| { |
| "best_metric": 3.490433931350708, |
| "best_model_checkpoint": "/scratch/cl5625/exceptions/models/100M__1208/checkpoint-30000", |
| "epoch": 3.234501347708895, |
| "eval_steps": 1000, |
| "global_step": 30000, |
| "is_hyper_param_search": false, |
| "is_local_process_zero": true, |
| "is_world_process_zero": true, |
| "log_history": [ |
| { |
| "epoch": 0.005390835579514825, |
| "grad_norm": 1.3872060775756836, |
| "learning_rate": 0.0003, |
| "loss": 8.7938, |
| "step": 50 |
| }, |
| { |
| "epoch": 0.01078167115902965, |
| "grad_norm": 3.891347885131836, |
| "learning_rate": 0.0006, |
| "loss": 6.992, |
| "step": 100 |
| }, |
| { |
| "epoch": 0.016172506738544475, |
| "grad_norm": 1.848716378211975, |
| "learning_rate": 0.0005996762007555315, |
| "loss": 6.5045, |
| "step": 150 |
| }, |
| { |
| "epoch": 0.0215633423180593, |
| "grad_norm": 1.3053127527236938, |
| "learning_rate": 0.000599352401511063, |
| "loss": 6.2433, |
| "step": 200 |
| }, |
| { |
| "epoch": 0.026954177897574125, |
| "grad_norm": 1.2040534019470215, |
| "learning_rate": 0.0005990286022665946, |
| "loss": 6.0719, |
| "step": 250 |
| }, |
| { |
| "epoch": 0.03234501347708895, |
| "grad_norm": 1.105197548866272, |
| "learning_rate": 0.0005987048030221263, |
| "loss": 5.9606, |
| "step": 300 |
| }, |
| { |
| "epoch": 0.03773584905660377, |
| "grad_norm": 1.502025842666626, |
| "learning_rate": 0.0005983810037776578, |
| "loss": 5.8818, |
| "step": 350 |
| }, |
| { |
| "epoch": 0.0431266846361186, |
| "grad_norm": 2.493607759475708, |
| "learning_rate": 0.0005980572045331894, |
| "loss": 5.8101, |
| "step": 400 |
| }, |
| { |
| "epoch": 0.04851752021563342, |
| "grad_norm": 1.3217182159423828, |
| "learning_rate": 0.0005977334052887209, |
| "loss": 5.7175, |
| "step": 450 |
| }, |
| { |
| "epoch": 0.05390835579514825, |
| "grad_norm": 1.215240240097046, |
| "learning_rate": 0.0005974096060442526, |
| "loss": 5.6639, |
| "step": 500 |
| }, |
| { |
| "epoch": 0.05929919137466307, |
| "grad_norm": 1.3821635246276855, |
| "learning_rate": 0.0005970858067997841, |
| "loss": 5.5744, |
| "step": 550 |
| }, |
| { |
| "epoch": 0.0646900269541779, |
| "grad_norm": 1.378275752067566, |
| "learning_rate": 0.0005967620075553157, |
| "loss": 5.5287, |
| "step": 600 |
| }, |
| { |
| "epoch": 0.07008086253369272, |
| "grad_norm": 1.1444066762924194, |
| "learning_rate": 0.0005964382083108472, |
| "loss": 5.4332, |
| "step": 650 |
| }, |
| { |
| "epoch": 0.07547169811320754, |
| "grad_norm": 1.657812476158142, |
| "learning_rate": 0.0005961144090663788, |
| "loss": 5.4098, |
| "step": 700 |
| }, |
| { |
| "epoch": 0.08086253369272237, |
| "grad_norm": 1.3864067792892456, |
| "learning_rate": 0.0005957906098219104, |
| "loss": 5.3168, |
| "step": 750 |
| }, |
| { |
| "epoch": 0.0862533692722372, |
| "grad_norm": 1.0762931108474731, |
| "learning_rate": 0.0005954668105774419, |
| "loss": 5.2678, |
| "step": 800 |
| }, |
| { |
| "epoch": 0.09164420485175202, |
| "grad_norm": 1.2499072551727295, |
| "learning_rate": 0.0005951430113329735, |
| "loss": 5.2107, |
| "step": 850 |
| }, |
| { |
| "epoch": 0.09703504043126684, |
| "grad_norm": 0.9282752275466919, |
| "learning_rate": 0.0005948192120885051, |
| "loss": 5.1849, |
| "step": 900 |
| }, |
| { |
| "epoch": 0.10242587601078167, |
| "grad_norm": 1.071282982826233, |
| "learning_rate": 0.0005944954128440366, |
| "loss": 5.1275, |
| "step": 950 |
| }, |
| { |
| "epoch": 0.1078167115902965, |
| "grad_norm": 1.3316810131072998, |
| "learning_rate": 0.0005941716135995682, |
| "loss": 5.1151, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.1078167115902965, |
| "eval_accuracy": 0.22672739349001877, |
| "eval_loss": 5.027144432067871, |
| "eval_runtime": 184.654, |
| "eval_samples_per_second": 97.539, |
| "eval_steps_per_second": 6.098, |
| "step": 1000 |
| }, |
| { |
| "epoch": 0.11320754716981132, |
| "grad_norm": 1.1736186742782593, |
| "learning_rate": 0.0005938478143550997, |
| "loss": 5.0742, |
| "step": 1050 |
| }, |
| { |
| "epoch": 0.11859838274932614, |
| "grad_norm": 1.217553973197937, |
| "learning_rate": 0.0005935240151106314, |
| "loss": 5.0002, |
| "step": 1100 |
| }, |
| { |
| "epoch": 0.12398921832884097, |
| "grad_norm": 1.3116732835769653, |
| "learning_rate": 0.0005932002158661629, |
| "loss": 4.9959, |
| "step": 1150 |
| }, |
| { |
| "epoch": 0.1293800539083558, |
| "grad_norm": 1.10372793674469, |
| "learning_rate": 0.0005928764166216945, |
| "loss": 4.9613, |
| "step": 1200 |
| }, |
| { |
| "epoch": 0.1347708894878706, |
| "grad_norm": 1.0251359939575195, |
| "learning_rate": 0.000592552617377226, |
| "loss": 4.9081, |
| "step": 1250 |
| }, |
| { |
| "epoch": 0.14016172506738545, |
| "grad_norm": 1.1212270259857178, |
| "learning_rate": 0.0005922288181327577, |
| "loss": 4.9098, |
| "step": 1300 |
| }, |
| { |
| "epoch": 0.14555256064690028, |
| "grad_norm": 1.0358482599258423, |
| "learning_rate": 0.0005919050188882893, |
| "loss": 4.8835, |
| "step": 1350 |
| }, |
| { |
| "epoch": 0.1509433962264151, |
| "grad_norm": 1.127200722694397, |
| "learning_rate": 0.0005915812196438207, |
| "loss": 4.85, |
| "step": 1400 |
| }, |
| { |
| "epoch": 0.15633423180592992, |
| "grad_norm": 0.9541453123092651, |
| "learning_rate": 0.0005912574203993524, |
| "loss": 4.8199, |
| "step": 1450 |
| }, |
| { |
| "epoch": 0.16172506738544473, |
| "grad_norm": 0.9905341863632202, |
| "learning_rate": 0.0005909336211548839, |
| "loss": 4.812, |
| "step": 1500 |
| }, |
| { |
| "epoch": 0.16711590296495957, |
| "grad_norm": 0.9883430600166321, |
| "learning_rate": 0.0005906098219104155, |
| "loss": 4.7776, |
| "step": 1550 |
| }, |
| { |
| "epoch": 0.1725067385444744, |
| "grad_norm": 1.14104425907135, |
| "learning_rate": 0.000590286022665947, |
| "loss": 4.7877, |
| "step": 1600 |
| }, |
| { |
| "epoch": 0.1778975741239892, |
| "grad_norm": 0.915554404258728, |
| "learning_rate": 0.0005899622234214787, |
| "loss": 4.7362, |
| "step": 1650 |
| }, |
| { |
| "epoch": 0.18328840970350405, |
| "grad_norm": 0.8896968364715576, |
| "learning_rate": 0.0005896384241770102, |
| "loss": 4.6944, |
| "step": 1700 |
| }, |
| { |
| "epoch": 0.18867924528301888, |
| "grad_norm": 0.7295605540275574, |
| "learning_rate": 0.0005893146249325418, |
| "loss": 4.679, |
| "step": 1750 |
| }, |
| { |
| "epoch": 0.1940700808625337, |
| "grad_norm": 0.8035542368888855, |
| "learning_rate": 0.0005889908256880733, |
| "loss": 4.669, |
| "step": 1800 |
| }, |
| { |
| "epoch": 0.19946091644204852, |
| "grad_norm": 0.9095346927642822, |
| "learning_rate": 0.0005886670264436049, |
| "loss": 4.6768, |
| "step": 1850 |
| }, |
| { |
| "epoch": 0.20485175202156333, |
| "grad_norm": 0.8672581315040588, |
| "learning_rate": 0.0005883432271991365, |
| "loss": 4.6219, |
| "step": 1900 |
| }, |
| { |
| "epoch": 0.21024258760107817, |
| "grad_norm": 0.7349112033843994, |
| "learning_rate": 0.0005880194279546681, |
| "loss": 4.6077, |
| "step": 1950 |
| }, |
| { |
| "epoch": 0.215633423180593, |
| "grad_norm": 0.9716205596923828, |
| "learning_rate": 0.0005876956287101996, |
| "loss": 4.5792, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.215633423180593, |
| "eval_accuracy": 0.27099464439050674, |
| "eval_loss": 4.504695415496826, |
| "eval_runtime": 184.1142, |
| "eval_samples_per_second": 97.825, |
| "eval_steps_per_second": 6.116, |
| "step": 2000 |
| }, |
| { |
| "epoch": 0.2210242587601078, |
| "grad_norm": 1.203096628189087, |
| "learning_rate": 0.0005873718294657312, |
| "loss": 4.5695, |
| "step": 2050 |
| }, |
| { |
| "epoch": 0.22641509433962265, |
| "grad_norm": 0.9952515959739685, |
| "learning_rate": 0.0005870480302212628, |
| "loss": 4.5412, |
| "step": 2100 |
| }, |
| { |
| "epoch": 0.23180592991913745, |
| "grad_norm": 0.8814882040023804, |
| "learning_rate": 0.0005867242309767943, |
| "loss": 4.5392, |
| "step": 2150 |
| }, |
| { |
| "epoch": 0.2371967654986523, |
| "grad_norm": 1.0191090106964111, |
| "learning_rate": 0.0005864004317323259, |
| "loss": 4.5151, |
| "step": 2200 |
| }, |
| { |
| "epoch": 0.24258760107816713, |
| "grad_norm": 0.8580870628356934, |
| "learning_rate": 0.0005860766324878575, |
| "loss": 4.5014, |
| "step": 2250 |
| }, |
| { |
| "epoch": 0.24797843665768193, |
| "grad_norm": 0.8280946016311646, |
| "learning_rate": 0.000585752833243389, |
| "loss": 4.4755, |
| "step": 2300 |
| }, |
| { |
| "epoch": 0.25336927223719674, |
| "grad_norm": 0.9512032866477966, |
| "learning_rate": 0.0005854290339989206, |
| "loss": 4.4624, |
| "step": 2350 |
| }, |
| { |
| "epoch": 0.2587601078167116, |
| "grad_norm": 1.000420093536377, |
| "learning_rate": 0.0005851052347544521, |
| "loss": 4.4523, |
| "step": 2400 |
| }, |
| { |
| "epoch": 0.2641509433962264, |
| "grad_norm": 0.9130716919898987, |
| "learning_rate": 0.0005847814355099838, |
| "loss": 4.4414, |
| "step": 2450 |
| }, |
| { |
| "epoch": 0.2695417789757412, |
| "grad_norm": 0.8131120800971985, |
| "learning_rate": 0.0005844576362655154, |
| "loss": 4.4556, |
| "step": 2500 |
| }, |
| { |
| "epoch": 0.2749326145552561, |
| "grad_norm": 0.9587961435317993, |
| "learning_rate": 0.0005841338370210469, |
| "loss": 4.4083, |
| "step": 2550 |
| }, |
| { |
| "epoch": 0.2803234501347709, |
| "grad_norm": 0.846928060054779, |
| "learning_rate": 0.0005838100377765785, |
| "loss": 4.3825, |
| "step": 2600 |
| }, |
| { |
| "epoch": 0.2857142857142857, |
| "grad_norm": 0.9020141363143921, |
| "learning_rate": 0.0005834862385321101, |
| "loss": 4.3869, |
| "step": 2650 |
| }, |
| { |
| "epoch": 0.29110512129380056, |
| "grad_norm": 1.0460323095321655, |
| "learning_rate": 0.0005831624392876417, |
| "loss": 4.381, |
| "step": 2700 |
| }, |
| { |
| "epoch": 0.29649595687331537, |
| "grad_norm": 0.8714868426322937, |
| "learning_rate": 0.0005828386400431731, |
| "loss": 4.3618, |
| "step": 2750 |
| }, |
| { |
| "epoch": 0.3018867924528302, |
| "grad_norm": 0.7531731724739075, |
| "learning_rate": 0.0005825148407987048, |
| "loss": 4.361, |
| "step": 2800 |
| }, |
| { |
| "epoch": 0.30727762803234504, |
| "grad_norm": 0.7624960541725159, |
| "learning_rate": 0.0005821910415542363, |
| "loss": 4.3642, |
| "step": 2850 |
| }, |
| { |
| "epoch": 0.31266846361185985, |
| "grad_norm": 0.8426908850669861, |
| "learning_rate": 0.0005818672423097679, |
| "loss": 4.3367, |
| "step": 2900 |
| }, |
| { |
| "epoch": 0.31805929919137466, |
| "grad_norm": 0.9197669625282288, |
| "learning_rate": 0.0005815434430652994, |
| "loss": 4.3496, |
| "step": 2950 |
| }, |
| { |
| "epoch": 0.32345013477088946, |
| "grad_norm": 0.8955463171005249, |
| "learning_rate": 0.0005812196438208311, |
| "loss": 4.2988, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.32345013477088946, |
| "eval_accuracy": 0.2984499253065723, |
| "eval_loss": 4.238968372344971, |
| "eval_runtime": 185.178, |
| "eval_samples_per_second": 97.263, |
| "eval_steps_per_second": 6.081, |
| "step": 3000 |
| }, |
| { |
| "epoch": 0.3288409703504043, |
| "grad_norm": 0.7810043096542358, |
| "learning_rate": 0.0005808958445763626, |
| "loss": 4.3082, |
| "step": 3050 |
| }, |
| { |
| "epoch": 0.33423180592991913, |
| "grad_norm": 0.7354643940925598, |
| "learning_rate": 0.0005805720453318942, |
| "loss": 4.2858, |
| "step": 3100 |
| }, |
| { |
| "epoch": 0.33962264150943394, |
| "grad_norm": 0.7374374270439148, |
| "learning_rate": 0.0005802482460874257, |
| "loss": 4.2772, |
| "step": 3150 |
| }, |
| { |
| "epoch": 0.3450134770889488, |
| "grad_norm": 0.7399438619613647, |
| "learning_rate": 0.0005799244468429573, |
| "loss": 4.2993, |
| "step": 3200 |
| }, |
| { |
| "epoch": 0.3504043126684636, |
| "grad_norm": 0.7688404321670532, |
| "learning_rate": 0.0005796006475984889, |
| "loss": 4.2841, |
| "step": 3250 |
| }, |
| { |
| "epoch": 0.3557951482479784, |
| "grad_norm": 0.7221906781196594, |
| "learning_rate": 0.0005792768483540205, |
| "loss": 4.2667, |
| "step": 3300 |
| }, |
| { |
| "epoch": 0.3611859838274933, |
| "grad_norm": 0.7419180274009705, |
| "learning_rate": 0.000578953049109552, |
| "loss": 4.2709, |
| "step": 3350 |
| }, |
| { |
| "epoch": 0.3665768194070081, |
| "grad_norm": 0.7906516790390015, |
| "learning_rate": 0.0005786292498650836, |
| "loss": 4.2394, |
| "step": 3400 |
| }, |
| { |
| "epoch": 0.3719676549865229, |
| "grad_norm": 0.7784256339073181, |
| "learning_rate": 0.0005783054506206152, |
| "loss": 4.2519, |
| "step": 3450 |
| }, |
| { |
| "epoch": 0.37735849056603776, |
| "grad_norm": 0.7829201221466064, |
| "learning_rate": 0.0005779816513761467, |
| "loss": 4.2314, |
| "step": 3500 |
| }, |
| { |
| "epoch": 0.38274932614555257, |
| "grad_norm": 0.8310431241989136, |
| "learning_rate": 0.0005776578521316782, |
| "loss": 4.2097, |
| "step": 3550 |
| }, |
| { |
| "epoch": 0.3881401617250674, |
| "grad_norm": 0.8430302739143372, |
| "learning_rate": 0.0005773340528872099, |
| "loss": 4.2306, |
| "step": 3600 |
| }, |
| { |
| "epoch": 0.3935309973045822, |
| "grad_norm": 0.733945906162262, |
| "learning_rate": 0.0005770102536427414, |
| "loss": 4.2221, |
| "step": 3650 |
| }, |
| { |
| "epoch": 0.39892183288409705, |
| "grad_norm": 0.6383342742919922, |
| "learning_rate": 0.000576686454398273, |
| "loss": 4.2028, |
| "step": 3700 |
| }, |
| { |
| "epoch": 0.40431266846361186, |
| "grad_norm": 0.6672252416610718, |
| "learning_rate": 0.0005763626551538045, |
| "loss": 4.1801, |
| "step": 3750 |
| }, |
| { |
| "epoch": 0.40970350404312667, |
| "grad_norm": 0.7545523643493652, |
| "learning_rate": 0.0005760388559093362, |
| "loss": 4.1881, |
| "step": 3800 |
| }, |
| { |
| "epoch": 0.41509433962264153, |
| "grad_norm": 0.7340240478515625, |
| "learning_rate": 0.0005757150566648678, |
| "loss": 4.171, |
| "step": 3850 |
| }, |
| { |
| "epoch": 0.42048517520215634, |
| "grad_norm": 0.7903727889060974, |
| "learning_rate": 0.0005753912574203993, |
| "loss": 4.1784, |
| "step": 3900 |
| }, |
| { |
| "epoch": 0.42587601078167114, |
| "grad_norm": 0.8018081188201904, |
| "learning_rate": 0.0005750674581759309, |
| "loss": 4.1838, |
| "step": 3950 |
| }, |
| { |
| "epoch": 0.431266846361186, |
| "grad_norm": 0.722425103187561, |
| "learning_rate": 0.0005747436589314624, |
| "loss": 4.1617, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.431266846361186, |
| "eval_accuracy": 0.31149631215800827, |
| "eval_loss": 4.093442440032959, |
| "eval_runtime": 185.3481, |
| "eval_samples_per_second": 97.174, |
| "eval_steps_per_second": 6.075, |
| "step": 4000 |
| }, |
| { |
| "epoch": 0.4366576819407008, |
| "grad_norm": 0.9414203763008118, |
| "learning_rate": 0.0005744198596869941, |
| "loss": 4.15, |
| "step": 4050 |
| }, |
| { |
| "epoch": 0.4420485175202156, |
| "grad_norm": 0.7414451241493225, |
| "learning_rate": 0.0005740960604425255, |
| "loss": 4.1367, |
| "step": 4100 |
| }, |
| { |
| "epoch": 0.4474393530997305, |
| "grad_norm": 0.693946361541748, |
| "learning_rate": 0.0005737722611980572, |
| "loss": 4.1503, |
| "step": 4150 |
| }, |
| { |
| "epoch": 0.4528301886792453, |
| "grad_norm": 0.8669963479042053, |
| "learning_rate": 0.0005734484619535887, |
| "loss": 4.1342, |
| "step": 4200 |
| }, |
| { |
| "epoch": 0.4582210242587601, |
| "grad_norm": 0.717522144317627, |
| "learning_rate": 0.0005731246627091203, |
| "loss": 4.1361, |
| "step": 4250 |
| }, |
| { |
| "epoch": 0.4636118598382749, |
| "grad_norm": 0.5836500525474548, |
| "learning_rate": 0.0005728008634646518, |
| "loss": 4.1184, |
| "step": 4300 |
| }, |
| { |
| "epoch": 0.46900269541778977, |
| "grad_norm": 0.6450395584106445, |
| "learning_rate": 0.0005724770642201835, |
| "loss": 4.1149, |
| "step": 4350 |
| }, |
| { |
| "epoch": 0.4743935309973046, |
| "grad_norm": 0.6113871335983276, |
| "learning_rate": 0.000572153264975715, |
| "loss": 4.1102, |
| "step": 4400 |
| }, |
| { |
| "epoch": 0.4797843665768194, |
| "grad_norm": 0.7142659425735474, |
| "learning_rate": 0.0005718294657312466, |
| "loss": 4.1382, |
| "step": 4450 |
| }, |
| { |
| "epoch": 0.48517520215633425, |
| "grad_norm": 0.6929596066474915, |
| "learning_rate": 0.0005715056664867781, |
| "loss": 4.0957, |
| "step": 4500 |
| }, |
| { |
| "epoch": 0.49056603773584906, |
| "grad_norm": 0.5346873998641968, |
| "learning_rate": 0.0005711818672423097, |
| "loss": 4.1046, |
| "step": 4550 |
| }, |
| { |
| "epoch": 0.49595687331536387, |
| "grad_norm": 0.702147901058197, |
| "learning_rate": 0.0005708580679978413, |
| "loss": 4.1, |
| "step": 4600 |
| }, |
| { |
| "epoch": 0.5013477088948787, |
| "grad_norm": 0.5867359042167664, |
| "learning_rate": 0.0005705342687533729, |
| "loss": 4.0953, |
| "step": 4650 |
| }, |
| { |
| "epoch": 0.5067385444743935, |
| "grad_norm": 0.6331294178962708, |
| "learning_rate": 0.0005702104695089044, |
| "loss": 4.0979, |
| "step": 4700 |
| }, |
| { |
| "epoch": 0.5121293800539084, |
| "grad_norm": 0.6199395656585693, |
| "learning_rate": 0.000569886670264436, |
| "loss": 4.0987, |
| "step": 4750 |
| }, |
| { |
| "epoch": 0.5175202156334232, |
| "grad_norm": 0.6783130764961243, |
| "learning_rate": 0.0005695628710199675, |
| "loss": 4.0776, |
| "step": 4800 |
| }, |
| { |
| "epoch": 0.522911051212938, |
| "grad_norm": 0.6518664956092834, |
| "learning_rate": 0.0005692390717754991, |
| "loss": 4.1048, |
| "step": 4850 |
| }, |
| { |
| "epoch": 0.5283018867924528, |
| "grad_norm": 0.6022597551345825, |
| "learning_rate": 0.0005689152725310306, |
| "loss": 4.0778, |
| "step": 4900 |
| }, |
| { |
| "epoch": 0.5336927223719676, |
| "grad_norm": 0.6618348360061646, |
| "learning_rate": 0.0005685914732865623, |
| "loss": 4.0617, |
| "step": 4950 |
| }, |
| { |
| "epoch": 0.5390835579514824, |
| "grad_norm": 0.5986215472221375, |
| "learning_rate": 0.0005682676740420939, |
| "loss": 4.0655, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.5390835579514824, |
| "eval_accuracy": 0.32159864036122304, |
| "eval_loss": 3.9903128147125244, |
| "eval_runtime": 185.2705, |
| "eval_samples_per_second": 97.215, |
| "eval_steps_per_second": 6.078, |
| "step": 5000 |
| }, |
| { |
| "epoch": 0.5444743935309974, |
| "grad_norm": 0.5965196490287781, |
| "learning_rate": 0.0005679438747976254, |
| "loss": 4.0629, |
| "step": 5050 |
| }, |
| { |
| "epoch": 0.5498652291105122, |
| "grad_norm": 0.618198812007904, |
| "learning_rate": 0.000567620075553157, |
| "loss": 4.048, |
| "step": 5100 |
| }, |
| { |
| "epoch": 0.555256064690027, |
| "grad_norm": 0.6078091263771057, |
| "learning_rate": 0.0005672962763086886, |
| "loss": 4.0601, |
| "step": 5150 |
| }, |
| { |
| "epoch": 0.5606469002695418, |
| "grad_norm": 0.5713509917259216, |
| "learning_rate": 0.0005669724770642202, |
| "loss": 4.054, |
| "step": 5200 |
| }, |
| { |
| "epoch": 0.5660377358490566, |
| "grad_norm": 0.7706556916236877, |
| "learning_rate": 0.0005666486778197517, |
| "loss": 4.0499, |
| "step": 5250 |
| }, |
| { |
| "epoch": 0.5714285714285714, |
| "grad_norm": 0.6719933748245239, |
| "learning_rate": 0.0005663248785752833, |
| "loss": 4.0337, |
| "step": 5300 |
| }, |
| { |
| "epoch": 0.5768194070080862, |
| "grad_norm": 0.6685216426849365, |
| "learning_rate": 0.0005660010793308148, |
| "loss": 4.0071, |
| "step": 5350 |
| }, |
| { |
| "epoch": 0.5822102425876011, |
| "grad_norm": 0.6304830312728882, |
| "learning_rate": 0.0005656772800863465, |
| "loss": 4.0476, |
| "step": 5400 |
| }, |
| { |
| "epoch": 0.5876010781671159, |
| "grad_norm": 0.6301653385162354, |
| "learning_rate": 0.0005653534808418779, |
| "loss": 4.0282, |
| "step": 5450 |
| }, |
| { |
| "epoch": 0.5929919137466307, |
| "grad_norm": 0.6767484545707703, |
| "learning_rate": 0.0005650296815974096, |
| "loss": 4.0349, |
| "step": 5500 |
| }, |
| { |
| "epoch": 0.5983827493261455, |
| "grad_norm": 0.6710530519485474, |
| "learning_rate": 0.0005647058823529411, |
| "loss": 4.0179, |
| "step": 5550 |
| }, |
| { |
| "epoch": 0.6037735849056604, |
| "grad_norm": 0.5611121654510498, |
| "learning_rate": 0.0005643820831084727, |
| "loss": 4.0202, |
| "step": 5600 |
| }, |
| { |
| "epoch": 0.6091644204851752, |
| "grad_norm": 0.6187557578086853, |
| "learning_rate": 0.0005640582838640042, |
| "loss": 3.9935, |
| "step": 5650 |
| }, |
| { |
| "epoch": 0.6145552560646901, |
| "grad_norm": 0.6690448522567749, |
| "learning_rate": 0.0005637344846195358, |
| "loss": 4.0082, |
| "step": 5700 |
| }, |
| { |
| "epoch": 0.6199460916442049, |
| "grad_norm": 0.6771255135536194, |
| "learning_rate": 0.0005634106853750674, |
| "loss": 4.0026, |
| "step": 5750 |
| }, |
| { |
| "epoch": 0.6253369272237197, |
| "grad_norm": 0.5823601484298706, |
| "learning_rate": 0.000563086886130599, |
| "loss": 4.014, |
| "step": 5800 |
| }, |
| { |
| "epoch": 0.6307277628032345, |
| "grad_norm": 0.6157869100570679, |
| "learning_rate": 0.0005627630868861305, |
| "loss": 3.9952, |
| "step": 5850 |
| }, |
| { |
| "epoch": 0.6361185983827493, |
| "grad_norm": 0.6894795894622803, |
| "learning_rate": 0.0005624392876416621, |
| "loss": 3.9947, |
| "step": 5900 |
| }, |
| { |
| "epoch": 0.6415094339622641, |
| "grad_norm": 0.6236327290534973, |
| "learning_rate": 0.0005621154883971937, |
| "loss": 3.996, |
| "step": 5950 |
| }, |
| { |
| "epoch": 0.6469002695417789, |
| "grad_norm": 0.6344878077507019, |
| "learning_rate": 0.0005617916891527253, |
| "loss": 3.9934, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.6469002695417789, |
| "eval_accuracy": 0.3280596843351111, |
| "eval_loss": 3.924257278442383, |
| "eval_runtime": 185.4361, |
| "eval_samples_per_second": 97.128, |
| "eval_steps_per_second": 6.072, |
| "step": 6000 |
| }, |
| { |
| "epoch": 0.6522911051212938, |
| "grad_norm": 0.5802189111709595, |
| "learning_rate": 0.0005614743658931463, |
| "loss": 3.9855, |
| "step": 6050 |
| }, |
| { |
| "epoch": 0.6576819407008087, |
| "grad_norm": 0.6164999604225159, |
| "learning_rate": 0.0005611505666486777, |
| "loss": 4.0042, |
| "step": 6100 |
| }, |
| { |
| "epoch": 0.6630727762803235, |
| "grad_norm": 0.6946608424186707, |
| "learning_rate": 0.0005608267674042094, |
| "loss": 3.989, |
| "step": 6150 |
| }, |
| { |
| "epoch": 0.6684636118598383, |
| "grad_norm": 0.6158466935157776, |
| "learning_rate": 0.0005605029681597409, |
| "loss": 3.9871, |
| "step": 6200 |
| }, |
| { |
| "epoch": 0.6738544474393531, |
| "grad_norm": 0.6288596391677856, |
| "learning_rate": 0.0005601791689152725, |
| "loss": 3.9773, |
| "step": 6250 |
| }, |
| { |
| "epoch": 0.6792452830188679, |
| "grad_norm": 0.7042039036750793, |
| "learning_rate": 0.000559855369670804, |
| "loss": 3.9791, |
| "step": 6300 |
| }, |
| { |
| "epoch": 0.6846361185983828, |
| "grad_norm": 0.5582059025764465, |
| "learning_rate": 0.0005595315704263357, |
| "loss": 3.9639, |
| "step": 6350 |
| }, |
| { |
| "epoch": 0.6900269541778976, |
| "grad_norm": 0.6060189604759216, |
| "learning_rate": 0.0005592077711818672, |
| "loss": 3.9775, |
| "step": 6400 |
| }, |
| { |
| "epoch": 0.6954177897574124, |
| "grad_norm": 0.7476151585578918, |
| "learning_rate": 0.0005588839719373988, |
| "loss": 3.9536, |
| "step": 6450 |
| }, |
| { |
| "epoch": 0.7008086253369272, |
| "grad_norm": 0.6221930384635925, |
| "learning_rate": 0.0005585601726929303, |
| "loss": 3.9583, |
| "step": 6500 |
| }, |
| { |
| "epoch": 0.706199460916442, |
| "grad_norm": 0.6130439639091492, |
| "learning_rate": 0.0005582363734484619, |
| "loss": 3.9417, |
| "step": 6550 |
| }, |
| { |
| "epoch": 0.7115902964959568, |
| "grad_norm": 0.5701190233230591, |
| "learning_rate": 0.0005579125742039935, |
| "loss": 3.9488, |
| "step": 6600 |
| }, |
| { |
| "epoch": 0.7169811320754716, |
| "grad_norm": 0.5955173969268799, |
| "learning_rate": 0.0005575887749595251, |
| "loss": 3.9511, |
| "step": 6650 |
| }, |
| { |
| "epoch": 0.7223719676549866, |
| "grad_norm": 0.6578053832054138, |
| "learning_rate": 0.0005572649757150566, |
| "loss": 3.9526, |
| "step": 6700 |
| }, |
| { |
| "epoch": 0.7277628032345014, |
| "grad_norm": 0.7318617105484009, |
| "learning_rate": 0.0005569411764705882, |
| "loss": 3.9511, |
| "step": 6750 |
| }, |
| { |
| "epoch": 0.7331536388140162, |
| "grad_norm": 0.6341660618782043, |
| "learning_rate": 0.0005566173772261198, |
| "loss": 3.918, |
| "step": 6800 |
| }, |
| { |
| "epoch": 0.738544474393531, |
| "grad_norm": 0.5913158059120178, |
| "learning_rate": 0.0005562935779816513, |
| "loss": 3.9336, |
| "step": 6850 |
| }, |
| { |
| "epoch": 0.7439353099730458, |
| "grad_norm": 0.6217501759529114, |
| "learning_rate": 0.0005559697787371828, |
| "loss": 3.9485, |
| "step": 6900 |
| }, |
| { |
| "epoch": 0.7493261455525606, |
| "grad_norm": 0.629990816116333, |
| "learning_rate": 0.0005556459794927145, |
| "loss": 3.9304, |
| "step": 6950 |
| }, |
| { |
| "epoch": 0.7547169811320755, |
| "grad_norm": 0.598209798336029, |
| "learning_rate": 0.000555322180248246, |
| "loss": 3.9432, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.7547169811320755, |
| "eval_accuracy": 0.33339541034990466, |
| "eval_loss": 3.867119312286377, |
| "eval_runtime": 185.4718, |
| "eval_samples_per_second": 97.109, |
| "eval_steps_per_second": 6.071, |
| "step": 7000 |
| }, |
| { |
| "epoch": 0.7601078167115903, |
| "grad_norm": 0.7586105465888977, |
| "learning_rate": 0.0005549983810037776, |
| "loss": 3.911, |
| "step": 7050 |
| }, |
| { |
| "epoch": 0.7654986522911051, |
| "grad_norm": 0.6012836694717407, |
| "learning_rate": 0.0005546745817593091, |
| "loss": 3.9126, |
| "step": 7100 |
| }, |
| { |
| "epoch": 0.77088948787062, |
| "grad_norm": 0.6788042187690735, |
| "learning_rate": 0.0005543507825148408, |
| "loss": 3.9149, |
| "step": 7150 |
| }, |
| { |
| "epoch": 0.7762803234501348, |
| "grad_norm": 0.60019451379776, |
| "learning_rate": 0.0005540269832703723, |
| "loss": 3.8948, |
| "step": 7200 |
| }, |
| { |
| "epoch": 0.7816711590296496, |
| "grad_norm": 0.5468128323554993, |
| "learning_rate": 0.0005537031840259039, |
| "loss": 3.9228, |
| "step": 7250 |
| }, |
| { |
| "epoch": 0.7870619946091644, |
| "grad_norm": 0.560189425945282, |
| "learning_rate": 0.0005533793847814354, |
| "loss": 3.8974, |
| "step": 7300 |
| }, |
| { |
| "epoch": 0.7924528301886793, |
| "grad_norm": 0.6238011717796326, |
| "learning_rate": 0.000553055585536967, |
| "loss": 3.9169, |
| "step": 7350 |
| }, |
| { |
| "epoch": 0.7978436657681941, |
| "grad_norm": 0.6309143304824829, |
| "learning_rate": 0.0005527317862924987, |
| "loss": 3.9167, |
| "step": 7400 |
| }, |
| { |
| "epoch": 0.8032345013477089, |
| "grad_norm": 0.5579202771186829, |
| "learning_rate": 0.0005524079870480301, |
| "loss": 3.9102, |
| "step": 7450 |
| }, |
| { |
| "epoch": 0.8086253369272237, |
| "grad_norm": 0.637144923210144, |
| "learning_rate": 0.0005520841878035618, |
| "loss": 3.9305, |
| "step": 7500 |
| }, |
| { |
| "epoch": 0.8140161725067385, |
| "grad_norm": 0.560736358165741, |
| "learning_rate": 0.0005517603885590933, |
| "loss": 3.9012, |
| "step": 7550 |
| }, |
| { |
| "epoch": 0.8194070080862533, |
| "grad_norm": 0.5624358057975769, |
| "learning_rate": 0.0005514365893146249, |
| "loss": 3.8973, |
| "step": 7600 |
| }, |
| { |
| "epoch": 0.8247978436657682, |
| "grad_norm": 0.5344393253326416, |
| "learning_rate": 0.0005511127900701564, |
| "loss": 3.8894, |
| "step": 7650 |
| }, |
| { |
| "epoch": 0.8301886792452831, |
| "grad_norm": 0.5487794280052185, |
| "learning_rate": 0.000550788990825688, |
| "loss": 3.9173, |
| "step": 7700 |
| }, |
| { |
| "epoch": 0.8355795148247979, |
| "grad_norm": 0.5416852831840515, |
| "learning_rate": 0.0005504651915812196, |
| "loss": 3.9131, |
| "step": 7750 |
| }, |
| { |
| "epoch": 0.8409703504043127, |
| "grad_norm": 0.6219443678855896, |
| "learning_rate": 0.0005501413923367512, |
| "loss": 3.8783, |
| "step": 7800 |
| }, |
| { |
| "epoch": 0.8463611859838275, |
| "grad_norm": 0.5484825968742371, |
| "learning_rate": 0.0005498175930922827, |
| "loss": 3.8889, |
| "step": 7850 |
| }, |
| { |
| "epoch": 0.8517520215633423, |
| "grad_norm": 0.5506182909011841, |
| "learning_rate": 0.0005494937938478143, |
| "loss": 3.8945, |
| "step": 7900 |
| }, |
| { |
| "epoch": 0.8571428571428571, |
| "grad_norm": 0.5658506155014038, |
| "learning_rate": 0.0005491699946033459, |
| "loss": 3.883, |
| "step": 7950 |
| }, |
| { |
| "epoch": 0.862533692722372, |
| "grad_norm": 0.5695346593856812, |
| "learning_rate": 0.0005488461953588775, |
| "loss": 3.8869, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.862533692722372, |
| "eval_accuracy": 0.33750498852571176, |
| "eval_loss": 3.818067789077759, |
| "eval_runtime": 185.1439, |
| "eval_samples_per_second": 97.281, |
| "eval_steps_per_second": 6.082, |
| "step": 8000 |
| }, |
| { |
| "epoch": 0.8679245283018868, |
| "grad_norm": 0.5362111926078796, |
| "learning_rate": 0.000548522396114409, |
| "loss": 3.8853, |
| "step": 8050 |
| }, |
| { |
| "epoch": 0.8733153638814016, |
| "grad_norm": 0.6255955696105957, |
| "learning_rate": 0.0005482050728548299, |
| "loss": 3.8955, |
| "step": 8100 |
| }, |
| { |
| "epoch": 0.8787061994609164, |
| "grad_norm": 0.4989166557788849, |
| "learning_rate": 0.0005478812736103615, |
| "loss": 3.8747, |
| "step": 8150 |
| }, |
| { |
| "epoch": 0.8840970350404312, |
| "grad_norm": 0.5338106751441956, |
| "learning_rate": 0.0005475574743658931, |
| "loss": 3.8772, |
| "step": 8200 |
| }, |
| { |
| "epoch": 0.889487870619946, |
| "grad_norm": 0.6587180495262146, |
| "learning_rate": 0.0005472336751214246, |
| "loss": 3.8723, |
| "step": 8250 |
| }, |
| { |
| "epoch": 0.894878706199461, |
| "grad_norm": 0.6544439196586609, |
| "learning_rate": 0.0005469098758769562, |
| "loss": 3.856, |
| "step": 8300 |
| }, |
| { |
| "epoch": 0.9002695417789758, |
| "grad_norm": 0.5896086692810059, |
| "learning_rate": 0.0005465860766324878, |
| "loss": 3.8439, |
| "step": 8350 |
| }, |
| { |
| "epoch": 0.9056603773584906, |
| "grad_norm": 0.579795241355896, |
| "learning_rate": 0.0005462622773880194, |
| "loss": 3.892, |
| "step": 8400 |
| }, |
| { |
| "epoch": 0.9110512129380054, |
| "grad_norm": 0.6916389465332031, |
| "learning_rate": 0.000545938478143551, |
| "loss": 3.8569, |
| "step": 8450 |
| }, |
| { |
| "epoch": 0.9164420485175202, |
| "grad_norm": 0.5650646686553955, |
| "learning_rate": 0.0005456146788990825, |
| "loss": 3.8475, |
| "step": 8500 |
| }, |
| { |
| "epoch": 0.921832884097035, |
| "grad_norm": 0.5492226481437683, |
| "learning_rate": 0.000545290879654614, |
| "loss": 3.86, |
| "step": 8550 |
| }, |
| { |
| "epoch": 0.9272237196765498, |
| "grad_norm": 0.6170901656150818, |
| "learning_rate": 0.0005449670804101457, |
| "loss": 3.86, |
| "step": 8600 |
| }, |
| { |
| "epoch": 0.9326145552560647, |
| "grad_norm": 0.605499804019928, |
| "learning_rate": 0.0005446432811656773, |
| "loss": 3.8616, |
| "step": 8650 |
| }, |
| { |
| "epoch": 0.9380053908355795, |
| "grad_norm": 0.5699788331985474, |
| "learning_rate": 0.0005443194819212088, |
| "loss": 3.8398, |
| "step": 8700 |
| }, |
| { |
| "epoch": 0.9433962264150944, |
| "grad_norm": 0.6459490656852722, |
| "learning_rate": 0.0005439956826767404, |
| "loss": 3.8561, |
| "step": 8750 |
| }, |
| { |
| "epoch": 0.9487870619946092, |
| "grad_norm": 0.5878280401229858, |
| "learning_rate": 0.000543671883432272, |
| "loss": 3.8437, |
| "step": 8800 |
| }, |
| { |
| "epoch": 0.954177897574124, |
| "grad_norm": 0.5819820165634155, |
| "learning_rate": 0.0005433480841878035, |
| "loss": 3.8392, |
| "step": 8850 |
| }, |
| { |
| "epoch": 0.9595687331536388, |
| "grad_norm": 0.6361739635467529, |
| "learning_rate": 0.000543024284943335, |
| "loss": 3.8312, |
| "step": 8900 |
| }, |
| { |
| "epoch": 0.9649595687331537, |
| "grad_norm": 0.6311874389648438, |
| "learning_rate": 0.0005427004856988667, |
| "loss": 3.8481, |
| "step": 8950 |
| }, |
| { |
| "epoch": 0.9703504043126685, |
| "grad_norm": 0.6704027056694031, |
| "learning_rate": 0.0005423766864543982, |
| "loss": 3.848, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.9703504043126685, |
| "eval_accuracy": 0.3414659295509887, |
| "eval_loss": 3.779182195663452, |
| "eval_runtime": 184.1503, |
| "eval_samples_per_second": 97.806, |
| "eval_steps_per_second": 6.115, |
| "step": 9000 |
| }, |
| { |
| "epoch": 0.9757412398921833, |
| "grad_norm": 0.5311286449432373, |
| "learning_rate": 0.0005420528872099298, |
| "loss": 3.8298, |
| "step": 9050 |
| }, |
| { |
| "epoch": 0.9811320754716981, |
| "grad_norm": 0.5362455248832703, |
| "learning_rate": 0.0005417290879654613, |
| "loss": 3.8289, |
| "step": 9100 |
| }, |
| { |
| "epoch": 0.9865229110512129, |
| "grad_norm": 0.6195308566093445, |
| "learning_rate": 0.000541405288720993, |
| "loss": 3.8465, |
| "step": 9150 |
| }, |
| { |
| "epoch": 0.9919137466307277, |
| "grad_norm": 0.5365903973579407, |
| "learning_rate": 0.0005410814894765245, |
| "loss": 3.8248, |
| "step": 9200 |
| }, |
| { |
| "epoch": 0.9973045822102425, |
| "grad_norm": 0.652845025062561, |
| "learning_rate": 0.0005407576902320561, |
| "loss": 3.8518, |
| "step": 9250 |
| }, |
| { |
| "epoch": 1.0026954177897573, |
| "grad_norm": 0.60700523853302, |
| "learning_rate": 0.0005404338909875876, |
| "loss": 3.7881, |
| "step": 9300 |
| }, |
| { |
| "epoch": 1.0080862533692723, |
| "grad_norm": 0.6487429738044739, |
| "learning_rate": 0.0005401100917431192, |
| "loss": 3.7647, |
| "step": 9350 |
| }, |
| { |
| "epoch": 1.013477088948787, |
| "grad_norm": 0.561632513999939, |
| "learning_rate": 0.0005397862924986508, |
| "loss": 3.7791, |
| "step": 9400 |
| }, |
| { |
| "epoch": 1.0188679245283019, |
| "grad_norm": 0.5459903478622437, |
| "learning_rate": 0.0005394624932541824, |
| "loss": 3.7855, |
| "step": 9450 |
| }, |
| { |
| "epoch": 1.0242587601078168, |
| "grad_norm": 0.5852130055427551, |
| "learning_rate": 0.0005391386940097139, |
| "loss": 3.7552, |
| "step": 9500 |
| }, |
| { |
| "epoch": 1.0296495956873315, |
| "grad_norm": 0.6032885909080505, |
| "learning_rate": 0.0005388148947652455, |
| "loss": 3.7671, |
| "step": 9550 |
| }, |
| { |
| "epoch": 1.0350404312668464, |
| "grad_norm": 0.6321649551391602, |
| "learning_rate": 0.000538491095520777, |
| "loss": 3.795, |
| "step": 9600 |
| }, |
| { |
| "epoch": 1.0404312668463611, |
| "grad_norm": 0.615997314453125, |
| "learning_rate": 0.0005381672962763086, |
| "loss": 3.7544, |
| "step": 9650 |
| }, |
| { |
| "epoch": 1.045822102425876, |
| "grad_norm": 0.6234314441680908, |
| "learning_rate": 0.0005378434970318403, |
| "loss": 3.7772, |
| "step": 9700 |
| }, |
| { |
| "epoch": 1.0512129380053907, |
| "grad_norm": 0.638446569442749, |
| "learning_rate": 0.0005375196977873718, |
| "loss": 3.753, |
| "step": 9750 |
| }, |
| { |
| "epoch": 1.0566037735849056, |
| "grad_norm": 0.6064075827598572, |
| "learning_rate": 0.0005371958985429034, |
| "loss": 3.765, |
| "step": 9800 |
| }, |
| { |
| "epoch": 1.0619946091644206, |
| "grad_norm": 0.572862982749939, |
| "learning_rate": 0.0005368720992984349, |
| "loss": 3.7706, |
| "step": 9850 |
| }, |
| { |
| "epoch": 1.0673854447439353, |
| "grad_norm": 0.6439206600189209, |
| "learning_rate": 0.0005365483000539665, |
| "loss": 3.7649, |
| "step": 9900 |
| }, |
| { |
| "epoch": 1.0727762803234502, |
| "grad_norm": 0.5527343153953552, |
| "learning_rate": 0.0005362245008094981, |
| "loss": 3.7804, |
| "step": 9950 |
| }, |
| { |
| "epoch": 1.0781671159029649, |
| "grad_norm": 0.5652170777320862, |
| "learning_rate": 0.0005359007015650297, |
| "loss": 3.7701, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.0781671159029649, |
| "eval_accuracy": 0.3446935722364057, |
| "eval_loss": 3.7489497661590576, |
| "eval_runtime": 184.1084, |
| "eval_samples_per_second": 97.828, |
| "eval_steps_per_second": 6.116, |
| "step": 10000 |
| }, |
| { |
| "epoch": 1.0835579514824798, |
| "grad_norm": 0.5648168325424194, |
| "learning_rate": 0.0005355769023205612, |
| "loss": 3.7698, |
| "step": 10050 |
| }, |
| { |
| "epoch": 1.0889487870619945, |
| "grad_norm": 0.6025657057762146, |
| "learning_rate": 0.0005352531030760928, |
| "loss": 3.7595, |
| "step": 10100 |
| }, |
| { |
| "epoch": 1.0943396226415094, |
| "grad_norm": 0.6286342144012451, |
| "learning_rate": 0.0005349357798165137, |
| "loss": 3.7772, |
| "step": 10150 |
| }, |
| { |
| "epoch": 1.0997304582210243, |
| "grad_norm": 0.565532386302948, |
| "learning_rate": 0.0005346119805720453, |
| "loss": 3.7646, |
| "step": 10200 |
| }, |
| { |
| "epoch": 1.105121293800539, |
| "grad_norm": 0.5853555202484131, |
| "learning_rate": 0.0005342881813275768, |
| "loss": 3.7637, |
| "step": 10250 |
| }, |
| { |
| "epoch": 1.110512129380054, |
| "grad_norm": 0.6412824988365173, |
| "learning_rate": 0.0005339708580679978, |
| "loss": 3.7642, |
| "step": 10300 |
| }, |
| { |
| "epoch": 1.1159029649595686, |
| "grad_norm": 0.5646089315414429, |
| "learning_rate": 0.0005336470588235293, |
| "loss": 3.7553, |
| "step": 10350 |
| }, |
| { |
| "epoch": 1.1212938005390836, |
| "grad_norm": 0.6126633882522583, |
| "learning_rate": 0.000533323259579061, |
| "loss": 3.7789, |
| "step": 10400 |
| }, |
| { |
| "epoch": 1.1266846361185983, |
| "grad_norm": 0.5688264966011047, |
| "learning_rate": 0.0005329994603345925, |
| "loss": 3.7638, |
| "step": 10450 |
| }, |
| { |
| "epoch": 1.1320754716981132, |
| "grad_norm": 0.5909513235092163, |
| "learning_rate": 0.0005326756610901241, |
| "loss": 3.7545, |
| "step": 10500 |
| }, |
| { |
| "epoch": 1.137466307277628, |
| "grad_norm": 0.6165957450866699, |
| "learning_rate": 0.0005323518618456556, |
| "loss": 3.7571, |
| "step": 10550 |
| }, |
| { |
| "epoch": 1.1428571428571428, |
| "grad_norm": 0.6512620449066162, |
| "learning_rate": 0.0005320280626011872, |
| "loss": 3.7638, |
| "step": 10600 |
| }, |
| { |
| "epoch": 1.1482479784366577, |
| "grad_norm": 0.5640995502471924, |
| "learning_rate": 0.0005317042633567188, |
| "loss": 3.7548, |
| "step": 10650 |
| }, |
| { |
| "epoch": 1.1536388140161726, |
| "grad_norm": 0.5489684343338013, |
| "learning_rate": 0.0005313804641122504, |
| "loss": 3.745, |
| "step": 10700 |
| }, |
| { |
| "epoch": 1.1590296495956873, |
| "grad_norm": 0.5724853873252869, |
| "learning_rate": 0.0005310566648677819, |
| "loss": 3.7553, |
| "step": 10750 |
| }, |
| { |
| "epoch": 1.1644204851752022, |
| "grad_norm": 0.6125072240829468, |
| "learning_rate": 0.0005307328656233135, |
| "loss": 3.7447, |
| "step": 10800 |
| }, |
| { |
| "epoch": 1.169811320754717, |
| "grad_norm": 0.551490843296051, |
| "learning_rate": 0.000530409066378845, |
| "loss": 3.7589, |
| "step": 10850 |
| }, |
| { |
| "epoch": 1.1752021563342319, |
| "grad_norm": 0.5600878000259399, |
| "learning_rate": 0.0005300852671343766, |
| "loss": 3.7586, |
| "step": 10900 |
| }, |
| { |
| "epoch": 1.1805929919137466, |
| "grad_norm": 0.5831260681152344, |
| "learning_rate": 0.0005297614678899081, |
| "loss": 3.7425, |
| "step": 10950 |
| }, |
| { |
| "epoch": 1.1859838274932615, |
| "grad_norm": 0.6004246473312378, |
| "learning_rate": 0.0005294376686454398, |
| "loss": 3.7519, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.1859838274932615, |
| "eval_accuracy": 0.3466813768189716, |
| "eval_loss": 3.723628520965576, |
| "eval_runtime": 183.7188, |
| "eval_samples_per_second": 98.036, |
| "eval_steps_per_second": 6.129, |
| "step": 11000 |
| }, |
| { |
| "epoch": 1.1913746630727764, |
| "grad_norm": 0.5393942594528198, |
| "learning_rate": 0.0005291138694009714, |
| "loss": 3.753, |
| "step": 11050 |
| }, |
| { |
| "epoch": 1.196765498652291, |
| "grad_norm": 0.5937897562980652, |
| "learning_rate": 0.0005287900701565029, |
| "loss": 3.7424, |
| "step": 11100 |
| }, |
| { |
| "epoch": 1.202156334231806, |
| "grad_norm": 0.5454901456832886, |
| "learning_rate": 0.0005284662709120345, |
| "loss": 3.7416, |
| "step": 11150 |
| }, |
| { |
| "epoch": 1.2075471698113207, |
| "grad_norm": 0.5742169618606567, |
| "learning_rate": 0.0005281424716675661, |
| "loss": 3.7442, |
| "step": 11200 |
| }, |
| { |
| "epoch": 1.2129380053908356, |
| "grad_norm": 0.5434097051620483, |
| "learning_rate": 0.0005278186724230977, |
| "loss": 3.7337, |
| "step": 11250 |
| }, |
| { |
| "epoch": 1.2183288409703503, |
| "grad_norm": 0.5956228971481323, |
| "learning_rate": 0.0005274948731786292, |
| "loss": 3.7535, |
| "step": 11300 |
| }, |
| { |
| "epoch": 1.2237196765498652, |
| "grad_norm": 0.5586687922477722, |
| "learning_rate": 0.0005271710739341608, |
| "loss": 3.7113, |
| "step": 11350 |
| }, |
| { |
| "epoch": 1.2291105121293802, |
| "grad_norm": 0.5536643266677856, |
| "learning_rate": 0.0005268472746896923, |
| "loss": 3.7524, |
| "step": 11400 |
| }, |
| { |
| "epoch": 1.2345013477088949, |
| "grad_norm": 0.5457449555397034, |
| "learning_rate": 0.000526523475445224, |
| "loss": 3.7376, |
| "step": 11450 |
| }, |
| { |
| "epoch": 1.2398921832884098, |
| "grad_norm": 0.6064246892929077, |
| "learning_rate": 0.0005261996762007554, |
| "loss": 3.7229, |
| "step": 11500 |
| }, |
| { |
| "epoch": 1.2452830188679245, |
| "grad_norm": 0.5753145217895508, |
| "learning_rate": 0.0005258758769562871, |
| "loss": 3.7512, |
| "step": 11550 |
| }, |
| { |
| "epoch": 1.2506738544474394, |
| "grad_norm": 0.6125341057777405, |
| "learning_rate": 0.0005255520777118186, |
| "loss": 3.7342, |
| "step": 11600 |
| }, |
| { |
| "epoch": 1.256064690026954, |
| "grad_norm": 0.589798629283905, |
| "learning_rate": 0.0005252282784673502, |
| "loss": 3.7213, |
| "step": 11650 |
| }, |
| { |
| "epoch": 1.261455525606469, |
| "grad_norm": 0.6119717359542847, |
| "learning_rate": 0.0005249044792228817, |
| "loss": 3.7299, |
| "step": 11700 |
| }, |
| { |
| "epoch": 1.266846361185984, |
| "grad_norm": 0.6549661159515381, |
| "learning_rate": 0.0005245806799784133, |
| "loss": 3.7474, |
| "step": 11750 |
| }, |
| { |
| "epoch": 1.2722371967654986, |
| "grad_norm": 0.5407906174659729, |
| "learning_rate": 0.0005242568807339449, |
| "loss": 3.7444, |
| "step": 11800 |
| }, |
| { |
| "epoch": 1.2776280323450135, |
| "grad_norm": 0.5647597312927246, |
| "learning_rate": 0.0005239330814894765, |
| "loss": 3.744, |
| "step": 11850 |
| }, |
| { |
| "epoch": 1.2830188679245282, |
| "grad_norm": 0.6143915057182312, |
| "learning_rate": 0.000523609282245008, |
| "loss": 3.7301, |
| "step": 11900 |
| }, |
| { |
| "epoch": 1.2884097035040432, |
| "grad_norm": 0.5904483795166016, |
| "learning_rate": 0.0005232854830005396, |
| "loss": 3.7286, |
| "step": 11950 |
| }, |
| { |
| "epoch": 1.2938005390835579, |
| "grad_norm": 0.5413955450057983, |
| "learning_rate": 0.0005229616837560712, |
| "loss": 3.7387, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.2938005390835579, |
| "eval_accuracy": 0.3488454163855726, |
| "eval_loss": 3.6995937824249268, |
| "eval_runtime": 184.1584, |
| "eval_samples_per_second": 97.802, |
| "eval_steps_per_second": 6.114, |
| "step": 12000 |
| }, |
| { |
| "epoch": 1.2991913746630728, |
| "grad_norm": 0.5844531059265137, |
| "learning_rate": 0.0005226378845116028, |
| "loss": 3.7271, |
| "step": 12050 |
| }, |
| { |
| "epoch": 1.3045822102425877, |
| "grad_norm": 0.692406415939331, |
| "learning_rate": 0.0005223140852671344, |
| "loss": 3.7326, |
| "step": 12100 |
| }, |
| { |
| "epoch": 1.3099730458221024, |
| "grad_norm": 0.531926691532135, |
| "learning_rate": 0.0005219902860226659, |
| "loss": 3.7207, |
| "step": 12150 |
| }, |
| { |
| "epoch": 1.3153638814016173, |
| "grad_norm": 0.5657473206520081, |
| "learning_rate": 0.0005216664867781975, |
| "loss": 3.7375, |
| "step": 12200 |
| }, |
| { |
| "epoch": 1.320754716981132, |
| "grad_norm": 0.5740625858306885, |
| "learning_rate": 0.000521342687533729, |
| "loss": 3.7009, |
| "step": 12250 |
| }, |
| { |
| "epoch": 1.326145552560647, |
| "grad_norm": 0.5976618528366089, |
| "learning_rate": 0.0005210188882892606, |
| "loss": 3.7069, |
| "step": 12300 |
| }, |
| { |
| "epoch": 1.3315363881401616, |
| "grad_norm": 0.5708423256874084, |
| "learning_rate": 0.0005206950890447922, |
| "loss": 3.7345, |
| "step": 12350 |
| }, |
| { |
| "epoch": 1.3369272237196765, |
| "grad_norm": 0.6060131192207336, |
| "learning_rate": 0.0005203712898003238, |
| "loss": 3.7245, |
| "step": 12400 |
| }, |
| { |
| "epoch": 1.3423180592991915, |
| "grad_norm": 0.5739725232124329, |
| "learning_rate": 0.0005200474905558553, |
| "loss": 3.7225, |
| "step": 12450 |
| }, |
| { |
| "epoch": 1.3477088948787062, |
| "grad_norm": 0.5318633913993835, |
| "learning_rate": 0.0005197236913113869, |
| "loss": 3.6992, |
| "step": 12500 |
| }, |
| { |
| "epoch": 1.353099730458221, |
| "grad_norm": 0.5786360502243042, |
| "learning_rate": 0.0005193998920669184, |
| "loss": 3.7161, |
| "step": 12550 |
| }, |
| { |
| "epoch": 1.3584905660377358, |
| "grad_norm": 0.5843545198440552, |
| "learning_rate": 0.0005190760928224501, |
| "loss": 3.7138, |
| "step": 12600 |
| }, |
| { |
| "epoch": 1.3638814016172507, |
| "grad_norm": 0.579613447189331, |
| "learning_rate": 0.0005187522935779816, |
| "loss": 3.7258, |
| "step": 12650 |
| }, |
| { |
| "epoch": 1.3692722371967654, |
| "grad_norm": 0.6323828101158142, |
| "learning_rate": 0.0005184284943335132, |
| "loss": 3.7231, |
| "step": 12700 |
| }, |
| { |
| "epoch": 1.3746630727762803, |
| "grad_norm": 0.5477507710456848, |
| "learning_rate": 0.0005181046950890447, |
| "loss": 3.7115, |
| "step": 12750 |
| }, |
| { |
| "epoch": 1.3800539083557952, |
| "grad_norm": 0.5417460203170776, |
| "learning_rate": 0.0005177808958445764, |
| "loss": 3.7014, |
| "step": 12800 |
| }, |
| { |
| "epoch": 1.38544474393531, |
| "grad_norm": 0.5849714279174805, |
| "learning_rate": 0.0005174635725849973, |
| "loss": 3.7173, |
| "step": 12850 |
| }, |
| { |
| "epoch": 1.3908355795148248, |
| "grad_norm": 0.5853115916252136, |
| "learning_rate": 0.0005171397733405288, |
| "loss": 3.7121, |
| "step": 12900 |
| }, |
| { |
| "epoch": 1.3962264150943398, |
| "grad_norm": 0.5460019111633301, |
| "learning_rate": 0.0005168159740960603, |
| "loss": 3.738, |
| "step": 12950 |
| }, |
| { |
| "epoch": 1.4016172506738545, |
| "grad_norm": 0.6247442960739136, |
| "learning_rate": 0.000516492174851592, |
| "loss": 3.7317, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.4016172506738545, |
| "eval_accuracy": 0.3512436029254138, |
| "eval_loss": 3.6774609088897705, |
| "eval_runtime": 183.7689, |
| "eval_samples_per_second": 98.009, |
| "eval_steps_per_second": 6.127, |
| "step": 13000 |
| }, |
| { |
| "epoch": 1.4070080862533692, |
| "grad_norm": 0.6057413220405579, |
| "learning_rate": 0.0005161683756071235, |
| "loss": 3.7133, |
| "step": 13050 |
| }, |
| { |
| "epoch": 1.412398921832884, |
| "grad_norm": 0.5644381642341614, |
| "learning_rate": 0.0005158445763626551, |
| "loss": 3.698, |
| "step": 13100 |
| }, |
| { |
| "epoch": 1.417789757412399, |
| "grad_norm": 0.5523656010627747, |
| "learning_rate": 0.0005155207771181866, |
| "loss": 3.6978, |
| "step": 13150 |
| }, |
| { |
| "epoch": 1.4231805929919137, |
| "grad_norm": 0.5388391613960266, |
| "learning_rate": 0.0005151969778737183, |
| "loss": 3.7136, |
| "step": 13200 |
| }, |
| { |
| "epoch": 1.4285714285714286, |
| "grad_norm": 0.6056439876556396, |
| "learning_rate": 0.0005148731786292498, |
| "loss": 3.7094, |
| "step": 13250 |
| }, |
| { |
| "epoch": 1.4339622641509435, |
| "grad_norm": 0.6014870405197144, |
| "learning_rate": 0.0005145493793847814, |
| "loss": 3.7149, |
| "step": 13300 |
| }, |
| { |
| "epoch": 1.4393530997304582, |
| "grad_norm": 0.6494507193565369, |
| "learning_rate": 0.000514225580140313, |
| "loss": 3.7201, |
| "step": 13350 |
| }, |
| { |
| "epoch": 1.444743935309973, |
| "grad_norm": 0.5282674431800842, |
| "learning_rate": 0.0005139017808958445, |
| "loss": 3.712, |
| "step": 13400 |
| }, |
| { |
| "epoch": 1.4501347708894878, |
| "grad_norm": 0.5607653856277466, |
| "learning_rate": 0.0005135779816513762, |
| "loss": 3.7172, |
| "step": 13450 |
| }, |
| { |
| "epoch": 1.4555256064690028, |
| "grad_norm": 0.5750332474708557, |
| "learning_rate": 0.0005132541824069076, |
| "loss": 3.7082, |
| "step": 13500 |
| }, |
| { |
| "epoch": 1.4609164420485174, |
| "grad_norm": 0.5525111556053162, |
| "learning_rate": 0.0005129303831624393, |
| "loss": 3.7189, |
| "step": 13550 |
| }, |
| { |
| "epoch": 1.4663072776280324, |
| "grad_norm": 0.5573806762695312, |
| "learning_rate": 0.0005126065839179708, |
| "loss": 3.7074, |
| "step": 13600 |
| }, |
| { |
| "epoch": 1.4716981132075473, |
| "grad_norm": 0.5327258706092834, |
| "learning_rate": 0.0005122827846735024, |
| "loss": 3.7108, |
| "step": 13650 |
| }, |
| { |
| "epoch": 1.477088948787062, |
| "grad_norm": 0.6094257831573486, |
| "learning_rate": 0.0005119589854290339, |
| "loss": 3.7015, |
| "step": 13700 |
| }, |
| { |
| "epoch": 1.482479784366577, |
| "grad_norm": 0.5354958772659302, |
| "learning_rate": 0.0005116351861845655, |
| "loss": 3.7019, |
| "step": 13750 |
| }, |
| { |
| "epoch": 1.4878706199460916, |
| "grad_norm": 0.5956083536148071, |
| "learning_rate": 0.0005113113869400971, |
| "loss": 3.696, |
| "step": 13800 |
| }, |
| { |
| "epoch": 1.4932614555256065, |
| "grad_norm": 0.5589125156402588, |
| "learning_rate": 0.0005109875876956287, |
| "loss": 3.6968, |
| "step": 13850 |
| }, |
| { |
| "epoch": 1.4986522911051212, |
| "grad_norm": 0.5720492601394653, |
| "learning_rate": 0.0005106637884511602, |
| "loss": 3.6979, |
| "step": 13900 |
| }, |
| { |
| "epoch": 1.5040431266846361, |
| "grad_norm": 0.5775404572486877, |
| "learning_rate": 0.0005103399892066918, |
| "loss": 3.6881, |
| "step": 13950 |
| }, |
| { |
| "epoch": 1.509433962264151, |
| "grad_norm": 0.5249178409576416, |
| "learning_rate": 0.0005100161899622234, |
| "loss": 3.6971, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.509433962264151, |
| "eval_accuracy": 0.35375946054275814, |
| "eval_loss": 3.6579151153564453, |
| "eval_runtime": 184.1099, |
| "eval_samples_per_second": 97.827, |
| "eval_steps_per_second": 6.116, |
| "step": 14000 |
| }, |
| { |
| "epoch": 1.5148247978436657, |
| "grad_norm": 0.5867551565170288, |
| "learning_rate": 0.000509692390717755, |
| "loss": 3.6801, |
| "step": 14050 |
| }, |
| { |
| "epoch": 1.5202156334231804, |
| "grad_norm": 0.5887311100959778, |
| "learning_rate": 0.0005093685914732865, |
| "loss": 3.6936, |
| "step": 14100 |
| }, |
| { |
| "epoch": 1.5256064690026954, |
| "grad_norm": 0.5415863394737244, |
| "learning_rate": 0.0005090447922288181, |
| "loss": 3.6752, |
| "step": 14150 |
| }, |
| { |
| "epoch": 1.5309973045822103, |
| "grad_norm": 0.5622326731681824, |
| "learning_rate": 0.0005087209929843496, |
| "loss": 3.6959, |
| "step": 14200 |
| }, |
| { |
| "epoch": 1.536388140161725, |
| "grad_norm": 0.5287529230117798, |
| "learning_rate": 0.0005083971937398812, |
| "loss": 3.7028, |
| "step": 14250 |
| }, |
| { |
| "epoch": 1.54177897574124, |
| "grad_norm": 0.5576687455177307, |
| "learning_rate": 0.0005080733944954127, |
| "loss": 3.7062, |
| "step": 14300 |
| }, |
| { |
| "epoch": 1.5471698113207548, |
| "grad_norm": 0.5946956872940063, |
| "learning_rate": 0.0005077495952509444, |
| "loss": 3.7062, |
| "step": 14350 |
| }, |
| { |
| "epoch": 1.5525606469002695, |
| "grad_norm": 0.5189834237098694, |
| "learning_rate": 0.0005074257960064759, |
| "loss": 3.6921, |
| "step": 14400 |
| }, |
| { |
| "epoch": 1.5579514824797842, |
| "grad_norm": 0.5719146728515625, |
| "learning_rate": 0.0005071019967620075, |
| "loss": 3.6974, |
| "step": 14450 |
| }, |
| { |
| "epoch": 1.5633423180592994, |
| "grad_norm": 0.5453420281410217, |
| "learning_rate": 0.000506778197517539, |
| "loss": 3.6744, |
| "step": 14500 |
| }, |
| { |
| "epoch": 1.568733153638814, |
| "grad_norm": 0.5827313661575317, |
| "learning_rate": 0.0005064543982730707, |
| "loss": 3.6753, |
| "step": 14550 |
| }, |
| { |
| "epoch": 1.5741239892183287, |
| "grad_norm": 0.5802609920501709, |
| "learning_rate": 0.0005061305990286023, |
| "loss": 3.6732, |
| "step": 14600 |
| }, |
| { |
| "epoch": 1.5795148247978437, |
| "grad_norm": 0.5037614703178406, |
| "learning_rate": 0.0005058067997841338, |
| "loss": 3.6892, |
| "step": 14650 |
| }, |
| { |
| "epoch": 1.5849056603773586, |
| "grad_norm": 0.5056049823760986, |
| "learning_rate": 0.0005054830005396654, |
| "loss": 3.6858, |
| "step": 14700 |
| }, |
| { |
| "epoch": 1.5902964959568733, |
| "grad_norm": 0.6599529385566711, |
| "learning_rate": 0.0005051592012951969, |
| "loss": 3.6799, |
| "step": 14750 |
| }, |
| { |
| "epoch": 1.595687331536388, |
| "grad_norm": 0.563842236995697, |
| "learning_rate": 0.0005048354020507286, |
| "loss": 3.6988, |
| "step": 14800 |
| }, |
| { |
| "epoch": 1.6010781671159031, |
| "grad_norm": 0.5434414744377136, |
| "learning_rate": 0.00050451160280626, |
| "loss": 3.6767, |
| "step": 14850 |
| }, |
| { |
| "epoch": 1.6064690026954178, |
| "grad_norm": 0.61106938123703, |
| "learning_rate": 0.0005041878035617917, |
| "loss": 3.6979, |
| "step": 14900 |
| }, |
| { |
| "epoch": 1.6118598382749325, |
| "grad_norm": 0.5468453764915466, |
| "learning_rate": 0.0005038640043173232, |
| "loss": 3.6902, |
| "step": 14950 |
| }, |
| { |
| "epoch": 1.6172506738544474, |
| "grad_norm": 0.589815616607666, |
| "learning_rate": 0.0005035402050728548, |
| "loss": 3.6885, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.6172506738544474, |
| "eval_accuracy": 0.3551121889960484, |
| "eval_loss": 3.6401309967041016, |
| "eval_runtime": 183.883, |
| "eval_samples_per_second": 97.948, |
| "eval_steps_per_second": 6.123, |
| "step": 15000 |
| }, |
| { |
| "epoch": 1.6226415094339623, |
| "grad_norm": 0.5671663284301758, |
| "learning_rate": 0.0005032164058283863, |
| "loss": 3.6831, |
| "step": 15050 |
| }, |
| { |
| "epoch": 1.628032345013477, |
| "grad_norm": 0.6244860887527466, |
| "learning_rate": 0.0005028926065839179, |
| "loss": 3.6613, |
| "step": 15100 |
| }, |
| { |
| "epoch": 1.633423180592992, |
| "grad_norm": 0.5171533823013306, |
| "learning_rate": 0.0005025688073394495, |
| "loss": 3.6778, |
| "step": 15150 |
| }, |
| { |
| "epoch": 1.6388140161725069, |
| "grad_norm": 0.5679621696472168, |
| "learning_rate": 0.0005022450080949811, |
| "loss": 3.6648, |
| "step": 15200 |
| }, |
| { |
| "epoch": 1.6442048517520216, |
| "grad_norm": 0.5312994718551636, |
| "learning_rate": 0.0005019212088505126, |
| "loss": 3.6837, |
| "step": 15250 |
| }, |
| { |
| "epoch": 1.6495956873315363, |
| "grad_norm": 0.5288447737693787, |
| "learning_rate": 0.0005015974096060442, |
| "loss": 3.6684, |
| "step": 15300 |
| }, |
| { |
| "epoch": 1.6549865229110512, |
| "grad_norm": 0.6825982928276062, |
| "learning_rate": 0.0005012736103615758, |
| "loss": 3.671, |
| "step": 15350 |
| }, |
| { |
| "epoch": 1.6603773584905661, |
| "grad_norm": 0.5444086790084839, |
| "learning_rate": 0.0005009498111171074, |
| "loss": 3.7001, |
| "step": 15400 |
| }, |
| { |
| "epoch": 1.6657681940700808, |
| "grad_norm": 0.5970629453659058, |
| "learning_rate": 0.0005006260118726389, |
| "loss": 3.6732, |
| "step": 15450 |
| }, |
| { |
| "epoch": 1.6711590296495957, |
| "grad_norm": 0.6333000063896179, |
| "learning_rate": 0.0005003022126281705, |
| "loss": 3.6567, |
| "step": 15500 |
| }, |
| { |
| "epoch": 1.6765498652291106, |
| "grad_norm": 0.5841792821884155, |
| "learning_rate": 0.000499978413383702, |
| "loss": 3.6866, |
| "step": 15550 |
| }, |
| { |
| "epoch": 1.6819407008086253, |
| "grad_norm": 0.5716648697853088, |
| "learning_rate": 0.0004996546141392336, |
| "loss": 3.6464, |
| "step": 15600 |
| }, |
| { |
| "epoch": 1.68733153638814, |
| "grad_norm": 0.6867700219154358, |
| "learning_rate": 0.0004993308148947651, |
| "loss": 3.667, |
| "step": 15650 |
| }, |
| { |
| "epoch": 1.692722371967655, |
| "grad_norm": 0.5503226518630981, |
| "learning_rate": 0.0004990070156502968, |
| "loss": 3.6594, |
| "step": 15700 |
| }, |
| { |
| "epoch": 1.6981132075471699, |
| "grad_norm": 0.5787608623504639, |
| "learning_rate": 0.0004986832164058284, |
| "loss": 3.6795, |
| "step": 15750 |
| }, |
| { |
| "epoch": 1.7035040431266846, |
| "grad_norm": 0.5347179770469666, |
| "learning_rate": 0.0004983594171613599, |
| "loss": 3.6652, |
| "step": 15800 |
| }, |
| { |
| "epoch": 1.7088948787061995, |
| "grad_norm": 0.5248307585716248, |
| "learning_rate": 0.0004980356179168915, |
| "loss": 3.6812, |
| "step": 15850 |
| }, |
| { |
| "epoch": 1.7142857142857144, |
| "grad_norm": 0.5549546480178833, |
| "learning_rate": 0.000497711818672423, |
| "loss": 3.6636, |
| "step": 15900 |
| }, |
| { |
| "epoch": 1.719676549865229, |
| "grad_norm": 0.6331003308296204, |
| "learning_rate": 0.0004973880194279547, |
| "loss": 3.6594, |
| "step": 15950 |
| }, |
| { |
| "epoch": 1.7250673854447438, |
| "grad_norm": 0.6510477662086487, |
| "learning_rate": 0.0004970642201834862, |
| "loss": 3.6771, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.7250673854447438, |
| "eval_accuracy": 0.3569796061789159, |
| "eval_loss": 3.619459867477417, |
| "eval_runtime": 183.9465, |
| "eval_samples_per_second": 97.914, |
| "eval_steps_per_second": 6.121, |
| "step": 16000 |
| }, |
| { |
| "epoch": 1.7304582210242587, |
| "grad_norm": 0.6069450378417969, |
| "learning_rate": 0.0004967404209390178, |
| "loss": 3.654, |
| "step": 16050 |
| }, |
| { |
| "epoch": 1.7358490566037736, |
| "grad_norm": 0.5242086052894592, |
| "learning_rate": 0.0004964166216945493, |
| "loss": 3.6722, |
| "step": 16100 |
| }, |
| { |
| "epoch": 1.7412398921832883, |
| "grad_norm": 0.612231433391571, |
| "learning_rate": 0.000496092822450081, |
| "loss": 3.6597, |
| "step": 16150 |
| }, |
| { |
| "epoch": 1.7466307277628033, |
| "grad_norm": 0.6022439002990723, |
| "learning_rate": 0.0004957690232056125, |
| "loss": 3.6801, |
| "step": 16200 |
| }, |
| { |
| "epoch": 1.7520215633423182, |
| "grad_norm": 0.5812779664993286, |
| "learning_rate": 0.0004954452239611441, |
| "loss": 3.6582, |
| "step": 16250 |
| }, |
| { |
| "epoch": 1.7574123989218329, |
| "grad_norm": 0.5933021306991577, |
| "learning_rate": 0.0004951214247166756, |
| "loss": 3.6652, |
| "step": 16300 |
| }, |
| { |
| "epoch": 1.7628032345013476, |
| "grad_norm": 0.5812273621559143, |
| "learning_rate": 0.0004947976254722072, |
| "loss": 3.6697, |
| "step": 16350 |
| }, |
| { |
| "epoch": 1.7681940700808625, |
| "grad_norm": 0.6487030386924744, |
| "learning_rate": 0.0004944738262277387, |
| "loss": 3.6677, |
| "step": 16400 |
| }, |
| { |
| "epoch": 1.7735849056603774, |
| "grad_norm": 0.548700749874115, |
| "learning_rate": 0.0004941500269832703, |
| "loss": 3.658, |
| "step": 16450 |
| }, |
| { |
| "epoch": 1.778975741239892, |
| "grad_norm": 0.6098374724388123, |
| "learning_rate": 0.0004938262277388019, |
| "loss": 3.6542, |
| "step": 16500 |
| }, |
| { |
| "epoch": 1.784366576819407, |
| "grad_norm": 0.6051768660545349, |
| "learning_rate": 0.0004935024284943335, |
| "loss": 3.6566, |
| "step": 16550 |
| }, |
| { |
| "epoch": 1.789757412398922, |
| "grad_norm": 0.6026385426521301, |
| "learning_rate": 0.000493178629249865, |
| "loss": 3.653, |
| "step": 16600 |
| }, |
| { |
| "epoch": 1.7951482479784366, |
| "grad_norm": 0.562525749206543, |
| "learning_rate": 0.0004928548300053966, |
| "loss": 3.6491, |
| "step": 16650 |
| }, |
| { |
| "epoch": 1.8005390835579513, |
| "grad_norm": 0.571699321269989, |
| "learning_rate": 0.0004925310307609282, |
| "loss": 3.6552, |
| "step": 16700 |
| }, |
| { |
| "epoch": 1.8059299191374663, |
| "grad_norm": 0.5961069464683533, |
| "learning_rate": 0.0004922072315164598, |
| "loss": 3.6492, |
| "step": 16750 |
| }, |
| { |
| "epoch": 1.8113207547169812, |
| "grad_norm": 0.5025362372398376, |
| "learning_rate": 0.0004918834322719913, |
| "loss": 3.6478, |
| "step": 16800 |
| }, |
| { |
| "epoch": 1.8167115902964959, |
| "grad_norm": 0.5551483631134033, |
| "learning_rate": 0.0004915661090124123, |
| "loss": 3.6596, |
| "step": 16850 |
| }, |
| { |
| "epoch": 1.8221024258760108, |
| "grad_norm": 0.5412172079086304, |
| "learning_rate": 0.0004912423097679439, |
| "loss": 3.6498, |
| "step": 16900 |
| }, |
| { |
| "epoch": 1.8274932614555257, |
| "grad_norm": 0.6038126945495605, |
| "learning_rate": 0.0004909185105234754, |
| "loss": 3.6542, |
| "step": 16950 |
| }, |
| { |
| "epoch": 1.8328840970350404, |
| "grad_norm": 0.5766414403915405, |
| "learning_rate": 0.000490594711279007, |
| "loss": 3.6503, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.8328840970350404, |
| "eval_accuracy": 0.35894415904348953, |
| "eval_loss": 3.6033706665039062, |
| "eval_runtime": 184.1497, |
| "eval_samples_per_second": 97.806, |
| "eval_steps_per_second": 6.115, |
| "step": 17000 |
| }, |
| { |
| "epoch": 1.838274932614555, |
| "grad_norm": 0.582373857498169, |
| "learning_rate": 0.0004902709120345385, |
| "loss": 3.641, |
| "step": 17050 |
| }, |
| { |
| "epoch": 1.8436657681940702, |
| "grad_norm": 0.5415629148483276, |
| "learning_rate": 0.00048994711279007, |
| "loss": 3.6456, |
| "step": 17100 |
| }, |
| { |
| "epoch": 1.849056603773585, |
| "grad_norm": 0.5694625973701477, |
| "learning_rate": 0.0004896233135456017, |
| "loss": 3.6568, |
| "step": 17150 |
| }, |
| { |
| "epoch": 1.8544474393530996, |
| "grad_norm": 0.547035276889801, |
| "learning_rate": 0.0004892995143011333, |
| "loss": 3.654, |
| "step": 17200 |
| }, |
| { |
| "epoch": 1.8598382749326146, |
| "grad_norm": 0.6010807752609253, |
| "learning_rate": 0.0004889757150566648, |
| "loss": 3.6348, |
| "step": 17250 |
| }, |
| { |
| "epoch": 1.8652291105121295, |
| "grad_norm": 0.5640713572502136, |
| "learning_rate": 0.0004886519158121964, |
| "loss": 3.6555, |
| "step": 17300 |
| }, |
| { |
| "epoch": 1.8706199460916442, |
| "grad_norm": 0.6125467419624329, |
| "learning_rate": 0.000488328116567728, |
| "loss": 3.656, |
| "step": 17350 |
| }, |
| { |
| "epoch": 1.8760107816711589, |
| "grad_norm": 0.6842520833015442, |
| "learning_rate": 0.0004880043173232595, |
| "loss": 3.6642, |
| "step": 17400 |
| }, |
| { |
| "epoch": 1.881401617250674, |
| "grad_norm": 0.5659988522529602, |
| "learning_rate": 0.0004876805180787911, |
| "loss": 3.6325, |
| "step": 17450 |
| }, |
| { |
| "epoch": 1.8867924528301887, |
| "grad_norm": 0.6425742506980896, |
| "learning_rate": 0.0004873567188343227, |
| "loss": 3.6474, |
| "step": 17500 |
| }, |
| { |
| "epoch": 1.8921832884097034, |
| "grad_norm": 0.5763974785804749, |
| "learning_rate": 0.0004870329195898542, |
| "loss": 3.6383, |
| "step": 17550 |
| }, |
| { |
| "epoch": 1.8975741239892183, |
| "grad_norm": 0.5723247528076172, |
| "learning_rate": 0.00048670912034538583, |
| "loss": 3.6529, |
| "step": 17600 |
| }, |
| { |
| "epoch": 1.9029649595687332, |
| "grad_norm": 0.5686931610107422, |
| "learning_rate": 0.0004863853211009174, |
| "loss": 3.6496, |
| "step": 17650 |
| }, |
| { |
| "epoch": 1.908355795148248, |
| "grad_norm": 0.5761986970901489, |
| "learning_rate": 0.000486061521856449, |
| "loss": 3.6397, |
| "step": 17700 |
| }, |
| { |
| "epoch": 1.9137466307277629, |
| "grad_norm": 0.6086380481719971, |
| "learning_rate": 0.00048573772261198054, |
| "loss": 3.6421, |
| "step": 17750 |
| }, |
| { |
| "epoch": 1.9191374663072778, |
| "grad_norm": 0.5803984999656677, |
| "learning_rate": 0.00048541392336751214, |
| "loss": 3.6658, |
| "step": 17800 |
| }, |
| { |
| "epoch": 1.9245283018867925, |
| "grad_norm": 0.6012025475502014, |
| "learning_rate": 0.0004850901241230437, |
| "loss": 3.6396, |
| "step": 17850 |
| }, |
| { |
| "epoch": 1.9299191374663072, |
| "grad_norm": 0.5470909476280212, |
| "learning_rate": 0.0004847663248785753, |
| "loss": 3.6366, |
| "step": 17900 |
| }, |
| { |
| "epoch": 1.935309973045822, |
| "grad_norm": 0.6374045014381409, |
| "learning_rate": 0.0004844425256341068, |
| "loss": 3.6164, |
| "step": 17950 |
| }, |
| { |
| "epoch": 1.940700808625337, |
| "grad_norm": 0.5931817889213562, |
| "learning_rate": 0.00048411872638963834, |
| "loss": 3.6523, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.940700808625337, |
| "eval_accuracy": 0.35995310975973477, |
| "eval_loss": 3.5907158851623535, |
| "eval_runtime": 183.7191, |
| "eval_samples_per_second": 98.036, |
| "eval_steps_per_second": 6.129, |
| "step": 18000 |
| }, |
| { |
| "epoch": 1.9460916442048517, |
| "grad_norm": 0.5779642462730408, |
| "learning_rate": 0.00048379492714516995, |
| "loss": 3.6288, |
| "step": 18050 |
| }, |
| { |
| "epoch": 1.9514824797843666, |
| "grad_norm": 0.5749161839485168, |
| "learning_rate": 0.0004834711279007015, |
| "loss": 3.6552, |
| "step": 18100 |
| }, |
| { |
| "epoch": 1.9568733153638815, |
| "grad_norm": 0.5829963684082031, |
| "learning_rate": 0.0004831473286562331, |
| "loss": 3.6656, |
| "step": 18150 |
| }, |
| { |
| "epoch": 1.9622641509433962, |
| "grad_norm": 0.547026515007019, |
| "learning_rate": 0.00048282352941176465, |
| "loss": 3.6443, |
| "step": 18200 |
| }, |
| { |
| "epoch": 1.967654986522911, |
| "grad_norm": 0.5671616792678833, |
| "learning_rate": 0.00048249973016729626, |
| "loss": 3.6491, |
| "step": 18250 |
| }, |
| { |
| "epoch": 1.9730458221024259, |
| "grad_norm": 0.5770370960235596, |
| "learning_rate": 0.0004821759309228278, |
| "loss": 3.6299, |
| "step": 18300 |
| }, |
| { |
| "epoch": 1.9784366576819408, |
| "grad_norm": 0.5550295114517212, |
| "learning_rate": 0.00048185213167835936, |
| "loss": 3.6396, |
| "step": 18350 |
| }, |
| { |
| "epoch": 1.9838274932614555, |
| "grad_norm": 0.5505232214927673, |
| "learning_rate": 0.00048152833243389096, |
| "loss": 3.6541, |
| "step": 18400 |
| }, |
| { |
| "epoch": 1.9892183288409704, |
| "grad_norm": 0.5260711908340454, |
| "learning_rate": 0.0004812045331894225, |
| "loss": 3.6188, |
| "step": 18450 |
| }, |
| { |
| "epoch": 1.9946091644204853, |
| "grad_norm": 0.5258703231811523, |
| "learning_rate": 0.0004808807339449541, |
| "loss": 3.6256, |
| "step": 18500 |
| }, |
| { |
| "epoch": 2.0, |
| "grad_norm": 1.088941216468811, |
| "learning_rate": 0.0004805569347004856, |
| "loss": 3.6356, |
| "step": 18550 |
| }, |
| { |
| "epoch": 2.0053908355795147, |
| "grad_norm": 0.5310294032096863, |
| "learning_rate": 0.0004802331354560173, |
| "loss": 3.542, |
| "step": 18600 |
| }, |
| { |
| "epoch": 2.01078167115903, |
| "grad_norm": 0.5242578387260437, |
| "learning_rate": 0.00047990933621154877, |
| "loss": 3.5532, |
| "step": 18650 |
| }, |
| { |
| "epoch": 2.0161725067385445, |
| "grad_norm": 0.5820230841636658, |
| "learning_rate": 0.0004795855369670804, |
| "loss": 3.5545, |
| "step": 18700 |
| }, |
| { |
| "epoch": 2.0215633423180592, |
| "grad_norm": 0.6573295593261719, |
| "learning_rate": 0.0004792617377226119, |
| "loss": 3.5472, |
| "step": 18750 |
| }, |
| { |
| "epoch": 2.026954177897574, |
| "grad_norm": 0.5828113555908203, |
| "learning_rate": 0.0004789379384781435, |
| "loss": 3.5532, |
| "step": 18800 |
| }, |
| { |
| "epoch": 2.032345013477089, |
| "grad_norm": 0.6123409867286682, |
| "learning_rate": 0.0004786141392336751, |
| "loss": 3.5318, |
| "step": 18850 |
| }, |
| { |
| "epoch": 2.0377358490566038, |
| "grad_norm": 0.5386969447135925, |
| "learning_rate": 0.00047829033998920663, |
| "loss": 3.5529, |
| "step": 18900 |
| }, |
| { |
| "epoch": 2.0431266846361185, |
| "grad_norm": 0.5555720329284668, |
| "learning_rate": 0.00047796654074473824, |
| "loss": 3.5571, |
| "step": 18950 |
| }, |
| { |
| "epoch": 2.0485175202156336, |
| "grad_norm": 0.5521024465560913, |
| "learning_rate": 0.0004776427415002698, |
| "loss": 3.5469, |
| "step": 19000 |
| }, |
| { |
| "epoch": 2.0485175202156336, |
| "eval_accuracy": 0.3615377034756212, |
| "eval_loss": 3.5789687633514404, |
| "eval_runtime": 184.2139, |
| "eval_samples_per_second": 97.772, |
| "eval_steps_per_second": 6.112, |
| "step": 19000 |
| }, |
| { |
| "epoch": 2.0539083557951483, |
| "grad_norm": 0.6329952478408813, |
| "learning_rate": 0.00047732541824069074, |
| "loss": 3.5373, |
| "step": 19050 |
| }, |
| { |
| "epoch": 2.059299191374663, |
| "grad_norm": 0.5793041586875916, |
| "learning_rate": 0.0004770016189962223, |
| "loss": 3.5498, |
| "step": 19100 |
| }, |
| { |
| "epoch": 2.0646900269541777, |
| "grad_norm": 0.5503807067871094, |
| "learning_rate": 0.0004766778197517539, |
| "loss": 3.5437, |
| "step": 19150 |
| }, |
| { |
| "epoch": 2.070080862533693, |
| "grad_norm": 0.5668014883995056, |
| "learning_rate": 0.0004763540205072854, |
| "loss": 3.5563, |
| "step": 19200 |
| }, |
| { |
| "epoch": 2.0754716981132075, |
| "grad_norm": 0.5484731197357178, |
| "learning_rate": 0.00047603022126281705, |
| "loss": 3.554, |
| "step": 19250 |
| }, |
| { |
| "epoch": 2.0808625336927222, |
| "grad_norm": 0.6245366930961609, |
| "learning_rate": 0.00047570642201834855, |
| "loss": 3.5468, |
| "step": 19300 |
| }, |
| { |
| "epoch": 2.0862533692722374, |
| "grad_norm": 0.6225091814994812, |
| "learning_rate": 0.00047538262277388015, |
| "loss": 3.5642, |
| "step": 19350 |
| }, |
| { |
| "epoch": 2.091644204851752, |
| "grad_norm": 0.5625278949737549, |
| "learning_rate": 0.0004750588235294117, |
| "loss": 3.5567, |
| "step": 19400 |
| }, |
| { |
| "epoch": 2.0970350404312668, |
| "grad_norm": 0.632989764213562, |
| "learning_rate": 0.0004747350242849433, |
| "loss": 3.5484, |
| "step": 19450 |
| }, |
| { |
| "epoch": 2.1024258760107815, |
| "grad_norm": 0.5744699239730835, |
| "learning_rate": 0.00047441122504047486, |
| "loss": 3.5617, |
| "step": 19500 |
| }, |
| { |
| "epoch": 2.1078167115902966, |
| "grad_norm": 0.5948736071586609, |
| "learning_rate": 0.0004740874257960064, |
| "loss": 3.542, |
| "step": 19550 |
| }, |
| { |
| "epoch": 2.1132075471698113, |
| "grad_norm": 0.5679342746734619, |
| "learning_rate": 0.000473763626551538, |
| "loss": 3.5556, |
| "step": 19600 |
| }, |
| { |
| "epoch": 2.118598382749326, |
| "grad_norm": 0.5541220307350159, |
| "learning_rate": 0.00047343982730706956, |
| "loss": 3.5516, |
| "step": 19650 |
| }, |
| { |
| "epoch": 2.123989218328841, |
| "grad_norm": 0.544877290725708, |
| "learning_rate": 0.00047311602806260117, |
| "loss": 3.5489, |
| "step": 19700 |
| }, |
| { |
| "epoch": 2.129380053908356, |
| "grad_norm": 0.5709410309791565, |
| "learning_rate": 0.0004727922288181327, |
| "loss": 3.5652, |
| "step": 19750 |
| }, |
| { |
| "epoch": 2.1347708894878705, |
| "grad_norm": 0.5732446908950806, |
| "learning_rate": 0.0004724684295736643, |
| "loss": 3.5745, |
| "step": 19800 |
| }, |
| { |
| "epoch": 2.1401617250673857, |
| "grad_norm": 0.6136924028396606, |
| "learning_rate": 0.0004721446303291959, |
| "loss": 3.5503, |
| "step": 19850 |
| }, |
| { |
| "epoch": 2.1455525606469004, |
| "grad_norm": 0.5857067108154297, |
| "learning_rate": 0.0004718208310847275, |
| "loss": 3.5585, |
| "step": 19900 |
| }, |
| { |
| "epoch": 2.150943396226415, |
| "grad_norm": 0.5760287046432495, |
| "learning_rate": 0.000471497031840259, |
| "loss": 3.5601, |
| "step": 19950 |
| }, |
| { |
| "epoch": 2.1563342318059298, |
| "grad_norm": 0.5563052296638489, |
| "learning_rate": 0.0004711732325957905, |
| "loss": 3.553, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.1563342318059298, |
| "eval_accuracy": 0.36297746289205085, |
| "eval_loss": 3.5687758922576904, |
| "eval_runtime": 183.8162, |
| "eval_samples_per_second": 97.984, |
| "eval_steps_per_second": 6.126, |
| "step": 20000 |
| }, |
| { |
| "epoch": 2.161725067385445, |
| "grad_norm": 0.6046416759490967, |
| "learning_rate": 0.00047084943335132213, |
| "loss": 3.5431, |
| "step": 20050 |
| }, |
| { |
| "epoch": 2.1671159029649596, |
| "grad_norm": 0.5706889033317566, |
| "learning_rate": 0.0004705256341068537, |
| "loss": 3.5522, |
| "step": 20100 |
| }, |
| { |
| "epoch": 2.1725067385444743, |
| "grad_norm": 0.6285442113876343, |
| "learning_rate": 0.0004702018348623853, |
| "loss": 3.5604, |
| "step": 20150 |
| }, |
| { |
| "epoch": 2.177897574123989, |
| "grad_norm": 0.567588746547699, |
| "learning_rate": 0.00046987803561791684, |
| "loss": 3.5649, |
| "step": 20200 |
| }, |
| { |
| "epoch": 2.183288409703504, |
| "grad_norm": 0.543428897857666, |
| "learning_rate": 0.00046955423637344844, |
| "loss": 3.5729, |
| "step": 20250 |
| }, |
| { |
| "epoch": 2.188679245283019, |
| "grad_norm": 0.6000525951385498, |
| "learning_rate": 0.00046923043712898, |
| "loss": 3.5503, |
| "step": 20300 |
| }, |
| { |
| "epoch": 2.1940700808625335, |
| "grad_norm": 0.6174874305725098, |
| "learning_rate": 0.0004689066378845116, |
| "loss": 3.5539, |
| "step": 20350 |
| }, |
| { |
| "epoch": 2.1994609164420487, |
| "grad_norm": 0.591032087802887, |
| "learning_rate": 0.00046858283864004315, |
| "loss": 3.5589, |
| "step": 20400 |
| }, |
| { |
| "epoch": 2.2048517520215634, |
| "grad_norm": 0.5443971753120422, |
| "learning_rate": 0.0004682590393955747, |
| "loss": 3.551, |
| "step": 20450 |
| }, |
| { |
| "epoch": 2.210242587601078, |
| "grad_norm": 0.6590529084205627, |
| "learning_rate": 0.0004679352401511063, |
| "loss": 3.5497, |
| "step": 20500 |
| }, |
| { |
| "epoch": 2.215633423180593, |
| "grad_norm": 0.592398464679718, |
| "learning_rate": 0.0004676114409066378, |
| "loss": 3.5583, |
| "step": 20550 |
| }, |
| { |
| "epoch": 2.221024258760108, |
| "grad_norm": 0.5806289911270142, |
| "learning_rate": 0.00046728764166216946, |
| "loss": 3.5656, |
| "step": 20600 |
| }, |
| { |
| "epoch": 2.2264150943396226, |
| "grad_norm": 0.5720568299293518, |
| "learning_rate": 0.00046696384241770095, |
| "loss": 3.5663, |
| "step": 20650 |
| }, |
| { |
| "epoch": 2.2318059299191373, |
| "grad_norm": 0.6045317649841309, |
| "learning_rate": 0.00046664004317323256, |
| "loss": 3.5707, |
| "step": 20700 |
| }, |
| { |
| "epoch": 2.2371967654986524, |
| "grad_norm": 0.587020218372345, |
| "learning_rate": 0.0004663162439287641, |
| "loss": 3.5501, |
| "step": 20750 |
| }, |
| { |
| "epoch": 2.242587601078167, |
| "grad_norm": 0.6044445037841797, |
| "learning_rate": 0.00046599244468429566, |
| "loss": 3.5433, |
| "step": 20800 |
| }, |
| { |
| "epoch": 2.247978436657682, |
| "grad_norm": 0.5762318968772888, |
| "learning_rate": 0.00046566864543982726, |
| "loss": 3.5415, |
| "step": 20850 |
| }, |
| { |
| "epoch": 2.2533692722371965, |
| "grad_norm": 0.5729345083236694, |
| "learning_rate": 0.0004653448461953588, |
| "loss": 3.5473, |
| "step": 20900 |
| }, |
| { |
| "epoch": 2.2587601078167117, |
| "grad_norm": 0.619462251663208, |
| "learning_rate": 0.0004650210469508904, |
| "loss": 3.5689, |
| "step": 20950 |
| }, |
| { |
| "epoch": 2.2641509433962264, |
| "grad_norm": 0.5913404226303101, |
| "learning_rate": 0.00046469724770642197, |
| "loss": 3.5623, |
| "step": 21000 |
| }, |
| { |
| "epoch": 2.2641509433962264, |
| "eval_accuracy": 0.3636772961424639, |
| "eval_loss": 3.5585999488830566, |
| "eval_runtime": 183.9266, |
| "eval_samples_per_second": 97.925, |
| "eval_steps_per_second": 6.122, |
| "step": 21000 |
| }, |
| { |
| "epoch": 2.269541778975741, |
| "grad_norm": 0.584413468837738, |
| "learning_rate": 0.0004643734484619536, |
| "loss": 3.5674, |
| "step": 21050 |
| }, |
| { |
| "epoch": 2.274932614555256, |
| "grad_norm": 0.5533130764961243, |
| "learning_rate": 0.0004640496492174851, |
| "loss": 3.5502, |
| "step": 21100 |
| }, |
| { |
| "epoch": 2.280323450134771, |
| "grad_norm": 0.5667862296104431, |
| "learning_rate": 0.0004637323259579061, |
| "loss": 3.5487, |
| "step": 21150 |
| }, |
| { |
| "epoch": 2.2857142857142856, |
| "grad_norm": 0.5526608824729919, |
| "learning_rate": 0.0004634085267134376, |
| "loss": 3.5586, |
| "step": 21200 |
| }, |
| { |
| "epoch": 2.2911051212938007, |
| "grad_norm": 0.6517942547798157, |
| "learning_rate": 0.00046308472746896923, |
| "loss": 3.5757, |
| "step": 21250 |
| }, |
| { |
| "epoch": 2.2964959568733154, |
| "grad_norm": 0.5420659184455872, |
| "learning_rate": 0.00046276092822450073, |
| "loss": 3.5505, |
| "step": 21300 |
| }, |
| { |
| "epoch": 2.30188679245283, |
| "grad_norm": 0.5619015097618103, |
| "learning_rate": 0.00046243712898003233, |
| "loss": 3.5671, |
| "step": 21350 |
| }, |
| { |
| "epoch": 2.3072776280323453, |
| "grad_norm": 0.5901092886924744, |
| "learning_rate": 0.0004621133297355639, |
| "loss": 3.5691, |
| "step": 21400 |
| }, |
| { |
| "epoch": 2.31266846361186, |
| "grad_norm": 0.5045002102851868, |
| "learning_rate": 0.0004617895304910955, |
| "loss": 3.5569, |
| "step": 21450 |
| }, |
| { |
| "epoch": 2.3180592991913747, |
| "grad_norm": 0.6043163537979126, |
| "learning_rate": 0.00046146573124662704, |
| "loss": 3.5506, |
| "step": 21500 |
| }, |
| { |
| "epoch": 2.3234501347708894, |
| "grad_norm": 0.5670344829559326, |
| "learning_rate": 0.00046114193200215864, |
| "loss": 3.5562, |
| "step": 21550 |
| }, |
| { |
| "epoch": 2.3288409703504045, |
| "grad_norm": 0.5510123372077942, |
| "learning_rate": 0.0004608181327576902, |
| "loss": 3.5494, |
| "step": 21600 |
| }, |
| { |
| "epoch": 2.334231805929919, |
| "grad_norm": 0.5462912917137146, |
| "learning_rate": 0.00046049433351322175, |
| "loss": 3.5591, |
| "step": 21650 |
| }, |
| { |
| "epoch": 2.339622641509434, |
| "grad_norm": 0.6254079341888428, |
| "learning_rate": 0.00046017053426875335, |
| "loss": 3.5383, |
| "step": 21700 |
| }, |
| { |
| "epoch": 2.3450134770889486, |
| "grad_norm": 0.5986773371696472, |
| "learning_rate": 0.0004598467350242849, |
| "loss": 3.5331, |
| "step": 21750 |
| }, |
| { |
| "epoch": 2.3504043126684637, |
| "grad_norm": 0.5816046595573425, |
| "learning_rate": 0.0004595229357798165, |
| "loss": 3.5548, |
| "step": 21800 |
| }, |
| { |
| "epoch": 2.3557951482479784, |
| "grad_norm": 0.5853009223937988, |
| "learning_rate": 0.00045919913653534806, |
| "loss": 3.552, |
| "step": 21850 |
| }, |
| { |
| "epoch": 2.361185983827493, |
| "grad_norm": 0.5687203407287598, |
| "learning_rate": 0.00045887533729087966, |
| "loss": 3.5502, |
| "step": 21900 |
| }, |
| { |
| "epoch": 2.3665768194070083, |
| "grad_norm": 0.604446291923523, |
| "learning_rate": 0.00045855153804641116, |
| "loss": 3.5531, |
| "step": 21950 |
| }, |
| { |
| "epoch": 2.371967654986523, |
| "grad_norm": 0.570324182510376, |
| "learning_rate": 0.0004582277388019427, |
| "loss": 3.5464, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.371967654986523, |
| "eval_accuracy": 0.36487475961906735, |
| "eval_loss": 3.5480477809906006, |
| "eval_runtime": 183.8636, |
| "eval_samples_per_second": 97.958, |
| "eval_steps_per_second": 6.124, |
| "step": 22000 |
| }, |
| { |
| "epoch": 2.3773584905660377, |
| "grad_norm": 0.6100995540618896, |
| "learning_rate": 0.0004579039395574743, |
| "loss": 3.5415, |
| "step": 22050 |
| }, |
| { |
| "epoch": 2.382749326145553, |
| "grad_norm": 0.5746794939041138, |
| "learning_rate": 0.00045758014031300586, |
| "loss": 3.557, |
| "step": 22100 |
| }, |
| { |
| "epoch": 2.3881401617250675, |
| "grad_norm": 0.6097766757011414, |
| "learning_rate": 0.00045725634106853747, |
| "loss": 3.5718, |
| "step": 22150 |
| }, |
| { |
| "epoch": 2.393530997304582, |
| "grad_norm": 0.5994210243225098, |
| "learning_rate": 0.000456932541824069, |
| "loss": 3.5689, |
| "step": 22200 |
| }, |
| { |
| "epoch": 2.398921832884097, |
| "grad_norm": 0.5749094486236572, |
| "learning_rate": 0.0004566087425796006, |
| "loss": 3.538, |
| "step": 22250 |
| }, |
| { |
| "epoch": 2.404312668463612, |
| "grad_norm": 0.549342691898346, |
| "learning_rate": 0.0004562849433351322, |
| "loss": 3.5508, |
| "step": 22300 |
| }, |
| { |
| "epoch": 2.4097035040431267, |
| "grad_norm": 0.5516054630279541, |
| "learning_rate": 0.0004559611440906638, |
| "loss": 3.5465, |
| "step": 22350 |
| }, |
| { |
| "epoch": 2.4150943396226414, |
| "grad_norm": 0.5950953364372253, |
| "learning_rate": 0.00045563734484619533, |
| "loss": 3.5643, |
| "step": 22400 |
| }, |
| { |
| "epoch": 2.420485175202156, |
| "grad_norm": 0.8571392297744751, |
| "learning_rate": 0.0004553135456017269, |
| "loss": 3.5725, |
| "step": 22450 |
| }, |
| { |
| "epoch": 2.4258760107816713, |
| "grad_norm": 0.6473979949951172, |
| "learning_rate": 0.0004549897463572585, |
| "loss": 3.5584, |
| "step": 22500 |
| }, |
| { |
| "epoch": 2.431266846361186, |
| "grad_norm": 0.5623269081115723, |
| "learning_rate": 0.00045466594711279, |
| "loss": 3.5454, |
| "step": 22550 |
| }, |
| { |
| "epoch": 2.4366576819407006, |
| "grad_norm": 0.5944833159446716, |
| "learning_rate": 0.00045434214786832164, |
| "loss": 3.5497, |
| "step": 22600 |
| }, |
| { |
| "epoch": 2.442048517520216, |
| "grad_norm": 0.5693549513816833, |
| "learning_rate": 0.00045401834862385314, |
| "loss": 3.552, |
| "step": 22650 |
| }, |
| { |
| "epoch": 2.4474393530997305, |
| "grad_norm": 0.5360573530197144, |
| "learning_rate": 0.00045369454937938474, |
| "loss": 3.5426, |
| "step": 22700 |
| }, |
| { |
| "epoch": 2.452830188679245, |
| "grad_norm": 0.5875699520111084, |
| "learning_rate": 0.0004533707501349163, |
| "loss": 3.5768, |
| "step": 22750 |
| }, |
| { |
| "epoch": 2.4582210242587603, |
| "grad_norm": 0.5625613331794739, |
| "learning_rate": 0.0004530469508904479, |
| "loss": 3.543, |
| "step": 22800 |
| }, |
| { |
| "epoch": 2.463611859838275, |
| "grad_norm": 0.5869686603546143, |
| "learning_rate": 0.0004527296276308688, |
| "loss": 3.5705, |
| "step": 22850 |
| }, |
| { |
| "epoch": 2.4690026954177897, |
| "grad_norm": 0.5737190842628479, |
| "learning_rate": 0.0004524058283864004, |
| "loss": 3.5454, |
| "step": 22900 |
| }, |
| { |
| "epoch": 2.4743935309973044, |
| "grad_norm": 0.5693169236183167, |
| "learning_rate": 0.00045208202914193195, |
| "loss": 3.5611, |
| "step": 22950 |
| }, |
| { |
| "epoch": 2.4797843665768196, |
| "grad_norm": 0.6664920449256897, |
| "learning_rate": 0.00045175822989746355, |
| "loss": 3.543, |
| "step": 23000 |
| }, |
| { |
| "epoch": 2.4797843665768196, |
| "eval_accuracy": 0.3658964227231869, |
| "eval_loss": 3.5396640300750732, |
| "eval_runtime": 184.0061, |
| "eval_samples_per_second": 97.883, |
| "eval_steps_per_second": 6.119, |
| "step": 23000 |
| }, |
| { |
| "epoch": 2.4851752021563343, |
| "grad_norm": 0.5835683345794678, |
| "learning_rate": 0.0004514344306529951, |
| "loss": 3.5382, |
| "step": 23050 |
| }, |
| { |
| "epoch": 2.490566037735849, |
| "grad_norm": 0.5643031001091003, |
| "learning_rate": 0.0004511106314085267, |
| "loss": 3.537, |
| "step": 23100 |
| }, |
| { |
| "epoch": 2.4959568733153636, |
| "grad_norm": 0.5767573118209839, |
| "learning_rate": 0.00045078683216405826, |
| "loss": 3.5492, |
| "step": 23150 |
| }, |
| { |
| "epoch": 2.501347708894879, |
| "grad_norm": 0.5756224989891052, |
| "learning_rate": 0.00045046303291958976, |
| "loss": 3.5463, |
| "step": 23200 |
| }, |
| { |
| "epoch": 2.5067385444743935, |
| "grad_norm": 0.6499918103218079, |
| "learning_rate": 0.0004501392336751214, |
| "loss": 3.561, |
| "step": 23250 |
| }, |
| { |
| "epoch": 2.512129380053908, |
| "grad_norm": 0.6103466153144836, |
| "learning_rate": 0.0004498154344306529, |
| "loss": 3.5578, |
| "step": 23300 |
| }, |
| { |
| "epoch": 2.5175202156334233, |
| "grad_norm": 0.6318963170051575, |
| "learning_rate": 0.0004494916351861845, |
| "loss": 3.5593, |
| "step": 23350 |
| }, |
| { |
| "epoch": 2.522911051212938, |
| "grad_norm": 0.6232575178146362, |
| "learning_rate": 0.00044916783594171607, |
| "loss": 3.5548, |
| "step": 23400 |
| }, |
| { |
| "epoch": 2.5283018867924527, |
| "grad_norm": 0.5690689086914062, |
| "learning_rate": 0.00044884403669724767, |
| "loss": 3.5452, |
| "step": 23450 |
| }, |
| { |
| "epoch": 2.533692722371968, |
| "grad_norm": 0.6115185618400574, |
| "learning_rate": 0.0004485202374527792, |
| "loss": 3.5415, |
| "step": 23500 |
| }, |
| { |
| "epoch": 2.5390835579514826, |
| "grad_norm": 0.5683715343475342, |
| "learning_rate": 0.00044819643820831083, |
| "loss": 3.5434, |
| "step": 23550 |
| }, |
| { |
| "epoch": 2.5444743935309972, |
| "grad_norm": 0.6168181896209717, |
| "learning_rate": 0.0004478726389638424, |
| "loss": 3.5498, |
| "step": 23600 |
| }, |
| { |
| "epoch": 2.5498652291105124, |
| "grad_norm": 0.6147149205207825, |
| "learning_rate": 0.00044754883971937393, |
| "loss": 3.5289, |
| "step": 23650 |
| }, |
| { |
| "epoch": 2.555256064690027, |
| "grad_norm": 0.5647293925285339, |
| "learning_rate": 0.00044722504047490553, |
| "loss": 3.5448, |
| "step": 23700 |
| }, |
| { |
| "epoch": 2.560646900269542, |
| "grad_norm": 0.6555283665657043, |
| "learning_rate": 0.0004469012412304371, |
| "loss": 3.5521, |
| "step": 23750 |
| }, |
| { |
| "epoch": 2.5660377358490565, |
| "grad_norm": 0.5926530361175537, |
| "learning_rate": 0.0004465774419859687, |
| "loss": 3.5618, |
| "step": 23800 |
| }, |
| { |
| "epoch": 2.571428571428571, |
| "grad_norm": 0.6275022029876709, |
| "learning_rate": 0.00044625364274150024, |
| "loss": 3.5548, |
| "step": 23850 |
| }, |
| { |
| "epoch": 2.5768194070080863, |
| "grad_norm": 0.5570287704467773, |
| "learning_rate": 0.00044592984349703184, |
| "loss": 3.5318, |
| "step": 23900 |
| }, |
| { |
| "epoch": 2.582210242587601, |
| "grad_norm": 0.5836941599845886, |
| "learning_rate": 0.00044560604425256334, |
| "loss": 3.5265, |
| "step": 23950 |
| }, |
| { |
| "epoch": 2.5876010781671157, |
| "grad_norm": 0.546286404132843, |
| "learning_rate": 0.000445282245008095, |
| "loss": 3.5598, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.5876010781671157, |
| "eval_accuracy": 0.36674695752899866, |
| "eval_loss": 3.528761386871338, |
| "eval_runtime": 184.0106, |
| "eval_samples_per_second": 97.88, |
| "eval_steps_per_second": 6.119, |
| "step": 24000 |
| }, |
| { |
| "epoch": 2.592991913746631, |
| "grad_norm": 0.5817551612854004, |
| "learning_rate": 0.0004449584457636265, |
| "loss": 3.554, |
| "step": 24050 |
| }, |
| { |
| "epoch": 2.5983827493261455, |
| "grad_norm": 0.6376777291297913, |
| "learning_rate": 0.00044463464651915805, |
| "loss": 3.5397, |
| "step": 24100 |
| }, |
| { |
| "epoch": 2.6037735849056602, |
| "grad_norm": 0.6055445671081543, |
| "learning_rate": 0.00044431084727468965, |
| "loss": 3.5507, |
| "step": 24150 |
| }, |
| { |
| "epoch": 2.6091644204851754, |
| "grad_norm": 0.5456017851829529, |
| "learning_rate": 0.0004439870480302212, |
| "loss": 3.526, |
| "step": 24200 |
| }, |
| { |
| "epoch": 2.61455525606469, |
| "grad_norm": 0.6268299221992493, |
| "learning_rate": 0.0004436632487857528, |
| "loss": 3.5493, |
| "step": 24250 |
| }, |
| { |
| "epoch": 2.6199460916442048, |
| "grad_norm": 0.6223170757293701, |
| "learning_rate": 0.00044333944954128436, |
| "loss": 3.5491, |
| "step": 24300 |
| }, |
| { |
| "epoch": 2.62533692722372, |
| "grad_norm": 0.580644428730011, |
| "learning_rate": 0.00044301565029681596, |
| "loss": 3.5434, |
| "step": 24350 |
| }, |
| { |
| "epoch": 2.6307277628032346, |
| "grad_norm": 0.5493316054344177, |
| "learning_rate": 0.0004426918510523475, |
| "loss": 3.5352, |
| "step": 24400 |
| }, |
| { |
| "epoch": 2.6361185983827493, |
| "grad_norm": 0.5446970462799072, |
| "learning_rate": 0.0004423680518078791, |
| "loss": 3.5151, |
| "step": 24450 |
| }, |
| { |
| "epoch": 2.641509433962264, |
| "grad_norm": 0.5663889646530151, |
| "learning_rate": 0.00044204425256341067, |
| "loss": 3.5352, |
| "step": 24500 |
| }, |
| { |
| "epoch": 2.6469002695417787, |
| "grad_norm": 0.5662615895271301, |
| "learning_rate": 0.0004417204533189422, |
| "loss": 3.5313, |
| "step": 24550 |
| }, |
| { |
| "epoch": 2.652291105121294, |
| "grad_norm": 0.5774913430213928, |
| "learning_rate": 0.0004413966540744738, |
| "loss": 3.5433, |
| "step": 24600 |
| }, |
| { |
| "epoch": 2.6576819407008085, |
| "grad_norm": 0.5947896838188171, |
| "learning_rate": 0.0004410728548300053, |
| "loss": 3.5317, |
| "step": 24650 |
| }, |
| { |
| "epoch": 2.6630727762803232, |
| "grad_norm": 0.5944833159446716, |
| "learning_rate": 0.0004407490555855369, |
| "loss": 3.519, |
| "step": 24700 |
| }, |
| { |
| "epoch": 2.6684636118598384, |
| "grad_norm": 0.6091347932815552, |
| "learning_rate": 0.0004404252563410685, |
| "loss": 3.5494, |
| "step": 24750 |
| }, |
| { |
| "epoch": 2.673854447439353, |
| "grad_norm": 0.6208272576332092, |
| "learning_rate": 0.0004401014570966001, |
| "loss": 3.5352, |
| "step": 24800 |
| }, |
| { |
| "epoch": 2.6792452830188678, |
| "grad_norm": 0.5658981204032898, |
| "learning_rate": 0.00043977765785213163, |
| "loss": 3.5444, |
| "step": 24850 |
| }, |
| { |
| "epoch": 2.684636118598383, |
| "grad_norm": 0.632360577583313, |
| "learning_rate": 0.0004394538586076632, |
| "loss": 3.5415, |
| "step": 24900 |
| }, |
| { |
| "epoch": 2.6900269541778976, |
| "grad_norm": 0.6172274351119995, |
| "learning_rate": 0.0004391300593631948, |
| "loss": 3.5341, |
| "step": 24950 |
| }, |
| { |
| "epoch": 2.6954177897574123, |
| "grad_norm": 0.5555505156517029, |
| "learning_rate": 0.00043880626011872634, |
| "loss": 3.533, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.6954177897574123, |
| "eval_accuracy": 0.3680360153900296, |
| "eval_loss": 3.5185306072235107, |
| "eval_runtime": 183.9436, |
| "eval_samples_per_second": 97.916, |
| "eval_steps_per_second": 6.121, |
| "step": 25000 |
| }, |
| { |
| "epoch": 2.7008086253369274, |
| "grad_norm": 0.6439934968948364, |
| "learning_rate": 0.00043848246087425794, |
| "loss": 3.5493, |
| "step": 25050 |
| }, |
| { |
| "epoch": 2.706199460916442, |
| "grad_norm": 0.6363284587860107, |
| "learning_rate": 0.0004381586616297895, |
| "loss": 3.5325, |
| "step": 25100 |
| }, |
| { |
| "epoch": 2.711590296495957, |
| "grad_norm": 0.6599604487419128, |
| "learning_rate": 0.0004378348623853211, |
| "loss": 3.5416, |
| "step": 25150 |
| }, |
| { |
| "epoch": 2.7169811320754715, |
| "grad_norm": 0.626559317111969, |
| "learning_rate": 0.00043751106314085265, |
| "loss": 3.5485, |
| "step": 25200 |
| }, |
| { |
| "epoch": 2.7223719676549867, |
| "grad_norm": 0.6159871816635132, |
| "learning_rate": 0.00043718726389638425, |
| "loss": 3.5455, |
| "step": 25250 |
| }, |
| { |
| "epoch": 2.7277628032345014, |
| "grad_norm": 0.6169484853744507, |
| "learning_rate": 0.00043686346465191575, |
| "loss": 3.5336, |
| "step": 25300 |
| }, |
| { |
| "epoch": 2.733153638814016, |
| "grad_norm": 0.6067425012588501, |
| "learning_rate": 0.0004365396654074473, |
| "loss": 3.5217, |
| "step": 25350 |
| }, |
| { |
| "epoch": 2.7385444743935308, |
| "grad_norm": 0.5563640594482422, |
| "learning_rate": 0.0004362158661629789, |
| "loss": 3.5439, |
| "step": 25400 |
| }, |
| { |
| "epoch": 2.743935309973046, |
| "grad_norm": 0.7104645371437073, |
| "learning_rate": 0.00043589206691851045, |
| "loss": 3.5473, |
| "step": 25450 |
| }, |
| { |
| "epoch": 2.7493261455525606, |
| "grad_norm": 0.5590559244155884, |
| "learning_rate": 0.00043556826767404206, |
| "loss": 3.5257, |
| "step": 25500 |
| }, |
| { |
| "epoch": 2.7547169811320753, |
| "grad_norm": 0.5929838418960571, |
| "learning_rate": 0.0004352444684295736, |
| "loss": 3.5374, |
| "step": 25550 |
| }, |
| { |
| "epoch": 2.7601078167115904, |
| "grad_norm": 0.6350831985473633, |
| "learning_rate": 0.0004349206691851052, |
| "loss": 3.5381, |
| "step": 25600 |
| }, |
| { |
| "epoch": 2.765498652291105, |
| "grad_norm": 0.7433158755302429, |
| "learning_rate": 0.00043459686994063676, |
| "loss": 3.5282, |
| "step": 25650 |
| }, |
| { |
| "epoch": 2.77088948787062, |
| "grad_norm": 0.6278929114341736, |
| "learning_rate": 0.00043427307069616837, |
| "loss": 3.5368, |
| "step": 25700 |
| }, |
| { |
| "epoch": 2.776280323450135, |
| "grad_norm": 0.6056564450263977, |
| "learning_rate": 0.0004339492714516999, |
| "loss": 3.5388, |
| "step": 25750 |
| }, |
| { |
| "epoch": 2.7816711590296497, |
| "grad_norm": 0.6050384640693665, |
| "learning_rate": 0.00043362547220723147, |
| "loss": 3.5446, |
| "step": 25800 |
| }, |
| { |
| "epoch": 2.7870619946091644, |
| "grad_norm": 0.5863341689109802, |
| "learning_rate": 0.0004333016729627631, |
| "loss": 3.5392, |
| "step": 25850 |
| }, |
| { |
| "epoch": 2.7924528301886795, |
| "grad_norm": 0.6051446795463562, |
| "learning_rate": 0.0004329778737182946, |
| "loss": 3.5314, |
| "step": 25900 |
| }, |
| { |
| "epoch": 2.797843665768194, |
| "grad_norm": 0.5601605772972107, |
| "learning_rate": 0.00043265407447382623, |
| "loss": 3.5512, |
| "step": 25950 |
| }, |
| { |
| "epoch": 2.803234501347709, |
| "grad_norm": 0.6006129384040833, |
| "learning_rate": 0.0004323302752293577, |
| "loss": 3.5272, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.803234501347709, |
| "eval_accuracy": 0.3687935433238722, |
| "eval_loss": 3.511807441711426, |
| "eval_runtime": 183.9921, |
| "eval_samples_per_second": 97.89, |
| "eval_steps_per_second": 6.12, |
| "step": 26000 |
| }, |
| { |
| "epoch": 2.8086253369272236, |
| "grad_norm": 0.5731135606765747, |
| "learning_rate": 0.00043200647598488933, |
| "loss": 3.5367, |
| "step": 26050 |
| }, |
| { |
| "epoch": 2.8140161725067383, |
| "grad_norm": 0.5950818061828613, |
| "learning_rate": 0.0004316826767404209, |
| "loss": 3.5272, |
| "step": 26100 |
| }, |
| { |
| "epoch": 2.8194070080862534, |
| "grad_norm": 0.6465256810188293, |
| "learning_rate": 0.00043136535348084183, |
| "loss": 3.5273, |
| "step": 26150 |
| }, |
| { |
| "epoch": 2.824797843665768, |
| "grad_norm": 0.6396045684814453, |
| "learning_rate": 0.0004310415542363734, |
| "loss": 3.5309, |
| "step": 26200 |
| }, |
| { |
| "epoch": 2.830188679245283, |
| "grad_norm": 0.575483500957489, |
| "learning_rate": 0.000430717754991905, |
| "loss": 3.5388, |
| "step": 26250 |
| }, |
| { |
| "epoch": 2.835579514824798, |
| "grad_norm": 0.5966452360153198, |
| "learning_rate": 0.00043039395574743654, |
| "loss": 3.523, |
| "step": 26300 |
| }, |
| { |
| "epoch": 2.8409703504043127, |
| "grad_norm": 0.5425668954849243, |
| "learning_rate": 0.00043007015650296814, |
| "loss": 3.5356, |
| "step": 26350 |
| }, |
| { |
| "epoch": 2.8463611859838274, |
| "grad_norm": 0.587317943572998, |
| "learning_rate": 0.0004297463572584997, |
| "loss": 3.5102, |
| "step": 26400 |
| }, |
| { |
| "epoch": 2.8517520215633425, |
| "grad_norm": 0.578007161617279, |
| "learning_rate": 0.0004294225580140313, |
| "loss": 3.542, |
| "step": 26450 |
| }, |
| { |
| "epoch": 2.857142857142857, |
| "grad_norm": 0.5723671913146973, |
| "learning_rate": 0.00042909875876956285, |
| "loss": 3.5329, |
| "step": 26500 |
| }, |
| { |
| "epoch": 2.862533692722372, |
| "grad_norm": 0.5588173866271973, |
| "learning_rate": 0.0004287749595250944, |
| "loss": 3.5404, |
| "step": 26550 |
| }, |
| { |
| "epoch": 2.867924528301887, |
| "grad_norm": 0.5535792112350464, |
| "learning_rate": 0.000428451160280626, |
| "loss": 3.5266, |
| "step": 26600 |
| }, |
| { |
| "epoch": 2.8733153638814017, |
| "grad_norm": 0.573268711566925, |
| "learning_rate": 0.0004281273610361575, |
| "loss": 3.5303, |
| "step": 26650 |
| }, |
| { |
| "epoch": 2.8787061994609164, |
| "grad_norm": 0.5750014185905457, |
| "learning_rate": 0.0004278035617916891, |
| "loss": 3.5032, |
| "step": 26700 |
| }, |
| { |
| "epoch": 2.884097035040431, |
| "grad_norm": 0.6313278079032898, |
| "learning_rate": 0.00042747976254722066, |
| "loss": 3.5281, |
| "step": 26750 |
| }, |
| { |
| "epoch": 2.889487870619946, |
| "grad_norm": 0.5894135236740112, |
| "learning_rate": 0.00042715596330275226, |
| "loss": 3.5066, |
| "step": 26800 |
| }, |
| { |
| "epoch": 2.894878706199461, |
| "grad_norm": 0.5986575484275818, |
| "learning_rate": 0.0004268321640582838, |
| "loss": 3.5441, |
| "step": 26850 |
| }, |
| { |
| "epoch": 2.9002695417789757, |
| "grad_norm": 0.5755992531776428, |
| "learning_rate": 0.0004265083648138154, |
| "loss": 3.5257, |
| "step": 26900 |
| }, |
| { |
| "epoch": 2.9056603773584904, |
| "grad_norm": 0.6677845120429993, |
| "learning_rate": 0.00042618456556934697, |
| "loss": 3.544, |
| "step": 26950 |
| }, |
| { |
| "epoch": 2.9110512129380055, |
| "grad_norm": 0.6155284643173218, |
| "learning_rate": 0.0004258607663248785, |
| "loss": 3.5211, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.9110512129380055, |
| "eval_accuracy": 0.3694859881779139, |
| "eval_loss": 3.50223445892334, |
| "eval_runtime": 183.9135, |
| "eval_samples_per_second": 97.932, |
| "eval_steps_per_second": 6.122, |
| "step": 27000 |
| }, |
| { |
| "epoch": 2.91644204851752, |
| "grad_norm": 0.6333293318748474, |
| "learning_rate": 0.0004255369670804101, |
| "loss": 3.5293, |
| "step": 27050 |
| }, |
| { |
| "epoch": 2.921832884097035, |
| "grad_norm": 0.5835155248641968, |
| "learning_rate": 0.0004252131678359417, |
| "loss": 3.5198, |
| "step": 27100 |
| }, |
| { |
| "epoch": 2.92722371967655, |
| "grad_norm": 0.5927255749702454, |
| "learning_rate": 0.0004248893685914733, |
| "loss": 3.5244, |
| "step": 27150 |
| }, |
| { |
| "epoch": 2.9326145552560647, |
| "grad_norm": 0.6329957842826843, |
| "learning_rate": 0.00042456556934700483, |
| "loss": 3.5264, |
| "step": 27200 |
| }, |
| { |
| "epoch": 2.9380053908355794, |
| "grad_norm": 0.6049855351448059, |
| "learning_rate": 0.00042424177010253643, |
| "loss": 3.5223, |
| "step": 27250 |
| }, |
| { |
| "epoch": 2.9433962264150946, |
| "grad_norm": 0.5478748679161072, |
| "learning_rate": 0.00042391797085806793, |
| "loss": 3.52, |
| "step": 27300 |
| }, |
| { |
| "epoch": 2.9487870619946093, |
| "grad_norm": 0.599579930305481, |
| "learning_rate": 0.0004235941716135995, |
| "loss": 3.5254, |
| "step": 27350 |
| }, |
| { |
| "epoch": 2.954177897574124, |
| "grad_norm": 0.597079873085022, |
| "learning_rate": 0.0004232703723691311, |
| "loss": 3.5192, |
| "step": 27400 |
| }, |
| { |
| "epoch": 2.9595687331536387, |
| "grad_norm": 0.5469141602516174, |
| "learning_rate": 0.00042294657312466264, |
| "loss": 3.5294, |
| "step": 27450 |
| }, |
| { |
| "epoch": 2.964959568733154, |
| "grad_norm": 0.6046788692474365, |
| "learning_rate": 0.00042262277388019424, |
| "loss": 3.5236, |
| "step": 27500 |
| }, |
| { |
| "epoch": 2.9703504043126685, |
| "grad_norm": 0.5884106159210205, |
| "learning_rate": 0.0004222989746357258, |
| "loss": 3.524, |
| "step": 27550 |
| }, |
| { |
| "epoch": 2.975741239892183, |
| "grad_norm": 0.634127676486969, |
| "learning_rate": 0.0004219751753912574, |
| "loss": 3.5287, |
| "step": 27600 |
| }, |
| { |
| "epoch": 2.981132075471698, |
| "grad_norm": 0.5886887311935425, |
| "learning_rate": 0.00042165137614678895, |
| "loss": 3.5248, |
| "step": 27650 |
| }, |
| { |
| "epoch": 2.986522911051213, |
| "grad_norm": 0.5682351589202881, |
| "learning_rate": 0.00042132757690232055, |
| "loss": 3.5282, |
| "step": 27700 |
| }, |
| { |
| "epoch": 2.9919137466307277, |
| "grad_norm": 0.5956007242202759, |
| "learning_rate": 0.0004210037776578521, |
| "loss": 3.5301, |
| "step": 27750 |
| }, |
| { |
| "epoch": 2.9973045822102424, |
| "grad_norm": 0.6513469815254211, |
| "learning_rate": 0.00042067997841338365, |
| "loss": 3.499, |
| "step": 27800 |
| }, |
| { |
| "epoch": 3.0026954177897576, |
| "grad_norm": 0.5566192865371704, |
| "learning_rate": 0.00042035617916891526, |
| "loss": 3.4543, |
| "step": 27850 |
| }, |
| { |
| "epoch": 3.0080862533692723, |
| "grad_norm": 0.6176077127456665, |
| "learning_rate": 0.0004200323799244468, |
| "loss": 3.423, |
| "step": 27900 |
| }, |
| { |
| "epoch": 3.013477088948787, |
| "grad_norm": 0.6107637286186218, |
| "learning_rate": 0.0004197085806799784, |
| "loss": 3.4323, |
| "step": 27950 |
| }, |
| { |
| "epoch": 3.018867924528302, |
| "grad_norm": 0.6046628355979919, |
| "learning_rate": 0.0004193847814355099, |
| "loss": 3.4424, |
| "step": 28000 |
| }, |
| { |
| "epoch": 3.018867924528302, |
| "eval_accuracy": 0.37057208244450746, |
| "eval_loss": 3.4983136653900146, |
| "eval_runtime": 183.6916, |
| "eval_samples_per_second": 98.05, |
| "eval_steps_per_second": 6.13, |
| "step": 28000 |
| }, |
| { |
| "epoch": 3.024258760107817, |
| "grad_norm": 0.6203792095184326, |
| "learning_rate": 0.0004190609821910415, |
| "loss": 3.4246, |
| "step": 28050 |
| }, |
| { |
| "epoch": 3.0296495956873315, |
| "grad_norm": 0.5849449038505554, |
| "learning_rate": 0.00041873718294657306, |
| "loss": 3.413, |
| "step": 28100 |
| }, |
| { |
| "epoch": 3.035040431266846, |
| "grad_norm": 0.6401404142379761, |
| "learning_rate": 0.00041841338370210467, |
| "loss": 3.4293, |
| "step": 28150 |
| }, |
| { |
| "epoch": 3.0404312668463613, |
| "grad_norm": 0.5675318241119385, |
| "learning_rate": 0.00041809606044252557, |
| "loss": 3.4324, |
| "step": 28200 |
| }, |
| { |
| "epoch": 3.045822102425876, |
| "grad_norm": 0.5723515152931213, |
| "learning_rate": 0.00041777226119805717, |
| "loss": 3.4211, |
| "step": 28250 |
| }, |
| { |
| "epoch": 3.0512129380053907, |
| "grad_norm": 0.5614615678787231, |
| "learning_rate": 0.0004174484619535887, |
| "loss": 3.4325, |
| "step": 28300 |
| }, |
| { |
| "epoch": 3.056603773584906, |
| "grad_norm": 0.5907319784164429, |
| "learning_rate": 0.00041712466270912033, |
| "loss": 3.4343, |
| "step": 28350 |
| }, |
| { |
| "epoch": 3.0619946091644206, |
| "grad_norm": 0.6139299869537354, |
| "learning_rate": 0.0004168008634646519, |
| "loss": 3.4437, |
| "step": 28400 |
| }, |
| { |
| "epoch": 3.0673854447439353, |
| "grad_norm": 0.6088677048683167, |
| "learning_rate": 0.0004164770642201835, |
| "loss": 3.4369, |
| "step": 28450 |
| }, |
| { |
| "epoch": 3.07277628032345, |
| "grad_norm": 0.6208965182304382, |
| "learning_rate": 0.00041615326497571503, |
| "loss": 3.4344, |
| "step": 28500 |
| }, |
| { |
| "epoch": 3.078167115902965, |
| "grad_norm": 0.6358165144920349, |
| "learning_rate": 0.0004158294657312466, |
| "loss": 3.4428, |
| "step": 28550 |
| }, |
| { |
| "epoch": 3.08355795148248, |
| "grad_norm": 0.6254170536994934, |
| "learning_rate": 0.0004155056664867782, |
| "loss": 3.4463, |
| "step": 28600 |
| }, |
| { |
| "epoch": 3.0889487870619945, |
| "grad_norm": 0.585701584815979, |
| "learning_rate": 0.0004151818672423097, |
| "loss": 3.4471, |
| "step": 28650 |
| }, |
| { |
| "epoch": 3.0943396226415096, |
| "grad_norm": 0.6081207394599915, |
| "learning_rate": 0.0004148580679978413, |
| "loss": 3.4386, |
| "step": 28700 |
| }, |
| { |
| "epoch": 3.0997304582210243, |
| "grad_norm": 0.6378210186958313, |
| "learning_rate": 0.00041453426875337284, |
| "loss": 3.4258, |
| "step": 28750 |
| }, |
| { |
| "epoch": 3.105121293800539, |
| "grad_norm": 0.6246366500854492, |
| "learning_rate": 0.00041421046950890445, |
| "loss": 3.4211, |
| "step": 28800 |
| }, |
| { |
| "epoch": 3.1105121293800537, |
| "grad_norm": 0.5723994374275208, |
| "learning_rate": 0.000413886670264436, |
| "loss": 3.4276, |
| "step": 28850 |
| }, |
| { |
| "epoch": 3.115902964959569, |
| "grad_norm": 0.6091293692588806, |
| "learning_rate": 0.0004135628710199676, |
| "loss": 3.4424, |
| "step": 28900 |
| }, |
| { |
| "epoch": 3.1212938005390836, |
| "grad_norm": 0.5650001168251038, |
| "learning_rate": 0.00041323907177549915, |
| "loss": 3.4413, |
| "step": 28950 |
| }, |
| { |
| "epoch": 3.1266846361185983, |
| "grad_norm": 0.6242942214012146, |
| "learning_rate": 0.0004129152725310307, |
| "loss": 3.4427, |
| "step": 29000 |
| }, |
| { |
| "epoch": 3.1266846361185983, |
| "eval_accuracy": 0.37073506177622917, |
| "eval_loss": 3.4975240230560303, |
| "eval_runtime": 183.9334, |
| "eval_samples_per_second": 97.921, |
| "eval_steps_per_second": 6.122, |
| "step": 29000 |
| }, |
| { |
| "epoch": 3.1320754716981134, |
| "grad_norm": 0.5982543230056763, |
| "learning_rate": 0.0004125914732865623, |
| "loss": 3.4407, |
| "step": 29050 |
| }, |
| { |
| "epoch": 3.137466307277628, |
| "grad_norm": 0.584873616695404, |
| "learning_rate": 0.00041226767404209386, |
| "loss": 3.4539, |
| "step": 29100 |
| }, |
| { |
| "epoch": 3.142857142857143, |
| "grad_norm": 0.6070249676704407, |
| "learning_rate": 0.00041194387479762546, |
| "loss": 3.4543, |
| "step": 29150 |
| }, |
| { |
| "epoch": 3.1482479784366575, |
| "grad_norm": 0.6300134658813477, |
| "learning_rate": 0.000411620075553157, |
| "loss": 3.4457, |
| "step": 29200 |
| }, |
| { |
| "epoch": 3.1536388140161726, |
| "grad_norm": 0.6087648868560791, |
| "learning_rate": 0.0004112962763086886, |
| "loss": 3.4461, |
| "step": 29250 |
| }, |
| { |
| "epoch": 3.1590296495956873, |
| "grad_norm": 0.6131951808929443, |
| "learning_rate": 0.0004109724770642201, |
| "loss": 3.4512, |
| "step": 29300 |
| }, |
| { |
| "epoch": 3.164420485175202, |
| "grad_norm": 0.6422128677368164, |
| "learning_rate": 0.00041064867781975177, |
| "loss": 3.4648, |
| "step": 29350 |
| }, |
| { |
| "epoch": 3.169811320754717, |
| "grad_norm": 0.7360274791717529, |
| "learning_rate": 0.00041032487857528327, |
| "loss": 3.4428, |
| "step": 29400 |
| }, |
| { |
| "epoch": 3.175202156334232, |
| "grad_norm": 0.5676162838935852, |
| "learning_rate": 0.0004100010793308148, |
| "loss": 3.4475, |
| "step": 29450 |
| }, |
| { |
| "epoch": 3.1805929919137466, |
| "grad_norm": 0.5553027987480164, |
| "learning_rate": 0.0004096772800863464, |
| "loss": 3.4486, |
| "step": 29500 |
| }, |
| { |
| "epoch": 3.1859838274932613, |
| "grad_norm": 0.6182715892791748, |
| "learning_rate": 0.000409353480841878, |
| "loss": 3.4488, |
| "step": 29550 |
| }, |
| { |
| "epoch": 3.1913746630727764, |
| "grad_norm": 0.5933767557144165, |
| "learning_rate": 0.0004090296815974096, |
| "loss": 3.4744, |
| "step": 29600 |
| }, |
| { |
| "epoch": 3.196765498652291, |
| "grad_norm": 0.5704579949378967, |
| "learning_rate": 0.00040870588235294113, |
| "loss": 3.4456, |
| "step": 29650 |
| }, |
| { |
| "epoch": 3.202156334231806, |
| "grad_norm": 0.6044793128967285, |
| "learning_rate": 0.00040838208310847273, |
| "loss": 3.455, |
| "step": 29700 |
| }, |
| { |
| "epoch": 3.207547169811321, |
| "grad_norm": 0.5789558291435242, |
| "learning_rate": 0.0004080582838640043, |
| "loss": 3.4687, |
| "step": 29750 |
| }, |
| { |
| "epoch": 3.2129380053908356, |
| "grad_norm": 0.6031609177589417, |
| "learning_rate": 0.0004077344846195359, |
| "loss": 3.4433, |
| "step": 29800 |
| }, |
| { |
| "epoch": 3.2183288409703503, |
| "grad_norm": 0.6147695183753967, |
| "learning_rate": 0.00040741068537506744, |
| "loss": 3.4398, |
| "step": 29850 |
| }, |
| { |
| "epoch": 3.223719676549865, |
| "grad_norm": 0.6561545133590698, |
| "learning_rate": 0.000407086886130599, |
| "loss": 3.4364, |
| "step": 29900 |
| }, |
| { |
| "epoch": 3.22911051212938, |
| "grad_norm": 0.6324929594993591, |
| "learning_rate": 0.0004067630868861306, |
| "loss": 3.4486, |
| "step": 29950 |
| }, |
| { |
| "epoch": 3.234501347708895, |
| "grad_norm": 0.6108676195144653, |
| "learning_rate": 0.0004064392876416621, |
| "loss": 3.4427, |
| "step": 30000 |
| }, |
| { |
| "epoch": 3.234501347708895, |
| "eval_accuracy": 0.3712915818676149, |
| "eval_loss": 3.490433931350708, |
| "eval_runtime": 183.7094, |
| "eval_samples_per_second": 98.041, |
| "eval_steps_per_second": 6.129, |
| "step": 30000 |
| } |
| ], |
| "logging_steps": 50, |
| "max_steps": 92750, |
| "num_input_tokens_seen": 0, |
| "num_train_epochs": 10, |
| "save_steps": 10000, |
| "stateful_callbacks": { |
| "TrainerControl": { |
| "args": { |
| "should_epoch_stop": false, |
| "should_evaluate": false, |
| "should_log": false, |
| "should_save": true, |
| "should_training_stop": false |
| }, |
| "attributes": {} |
| } |
| }, |
| "total_flos": 2.508207538176e+17, |
| "train_batch_size": 32, |
| "trial_name": null, |
| "trial_params": null |
| } |
|
|
