| { | |
| "best_metric": null, | |
| "best_model_checkpoint": null, | |
| "epoch": 0.9992414664981036, | |
| "eval_steps": 500, | |
| "global_step": 988, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.010113780025284451, | |
| "grad_norm": 0.25880733132362366, | |
| "learning_rate": 5.050505050505051e-06, | |
| "loss": 2.1545, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.020227560050568902, | |
| "grad_norm": 0.22247332334518433, | |
| "learning_rate": 1.0101010101010101e-05, | |
| "loss": 2.0716, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03034134007585335, | |
| "grad_norm": 0.18701794743537903, | |
| "learning_rate": 1.5151515151515153e-05, | |
| "loss": 2.0514, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.040455120101137804, | |
| "grad_norm": 0.2414943277835846, | |
| "learning_rate": 2.0202020202020203e-05, | |
| "loss": 2.0152, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05056890012642225, | |
| "grad_norm": 0.35905972123146057, | |
| "learning_rate": 2.5252525252525256e-05, | |
| "loss": 1.9956, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.0606826801517067, | |
| "grad_norm": 0.25367870926856995, | |
| "learning_rate": 3.0303030303030306e-05, | |
| "loss": 1.9278, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07079646017699115, | |
| "grad_norm": 0.3600216805934906, | |
| "learning_rate": 3.535353535353535e-05, | |
| "loss": 1.9378, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08091024020227561, | |
| "grad_norm": 0.2232195883989334, | |
| "learning_rate": 4.0404040404040405e-05, | |
| "loss": 1.9108, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09102402022756005, | |
| "grad_norm": 0.21375969052314758, | |
| "learning_rate": 4.545454545454546e-05, | |
| "loss": 1.9112, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1011378002528445, | |
| "grad_norm": 0.2019467055797577, | |
| "learning_rate": 4.999984389896431e-05, | |
| "loss": 1.9012, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11125158027812895, | |
| "grad_norm": 0.23392678797245026, | |
| "learning_rate": 4.998111413334416e-05, | |
| "loss": 1.9264, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.1213653603034134, | |
| "grad_norm": 0.25306347012519836, | |
| "learning_rate": 4.993119095936937e-05, | |
| "loss": 1.8857, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13147914032869784, | |
| "grad_norm": 0.20288558304309845, | |
| "learning_rate": 4.985013671509034e-05, | |
| "loss": 1.9397, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.1415929203539823, | |
| "grad_norm": 0.23422028124332428, | |
| "learning_rate": 4.9738052611290836e-05, | |
| "loss": 1.8809, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.15170670037926676, | |
| "grad_norm": 0.21707236766815186, | |
| "learning_rate": 4.959507860510813e-05, | |
| "loss": 1.8604, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16182048040455121, | |
| "grad_norm": 0.2721821367740631, | |
| "learning_rate": 4.9421393225271375e-05, | |
| "loss": 1.8316, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17193426042983564, | |
| "grad_norm": 0.2519826591014862, | |
| "learning_rate": 4.921721334917639e-05, | |
| "loss": 1.8368, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.1820480404551201, | |
| "grad_norm": 0.296394020318985, | |
| "learning_rate": 4.8982793932075356e-05, | |
| "loss": 1.851, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19216182048040456, | |
| "grad_norm": 0.36436939239501953, | |
| "learning_rate": 4.871842768871928e-05, | |
| "loss": 1.8381, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.202275600505689, | |
| "grad_norm": 0.3738105595111847, | |
| "learning_rate": 4.8424444727851126e-05, | |
| "loss": 1.8526, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21238938053097345, | |
| "grad_norm": 0.2505233883857727, | |
| "learning_rate": 4.81012121400057e-05, | |
| "loss": 1.8241, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.2225031605562579, | |
| "grad_norm": 0.22470301389694214, | |
| "learning_rate": 4.774913353913124e-05, | |
| "loss": 1.8848, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.23261694058154236, | |
| "grad_norm": 0.22071461379528046, | |
| "learning_rate": 4.7368648558604836e-05, | |
| "loss": 1.8596, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.2427307206068268, | |
| "grad_norm": 0.2559085190296173, | |
| "learning_rate": 4.69602323022712e-05, | |
| "loss": 1.8337, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.2528445006321112, | |
| "grad_norm": 0.42372068762779236, | |
| "learning_rate": 4.6524394751190215e-05, | |
| "loss": 1.8369, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.2629582806573957, | |
| "grad_norm": 0.2989506423473358, | |
| "learning_rate": 4.606168012683394e-05, | |
| "loss": 1.8556, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.27307206068268014, | |
| "grad_norm": 0.30372118949890137, | |
| "learning_rate": 4.5572666211528324e-05, | |
| "loss": 1.8343, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.2831858407079646, | |
| "grad_norm": 0.23717807233333588, | |
| "learning_rate": 4.5057963626988255e-05, | |
| "loss": 1.7969, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.29329962073324906, | |
| "grad_norm": 0.3980520963668823, | |
| "learning_rate": 4.451821507184656e-05, | |
| "loss": 1.8078, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3034134007585335, | |
| "grad_norm": 0.2746827006340027, | |
| "learning_rate": 4.395409451912942e-05, | |
| "loss": 1.8017, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.31352718078381797, | |
| "grad_norm": 0.24784184992313385, | |
| "learning_rate": 4.336630637467991e-05, | |
| "loss": 1.8024, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.32364096080910243, | |
| "grad_norm": 0.23716619610786438, | |
| "learning_rate": 4.275558459758079e-05, | |
| "loss": 1.828, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.33375474083438683, | |
| "grad_norm": 0.32575511932373047, | |
| "learning_rate": 4.2122691783674786e-05, | |
| "loss": 1.8078, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.3438685208596713, | |
| "grad_norm": 0.2531313896179199, | |
| "learning_rate": 4.14684182133266e-05, | |
| "loss": 1.7991, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.35398230088495575, | |
| "grad_norm": 0.2725628614425659, | |
| "learning_rate": 4.079358086461605e-05, | |
| "loss": 1.812, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.3640960809102402, | |
| "grad_norm": 0.2779051661491394, | |
| "learning_rate": 4.009902239319405e-05, | |
| "loss": 1.7885, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.37420986093552466, | |
| "grad_norm": 0.4107356071472168, | |
| "learning_rate": 3.938561008007578e-05, | |
| "loss": 1.8163, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.3843236409608091, | |
| "grad_norm": 0.31746843457221985, | |
| "learning_rate": 3.8654234748684446e-05, | |
| "loss": 1.8151, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.3944374209860936, | |
| "grad_norm": 0.31331196427345276, | |
| "learning_rate": 3.7905809652498316e-05, | |
| "loss": 1.7881, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.404551201011378, | |
| "grad_norm": 0.2541310787200928, | |
| "learning_rate": 3.714126933468959e-05, | |
| "loss": 1.8035, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.41466498103666244, | |
| "grad_norm": 0.3339349925518036, | |
| "learning_rate": 3.6361568461179516e-05, | |
| "loss": 1.8021, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.4247787610619469, | |
| "grad_norm": 0.30160820484161377, | |
| "learning_rate": 3.5567680628566366e-05, | |
| "loss": 1.7554, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.43489254108723135, | |
| "grad_norm": 0.2795872986316681, | |
| "learning_rate": 3.476059714841529e-05, | |
| "loss": 1.7935, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.4450063211125158, | |
| "grad_norm": 0.2831222712993622, | |
| "learning_rate": 3.3941325809427715e-05, | |
| "loss": 1.7723, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.45512010113780027, | |
| "grad_norm": 0.31489044427871704, | |
| "learning_rate": 3.311088961903613e-05, | |
| "loss": 1.7395, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.46523388116308473, | |
| "grad_norm": 0.26574623584747314, | |
| "learning_rate": 3.227032552599555e-05, | |
| "loss": 1.78, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.47534766118836913, | |
| "grad_norm": 0.24156448245048523, | |
| "learning_rate": 3.14206831255667e-05, | |
| "loss": 1.7964, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.4854614412136536, | |
| "grad_norm": 0.32304009795188904, | |
| "learning_rate": 3.056302334890786e-05, | |
| "loss": 1.7529, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.49557522123893805, | |
| "grad_norm": 0.27490848302841187, | |
| "learning_rate": 2.9698417138311658e-05, | |
| "loss": 1.7542, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.5056890012642224, | |
| "grad_norm": 0.2414240837097168, | |
| "learning_rate": 2.88279441099413e-05, | |
| "loss": 1.7265, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.515802781289507, | |
| "grad_norm": 0.24709856510162354, | |
| "learning_rate": 2.7952691205735832e-05, | |
| "loss": 1.7551, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.5259165613147914, | |
| "grad_norm": 0.2721826732158661, | |
| "learning_rate": 2.7073751336167886e-05, | |
| "loss": 1.7354, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.5360303413400759, | |
| "grad_norm": 0.26764270663261414, | |
| "learning_rate": 2.6192222015548645e-05, | |
| "loss": 1.7715, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.5461441213653603, | |
| "grad_norm": 0.25439366698265076, | |
| "learning_rate": 2.5309203991584073e-05, | |
| "loss": 1.7809, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.5562579013906448, | |
| "grad_norm": 0.30701911449432373, | |
| "learning_rate": 2.4425799870893682e-05, | |
| "loss": 1.7644, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.5663716814159292, | |
| "grad_norm": 0.24427258968353271, | |
| "learning_rate": 2.3543112742208078e-05, | |
| "loss": 1.7809, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.5764854614412137, | |
| "grad_norm": 0.31044137477874756, | |
| "learning_rate": 2.266224479896458e-05, | |
| "loss": 1.756, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.5865992414664981, | |
| "grad_norm": 0.27852073311805725, | |
| "learning_rate": 2.17842959630207e-05, | |
| "loss": 1.7666, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.5967130214917825, | |
| "grad_norm": 0.2860328257083893, | |
| "learning_rate": 2.091036251120411e-05, | |
| "loss": 1.7541, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.606826801517067, | |
| "grad_norm": 0.2761785089969635, | |
| "learning_rate": 2.0041535706414138e-05, | |
| "loss": 1.7656, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6169405815423514, | |
| "grad_norm": 0.25353536009788513, | |
| "learning_rate": 1.917890043498397e-05, | |
| "loss": 1.7586, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.6270543615676359, | |
| "grad_norm": 0.2852698564529419, | |
| "learning_rate": 1.832353385200522e-05, | |
| "loss": 1.7354, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.6371681415929203, | |
| "grad_norm": 0.27082163095474243, | |
| "learning_rate": 1.747650403630629e-05, | |
| "loss": 1.762, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.6472819216182049, | |
| "grad_norm": 0.24518464505672455, | |
| "learning_rate": 1.66388686567641e-05, | |
| "loss": 1.7514, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.6573957016434893, | |
| "grad_norm": 0.3101823031902313, | |
| "learning_rate": 1.5811673651614514e-05, | |
| "loss": 1.7323, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.6675094816687737, | |
| "grad_norm": 0.27251914143562317, | |
| "learning_rate": 1.4995951922410551e-05, | |
| "loss": 1.7726, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.6776232616940582, | |
| "grad_norm": 0.2582049071788788, | |
| "learning_rate": 1.4192722044259275e-05, | |
| "loss": 1.7459, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.6877370417193426, | |
| "grad_norm": 0.2840460240840912, | |
| "learning_rate": 1.340298699394777e-05, | |
| "loss": 1.729, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.6978508217446271, | |
| "grad_norm": 0.2834312915802002, | |
| "learning_rate": 1.2627732897546535e-05, | |
| "loss": 1.7216, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.7079646017699115, | |
| "grad_norm": 0.27712661027908325, | |
| "learning_rate": 1.186792779905386e-05, | |
| "loss": 1.7572, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.718078381795196, | |
| "grad_norm": 0.27337151765823364, | |
| "learning_rate": 1.1124520451619048e-05, | |
| "loss": 1.741, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.7281921618204804, | |
| "grad_norm": 0.2533912658691406, | |
| "learning_rate": 1.0398439132853696e-05, | |
| "loss": 1.771, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.7383059418457648, | |
| "grad_norm": 0.24308809638023376, | |
| "learning_rate": 9.690590485710311e-06, | |
| "loss": 1.7094, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.7484197218710493, | |
| "grad_norm": 0.2828107476234436, | |
| "learning_rate": 9.001858386375744e-06, | |
| "loss": 1.7145, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.7585335018963337, | |
| "grad_norm": 0.26573511958122253, | |
| "learning_rate": 8.333102840593015e-06, | |
| "loss": 1.7757, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.7686472819216182, | |
| "grad_norm": 0.24994178116321564, | |
| "learning_rate": 7.68515890978963e-06, | |
| "loss": 1.7445, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.7787610619469026, | |
| "grad_norm": 0.2822398841381073, | |
| "learning_rate": 7.058835668353409e-06, | |
| "loss": 1.7122, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.7888748419721872, | |
| "grad_norm": 0.3666843771934509, | |
| "learning_rate": 6.454915193357772e-06, | |
| "loss": 1.7412, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.7989886219974716, | |
| "grad_norm": 0.27069199085235596, | |
| "learning_rate": 5.874151587998023e-06, | |
| "loss": 1.7443, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.809102402022756, | |
| "grad_norm": 0.2951764464378357, | |
| "learning_rate": 5.317270039958058e-06, | |
| "loss": 1.7275, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8192161820480405, | |
| "grad_norm": 0.27236202359199524, | |
| "learning_rate": 4.784965915883274e-06, | |
| "loss": 1.7312, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.8293299620733249, | |
| "grad_norm": 0.2640001177787781, | |
| "learning_rate": 4.277903893090407e-06, | |
| "loss": 1.7096, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.8394437420986094, | |
| "grad_norm": 0.30069518089294434, | |
| "learning_rate": 3.7967171295984925e-06, | |
| "loss": 1.7635, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.8495575221238938, | |
| "grad_norm": 0.2520219683647156, | |
| "learning_rate": 3.342006473517362e-06, | |
| "loss": 1.7529, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.8596713021491783, | |
| "grad_norm": 0.26766538619995117, | |
| "learning_rate": 2.9143397127808393e-06, | |
| "loss": 1.7302, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.8697850821744627, | |
| "grad_norm": 0.2785772383213043, | |
| "learning_rate": 2.5142508661615077e-06, | |
| "loss": 1.7374, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.8798988621997471, | |
| "grad_norm": 0.3111814558506012, | |
| "learning_rate": 2.1422395164523573e-06, | |
| "loss": 1.7211, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.8900126422250316, | |
| "grad_norm": 0.2579687535762787, | |
| "learning_rate": 1.7987701866479329e-06, | |
| "loss": 1.7113, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.900126422250316, | |
| "grad_norm": 0.38696742057800293, | |
| "learning_rate": 1.4842717599039047e-06, | |
| "loss": 1.736, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.9102402022756005, | |
| "grad_norm": 0.2710600197315216, | |
| "learning_rate": 1.1991369439994349e-06, | |
| "loss": 1.7054, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9203539823008849, | |
| "grad_norm": 0.27956292033195496, | |
| "learning_rate": 9.437217809709714e-07, | |
| "loss": 1.7388, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.9304677623261695, | |
| "grad_norm": 0.2631407082080841, | |
| "learning_rate": 7.183452025297937e-07, | |
| "loss": 1.7099, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.9405815423514539, | |
| "grad_norm": 0.2510451376438141, | |
| "learning_rate": 5.232886318184971e-07, | |
| "loss": 1.7122, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.9506953223767383, | |
| "grad_norm": 0.2797139883041382, | |
| "learning_rate": 3.587956320036362e-07, | |
| "loss": 1.7152, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.9608091024020228, | |
| "grad_norm": 0.2599884271621704, | |
| "learning_rate": 2.250716021433702e-07, | |
| "loss": 1.7371, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.9709228824273072, | |
| "grad_norm": 0.31965357065200806, | |
| "learning_rate": 1.2228352070983719e-07, | |
| "loss": 1.7437, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.9810366624525917, | |
| "grad_norm": 0.33041146397590637, | |
| "learning_rate": 5.05597370865335e-08, | |
| "loss": 1.7488, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.9911504424778761, | |
| "grad_norm": 0.2697661519050598, | |
| "learning_rate": 9.98981130106158e-09, | |
| "loss": 1.718, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.9992414664981036, | |
| "step": 988, | |
| "total_flos": 1.4618955817345352e+18, | |
| "train_loss": 1.7979252251536257, | |
| "train_runtime": 11520.2382, | |
| "train_samples_per_second": 2.746, | |
| "train_steps_per_second": 0.086 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 988, | |
| "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": 1.4618955817345352e+18, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |