| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 1.0, | |
| "eval_steps": 500, | |
| "global_step": 671, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0014903129657228018, | |
| "grad_norm": 0.46681726349989877, | |
| "learning_rate": 4.411764705882353e-08, | |
| "loss": 1.7639, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.007451564828614009, | |
| "grad_norm": 0.41873451593648303, | |
| "learning_rate": 2.2058823529411765e-07, | |
| "loss": 1.7304, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.014903129657228018, | |
| "grad_norm": 0.4344144099098602, | |
| "learning_rate": 4.411764705882353e-07, | |
| "loss": 1.7792, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.022354694485842028, | |
| "grad_norm": 0.465770292151279, | |
| "learning_rate": 6.61764705882353e-07, | |
| "loss": 1.7689, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.029806259314456036, | |
| "grad_norm": 0.43483029835423426, | |
| "learning_rate": 8.823529411764706e-07, | |
| "loss": 1.7205, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.037257824143070044, | |
| "grad_norm": 0.44740719809064367, | |
| "learning_rate": 1.1029411764705884e-06, | |
| "loss": 1.7651, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.044709388971684055, | |
| "grad_norm": 0.43250694443997345, | |
| "learning_rate": 1.323529411764706e-06, | |
| "loss": 1.7926, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.05216095380029806, | |
| "grad_norm": 0.43541824597905016, | |
| "learning_rate": 1.5441176470588234e-06, | |
| "loss": 1.7368, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.05961251862891207, | |
| "grad_norm": 0.44779413861684675, | |
| "learning_rate": 1.7647058823529412e-06, | |
| "loss": 1.7887, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.06706408345752608, | |
| "grad_norm": 0.46078766735704463, | |
| "learning_rate": 1.9852941176470586e-06, | |
| "loss": 1.6828, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.07451564828614009, | |
| "grad_norm": 0.43875939336780606, | |
| "learning_rate": 2.2058823529411767e-06, | |
| "loss": 1.731, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.08196721311475409, | |
| "grad_norm": 0.4229850780249163, | |
| "learning_rate": 2.4264705882352943e-06, | |
| "loss": 1.7005, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.08941877794336811, | |
| "grad_norm": 0.4045297221644775, | |
| "learning_rate": 2.647058823529412e-06, | |
| "loss": 1.7108, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.09687034277198212, | |
| "grad_norm": 0.4373623437488123, | |
| "learning_rate": 2.8676470588235296e-06, | |
| "loss": 1.7132, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.10432190760059612, | |
| "grad_norm": 0.37968682487624034, | |
| "learning_rate": 2.999918570372821e-06, | |
| "loss": 1.6389, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.11177347242921014, | |
| "grad_norm": 0.3852394407171853, | |
| "learning_rate": 2.9990025885979037e-06, | |
| "loss": 1.6099, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.11922503725782414, | |
| "grad_norm": 0.3557069482806777, | |
| "learning_rate": 2.997069461623824e-06, | |
| "loss": 1.6954, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.12667660208643816, | |
| "grad_norm": 0.3079289274590612, | |
| "learning_rate": 2.9941205011700118e-06, | |
| "loss": 1.6172, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.13412816691505217, | |
| "grad_norm": 0.29857210375623594, | |
| "learning_rate": 2.990157708247667e-06, | |
| "loss": 1.595, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.14157973174366617, | |
| "grad_norm": 0.2842045412377899, | |
| "learning_rate": 2.9851837718019762e-06, | |
| "loss": 1.6739, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.14903129657228018, | |
| "grad_norm": 0.24687715976280528, | |
| "learning_rate": 2.9792020668875367e-06, | |
| "loss": 1.5807, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.15648286140089418, | |
| "grad_norm": 0.2369983753260976, | |
| "learning_rate": 2.9722166523782167e-06, | |
| "loss": 1.5851, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.16393442622950818, | |
| "grad_norm": 0.20261276348197832, | |
| "learning_rate": 2.964232268213018e-06, | |
| "loss": 1.5384, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.17138599105812222, | |
| "grad_norm": 0.20698300200844078, | |
| "learning_rate": 2.955254332179797e-06, | |
| "loss": 1.5694, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.17883755588673622, | |
| "grad_norm": 0.2043135439690471, | |
| "learning_rate": 2.9452889362390366e-06, | |
| "loss": 1.6602, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.18628912071535023, | |
| "grad_norm": 0.17624295165768436, | |
| "learning_rate": 2.9343428423901614e-06, | |
| "loss": 1.5703, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.19374068554396423, | |
| "grad_norm": 0.15837032215765645, | |
| "learning_rate": 2.9224234780831905e-06, | |
| "loss": 1.5531, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.20119225037257824, | |
| "grad_norm": 0.1811372542204385, | |
| "learning_rate": 2.9095389311788626e-06, | |
| "loss": 1.577, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.20864381520119224, | |
| "grad_norm": 0.15759029786619647, | |
| "learning_rate": 2.8956979444606303e-06, | |
| "loss": 1.5101, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.21609538002980627, | |
| "grad_norm": 0.1520856154563207, | |
| "learning_rate": 2.8809099097022624e-06, | |
| "loss": 1.5547, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.22354694485842028, | |
| "grad_norm": 0.14966084871837593, | |
| "learning_rate": 2.8651848612950768e-06, | |
| "loss": 1.5801, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.23099850968703428, | |
| "grad_norm": 0.1378402679865988, | |
| "learning_rate": 2.848533469439122e-06, | |
| "loss": 1.4604, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.23845007451564829, | |
| "grad_norm": 0.1464647323237256, | |
| "learning_rate": 2.8309670329029358e-06, | |
| "loss": 1.4894, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.2459016393442623, | |
| "grad_norm": 0.1437051133777023, | |
| "learning_rate": 2.8124974713567872e-06, | |
| "loss": 1.5269, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.2533532041728763, | |
| "grad_norm": 0.14789288282077315, | |
| "learning_rate": 2.79313731728461e-06, | |
| "loss": 1.5537, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.2608047690014903, | |
| "grad_norm": 0.14308275673434606, | |
| "learning_rate": 2.772899707480108e-06, | |
| "loss": 1.4818, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.26825633383010433, | |
| "grad_norm": 0.13899835934264645, | |
| "learning_rate": 2.7517983741328146e-06, | |
| "loss": 1.4718, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.2757078986587183, | |
| "grad_norm": 0.13342232666402679, | |
| "learning_rate": 2.729847635510137e-06, | |
| "loss": 1.5693, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.28315946348733234, | |
| "grad_norm": 0.13133306157733085, | |
| "learning_rate": 2.70706238624173e-06, | |
| "loss": 1.5144, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2906110283159464, | |
| "grad_norm": 0.12346136386902845, | |
| "learning_rate": 2.6834580872127733e-06, | |
| "loss": 1.4706, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.29806259314456035, | |
| "grad_norm": 0.12962940757781438, | |
| "learning_rate": 2.6590507550730175e-06, | |
| "loss": 1.4595, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.3055141579731744, | |
| "grad_norm": 0.1256519676373563, | |
| "learning_rate": 2.6338569513687182e-06, | |
| "loss": 1.4993, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.31296572280178836, | |
| "grad_norm": 0.12374610253004059, | |
| "learning_rate": 2.6078937713048357e-06, | |
| "loss": 1.459, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.3204172876304024, | |
| "grad_norm": 0.13618026502855957, | |
| "learning_rate": 2.581178832145114e-06, | |
| "loss": 1.5273, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.32786885245901637, | |
| "grad_norm": 0.12673229090948254, | |
| "learning_rate": 2.553730261257924e-06, | |
| "loss": 1.4442, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.3353204172876304, | |
| "grad_norm": 0.11648035880727567, | |
| "learning_rate": 2.525566683815973e-06, | |
| "loss": 1.4502, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.34277198211624443, | |
| "grad_norm": 0.12644477441341948, | |
| "learning_rate": 2.496707210158233e-06, | |
| "loss": 1.5173, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.3502235469448584, | |
| "grad_norm": 0.11586720234125189, | |
| "learning_rate": 2.4671714228226542e-06, | |
| "loss": 1.46, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.35767511177347244, | |
| "grad_norm": 0.11883054977090766, | |
| "learning_rate": 2.4369793632584796e-06, | |
| "loss": 1.4847, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.3651266766020864, | |
| "grad_norm": 0.13665955324099738, | |
| "learning_rate": 2.4061515182271535e-06, | |
| "loss": 1.5317, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.37257824143070045, | |
| "grad_norm": 0.11799056969645393, | |
| "learning_rate": 2.3747088059010745e-06, | |
| "loss": 1.4061, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.38002980625931443, | |
| "grad_norm": 0.12197294071650851, | |
| "learning_rate": 2.342672561669611e-06, | |
| "loss": 1.5242, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.38748137108792846, | |
| "grad_norm": 0.12328952083324084, | |
| "learning_rate": 2.3100645236620133e-06, | |
| "loss": 1.4472, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.3949329359165425, | |
| "grad_norm": 0.13371579203837833, | |
| "learning_rate": 2.276906817997054e-06, | |
| "loss": 1.4578, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.40238450074515647, | |
| "grad_norm": 0.11797510100880551, | |
| "learning_rate": 2.2432219437693897e-06, | |
| "loss": 1.4541, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.4098360655737705, | |
| "grad_norm": 0.11652500829878425, | |
| "learning_rate": 2.209032757782848e-06, | |
| "loss": 1.428, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.4172876304023845, | |
| "grad_norm": 0.11784740588346589, | |
| "learning_rate": 2.174362459040989e-06, | |
| "loss": 1.4765, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.4247391952309985, | |
| "grad_norm": 0.13214434156180432, | |
| "learning_rate": 2.139234573005468e-06, | |
| "loss": 1.4823, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.43219076005961254, | |
| "grad_norm": 0.12325194724942345, | |
| "learning_rate": 2.1036729356328806e-06, | |
| "loss": 1.4587, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.4396423248882265, | |
| "grad_norm": 0.11758350672051872, | |
| "learning_rate": 2.06770167720092e-06, | |
| "loss": 1.427, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.44709388971684055, | |
| "grad_norm": 0.11607090772093101, | |
| "learning_rate": 2.0313452059348308e-06, | |
| "loss": 1.4912, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.45454545454545453, | |
| "grad_norm": 0.11742494412844874, | |
| "learning_rate": 1.99462819144525e-06, | |
| "loss": 1.4829, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.46199701937406856, | |
| "grad_norm": 0.12666171818837882, | |
| "learning_rate": 1.957575547988697e-06, | |
| "loss": 1.4497, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.46944858420268254, | |
| "grad_norm": 0.125963894413993, | |
| "learning_rate": 1.9202124175620545e-06, | |
| "loss": 1.4255, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.47690014903129657, | |
| "grad_norm": 0.11914613772013677, | |
| "learning_rate": 1.8825641528425148e-06, | |
| "loss": 1.4425, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.4843517138599106, | |
| "grad_norm": 0.12972400981235876, | |
| "learning_rate": 1.8446562999845715e-06, | |
| "loss": 1.4829, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.4918032786885246, | |
| "grad_norm": 0.12709044180361684, | |
| "learning_rate": 1.8065145812857305e-06, | |
| "loss": 1.4178, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.4992548435171386, | |
| "grad_norm": 0.1202237982692701, | |
| "learning_rate": 1.7681648777326943e-06, | |
| "loss": 1.4876, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.5067064083457526, | |
| "grad_norm": 0.12048604340467019, | |
| "learning_rate": 1.7296332114398704e-06, | |
| "loss": 1.4076, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.5141579731743666, | |
| "grad_norm": 0.1337777649679242, | |
| "learning_rate": 1.6909457279921186e-06, | |
| "loss": 1.4171, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.5216095380029806, | |
| "grad_norm": 0.12802906461759175, | |
| "learning_rate": 1.6521286787037178e-06, | |
| "loss": 1.4569, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.5290611028315947, | |
| "grad_norm": 0.11319710800264886, | |
| "learning_rate": 1.613208402805586e-06, | |
| "loss": 1.4541, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.5365126676602087, | |
| "grad_norm": 0.117227595325786, | |
| "learning_rate": 1.5742113095728515e-06, | |
| "loss": 1.4812, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.5439642324888226, | |
| "grad_norm": 0.12215641426475456, | |
| "learning_rate": 1.535163860404891e-06, | |
| "loss": 1.4964, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.5514157973174366, | |
| "grad_norm": 0.13083431124739964, | |
| "learning_rate": 1.4960925508699984e-06, | |
| "loss": 1.4217, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.5588673621460507, | |
| "grad_norm": 0.1409576348850734, | |
| "learning_rate": 1.4570238927268746e-06, | |
| "loss": 1.4805, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.5663189269746647, | |
| "grad_norm": 0.11615933961706335, | |
| "learning_rate": 1.4179843959351213e-06, | |
| "loss": 1.5039, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.5737704918032787, | |
| "grad_norm": 0.12739604211462074, | |
| "learning_rate": 1.3790005506669643e-06, | |
| "loss": 1.4761, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.5812220566318927, | |
| "grad_norm": 0.1164562209744072, | |
| "learning_rate": 1.340098809332401e-06, | |
| "loss": 1.4594, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.5886736214605067, | |
| "grad_norm": 0.1264012532912417, | |
| "learning_rate": 1.3013055686299683e-06, | |
| "loss": 1.4098, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.5961251862891207, | |
| "grad_norm": 0.1165323108273604, | |
| "learning_rate": 1.2626471516353158e-06, | |
| "loss": 1.4998, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.6035767511177347, | |
| "grad_norm": 0.12051657476409453, | |
| "learning_rate": 1.22414978993974e-06, | |
| "loss": 1.3897, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.6110283159463488, | |
| "grad_norm": 0.1326292969168002, | |
| "learning_rate": 1.1858396058507837e-06, | |
| "loss": 1.543, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.6184798807749627, | |
| "grad_norm": 0.12455825193543946, | |
| "learning_rate": 1.1477425946670016e-06, | |
| "loss": 1.4503, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.6259314456035767, | |
| "grad_norm": 0.12370626954874395, | |
| "learning_rate": 1.1098846070389027e-06, | |
| "loss": 1.4326, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.6333830104321908, | |
| "grad_norm": 0.12765662301993327, | |
| "learning_rate": 1.0722913314280395e-06, | |
| "loss": 1.4533, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.6408345752608048, | |
| "grad_norm": 0.11436035436770554, | |
| "learning_rate": 1.0349882766761573e-06, | |
| "loss": 1.3693, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.6482861400894188, | |
| "grad_norm": 0.11842787699252887, | |
| "learning_rate": 9.980007546962206e-07, | |
| "loss": 1.4324, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.6557377049180327, | |
| "grad_norm": 0.11630059586095602, | |
| "learning_rate": 9.613538632970634e-07, | |
| "loss": 1.4374, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.6631892697466468, | |
| "grad_norm": 0.11529623096240363, | |
| "learning_rate": 9.250724691533223e-07, | |
| "loss": 1.4225, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.6706408345752608, | |
| "grad_norm": 0.11523744193256093, | |
| "learning_rate": 8.891811909322058e-07, | |
| "loss": 1.4843, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.6780923994038748, | |
| "grad_norm": 0.13507595363725217, | |
| "learning_rate": 8.537043825885445e-07, | |
| "loss": 1.4891, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.6855439642324889, | |
| "grad_norm": 0.11541281167932681, | |
| "learning_rate": 8.186661168394658e-07, | |
| "loss": 1.4318, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.6929955290611028, | |
| "grad_norm": 0.12007185600365232, | |
| "learning_rate": 7.840901688299e-07, | |
| "loss": 1.4914, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.7004470938897168, | |
| "grad_norm": 0.10337470443027447, | |
| "learning_rate": 7.500000000000003e-07, | |
| "loss": 1.3677, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.7078986587183308, | |
| "grad_norm": 0.1181215216758366, | |
| "learning_rate": 7.16418742165435e-07, | |
| "loss": 1.4494, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.7153502235469449, | |
| "grad_norm": 0.11631524814806833, | |
| "learning_rate": 6.83369181821336e-07, | |
| "loss": 1.4408, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.7228017883755589, | |
| "grad_norm": 0.11883945369899837, | |
| "learning_rate": 6.508737446805704e-07, | |
| "loss": 1.4921, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.7302533532041728, | |
| "grad_norm": 0.12334235443369139, | |
| "learning_rate": 6.189544804568165e-07, | |
| "loss": 1.4098, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.7377049180327869, | |
| "grad_norm": 0.11865941207884222, | |
| "learning_rate": 5.876330479027766e-07, | |
| "loss": 1.436, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.7451564828614009, | |
| "grad_norm": 0.11288437598350734, | |
| "learning_rate": 5.56930700113673e-07, | |
| "loss": 1.349, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.7526080476900149, | |
| "grad_norm": 0.12263443449722483, | |
| "learning_rate": 5.268682701060012e-07, | |
| "loss": 1.3842, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.7600596125186289, | |
| "grad_norm": 0.13248150696752248, | |
| "learning_rate": 4.974661566813315e-07, | |
| "loss": 1.4917, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.767511177347243, | |
| "grad_norm": 0.1235746701113258, | |
| "learning_rate": 4.6874431058474127e-07, | |
| "loss": 1.3985, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.7749627421758569, | |
| "grad_norm": 0.12435997349736658, | |
| "learning_rate": 4.4072222096727663e-07, | |
| "loss": 1.3998, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.7824143070044709, | |
| "grad_norm": 0.12146095464339451, | |
| "learning_rate": 4.1341890216162934e-07, | |
| "loss": 1.4523, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.789865871833085, | |
| "grad_norm": 0.11678985797029438, | |
| "learning_rate": 3.868528807799988e-07, | |
| "loss": 1.4505, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.797317436661699, | |
| "grad_norm": 0.1271289808154804, | |
| "learning_rate": 3.610421831428953e-07, | |
| "loss": 1.4415, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.8047690014903129, | |
| "grad_norm": 0.12203618843739349, | |
| "learning_rate": 3.36004323047419e-07, | |
| "loss": 1.4415, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.812220566318927, | |
| "grad_norm": 0.12970737402513247, | |
| "learning_rate": 3.11756289883306e-07, | |
| "loss": 1.4382, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.819672131147541, | |
| "grad_norm": 0.1290193224680504, | |
| "learning_rate": 2.883145371048133e-07, | |
| "loss": 1.4756, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.827123695976155, | |
| "grad_norm": 0.11595098181221819, | |
| "learning_rate": 2.656949710662591e-07, | |
| "loss": 1.4355, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.834575260804769, | |
| "grad_norm": 0.11852958967235772, | |
| "learning_rate": 2.4391294022879947e-07, | |
| "loss": 1.4338, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.842026825633383, | |
| "grad_norm": 0.12831441033578755, | |
| "learning_rate": 2.2298322474575838e-07, | |
| "loss": 1.4586, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.849478390461997, | |
| "grad_norm": 0.1226839578398384, | |
| "learning_rate": 2.0292002643358892e-07, | |
| "loss": 1.394, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.856929955290611, | |
| "grad_norm": 0.1321673633340756, | |
| "learning_rate": 1.8373695913525317e-07, | |
| "loss": 1.4389, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.8643815201192251, | |
| "grad_norm": 0.13356949132529727, | |
| "learning_rate": 1.6544703948258172e-07, | |
| "loss": 1.4139, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.8718330849478391, | |
| "grad_norm": 0.12214303306764548, | |
| "learning_rate": 1.4806267806386093e-07, | |
| "loss": 1.4682, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.879284649776453, | |
| "grad_norm": 0.1127695145084169, | |
| "learning_rate": 1.3159567100265506e-07, | |
| "loss": 1.3907, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.886736214605067, | |
| "grad_norm": 0.11667918336134676, | |
| "learning_rate": 1.1605719195356806e-07, | |
| "loss": 1.4541, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.8941877794336811, | |
| "grad_norm": 0.12072677981772588, | |
| "learning_rate": 1.0145778452038629e-07, | |
| "loss": 1.3466, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.9016393442622951, | |
| "grad_norm": 0.1127152867709536, | |
| "learning_rate": 8.780735510173316e-08, | |
| "loss": 1.4243, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.9090909090909091, | |
| "grad_norm": 0.12054633218635875, | |
| "learning_rate": 7.51151661691048e-08, | |
| "loss": 1.4678, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.9165424739195231, | |
| "grad_norm": 0.1296306157204084, | |
| "learning_rate": 6.338982998183856e-08, | |
| "loss": 1.4972, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.9239940387481371, | |
| "grad_norm": 0.12472731735613568, | |
| "learning_rate": 5.263930274328044e-08, | |
| "loss": 1.4058, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.9314456035767511, | |
| "grad_norm": 0.10862305460842242, | |
| "learning_rate": 4.287087920212035e-08, | |
| "loss": 1.3878, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.9388971684053651, | |
| "grad_norm": 0.11400949590495348, | |
| "learning_rate": 3.4091187702554485e-08, | |
| "loss": 1.4539, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.9463487332339792, | |
| "grad_norm": 0.12503890566941747, | |
| "learning_rate": 2.630618568663584e-08, | |
| "loss": 1.4149, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.9538002980625931, | |
| "grad_norm": 0.11857672598071958, | |
| "learning_rate": 1.9521155651863854e-08, | |
| "loss": 1.4774, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.9612518628912071, | |
| "grad_norm": 0.12362094771502449, | |
| "learning_rate": 1.3740701566756276e-08, | |
| "loss": 1.4155, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.9687034277198212, | |
| "grad_norm": 0.12708375923965276, | |
| "learning_rate": 8.968745746835983e-09, | |
| "loss": 1.3642, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.9761549925484352, | |
| "grad_norm": 0.1196512207904661, | |
| "learning_rate": 5.208526193150764e-09, | |
| "loss": 1.4681, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.9836065573770492, | |
| "grad_norm": 0.12024283259629832, | |
| "learning_rate": 2.462594395134854e-09, | |
| "loss": 1.4664, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.9910581222056631, | |
| "grad_norm": 0.11574911429092992, | |
| "learning_rate": 7.328135993011631e-10, | |
| "loss": 1.4406, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.9985096870342772, | |
| "grad_norm": 0.11828232321325595, | |
| "learning_rate": 2.035754493812103e-11, | |
| "loss": 1.4711, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_loss": 1.4359323978424072, | |
| "eval_runtime": 564.1313, | |
| "eval_samples_per_second": 4.23, | |
| "eval_steps_per_second": 0.133, | |
| "step": 671 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "step": 671, | |
| "total_flos": 2239826553733120.0, | |
| "train_loss": 1.4986209796544514, | |
| "train_runtime": 17014.7656, | |
| "train_samples_per_second": 1.262, | |
| "train_steps_per_second": 0.039 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 671, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "save_steps": 200, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 2239826553733120.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |