| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.991735537190083, | |
| "eval_steps": 500, | |
| "global_step": 980, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.02542911633820725, | |
| "grad_norm": 0.7163803577423096, | |
| "learning_rate": 4.9996788644998285e-05, | |
| "loss": 1.6242, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.0508582326764145, | |
| "grad_norm": 0.5585092306137085, | |
| "learning_rate": 4.99871554050172e-05, | |
| "loss": 1.4834, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07628734901462174, | |
| "grad_norm": 0.3108583688735962, | |
| "learning_rate": 4.997110275491702e-05, | |
| "loss": 1.3901, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.101716465352829, | |
| "grad_norm": 0.20925107598304749, | |
| "learning_rate": 4.994863481875841e-05, | |
| "loss": 1.2635, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.12714558169103624, | |
| "grad_norm": 0.2436874657869339, | |
| "learning_rate": 4.991975736874289e-05, | |
| "loss": 1.2457, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.15257469802924348, | |
| "grad_norm": 0.19612200558185577, | |
| "learning_rate": 4.9884477823729956e-05, | |
| "loss": 1.2602, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.17800381436745072, | |
| "grad_norm": 0.1591530740261078, | |
| "learning_rate": 4.984280524733107e-05, | |
| "loss": 1.2238, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.203432930705658, | |
| "grad_norm": 0.15396293997764587, | |
| "learning_rate": 4.979475034558115e-05, | |
| "loss": 1.2456, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.22886204704386523, | |
| "grad_norm": 0.17088189721107483, | |
| "learning_rate": 4.9740325464188154e-05, | |
| "loss": 1.2554, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.25429116338207247, | |
| "grad_norm": 0.14276106655597687, | |
| "learning_rate": 4.967954458536126e-05, | |
| "loss": 1.248, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.27972027972027974, | |
| "grad_norm": 0.13971547782421112, | |
| "learning_rate": 4.961242332421882e-05, | |
| "loss": 1.1816, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.30514939605848695, | |
| "grad_norm": 0.14659512042999268, | |
| "learning_rate": 4.9538978924776634e-05, | |
| "loss": 1.1962, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.3305785123966942, | |
| "grad_norm": 0.1615353226661682, | |
| "learning_rate": 4.945923025551789e-05, | |
| "loss": 1.1989, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.35600762873490144, | |
| "grad_norm": 0.1626625955104828, | |
| "learning_rate": 4.937319780454559e-05, | |
| "loss": 1.2023, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.3814367450731087, | |
| "grad_norm": 0.1588200479745865, | |
| "learning_rate": 4.92809036743191e-05, | |
| "loss": 1.2328, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.406865861411316, | |
| "grad_norm": 0.1648840457201004, | |
| "learning_rate": 4.9182371575975736e-05, | |
| "loss": 1.1974, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.4322949777495232, | |
| "grad_norm": 0.16755953431129456, | |
| "learning_rate": 4.907762682323927e-05, | |
| "loss": 1.1876, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.45772409408773046, | |
| "grad_norm": 0.19065892696380615, | |
| "learning_rate": 4.8966696325916515e-05, | |
| "loss": 1.2396, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.4831532104259377, | |
| "grad_norm": 0.18359169363975525, | |
| "learning_rate": 4.8849608582984096e-05, | |
| "loss": 1.1743, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.5085823267641449, | |
| "grad_norm": 0.18455757200717926, | |
| "learning_rate": 4.8726393675266716e-05, | |
| "loss": 1.1566, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5340114431023522, | |
| "grad_norm": 0.18412873148918152, | |
| "learning_rate": 4.8597083257709194e-05, | |
| "loss": 1.1998, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.5594405594405595, | |
| "grad_norm": 0.2001655399799347, | |
| "learning_rate": 4.846171055124401e-05, | |
| "loss": 1.172, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.5848696757787667, | |
| "grad_norm": 0.1914481222629547, | |
| "learning_rate": 4.832031033425662e-05, | |
| "loss": 1.1514, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.6102987921169739, | |
| "grad_norm": 0.19641566276550293, | |
| "learning_rate": 4.817291893365055e-05, | |
| "loss": 1.1887, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6357279084551812, | |
| "grad_norm": 0.20107164978981018, | |
| "learning_rate": 4.8019574215514705e-05, | |
| "loss": 1.1932, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.6611570247933884, | |
| "grad_norm": 0.21638870239257812, | |
| "learning_rate": 4.7860315575395316e-05, | |
| "loss": 1.166, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.6865861411315957, | |
| "grad_norm": 0.22959065437316895, | |
| "learning_rate": 4.7695183928174804e-05, | |
| "loss": 1.1645, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.7120152574698029, | |
| "grad_norm": 0.2240159660577774, | |
| "learning_rate": 4.752422169756048e-05, | |
| "loss": 1.167, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7374443738080102, | |
| "grad_norm": 0.22989672422409058, | |
| "learning_rate": 4.734747280518549e-05, | |
| "loss": 1.187, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.7628734901462174, | |
| "grad_norm": 0.22513297200202942, | |
| "learning_rate": 4.716498265932501e-05, | |
| "loss": 1.18, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.7883026064844246, | |
| "grad_norm": 0.23522530496120453, | |
| "learning_rate": 4.6976798143230434e-05, | |
| "loss": 1.1532, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.813731722822632, | |
| "grad_norm": 0.2377459853887558, | |
| "learning_rate": 4.678296760308474e-05, | |
| "loss": 1.1726, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8391608391608392, | |
| "grad_norm": 0.25356563925743103, | |
| "learning_rate": 4.658354083558188e-05, | |
| "loss": 1.2446, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.8645899554990464, | |
| "grad_norm": 0.24188080430030823, | |
| "learning_rate": 4.637856907513366e-05, | |
| "loss": 1.1905, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.8900190718372537, | |
| "grad_norm": 0.2587025761604309, | |
| "learning_rate": 4.6168104980707107e-05, | |
| "loss": 1.1594, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.9154481881754609, | |
| "grad_norm": 0.25419944524765015, | |
| "learning_rate": 4.595220262229601e-05, | |
| "loss": 1.1709, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.9408773045136681, | |
| "grad_norm": 0.2684890031814575, | |
| "learning_rate": 4.573091746702988e-05, | |
| "loss": 1.1936, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.9663064208518753, | |
| "grad_norm": 0.2676820755004883, | |
| "learning_rate": 4.55043063649239e-05, | |
| "loss": 1.1692, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.9917355371900827, | |
| "grad_norm": 0.27281874418258667, | |
| "learning_rate": 4.5272427534273774e-05, | |
| "loss": 1.1812, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 1.0190718372536554, | |
| "grad_norm": 0.26799142360687256, | |
| "learning_rate": 4.503534054669892e-05, | |
| "loss": 1.3447, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0445009535918626, | |
| "grad_norm": 0.29342585802078247, | |
| "learning_rate": 4.4793106311838e-05, | |
| "loss": 1.1098, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 1.06993006993007, | |
| "grad_norm": 0.29103586077690125, | |
| "learning_rate": 4.454578706170075e-05, | |
| "loss": 1.1913, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.0953591862682772, | |
| "grad_norm": 0.28619349002838135, | |
| "learning_rate": 4.429344633468004e-05, | |
| "loss": 1.1936, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 1.1207883026064844, | |
| "grad_norm": 0.3098839521408081, | |
| "learning_rate": 4.4036148959228365e-05, | |
| "loss": 1.1187, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.1462174189446916, | |
| "grad_norm": 0.30148303508758545, | |
| "learning_rate": 4.377396103720278e-05, | |
| "loss": 1.157, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 1.1716465352828989, | |
| "grad_norm": 0.31909579038619995, | |
| "learning_rate": 4.350694992688289e-05, | |
| "loss": 1.1618, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.197075651621106, | |
| "grad_norm": 0.327538400888443, | |
| "learning_rate": 4.323518422566586e-05, | |
| "loss": 1.1245, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 1.2225047679593135, | |
| "grad_norm": 0.334259033203125, | |
| "learning_rate": 4.2958733752443195e-05, | |
| "loss": 1.1108, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.2479338842975207, | |
| "grad_norm": 0.3590494692325592, | |
| "learning_rate": 4.267766952966369e-05, | |
| "loss": 1.1006, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 1.273363000635728, | |
| "grad_norm": 0.32646894454956055, | |
| "learning_rate": 4.239206376508717e-05, | |
| "loss": 1.1424, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.2987921169739352, | |
| "grad_norm": 0.3532828688621521, | |
| "learning_rate": 4.210198983323366e-05, | |
| "loss": 1.141, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 1.3242212333121424, | |
| "grad_norm": 0.33540236949920654, | |
| "learning_rate": 4.180752225653292e-05, | |
| "loss": 1.0976, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.3496503496503496, | |
| "grad_norm": 0.3587055504322052, | |
| "learning_rate": 4.150873668617898e-05, | |
| "loss": 1.1019, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 1.375079465988557, | |
| "grad_norm": 0.35621777176856995, | |
| "learning_rate": 4.120570988269472e-05, | |
| "loss": 1.1075, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.400508582326764, | |
| "grad_norm": 0.3648779094219208, | |
| "learning_rate": 4.089851969621138e-05, | |
| "loss": 1.1422, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 1.4259376986649714, | |
| "grad_norm": 0.3523823916912079, | |
| "learning_rate": 4.058724504646834e-05, | |
| "loss": 1.115, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.4513668150031787, | |
| "grad_norm": 0.36515510082244873, | |
| "learning_rate": 4.027196590253786e-05, | |
| "loss": 1.1293, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 1.4767959313413859, | |
| "grad_norm": 0.37401697039604187, | |
| "learning_rate": 3.9952763262280405e-05, | |
| "loss": 1.1186, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.502225047679593, | |
| "grad_norm": 0.3709016442298889, | |
| "learning_rate": 3.9629719131535594e-05, | |
| "loss": 1.149, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 1.5276541640178003, | |
| "grad_norm": 0.38189002871513367, | |
| "learning_rate": 3.9302916503054246e-05, | |
| "loss": 1.1267, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.5530832803560077, | |
| "grad_norm": 0.36769306659698486, | |
| "learning_rate": 3.897243933517679e-05, | |
| "loss": 1.108, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 1.5785123966942147, | |
| "grad_norm": 0.398843377828598, | |
| "learning_rate": 3.8638372530263715e-05, | |
| "loss": 1.1324, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.6039415130324222, | |
| "grad_norm": 0.37085267901420593, | |
| "learning_rate": 3.830080191288342e-05, | |
| "loss": 1.1051, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 1.6293706293706294, | |
| "grad_norm": 0.39182424545288086, | |
| "learning_rate": 3.7959814207763135e-05, | |
| "loss": 1.1622, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.6547997457088366, | |
| "grad_norm": 0.3733777105808258, | |
| "learning_rate": 3.761549701750865e-05, | |
| "loss": 1.1146, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 1.680228862047044, | |
| "grad_norm": 0.3874206840991974, | |
| "learning_rate": 3.726793880009845e-05, | |
| "loss": 1.0812, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.705657978385251, | |
| "grad_norm": 0.4049334228038788, | |
| "learning_rate": 3.6917228846158134e-05, | |
| "loss": 1.1379, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 1.7310870947234585, | |
| "grad_norm": 0.4189164340496063, | |
| "learning_rate": 3.656345725602089e-05, | |
| "loss": 1.1183, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.7565162110616654, | |
| "grad_norm": 0.38738593459129333, | |
| "learning_rate": 3.620671491657993e-05, | |
| "loss": 1.1094, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 1.7819453273998729, | |
| "grad_norm": 0.40121838450431824, | |
| "learning_rate": 3.5847093477938956e-05, | |
| "loss": 1.1048, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.80737444373808, | |
| "grad_norm": 0.42995157837867737, | |
| "learning_rate": 3.5484685329866425e-05, | |
| "loss": 1.1209, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 1.8328035600762873, | |
| "grad_norm": 0.4180549085140228, | |
| "learning_rate": 3.5119583578059846e-05, | |
| "loss": 1.1284, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.8582326764144947, | |
| "grad_norm": 0.40642648935317993, | |
| "learning_rate": 3.475188202022617e-05, | |
| "loss": 1.1239, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 1.8836617927527017, | |
| "grad_norm": 0.4026438593864441, | |
| "learning_rate": 3.438167512198436e-05, | |
| "loss": 1.1068, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.9090909090909092, | |
| "grad_norm": 0.4229359030723572, | |
| "learning_rate": 3.400905799259634e-05, | |
| "loss": 1.1046, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 1.9345200254291164, | |
| "grad_norm": 0.4220713973045349, | |
| "learning_rate": 3.363412636053269e-05, | |
| "loss": 1.1127, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.9599491417673236, | |
| "grad_norm": 0.40782222151756287, | |
| "learning_rate": 3.3256976548879184e-05, | |
| "loss": 1.117, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 1.9853782581055308, | |
| "grad_norm": 0.43011417984962463, | |
| "learning_rate": 3.2877705450590526e-05, | |
| "loss": 1.1105, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.0127145581691037, | |
| "grad_norm": 0.4370057284832001, | |
| "learning_rate": 3.249641050359779e-05, | |
| "loss": 1.2868, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 2.0381436745073107, | |
| "grad_norm": 0.42571040987968445, | |
| "learning_rate": 3.211318966577581e-05, | |
| "loss": 1.082, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.063572790845518, | |
| "grad_norm": 0.45124417543411255, | |
| "learning_rate": 3.172814138977692e-05, | |
| "loss": 1.0788, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 2.089001907183725, | |
| "grad_norm": 0.4748118817806244, | |
| "learning_rate": 3.1341364597737686e-05, | |
| "loss": 1.0658, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.1144310235219326, | |
| "grad_norm": 0.44493603706359863, | |
| "learning_rate": 3.0952958655864955e-05, | |
| "loss": 1.0914, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 2.13986013986014, | |
| "grad_norm": 0.4867158532142639, | |
| "learning_rate": 3.056302334890786e-05, | |
| "loss": 1.0719, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.165289256198347, | |
| "grad_norm": 0.494890034198761, | |
| "learning_rate": 3.0171658854522273e-05, | |
| "loss": 1.0929, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 2.1907183725365544, | |
| "grad_norm": 0.489145964384079, | |
| "learning_rate": 2.9778965717534313e-05, | |
| "loss": 1.0833, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.2161474888747614, | |
| "grad_norm": 0.4823625087738037, | |
| "learning_rate": 2.9385044824109543e-05, | |
| "loss": 1.0836, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 2.241576605212969, | |
| "grad_norm": 0.5059676170349121, | |
| "learning_rate": 2.8989997375834482e-05, | |
| "loss": 1.0492, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.2670057215511763, | |
| "grad_norm": 0.5108452439308167, | |
| "learning_rate": 2.8593924863717046e-05, | |
| "loss": 1.0746, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 2.2924348378893833, | |
| "grad_norm": 0.516611635684967, | |
| "learning_rate": 2.8196929042112652e-05, | |
| "loss": 1.0972, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.3178639542275907, | |
| "grad_norm": 0.5169530510902405, | |
| "learning_rate": 2.7799111902582696e-05, | |
| "loss": 1.0524, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 2.3432930705657977, | |
| "grad_norm": 0.5156260132789612, | |
| "learning_rate": 2.7400575647692046e-05, | |
| "loss": 1.0402, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.368722186904005, | |
| "grad_norm": 0.5396901965141296, | |
| "learning_rate": 2.7001422664752333e-05, | |
| "loss": 1.0956, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 2.394151303242212, | |
| "grad_norm": 0.5307148694992065, | |
| "learning_rate": 2.6601755499517826e-05, | |
| "loss": 1.0803, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.4195804195804196, | |
| "grad_norm": 0.5289092063903809, | |
| "learning_rate": 2.620167682984052e-05, | |
| "loss": 1.0657, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 2.445009535918627, | |
| "grad_norm": 0.5314033031463623, | |
| "learning_rate": 2.5801289439291388e-05, | |
| "loss": 1.07, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.470438652256834, | |
| "grad_norm": 0.5466395616531372, | |
| "learning_rate": 2.540069619075435e-05, | |
| "loss": 1.0835, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 2.4958677685950414, | |
| "grad_norm": 0.5496529340744019, | |
| "learning_rate": 2.5e-05, | |
| "loss": 1.0571, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.5212968849332484, | |
| "grad_norm": 0.5266883373260498, | |
| "learning_rate": 2.459930380924566e-05, | |
| "loss": 1.0984, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 2.546726001271456, | |
| "grad_norm": 0.5321890711784363, | |
| "learning_rate": 2.419871056070862e-05, | |
| "loss": 1.0453, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.572155117609663, | |
| "grad_norm": 0.5523152947425842, | |
| "learning_rate": 2.3798323170159488e-05, | |
| "loss": 1.0697, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 2.5975842339478703, | |
| "grad_norm": 0.5818379521369934, | |
| "learning_rate": 2.339824450048218e-05, | |
| "loss": 1.1192, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.6230133502860777, | |
| "grad_norm": 0.5280163884162903, | |
| "learning_rate": 2.299857733524767e-05, | |
| "loss": 1.0263, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 2.6484424666242847, | |
| "grad_norm": 0.5316183567047119, | |
| "learning_rate": 2.2599424352307957e-05, | |
| "loss": 1.0851, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.673871582962492, | |
| "grad_norm": 0.5420896410942078, | |
| "learning_rate": 2.2200888097417307e-05, | |
| "loss": 1.0691, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 2.699300699300699, | |
| "grad_norm": 0.5518523454666138, | |
| "learning_rate": 2.1803070957887347e-05, | |
| "loss": 1.0735, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.7247298156389066, | |
| "grad_norm": 0.5464677810668945, | |
| "learning_rate": 2.140607513628296e-05, | |
| "loss": 1.0905, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 2.750158931977114, | |
| "grad_norm": 0.5611386895179749, | |
| "learning_rate": 2.1010002624165527e-05, | |
| "loss": 1.0789, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.775588048315321, | |
| "grad_norm": 0.576288104057312, | |
| "learning_rate": 2.0614955175890463e-05, | |
| "loss": 1.0551, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 2.801017164653528, | |
| "grad_norm": 0.5907012820243835, | |
| "learning_rate": 2.02210342824657e-05, | |
| "loss": 1.0641, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.8264462809917354, | |
| "grad_norm": 0.5511234998703003, | |
| "learning_rate": 1.982834114547773e-05, | |
| "loss": 1.0781, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 2.851875397329943, | |
| "grad_norm": 0.5715182423591614, | |
| "learning_rate": 1.9436976651092144e-05, | |
| "loss": 1.0377, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.87730451366815, | |
| "grad_norm": 0.5606087446212769, | |
| "learning_rate": 1.9047041344135044e-05, | |
| "loss": 1.037, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 2.9027336300063573, | |
| "grad_norm": 0.5559452176094055, | |
| "learning_rate": 1.865863540226232e-05, | |
| "loss": 1.0786, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.9281627463445643, | |
| "grad_norm": 0.5585038065910339, | |
| "learning_rate": 1.827185861022308e-05, | |
| "loss": 1.058, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 2.9535918626827717, | |
| "grad_norm": 0.6056733727455139, | |
| "learning_rate": 1.7886810334224192e-05, | |
| "loss": 1.0654, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.979020979020979, | |
| "grad_norm": 0.5935595035552979, | |
| "learning_rate": 1.7503589496402208e-05, | |
| "loss": 1.0603, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 3.006357279084552, | |
| "grad_norm": 0.5643672943115234, | |
| "learning_rate": 1.7122294549409484e-05, | |
| "loss": 1.2515, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 3.031786395422759, | |
| "grad_norm": 0.5927174687385559, | |
| "learning_rate": 1.6743023451120832e-05, | |
| "loss": 1.0486, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 3.0572155117609663, | |
| "grad_norm": 0.6135289072990417, | |
| "learning_rate": 1.6365873639467315e-05, | |
| "loss": 1.0647, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.0826446280991737, | |
| "grad_norm": 0.6193404197692871, | |
| "learning_rate": 1.599094200740367e-05, | |
| "loss": 1.0201, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 3.1080737444373807, | |
| "grad_norm": 0.5914512872695923, | |
| "learning_rate": 1.561832487801565e-05, | |
| "loss": 1.0361, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 3.133502860775588, | |
| "grad_norm": 0.6324409246444702, | |
| "learning_rate": 1.5248117979773829e-05, | |
| "loss": 1.0334, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 3.158931977113795, | |
| "grad_norm": 0.6424704790115356, | |
| "learning_rate": 1.4880416421940155e-05, | |
| "loss": 0.9946, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 3.1843610934520026, | |
| "grad_norm": 0.6515686511993408, | |
| "learning_rate": 1.4515314670133582e-05, | |
| "loss": 1.0242, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 3.20979020979021, | |
| "grad_norm": 0.6387981176376343, | |
| "learning_rate": 1.4152906522061048e-05, | |
| "loss": 1.051, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 3.235219326128417, | |
| "grad_norm": 0.6447208523750305, | |
| "learning_rate": 1.3793285083420077e-05, | |
| "loss": 1.0407, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 3.2606484424666244, | |
| "grad_norm": 0.6469278335571289, | |
| "learning_rate": 1.3436542743979125e-05, | |
| "loss": 1.0262, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 3.2860775588048314, | |
| "grad_norm": 0.6370428204536438, | |
| "learning_rate": 1.3082771153841871e-05, | |
| "loss": 1.0428, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 3.311506675143039, | |
| "grad_norm": 0.6550537943840027, | |
| "learning_rate": 1.2732061199901562e-05, | |
| "loss": 1.0236, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.336935791481246, | |
| "grad_norm": 0.6451073288917542, | |
| "learning_rate": 1.2384502982491358e-05, | |
| "loss": 1.0346, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 3.3623649078194533, | |
| "grad_norm": 0.6514309048652649, | |
| "learning_rate": 1.2040185792236874e-05, | |
| "loss": 1.048, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 3.3877940241576603, | |
| "grad_norm": 0.6752684116363525, | |
| "learning_rate": 1.1699198087116589e-05, | |
| "loss": 1.0338, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 3.4132231404958677, | |
| "grad_norm": 0.6573219895362854, | |
| "learning_rate": 1.1361627469736285e-05, | |
| "loss": 1.0606, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 3.438652256834075, | |
| "grad_norm": 0.6186379194259644, | |
| "learning_rate": 1.1027560664823209e-05, | |
| "loss": 1.0397, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 3.464081373172282, | |
| "grad_norm": 0.6716841459274292, | |
| "learning_rate": 1.0697083496945765e-05, | |
| "loss": 1.0613, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 3.4895104895104896, | |
| "grad_norm": 0.6756547093391418, | |
| "learning_rate": 1.0370280868464405e-05, | |
| "loss": 1.0409, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 3.5149396058486966, | |
| "grad_norm": 0.6823243498802185, | |
| "learning_rate": 1.0047236737719601e-05, | |
| "loss": 1.0539, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 3.540368722186904, | |
| "grad_norm": 0.6413143277168274, | |
| "learning_rate": 9.728034097462144e-06, | |
| "loss": 1.0292, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 3.5657978385251115, | |
| "grad_norm": 0.6690115928649902, | |
| "learning_rate": 9.412754953531663e-06, | |
| "loss": 1.0183, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.5912269548633184, | |
| "grad_norm": 0.684471070766449, | |
| "learning_rate": 9.101480303788623e-06, | |
| "loss": 1.022, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 3.616656071201526, | |
| "grad_norm": 0.6697286367416382, | |
| "learning_rate": 8.794290117305296e-06, | |
| "loss": 1.0101, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 3.642085187539733, | |
| "grad_norm": 0.6681500673294067, | |
| "learning_rate": 8.491263313821021e-06, | |
| "loss": 1.0113, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 3.6675143038779403, | |
| "grad_norm": 0.6634730100631714, | |
| "learning_rate": 8.192477743467078e-06, | |
| "loss": 1.0016, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.6929434202161477, | |
| "grad_norm": 0.645428478717804, | |
| "learning_rate": 7.898010166766347e-06, | |
| "loss": 1.0227, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 3.7183725365543547, | |
| "grad_norm": 0.678112804889679, | |
| "learning_rate": 7.607936234912841e-06, | |
| "loss": 1.0339, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 3.7438016528925617, | |
| "grad_norm": 0.691081166267395, | |
| "learning_rate": 7.3223304703363135e-06, | |
| "loss": 0.9891, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 3.769230769230769, | |
| "grad_norm": 0.6777130961418152, | |
| "learning_rate": 7.041266247556813e-06, | |
| "loss": 1.0427, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.7946598855689766, | |
| "grad_norm": 0.6708715558052063, | |
| "learning_rate": 6.764815774334149e-06, | |
| "loss": 1.0068, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 3.8200890019071836, | |
| "grad_norm": 0.685946524143219, | |
| "learning_rate": 6.493050073117116e-06, | |
| "loss": 1.0673, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.845518118245391, | |
| "grad_norm": 0.654752790927887, | |
| "learning_rate": 6.226038962797218e-06, | |
| "loss": 1.015, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 3.870947234583598, | |
| "grad_norm": 0.679111897945404, | |
| "learning_rate": 5.9638510407716394e-06, | |
| "loss": 1.0068, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.8963763509218055, | |
| "grad_norm": 0.6768970489501953, | |
| "learning_rate": 5.706553665319955e-06, | |
| "loss": 1.0424, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 3.921805467260013, | |
| "grad_norm": 0.6532475352287292, | |
| "learning_rate": 5.454212938299255e-06, | |
| "loss": 1.0668, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.94723458359822, | |
| "grad_norm": 0.646823525428772, | |
| "learning_rate": 5.2068936881620095e-06, | |
| "loss": 1.0185, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 3.9726636999364273, | |
| "grad_norm": 0.6898627281188965, | |
| "learning_rate": 4.9646594533010875e-06, | |
| "loss": 1.036, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 3.9980928162746343, | |
| "grad_norm": 1.4582291841506958, | |
| "learning_rate": 4.72757246572623e-06, | |
| "loss": 1.1749, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 4.025429116338207, | |
| "grad_norm": 0.6636221408843994, | |
| "learning_rate": 4.495693635076101e-06, | |
| "loss": 1.0003, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 4.050858232676415, | |
| "grad_norm": 0.679002046585083, | |
| "learning_rate": 4.2690825329701315e-06, | |
| "loss": 0.9937, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 4.076287349014621, | |
| "grad_norm": 0.6657451391220093, | |
| "learning_rate": 4.047797377703985e-06, | |
| "loss": 1.0328, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.101716465352829, | |
| "grad_norm": 0.6936158537864685, | |
| "learning_rate": 3.831895019292897e-06, | |
| "loss": 1.0453, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 4.127145581691036, | |
| "grad_norm": 0.7010984420776367, | |
| "learning_rate": 3.621430924866348e-06, | |
| "loss": 0.9894, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 4.152574698029244, | |
| "grad_norm": 0.7202966809272766, | |
| "learning_rate": 3.4164591644181233e-06, | |
| "loss": 1.0028, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 4.17800381436745, | |
| "grad_norm": 0.7169246077537537, | |
| "learning_rate": 3.217032396915265e-06, | |
| "loss": 1.0268, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 4.203432930705658, | |
| "grad_norm": 0.68609219789505, | |
| "learning_rate": 3.0232018567695697e-06, | |
| "loss": 1.008, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 4.228862047043865, | |
| "grad_norm": 0.6944385170936584, | |
| "learning_rate": 2.8350173406749973e-06, | |
| "loss": 0.9933, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 4.254291163382073, | |
| "grad_norm": 0.6846181750297546, | |
| "learning_rate": 2.652527194814511e-06, | |
| "loss": 1.0147, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 4.27972027972028, | |
| "grad_norm": 0.6891025304794312, | |
| "learning_rate": 2.475778302439524e-06, | |
| "loss": 1.0619, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 4.305149396058487, | |
| "grad_norm": 0.7004478573799133, | |
| "learning_rate": 2.3048160718251997e-06, | |
| "loss": 0.9766, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 4.330578512396694, | |
| "grad_norm": 0.681630551815033, | |
| "learning_rate": 2.1396844246046903e-06, | |
| "loss": 1.0194, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 4.356007628734901, | |
| "grad_norm": 0.7060221433639526, | |
| "learning_rate": 1.980425784485293e-06, | |
| "loss": 1.0079, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 4.381436745073109, | |
| "grad_norm": 0.7196823954582214, | |
| "learning_rate": 1.827081066349459e-06, | |
| "loss": 1.0269, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 4.406865861411316, | |
| "grad_norm": 0.7184461951255798, | |
| "learning_rate": 1.6796896657433808e-06, | |
| "loss": 0.9593, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 4.432294977749523, | |
| "grad_norm": 0.7016611099243164, | |
| "learning_rate": 1.538289448755989e-06, | |
| "loss": 1.005, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 4.45772409408773, | |
| "grad_norm": 0.7026543021202087, | |
| "learning_rate": 1.4029167422908107e-06, | |
| "loss": 1.0268, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 4.483153210425938, | |
| "grad_norm": 0.722529411315918, | |
| "learning_rate": 1.273606324733284e-06, | |
| "loss": 1.0258, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 4.508582326764145, | |
| "grad_norm": 0.6864616274833679, | |
| "learning_rate": 1.1503914170159057e-06, | |
| "loss": 1.0159, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 4.534011443102353, | |
| "grad_norm": 0.6917942762374878, | |
| "learning_rate": 1.0333036740834856e-06, | |
| "loss": 1.0288, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 4.559440559440559, | |
| "grad_norm": 0.7076994180679321, | |
| "learning_rate": 9.223731767607435e-07, | |
| "loss": 1.0102, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 4.584869675778767, | |
| "grad_norm": 0.6951266527175903, | |
| "learning_rate": 8.176284240242638e-07, | |
| "loss": 1.0253, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.610298792116974, | |
| "grad_norm": 0.6959212422370911, | |
| "learning_rate": 7.19096325680907e-07, | |
| "loss": 1.0014, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 4.6357279084551815, | |
| "grad_norm": 0.7114453911781311, | |
| "learning_rate": 6.268021954544096e-07, | |
| "loss": 0.9987, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 4.661157024793388, | |
| "grad_norm": 0.7204643487930298, | |
| "learning_rate": 5.407697444821169e-07, | |
| "loss": 0.9987, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 4.686586141131595, | |
| "grad_norm": 0.7097218632698059, | |
| "learning_rate": 4.6102107522336403e-07, | |
| "loss": 1.0043, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 4.712015257469803, | |
| "grad_norm": 0.7494398951530457, | |
| "learning_rate": 3.8757667578118994e-07, | |
| "loss": 1.0321, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 4.73744437380801, | |
| "grad_norm": 0.6792721748352051, | |
| "learning_rate": 3.204554146387456e-07, | |
| "loss": 1.0747, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 4.762873490146218, | |
| "grad_norm": 0.6858909130096436, | |
| "learning_rate": 2.5967453581185184e-07, | |
| "loss": 0.9982, | |
| "step": 935 | |
| }, | |
| { | |
| "epoch": 4.788302606484424, | |
| "grad_norm": 0.7001191973686218, | |
| "learning_rate": 2.052496544188487e-07, | |
| "loss": 1.0616, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 4.813731722822632, | |
| "grad_norm": 0.7212966084480286, | |
| "learning_rate": 1.571947526689349e-07, | |
| "loss": 1.0264, | |
| "step": 945 | |
| }, | |
| { | |
| "epoch": 4.839160839160839, | |
| "grad_norm": 0.670038104057312, | |
| "learning_rate": 1.1552217627004425e-07, | |
| "loss": 1.0237, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.864589955499047, | |
| "grad_norm": 0.7079971432685852, | |
| "learning_rate": 8.024263125710752e-08, | |
| "loss": 0.992, | |
| "step": 955 | |
| }, | |
| { | |
| "epoch": 4.890019071837254, | |
| "grad_norm": 0.7235581874847412, | |
| "learning_rate": 5.136518124159162e-08, | |
| "loss": 0.9947, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 4.915448188175461, | |
| "grad_norm": 0.6986010074615479, | |
| "learning_rate": 2.8897245082978864e-08, | |
| "loss": 0.9907, | |
| "step": 965 | |
| }, | |
| { | |
| "epoch": 4.940877304513668, | |
| "grad_norm": 0.7248331904411316, | |
| "learning_rate": 1.284459498280266e-08, | |
| "loss": 0.9958, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 4.9663064208518755, | |
| "grad_norm": 0.7134162783622742, | |
| "learning_rate": 3.2113550017198735e-09, | |
| "loss": 1.0067, | |
| "step": 975 | |
| }, | |
| { | |
| "epoch": 4.991735537190083, | |
| "grad_norm": 0.7377886176109314, | |
| "learning_rate": 0.0, | |
| "loss": 0.9834, | |
| "step": 980 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 980, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 1.2928446514593792e+18, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |