| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9990182603573532, | |
| "eval_steps": 500, | |
| "global_step": 636, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.007853917141174161, | |
| "grad_norm": 9.454776697912866, | |
| "learning_rate": 1.25e-06, | |
| "loss": 1.9975, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.015707834282348322, | |
| "grad_norm": 7.36591132940049, | |
| "learning_rate": 2.8125e-06, | |
| "loss": 1.8813, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.023561751423522483, | |
| "grad_norm": 7.154615745335707, | |
| "learning_rate": 4.3750000000000005e-06, | |
| "loss": 1.4274, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.031415668564696644, | |
| "grad_norm": 8.822500376414714, | |
| "learning_rate": 5.9375e-06, | |
| "loss": 0.8007, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0392695857058708, | |
| "grad_norm": 1.5626818502455462, | |
| "learning_rate": 7.500000000000001e-06, | |
| "loss": 0.2708, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.04712350284704497, | |
| "grad_norm": 0.95138015574769, | |
| "learning_rate": 9.0625e-06, | |
| "loss": 0.1895, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.054977419988219124, | |
| "grad_norm": 0.632172201703669, | |
| "learning_rate": 1.0625e-05, | |
| "loss": 0.161, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.06283133712939329, | |
| "grad_norm": 0.6087184175934243, | |
| "learning_rate": 1.2187500000000001e-05, | |
| "loss": 0.1514, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.07068525427056745, | |
| "grad_norm": 0.43002905751035403, | |
| "learning_rate": 1.375e-05, | |
| "loss": 0.1405, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.0785391714117416, | |
| "grad_norm": 0.4316303477126074, | |
| "learning_rate": 1.5312500000000003e-05, | |
| "loss": 0.1323, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08639308855291576, | |
| "grad_norm": 0.41295406365002396, | |
| "learning_rate": 1.6875e-05, | |
| "loss": 0.1232, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.09424700569408993, | |
| "grad_norm": 0.3869106906435148, | |
| "learning_rate": 1.84375e-05, | |
| "loss": 0.1185, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.10210092283526409, | |
| "grad_norm": 0.38396751864248035, | |
| "learning_rate": 2e-05, | |
| "loss": 0.1127, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.10995483997643825, | |
| "grad_norm": 0.31600318922334614, | |
| "learning_rate": 1.9996229574120564e-05, | |
| "loss": 0.1141, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.1178087571176124, | |
| "grad_norm": 0.2807626851547414, | |
| "learning_rate": 1.998492113970451e-05, | |
| "loss": 0.1091, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.12566267425878658, | |
| "grad_norm": 0.2712214079154975, | |
| "learning_rate": 1.99660832242746e-05, | |
| "loss": 0.1062, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.13351659139996072, | |
| "grad_norm": 0.3514322765970332, | |
| "learning_rate": 1.99397300332236e-05, | |
| "loss": 0.1043, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.1413705085411349, | |
| "grad_norm": 0.2565181736697417, | |
| "learning_rate": 1.9905881439102222e-05, | |
| "loss": 0.1012, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.14922442568230906, | |
| "grad_norm": 0.22648554493196996, | |
| "learning_rate": 1.9864562966633517e-05, | |
| "loss": 0.1091, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.1570783428234832, | |
| "grad_norm": 0.3160797286348635, | |
| "learning_rate": 1.9815805773465064e-05, | |
| "loss": 0.1003, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.16493225996465738, | |
| "grad_norm": 0.24530695236987562, | |
| "learning_rate": 1.9759646626673445e-05, | |
| "loss": 0.1054, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.17278617710583152, | |
| "grad_norm": 0.3308392923974687, | |
| "learning_rate": 1.9696127875038753e-05, | |
| "loss": 0.1021, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1806400942470057, | |
| "grad_norm": 0.281121787116649, | |
| "learning_rate": 1.9625297417109982e-05, | |
| "loss": 0.0967, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.18849401138817987, | |
| "grad_norm": 0.20020373984472567, | |
| "learning_rate": 1.954720866508546e-05, | |
| "loss": 0.1, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.196347928529354, | |
| "grad_norm": 0.27280707245432356, | |
| "learning_rate": 1.946192050453549e-05, | |
| "loss": 0.0999, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.20420184567052818, | |
| "grad_norm": 0.24726594932319515, | |
| "learning_rate": 1.936949724999762e-05, | |
| "loss": 0.0947, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.21205576281170233, | |
| "grad_norm": 0.2383332440997248, | |
| "learning_rate": 1.9270008596478008e-05, | |
| "loss": 0.0982, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.2199096799528765, | |
| "grad_norm": 0.2203997646125469, | |
| "learning_rate": 1.916352956689544e-05, | |
| "loss": 0.1017, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.22776359709405067, | |
| "grad_norm": 0.2362083987436209, | |
| "learning_rate": 1.905014045550767e-05, | |
| "loss": 0.0987, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.2356175142352248, | |
| "grad_norm": 0.23431067680898818, | |
| "learning_rate": 1.89299267673627e-05, | |
| "loss": 0.1055, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.24347143137639898, | |
| "grad_norm": 0.21183859562424995, | |
| "learning_rate": 1.88029791538207e-05, | |
| "loss": 0.0972, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.25132534851757315, | |
| "grad_norm": 0.22397328492306617, | |
| "learning_rate": 1.8669393344195154e-05, | |
| "loss": 0.1009, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2591792656587473, | |
| "grad_norm": 0.22197633586551052, | |
| "learning_rate": 1.852927007356481e-05, | |
| "loss": 0.0936, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.26703318279992144, | |
| "grad_norm": 0.21347022910458285, | |
| "learning_rate": 1.8382715006810853e-05, | |
| "loss": 0.0957, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2748870999410956, | |
| "grad_norm": 0.19957938768669062, | |
| "learning_rate": 1.8229838658936566e-05, | |
| "loss": 0.0873, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.2827410170822698, | |
| "grad_norm": 0.21372555100198343, | |
| "learning_rate": 1.807075631172963e-05, | |
| "loss": 0.0931, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.29059493422344396, | |
| "grad_norm": 0.24190910676139596, | |
| "learning_rate": 1.7905587926829815e-05, | |
| "loss": 0.0954, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.29844885136461813, | |
| "grad_norm": 0.2348573879122653, | |
| "learning_rate": 1.77344580552677e-05, | |
| "loss": 0.092, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.30630276850579224, | |
| "grad_norm": 0.21204667498616597, | |
| "learning_rate": 1.7557495743542586e-05, | |
| "loss": 0.0865, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.3141566856469664, | |
| "grad_norm": 0.23025633594185083, | |
| "learning_rate": 1.7374834436310427e-05, | |
| "loss": 0.0892, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3220106027881406, | |
| "grad_norm": 0.1958893222829977, | |
| "learning_rate": 1.7186611875755227e-05, | |
| "loss": 0.0896, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.32986451992931476, | |
| "grad_norm": 0.22589475386438357, | |
| "learning_rate": 1.6992969997719658e-05, | |
| "loss": 0.0945, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.33771843707048893, | |
| "grad_norm": 0.20789776750177996, | |
| "learning_rate": 1.679405482467338e-05, | |
| "loss": 0.0904, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.34557235421166305, | |
| "grad_norm": 0.22053512123077823, | |
| "learning_rate": 1.6590016355599653e-05, | |
| "loss": 0.0882, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3534262713528372, | |
| "grad_norm": 0.20437871058955417, | |
| "learning_rate": 1.638100845288331e-05, | |
| "loss": 0.0943, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.3612801884940114, | |
| "grad_norm": 0.2090640789388261, | |
| "learning_rate": 1.6167188726285433e-05, | |
| "loss": 0.0924, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.36913410563518556, | |
| "grad_norm": 0.21993457381708645, | |
| "learning_rate": 1.5948718414092163e-05, | |
| "loss": 0.082, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.37698802277635973, | |
| "grad_norm": 0.21951962440906275, | |
| "learning_rate": 1.5725762261527295e-05, | |
| "loss": 0.0918, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.38484193991753385, | |
| "grad_norm": 0.19872240145820674, | |
| "learning_rate": 1.549848839652035e-05, | |
| "loss": 0.0955, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.392695857058708, | |
| "grad_norm": 0.2184861301177773, | |
| "learning_rate": 1.5267068202923802e-05, | |
| "loss": 0.0899, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.4005497741998822, | |
| "grad_norm": 0.21059689355945896, | |
| "learning_rate": 1.503167619127504e-05, | |
| "loss": 0.0919, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.40840369134105636, | |
| "grad_norm": 0.21063064163571898, | |
| "learning_rate": 1.479248986720057e-05, | |
| "loss": 0.086, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.41625760848223053, | |
| "grad_norm": 0.19969666720995202, | |
| "learning_rate": 1.4549689597561652e-05, | |
| "loss": 0.0949, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.42411152562340465, | |
| "grad_norm": 0.21154526388191275, | |
| "learning_rate": 1.4303458474442325e-05, | |
| "loss": 0.0869, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4319654427645788, | |
| "grad_norm": 0.2305997701221425, | |
| "learning_rate": 1.4053982177082369e-05, | |
| "loss": 0.09, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.439819359905753, | |
| "grad_norm": 0.18766215420046584, | |
| "learning_rate": 1.3801448831859363e-05, | |
| "loss": 0.0869, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.44767327704692716, | |
| "grad_norm": 0.22942333663793876, | |
| "learning_rate": 1.3546048870425356e-05, | |
| "loss": 0.0929, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.45552719418810134, | |
| "grad_norm": 0.24300158612250203, | |
| "learning_rate": 1.328797488610519e-05, | |
| "loss": 0.0865, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.46338111132927545, | |
| "grad_norm": 0.20752136799187096, | |
| "learning_rate": 1.3027421488664723e-05, | |
| "loss": 0.0927, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.4712350284704496, | |
| "grad_norm": 0.21217812864504576, | |
| "learning_rate": 1.2764585157558486e-05, | |
| "loss": 0.0879, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.4790889456116238, | |
| "grad_norm": 0.22299949763483096, | |
| "learning_rate": 1.2499664093767458e-05, | |
| "loss": 0.0938, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.48694286275279797, | |
| "grad_norm": 0.23957557508813998, | |
| "learning_rate": 1.2232858070338618e-05, | |
| "loss": 0.0904, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.49479677989397214, | |
| "grad_norm": 0.2023486493539765, | |
| "learning_rate": 1.1964368281739078e-05, | |
| "loss": 0.0874, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.5026506970351463, | |
| "grad_norm": 0.1861038104854032, | |
| "learning_rate": 1.1694397192138295e-05, | |
| "loss": 0.0816, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.5105046141763204, | |
| "grad_norm": 0.2014117982786214, | |
| "learning_rate": 1.1423148382732854e-05, | |
| "loss": 0.0823, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.5183585313174947, | |
| "grad_norm": 0.22960664067039807, | |
| "learning_rate": 1.1150826398228904e-05, | |
| "loss": 0.0882, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.5262124484586688, | |
| "grad_norm": 0.222023007250355, | |
| "learning_rate": 1.087763659259803e-05, | |
| "loss": 0.0863, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.5340663655998429, | |
| "grad_norm": 0.19468935638754573, | |
| "learning_rate": 1.0603784974222862e-05, | |
| "loss": 0.0884, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5419202827410171, | |
| "grad_norm": 0.2047369290635767, | |
| "learning_rate": 1.0329478050549208e-05, | |
| "loss": 0.0781, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.5497741998821912, | |
| "grad_norm": 0.2057117950843339, | |
| "learning_rate": 1.0054922672361858e-05, | |
| "loss": 0.0813, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5576281170233655, | |
| "grad_norm": 0.23197127290798342, | |
| "learning_rate": 9.780325877801455e-06, | |
| "loss": 0.0989, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5654820341645396, | |
| "grad_norm": 0.3305907734524743, | |
| "learning_rate": 9.50589473624013e-06, | |
| "loss": 0.0816, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5733359513057137, | |
| "grad_norm": 0.215453345630245, | |
| "learning_rate": 9.231836192133532e-06, | |
| "loss": 0.0861, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5811898684468879, | |
| "grad_norm": 0.1908748794995548, | |
| "learning_rate": 8.958356908967104e-06, | |
| "loss": 0.0855, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.589043785588062, | |
| "grad_norm": 0.28572080285977014, | |
| "learning_rate": 8.685663113414186e-06, | |
| "loss": 0.0841, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5968977027292363, | |
| "grad_norm": 0.29366381772227595, | |
| "learning_rate": 8.413960439823567e-06, | |
| "loss": 0.0824, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.6047516198704104, | |
| "grad_norm": 0.21387609796609025, | |
| "learning_rate": 8.143453775153646e-06, | |
| "loss": 0.0846, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.6126055370115845, | |
| "grad_norm": 0.18382031478587926, | |
| "learning_rate": 7.874347104470234e-06, | |
| "loss": 0.0886, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.6204594541527587, | |
| "grad_norm": 0.20443500947333, | |
| "learning_rate": 7.606843357124426e-06, | |
| "loss": 0.0818, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.6283133712939328, | |
| "grad_norm": 0.1868753348119755, | |
| "learning_rate": 7.341144253726583e-06, | |
| "loss": 0.0801, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6361672884351071, | |
| "grad_norm": 0.1919110684349949, | |
| "learning_rate": 7.0774501540318305e-06, | |
| "loss": 0.0837, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6440212055762812, | |
| "grad_norm": 0.19733646385316747, | |
| "learning_rate": 6.815959905851715e-06, | |
| "loss": 0.0907, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6518751227174553, | |
| "grad_norm": 0.20695118469945245, | |
| "learning_rate": 6.556870695106028e-06, | |
| "loss": 0.0839, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.6597290398586295, | |
| "grad_norm": 0.206342467457721, | |
| "learning_rate": 6.300377897127825e-06, | |
| "loss": 0.0817, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6675829569998036, | |
| "grad_norm": 0.19498340314833404, | |
| "learning_rate": 6.046674929333787e-06, | |
| "loss": 0.0832, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6754368741409779, | |
| "grad_norm": 0.18423257358883185, | |
| "learning_rate": 5.795953105371e-06, | |
| "loss": 0.0894, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.683290791282152, | |
| "grad_norm": 0.19500438735567713, | |
| "learning_rate": 5.548401490850193e-06, | |
| "loss": 0.0841, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6911447084233261, | |
| "grad_norm": 0.1943344540910271, | |
| "learning_rate": 5.304206760774139e-06, | |
| "loss": 0.0857, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6989986255645003, | |
| "grad_norm": 0.18694442634566882, | |
| "learning_rate": 5.063553058768814e-06, | |
| "loss": 0.0873, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.7068525427056744, | |
| "grad_norm": 0.2528592095445329, | |
| "learning_rate": 4.826621858223431e-06, | |
| "loss": 0.0817, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.7147064598468487, | |
| "grad_norm": 0.2090763236109062, | |
| "learning_rate": 4.593591825444028e-06, | |
| "loss": 0.0845, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.7225603769880228, | |
| "grad_norm": 0.18944053970106092, | |
| "learning_rate": 4.364638684923848e-06, | |
| "loss": 0.0871, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.7304142941291969, | |
| "grad_norm": 0.20174647482444266, | |
| "learning_rate": 4.13993508683214e-06, | |
| "loss": 0.084, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.7382682112703711, | |
| "grad_norm": 0.19890794882086518, | |
| "learning_rate": 3.919650476821192e-06, | |
| "loss": 0.0848, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7461221284115452, | |
| "grad_norm": 0.17866134537019918, | |
| "learning_rate": 3.7039509682498887e-06, | |
| "loss": 0.0767, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7539760455527195, | |
| "grad_norm": 0.1758570225271436, | |
| "learning_rate": 3.4929992169200865e-06, | |
| "loss": 0.079, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7618299626938936, | |
| "grad_norm": 0.19532675849601197, | |
| "learning_rate": 3.2869542984202974e-06, | |
| "loss": 0.0872, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7696838798350677, | |
| "grad_norm": 0.18336677624616388, | |
| "learning_rate": 3.0859715881691267e-06, | |
| "loss": 0.078, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7775377969762419, | |
| "grad_norm": 0.19477530024566625, | |
| "learning_rate": 2.890202644248983e-06, | |
| "loss": 0.0888, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.785391714117416, | |
| "grad_norm": 0.18612540878842743, | |
| "learning_rate": 2.6997950931183736e-06, | |
| "loss": 0.0806, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.785391714117416, | |
| "eval_loss": 0.07095803320407867, | |
| "eval_runtime": 199.4165, | |
| "eval_samples_per_second": 9.046, | |
| "eval_steps_per_second": 2.262, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7932456312585903, | |
| "grad_norm": 0.20861974562159583, | |
| "learning_rate": 2.514892518288988e-06, | |
| "loss": 0.0816, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.8010995483997644, | |
| "grad_norm": 0.19828891809733562, | |
| "learning_rate": 2.335634352051488e-06, | |
| "loss": 0.0927, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.8089534655409385, | |
| "grad_norm": 0.17567422510168001, | |
| "learning_rate": 2.1621557703316876e-06, | |
| "loss": 0.077, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.8168073826821127, | |
| "grad_norm": 0.43901734050274105, | |
| "learning_rate": 1.994587590756397e-06, | |
| "loss": 0.0879, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.8246612998232868, | |
| "grad_norm": 0.19303609197478824, | |
| "learning_rate": 1.8330561740057839e-06, | |
| "loss": 0.0798, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.8325152169644611, | |
| "grad_norm": 0.21014700571002942, | |
| "learning_rate": 1.6776833285266602e-06, | |
| "loss": 0.0846, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.8403691341056352, | |
| "grad_norm": 0.19381351445718112, | |
| "learning_rate": 1.528586218678535e-06, | |
| "loss": 0.0828, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.8482230512468093, | |
| "grad_norm": 0.20006096517086086, | |
| "learning_rate": 1.3858772763817174e-06, | |
| "loss": 0.0817, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.8560769683879835, | |
| "grad_norm": 0.25891905601753346, | |
| "learning_rate": 1.2496641163340562e-06, | |
| "loss": 0.0815, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.8639308855291576, | |
| "grad_norm": 0.20338109286639236, | |
| "learning_rate": 1.120049454860307e-06, | |
| "loss": 0.0827, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.8717848026703319, | |
| "grad_norm": 0.1817547896016754, | |
| "learning_rate": 9.971310324552597e-07, | |
| "loss": 0.0845, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.879638719811506, | |
| "grad_norm": 0.20333004229909032, | |
| "learning_rate": 8.810015400790994e-07, | |
| "loss": 0.0836, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.8874926369526801, | |
| "grad_norm": 0.2005258856819357, | |
| "learning_rate": 7.71748549260507e-07, | |
| "loss": 0.0813, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.8953465540938543, | |
| "grad_norm": 0.17505423789580266, | |
| "learning_rate": 6.694544460602825e-07, | |
| "loss": 0.0881, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.9032004712350284, | |
| "grad_norm": 0.19509475957323322, | |
| "learning_rate": 5.741963689452268e-07, | |
| "loss": 0.082, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.9110543883762027, | |
| "grad_norm": 0.19518986043069664, | |
| "learning_rate": 4.860461506191782e-07, | |
| "loss": 0.0842, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.9189083055173768, | |
| "grad_norm": 0.1925064249529951, | |
| "learning_rate": 4.0507026385502747e-07, | |
| "loss": 0.0839, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.9267622226585509, | |
| "grad_norm": 0.18327816158141888, | |
| "learning_rate": 3.313297713685859e-07, | |
| "loss": 0.087, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.9346161397997251, | |
| "grad_norm": 0.20588287191995575, | |
| "learning_rate": 2.6488027977210175e-07, | |
| "loss": 0.0827, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.9424700569408992, | |
| "grad_norm": 0.197357389547204, | |
| "learning_rate": 2.057718976421341e-07, | |
| "loss": 0.0815, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9503239740820735, | |
| "grad_norm": 0.18513291000922844, | |
| "learning_rate": 1.5404919773341576e-07, | |
| "loss": 0.082, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.9581778912232476, | |
| "grad_norm": 0.185473668497204, | |
| "learning_rate": 1.0975118336720603e-07, | |
| "loss": 0.0846, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9660318083644217, | |
| "grad_norm": 0.1803986934893818, | |
| "learning_rate": 7.291125901946027e-08, | |
| "loss": 0.0777, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9738857255055959, | |
| "grad_norm": 0.18308958490608873, | |
| "learning_rate": 4.3557205131008475e-08, | |
| "loss": 0.0771, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.98173964264677, | |
| "grad_norm": 0.2968041785813041, | |
| "learning_rate": 2.171115715874139e-08, | |
| "loss": 0.0872, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.9895935597879443, | |
| "grad_norm": 0.19838566337698643, | |
| "learning_rate": 7.389588883585097e-09, | |
| "loss": 0.0789, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9974474769291184, | |
| "grad_norm": 0.17786747719054677, | |
| "learning_rate": 6.032999878735624e-10, | |
| "loss": 0.0821, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.9990182603573532, | |
| "step": 636, | |
| "total_flos": 3801667199303680.0, | |
| "train_loss": 0.13866351419509207, | |
| "train_runtime": 19174.2073, | |
| "train_samples_per_second": 2.125, | |
| "train_steps_per_second": 0.033 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 636, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 1000, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 3801667199303680.0, | |
| "train_batch_size": 1, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |