| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9994632313472893, | |
| "eval_steps": 500, | |
| "global_step": 931, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.0010735373054213634, | |
| "grad_norm": 22.07959914956398, | |
| "learning_rate": 1.0638297872340426e-07, | |
| "loss": 1.3304, | |
| "step": 1 | |
| }, | |
| { | |
| "epoch": 0.005367686527106817, | |
| "grad_norm": 20.362741832883, | |
| "learning_rate": 5.319148936170213e-07, | |
| "loss": 1.3199, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.010735373054213635, | |
| "grad_norm": 9.022624585805783, | |
| "learning_rate": 1.0638297872340427e-06, | |
| "loss": 1.1933, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.01610305958132045, | |
| "grad_norm": 7.978147779985442, | |
| "learning_rate": 1.595744680851064e-06, | |
| "loss": 1.0373, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.02147074610842727, | |
| "grad_norm": 2.9506039855166453, | |
| "learning_rate": 2.1276595744680853e-06, | |
| "loss": 0.9114, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.026838432635534086, | |
| "grad_norm": 2.3851062480608025, | |
| "learning_rate": 2.6595744680851065e-06, | |
| "loss": 0.8679, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.0322061191626409, | |
| "grad_norm": 2.2663674501530258, | |
| "learning_rate": 3.191489361702128e-06, | |
| "loss": 0.8324, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.03757380568974772, | |
| "grad_norm": 2.12295130188594, | |
| "learning_rate": 3.723404255319149e-06, | |
| "loss": 0.8086, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.04294149221685454, | |
| "grad_norm": 2.156011284979002, | |
| "learning_rate": 4.255319148936171e-06, | |
| "loss": 0.8121, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.04830917874396135, | |
| "grad_norm": 2.4610831728611906, | |
| "learning_rate": 4.787234042553192e-06, | |
| "loss": 0.7828, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 0.05367686527106817, | |
| "grad_norm": 2.3689072302884027, | |
| "learning_rate": 5.319148936170213e-06, | |
| "loss": 0.7646, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.059044551798174985, | |
| "grad_norm": 2.429397980263518, | |
| "learning_rate": 5.851063829787235e-06, | |
| "loss": 0.7568, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 0.0644122383252818, | |
| "grad_norm": 2.37804165752358, | |
| "learning_rate": 6.382978723404256e-06, | |
| "loss": 0.7398, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.06977992485238863, | |
| "grad_norm": 2.524295711780105, | |
| "learning_rate": 6.914893617021278e-06, | |
| "loss": 0.7251, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 0.07514761137949544, | |
| "grad_norm": 2.4359530361530624, | |
| "learning_rate": 7.446808510638298e-06, | |
| "loss": 0.7183, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08051529790660225, | |
| "grad_norm": 2.3732088149554795, | |
| "learning_rate": 7.97872340425532e-06, | |
| "loss": 0.7165, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 0.08588298443370908, | |
| "grad_norm": 2.319404881166218, | |
| "learning_rate": 8.510638297872341e-06, | |
| "loss": 0.7049, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09125067096081589, | |
| "grad_norm": 2.467420301262588, | |
| "learning_rate": 9.042553191489362e-06, | |
| "loss": 0.6982, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 0.0966183574879227, | |
| "grad_norm": 2.327678008320516, | |
| "learning_rate": 9.574468085106385e-06, | |
| "loss": 0.7016, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.10198604401502952, | |
| "grad_norm": 2.336742003872521, | |
| "learning_rate": 9.999964780082996e-06, | |
| "loss": 0.6895, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 0.10735373054213634, | |
| "grad_norm": 2.261847003362599, | |
| "learning_rate": 9.998732135085665e-06, | |
| "loss": 0.7028, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11272141706924316, | |
| "grad_norm": 2.0345743029930397, | |
| "learning_rate": 9.995738990383743e-06, | |
| "loss": 0.6877, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 0.11808910359634997, | |
| "grad_norm": 2.424929767270555, | |
| "learning_rate": 9.990986400130607e-06, | |
| "loss": 0.676, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12345679012345678, | |
| "grad_norm": 2.1689597645668814, | |
| "learning_rate": 9.984476038137437e-06, | |
| "loss": 0.6863, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 0.1288244766505636, | |
| "grad_norm": 2.0449283955231192, | |
| "learning_rate": 9.97621019728372e-06, | |
| "loss": 0.6746, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13419216317767044, | |
| "grad_norm": 2.0697918308040144, | |
| "learning_rate": 9.966191788709716e-06, | |
| "loss": 0.6741, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 0.13955984970477725, | |
| "grad_norm": 2.0070574066142184, | |
| "learning_rate": 9.954424340791195e-06, | |
| "loss": 0.6586, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14492753623188406, | |
| "grad_norm": 2.0705850233290386, | |
| "learning_rate": 9.940911997896774e-06, | |
| "loss": 0.6696, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 0.15029522275899088, | |
| "grad_norm": 2.0521551046671624, | |
| "learning_rate": 9.925659518928316e-06, | |
| "loss": 0.6581, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.1556629092860977, | |
| "grad_norm": 2.085647742610215, | |
| "learning_rate": 9.908672275644898e-06, | |
| "loss": 0.6538, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 0.1610305958132045, | |
| "grad_norm": 2.0748876889210024, | |
| "learning_rate": 9.889956250770933e-06, | |
| "loss": 0.6582, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16639828234031132, | |
| "grad_norm": 1.9231602777551962, | |
| "learning_rate": 9.86951803588912e-06, | |
| "loss": 0.6547, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 0.17176596886741816, | |
| "grad_norm": 2.1450314001651023, | |
| "learning_rate": 9.847364829118963e-06, | |
| "loss": 0.6437, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17713365539452497, | |
| "grad_norm": 2.100281454841468, | |
| "learning_rate": 9.82350443258166e-06, | |
| "loss": 0.6369, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 0.18250134192163178, | |
| "grad_norm": 1.911217221900268, | |
| "learning_rate": 9.797945249652295e-06, | |
| "loss": 0.6348, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1878690284487386, | |
| "grad_norm": 2.09075575294963, | |
| "learning_rate": 9.770696282000245e-06, | |
| "loss": 0.6488, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 0.1932367149758454, | |
| "grad_norm": 2.0986149327599795, | |
| "learning_rate": 9.741767126418898e-06, | |
| "loss": 0.6301, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19860440150295222, | |
| "grad_norm": 1.9725424031547765, | |
| "learning_rate": 9.711167971445766e-06, | |
| "loss": 0.6215, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 0.20397208803005903, | |
| "grad_norm": 1.9386685500016003, | |
| "learning_rate": 9.67890959377418e-06, | |
| "loss": 0.6316, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.20933977455716588, | |
| "grad_norm": 2.00224716719208, | |
| "learning_rate": 9.645003354457872e-06, | |
| "loss": 0.624, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 0.2147074610842727, | |
| "grad_norm": 1.9315585498716268, | |
| "learning_rate": 9.60946119490972e-06, | |
| "loss": 0.6194, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.2200751476113795, | |
| "grad_norm": 2.140910573853744, | |
| "learning_rate": 9.57229563269612e-06, | |
| "loss": 0.6174, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 0.22544283413848631, | |
| "grad_norm": 2.1739763171596884, | |
| "learning_rate": 9.533519757128426e-06, | |
| "loss": 0.6231, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.23081052066559313, | |
| "grad_norm": 1.9791887683055185, | |
| "learning_rate": 9.49314722465304e-06, | |
| "loss": 0.5963, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 0.23617820719269994, | |
| "grad_norm": 2.0935279890007785, | |
| "learning_rate": 9.451192254041759e-06, | |
| "loss": 0.6165, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.24154589371980675, | |
| "grad_norm": 1.9389350149885696, | |
| "learning_rate": 9.407669621384073e-06, | |
| "loss": 0.6077, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 0.24691358024691357, | |
| "grad_norm": 1.9673443855756516, | |
| "learning_rate": 9.362594654883185e-06, | |
| "loss": 0.5986, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2522812667740204, | |
| "grad_norm": 1.9267038189193195, | |
| "learning_rate": 9.31598322945759e-06, | |
| "loss": 0.6209, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 0.2576489533011272, | |
| "grad_norm": 1.9705398985631108, | |
| "learning_rate": 9.267851761150092e-06, | |
| "loss": 0.5949, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.26301663982823403, | |
| "grad_norm": 2.763348732615502, | |
| "learning_rate": 9.218217201346251e-06, | |
| "loss": 0.6003, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 0.2683843263553409, | |
| "grad_norm": 1.9828807479273434, | |
| "learning_rate": 9.167097030804289e-06, | |
| "loss": 0.5967, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.27375201288244766, | |
| "grad_norm": 2.067267771380607, | |
| "learning_rate": 9.114509253498554e-06, | |
| "loss": 0.5906, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 0.2791196994095545, | |
| "grad_norm": 2.1736682143187553, | |
| "learning_rate": 9.060472390278717e-06, | |
| "loss": 0.5786, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.2844873859366613, | |
| "grad_norm": 2.3050896737016333, | |
| "learning_rate": 9.005005472346923e-06, | |
| "loss": 0.5752, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 0.2898550724637681, | |
| "grad_norm": 2.1110962605389147, | |
| "learning_rate": 8.948128034555212e-06, | |
| "loss": 0.5941, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2952227589908749, | |
| "grad_norm": 2.276509382528488, | |
| "learning_rate": 8.889860108525544e-06, | |
| "loss": 0.5666, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 0.30059044551798175, | |
| "grad_norm": 2.0249679866823644, | |
| "learning_rate": 8.83022221559489e-06, | |
| "loss": 0.5834, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.3059581320450886, | |
| "grad_norm": 2.0546291551649976, | |
| "learning_rate": 8.76923535958783e-06, | |
| "loss": 0.5675, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 0.3113258185721954, | |
| "grad_norm": 2.2236300247259697, | |
| "learning_rate": 8.706921019419237e-06, | |
| "loss": 0.563, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3166935050993022, | |
| "grad_norm": 1.8037385365479262, | |
| "learning_rate": 8.643301141529619e-06, | |
| "loss": 0.5748, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 0.322061191626409, | |
| "grad_norm": 1.985529414306051, | |
| "learning_rate": 8.578398132155846e-06, | |
| "loss": 0.5578, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.32742887815351585, | |
| "grad_norm": 2.0083187215038674, | |
| "learning_rate": 8.512234849439887e-06, | |
| "loss": 0.5722, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 0.33279656468062263, | |
| "grad_norm": 1.968155782313349, | |
| "learning_rate": 8.444834595378434e-06, | |
| "loss": 0.5674, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.33816425120772947, | |
| "grad_norm": 2.0307121236920613, | |
| "learning_rate": 8.376221107616187e-06, | |
| "loss": 0.5596, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 0.3435319377348363, | |
| "grad_norm": 2.150965816694603, | |
| "learning_rate": 8.306418551085707e-06, | |
| "loss": 0.5505, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.3488996242619431, | |
| "grad_norm": 1.9333182701879525, | |
| "learning_rate": 8.23545150949679e-06, | |
| "loss": 0.5407, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 0.35426731078904994, | |
| "grad_norm": 2.0408489065141673, | |
| "learning_rate": 8.163344976678342e-06, | |
| "loss": 0.5341, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3596349973161567, | |
| "grad_norm": 2.039571243808379, | |
| "learning_rate": 8.090124347775837e-06, | |
| "loss": 0.5386, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 0.36500268384326356, | |
| "grad_norm": 1.907167090419826, | |
| "learning_rate": 8.0158154103074e-06, | |
| "loss": 0.5481, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.37037037037037035, | |
| "grad_norm": 1.960372580403973, | |
| "learning_rate": 7.940444335081733e-06, | |
| "loss": 0.5399, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 0.3757380568974772, | |
| "grad_norm": 2.055565141792339, | |
| "learning_rate": 7.864037666981037e-06, | |
| "loss": 0.5329, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.38110574342458403, | |
| "grad_norm": 1.9530633167577272, | |
| "learning_rate": 7.786622315612182e-06, | |
| "loss": 0.5281, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 0.3864734299516908, | |
| "grad_norm": 1.9983415429313947, | |
| "learning_rate": 7.708225545829446e-06, | |
| "loss": 0.5332, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.39184111647879766, | |
| "grad_norm": 2.074941439605804, | |
| "learning_rate": 7.6288749681321115e-06, | |
| "loss": 0.5182, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 0.39720880300590444, | |
| "grad_norm": 1.9934314007469667, | |
| "learning_rate": 7.548598528940354e-06, | |
| "loss": 0.526, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.4025764895330113, | |
| "grad_norm": 2.033407744493333, | |
| "learning_rate": 7.4674245007528135e-06, | |
| "loss": 0.5263, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 0.40794417606011807, | |
| "grad_norm": 1.9620705282952844, | |
| "learning_rate": 7.385381472189321e-06, | |
| "loss": 0.518, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.4133118625872249, | |
| "grad_norm": 1.9020125127771519, | |
| "learning_rate": 7.302498337922293e-06, | |
| "loss": 0.5154, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 0.41867954911433175, | |
| "grad_norm": 2.0750218270909966, | |
| "learning_rate": 7.218804288500343e-06, | |
| "loss": 0.5147, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.42404723564143854, | |
| "grad_norm": 2.040651657120084, | |
| "learning_rate": 7.134328800067684e-06, | |
| "loss": 0.5099, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 0.4294149221685454, | |
| "grad_norm": 1.9898945970825554, | |
| "learning_rate": 7.049101623982938e-06, | |
| "loss": 0.4993, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.43478260869565216, | |
| "grad_norm": 1.9951295882452424, | |
| "learning_rate": 6.963152776341044e-06, | |
| "loss": 0.5104, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 0.440150295222759, | |
| "grad_norm": 1.8963421029024186, | |
| "learning_rate": 6.876512527401897e-06, | |
| "loss": 0.4934, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4455179817498658, | |
| "grad_norm": 2.0348217650465785, | |
| "learning_rate": 6.789211390929497e-06, | |
| "loss": 0.5073, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 0.45088566827697263, | |
| "grad_norm": 2.1752799116340986, | |
| "learning_rate": 6.701280113445324e-06, | |
| "loss": 0.4968, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.4562533548040794, | |
| "grad_norm": 1.9168080270884666, | |
| "learning_rate": 6.6127496633997475e-06, | |
| "loss": 0.4881, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 0.46162104133118625, | |
| "grad_norm": 1.8791343828515361, | |
| "learning_rate": 6.523651220265269e-06, | |
| "loss": 0.4917, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4669887278582931, | |
| "grad_norm": 1.9398738713179469, | |
| "learning_rate": 6.434016163555452e-06, | |
| "loss": 0.4842, | |
| "step": 435 | |
| }, | |
| { | |
| "epoch": 0.4723564143853999, | |
| "grad_norm": 1.8681466257367256, | |
| "learning_rate": 6.343876061773385e-06, | |
| "loss": 0.4888, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.4777241009125067, | |
| "grad_norm": 1.8826036572462372, | |
| "learning_rate": 6.2532626612936035e-06, | |
| "loss": 0.4922, | |
| "step": 445 | |
| }, | |
| { | |
| "epoch": 0.4830917874396135, | |
| "grad_norm": 1.9400538781987646, | |
| "learning_rate": 6.162207875181354e-06, | |
| "loss": 0.4783, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.48845947396672035, | |
| "grad_norm": 1.9864272178376912, | |
| "learning_rate": 6.070743771953157e-06, | |
| "loss": 0.4754, | |
| "step": 455 | |
| }, | |
| { | |
| "epoch": 0.49382716049382713, | |
| "grad_norm": 1.943057752510615, | |
| "learning_rate": 5.978902564282616e-06, | |
| "loss": 0.4811, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.499194847020934, | |
| "grad_norm": 2.0041710832979547, | |
| "learning_rate": 5.886716597655472e-06, | |
| "loss": 0.4717, | |
| "step": 465 | |
| }, | |
| { | |
| "epoch": 0.5045625335480408, | |
| "grad_norm": 1.9473535928377121, | |
| "learning_rate": 5.7942183389778536e-06, | |
| "loss": 0.4766, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.5099302200751477, | |
| "grad_norm": 2.061547109286419, | |
| "learning_rate": 5.701440365141799e-06, | |
| "loss": 0.4667, | |
| "step": 475 | |
| }, | |
| { | |
| "epoch": 0.5152979066022544, | |
| "grad_norm": 1.9952943514062236, | |
| "learning_rate": 5.608415351552014e-06, | |
| "loss": 0.461, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.5206655931293612, | |
| "grad_norm": 2.159739403315762, | |
| "learning_rate": 5.515176060617945e-06, | |
| "loss": 0.4817, | |
| "step": 485 | |
| }, | |
| { | |
| "epoch": 0.5260332796564681, | |
| "grad_norm": 2.098839883318929, | |
| "learning_rate": 5.421755330215223e-06, | |
| "loss": 0.4626, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5314009661835749, | |
| "grad_norm": 1.893289381010737, | |
| "learning_rate": 5.328186062120509e-06, | |
| "loss": 0.4493, | |
| "step": 495 | |
| }, | |
| { | |
| "epoch": 0.5367686527106817, | |
| "grad_norm": 2.034865472461479, | |
| "learning_rate": 5.23450121042383e-06, | |
| "loss": 0.461, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5421363392377885, | |
| "grad_norm": 1.913832222398215, | |
| "learning_rate": 5.140733769922525e-06, | |
| "loss": 0.4555, | |
| "step": 505 | |
| }, | |
| { | |
| "epoch": 0.5475040257648953, | |
| "grad_norm": 1.9315440299324362, | |
| "learning_rate": 5.0469167645008245e-06, | |
| "loss": 0.4594, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5528717122920022, | |
| "grad_norm": 2.0634590710960263, | |
| "learning_rate": 4.953083235499177e-06, | |
| "loss": 0.4485, | |
| "step": 515 | |
| }, | |
| { | |
| "epoch": 0.558239398819109, | |
| "grad_norm": 1.8668954388857975, | |
| "learning_rate": 4.859266230077474e-06, | |
| "loss": 0.4483, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5636070853462157, | |
| "grad_norm": 1.844866191674477, | |
| "learning_rate": 4.7654987895761705e-06, | |
| "loss": 0.4436, | |
| "step": 525 | |
| }, | |
| { | |
| "epoch": 0.5689747718733226, | |
| "grad_norm": 1.9794533025528265, | |
| "learning_rate": 4.671813937879494e-06, | |
| "loss": 0.441, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5743424584004294, | |
| "grad_norm": 1.883767510221611, | |
| "learning_rate": 4.5782446697847775e-06, | |
| "loss": 0.4432, | |
| "step": 535 | |
| }, | |
| { | |
| "epoch": 0.5797101449275363, | |
| "grad_norm": 1.976673677149353, | |
| "learning_rate": 4.484823939382056e-06, | |
| "loss": 0.4358, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5850778314546431, | |
| "grad_norm": 1.9433175492111152, | |
| "learning_rate": 4.391584648447989e-06, | |
| "loss": 0.4319, | |
| "step": 545 | |
| }, | |
| { | |
| "epoch": 0.5904455179817498, | |
| "grad_norm": 1.985070698744384, | |
| "learning_rate": 4.298559634858202e-06, | |
| "loss": 0.4354, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5958132045088567, | |
| "grad_norm": 2.01070608130013, | |
| "learning_rate": 4.205781661022146e-06, | |
| "loss": 0.4267, | |
| "step": 555 | |
| }, | |
| { | |
| "epoch": 0.6011808910359635, | |
| "grad_norm": 1.9373278557153377, | |
| "learning_rate": 4.1132834023445304e-06, | |
| "loss": 0.4208, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.6065485775630703, | |
| "grad_norm": 1.896977030964561, | |
| "learning_rate": 4.021097435717386e-06, | |
| "loss": 0.4254, | |
| "step": 565 | |
| }, | |
| { | |
| "epoch": 0.6119162640901772, | |
| "grad_norm": 1.9694900821692567, | |
| "learning_rate": 3.929256228046845e-06, | |
| "loss": 0.4178, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.6172839506172839, | |
| "grad_norm": 1.9130375537620703, | |
| "learning_rate": 3.837792124818647e-06, | |
| "loss": 0.4256, | |
| "step": 575 | |
| }, | |
| { | |
| "epoch": 0.6226516371443908, | |
| "grad_norm": 2.148799842524908, | |
| "learning_rate": 3.7467373387063973e-06, | |
| "loss": 0.4172, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.6280193236714976, | |
| "grad_norm": 1.9520935395346655, | |
| "learning_rate": 3.656123938226618e-06, | |
| "loss": 0.4209, | |
| "step": 585 | |
| }, | |
| { | |
| "epoch": 0.6333870101986044, | |
| "grad_norm": 1.9284426230497334, | |
| "learning_rate": 3.5659838364445505e-06, | |
| "loss": 0.4135, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.6387546967257112, | |
| "grad_norm": 3.083878858254763, | |
| "learning_rate": 3.476348779734732e-06, | |
| "loss": 0.4067, | |
| "step": 595 | |
| }, | |
| { | |
| "epoch": 0.644122383252818, | |
| "grad_norm": 1.956096320950966, | |
| "learning_rate": 3.387250336600254e-06, | |
| "loss": 0.4193, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6494900697799249, | |
| "grad_norm": 1.7978680707200165, | |
| "learning_rate": 3.298719886554677e-06, | |
| "loss": 0.4082, | |
| "step": 605 | |
| }, | |
| { | |
| "epoch": 0.6548577563070317, | |
| "grad_norm": 1.8632209883906246, | |
| "learning_rate": 3.2107886090705035e-06, | |
| "loss": 0.4009, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6602254428341385, | |
| "grad_norm": 1.9564221459543165, | |
| "learning_rate": 3.1234874725981045e-06, | |
| "loss": 0.4116, | |
| "step": 615 | |
| }, | |
| { | |
| "epoch": 0.6655931293612453, | |
| "grad_norm": 2.1099975602879217, | |
| "learning_rate": 3.036847223658958e-06, | |
| "loss": 0.4046, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6709608158883521, | |
| "grad_norm": 1.923676982794732, | |
| "learning_rate": 2.950898376017064e-06, | |
| "loss": 0.4022, | |
| "step": 625 | |
| }, | |
| { | |
| "epoch": 0.6763285024154589, | |
| "grad_norm": 1.9781538947455772, | |
| "learning_rate": 2.865671199932318e-06, | |
| "loss": 0.4084, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6816961889425658, | |
| "grad_norm": 1.8903547449677651, | |
| "learning_rate": 2.781195711499658e-06, | |
| "loss": 0.399, | |
| "step": 635 | |
| }, | |
| { | |
| "epoch": 0.6870638754696726, | |
| "grad_norm": 1.812136545065083, | |
| "learning_rate": 2.697501662077707e-06, | |
| "loss": 0.3981, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6924315619967794, | |
| "grad_norm": 1.8634900584603342, | |
| "learning_rate": 2.6146185278106807e-06, | |
| "loss": 0.4009, | |
| "step": 645 | |
| }, | |
| { | |
| "epoch": 0.6977992485238862, | |
| "grad_norm": 1.8908371985643206, | |
| "learning_rate": 2.5325754992471886e-06, | |
| "loss": 0.3873, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.703166935050993, | |
| "grad_norm": 1.825726975173959, | |
| "learning_rate": 2.4514014710596467e-06, | |
| "loss": 0.3956, | |
| "step": 655 | |
| }, | |
| { | |
| "epoch": 0.7085346215780999, | |
| "grad_norm": 1.8145985790587735, | |
| "learning_rate": 2.3711250318678906e-06, | |
| "loss": 0.387, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.7139023081052066, | |
| "grad_norm": 1.9318642386352745, | |
| "learning_rate": 2.2917744541705544e-06, | |
| "loss": 0.3915, | |
| "step": 665 | |
| }, | |
| { | |
| "epoch": 0.7192699946323134, | |
| "grad_norm": 1.8371335401848725, | |
| "learning_rate": 2.2133776843878185e-06, | |
| "loss": 0.3913, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.7246376811594203, | |
| "grad_norm": 1.8857468128113828, | |
| "learning_rate": 2.1359623330189655e-06, | |
| "loss": 0.382, | |
| "step": 675 | |
| }, | |
| { | |
| "epoch": 0.7300053676865271, | |
| "grad_norm": 1.847380544620127, | |
| "learning_rate": 2.059555664918268e-06, | |
| "loss": 0.3768, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.735373054213634, | |
| "grad_norm": 1.7802526887131536, | |
| "learning_rate": 1.9841845896926022e-06, | |
| "loss": 0.3842, | |
| "step": 685 | |
| }, | |
| { | |
| "epoch": 0.7407407407407407, | |
| "grad_norm": 1.8355700135485964, | |
| "learning_rate": 1.9098756522241634e-06, | |
| "loss": 0.3803, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7461084272678475, | |
| "grad_norm": 1.8378538582720527, | |
| "learning_rate": 1.8366550233216584e-06, | |
| "loss": 0.3774, | |
| "step": 695 | |
| }, | |
| { | |
| "epoch": 0.7514761137949544, | |
| "grad_norm": 1.8843628517686122, | |
| "learning_rate": 1.7645484905032129e-06, | |
| "loss": 0.3804, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7568438003220612, | |
| "grad_norm": 1.842487842009676, | |
| "learning_rate": 1.6935814489142937e-06, | |
| "loss": 0.3697, | |
| "step": 705 | |
| }, | |
| { | |
| "epoch": 0.7622114868491681, | |
| "grad_norm": 1.8129216270941833, | |
| "learning_rate": 1.6237788923838149e-06, | |
| "loss": 0.374, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7675791733762748, | |
| "grad_norm": 1.9032083523775989, | |
| "learning_rate": 1.555165404621567e-06, | |
| "loss": 0.3755, | |
| "step": 715 | |
| }, | |
| { | |
| "epoch": 0.7729468599033816, | |
| "grad_norm": 1.8048406689197223, | |
| "learning_rate": 1.487765150560116e-06, | |
| "loss": 0.37, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7783145464304885, | |
| "grad_norm": 1.7679512404980078, | |
| "learning_rate": 1.4216018678441558e-06, | |
| "loss": 0.371, | |
| "step": 725 | |
| }, | |
| { | |
| "epoch": 0.7836822329575953, | |
| "grad_norm": 1.7580438160255716, | |
| "learning_rate": 1.3566988584703817e-06, | |
| "loss": 0.3744, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.789049919484702, | |
| "grad_norm": 1.9423775433056052, | |
| "learning_rate": 1.293078980580766e-06, | |
| "loss": 0.3648, | |
| "step": 735 | |
| }, | |
| { | |
| "epoch": 0.7944176060118089, | |
| "grad_norm": 1.8075229160199813, | |
| "learning_rate": 1.2307646404121692e-06, | |
| "loss": 0.3793, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7997852925389157, | |
| "grad_norm": 1.7960898775042207, | |
| "learning_rate": 1.1697777844051105e-06, | |
| "loss": 0.3658, | |
| "step": 745 | |
| }, | |
| { | |
| "epoch": 0.8051529790660226, | |
| "grad_norm": 1.714976428918269, | |
| "learning_rate": 1.1101398914744565e-06, | |
| "loss": 0.3539, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.8105206655931294, | |
| "grad_norm": 1.9437004560812625, | |
| "learning_rate": 1.0518719654447896e-06, | |
| "loss": 0.3672, | |
| "step": 755 | |
| }, | |
| { | |
| "epoch": 0.8158883521202361, | |
| "grad_norm": 1.8231928241724982, | |
| "learning_rate": 9.949945276530782e-07, | |
| "loss": 0.3623, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.821256038647343, | |
| "grad_norm": 1.768854699968928, | |
| "learning_rate": 9.395276097212841e-07, | |
| "loss": 0.3644, | |
| "step": 765 | |
| }, | |
| { | |
| "epoch": 0.8266237251744498, | |
| "grad_norm": 1.7546007013596414, | |
| "learning_rate": 8.854907465014479e-07, | |
| "loss": 0.3612, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.8319914117015567, | |
| "grad_norm": 1.7931543962107859, | |
| "learning_rate": 8.329029691957124e-07, | |
| "loss": 0.3651, | |
| "step": 775 | |
| }, | |
| { | |
| "epoch": 0.8373590982286635, | |
| "grad_norm": 1.8334730517443687, | |
| "learning_rate": 7.817827986537508e-07, | |
| "loss": 0.3572, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.8427267847557702, | |
| "grad_norm": 1.9211319549733363, | |
| "learning_rate": 7.321482388499096e-07, | |
| "loss": 0.3547, | |
| "step": 785 | |
| }, | |
| { | |
| "epoch": 0.8480944712828771, | |
| "grad_norm": 1.8336015586916503, | |
| "learning_rate": 6.840167705424106e-07, | |
| "loss": 0.3591, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.8534621578099839, | |
| "grad_norm": 1.7443053171841683, | |
| "learning_rate": 6.374053451168166e-07, | |
| "loss": 0.3557, | |
| "step": 795 | |
| }, | |
| { | |
| "epoch": 0.8588298443370908, | |
| "grad_norm": 1.8189610501841909, | |
| "learning_rate": 5.92330378615929e-07, | |
| "loss": 0.3629, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8641975308641975, | |
| "grad_norm": 1.8372980732446615, | |
| "learning_rate": 5.488077459582425e-07, | |
| "loss": 0.355, | |
| "step": 805 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 1.7662263372996434, | |
| "learning_rate": 5.068527753469604e-07, | |
| "loss": 0.3457, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8749329039184112, | |
| "grad_norm": 1.884357397297957, | |
| "learning_rate": 4.664802428715753e-07, | |
| "loss": 0.3555, | |
| "step": 815 | |
| }, | |
| { | |
| "epoch": 0.880300590445518, | |
| "grad_norm": 1.7648193673189987, | |
| "learning_rate": 4.2770436730388166e-07, | |
| "loss": 0.3533, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8856682769726248, | |
| "grad_norm": 1.8350106858958837, | |
| "learning_rate": 3.9053880509028086e-07, | |
| "loss": 0.35, | |
| "step": 825 | |
| }, | |
| { | |
| "epoch": 0.8910359634997316, | |
| "grad_norm": 1.860048919453916, | |
| "learning_rate": 3.549966455421305e-07, | |
| "loss": 0.3513, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8964036500268384, | |
| "grad_norm": 1.7816738115824102, | |
| "learning_rate": 3.2109040622582186e-07, | |
| "loss": 0.3491, | |
| "step": 835 | |
| }, | |
| { | |
| "epoch": 0.9017713365539453, | |
| "grad_norm": 1.7638165624694198, | |
| "learning_rate": 2.8883202855423676e-07, | |
| "loss": 0.3533, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.9071390230810521, | |
| "grad_norm": 1.8839660726806369, | |
| "learning_rate": 2.582328735811029e-07, | |
| "loss": 0.3553, | |
| "step": 845 | |
| }, | |
| { | |
| "epoch": 0.9125067096081588, | |
| "grad_norm": 1.8407087550617782, | |
| "learning_rate": 2.2930371799975593e-07, | |
| "loss": 0.3627, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.9178743961352657, | |
| "grad_norm": 1.801970350663046, | |
| "learning_rate": 2.0205475034770606e-07, | |
| "loss": 0.3552, | |
| "step": 855 | |
| }, | |
| { | |
| "epoch": 0.9232420826623725, | |
| "grad_norm": 1.685025912596546, | |
| "learning_rate": 1.7649556741833995e-07, | |
| "loss": 0.3432, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.9286097691894794, | |
| "grad_norm": 1.675226625798938, | |
| "learning_rate": 1.5263517088103862e-07, | |
| "loss": 0.3531, | |
| "step": 865 | |
| }, | |
| { | |
| "epoch": 0.9339774557165862, | |
| "grad_norm": 1.7824714528651928, | |
| "learning_rate": 1.304819641108801e-07, | |
| "loss": 0.3408, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.9393451422436929, | |
| "grad_norm": 1.7589068127764829, | |
| "learning_rate": 1.1004374922906846e-07, | |
| "loss": 0.3436, | |
| "step": 875 | |
| }, | |
| { | |
| "epoch": 0.9447128287707998, | |
| "grad_norm": 1.8043041133322923, | |
| "learning_rate": 9.132772435510362e-08, | |
| "loss": 0.3557, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.9500805152979066, | |
| "grad_norm": 1.8271300591957, | |
| "learning_rate": 7.434048107168523e-08, | |
| "loss": 0.3413, | |
| "step": 885 | |
| }, | |
| { | |
| "epoch": 0.9554482018250134, | |
| "grad_norm": 1.8377353760781505, | |
| "learning_rate": 5.908800210322696e-08, | |
| "loss": 0.3477, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9608158883521203, | |
| "grad_norm": 1.7930991363661997, | |
| "learning_rate": 4.55756592088058e-08, | |
| "loss": 0.3494, | |
| "step": 895 | |
| }, | |
| { | |
| "epoch": 0.966183574879227, | |
| "grad_norm": 1.7208695719091334, | |
| "learning_rate": 3.3808211290284886e-08, | |
| "loss": 0.3537, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9715512614063339, | |
| "grad_norm": 1.7750738665351566, | |
| "learning_rate": 2.378980271628195e-08, | |
| "loss": 0.3455, | |
| "step": 905 | |
| }, | |
| { | |
| "epoch": 0.9769189479334407, | |
| "grad_norm": 1.7676906709440423, | |
| "learning_rate": 1.552396186256411e-08, | |
| "loss": 0.3522, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9822866344605475, | |
| "grad_norm": 1.7765229197521504, | |
| "learning_rate": 9.013599869394096e-09, | |
| "loss": 0.3372, | |
| "step": 915 | |
| }, | |
| { | |
| "epoch": 0.9876543209876543, | |
| "grad_norm": 1.7778750895811917, | |
| "learning_rate": 4.261009616257638e-09, | |
| "loss": 0.354, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9930220075147611, | |
| "grad_norm": 1.7318086433483029, | |
| "learning_rate": 1.2678649143349485e-09, | |
| "loss": 0.3479, | |
| "step": 925 | |
| }, | |
| { | |
| "epoch": 0.998389694041868, | |
| "grad_norm": 1.7750655216541529, | |
| "learning_rate": 3.5219917003948003e-11, | |
| "loss": 0.3526, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9994632313472893, | |
| "eval_loss": 0.29608339071273804, | |
| "eval_runtime": 1.1911, | |
| "eval_samples_per_second": 1.679, | |
| "eval_steps_per_second": 0.84, | |
| "step": 931 | |
| }, | |
| { | |
| "epoch": 0.9994632313472893, | |
| "step": 931, | |
| "total_flos": 194880114524160.0, | |
| "train_loss": 0.5082121759685102, | |
| "train_runtime": 20748.1716, | |
| "train_samples_per_second": 1.437, | |
| "train_steps_per_second": 0.045 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 931, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 1, | |
| "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": 194880114524160.0, | |
| "train_batch_size": 2, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |