| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 3930, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.002544529262086514, | |
| "grad_norm": 2.2193164825439453, | |
| "learning_rate": 1.1450381679389313e-06, | |
| "loss": 0.1149, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.005089058524173028, | |
| "grad_norm": 2.2632293701171875, | |
| "learning_rate": 2.4173027989821884e-06, | |
| "loss": 0.1642, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.007633587786259542, | |
| "grad_norm": 3.3879759311676025, | |
| "learning_rate": 3.6895674300254453e-06, | |
| "loss": 0.1337, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.010178117048346057, | |
| "grad_norm": 3.659600257873535, | |
| "learning_rate": 4.961832061068703e-06, | |
| "loss": 0.1454, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.01272264631043257, | |
| "grad_norm": 2.5636653900146484, | |
| "learning_rate": 6.2340966921119596e-06, | |
| "loss": 0.0598, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.015267175572519083, | |
| "grad_norm": 3.9301695823669434, | |
| "learning_rate": 7.506361323155217e-06, | |
| "loss": 0.0593, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.017811704834605598, | |
| "grad_norm": 0.9530353546142578, | |
| "learning_rate": 8.778625954198473e-06, | |
| "loss": 0.0368, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.020356234096692113, | |
| "grad_norm": 2.2630889415740967, | |
| "learning_rate": 1.0050890585241731e-05, | |
| "loss": 0.0398, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.022900763358778626, | |
| "grad_norm": 1.9916532039642334, | |
| "learning_rate": 1.1323155216284988e-05, | |
| "loss": 0.0319, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.02544529262086514, | |
| "grad_norm": 2.0590147972106934, | |
| "learning_rate": 1.2595419847328243e-05, | |
| "loss": 0.0432, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.027989821882951654, | |
| "grad_norm": 2.5208053588867188, | |
| "learning_rate": 1.38676844783715e-05, | |
| "loss": 0.0291, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.030534351145038167, | |
| "grad_norm": 1.356877088546753, | |
| "learning_rate": 1.5139949109414759e-05, | |
| "loss": 0.0202, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.03307888040712468, | |
| "grad_norm": 0.7600014805793762, | |
| "learning_rate": 1.6412213740458016e-05, | |
| "loss": 0.0259, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.035623409669211195, | |
| "grad_norm": 0.018205782398581505, | |
| "learning_rate": 1.7684478371501274e-05, | |
| "loss": 0.0163, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.03816793893129771, | |
| "grad_norm": 0.918929398059845, | |
| "learning_rate": 1.895674300254453e-05, | |
| "loss": 0.0244, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.04071246819338423, | |
| "grad_norm": 1.1537033319473267, | |
| "learning_rate": 2.0229007633587788e-05, | |
| "loss": 0.0184, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.043256997455470736, | |
| "grad_norm": 1.2687195539474487, | |
| "learning_rate": 2.1501272264631043e-05, | |
| "loss": 0.012, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.04580152671755725, | |
| "grad_norm": 0.24528977274894714, | |
| "learning_rate": 2.2773536895674302e-05, | |
| "loss": 0.0197, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.04834605597964377, | |
| "grad_norm": 1.1114501953125, | |
| "learning_rate": 2.404580152671756e-05, | |
| "loss": 0.0119, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.05089058524173028, | |
| "grad_norm": 0.843619704246521, | |
| "learning_rate": 2.5318066157760816e-05, | |
| "loss": 0.017, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.05343511450381679, | |
| "grad_norm": 0.6730576753616333, | |
| "learning_rate": 2.6590330788804074e-05, | |
| "loss": 0.021, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.05597964376590331, | |
| "grad_norm": 0.7756699323654175, | |
| "learning_rate": 2.7862595419847333e-05, | |
| "loss": 0.0164, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.058524173027989825, | |
| "grad_norm": 0.18474726378917694, | |
| "learning_rate": 2.9134860050890588e-05, | |
| "loss": 0.0198, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.061068702290076333, | |
| "grad_norm": 0.012082410044968128, | |
| "learning_rate": 3.0407124681933847e-05, | |
| "loss": 0.0189, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.06361323155216285, | |
| "grad_norm": 0.9460607767105103, | |
| "learning_rate": 3.16793893129771e-05, | |
| "loss": 0.0069, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.06615776081424936, | |
| "grad_norm": 0.15384167432785034, | |
| "learning_rate": 3.295165394402036e-05, | |
| "loss": 0.0192, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.06870229007633588, | |
| "grad_norm": 0.16621553897857666, | |
| "learning_rate": 3.4223918575063616e-05, | |
| "loss": 0.0262, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.07124681933842239, | |
| "grad_norm": 0.08461080491542816, | |
| "learning_rate": 3.549618320610687e-05, | |
| "loss": 0.0127, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.0737913486005089, | |
| "grad_norm": 0.6972628235816956, | |
| "learning_rate": 3.6768447837150126e-05, | |
| "loss": 0.0123, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.07633587786259542, | |
| "grad_norm": 0.3922271132469177, | |
| "learning_rate": 3.8040712468193385e-05, | |
| "loss": 0.009, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.07888040712468193, | |
| "grad_norm": 0.9791713356971741, | |
| "learning_rate": 3.9312977099236644e-05, | |
| "loss": 0.0225, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.08142493638676845, | |
| "grad_norm": 0.32885971665382385, | |
| "learning_rate": 4.05852417302799e-05, | |
| "loss": 0.0143, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.08396946564885496, | |
| "grad_norm": 0.09194409102201462, | |
| "learning_rate": 4.1857506361323154e-05, | |
| "loss": 0.0112, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.08651399491094147, | |
| "grad_norm": 0.03973765671253204, | |
| "learning_rate": 4.312977099236641e-05, | |
| "loss": 0.0127, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.089058524173028, | |
| "grad_norm": 1.0717049837112427, | |
| "learning_rate": 4.440203562340967e-05, | |
| "loss": 0.0101, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.0916030534351145, | |
| "grad_norm": 0.5723052620887756, | |
| "learning_rate": 4.567430025445293e-05, | |
| "loss": 0.0179, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.09414758269720101, | |
| "grad_norm": 0.7891315817832947, | |
| "learning_rate": 4.694656488549619e-05, | |
| "loss": 0.0061, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.09669211195928754, | |
| "grad_norm": 0.7999855875968933, | |
| "learning_rate": 4.821882951653944e-05, | |
| "loss": 0.023, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.09923664122137404, | |
| "grad_norm": 0.07333844900131226, | |
| "learning_rate": 4.94910941475827e-05, | |
| "loss": 0.01, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.10178117048346055, | |
| "grad_norm": 1.1714524030685425, | |
| "learning_rate": 4.991518235793045e-05, | |
| "loss": 0.0191, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.10432569974554708, | |
| "grad_norm": 1.1736589670181274, | |
| "learning_rate": 4.977381962114787e-05, | |
| "loss": 0.016, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.10687022900763359, | |
| "grad_norm": 0.5235586166381836, | |
| "learning_rate": 4.963245688436528e-05, | |
| "loss": 0.0197, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.10941475826972011, | |
| "grad_norm": 0.3463246822357178, | |
| "learning_rate": 4.94910941475827e-05, | |
| "loss": 0.0147, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.11195928753180662, | |
| "grad_norm": 0.03407707437872887, | |
| "learning_rate": 4.9349731410800115e-05, | |
| "loss": 0.004, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.11450381679389313, | |
| "grad_norm": 0.8756362795829773, | |
| "learning_rate": 4.920836867401753e-05, | |
| "loss": 0.008, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.11704834605597965, | |
| "grad_norm": 0.12369545549154282, | |
| "learning_rate": 4.906700593723495e-05, | |
| "loss": 0.0055, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.11959287531806616, | |
| "grad_norm": 0.21906960010528564, | |
| "learning_rate": 4.892564320045236e-05, | |
| "loss": 0.0064, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.12213740458015267, | |
| "grad_norm": 0.02911725640296936, | |
| "learning_rate": 4.878428046366978e-05, | |
| "loss": 0.0103, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.12468193384223919, | |
| "grad_norm": 0.037161994725465775, | |
| "learning_rate": 4.864291772688719e-05, | |
| "loss": 0.011, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.1272264631043257, | |
| "grad_norm": 0.12944084405899048, | |
| "learning_rate": 4.850155499010461e-05, | |
| "loss": 0.0187, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1272264631043257, | |
| "eval_loss": 0.0358583927154541, | |
| "eval_runtime": 133.9654, | |
| "eval_samples_per_second": 59.583, | |
| "eval_steps_per_second": 0.47, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.1297709923664122, | |
| "grad_norm": 1.3357224464416504, | |
| "learning_rate": 4.836019225332203e-05, | |
| "loss": 0.0209, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.13231552162849872, | |
| "grad_norm": 0.08715047687292099, | |
| "learning_rate": 4.821882951653944e-05, | |
| "loss": 0.0095, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.13486005089058525, | |
| "grad_norm": 0.2528047561645508, | |
| "learning_rate": 4.8077466779756856e-05, | |
| "loss": 0.0055, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.13740458015267176, | |
| "grad_norm": 0.5453704595565796, | |
| "learning_rate": 4.793610404297427e-05, | |
| "loss": 0.0097, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.13994910941475827, | |
| "grad_norm": 0.9273648858070374, | |
| "learning_rate": 4.7794741306191694e-05, | |
| "loss": 0.0091, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.14249363867684478, | |
| "grad_norm": 0.7525148987770081, | |
| "learning_rate": 4.765337856940911e-05, | |
| "loss": 0.0109, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.1450381679389313, | |
| "grad_norm": 0.23284463584423065, | |
| "learning_rate": 4.751201583262652e-05, | |
| "loss": 0.0084, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.1475826972010178, | |
| "grad_norm": 0.2786177396774292, | |
| "learning_rate": 4.7370653095843935e-05, | |
| "loss": 0.0049, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.15012722646310434, | |
| "grad_norm": 0.004754980094730854, | |
| "learning_rate": 4.722929035906135e-05, | |
| "loss": 0.0114, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.15267175572519084, | |
| "grad_norm": 0.16983787715435028, | |
| "learning_rate": 4.708792762227877e-05, | |
| "loss": 0.0026, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.15521628498727735, | |
| "grad_norm": 0.35884377360343933, | |
| "learning_rate": 4.694656488549619e-05, | |
| "loss": 0.0117, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.15776081424936386, | |
| "grad_norm": 0.011907716281712055, | |
| "learning_rate": 4.68052021487136e-05, | |
| "loss": 0.0148, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.16030534351145037, | |
| "grad_norm": 0.07890170812606812, | |
| "learning_rate": 4.666383941193101e-05, | |
| "loss": 0.0117, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.1628498727735369, | |
| "grad_norm": 0.37597450613975525, | |
| "learning_rate": 4.6522476675148436e-05, | |
| "loss": 0.0098, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.16539440203562342, | |
| "grad_norm": 0.09366422146558762, | |
| "learning_rate": 4.638111393836585e-05, | |
| "loss": 0.0035, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.16793893129770993, | |
| "grad_norm": 0.5103874206542969, | |
| "learning_rate": 4.623975120158327e-05, | |
| "loss": 0.0088, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.17048346055979643, | |
| "grad_norm": 0.18060427904129028, | |
| "learning_rate": 4.6098388464800676e-05, | |
| "loss": 0.0085, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.17302798982188294, | |
| "grad_norm": 0.01927395723760128, | |
| "learning_rate": 4.59570257280181e-05, | |
| "loss": 0.0156, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.17557251908396945, | |
| "grad_norm": 0.26210394501686096, | |
| "learning_rate": 4.5815662991235514e-05, | |
| "loss": 0.0227, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.178117048346056, | |
| "grad_norm": 0.5203946232795715, | |
| "learning_rate": 4.567430025445293e-05, | |
| "loss": 0.01, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.1806615776081425, | |
| "grad_norm": 0.4313431680202484, | |
| "learning_rate": 4.5532937517670346e-05, | |
| "loss": 0.0057, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.183206106870229, | |
| "grad_norm": 0.049217287451028824, | |
| "learning_rate": 4.5391574780887755e-05, | |
| "loss": 0.0099, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.18575063613231552, | |
| "grad_norm": 0.6959419250488281, | |
| "learning_rate": 4.525021204410518e-05, | |
| "loss": 0.0087, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.18829516539440203, | |
| "grad_norm": 0.35286396741867065, | |
| "learning_rate": 4.510884930732259e-05, | |
| "loss": 0.0082, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.19083969465648856, | |
| "grad_norm": 0.028058692812919617, | |
| "learning_rate": 4.496748657054001e-05, | |
| "loss": 0.0096, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.19338422391857507, | |
| "grad_norm": 1.0182298421859741, | |
| "learning_rate": 4.4826123833757424e-05, | |
| "loss": 0.0072, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.19592875318066158, | |
| "grad_norm": 0.929204523563385, | |
| "learning_rate": 4.468476109697484e-05, | |
| "loss": 0.0059, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.1984732824427481, | |
| "grad_norm": 0.18072165548801422, | |
| "learning_rate": 4.4543398360192256e-05, | |
| "loss": 0.0101, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.2010178117048346, | |
| "grad_norm": 1.1691681146621704, | |
| "learning_rate": 4.440203562340967e-05, | |
| "loss": 0.007, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.2035623409669211, | |
| "grad_norm": 0.6434471607208252, | |
| "learning_rate": 4.426067288662709e-05, | |
| "loss": 0.0165, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.20610687022900764, | |
| "grad_norm": 0.1366460621356964, | |
| "learning_rate": 4.41193101498445e-05, | |
| "loss": 0.0124, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.20865139949109415, | |
| "grad_norm": 0.10479103028774261, | |
| "learning_rate": 4.397794741306192e-05, | |
| "loss": 0.0033, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.21119592875318066, | |
| "grad_norm": 0.6483312249183655, | |
| "learning_rate": 4.3836584676279334e-05, | |
| "loss": 0.0072, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.21374045801526717, | |
| "grad_norm": 0.9132885336875916, | |
| "learning_rate": 4.369522193949675e-05, | |
| "loss": 0.0157, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.21628498727735368, | |
| "grad_norm": 0.007977353408932686, | |
| "learning_rate": 4.3553859202714166e-05, | |
| "loss": 0.0042, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.21882951653944022, | |
| "grad_norm": 0.5941423773765564, | |
| "learning_rate": 4.341249646593159e-05, | |
| "loss": 0.0182, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.22137404580152673, | |
| "grad_norm": 0.03638213500380516, | |
| "learning_rate": 4.3271133729149e-05, | |
| "loss": 0.0035, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.22391857506361323, | |
| "grad_norm": 0.018258914351463318, | |
| "learning_rate": 4.312977099236641e-05, | |
| "loss": 0.0096, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.22646310432569974, | |
| "grad_norm": 0.7593536376953125, | |
| "learning_rate": 4.298840825558383e-05, | |
| "loss": 0.0106, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.22900763358778625, | |
| "grad_norm": 0.01460027601569891, | |
| "learning_rate": 4.284704551880125e-05, | |
| "loss": 0.0112, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.23155216284987276, | |
| "grad_norm": 0.05933653190732002, | |
| "learning_rate": 4.2705682782018667e-05, | |
| "loss": 0.0107, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.2340966921119593, | |
| "grad_norm": 0.09628820419311523, | |
| "learning_rate": 4.2564320045236076e-05, | |
| "loss": 0.0098, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.2366412213740458, | |
| "grad_norm": 0.0987539142370224, | |
| "learning_rate": 4.242295730845349e-05, | |
| "loss": 0.007, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.23918575063613232, | |
| "grad_norm": 0.0038352105766534805, | |
| "learning_rate": 4.228159457167091e-05, | |
| "loss": 0.0021, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.24173027989821882, | |
| "grad_norm": 0.022375449538230896, | |
| "learning_rate": 4.214023183488833e-05, | |
| "loss": 0.0096, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.24427480916030533, | |
| "grad_norm": 0.036569077521562576, | |
| "learning_rate": 4.1998869098105745e-05, | |
| "loss": 0.002, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.24681933842239187, | |
| "grad_norm": 0.537909746170044, | |
| "learning_rate": 4.1857506361323154e-05, | |
| "loss": 0.0094, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.24936386768447838, | |
| "grad_norm": 0.046400997787714005, | |
| "learning_rate": 4.171614362454057e-05, | |
| "loss": 0.0096, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.25190839694656486, | |
| "grad_norm": 0.06931813061237335, | |
| "learning_rate": 4.157478088775799e-05, | |
| "loss": 0.0049, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.2544529262086514, | |
| "grad_norm": 0.47014328837394714, | |
| "learning_rate": 4.143341815097541e-05, | |
| "loss": 0.0083, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.2544529262086514, | |
| "eval_loss": 0.03271668031811714, | |
| "eval_runtime": 133.8168, | |
| "eval_samples_per_second": 59.649, | |
| "eval_steps_per_second": 0.471, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 0.25699745547073793, | |
| "grad_norm": 0.10135433822870255, | |
| "learning_rate": 4.1292055414192824e-05, | |
| "loss": 0.0046, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 0.2595419847328244, | |
| "grad_norm": 0.18397603929042816, | |
| "learning_rate": 4.115069267741023e-05, | |
| "loss": 0.0049, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 0.26208651399491095, | |
| "grad_norm": 0.4814232587814331, | |
| "learning_rate": 4.100932994062765e-05, | |
| "loss": 0.006, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 0.26463104325699743, | |
| "grad_norm": 0.36830148100852966, | |
| "learning_rate": 4.086796720384507e-05, | |
| "loss": 0.0145, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 0.26717557251908397, | |
| "grad_norm": 0.3159887194633484, | |
| "learning_rate": 4.0726604467062487e-05, | |
| "loss": 0.0043, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 0.2697201017811705, | |
| "grad_norm": 0.25662675499916077, | |
| "learning_rate": 4.05852417302799e-05, | |
| "loss": 0.0029, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 0.272264631043257, | |
| "grad_norm": 0.8084707856178284, | |
| "learning_rate": 4.044387899349731e-05, | |
| "loss": 0.0077, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 0.2748091603053435, | |
| "grad_norm": 0.35806888341903687, | |
| "learning_rate": 4.0302516256714734e-05, | |
| "loss": 0.0086, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 0.27735368956743, | |
| "grad_norm": 0.8191736340522766, | |
| "learning_rate": 4.016115351993215e-05, | |
| "loss": 0.0077, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 0.27989821882951654, | |
| "grad_norm": 0.08160542696714401, | |
| "learning_rate": 4.0019790783149565e-05, | |
| "loss": 0.0098, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 0.2824427480916031, | |
| "grad_norm": 0.003803673665970564, | |
| "learning_rate": 3.987842804636698e-05, | |
| "loss": 0.0073, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 0.28498727735368956, | |
| "grad_norm": 0.9810532331466675, | |
| "learning_rate": 3.9737065309584396e-05, | |
| "loss": 0.0154, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 0.2875318066157761, | |
| "grad_norm": 0.05929688364267349, | |
| "learning_rate": 3.959570257280181e-05, | |
| "loss": 0.0078, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 0.2900763358778626, | |
| "grad_norm": 0.04006582498550415, | |
| "learning_rate": 3.945433983601923e-05, | |
| "loss": 0.0055, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 0.2926208651399491, | |
| "grad_norm": 0.026924826204776764, | |
| "learning_rate": 3.9312977099236644e-05, | |
| "loss": 0.0109, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 0.2951653944020356, | |
| "grad_norm": 0.2333918809890747, | |
| "learning_rate": 3.917161436245406e-05, | |
| "loss": 0.0059, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 0.29770992366412213, | |
| "grad_norm": 0.5308976769447327, | |
| "learning_rate": 3.9030251625671475e-05, | |
| "loss": 0.0123, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 0.30025445292620867, | |
| "grad_norm": 0.03540479764342308, | |
| "learning_rate": 3.888888888888889e-05, | |
| "loss": 0.0042, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 0.30279898218829515, | |
| "grad_norm": 0.2590758204460144, | |
| "learning_rate": 3.8747526152106306e-05, | |
| "loss": 0.0112, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 0.3053435114503817, | |
| "grad_norm": 0.07383149117231369, | |
| "learning_rate": 3.860616341532372e-05, | |
| "loss": 0.0042, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 0.30788804071246817, | |
| "grad_norm": 0.008312661200761795, | |
| "learning_rate": 3.846480067854114e-05, | |
| "loss": 0.0073, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 0.3104325699745547, | |
| "grad_norm": 0.004382140003144741, | |
| "learning_rate": 3.8323437941758554e-05, | |
| "loss": 0.0099, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 0.31297709923664124, | |
| "grad_norm": 0.08335595577955246, | |
| "learning_rate": 3.818207520497597e-05, | |
| "loss": 0.011, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 0.3155216284987277, | |
| "grad_norm": 0.010813195258378983, | |
| "learning_rate": 3.8040712468193385e-05, | |
| "loss": 0.0045, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 0.31806615776081426, | |
| "grad_norm": 0.005388839170336723, | |
| "learning_rate": 3.78993497314108e-05, | |
| "loss": 0.0081, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 0.32061068702290074, | |
| "grad_norm": 0.011531420052051544, | |
| "learning_rate": 3.7757986994628216e-05, | |
| "loss": 0.0204, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 0.3231552162849873, | |
| "grad_norm": 0.01907213032245636, | |
| "learning_rate": 3.761662425784563e-05, | |
| "loss": 0.0056, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 0.3256997455470738, | |
| "grad_norm": 0.0880994126200676, | |
| "learning_rate": 3.747526152106305e-05, | |
| "loss": 0.0059, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 0.3282442748091603, | |
| "grad_norm": 0.24218730628490448, | |
| "learning_rate": 3.7333898784280464e-05, | |
| "loss": 0.0053, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 0.33078880407124683, | |
| "grad_norm": 0.03127405792474747, | |
| "learning_rate": 3.7192536047497886e-05, | |
| "loss": 0.0058, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 0.3333333333333333, | |
| "grad_norm": 0.19848968088626862, | |
| "learning_rate": 3.7051173310715295e-05, | |
| "loss": 0.0095, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 0.33587786259541985, | |
| "grad_norm": 0.01141483522951603, | |
| "learning_rate": 3.690981057393271e-05, | |
| "loss": 0.0089, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 0.3384223918575064, | |
| "grad_norm": 0.10185040533542633, | |
| "learning_rate": 3.6768447837150126e-05, | |
| "loss": 0.0034, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 0.34096692111959287, | |
| "grad_norm": 0.715850830078125, | |
| "learning_rate": 3.662708510036755e-05, | |
| "loss": 0.0101, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 0.3435114503816794, | |
| "grad_norm": 0.06744267046451569, | |
| "learning_rate": 3.6485722363584965e-05, | |
| "loss": 0.011, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 0.3460559796437659, | |
| "grad_norm": 1.0349208116531372, | |
| "learning_rate": 3.6344359626802374e-05, | |
| "loss": 0.0125, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 0.3486005089058524, | |
| "grad_norm": 0.01572897657752037, | |
| "learning_rate": 3.620299689001979e-05, | |
| "loss": 0.0085, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 0.3511450381679389, | |
| "grad_norm": 0.5127367377281189, | |
| "learning_rate": 3.6061634153237205e-05, | |
| "loss": 0.0089, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 0.35368956743002544, | |
| "grad_norm": 0.6391226053237915, | |
| "learning_rate": 3.592027141645463e-05, | |
| "loss": 0.0107, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 0.356234096692112, | |
| "grad_norm": 0.0599433109164238, | |
| "learning_rate": 3.577890867967204e-05, | |
| "loss": 0.0046, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 0.35877862595419846, | |
| "grad_norm": 0.01737218163907528, | |
| "learning_rate": 3.563754594288945e-05, | |
| "loss": 0.0098, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 0.361323155216285, | |
| "grad_norm": 0.5793020725250244, | |
| "learning_rate": 3.549618320610687e-05, | |
| "loss": 0.0094, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 0.3638676844783715, | |
| "grad_norm": 0.08681387454271317, | |
| "learning_rate": 3.535482046932429e-05, | |
| "loss": 0.0041, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 0.366412213740458, | |
| "grad_norm": 0.06883745640516281, | |
| "learning_rate": 3.5213457732541706e-05, | |
| "loss": 0.0125, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 0.36895674300254455, | |
| "grad_norm": 0.016956673935055733, | |
| "learning_rate": 3.507209499575912e-05, | |
| "loss": 0.0076, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 0.37150127226463103, | |
| "grad_norm": 0.033391110599040985, | |
| "learning_rate": 3.493073225897653e-05, | |
| "loss": 0.0083, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 0.37404580152671757, | |
| "grad_norm": 0.23228399455547333, | |
| "learning_rate": 3.4789369522193946e-05, | |
| "loss": 0.0043, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 0.37659033078880405, | |
| "grad_norm": 0.13741058111190796, | |
| "learning_rate": 3.464800678541137e-05, | |
| "loss": 0.0151, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 0.3791348600508906, | |
| "grad_norm": 0.05785220116376877, | |
| "learning_rate": 3.4506644048628785e-05, | |
| "loss": 0.0051, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 0.3816793893129771, | |
| "grad_norm": 0.026762302964925766, | |
| "learning_rate": 3.43652813118462e-05, | |
| "loss": 0.0106, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.3816793893129771, | |
| "eval_loss": 0.03001326136291027, | |
| "eval_runtime": 133.9379, | |
| "eval_samples_per_second": 59.595, | |
| "eval_steps_per_second": 0.47, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 0.3842239185750636, | |
| "grad_norm": 0.818063497543335, | |
| "learning_rate": 3.4223918575063616e-05, | |
| "loss": 0.0157, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 0.38676844783715014, | |
| "grad_norm": 0.2497348040342331, | |
| "learning_rate": 3.408255583828103e-05, | |
| "loss": 0.0035, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 0.3893129770992366, | |
| "grad_norm": 0.20524458587169647, | |
| "learning_rate": 3.394119310149845e-05, | |
| "loss": 0.0065, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 0.39185750636132316, | |
| "grad_norm": 0.02035215124487877, | |
| "learning_rate": 3.379983036471586e-05, | |
| "loss": 0.0033, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 0.3944020356234097, | |
| "grad_norm": 0.20024162530899048, | |
| "learning_rate": 3.365846762793328e-05, | |
| "loss": 0.0045, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 0.3969465648854962, | |
| "grad_norm": 0.22394417226314545, | |
| "learning_rate": 3.3517104891150695e-05, | |
| "loss": 0.0101, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 0.3994910941475827, | |
| "grad_norm": 0.22689932584762573, | |
| "learning_rate": 3.337574215436811e-05, | |
| "loss": 0.0035, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 0.4020356234096692, | |
| "grad_norm": 0.007624524179846048, | |
| "learning_rate": 3.3234379417585526e-05, | |
| "loss": 0.0038, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 0.40458015267175573, | |
| "grad_norm": 0.6771214008331299, | |
| "learning_rate": 3.309301668080294e-05, | |
| "loss": 0.0077, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 0.4071246819338422, | |
| "grad_norm": 0.04878820851445198, | |
| "learning_rate": 3.295165394402036e-05, | |
| "loss": 0.005, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 0.40966921119592875, | |
| "grad_norm": 0.005000208970159292, | |
| "learning_rate": 3.281029120723777e-05, | |
| "loss": 0.0034, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 0.4122137404580153, | |
| "grad_norm": 0.11921621114015579, | |
| "learning_rate": 3.266892847045519e-05, | |
| "loss": 0.0012, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 0.41475826972010177, | |
| "grad_norm": 0.8879194855690002, | |
| "learning_rate": 3.2527565733672604e-05, | |
| "loss": 0.0039, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 0.4173027989821883, | |
| "grad_norm": 0.7251662015914917, | |
| "learning_rate": 3.238620299689002e-05, | |
| "loss": 0.0097, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 0.4198473282442748, | |
| "grad_norm": 0.0009248352725990117, | |
| "learning_rate": 3.224484026010744e-05, | |
| "loss": 0.0017, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 0.4223918575063613, | |
| "grad_norm": 0.018004465848207474, | |
| "learning_rate": 3.210347752332485e-05, | |
| "loss": 0.0122, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 0.42493638676844786, | |
| "grad_norm": 0.01126956008374691, | |
| "learning_rate": 3.196211478654227e-05, | |
| "loss": 0.0053, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 0.42748091603053434, | |
| "grad_norm": 0.08610785752534866, | |
| "learning_rate": 3.182075204975968e-05, | |
| "loss": 0.0068, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 0.4300254452926209, | |
| "grad_norm": 0.08730663359165192, | |
| "learning_rate": 3.16793893129771e-05, | |
| "loss": 0.0108, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 0.43256997455470736, | |
| "grad_norm": 0.16888225078582764, | |
| "learning_rate": 3.153802657619452e-05, | |
| "loss": 0.0066, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 0.4351145038167939, | |
| "grad_norm": 0.6367448568344116, | |
| "learning_rate": 3.139666383941193e-05, | |
| "loss": 0.0106, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 0.43765903307888043, | |
| "grad_norm": 0.00321201141923666, | |
| "learning_rate": 3.1255301102629346e-05, | |
| "loss": 0.0067, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 0.4402035623409669, | |
| "grad_norm": 0.8670584559440613, | |
| "learning_rate": 3.111393836584676e-05, | |
| "loss": 0.009, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 0.44274809160305345, | |
| "grad_norm": 0.011140185408294201, | |
| "learning_rate": 3.0972575629064184e-05, | |
| "loss": 0.0027, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 0.44529262086513993, | |
| "grad_norm": 0.481245756149292, | |
| "learning_rate": 3.08312128922816e-05, | |
| "loss": 0.0069, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 0.44783715012722647, | |
| "grad_norm": 0.005612820386886597, | |
| "learning_rate": 3.068985015549901e-05, | |
| "loss": 0.0056, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 0.45038167938931295, | |
| "grad_norm": 0.09036795794963837, | |
| "learning_rate": 3.0548487418716424e-05, | |
| "loss": 0.0066, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 0.4529262086513995, | |
| "grad_norm": 1.0461736917495728, | |
| "learning_rate": 3.0407124681933847e-05, | |
| "loss": 0.0149, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 0.455470737913486, | |
| "grad_norm": 0.030460286885499954, | |
| "learning_rate": 3.026576194515126e-05, | |
| "loss": 0.0054, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 0.4580152671755725, | |
| "grad_norm": 0.6616231203079224, | |
| "learning_rate": 3.0124399208368675e-05, | |
| "loss": 0.0103, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 0.46055979643765904, | |
| "grad_norm": 0.02124844118952751, | |
| "learning_rate": 2.998303647158609e-05, | |
| "loss": 0.0097, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 0.4631043256997455, | |
| "grad_norm": 0.11012452095746994, | |
| "learning_rate": 2.9841673734803506e-05, | |
| "loss": 0.0065, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 0.46564885496183206, | |
| "grad_norm": 0.01103823259472847, | |
| "learning_rate": 2.9700310998020925e-05, | |
| "loss": 0.0073, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 0.4681933842239186, | |
| "grad_norm": 0.019804174080491066, | |
| "learning_rate": 2.955894826123834e-05, | |
| "loss": 0.0058, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 0.4707379134860051, | |
| "grad_norm": 0.11179743707180023, | |
| "learning_rate": 2.9417585524455753e-05, | |
| "loss": 0.0058, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 0.4732824427480916, | |
| "grad_norm": 0.2278183400630951, | |
| "learning_rate": 2.927622278767317e-05, | |
| "loss": 0.0128, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 0.4758269720101781, | |
| "grad_norm": 0.0034842302557080984, | |
| "learning_rate": 2.9134860050890588e-05, | |
| "loss": 0.0078, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 0.47837150127226463, | |
| "grad_norm": 0.34698769450187683, | |
| "learning_rate": 2.8993497314108004e-05, | |
| "loss": 0.0034, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 0.48091603053435117, | |
| "grad_norm": 0.883517861366272, | |
| "learning_rate": 2.885213457732542e-05, | |
| "loss": 0.0134, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 0.48346055979643765, | |
| "grad_norm": 0.06100672483444214, | |
| "learning_rate": 2.8710771840542832e-05, | |
| "loss": 0.0088, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 0.4860050890585242, | |
| "grad_norm": 0.1570027470588684, | |
| "learning_rate": 2.8569409103760255e-05, | |
| "loss": 0.0072, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 0.48854961832061067, | |
| "grad_norm": 0.542742908000946, | |
| "learning_rate": 2.8428046366977667e-05, | |
| "loss": 0.0043, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 0.4910941475826972, | |
| "grad_norm": 0.9881167411804199, | |
| "learning_rate": 2.8286683630195083e-05, | |
| "loss": 0.01, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 0.49363867684478374, | |
| "grad_norm": 0.14431729912757874, | |
| "learning_rate": 2.8145320893412498e-05, | |
| "loss": 0.0056, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 0.4961832061068702, | |
| "grad_norm": 0.006597043946385384, | |
| "learning_rate": 2.800395815662991e-05, | |
| "loss": 0.0078, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 0.49872773536895676, | |
| "grad_norm": 0.08411956578493118, | |
| "learning_rate": 2.7862595419847333e-05, | |
| "loss": 0.004, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 0.5012722646310432, | |
| "grad_norm": 0.49558550119400024, | |
| "learning_rate": 2.7721232683064745e-05, | |
| "loss": 0.0058, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 0.5038167938931297, | |
| "grad_norm": 0.27857571840286255, | |
| "learning_rate": 2.757986994628216e-05, | |
| "loss": 0.0078, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 0.5063613231552163, | |
| "grad_norm": 0.2860904037952423, | |
| "learning_rate": 2.7438507209499577e-05, | |
| "loss": 0.0038, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 0.5089058524173028, | |
| "grad_norm": 0.14743930101394653, | |
| "learning_rate": 2.7297144472716996e-05, | |
| "loss": 0.0018, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5089058524173028, | |
| "eval_loss": 0.026436129584908485, | |
| "eval_runtime": 133.858, | |
| "eval_samples_per_second": 59.63, | |
| "eval_steps_per_second": 0.471, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 0.5114503816793893, | |
| "grad_norm": 0.0035872168373316526, | |
| "learning_rate": 2.715578173593441e-05, | |
| "loss": 0.0107, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 0.5139949109414759, | |
| "grad_norm": 0.34011369943618774, | |
| "learning_rate": 2.7014418999151824e-05, | |
| "loss": 0.0055, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 0.5165394402035624, | |
| "grad_norm": 0.16972984373569489, | |
| "learning_rate": 2.687305626236924e-05, | |
| "loss": 0.0064, | |
| "step": 2030 | |
| }, | |
| { | |
| "epoch": 0.5190839694656488, | |
| "grad_norm": 0.4104236960411072, | |
| "learning_rate": 2.6731693525586655e-05, | |
| "loss": 0.0085, | |
| "step": 2040 | |
| }, | |
| { | |
| "epoch": 0.5216284987277354, | |
| "grad_norm": 0.5132906436920166, | |
| "learning_rate": 2.6590330788804074e-05, | |
| "loss": 0.0101, | |
| "step": 2050 | |
| }, | |
| { | |
| "epoch": 0.5241730279898219, | |
| "grad_norm": 0.6972320079803467, | |
| "learning_rate": 2.644896805202149e-05, | |
| "loss": 0.0072, | |
| "step": 2060 | |
| }, | |
| { | |
| "epoch": 0.5267175572519084, | |
| "grad_norm": 0.6260659694671631, | |
| "learning_rate": 2.6307605315238902e-05, | |
| "loss": 0.0149, | |
| "step": 2070 | |
| }, | |
| { | |
| "epoch": 0.5292620865139949, | |
| "grad_norm": 0.19219329953193665, | |
| "learning_rate": 2.6166242578456318e-05, | |
| "loss": 0.0105, | |
| "step": 2080 | |
| }, | |
| { | |
| "epoch": 0.5318066157760815, | |
| "grad_norm": 0.2450764775276184, | |
| "learning_rate": 2.6024879841673737e-05, | |
| "loss": 0.0114, | |
| "step": 2090 | |
| }, | |
| { | |
| "epoch": 0.5343511450381679, | |
| "grad_norm": 0.70209139585495, | |
| "learning_rate": 2.5883517104891153e-05, | |
| "loss": 0.0123, | |
| "step": 2100 | |
| }, | |
| { | |
| "epoch": 0.5368956743002544, | |
| "grad_norm": 0.04102372005581856, | |
| "learning_rate": 2.574215436810857e-05, | |
| "loss": 0.0102, | |
| "step": 2110 | |
| }, | |
| { | |
| "epoch": 0.539440203562341, | |
| "grad_norm": 0.0783892348408699, | |
| "learning_rate": 2.560079163132598e-05, | |
| "loss": 0.007, | |
| "step": 2120 | |
| }, | |
| { | |
| "epoch": 0.5419847328244275, | |
| "grad_norm": 0.021079327911138535, | |
| "learning_rate": 2.5459428894543404e-05, | |
| "loss": 0.0046, | |
| "step": 2130 | |
| }, | |
| { | |
| "epoch": 0.544529262086514, | |
| "grad_norm": 0.03712490573525429, | |
| "learning_rate": 2.5318066157760816e-05, | |
| "loss": 0.0106, | |
| "step": 2140 | |
| }, | |
| { | |
| "epoch": 0.5470737913486005, | |
| "grad_norm": 0.07305669784545898, | |
| "learning_rate": 2.517670342097823e-05, | |
| "loss": 0.0101, | |
| "step": 2150 | |
| }, | |
| { | |
| "epoch": 0.549618320610687, | |
| "grad_norm": 0.33150365948677063, | |
| "learning_rate": 2.5035340684195647e-05, | |
| "loss": 0.0035, | |
| "step": 2160 | |
| }, | |
| { | |
| "epoch": 0.5521628498727735, | |
| "grad_norm": 0.16319264471530914, | |
| "learning_rate": 2.4893977947413063e-05, | |
| "loss": 0.0054, | |
| "step": 2170 | |
| }, | |
| { | |
| "epoch": 0.55470737913486, | |
| "grad_norm": 0.5374312400817871, | |
| "learning_rate": 2.475261521063048e-05, | |
| "loss": 0.0089, | |
| "step": 2180 | |
| }, | |
| { | |
| "epoch": 0.5572519083969466, | |
| "grad_norm": 0.008280829526484013, | |
| "learning_rate": 2.4611252473847894e-05, | |
| "loss": 0.0052, | |
| "step": 2190 | |
| }, | |
| { | |
| "epoch": 0.5597964376590331, | |
| "grad_norm": 0.006080132443457842, | |
| "learning_rate": 2.446988973706531e-05, | |
| "loss": 0.0097, | |
| "step": 2200 | |
| }, | |
| { | |
| "epoch": 0.5623409669211196, | |
| "grad_norm": 1.0284223556518555, | |
| "learning_rate": 2.432852700028273e-05, | |
| "loss": 0.005, | |
| "step": 2210 | |
| }, | |
| { | |
| "epoch": 0.5648854961832062, | |
| "grad_norm": 0.5061278343200684, | |
| "learning_rate": 2.418716426350014e-05, | |
| "loss": 0.0059, | |
| "step": 2220 | |
| }, | |
| { | |
| "epoch": 0.5674300254452926, | |
| "grad_norm": 0.8988016843795776, | |
| "learning_rate": 2.404580152671756e-05, | |
| "loss": 0.0155, | |
| "step": 2230 | |
| }, | |
| { | |
| "epoch": 0.5699745547073791, | |
| "grad_norm": 0.18407580256462097, | |
| "learning_rate": 2.3904438789934973e-05, | |
| "loss": 0.0026, | |
| "step": 2240 | |
| }, | |
| { | |
| "epoch": 0.5725190839694656, | |
| "grad_norm": 0.15080343186855316, | |
| "learning_rate": 2.376307605315239e-05, | |
| "loss": 0.0092, | |
| "step": 2250 | |
| }, | |
| { | |
| "epoch": 0.5750636132315522, | |
| "grad_norm": 0.02580890618264675, | |
| "learning_rate": 2.3621713316369808e-05, | |
| "loss": 0.0055, | |
| "step": 2260 | |
| }, | |
| { | |
| "epoch": 0.5776081424936387, | |
| "grad_norm": 0.46047937870025635, | |
| "learning_rate": 2.348035057958722e-05, | |
| "loss": 0.0075, | |
| "step": 2270 | |
| }, | |
| { | |
| "epoch": 0.5801526717557252, | |
| "grad_norm": 0.457792729139328, | |
| "learning_rate": 2.333898784280464e-05, | |
| "loss": 0.0071, | |
| "step": 2280 | |
| }, | |
| { | |
| "epoch": 0.5826972010178118, | |
| "grad_norm": 0.7703973650932312, | |
| "learning_rate": 2.319762510602205e-05, | |
| "loss": 0.0082, | |
| "step": 2290 | |
| }, | |
| { | |
| "epoch": 0.5852417302798982, | |
| "grad_norm": 0.015261597000062466, | |
| "learning_rate": 2.305626236923947e-05, | |
| "loss": 0.0005, | |
| "step": 2300 | |
| }, | |
| { | |
| "epoch": 0.5877862595419847, | |
| "grad_norm": 0.3900598883628845, | |
| "learning_rate": 2.2914899632456886e-05, | |
| "loss": 0.0031, | |
| "step": 2310 | |
| }, | |
| { | |
| "epoch": 0.5903307888040712, | |
| "grad_norm": 0.027707261964678764, | |
| "learning_rate": 2.2773536895674302e-05, | |
| "loss": 0.0122, | |
| "step": 2320 | |
| }, | |
| { | |
| "epoch": 0.5928753180661578, | |
| "grad_norm": 0.15728776156902313, | |
| "learning_rate": 2.2632174158891718e-05, | |
| "loss": 0.0057, | |
| "step": 2330 | |
| }, | |
| { | |
| "epoch": 0.5954198473282443, | |
| "grad_norm": 0.21821542084217072, | |
| "learning_rate": 2.2490811422109133e-05, | |
| "loss": 0.0051, | |
| "step": 2340 | |
| }, | |
| { | |
| "epoch": 0.5979643765903307, | |
| "grad_norm": 0.002715646056458354, | |
| "learning_rate": 2.234944868532655e-05, | |
| "loss": 0.0084, | |
| "step": 2350 | |
| }, | |
| { | |
| "epoch": 0.6005089058524173, | |
| "grad_norm": 0.3513769805431366, | |
| "learning_rate": 2.2208085948543965e-05, | |
| "loss": 0.0032, | |
| "step": 2360 | |
| }, | |
| { | |
| "epoch": 0.6030534351145038, | |
| "grad_norm": 0.1429492086172104, | |
| "learning_rate": 2.206672321176138e-05, | |
| "loss": 0.0037, | |
| "step": 2370 | |
| }, | |
| { | |
| "epoch": 0.6055979643765903, | |
| "grad_norm": 0.02939917892217636, | |
| "learning_rate": 2.1925360474978796e-05, | |
| "loss": 0.0102, | |
| "step": 2380 | |
| }, | |
| { | |
| "epoch": 0.6081424936386769, | |
| "grad_norm": 0.3557289242744446, | |
| "learning_rate": 2.1783997738196212e-05, | |
| "loss": 0.0062, | |
| "step": 2390 | |
| }, | |
| { | |
| "epoch": 0.6106870229007634, | |
| "grad_norm": 0.094198077917099, | |
| "learning_rate": 2.1642635001413628e-05, | |
| "loss": 0.0123, | |
| "step": 2400 | |
| }, | |
| { | |
| "epoch": 0.6132315521628499, | |
| "grad_norm": 0.13312627375125885, | |
| "learning_rate": 2.1501272264631043e-05, | |
| "loss": 0.0059, | |
| "step": 2410 | |
| }, | |
| { | |
| "epoch": 0.6157760814249363, | |
| "grad_norm": 0.7499666810035706, | |
| "learning_rate": 2.135990952784846e-05, | |
| "loss": 0.013, | |
| "step": 2420 | |
| }, | |
| { | |
| "epoch": 0.6183206106870229, | |
| "grad_norm": 0.04554931819438934, | |
| "learning_rate": 2.1218546791065878e-05, | |
| "loss": 0.0034, | |
| "step": 2430 | |
| }, | |
| { | |
| "epoch": 0.6208651399491094, | |
| "grad_norm": 0.01164567656815052, | |
| "learning_rate": 2.107718405428329e-05, | |
| "loss": 0.0091, | |
| "step": 2440 | |
| }, | |
| { | |
| "epoch": 0.6234096692111959, | |
| "grad_norm": 1.0809807777404785, | |
| "learning_rate": 2.093582131750071e-05, | |
| "loss": 0.0049, | |
| "step": 2450 | |
| }, | |
| { | |
| "epoch": 0.6259541984732825, | |
| "grad_norm": 0.679319441318512, | |
| "learning_rate": 2.0794458580718122e-05, | |
| "loss": 0.0103, | |
| "step": 2460 | |
| }, | |
| { | |
| "epoch": 0.628498727735369, | |
| "grad_norm": 0.37474673986434937, | |
| "learning_rate": 2.0653095843935538e-05, | |
| "loss": 0.0053, | |
| "step": 2470 | |
| }, | |
| { | |
| "epoch": 0.6310432569974554, | |
| "grad_norm": 0.3449637293815613, | |
| "learning_rate": 2.0511733107152957e-05, | |
| "loss": 0.0232, | |
| "step": 2480 | |
| }, | |
| { | |
| "epoch": 0.6335877862595419, | |
| "grad_norm": 0.022124435752630234, | |
| "learning_rate": 2.037037037037037e-05, | |
| "loss": 0.0089, | |
| "step": 2490 | |
| }, | |
| { | |
| "epoch": 0.6361323155216285, | |
| "grad_norm": 0.4030318856239319, | |
| "learning_rate": 2.0229007633587788e-05, | |
| "loss": 0.0043, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.6361323155216285, | |
| "eval_loss": 0.026291608810424805, | |
| "eval_runtime": 133.866, | |
| "eval_samples_per_second": 59.627, | |
| "eval_steps_per_second": 0.471, | |
| "step": 2500 | |
| }, | |
| { | |
| "epoch": 0.638676844783715, | |
| "grad_norm": 0.2445756196975708, | |
| "learning_rate": 2.0087644896805204e-05, | |
| "loss": 0.0124, | |
| "step": 2510 | |
| }, | |
| { | |
| "epoch": 0.6412213740458015, | |
| "grad_norm": 0.42559316754341125, | |
| "learning_rate": 1.994628216002262e-05, | |
| "loss": 0.0063, | |
| "step": 2520 | |
| }, | |
| { | |
| "epoch": 0.6437659033078881, | |
| "grad_norm": 0.18691177666187286, | |
| "learning_rate": 1.9804919423240035e-05, | |
| "loss": 0.0091, | |
| "step": 2530 | |
| }, | |
| { | |
| "epoch": 0.6463104325699746, | |
| "grad_norm": 0.006653166376054287, | |
| "learning_rate": 1.966355668645745e-05, | |
| "loss": 0.0058, | |
| "step": 2540 | |
| }, | |
| { | |
| "epoch": 0.648854961832061, | |
| "grad_norm": 0.6509602069854736, | |
| "learning_rate": 1.9522193949674867e-05, | |
| "loss": 0.0097, | |
| "step": 2550 | |
| }, | |
| { | |
| "epoch": 0.6513994910941476, | |
| "grad_norm": 0.7203097939491272, | |
| "learning_rate": 1.9380831212892282e-05, | |
| "loss": 0.0283, | |
| "step": 2560 | |
| }, | |
| { | |
| "epoch": 0.6539440203562341, | |
| "grad_norm": 0.4940910041332245, | |
| "learning_rate": 1.9239468476109698e-05, | |
| "loss": 0.0112, | |
| "step": 2570 | |
| }, | |
| { | |
| "epoch": 0.6564885496183206, | |
| "grad_norm": 0.04976008087396622, | |
| "learning_rate": 1.9098105739327117e-05, | |
| "loss": 0.0088, | |
| "step": 2580 | |
| }, | |
| { | |
| "epoch": 0.6590330788804071, | |
| "grad_norm": 0.5175666213035583, | |
| "learning_rate": 1.895674300254453e-05, | |
| "loss": 0.0128, | |
| "step": 2590 | |
| }, | |
| { | |
| "epoch": 0.6615776081424937, | |
| "grad_norm": 0.11857312172651291, | |
| "learning_rate": 1.8815380265761945e-05, | |
| "loss": 0.0043, | |
| "step": 2600 | |
| }, | |
| { | |
| "epoch": 0.6641221374045801, | |
| "grad_norm": 0.020734384655952454, | |
| "learning_rate": 1.867401752897936e-05, | |
| "loss": 0.0124, | |
| "step": 2610 | |
| }, | |
| { | |
| "epoch": 0.6666666666666666, | |
| "grad_norm": 0.013484890572726727, | |
| "learning_rate": 1.8532654792196777e-05, | |
| "loss": 0.0058, | |
| "step": 2620 | |
| }, | |
| { | |
| "epoch": 0.6692111959287532, | |
| "grad_norm": 0.016002101823687553, | |
| "learning_rate": 1.8391292055414196e-05, | |
| "loss": 0.003, | |
| "step": 2630 | |
| }, | |
| { | |
| "epoch": 0.6717557251908397, | |
| "grad_norm": 0.5079576969146729, | |
| "learning_rate": 1.8249929318631608e-05, | |
| "loss": 0.0069, | |
| "step": 2640 | |
| }, | |
| { | |
| "epoch": 0.6743002544529262, | |
| "grad_norm": 0.10626394301652908, | |
| "learning_rate": 1.8108566581849027e-05, | |
| "loss": 0.0046, | |
| "step": 2650 | |
| }, | |
| { | |
| "epoch": 0.6768447837150128, | |
| "grad_norm": 0.29682421684265137, | |
| "learning_rate": 1.796720384506644e-05, | |
| "loss": 0.0016, | |
| "step": 2660 | |
| }, | |
| { | |
| "epoch": 0.6793893129770993, | |
| "grad_norm": 0.18330760300159454, | |
| "learning_rate": 1.782584110828386e-05, | |
| "loss": 0.0104, | |
| "step": 2670 | |
| }, | |
| { | |
| "epoch": 0.6819338422391857, | |
| "grad_norm": 0.6251747012138367, | |
| "learning_rate": 1.7684478371501274e-05, | |
| "loss": 0.0256, | |
| "step": 2680 | |
| }, | |
| { | |
| "epoch": 0.6844783715012722, | |
| "grad_norm": 0.8732611536979675, | |
| "learning_rate": 1.754311563471869e-05, | |
| "loss": 0.0154, | |
| "step": 2690 | |
| }, | |
| { | |
| "epoch": 0.6870229007633588, | |
| "grad_norm": 0.9835492372512817, | |
| "learning_rate": 1.7401752897936106e-05, | |
| "loss": 0.0069, | |
| "step": 2700 | |
| }, | |
| { | |
| "epoch": 0.6895674300254453, | |
| "grad_norm": 0.8463011980056763, | |
| "learning_rate": 1.7260390161153518e-05, | |
| "loss": 0.0062, | |
| "step": 2710 | |
| }, | |
| { | |
| "epoch": 0.6921119592875318, | |
| "grad_norm": 0.034930918365716934, | |
| "learning_rate": 1.7119027424370937e-05, | |
| "loss": 0.0101, | |
| "step": 2720 | |
| }, | |
| { | |
| "epoch": 0.6946564885496184, | |
| "grad_norm": 0.022007199004292488, | |
| "learning_rate": 1.6977664687588353e-05, | |
| "loss": 0.0094, | |
| "step": 2730 | |
| }, | |
| { | |
| "epoch": 0.6972010178117048, | |
| "grad_norm": 0.0753493681550026, | |
| "learning_rate": 1.683630195080577e-05, | |
| "loss": 0.0102, | |
| "step": 2740 | |
| }, | |
| { | |
| "epoch": 0.6997455470737913, | |
| "grad_norm": 0.2625516653060913, | |
| "learning_rate": 1.6694939214023184e-05, | |
| "loss": 0.0051, | |
| "step": 2750 | |
| }, | |
| { | |
| "epoch": 0.7022900763358778, | |
| "grad_norm": 0.11348223686218262, | |
| "learning_rate": 1.65535764772406e-05, | |
| "loss": 0.0057, | |
| "step": 2760 | |
| }, | |
| { | |
| "epoch": 0.7048346055979644, | |
| "grad_norm": 0.299972265958786, | |
| "learning_rate": 1.6412213740458016e-05, | |
| "loss": 0.0078, | |
| "step": 2770 | |
| }, | |
| { | |
| "epoch": 0.7073791348600509, | |
| "grad_norm": 0.019401008263230324, | |
| "learning_rate": 1.627085100367543e-05, | |
| "loss": 0.0011, | |
| "step": 2780 | |
| }, | |
| { | |
| "epoch": 0.7099236641221374, | |
| "grad_norm": 0.0028962348587810993, | |
| "learning_rate": 1.6129488266892847e-05, | |
| "loss": 0.0026, | |
| "step": 2790 | |
| }, | |
| { | |
| "epoch": 0.712468193384224, | |
| "grad_norm": 0.22273877263069153, | |
| "learning_rate": 1.5988125530110266e-05, | |
| "loss": 0.005, | |
| "step": 2800 | |
| }, | |
| { | |
| "epoch": 0.7150127226463104, | |
| "grad_norm": 0.005791067145764828, | |
| "learning_rate": 1.584676279332768e-05, | |
| "loss": 0.0062, | |
| "step": 2810 | |
| }, | |
| { | |
| "epoch": 0.7175572519083969, | |
| "grad_norm": 0.5451294779777527, | |
| "learning_rate": 1.5705400056545094e-05, | |
| "loss": 0.0054, | |
| "step": 2820 | |
| }, | |
| { | |
| "epoch": 0.7201017811704835, | |
| "grad_norm": 0.009409861639142036, | |
| "learning_rate": 1.556403731976251e-05, | |
| "loss": 0.0019, | |
| "step": 2830 | |
| }, | |
| { | |
| "epoch": 0.72264631043257, | |
| "grad_norm": 0.016224944964051247, | |
| "learning_rate": 1.5422674582979926e-05, | |
| "loss": 0.0136, | |
| "step": 2840 | |
| }, | |
| { | |
| "epoch": 0.7251908396946565, | |
| "grad_norm": 0.0011498507810756564, | |
| "learning_rate": 1.5281311846197345e-05, | |
| "loss": 0.0037, | |
| "step": 2850 | |
| }, | |
| { | |
| "epoch": 0.727735368956743, | |
| "grad_norm": 0.2645626366138458, | |
| "learning_rate": 1.5139949109414759e-05, | |
| "loss": 0.0136, | |
| "step": 2860 | |
| }, | |
| { | |
| "epoch": 0.7302798982188295, | |
| "grad_norm": 0.3801277279853821, | |
| "learning_rate": 1.4998586372632176e-05, | |
| "loss": 0.001, | |
| "step": 2870 | |
| }, | |
| { | |
| "epoch": 0.732824427480916, | |
| "grad_norm": 0.8078188300132751, | |
| "learning_rate": 1.485722363584959e-05, | |
| "loss": 0.0036, | |
| "step": 2880 | |
| }, | |
| { | |
| "epoch": 0.7353689567430025, | |
| "grad_norm": 0.27664822340011597, | |
| "learning_rate": 1.4715860899067008e-05, | |
| "loss": 0.006, | |
| "step": 2890 | |
| }, | |
| { | |
| "epoch": 0.7379134860050891, | |
| "grad_norm": 0.3722473978996277, | |
| "learning_rate": 1.4574498162284422e-05, | |
| "loss": 0.0129, | |
| "step": 2900 | |
| }, | |
| { | |
| "epoch": 0.7404580152671756, | |
| "grad_norm": 0.028823425993323326, | |
| "learning_rate": 1.4433135425501839e-05, | |
| "loss": 0.0109, | |
| "step": 2910 | |
| }, | |
| { | |
| "epoch": 0.7430025445292621, | |
| "grad_norm": 0.15513518452644348, | |
| "learning_rate": 1.4291772688719255e-05, | |
| "loss": 0.0055, | |
| "step": 2920 | |
| }, | |
| { | |
| "epoch": 0.7455470737913485, | |
| "grad_norm": 0.04313421621918678, | |
| "learning_rate": 1.4150409951936669e-05, | |
| "loss": 0.0022, | |
| "step": 2930 | |
| }, | |
| { | |
| "epoch": 0.7480916030534351, | |
| "grad_norm": 0.8002997636795044, | |
| "learning_rate": 1.4009047215154086e-05, | |
| "loss": 0.0063, | |
| "step": 2940 | |
| }, | |
| { | |
| "epoch": 0.7506361323155216, | |
| "grad_norm": 0.4388156235218048, | |
| "learning_rate": 1.38676844783715e-05, | |
| "loss": 0.0083, | |
| "step": 2950 | |
| }, | |
| { | |
| "epoch": 0.7531806615776081, | |
| "grad_norm": 0.5030899047851562, | |
| "learning_rate": 1.3726321741588918e-05, | |
| "loss": 0.0048, | |
| "step": 2960 | |
| }, | |
| { | |
| "epoch": 0.7557251908396947, | |
| "grad_norm": 0.551201581954956, | |
| "learning_rate": 1.3584959004806333e-05, | |
| "loss": 0.0074, | |
| "step": 2970 | |
| }, | |
| { | |
| "epoch": 0.7582697201017812, | |
| "grad_norm": 0.08817600458860397, | |
| "learning_rate": 1.344359626802375e-05, | |
| "loss": 0.0089, | |
| "step": 2980 | |
| }, | |
| { | |
| "epoch": 0.7608142493638677, | |
| "grad_norm": 0.1626739203929901, | |
| "learning_rate": 1.3302233531241165e-05, | |
| "loss": 0.0037, | |
| "step": 2990 | |
| }, | |
| { | |
| "epoch": 0.7633587786259542, | |
| "grad_norm": 0.5182027816772461, | |
| "learning_rate": 1.3160870794458582e-05, | |
| "loss": 0.0079, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7633587786259542, | |
| "eval_loss": 0.026297220960259438, | |
| "eval_runtime": 134.251, | |
| "eval_samples_per_second": 59.456, | |
| "eval_steps_per_second": 0.469, | |
| "step": 3000 | |
| }, | |
| { | |
| "epoch": 0.7659033078880407, | |
| "grad_norm": 0.015984637662768364, | |
| "learning_rate": 1.3019508057675996e-05, | |
| "loss": 0.008, | |
| "step": 3010 | |
| }, | |
| { | |
| "epoch": 0.7684478371501272, | |
| "grad_norm": 0.05166172236204147, | |
| "learning_rate": 1.2878145320893414e-05, | |
| "loss": 0.008, | |
| "step": 3020 | |
| }, | |
| { | |
| "epoch": 0.7709923664122137, | |
| "grad_norm": 1.257007360458374, | |
| "learning_rate": 1.273678258411083e-05, | |
| "loss": 0.0077, | |
| "step": 3030 | |
| }, | |
| { | |
| "epoch": 0.7735368956743003, | |
| "grad_norm": 0.3560980558395386, | |
| "learning_rate": 1.2595419847328243e-05, | |
| "loss": 0.0061, | |
| "step": 3040 | |
| }, | |
| { | |
| "epoch": 0.7760814249363868, | |
| "grad_norm": 0.025006460025906563, | |
| "learning_rate": 1.245405711054566e-05, | |
| "loss": 0.0019, | |
| "step": 3050 | |
| }, | |
| { | |
| "epoch": 0.7786259541984732, | |
| "grad_norm": 0.7989102602005005, | |
| "learning_rate": 1.2312694373763076e-05, | |
| "loss": 0.0081, | |
| "step": 3060 | |
| }, | |
| { | |
| "epoch": 0.7811704834605598, | |
| "grad_norm": 0.033314093947410583, | |
| "learning_rate": 1.2171331636980494e-05, | |
| "loss": 0.0195, | |
| "step": 3070 | |
| }, | |
| { | |
| "epoch": 0.7837150127226463, | |
| "grad_norm": 0.24344636499881744, | |
| "learning_rate": 1.202996890019791e-05, | |
| "loss": 0.0036, | |
| "step": 3080 | |
| }, | |
| { | |
| "epoch": 0.7862595419847328, | |
| "grad_norm": 0.02246682159602642, | |
| "learning_rate": 1.1888606163415324e-05, | |
| "loss": 0.0062, | |
| "step": 3090 | |
| }, | |
| { | |
| "epoch": 0.7888040712468194, | |
| "grad_norm": 0.40515023469924927, | |
| "learning_rate": 1.174724342663274e-05, | |
| "loss": 0.006, | |
| "step": 3100 | |
| }, | |
| { | |
| "epoch": 0.7913486005089059, | |
| "grad_norm": 0.08088365942239761, | |
| "learning_rate": 1.1605880689850155e-05, | |
| "loss": 0.0065, | |
| "step": 3110 | |
| }, | |
| { | |
| "epoch": 0.7938931297709924, | |
| "grad_norm": 0.08734181523323059, | |
| "learning_rate": 1.1464517953067572e-05, | |
| "loss": 0.0047, | |
| "step": 3120 | |
| }, | |
| { | |
| "epoch": 0.7964376590330788, | |
| "grad_norm": 0.033424489200115204, | |
| "learning_rate": 1.1323155216284988e-05, | |
| "loss": 0.0099, | |
| "step": 3130 | |
| }, | |
| { | |
| "epoch": 0.7989821882951654, | |
| "grad_norm": 0.08419552445411682, | |
| "learning_rate": 1.1181792479502404e-05, | |
| "loss": 0.0073, | |
| "step": 3140 | |
| }, | |
| { | |
| "epoch": 0.8015267175572519, | |
| "grad_norm": 0.01626414805650711, | |
| "learning_rate": 1.104042974271982e-05, | |
| "loss": 0.007, | |
| "step": 3150 | |
| }, | |
| { | |
| "epoch": 0.8040712468193384, | |
| "grad_norm": 0.5791566371917725, | |
| "learning_rate": 1.0899067005937235e-05, | |
| "loss": 0.0158, | |
| "step": 3160 | |
| }, | |
| { | |
| "epoch": 0.806615776081425, | |
| "grad_norm": 0.03521082550287247, | |
| "learning_rate": 1.0757704269154651e-05, | |
| "loss": 0.0094, | |
| "step": 3170 | |
| }, | |
| { | |
| "epoch": 0.8091603053435115, | |
| "grad_norm": 0.31241437792778015, | |
| "learning_rate": 1.0616341532372068e-05, | |
| "loss": 0.0053, | |
| "step": 3180 | |
| }, | |
| { | |
| "epoch": 0.811704834605598, | |
| "grad_norm": 0.019581139087677002, | |
| "learning_rate": 1.0474978795589484e-05, | |
| "loss": 0.0123, | |
| "step": 3190 | |
| }, | |
| { | |
| "epoch": 0.8142493638676844, | |
| "grad_norm": 0.9257656931877136, | |
| "learning_rate": 1.0333616058806898e-05, | |
| "loss": 0.0266, | |
| "step": 3200 | |
| }, | |
| { | |
| "epoch": 0.816793893129771, | |
| "grad_norm": 0.026714293286204338, | |
| "learning_rate": 1.0192253322024314e-05, | |
| "loss": 0.0009, | |
| "step": 3210 | |
| }, | |
| { | |
| "epoch": 0.8193384223918575, | |
| "grad_norm": 0.08150497823953629, | |
| "learning_rate": 1.0050890585241731e-05, | |
| "loss": 0.0018, | |
| "step": 3220 | |
| }, | |
| { | |
| "epoch": 0.821882951653944, | |
| "grad_norm": 0.291150838136673, | |
| "learning_rate": 9.909527848459147e-06, | |
| "loss": 0.01, | |
| "step": 3230 | |
| }, | |
| { | |
| "epoch": 0.8244274809160306, | |
| "grad_norm": 0.20612020790576935, | |
| "learning_rate": 9.768165111676563e-06, | |
| "loss": 0.0041, | |
| "step": 3240 | |
| }, | |
| { | |
| "epoch": 0.8269720101781171, | |
| "grad_norm": 0.7024424076080322, | |
| "learning_rate": 9.626802374893978e-06, | |
| "loss": 0.0045, | |
| "step": 3250 | |
| }, | |
| { | |
| "epoch": 0.8295165394402035, | |
| "grad_norm": 0.27987831830978394, | |
| "learning_rate": 9.485439638111394e-06, | |
| "loss": 0.0032, | |
| "step": 3260 | |
| }, | |
| { | |
| "epoch": 0.8320610687022901, | |
| "grad_norm": 0.46367907524108887, | |
| "learning_rate": 9.34407690132881e-06, | |
| "loss": 0.0074, | |
| "step": 3270 | |
| }, | |
| { | |
| "epoch": 0.8346055979643766, | |
| "grad_norm": 0.6092146635055542, | |
| "learning_rate": 9.202714164546227e-06, | |
| "loss": 0.0174, | |
| "step": 3280 | |
| }, | |
| { | |
| "epoch": 0.8371501272264631, | |
| "grad_norm": 0.22869914770126343, | |
| "learning_rate": 9.061351427763643e-06, | |
| "loss": 0.0075, | |
| "step": 3290 | |
| }, | |
| { | |
| "epoch": 0.8396946564885496, | |
| "grad_norm": 0.5093429088592529, | |
| "learning_rate": 8.919988690981059e-06, | |
| "loss": 0.0088, | |
| "step": 3300 | |
| }, | |
| { | |
| "epoch": 0.8422391857506362, | |
| "grad_norm": 0.07150623947381973, | |
| "learning_rate": 8.778625954198473e-06, | |
| "loss": 0.0012, | |
| "step": 3310 | |
| }, | |
| { | |
| "epoch": 0.8447837150127226, | |
| "grad_norm": 0.048805948346853256, | |
| "learning_rate": 8.637263217415888e-06, | |
| "loss": 0.0057, | |
| "step": 3320 | |
| }, | |
| { | |
| "epoch": 0.8473282442748091, | |
| "grad_norm": 0.21837657690048218, | |
| "learning_rate": 8.495900480633306e-06, | |
| "loss": 0.0044, | |
| "step": 3330 | |
| }, | |
| { | |
| "epoch": 0.8498727735368957, | |
| "grad_norm": 0.054293371737003326, | |
| "learning_rate": 8.354537743850721e-06, | |
| "loss": 0.0105, | |
| "step": 3340 | |
| }, | |
| { | |
| "epoch": 0.8524173027989822, | |
| "grad_norm": 0.4355064332485199, | |
| "learning_rate": 8.213175007068137e-06, | |
| "loss": 0.0047, | |
| "step": 3350 | |
| }, | |
| { | |
| "epoch": 0.8549618320610687, | |
| "grad_norm": 0.2688146233558655, | |
| "learning_rate": 8.071812270285553e-06, | |
| "loss": 0.0127, | |
| "step": 3360 | |
| }, | |
| { | |
| "epoch": 0.8575063613231552, | |
| "grad_norm": 0.12613950669765472, | |
| "learning_rate": 7.930449533502969e-06, | |
| "loss": 0.0064, | |
| "step": 3370 | |
| }, | |
| { | |
| "epoch": 0.8600508905852418, | |
| "grad_norm": 0.3896316885948181, | |
| "learning_rate": 7.789086796720386e-06, | |
| "loss": 0.0073, | |
| "step": 3380 | |
| }, | |
| { | |
| "epoch": 0.8625954198473282, | |
| "grad_norm": 0.0075408038683235645, | |
| "learning_rate": 7.647724059937802e-06, | |
| "loss": 0.0039, | |
| "step": 3390 | |
| }, | |
| { | |
| "epoch": 0.8651399491094147, | |
| "grad_norm": 0.11392281204462051, | |
| "learning_rate": 7.506361323155217e-06, | |
| "loss": 0.0046, | |
| "step": 3400 | |
| }, | |
| { | |
| "epoch": 0.8676844783715013, | |
| "grad_norm": 0.5989497303962708, | |
| "learning_rate": 7.364998586372633e-06, | |
| "loss": 0.0119, | |
| "step": 3410 | |
| }, | |
| { | |
| "epoch": 0.8702290076335878, | |
| "grad_norm": 0.016849737614393234, | |
| "learning_rate": 7.223635849590048e-06, | |
| "loss": 0.0044, | |
| "step": 3420 | |
| }, | |
| { | |
| "epoch": 0.8727735368956743, | |
| "grad_norm": 0.4961392879486084, | |
| "learning_rate": 7.082273112807464e-06, | |
| "loss": 0.0038, | |
| "step": 3430 | |
| }, | |
| { | |
| "epoch": 0.8753180661577609, | |
| "grad_norm": 0.003794967196881771, | |
| "learning_rate": 6.94091037602488e-06, | |
| "loss": 0.0041, | |
| "step": 3440 | |
| }, | |
| { | |
| "epoch": 0.8778625954198473, | |
| "grad_norm": 0.003887136932462454, | |
| "learning_rate": 6.799547639242296e-06, | |
| "loss": 0.0104, | |
| "step": 3450 | |
| }, | |
| { | |
| "epoch": 0.8804071246819338, | |
| "grad_norm": 0.0005345813115127385, | |
| "learning_rate": 6.658184902459712e-06, | |
| "loss": 0.0039, | |
| "step": 3460 | |
| }, | |
| { | |
| "epoch": 0.8829516539440203, | |
| "grad_norm": 0.06041156128048897, | |
| "learning_rate": 6.516822165677128e-06, | |
| "loss": 0.0029, | |
| "step": 3470 | |
| }, | |
| { | |
| "epoch": 0.8854961832061069, | |
| "grad_norm": 0.19050799310207367, | |
| "learning_rate": 6.375459428894544e-06, | |
| "loss": 0.004, | |
| "step": 3480 | |
| }, | |
| { | |
| "epoch": 0.8880407124681934, | |
| "grad_norm": 0.46865060925483704, | |
| "learning_rate": 6.2340966921119596e-06, | |
| "loss": 0.0097, | |
| "step": 3490 | |
| }, | |
| { | |
| "epoch": 0.8905852417302799, | |
| "grad_norm": 0.10857687890529633, | |
| "learning_rate": 6.092733955329376e-06, | |
| "loss": 0.0158, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8905852417302799, | |
| "eval_loss": 0.024917850270867348, | |
| "eval_runtime": 133.9245, | |
| "eval_samples_per_second": 59.601, | |
| "eval_steps_per_second": 0.47, | |
| "step": 3500 | |
| }, | |
| { | |
| "epoch": 0.8931297709923665, | |
| "grad_norm": 0.003964769653975964, | |
| "learning_rate": 5.951371218546791e-06, | |
| "loss": 0.0026, | |
| "step": 3510 | |
| }, | |
| { | |
| "epoch": 0.8956743002544529, | |
| "grad_norm": 0.36591967940330505, | |
| "learning_rate": 5.810008481764207e-06, | |
| "loss": 0.0037, | |
| "step": 3520 | |
| }, | |
| { | |
| "epoch": 0.8982188295165394, | |
| "grad_norm": 0.2202427238225937, | |
| "learning_rate": 5.668645744981623e-06, | |
| "loss": 0.0038, | |
| "step": 3530 | |
| }, | |
| { | |
| "epoch": 0.9007633587786259, | |
| "grad_norm": 0.6670795679092407, | |
| "learning_rate": 5.527283008199039e-06, | |
| "loss": 0.0083, | |
| "step": 3540 | |
| }, | |
| { | |
| "epoch": 0.9033078880407125, | |
| "grad_norm": 0.009889673441648483, | |
| "learning_rate": 5.3859202714164555e-06, | |
| "loss": 0.0045, | |
| "step": 3550 | |
| }, | |
| { | |
| "epoch": 0.905852417302799, | |
| "grad_norm": 0.1490902602672577, | |
| "learning_rate": 5.24455753463387e-06, | |
| "loss": 0.0032, | |
| "step": 3560 | |
| }, | |
| { | |
| "epoch": 0.9083969465648855, | |
| "grad_norm": 0.011089697480201721, | |
| "learning_rate": 5.103194797851286e-06, | |
| "loss": 0.0058, | |
| "step": 3570 | |
| }, | |
| { | |
| "epoch": 0.910941475826972, | |
| "grad_norm": 0.015895185992121696, | |
| "learning_rate": 4.961832061068703e-06, | |
| "loss": 0.0064, | |
| "step": 3580 | |
| }, | |
| { | |
| "epoch": 0.9134860050890585, | |
| "grad_norm": 0.27433764934539795, | |
| "learning_rate": 4.820469324286118e-06, | |
| "loss": 0.0061, | |
| "step": 3590 | |
| }, | |
| { | |
| "epoch": 0.916030534351145, | |
| "grad_norm": 0.5579109787940979, | |
| "learning_rate": 4.679106587503534e-06, | |
| "loss": 0.0042, | |
| "step": 3600 | |
| }, | |
| { | |
| "epoch": 0.9185750636132316, | |
| "grad_norm": 0.0032702668104320765, | |
| "learning_rate": 4.537743850720951e-06, | |
| "loss": 0.0005, | |
| "step": 3610 | |
| }, | |
| { | |
| "epoch": 0.9211195928753181, | |
| "grad_norm": 0.012057729065418243, | |
| "learning_rate": 4.3963811139383655e-06, | |
| "loss": 0.0091, | |
| "step": 3620 | |
| }, | |
| { | |
| "epoch": 0.9236641221374046, | |
| "grad_norm": 0.04750519245862961, | |
| "learning_rate": 4.255018377155782e-06, | |
| "loss": 0.0094, | |
| "step": 3630 | |
| }, | |
| { | |
| "epoch": 0.926208651399491, | |
| "grad_norm": 0.015179237350821495, | |
| "learning_rate": 4.113655640373198e-06, | |
| "loss": 0.0079, | |
| "step": 3640 | |
| }, | |
| { | |
| "epoch": 0.9287531806615776, | |
| "grad_norm": 0.01407552883028984, | |
| "learning_rate": 3.9722929035906135e-06, | |
| "loss": 0.0048, | |
| "step": 3650 | |
| }, | |
| { | |
| "epoch": 0.9312977099236641, | |
| "grad_norm": 0.009867013432085514, | |
| "learning_rate": 3.83093016680803e-06, | |
| "loss": 0.0054, | |
| "step": 3660 | |
| }, | |
| { | |
| "epoch": 0.9338422391857506, | |
| "grad_norm": 0.06089715287089348, | |
| "learning_rate": 3.6895674300254453e-06, | |
| "loss": 0.0013, | |
| "step": 3670 | |
| }, | |
| { | |
| "epoch": 0.9363867684478372, | |
| "grad_norm": 0.03142721951007843, | |
| "learning_rate": 3.548204693242861e-06, | |
| "loss": 0.0026, | |
| "step": 3680 | |
| }, | |
| { | |
| "epoch": 0.9389312977099237, | |
| "grad_norm": 0.16199061274528503, | |
| "learning_rate": 3.406841956460277e-06, | |
| "loss": 0.0057, | |
| "step": 3690 | |
| }, | |
| { | |
| "epoch": 0.9414758269720102, | |
| "grad_norm": 0.1083199754357338, | |
| "learning_rate": 3.2654792196776933e-06, | |
| "loss": 0.0085, | |
| "step": 3700 | |
| }, | |
| { | |
| "epoch": 0.9440203562340967, | |
| "grad_norm": 0.006105615291744471, | |
| "learning_rate": 3.124116482895109e-06, | |
| "loss": 0.0027, | |
| "step": 3710 | |
| }, | |
| { | |
| "epoch": 0.9465648854961832, | |
| "grad_norm": 0.14423967897891998, | |
| "learning_rate": 2.9827537461125247e-06, | |
| "loss": 0.0053, | |
| "step": 3720 | |
| }, | |
| { | |
| "epoch": 0.9491094147582697, | |
| "grad_norm": 0.0012219453928992152, | |
| "learning_rate": 2.841391009329941e-06, | |
| "loss": 0.0098, | |
| "step": 3730 | |
| }, | |
| { | |
| "epoch": 0.9516539440203562, | |
| "grad_norm": 0.261630654335022, | |
| "learning_rate": 2.7000282725473566e-06, | |
| "loss": 0.0043, | |
| "step": 3740 | |
| }, | |
| { | |
| "epoch": 0.9541984732824428, | |
| "grad_norm": 0.056524503976106644, | |
| "learning_rate": 2.5586655357647727e-06, | |
| "loss": 0.0024, | |
| "step": 3750 | |
| }, | |
| { | |
| "epoch": 0.9567430025445293, | |
| "grad_norm": 0.02202003449201584, | |
| "learning_rate": 2.4173027989821884e-06, | |
| "loss": 0.0055, | |
| "step": 3760 | |
| }, | |
| { | |
| "epoch": 0.9592875318066157, | |
| "grad_norm": 0.03880184143781662, | |
| "learning_rate": 2.275940062199604e-06, | |
| "loss": 0.0062, | |
| "step": 3770 | |
| }, | |
| { | |
| "epoch": 0.9618320610687023, | |
| "grad_norm": 0.06771141290664673, | |
| "learning_rate": 2.1345773254170202e-06, | |
| "loss": 0.0073, | |
| "step": 3780 | |
| }, | |
| { | |
| "epoch": 0.9643765903307888, | |
| "grad_norm": 0.10680809617042542, | |
| "learning_rate": 1.9932145886344364e-06, | |
| "loss": 0.004, | |
| "step": 3790 | |
| }, | |
| { | |
| "epoch": 0.9669211195928753, | |
| "grad_norm": 0.02957719936966896, | |
| "learning_rate": 1.8518518518518519e-06, | |
| "loss": 0.0044, | |
| "step": 3800 | |
| }, | |
| { | |
| "epoch": 0.9694656488549618, | |
| "grad_norm": 0.29571524262428284, | |
| "learning_rate": 1.7104891150692678e-06, | |
| "loss": 0.0099, | |
| "step": 3810 | |
| }, | |
| { | |
| "epoch": 0.9720101781170484, | |
| "grad_norm": 0.0071874321438372135, | |
| "learning_rate": 1.569126378286684e-06, | |
| "loss": 0.0027, | |
| "step": 3820 | |
| }, | |
| { | |
| "epoch": 0.9745547073791349, | |
| "grad_norm": 0.2392864227294922, | |
| "learning_rate": 1.4277636415040994e-06, | |
| "loss": 0.0177, | |
| "step": 3830 | |
| }, | |
| { | |
| "epoch": 0.9770992366412213, | |
| "grad_norm": 0.015065650455653667, | |
| "learning_rate": 1.2864009047215156e-06, | |
| "loss": 0.012, | |
| "step": 3840 | |
| }, | |
| { | |
| "epoch": 0.9796437659033079, | |
| "grad_norm": 0.32246309518814087, | |
| "learning_rate": 1.1450381679389313e-06, | |
| "loss": 0.0083, | |
| "step": 3850 | |
| }, | |
| { | |
| "epoch": 0.9821882951653944, | |
| "grad_norm": 0.005143094342201948, | |
| "learning_rate": 1.0036754311563472e-06, | |
| "loss": 0.0052, | |
| "step": 3860 | |
| }, | |
| { | |
| "epoch": 0.9847328244274809, | |
| "grad_norm": 0.08060797303915024, | |
| "learning_rate": 8.623126943737632e-07, | |
| "loss": 0.0012, | |
| "step": 3870 | |
| }, | |
| { | |
| "epoch": 0.9872773536895675, | |
| "grad_norm": 0.09097331762313843, | |
| "learning_rate": 7.20949957591179e-07, | |
| "loss": 0.0064, | |
| "step": 3880 | |
| }, | |
| { | |
| "epoch": 0.989821882951654, | |
| "grad_norm": 0.03479582816362381, | |
| "learning_rate": 5.795872208085948e-07, | |
| "loss": 0.0036, | |
| "step": 3890 | |
| }, | |
| { | |
| "epoch": 0.9923664122137404, | |
| "grad_norm": 0.00452104676514864, | |
| "learning_rate": 4.3822448402601077e-07, | |
| "loss": 0.0036, | |
| "step": 3900 | |
| }, | |
| { | |
| "epoch": 0.9949109414758269, | |
| "grad_norm": 0.12740817666053772, | |
| "learning_rate": 2.9686174724342664e-07, | |
| "loss": 0.0049, | |
| "step": 3910 | |
| }, | |
| { | |
| "epoch": 0.9974554707379135, | |
| "grad_norm": 0.08695193380117416, | |
| "learning_rate": 1.5549901046084253e-07, | |
| "loss": 0.0011, | |
| "step": 3920 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 8.531525236321613e-05, | |
| "learning_rate": 1.413627367825841e-08, | |
| "loss": 0.0044, | |
| "step": 3930 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 3930, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2.273606668827065e+18, | |
| "train_batch_size": 128, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |