Invalid JSON:Unexpected token 'I', ..."ad_norm": Infinity,
"... is not valid JSON
| { | |
| "best_global_step": null, | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.034680677660441486, | |
| "eval_steps": 500, | |
| "global_step": 1000, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.00034680677660441485, | |
| "grad_norm": 8.286304473876953, | |
| "learning_rate": 2.6999999999999996e-06, | |
| "loss": 36.023321533203124, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.0006936135532088297, | |
| "grad_norm": 5.303624153137207, | |
| "learning_rate": 5.7e-06, | |
| "loss": 35.81325988769531, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.0010404203298132447, | |
| "grad_norm": 14.0803861618042, | |
| "learning_rate": 8.7e-06, | |
| "loss": 35.61415100097656, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.0013872271064176594, | |
| "grad_norm": 42.10546875, | |
| "learning_rate": 1.17e-05, | |
| "loss": 35.36910400390625, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.0017340338830220744, | |
| "grad_norm": 43.37276840209961, | |
| "learning_rate": 1.47e-05, | |
| "loss": 34.86077880859375, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0020808406596264893, | |
| "grad_norm": Infinity, | |
| "learning_rate": 1.7699999999999997e-05, | |
| "loss": 34.67452697753906, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.002427647436230904, | |
| "grad_norm": 86.09446716308594, | |
| "learning_rate": 2.04e-05, | |
| "loss": 34.29078979492188, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.002774454212835319, | |
| "grad_norm": 43.50123977661133, | |
| "learning_rate": 2.34e-05, | |
| "loss": 33.816766357421876, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.0031212609894397335, | |
| "grad_norm": 104.7258529663086, | |
| "learning_rate": 2.6399999999999995e-05, | |
| "loss": 33.36159057617188, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.0034680677660441487, | |
| "grad_norm": 66.46515655517578, | |
| "learning_rate": 2.91e-05, | |
| "loss": 33.181732177734375, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.0038148745426485635, | |
| "grad_norm": 127.66901397705078, | |
| "learning_rate": 3.2099999999999994e-05, | |
| "loss": 32.83320007324219, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.004161681319252979, | |
| "grad_norm": 71.67156982421875, | |
| "learning_rate": 3.51e-05, | |
| "loss": 32.48929138183594, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.004508488095857393, | |
| "grad_norm": 76.15655517578125, | |
| "learning_rate": 3.81e-05, | |
| "loss": 32.10602722167969, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.004855294872461808, | |
| "grad_norm": 38.65349578857422, | |
| "learning_rate": 4.11e-05, | |
| "loss": 31.449526977539062, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.005202101649066222, | |
| "grad_norm": 59.55622863769531, | |
| "learning_rate": 4.4099999999999995e-05, | |
| "loss": 31.01339111328125, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.005548908425670638, | |
| "grad_norm": 130.26136779785156, | |
| "learning_rate": 4.709999999999999e-05, | |
| "loss": 30.92586669921875, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.005895715202275053, | |
| "grad_norm": 77.2778549194336, | |
| "learning_rate": 5.01e-05, | |
| "loss": 30.492510986328124, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.006242521978879467, | |
| "grad_norm": 29.711563110351562, | |
| "learning_rate": 5.309999999999999e-05, | |
| "loss": 29.836746215820312, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.006589328755483882, | |
| "grad_norm": 29.891956329345703, | |
| "learning_rate": 5.6099999999999995e-05, | |
| "loss": 29.291275024414062, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.006936135532088297, | |
| "grad_norm": 30.56305503845215, | |
| "learning_rate": 5.91e-05, | |
| "loss": 29.14805908203125, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.007282942308692712, | |
| "grad_norm": 29.489208221435547, | |
| "learning_rate": 6.209999999999999e-05, | |
| "loss": 28.80553283691406, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.007629749085297127, | |
| "grad_norm": 37.147377014160156, | |
| "learning_rate": 6.51e-05, | |
| "loss": 28.694085693359376, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.007976555861901541, | |
| "grad_norm": 28.496992111206055, | |
| "learning_rate": 6.81e-05, | |
| "loss": 28.166943359375, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.008323362638505957, | |
| "grad_norm": 29.42328453063965, | |
| "learning_rate": 7.11e-05, | |
| "loss": 27.878482055664062, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.008670169415110372, | |
| "grad_norm": 27.867637634277344, | |
| "learning_rate": 7.41e-05, | |
| "loss": 26.99749755859375, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.009016976191714786, | |
| "grad_norm": 29.695758819580078, | |
| "learning_rate": 7.709999999999999e-05, | |
| "loss": 27.208102416992187, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.0093637829683192, | |
| "grad_norm": 34.605934143066406, | |
| "learning_rate": 8.01e-05, | |
| "loss": 26.798681640625, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.009710589744923616, | |
| "grad_norm": 35.05394744873047, | |
| "learning_rate": 8.31e-05, | |
| "loss": 27.192486572265626, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.01005739652152803, | |
| "grad_norm": 19.137067794799805, | |
| "learning_rate": 8.609999999999999e-05, | |
| "loss": 26.89140930175781, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.010404203298132445, | |
| "grad_norm": 24.543006896972656, | |
| "learning_rate": 8.879999999999999e-05, | |
| "loss": 26.167477416992188, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.010751010074736861, | |
| "grad_norm": 18.849651336669922, | |
| "learning_rate": 9.18e-05, | |
| "loss": 26.41388854980469, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.011097816851341275, | |
| "grad_norm": 21.179908752441406, | |
| "learning_rate": 9.479999999999999e-05, | |
| "loss": 26.003546142578124, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.01144462362794569, | |
| "grad_norm": 19.344528198242188, | |
| "learning_rate": 9.779999999999999e-05, | |
| "loss": 26.271051025390626, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.011791430404550106, | |
| "grad_norm": 19.74913215637207, | |
| "learning_rate": 0.0001008, | |
| "loss": 26.064590454101562, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.01213823718115452, | |
| "grad_norm": 17.638103485107422, | |
| "learning_rate": 0.00010379999999999999, | |
| "loss": 25.863119506835936, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.012485043957758934, | |
| "grad_norm": 24.873981475830078, | |
| "learning_rate": 0.00010679999999999998, | |
| "loss": 26.28948059082031, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.01283185073436335, | |
| "grad_norm": 17.048828125, | |
| "learning_rate": 0.00010979999999999999, | |
| "loss": 26.215179443359375, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.013178657510967764, | |
| "grad_norm": 21.703975677490234, | |
| "learning_rate": 0.00011279999999999999, | |
| "loss": 26.152392578125, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.013525464287572179, | |
| "grad_norm": 22.66498565673828, | |
| "learning_rate": 0.0001158, | |
| "loss": 26.0640625, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.013872271064176595, | |
| "grad_norm": 16.945016860961914, | |
| "learning_rate": 0.0001188, | |
| "loss": 26.023715209960937, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.01421907784078101, | |
| "grad_norm": 27.866268157958984, | |
| "learning_rate": 0.00012179999999999999, | |
| "loss": 25.43433380126953, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.014565884617385423, | |
| "grad_norm": 22.481279373168945, | |
| "learning_rate": 0.00012479999999999997, | |
| "loss": 25.464639282226564, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.014912691393989838, | |
| "grad_norm": 14.993969917297363, | |
| "learning_rate": 0.0001278, | |
| "loss": 25.218385314941408, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.015259498170594254, | |
| "grad_norm": 14.426427841186523, | |
| "learning_rate": 0.00013079999999999998, | |
| "loss": 24.94810333251953, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.015606304947198668, | |
| "grad_norm": 15.592818260192871, | |
| "learning_rate": 0.0001338, | |
| "loss": 25.38585662841797, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.015953111723803082, | |
| "grad_norm": 14.709174156188965, | |
| "learning_rate": 0.0001368, | |
| "loss": 24.31098175048828, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.016299918500407497, | |
| "grad_norm": 13.651994705200195, | |
| "learning_rate": 0.00013979999999999998, | |
| "loss": 24.82652893066406, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.016646725277011915, | |
| "grad_norm": 17.521072387695312, | |
| "learning_rate": 0.00014279999999999997, | |
| "loss": 25.121923828125, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.01699353205361633, | |
| "grad_norm": 13.458234786987305, | |
| "learning_rate": 0.0001458, | |
| "loss": 24.447775268554686, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.017340338830220743, | |
| "grad_norm": 13.57589340209961, | |
| "learning_rate": 0.00014879999999999998, | |
| "loss": 24.76665802001953, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.017687145606825157, | |
| "grad_norm": 12.934654235839844, | |
| "learning_rate": 0.00015179999999999998, | |
| "loss": 24.170372009277344, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.018033952383429572, | |
| "grad_norm": 13.884750366210938, | |
| "learning_rate": 0.0001548, | |
| "loss": 24.240016174316406, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.018380759160033986, | |
| "grad_norm": 13.660422325134277, | |
| "learning_rate": 0.0001578, | |
| "loss": 23.897187805175783, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.0187275659366384, | |
| "grad_norm": 11.849883079528809, | |
| "learning_rate": 0.0001608, | |
| "loss": 23.574449157714845, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.019074372713242818, | |
| "grad_norm": 10.595759391784668, | |
| "learning_rate": 0.0001638, | |
| "loss": 23.569731140136717, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.019421179489847232, | |
| "grad_norm": 12.363926887512207, | |
| "learning_rate": 0.0001668, | |
| "loss": 23.617283630371094, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.019767986266451647, | |
| "grad_norm": 10.968243598937988, | |
| "learning_rate": 0.00016979999999999998, | |
| "loss": 23.990586853027345, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.02011479304305606, | |
| "grad_norm": 12.141168594360352, | |
| "learning_rate": 0.00017279999999999997, | |
| "loss": 23.807048034667968, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.020461599819660475, | |
| "grad_norm": 14.928491592407227, | |
| "learning_rate": 0.00017579999999999996, | |
| "loss": 22.980938720703126, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.02080840659626489, | |
| "grad_norm": 11.39072036743164, | |
| "learning_rate": 0.00017879999999999998, | |
| "loss": 23.521900939941407, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.021155213372869307, | |
| "grad_norm": 10.157258033752441, | |
| "learning_rate": 0.00018179999999999997, | |
| "loss": 23.075851440429688, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.021502020149473722, | |
| "grad_norm": 12.673665046691895, | |
| "learning_rate": 0.0001848, | |
| "loss": 23.70465850830078, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.021848826926078136, | |
| "grad_norm": 11.242607116699219, | |
| "learning_rate": 0.00018779999999999998, | |
| "loss": 22.982296752929688, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.02219563370268255, | |
| "grad_norm": 25.496898651123047, | |
| "learning_rate": 0.00019079999999999998, | |
| "loss": 22.516278076171876, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.022542440479286965, | |
| "grad_norm": 11.215885162353516, | |
| "learning_rate": 0.0001938, | |
| "loss": 22.14655456542969, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.02288924725589138, | |
| "grad_norm": 9.149953842163086, | |
| "learning_rate": 0.00019679999999999999, | |
| "loss": 22.114730834960938, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.023236054032495793, | |
| "grad_norm": 14.604310035705566, | |
| "learning_rate": 0.0001995, | |
| "loss": 22.670735168457032, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.02358286080910021, | |
| "grad_norm": 11.181646347045898, | |
| "learning_rate": 0.0002025, | |
| "loss": 22.34198455810547, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.023929667585704625, | |
| "grad_norm": 10.59487533569336, | |
| "learning_rate": 0.0002055, | |
| "loss": 22.268771362304687, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.02427647436230904, | |
| "grad_norm": 11.758952140808105, | |
| "learning_rate": 0.00020849999999999997, | |
| "loss": 21.350621032714844, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.024623281138913454, | |
| "grad_norm": 18.59368896484375, | |
| "learning_rate": 0.00021149999999999996, | |
| "loss": 21.507183837890626, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.02497008791551787, | |
| "grad_norm": 9.710185050964355, | |
| "learning_rate": 0.00021449999999999998, | |
| "loss": 21.502236938476564, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.025316894692122283, | |
| "grad_norm": 20.68129539489746, | |
| "learning_rate": 0.00021749999999999997, | |
| "loss": 21.232615661621093, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.0256637014687267, | |
| "grad_norm": 9.399124145507812, | |
| "learning_rate": 0.00022049999999999997, | |
| "loss": 21.146585083007814, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.026010508245331115, | |
| "grad_norm": 9.145174026489258, | |
| "learning_rate": 0.00022349999999999998, | |
| "loss": 21.06390838623047, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.02635731502193553, | |
| "grad_norm": 11.161269187927246, | |
| "learning_rate": 0.00022649999999999998, | |
| "loss": 20.896176147460938, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.026704121798539943, | |
| "grad_norm": 8.362421035766602, | |
| "learning_rate": 0.0002295, | |
| "loss": 20.884358215332032, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.027050928575144358, | |
| "grad_norm": 9.504199981689453, | |
| "learning_rate": 0.00023249999999999999, | |
| "loss": 20.086947631835937, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.027397735351748772, | |
| "grad_norm": 8.848526954650879, | |
| "learning_rate": 0.00023549999999999998, | |
| "loss": 20.411599731445314, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.02774454212835319, | |
| "grad_norm": 9.612874984741211, | |
| "learning_rate": 0.0002385, | |
| "loss": 20.284182739257812, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.028091348904957604, | |
| "grad_norm": 8.688453674316406, | |
| "learning_rate": 0.0002415, | |
| "loss": 20.024198913574217, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.02843815568156202, | |
| "grad_norm": 9.779630661010742, | |
| "learning_rate": 0.0002445, | |
| "loss": 19.91345977783203, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.028784962458166433, | |
| "grad_norm": 12.37082576751709, | |
| "learning_rate": 0.0002472, | |
| "loss": 19.54875030517578, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.029131769234770847, | |
| "grad_norm": 8.736369132995605, | |
| "learning_rate": 0.00025019999999999996, | |
| "loss": 19.499098205566405, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.02947857601137526, | |
| "grad_norm": 10.752023696899414, | |
| "learning_rate": 0.0002532, | |
| "loss": 19.327198791503907, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.029825382787979676, | |
| "grad_norm": 9.755398750305176, | |
| "learning_rate": 0.0002562, | |
| "loss": 19.291722106933594, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.030172189564584093, | |
| "grad_norm": 8.066308975219727, | |
| "learning_rate": 0.00025919999999999996, | |
| "loss": 18.516694641113283, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.030518996341188508, | |
| "grad_norm": 8.660676002502441, | |
| "learning_rate": 0.0002622, | |
| "loss": 18.70814666748047, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.030865803117792922, | |
| "grad_norm": 8.439098358154297, | |
| "learning_rate": 0.0002652, | |
| "loss": 18.787985229492186, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.031212609894397336, | |
| "grad_norm": 8.525120735168457, | |
| "learning_rate": 0.00026819999999999996, | |
| "loss": 18.356887817382812, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.031559416671001754, | |
| "grad_norm": 9.237743377685547, | |
| "learning_rate": 0.0002712, | |
| "loss": 18.07917022705078, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.031906223447606165, | |
| "grad_norm": 8.137825965881348, | |
| "learning_rate": 0.0002742, | |
| "loss": 18.066172790527343, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.03225303022421058, | |
| "grad_norm": 20.67136001586914, | |
| "learning_rate": 0.0002772, | |
| "loss": 18.19664306640625, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.032599837000814993, | |
| "grad_norm": 8.68689250946045, | |
| "learning_rate": 0.0002802, | |
| "loss": 17.834249877929686, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.03294664377741941, | |
| "grad_norm": 8.076814651489258, | |
| "learning_rate": 0.00028319999999999994, | |
| "loss": 17.645645141601562, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.03329345055402383, | |
| "grad_norm": 8.278287887573242, | |
| "learning_rate": 0.00028619999999999996, | |
| "loss": 17.20880889892578, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.03364025733062824, | |
| "grad_norm": 8.195657730102539, | |
| "learning_rate": 0.0002892, | |
| "loss": 17.02202453613281, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.03398706410723266, | |
| "grad_norm": 35.29956817626953, | |
| "learning_rate": 0.00029219999999999995, | |
| "loss": 17.46351013183594, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.03433387088383707, | |
| "grad_norm": 14.36285400390625, | |
| "learning_rate": 0.00029519999999999997, | |
| "loss": 17.128692626953125, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.034680677660441486, | |
| "grad_norm": 117.96296691894531, | |
| "learning_rate": 0.0002979, | |
| "loss": 16.577227783203124, | |
| "step": 1000 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 100000, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 4, | |
| "save_steps": 100, | |
| "stateful_callbacks": { | |
| "TrainerControl": { | |
| "args": { | |
| "should_epoch_stop": false, | |
| "should_evaluate": false, | |
| "should_log": false, | |
| "should_save": true, | |
| "should_training_stop": false | |
| }, | |
| "attributes": {} | |
| } | |
| }, | |
| "total_flos": 6219913494528000.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |