| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 4.987212276214834, | |
| "eval_steps": 500, | |
| "global_step": 975, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.05115089514066496, | |
| "grad_norm": 214.7565900278413, | |
| "learning_rate": 1.0204081632653063e-06, | |
| "loss": 4.0075, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.10230179028132992, | |
| "grad_norm": 22.211069433232247, | |
| "learning_rate": 2.0408163265306125e-06, | |
| "loss": 0.6282, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.1534526854219949, | |
| "grad_norm": 12.11809889509627, | |
| "learning_rate": 3.0612244897959185e-06, | |
| "loss": 0.2107, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.20460358056265984, | |
| "grad_norm": 2.8569999074326495, | |
| "learning_rate": 4.081632653061225e-06, | |
| "loss": 0.2135, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.2557544757033248, | |
| "grad_norm": 6.867087399363433, | |
| "learning_rate": 5.1020408163265315e-06, | |
| "loss": 0.2072, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.3069053708439898, | |
| "grad_norm": 3.4564712619293663, | |
| "learning_rate": 6.122448979591837e-06, | |
| "loss": 0.2002, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.35805626598465473, | |
| "grad_norm": 3.69055204043271, | |
| "learning_rate": 7.1428571428571436e-06, | |
| "loss": 0.2, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.4092071611253197, | |
| "grad_norm": 1.0696324041329028, | |
| "learning_rate": 8.16326530612245e-06, | |
| "loss": 0.1964, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.46035805626598464, | |
| "grad_norm": 2.964347091843066, | |
| "learning_rate": 9.183673469387756e-06, | |
| "loss": 0.1952, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.5115089514066496, | |
| "grad_norm": 3.646664998250867, | |
| "learning_rate": 9.99987167871469e-06, | |
| "loss": 0.2045, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.5626598465473146, | |
| "grad_norm": 1.6191665925446022, | |
| "learning_rate": 9.995381125277747e-06, | |
| "loss": 0.2018, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.6138107416879796, | |
| "grad_norm": 4.017961040145252, | |
| "learning_rate": 9.984481092596683e-06, | |
| "loss": 0.1975, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.6649616368286445, | |
| "grad_norm": 2.9108813585819817, | |
| "learning_rate": 9.967185566297713e-06, | |
| "loss": 0.199, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.7161125319693095, | |
| "grad_norm": 3.1021824472509767, | |
| "learning_rate": 9.94351673794421e-06, | |
| "loss": 0.1978, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.7672634271099744, | |
| "grad_norm": 3.497183421035605, | |
| "learning_rate": 9.913504976563138e-06, | |
| "loss": 0.1972, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.8184143222506394, | |
| "grad_norm": 2.667271306285725, | |
| "learning_rate": 9.877188789679111e-06, | |
| "loss": 0.1972, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.8695652173913043, | |
| "grad_norm": 1.939938871403963, | |
| "learning_rate": 9.834614773906125e-06, | |
| "loss": 0.1983, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.9207161125319693, | |
| "grad_norm": 3.277095189836375, | |
| "learning_rate": 9.785837555160322e-06, | |
| "loss": 0.2001, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.9718670076726342, | |
| "grad_norm": 2.1349632530669784, | |
| "learning_rate": 9.730919718570501e-06, | |
| "loss": 0.1985, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.0230179028132993, | |
| "grad_norm": 1.9921171293582758, | |
| "learning_rate": 9.669931728176329e-06, | |
| "loss": 0.2044, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.0741687979539642, | |
| "grad_norm": 1.6800643640667052, | |
| "learning_rate": 9.602951836517255e-06, | |
| "loss": 0.1968, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.1253196930946292, | |
| "grad_norm": 3.6566429458754515, | |
| "learning_rate": 9.530065984228156e-06, | |
| "loss": 0.1972, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.1764705882352942, | |
| "grad_norm": 1.0633803952953227, | |
| "learning_rate": 9.451367689770532e-06, | |
| "loss": 0.1957, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.227621483375959, | |
| "grad_norm": 1.4368952344952126, | |
| "learning_rate": 9.366957929440731e-06, | |
| "loss": 0.1942, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.278772378516624, | |
| "grad_norm": 2.8115853641591015, | |
| "learning_rate": 9.276945007809167e-06, | |
| "loss": 0.1975, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 1.329923273657289, | |
| "grad_norm": 2.2368726208053515, | |
| "learning_rate": 9.181444418756774e-06, | |
| "loss": 0.1941, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 1.381074168797954, | |
| "grad_norm": 1.718287417414397, | |
| "learning_rate": 9.080578697286986e-06, | |
| "loss": 0.1955, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.432225063938619, | |
| "grad_norm": 2.6480547516752373, | |
| "learning_rate": 8.974477262303382e-06, | |
| "loss": 0.1965, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.4833759590792839, | |
| "grad_norm": 2.1323772112759856, | |
| "learning_rate": 8.86327625055474e-06, | |
| "loss": 0.1928, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.5345268542199488, | |
| "grad_norm": 1.3695013623025927, | |
| "learning_rate": 8.747118341960542e-06, | |
| "loss": 0.1964, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.5856777493606138, | |
| "grad_norm": 2.816490216879556, | |
| "learning_rate": 8.626152576541059e-06, | |
| "loss": 0.1926, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.6368286445012787, | |
| "grad_norm": 2.63475775763493, | |
| "learning_rate": 8.500534163186914e-06, | |
| "loss": 0.1944, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.6879795396419437, | |
| "grad_norm": 1.2976999470451909, | |
| "learning_rate": 8.37042428051349e-06, | |
| "loss": 0.1935, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.7391304347826086, | |
| "grad_norm": 3.0528789638369314, | |
| "learning_rate": 8.235989870055684e-06, | |
| "loss": 0.1931, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.7902813299232738, | |
| "grad_norm": 2.761510505813076, | |
| "learning_rate": 8.097403422068377e-06, | |
| "loss": 0.1901, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.8414322250639388, | |
| "grad_norm": 1.2860400398022556, | |
| "learning_rate": 7.954842754207442e-06, | |
| "loss": 0.1846, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.8925831202046037, | |
| "grad_norm": 1.640012696939897, | |
| "learning_rate": 7.80849078337527e-06, | |
| "loss": 0.1801, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.9437340153452687, | |
| "grad_norm": 2.959712367273353, | |
| "learning_rate": 7.658535291023533e-06, | |
| "loss": 0.1766, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.9948849104859336, | |
| "grad_norm": 2.582572305335923, | |
| "learning_rate": 7.5051686822143614e-06, | |
| "loss": 0.1592, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 2.0460358056265986, | |
| "grad_norm": 2.8473557994991165, | |
| "learning_rate": 7.348587738749037e-06, | |
| "loss": 0.1467, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 2.0971867007672635, | |
| "grad_norm": 2.246956202126538, | |
| "learning_rate": 7.18899336668097e-06, | |
| "loss": 0.1102, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 2.1483375959079285, | |
| "grad_norm": 2.3922923060664685, | |
| "learning_rate": 7.0265903385369385e-06, | |
| "loss": 0.0799, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 2.1994884910485935, | |
| "grad_norm": 2.235940967501667, | |
| "learning_rate": 6.861587030577326e-06, | |
| "loss": 0.0643, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 2.2506393861892584, | |
| "grad_norm": 2.1749048219884615, | |
| "learning_rate": 6.694195155432473e-06, | |
| "loss": 0.039, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 2.3017902813299234, | |
| "grad_norm": 2.4706549250013494, | |
| "learning_rate": 6.524629490458193e-06, | |
| "loss": 0.0287, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 2.3529411764705883, | |
| "grad_norm": 2.1875953818358975, | |
| "learning_rate": 6.353107602159005e-06, | |
| "loss": 0.0211, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 2.4040920716112533, | |
| "grad_norm": 1.1932819949506583, | |
| "learning_rate": 6.179849567032658e-06, | |
| "loss": 0.0199, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 2.455242966751918, | |
| "grad_norm": 3.043082138284078, | |
| "learning_rate": 6.0050776891941266e-06, | |
| "loss": 0.0138, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 2.506393861892583, | |
| "grad_norm": 1.4066811981696432, | |
| "learning_rate": 5.8290162151414e-06, | |
| "loss": 0.0102, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 2.557544757033248, | |
| "grad_norm": 1.089275292286627, | |
| "learning_rate": 5.651891046029031e-06, | |
| "loss": 0.0091, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 2.608695652173913, | |
| "grad_norm": 1.4386597224188378, | |
| "learning_rate": 5.473929447818628e-06, | |
| "loss": 0.0093, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 2.659846547314578, | |
| "grad_norm": 1.6800741407957291, | |
| "learning_rate": 5.295359759678187e-06, | |
| "loss": 0.0062, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 2.710997442455243, | |
| "grad_norm": 2.0143218617779004, | |
| "learning_rate": 5.1164111010044225e-06, | |
| "loss": 0.0058, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 2.762148337595908, | |
| "grad_norm": 2.927032625013264, | |
| "learning_rate": 4.937313077443985e-06, | |
| "loss": 0.0071, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.813299232736573, | |
| "grad_norm": 0.7049312278321461, | |
| "learning_rate": 4.7582954862907896e-06, | |
| "loss": 0.0041, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.864450127877238, | |
| "grad_norm": 0.39393926121790496, | |
| "learning_rate": 4.579588021637448e-06, | |
| "loss": 0.0029, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.915601023017903, | |
| "grad_norm": 0.7023810428810136, | |
| "learning_rate": 4.401419979659119e-06, | |
| "loss": 0.0036, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.9667519181585678, | |
| "grad_norm": 1.7272328840467275, | |
| "learning_rate": 4.224019964407902e-06, | |
| "loss": 0.0028, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 3.0179028132992327, | |
| "grad_norm": 0.19616140979782443, | |
| "learning_rate": 4.047615594495299e-06, | |
| "loss": 0.0017, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 3.0690537084398977, | |
| "grad_norm": 0.060475506150692124, | |
| "learning_rate": 3.872433211039067e-06, | |
| "loss": 0.0013, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 3.1202046035805626, | |
| "grad_norm": 0.060879802064906625, | |
| "learning_rate": 3.698697587249208e-06, | |
| "loss": 0.0012, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 3.1713554987212276, | |
| "grad_norm": 0.8953617552425924, | |
| "learning_rate": 3.526631640025697e-06, | |
| "loss": 0.0009, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 3.2225063938618925, | |
| "grad_norm": 0.400415376801392, | |
| "learning_rate": 3.3564561439380206e-06, | |
| "loss": 0.0006, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 3.2736572890025575, | |
| "grad_norm": 0.13846989650131333, | |
| "learning_rate": 3.1883894479534882e-06, | |
| "loss": 0.0005, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 3.3248081841432224, | |
| "grad_norm": 0.03190740929655757, | |
| "learning_rate": 3.022647195277799e-06, | |
| "loss": 0.0009, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 3.3759590792838874, | |
| "grad_norm": 0.01044371613769382, | |
| "learning_rate": 2.8594420466673133e-06, | |
| "loss": 0.0001, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 3.4271099744245523, | |
| "grad_norm": 0.021747528690115973, | |
| "learning_rate": 2.6989834075680452e-06, | |
| "loss": 0.0003, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 3.4782608695652173, | |
| "grad_norm": 0.003734147207667446, | |
| "learning_rate": 2.541477159431489e-06, | |
| "loss": 0.0002, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 3.5294117647058822, | |
| "grad_norm": 0.02578812155709838, | |
| "learning_rate": 2.3871253955520163e-06, | |
| "loss": 0.0003, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 3.580562659846547, | |
| "grad_norm": 0.07582199264749429, | |
| "learning_rate": 2.2361261617647873e-06, | |
| "loss": 0.0001, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 3.631713554987212, | |
| "grad_norm": 0.003064177461855074, | |
| "learning_rate": 2.0886732023368637e-06, | |
| "loss": 0.0001, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 3.682864450127877, | |
| "grad_norm": 0.0034307231208645776, | |
| "learning_rate": 1.9449557113776157e-06, | |
| "loss": 0.0, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 3.734015345268542, | |
| "grad_norm": 0.003406179275735569, | |
| "learning_rate": 1.8051580900873112e-06, | |
| "loss": 0.0003, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 3.785166240409207, | |
| "grad_norm": 0.0017568341440027282, | |
| "learning_rate": 1.6694597101554354e-06, | |
| "loss": 0.0001, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 3.836317135549872, | |
| "grad_norm": 0.001639728689373477, | |
| "learning_rate": 1.5380346836122429e-06, | |
| "loss": 0.0001, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 3.887468030690537, | |
| "grad_norm": 0.001901581956312772, | |
| "learning_rate": 1.4110516394289103e-06, | |
| "loss": 0.0001, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 3.938618925831202, | |
| "grad_norm": 0.001484483455691581, | |
| "learning_rate": 1.2886735071528829e-06, | |
| "loss": 0.0001, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 3.9897698209718673, | |
| "grad_norm": 0.0018431338087995083, | |
| "learning_rate": 1.1710573078560406e-06, | |
| "loss": 0.0001, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 4.040920716112532, | |
| "grad_norm": 0.06550643685551072, | |
| "learning_rate": 1.058353952663937e-06, | |
| "loss": 0.0002, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 4.092071611253197, | |
| "grad_norm": 0.06509818369816833, | |
| "learning_rate": 9.50708049124568e-07, | |
| "loss": 0.0001, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 4.143222506393862, | |
| "grad_norm": 0.001965866467798876, | |
| "learning_rate": 8.482577156651683e-07, | |
| "loss": 0.0001, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 4.194373401534527, | |
| "grad_norm": 0.003659181883344629, | |
| "learning_rate": 7.511344043750546e-07, | |
| "loss": 0.0001, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 4.245524296675192, | |
| "grad_norm": 0.0008421352975079428, | |
| "learning_rate": 6.594627323419383e-07, | |
| "loss": 0.0001, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 4.296675191815857, | |
| "grad_norm": 0.0015741296233894467, | |
| "learning_rate": 5.733603217580885e-07, | |
| "loss": 0.0001, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 4.3478260869565215, | |
| "grad_norm": 0.07134647602151516, | |
| "learning_rate": 4.929376490015148e-07, | |
| "loss": 0.0001, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 4.398976982097187, | |
| "grad_norm": 0.0009193314183040581, | |
| "learning_rate": 4.1829790288581694e-07, | |
| "loss": 0.0, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 4.450127877237851, | |
| "grad_norm": 0.0011078081696583465, | |
| "learning_rate": 3.495368522605602e-07, | |
| "loss": 0.0, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 4.501278772378517, | |
| "grad_norm": 0.002913655514720279, | |
| "learning_rate": 2.867427231320774e-07, | |
| "loss": 0.0001, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 4.552429667519181, | |
| "grad_norm": 0.0011670444227254474, | |
| "learning_rate": 2.2999608546234063e-07, | |
| "loss": 0.0001, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 4.603580562659847, | |
| "grad_norm": 0.0010703222073159705, | |
| "learning_rate": 1.7936974979116505e-07, | |
| "loss": 0.0001, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 4.654731457800511, | |
| "grad_norm": 0.0010341753002907296, | |
| "learning_rate": 1.349286738143829e-07, | |
| "loss": 0.0001, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 4.705882352941177, | |
| "grad_norm": 0.12688243780881078, | |
| "learning_rate": 9.672987903784237e-08, | |
| "loss": 0.0002, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 4.757033248081841, | |
| "grad_norm": 0.0013506462774850975, | |
| "learning_rate": 6.482237761419652e-08, | |
| "loss": 0.0001, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 4.8081841432225065, | |
| "grad_norm": 0.009021779442771162, | |
| "learning_rate": 3.9247109456332925e-08, | |
| "loss": 0.0002, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 4.859335038363171, | |
| "grad_norm": 0.0010536719662307653, | |
| "learning_rate": 2.003688970814377e-08, | |
| "loss": 0.0001, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 4.910485933503836, | |
| "grad_norm": 0.0010263273253712944, | |
| "learning_rate": 7.216366640032668e-09, | |
| "loss": 0.0001, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 4.961636828644501, | |
| "grad_norm": 0.060715491871115294, | |
| "learning_rate": 8.019900231881483e-10, | |
| "loss": 0.0001, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 4.987212276214834, | |
| "step": 975, | |
| "total_flos": 261426857902080.0, | |
| "train_loss": 0.12791604029011908, | |
| "train_runtime": 13376.8379, | |
| "train_samples_per_second": 18.689, | |
| "train_steps_per_second": 0.073 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 975, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": true | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 261426857902080.0, | |
| "train_batch_size": 16, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |