loss": 0.2068, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.5747126436781609, | |
| "grad_norm": 1.4933586120605469, | |
| "learning_rate": 0.00011494252873563218, | |
| "loss": 0.1671, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.6896551724137931, | |
| "grad_norm": 3.461280584335327, | |
| "learning_rate": 0.00013793103448275863, | |
| "loss": 0.163, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.8045977011494253, | |
| "grad_norm": 0.8741048574447632, | |
| "learning_rate": 0.00016091954022988506, | |
| "loss": 0.1625, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.9195402298850575, | |
| "grad_norm": 0.9238091111183167, | |
| "learning_rate": 0.0001839080459770115, | |
| "loss": 0.1292, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.0344827586206897, | |
| "grad_norm": 1.1964523792266846, | |
| "learning_rate": 0.00019999835873288997, | |
| "loss": 0.1137, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.1494252873563218, | |
| "grad_norm": 0.6014072895050049, | |
| "learning_rate": 0.0001999691821496584, | |
| "loss": 0.1066, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 1.264367816091954, | |
| "grad_norm": 0.6493586897850037, | |
| "learning_rate": 0.00019990354521250803, | |
| "loss": 0.1051, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 1.3793103448275863, | |
| "grad_norm": 0.6839510202407837, | |
| "learning_rate": 0.00019980147186027586, | |
| "loss": 0.0926, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 1.4942528735632183, | |
| "grad_norm": 0.5401099920272827, | |
| "learning_rate": 0.00019966299932074023, | |
| "loss": 0.0895, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.6091954022988506, | |
| "grad_norm": 1.1563774347305298, | |
| "learning_rate": 0.000199488178097043, | |
| "loss": 0.0915, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.7241379310344827, | |
| "grad_norm": 2.1474926471710205, | |
| "learning_rate": 0.00019927707194927066, | |
| "loss": 0.0853, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.839080459770115, | |
| "grad_norm": 0.7267495393753052, | |
| "learning_rate": 0.00019902975787119956, | |
| "loss": 0.0873, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.9540229885057472, | |
| "grad_norm": 0.5461205244064331, | |
| "learning_rate": 0.00019874632606221545, | |
| "loss": 0.0739, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 2.0689655172413794, | |
| "grad_norm": 0.43381860852241516, | |
| "learning_rate": 0.00019842687989441604, | |
| "loss": 0.0682, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 2.1839080459770113, | |
| "grad_norm": 0.619968593120575, | |
| "learning_rate": 0.00019807153587490963, | |
| "loss": 0.0725, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 2.2988505747126435, | |
| "grad_norm": 0.5531813502311707, | |
| "learning_rate": 0.00019768042360332325, | |
| "loss": 0.0649, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 2.413793103448276, | |
| "grad_norm": 0.4325454831123352, | |
| "learning_rate": 0.00019725368572453539, | |
| "loss": 0.0629, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 2.528735632183908, | |
| "grad_norm": 1.132866382598877, | |
| "learning_rate": 0.00019679147787665126, | |
| "loss": 0.0597, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 2.6436781609195403, | |
| "grad_norm": 0.5158783793449402, | |
| "learning_rate": 0.00019629396863423911, | |
| "loss": 0.0658, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 2.7586206896551726, | |
| "grad_norm": 0.5275442600250244, | |
| "learning_rate": 0.0001957613394468484, | |
| "loss": 0.0624, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 2.873563218390805, | |
| "grad_norm": 0.26212960481643677, | |
| "learning_rate": 0.0001951937845728321, | |
| "loss": 0.0565, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.9885057471264367, | |
| "grad_norm": 0.4064450263977051, | |
| "learning_rate": 0.00019459151100849784, | |
| "loss": 0.0586, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 3.103448275862069, | |
| "grad_norm": 0.497156023979187, | |
| "learning_rate": 0.0001939547384126128, | |
| "loss": 0.0569, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 3.218390804597701, | |
| "grad_norm": 0.23238833248615265, | |
| "learning_rate": 0.00019328369902629083, | |
| "loss": 0.048, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 3.3333333333333335, | |
| "grad_norm": 0.33057811856269836, | |
| "learning_rate": 0.00019257863758829035, | |
| "loss": 0.0508, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 3.4482758620689653, | |
| "grad_norm": 0.2923976182937622, | |
| "learning_rate": 0.00019183981124575418, | |
| "loss": 0.059, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 3.5632183908045976, | |
| "grad_norm": 0.40444961190223694, | |
| "learning_rate": 0.00019106748946042407, | |
| "loss": 0.0589, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 3.67816091954023, | |
| "grad_norm": 0.3618530333042145, | |
| "learning_rate": 0.00019026195391036338, | |
| "loss": 0.0493, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 3.793103448275862, | |
| "grad_norm": 0.2655580937862396, | |
| "learning_rate": 0.00018942349838722486, | |
| "loss": 0.0502, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 3.9080459770114944, | |
| "grad_norm": 0.30788642168045044, | |
| "learning_rate": 0.0001885524286891002, | |
| "loss": 0.0562, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 4.022988505747127, | |
| "grad_norm": 0.3801023066043854, | |
| "learning_rate": 0.00018764906250899027, | |
| "loss": 0.0536, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 4.137931034482759, | |
| "grad_norm": 0.34299996495246887, | |
| "learning_rate": 0.00018671372931893773, | |
| "loss": 0.0583, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 4.252873563218391, | |
| "grad_norm": 0.5825142860412598, | |
| "learning_rate": 0.0001857467702498633, | |
| "loss": 0.0465, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 4.3678160919540225, | |
| "grad_norm": 0.46258264780044556, | |
| "learning_rate": 0.0001847485379671496, | |
| "loss": 0.0469, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 4.482758620689655, | |
| "grad_norm": 0.23550616204738617, | |
| "learning_rate": 0.0001837193965420188, | |
| "loss": 0.0477, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 4.597701149425287, | |
| "grad_norm": 0.609255850315094, | |
| "learning_rate": 0.00018265972131874987, | |
| "loss": 0.0467, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 4.712643678160919, | |
| "grad_norm": 0.3701482117176056, | |
| "learning_rate": 0.00018156989877778461, | |
| "loss": 0.0471, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 4.827586206896552, | |
| "grad_norm": 0.4651474356651306, | |
| "learning_rate": 0.00018045032639477194, | |
| "loss": 0.0434, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 4.942528735632184, | |
| "grad_norm": 0.34028705954551697, | |
| "learning_rate": 0.00017930141249560233, | |
| "loss": 0.0452, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 5.057471264367816, | |
| "grad_norm": 0.2748933732509613, | |
| "learning_rate": 0.00017812357610748488, | |
| "loss": 0.0413, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 5.172413793103448, | |
| "grad_norm": 0.4612014591693878, | |
| "learning_rate": 0.00017691724680612118, | |
| "loss": 0.0423, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 5.287356321839081, | |
| "grad_norm": 0.37961891293525696, | |
| "learning_rate": 0.00017568286455903258, | |
| "loss": 0.0422, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 5.402298850574713, | |
| "grad_norm": 0.3245999813079834, | |
| "learning_rate": 0.00017442087956509665, | |
| "loss": 0.0396, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 5.517241379310345, | |
| "grad_norm": 0.5230941772460938, | |
| "learning_rate": 0.00017313175209035268, | |
| "loss": 0.0405, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 5.6321839080459775, | |
| "grad_norm": 0.2870311737060547, | |
| "learning_rate": 0.00017181595230013525, | |
| "loss": 0.0343, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 5.747126436781609, | |
| "grad_norm": 0.2876773774623871, | |
| "learning_rate": 0.00017047396008759754, | |
| "loss": 0.0436, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 5.862068965517241, | |
| "grad_norm": 0.4095667898654938, | |
| "learning_rate": 0.00016910626489868649, | |
| "loss": 0.0408, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 5.977011494252873, | |
| "grad_norm": 0.377605140209198, | |
| "learning_rate": 0.00016771336555363418, | |
| "loss": 0.0415, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 6.091954022988506, | |
| "grad_norm": 0.28248798847198486, | |
| "learning_rate": 0.00016629577006503009, | |
| "loss": 0.0386, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 6.206896551724138, | |
| "grad_norm": 0.36199840903282166, | |
| "learning_rate": 0.0001648539954525409, | |
| "loss": 0.0405, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 6.32183908045977, | |
| "grad_norm": 0.2778664529323578, | |
| "learning_rate": 0.00016338856755434503, | |
| "loss": 0.0359, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 6.436781609195402, | |
| "grad_norm": 0.23418012261390686, | |
| "learning_rate": 0.00016190002083535122, | |
| "loss": 0.0382, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 6.551724137931035, | |
| "grad_norm": 0.3027312457561493, | |
| "learning_rate": 0.00016038889819227045, | |
| "loss": 0.0394, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 6.666666666666667, | |
| "grad_norm": 0.2858007550239563, | |
| "learning_rate": 0.00015885575075561326, | |
| "loss": 0.042, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 6.781609195402299, | |
| "grad_norm": 0.2762337923049927, | |
| "learning_rate": 0.00015730113768868312, | |
| "loss": 0.039, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 6.896551724137931, | |
| "grad_norm": 0.40732237696647644, | |
| "learning_rate": 0.0001557256259836412, | |
| "loss": 0.0404, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 7.011494252873563, | |
| "grad_norm": 0.36847805976867676, | |
| "learning_rate": 0.00015412979025471488, | |
| "loss": 0.0368, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 7.126436781609195, | |
| "grad_norm": 0.2492237538099289, | |
| "learning_rate": 0.00015251421252862707, | |
| "loss": 0.0336, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 7.241379310344827, | |
| "grad_norm": 0.2626156210899353, | |
| "learning_rate": 0.00015087948203232156, | |
| "loss": 0.0352, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 7.35632183908046, | |
| "grad_norm": 0.6365396976470947, | |
| "learning_rate": 0.00014922619497806277, | |
| "loss": 0.0342, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 7.471264367816092, | |
| "grad_norm": 0.3000635802745819, | |
| "learning_rate": 0.00014755495434598745, | |
| "loss": 0.037, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 7.586206896551724, | |
| "grad_norm": 0.21869853138923645, | |
| "learning_rate": 0.0001458663696641884, | |
| "loss": 0.0365, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 7.7011494252873565, | |
| "grad_norm": 0.22284150123596191, | |
| "learning_rate": 0.0001441610567864096, | |
| "loss": 0.035, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 7.816091954022989, | |
| "grad_norm": 0.267621248960495, | |
| "learning_rate": 0.00014243963766743495, | |
| "loss": 0.029, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 7.931034482758621, | |
| "grad_norm": 0.2817297875881195, | |
| "learning_rate": 0.00014070274013625096, | |
| "loss": 0.0303, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 8.045977011494253, | |
| "grad_norm": 0.3712492287158966, | |
| "learning_rate": 0.00013895099766706784, | |
| "loss": 0.0297, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 8.160919540229886, | |
| "grad_norm": 0.4549995958805084, | |
| "learning_rate": 0.00013718504914828135, | |
| "loss": 0.033, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 8.275862068965518, | |
| "grad_norm": 0.28695234656333923, | |
| "learning_rate": 0.00013540553864945976, | |
| "loss": 0.0306, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 8.39080459770115, | |
| "grad_norm": 0.34577062726020813, | |
| "learning_rate": 0.00013361311518644172, | |
| "loss": 0.0325, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 8.505747126436782, | |
| "grad_norm": 0.3214464783668518, | |
| "learning_rate": 0.00013180843248462973, | |
| "loss": 0.0337, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 8.620689655172415, | |
| "grad_norm": 0.33310961723327637, | |
| "learning_rate": 0.00012999214874056595, | |
| "loss": 0.0344, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 8.735632183908045, | |
| "grad_norm": 0.25606226921081543, | |
| "learning_rate": 0.00012816492638187762, | |
| "loss": 0.0396, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 8.850574712643677, | |
| "grad_norm": 0.36330148577690125, | |
| "learning_rate": 0.00012632743182567905, | |
| "loss": 0.0348, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 8.96551724137931, | |
| "grad_norm": 0.314394474029541, | |
| "learning_rate": 0.00012448033523551865, | |
| "loss": 0.0399, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 9.080459770114942, | |
| "grad_norm": 0.28129351139068604, | |
| "learning_rate": 0.00012262431027695964, | |
| "loss": 0.0298, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 9.195402298850574, | |
| "grad_norm": 0.256881982088089, | |
| "learning_rate": 0.00012076003387188353, | |
| "loss": 0.0292, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 9.310344827586206, | |
| "grad_norm": 0.1919921189546585, | |
| "learning_rate": 0.00011888818595160584, | |
| "loss": 0.0269, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 9.425287356321839, | |
| "grad_norm": 0.2719796895980835, | |
| "learning_rate": 0.00011700944920889436, | |
| "loss": 0.0265, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 9.540229885057471, | |
| "grad_norm": 0.2269754707813263, | |
| "learning_rate": 0.00011512450884898022, | |
| "loss": 0.0316, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 9.655172413793103, | |
| "grad_norm": 0.23504453897476196, | |
| "learning_rate": 0.00011323405233965256, | |
| "loss": 0.0273, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 9.770114942528735, | |
| "grad_norm": 0.22570957243442535, | |
| "learning_rate": 0.00011133876916052821, | |
| "loss": 0.0304, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 9.885057471264368, | |
| "grad_norm": 0.19824576377868652, | |
| "learning_rate": 0.00010943935055158734, | |
| "loss": 0.0283, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 10.0, | |
| "grad_norm": 0.41852012276649475, | |
| "learning_rate": 0.00010753648926106723, | |
| "loss": 0.0319, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 10.114942528735632, | |
| "grad_norm": 0.20548714697360992, | |
| "learning_rate": 0.00010563087929280613, | |
| "loss": 0.0285, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 10.229885057471265, | |
| "grad_norm": 0.22767336666584015, | |
| "learning_rate": 0.00010372321565312872, | |
| "loss": 0.031, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 10.344827586206897, | |
| "grad_norm": 0.20542040467262268, | |
| "learning_rate": 0.00010181419409736647, | |
| "loss": 0.0316, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 10.459770114942529, | |
| "grad_norm": 0.3105849027633667, | |
| "learning_rate": 9.990451087610448e-05, | |
| "loss": 0.027, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 10.574712643678161, | |
| "grad_norm": 0.31816890835762024, | |
| "learning_rate": 9.799486248124775e-05, | |
| "loss": 0.025, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 10.689655172413794, | |
| "grad_norm": 0.3295416235923767, | |
| "learning_rate": 9.608594539199957e-05, | |
| "loss": 0.0247, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 10.804597701149426, | |
| "grad_norm": 0.17071272432804108, | |
| "learning_rate": 9.417845582084448e-05, | |
| "loss": 0.0291, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 10.919540229885058, | |
| "grad_norm": 0.189552441239357, | |
| "learning_rate": 9.227308945962827e-05, | |
| "loss": 0.0243, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 11.03448275862069, | |
| "grad_norm": 0.3179641664028168, | |
| "learning_rate": 9.037054122582839e-05, | |
| "loss": 0.0308, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 11.149425287356323, | |
| "grad_norm": 0.3051457703113556, | |
| "learning_rate": 8.847150500910618e-05, | |
| "loss": 0.0275, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 11.264367816091955, | |
| "grad_norm": 0.29757606983184814, | |
| "learning_rate": 8.657667341823448e-05, | |
| "loss": 0.0264, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 11.379310344827585, | |
| "grad_norm": 0.2845855951309204, | |
| "learning_rate": 8.4686737528492e-05, | |
| "loss": 0.0249, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 11.494252873563218, | |
| "grad_norm": 0.239737406373024, | |
| "learning_rate": 8.280238662961728e-05, | |
| "loss": 0.027, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 11.60919540229885, | |
| "grad_norm": 0.2692360281944275, | |
| "learning_rate": 8.092430797441364e-05, | |
| "loss": 0.0216, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 11.724137931034482, | |
| "grad_norm": 0.18495500087738037, | |
| "learning_rate": 7.905318652809728e-05, | |
| "loss": 0.0255, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 11.839080459770114, | |
| "grad_norm": 0.2230875939130783, | |
| "learning_rate": 7.718970471847923e-05, | |
| "loss": 0.0262, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 11.954022988505747, | |
| "grad_norm": 0.14376775920391083, | |
| "learning_rate": 7.53345421870735e-05, | |
| "loss": 0.0209, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 12.068965517241379, | |
| "grad_norm": 0.20623371005058289, | |
| "learning_rate": 7.348837554122057e-05, | |
| "loss": 0.0192, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 12.183908045977011, | |
| "grad_norm": 0.27209600806236267, | |
| "learning_rate": 7.165187810731823e-05, | |
| "loss": 0.0208, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 12.298850574712644, | |
| "grad_norm": 0.19447851181030273, | |
| "learning_rate": 6.982571968524847e-05, | |
| "loss": 0.0201, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 12.413793103448276, | |
| "grad_norm": 0.18613241612911224, | |
| "learning_rate": 6.801056630409098e-05, | |
| "loss": 0.0248, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 12.528735632183908, | |
| "grad_norm": 0.24156583845615387, | |
| "learning_rate": 6.620707997921157e-05, | |
| "loss": 0.0197, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 12.64367816091954, | |
| "grad_norm": 0.16912145912647247, | |
| "learning_rate": 6.441591847081476e-05, | |
| "loss": 0.022, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 12.758620689655173, | |
| "grad_norm": 0.14165754616260529, | |
| "learning_rate": 6.263773504404801e-05, | |
| "loss": 0.0199, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 12.873563218390805, | |
| "grad_norm": 0.3424724340438843, | |
| "learning_rate": 6.087317823074565e-05, | |
| "loss": 0.0209, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 12.988505747126437, | |
| "grad_norm": 0.2658204138278961, | |
| "learning_rate": 5.912289159289883e-05, | |
| "loss": 0.0242, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 13.10344827586207, | |
| "grad_norm": 0.21321730315685272, | |
| "learning_rate": 5.7387513487938386e-05, | |
| "loss": 0.0216, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 13.218390804597702, | |
| "grad_norm": 0.2854823172092438, | |
| "learning_rate": 5.566767683591553e-05, | |
| "loss": 0.0227, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 13.333333333333334, | |
| "grad_norm": 0.28919658064842224, | |
| "learning_rate": 5.396400888866601e-05, | |
| "loss": 0.0195, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 13.448275862068966, | |
| "grad_norm": 0.22510255873203278, | |
| "learning_rate": 5.2277131001041125e-05, | |
| "loss": 0.0241, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 13.563218390804598, | |
| "grad_norm": 0.21545900404453278, | |
| "learning_rate": 5.060765840429e-05, | |
| "loss": 0.023, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 13.678160919540229, | |
| "grad_norm": 0.20618782937526703, | |
| "learning_rate": 4.8956199981674656e-05, | |
| "loss": 0.0181, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 13.793103448275861, | |
| "grad_norm": 0.22331970930099487, | |
| "learning_rate": 4.7323358046400844e-05, | |
| "loss": 0.0212, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 13.908045977011493, | |
| "grad_norm": 0.14791706204414368, | |
| "learning_rate": 4.570972812194457e-05, | |
| "loss": 0.0195, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 14.022988505747126, | |
| "grad_norm": 0.1526448130607605, | |
| "learning_rate": 4.4115898724855455e-05, | |
| "loss": 0.0188, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 14.137931034482758, | |
| "grad_norm": 0.18956783413887024, | |
| "learning_rate": 4.254245115011506e-05, | |
| "loss": 0.0188, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 14.25287356321839, | |
| "grad_norm": 0.1313301920890808, | |
| "learning_rate": 4.098995925912972e-05, | |
| "loss": 0.019, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 14.367816091954023, | |
| "grad_norm": 0.13764789700508118, | |
| "learning_rate": 3.945898927043372e-05, | |
| "loss": 0.0175, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 14.482758620689655, | |
| "grad_norm": 0.19556942582130432, | |
| "learning_rate": 3.7950099553180804e-05, | |
| "loss": 0.0196, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 14.597701149425287, | |
| "grad_norm": 0.14027345180511475, | |
| "learning_rate": 3.646384042349764e-05, | |
| "loss": 0.0177, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 14.71264367816092, | |
| "grad_norm": 0.2918284833431244, | |
| "learning_rate": 3.500075394377511e-05, | |
| "loss": 0.0204, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 14.827586206896552, | |
| "grad_norm": 0.12948164343833923, | |
| "learning_rate": 3.3561373724969224e-05, | |
| "loss": 0.0188, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 14.942528735632184, | |
| "grad_norm": 0.1773224174976349, | |
| "learning_rate": 3.214622473198492e-05, | |
| "loss": 0.0212, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 15.057471264367816, | |
| "grad_norm": 0.29863160848617554, | |
| "learning_rate": 3.075582309221289e-05, | |
| "loss": 0.0157, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 15.172413793103448, | |
| "grad_norm": 0.18764474987983704, | |
| "learning_rate": 2.939067590728972e-05, | |
| "loss": 0.0175, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 15.28735632183908, | |
| "grad_norm": 0.16273990273475647, | |
| "learning_rate": 2.8051281068149803e-05, | |
| "loss": 0.0135, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 15.402298850574713, | |
| "grad_norm": 0.25088945031166077, | |
| "learning_rate": 2.673812707343669e-05, | |
| "loss": 0.0242, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 15.517241379310345, | |
| "grad_norm": 0.25521960854530334, | |
| "learning_rate": 2.545169285133965e-05, | |
| "loss": 0.0188, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 15.632183908045977, | |
| "grad_norm": 0.15780223906040192, | |
| "learning_rate": 2.4192447584921195e-05, | |
| "loss": 0.0194, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 15.74712643678161, | |
| "grad_norm": 0.13658417761325836, | |
| "learning_rate": 2.296085054099828e-05, | |
| "loss": 0.0179, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 15.862068965517242, | |
| "grad_norm": 0.14593394100666046, | |
| "learning_rate": 2.175735090264058e-05, | |
| "loss": 0.016, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 15.977011494252874, | |
| "grad_norm": 0.20093883574008942, | |
| "learning_rate": 2.0582387605346088e-05, | |
| "loss": 0.0157, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 16.091954022988507, | |
| "grad_norm": 0.22261527180671692, | |
| "learning_rate": 1.943638917695453e-05, | |
| "loss": 0.0175, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 16.20689655172414, | |
| "grad_norm": 0.17486433684825897, | |
| "learning_rate": 1.831977358135625e-05, | |
| "loss": 0.0166, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 16.32183908045977, | |
| "grad_norm": 0.2138216346502304, | |
| "learning_rate": 1.723294806605428e-05, | |
| "loss": 0.0146, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 16.436781609195403, | |
| "grad_norm": 0.20112960040569305, | |
| "learning_rate": 1.6176309013634517e-05, | |
| "loss": 0.0159, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 16.551724137931036, | |
| "grad_norm": 0.15377485752105713, | |
| "learning_rate": 1.5150241797198883e-05, | |
| "loss": 0.016, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 16.666666666666668, | |
| "grad_norm": 0.23132722079753876, | |
| "learning_rate": 1.415512063981339e-05, | |
| "loss": 0.0134, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 16.7816091954023, | |
| "grad_norm": 0.15262600779533386, | |
| "learning_rate": 1.3191308478023212e-05, | |
| "loss": 0.017, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 16.896551724137932, | |
| "grad_norm": 0.0991855040192604, | |
| "learning_rate": 1.2259156829483842e-05, | |
| "loss": 0.0167, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 17.011494252873565, | |
| "grad_norm": 0.12278055399656296, | |
| "learning_rate": 1.1359005664756994e-05, | |
| "loss": 0.0146, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 17.126436781609197, | |
| "grad_norm": 0.17124158143997192, | |
| "learning_rate": 1.0491183283317997e-05, | |
| "loss": 0.0173, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 17.24137931034483, | |
| "grad_norm": 0.1300356686115265, | |
| "learning_rate": 9.656006193819633e-06, | |
| "loss": 0.0143, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 17.35632183908046, | |
| "grad_norm": 0.17946338653564453, | |
| "learning_rate": 8.853778998656537e-06, | |
| "loss": 0.0154, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 17.47126436781609, | |
| "grad_norm": 0.28736400604248047, | |
| "learning_rate": 8.084794282871689e-06, | |
| "loss": 0.0166, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 17.586206896551722, | |
| "grad_norm": 0.13112574815750122, | |
| "learning_rate": 7.3493325074461165e-06, | |
| "loss": 0.0132, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 17.701149425287355, | |
| "grad_norm": 0.12864838540554047, | |
| "learning_rate": 6.647661907010183e-06, | |
| "loss": 0.0171, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 17.816091954022987, | |
| "grad_norm": 0.16958807408809662, | |
| "learning_rate": 5.980038392014309e-06, | |
| "loss": 0.0161, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 17.93103448275862, | |
| "grad_norm": 0.36121729016304016, | |
| "learning_rate": 5.3467054553941405e-06, | |
| "loss": 0.0158, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 18.04597701149425, | |
| "grad_norm": 0.2107989490032196, | |
| "learning_rate": 4.7478940837649924e-06, | |
| "loss": 0.0147, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 18.160919540229884, | |
| "grad_norm": 0.15654149651527405, | |
| "learning_rate": 4.183822673177229e-06, | |
| "loss": 0.0164, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 18.275862068965516, | |
| "grad_norm": 0.1438828557729721, | |
| "learning_rate": 3.6546969494637986e-06, | |
| "loss": 0.0131, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 18.39080459770115, | |
| "grad_norm": 0.2543192207813263, | |
| "learning_rate": 3.16070989320868e-06, | |
| "loss": 0.0157, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 18.50574712643678, | |
| "grad_norm": 0.13453112542629242, | |
| "learning_rate": 2.702041669363875e-06, | |
| "loss": 0.017, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 18.620689655172413, | |
| "grad_norm": 0.16369780898094177, | |
| "learning_rate": 2.2788595615403474e-06, | |
| "loss": 0.0157, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 18.735632183908045, | |
| "grad_norm": 0.14639818668365479, | |
| "learning_rate": 1.8913179109969482e-06, | |
| "loss": 0.0122, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 18.850574712643677, | |
| "grad_norm": 0.23813354969024658, | |
| "learning_rate": 1.5395580603498328e-06, | |
| "loss": 0.0157, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 18.96551724137931, | |
| "grad_norm": 0.15577834844589233, | |
| "learning_rate": 1.2237083020224526e-06, | |
| "loss": 0.0144, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 19.080459770114942, | |
| "grad_norm": 0.06880059838294983, | |
| "learning_rate": 9.438838314553056e-07, | |
| "loss": 0.0109, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 19.195402298850574, | |
| "grad_norm": 0.19819270074367523, | |
| "learning_rate": 7.001867050923095e-07, | |
| "loss": 0.0134, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 19.310344827586206, | |
| "grad_norm": 0.10673543065786362, | |
| "learning_rate": 4.92705803159188e-07, | |
| "loss": 0.0155, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 19.42528735632184, | |
| "grad_norm": 0.16529639065265656, | |
| "learning_rate": 3.2151679724748975e-07, | |
| "loss": 0.0175, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 19.54022988505747, | |
| "grad_norm": 0.1206677109003067, | |
| "learning_rate": 1.8668212271585327e-07, | |
| "loss": 0.0188, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 19.655172413793103, | |
| "grad_norm": 0.10180158913135529, | |
| "learning_rate": 8.825095591891152e-08, | |
| "loss": 0.0158, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 19.770114942528735, | |
| "grad_norm": 0.14544282853603363, | |
| "learning_rate": 2.625919627188278e-08, | |
| "loss": 0.015, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 19.885057471264368, | |
| "grad_norm": 0.14054107666015625, | |
| "learning_rate": 7.294531574553176e-10, | |
| "loss": 0.0139, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 19.908045977011493, | |
| "step": 1732, | |
| "total_flos": 2.431760592612004e+17, | |
| "train_loss": 0.04785273781403314, | |
| "train_runtime": 1980.9382, | |
| "train_samples_per_second": 55.957, | |
| "train_steps_per_second": 0.874 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 1732, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 20, | |
| "save_steps": 10000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.431760592612004e+17, | |
| "train_batch_size": 64, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |