| { | |
| "best_metric": 0.29854172468185425, | |
| "best_model_checkpoint": "./results/checkpoint-2026", | |
| "epoch": 2.0, | |
| "eval_steps": 500, | |
| "global_step": 2026, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.01, | |
| "grad_norm": 15.60984992980957, | |
| "learning_rate": 4.990128331688055e-05, | |
| "loss": 0.5627, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.02, | |
| "grad_norm": 1.2870399951934814, | |
| "learning_rate": 4.9802566633761114e-05, | |
| "loss": 0.2714, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.03, | |
| "grad_norm": 123.16035461425781, | |
| "learning_rate": 4.970384995064166e-05, | |
| "loss": 0.5078, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 11.994287490844727, | |
| "learning_rate": 4.960513326752221e-05, | |
| "loss": 0.4197, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.05, | |
| "grad_norm": 3.41953182220459, | |
| "learning_rate": 4.950641658440277e-05, | |
| "loss": 0.441, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.06, | |
| "grad_norm": 26.95296287536621, | |
| "learning_rate": 4.940769990128332e-05, | |
| "loss": 0.6958, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 28.05646324157715, | |
| "learning_rate": 4.930898321816387e-05, | |
| "loss": 0.3109, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 28.755550384521484, | |
| "learning_rate": 4.921026653504443e-05, | |
| "loss": 0.5321, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.09, | |
| "grad_norm": 46.09171676635742, | |
| "learning_rate": 4.9111549851924976e-05, | |
| "loss": 0.4216, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 3.234527587890625, | |
| "learning_rate": 4.901283316880553e-05, | |
| "loss": 0.3427, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 36.78240966796875, | |
| "learning_rate": 4.891411648568609e-05, | |
| "loss": 0.5259, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.12, | |
| "grad_norm": 2.4952964782714844, | |
| "learning_rate": 4.8815399802566636e-05, | |
| "loss": 0.525, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.13, | |
| "grad_norm": 6.107447147369385, | |
| "learning_rate": 4.8716683119447184e-05, | |
| "loss": 0.5349, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.14, | |
| "grad_norm": 17.599472045898438, | |
| "learning_rate": 4.861796643632775e-05, | |
| "loss": 0.3194, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 67.53023529052734, | |
| "learning_rate": 4.8519249753208296e-05, | |
| "loss": 0.4738, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 61.95085525512695, | |
| "learning_rate": 4.8420533070088844e-05, | |
| "loss": 0.4151, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.17, | |
| "grad_norm": 377.9793701171875, | |
| "learning_rate": 4.83218163869694e-05, | |
| "loss": 0.3219, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 9.71474838256836, | |
| "learning_rate": 4.8223099703849955e-05, | |
| "loss": 0.2931, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.19, | |
| "grad_norm": 54.442691802978516, | |
| "learning_rate": 4.8124383020730504e-05, | |
| "loss": 0.3802, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.2, | |
| "grad_norm": 111.41837310791016, | |
| "learning_rate": 4.802566633761106e-05, | |
| "loss": 0.4909, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 28.207542419433594, | |
| "learning_rate": 4.792694965449161e-05, | |
| "loss": 0.392, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 65.766357421875, | |
| "learning_rate": 4.7828232971372164e-05, | |
| "loss": 0.3002, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.23, | |
| "grad_norm": 127.14469909667969, | |
| "learning_rate": 4.772951628825272e-05, | |
| "loss": 0.3654, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 0.19254250824451447, | |
| "learning_rate": 4.763079960513327e-05, | |
| "loss": 0.2953, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.25, | |
| "grad_norm": 30.106840133666992, | |
| "learning_rate": 4.753208292201382e-05, | |
| "loss": 0.4019, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 4.884279727935791, | |
| "learning_rate": 4.743336623889438e-05, | |
| "loss": 0.4152, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.27, | |
| "grad_norm": 71.0513916015625, | |
| "learning_rate": 4.733464955577493e-05, | |
| "loss": 0.6281, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 0.28, | |
| "grad_norm": 1.381753921508789, | |
| "learning_rate": 4.723593287265548e-05, | |
| "loss": 0.3224, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 1.0361205339431763, | |
| "learning_rate": 4.713721618953603e-05, | |
| "loss": 0.6375, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 0.3, | |
| "grad_norm": 1.2360197305679321, | |
| "learning_rate": 4.703849950641659e-05, | |
| "loss": 0.419, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 188.26495361328125, | |
| "learning_rate": 4.693978282329714e-05, | |
| "loss": 0.5303, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 0.14256739616394043, | |
| "learning_rate": 4.684106614017769e-05, | |
| "loss": 0.2615, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 63.93450927734375, | |
| "learning_rate": 4.674234945705824e-05, | |
| "loss": 0.4678, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 0.34, | |
| "grad_norm": 31.07522201538086, | |
| "learning_rate": 4.66436327739388e-05, | |
| "loss": 0.607, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 0.35, | |
| "grad_norm": 12.982345581054688, | |
| "learning_rate": 4.654491609081935e-05, | |
| "loss": 0.189, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 0.36, | |
| "grad_norm": 14.37088394165039, | |
| "learning_rate": 4.64461994076999e-05, | |
| "loss": 0.4341, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 8.185881614685059, | |
| "learning_rate": 4.634748272458046e-05, | |
| "loss": 0.5132, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 1.082980751991272, | |
| "learning_rate": 4.624876604146101e-05, | |
| "loss": 0.3491, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 0.38, | |
| "grad_norm": 12.5576753616333, | |
| "learning_rate": 4.615004935834156e-05, | |
| "loss": 0.7587, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 5.933102130889893, | |
| "learning_rate": 4.605133267522212e-05, | |
| "loss": 0.579, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 1.3454967737197876, | |
| "learning_rate": 4.5952615992102666e-05, | |
| "loss": 0.1932, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 0.41, | |
| "grad_norm": 17.171228408813477, | |
| "learning_rate": 4.585389930898322e-05, | |
| "loss": 0.3217, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 3.5590412616729736, | |
| "learning_rate": 4.575518262586377e-05, | |
| "loss": 0.2279, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 0.43, | |
| "grad_norm": 17.661069869995117, | |
| "learning_rate": 4.5656465942744326e-05, | |
| "loss": 0.2676, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 48.93571853637695, | |
| "learning_rate": 4.5557749259624875e-05, | |
| "loss": 0.5449, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 0.45, | |
| "grad_norm": 12.7286376953125, | |
| "learning_rate": 4.545903257650543e-05, | |
| "loss": 0.5127, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 0.46, | |
| "grad_norm": 51.88860321044922, | |
| "learning_rate": 4.5360315893385986e-05, | |
| "loss": 0.4794, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 18.063552856445312, | |
| "learning_rate": 4.5261599210266535e-05, | |
| "loss": 0.3728, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 2.861877918243408, | |
| "learning_rate": 4.516288252714709e-05, | |
| "loss": 0.3038, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 0.49, | |
| "grad_norm": 6.476074695587158, | |
| "learning_rate": 4.5064165844027646e-05, | |
| "loss": 0.2592, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 0.5, | |
| "grad_norm": 62.48997497558594, | |
| "learning_rate": 4.4965449160908195e-05, | |
| "loss": 0.4779, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 0.5959272384643555, | |
| "learning_rate": 4.486673247778875e-05, | |
| "loss": 0.3865, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 32.949684143066406, | |
| "learning_rate": 4.47680157946693e-05, | |
| "loss": 0.5077, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 0.53, | |
| "grad_norm": 8.09738826751709, | |
| "learning_rate": 4.4669299111549855e-05, | |
| "loss": 0.3352, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 0.54, | |
| "grad_norm": 23.277297973632812, | |
| "learning_rate": 4.457058242843041e-05, | |
| "loss": 0.5204, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 89.32869720458984, | |
| "learning_rate": 4.447186574531096e-05, | |
| "loss": 0.3888, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 0.56, | |
| "grad_norm": 2.6795363426208496, | |
| "learning_rate": 4.437314906219151e-05, | |
| "loss": 0.5252, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 0.57, | |
| "grad_norm": 37.583744049072266, | |
| "learning_rate": 4.427443237907207e-05, | |
| "loss": 0.3881, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 0.58, | |
| "grad_norm": 1.256844162940979, | |
| "learning_rate": 4.417571569595262e-05, | |
| "loss": 0.1872, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 4.5737786293029785, | |
| "learning_rate": 4.407699901283317e-05, | |
| "loss": 0.2536, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 0.6, | |
| "grad_norm": 45.64347839355469, | |
| "learning_rate": 4.3978282329713724e-05, | |
| "loss": 0.3777, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 0.61, | |
| "grad_norm": 0.4227633774280548, | |
| "learning_rate": 4.387956564659428e-05, | |
| "loss": 0.2028, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 4.602664947509766, | |
| "learning_rate": 4.378084896347483e-05, | |
| "loss": 0.5563, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 0.7803702354431152, | |
| "learning_rate": 4.3682132280355384e-05, | |
| "loss": 0.3636, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 0.64, | |
| "grad_norm": 70.02734375, | |
| "learning_rate": 4.358341559723593e-05, | |
| "loss": 0.4558, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 0.65, | |
| "grad_norm": 39.45964050292969, | |
| "learning_rate": 4.348469891411649e-05, | |
| "loss": 0.4592, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 22.5675106048584, | |
| "learning_rate": 4.3385982230997044e-05, | |
| "loss": 0.3082, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 0.67, | |
| "grad_norm": 4.789850234985352, | |
| "learning_rate": 4.328726554787759e-05, | |
| "loss": 0.2404, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 0.68, | |
| "grad_norm": 4.671356678009033, | |
| "learning_rate": 4.318854886475814e-05, | |
| "loss": 0.2864, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 1.803113341331482, | |
| "learning_rate": 4.3089832181638704e-05, | |
| "loss": 0.2627, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 0.7, | |
| "grad_norm": 0.38143932819366455, | |
| "learning_rate": 4.299111549851925e-05, | |
| "loss": 0.1678, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 0.396694540977478, | |
| "learning_rate": 4.28923988153998e-05, | |
| "loss": 0.181, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 0.72, | |
| "grad_norm": 26.724634170532227, | |
| "learning_rate": 4.279368213228036e-05, | |
| "loss": 0.5595, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 179.3428497314453, | |
| "learning_rate": 4.269496544916091e-05, | |
| "loss": 0.3613, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 0.74, | |
| "grad_norm": 4.721936225891113, | |
| "learning_rate": 4.259624876604146e-05, | |
| "loss": 0.4182, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 0.75, | |
| "grad_norm": 1.8950241804122925, | |
| "learning_rate": 4.249753208292202e-05, | |
| "loss": 0.3623, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 0.76, | |
| "grad_norm": 5.388864994049072, | |
| "learning_rate": 4.2398815399802566e-05, | |
| "loss": 0.4246, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 0.41123124957084656, | |
| "learning_rate": 4.230009871668312e-05, | |
| "loss": 0.2425, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 0.78, | |
| "grad_norm": 0.3556106388568878, | |
| "learning_rate": 4.220138203356368e-05, | |
| "loss": 0.3751, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 0.899945080280304, | |
| "learning_rate": 4.2102665350444226e-05, | |
| "loss": 0.3994, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 4.583869934082031, | |
| "learning_rate": 4.2003948667324774e-05, | |
| "loss": 0.3681, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 0.81, | |
| "grad_norm": 0.3905455768108368, | |
| "learning_rate": 4.190523198420534e-05, | |
| "loss": 0.1491, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 0.82, | |
| "grad_norm": 36.8359260559082, | |
| "learning_rate": 4.1806515301085886e-05, | |
| "loss": 0.2609, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 34.53616714477539, | |
| "learning_rate": 4.1707798617966434e-05, | |
| "loss": 0.5495, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 14.104715347290039, | |
| "learning_rate": 4.160908193484699e-05, | |
| "loss": 0.33, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 0.85, | |
| "grad_norm": 30.295068740844727, | |
| "learning_rate": 4.1510365251727546e-05, | |
| "loss": 1.0008, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 0.86, | |
| "grad_norm": 93.3653793334961, | |
| "learning_rate": 4.1411648568608094e-05, | |
| "loss": 1.0401, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 114.31365966796875, | |
| "learning_rate": 4.131293188548865e-05, | |
| "loss": 0.4156, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 134.54774475097656, | |
| "learning_rate": 4.12142152023692e-05, | |
| "loss": 0.5463, | |
| "step": 890 | |
| }, | |
| { | |
| "epoch": 0.89, | |
| "grad_norm": 3.021076202392578, | |
| "learning_rate": 4.1115498519249754e-05, | |
| "loss": 0.2947, | |
| "step": 900 | |
| }, | |
| { | |
| "epoch": 0.9, | |
| "grad_norm": 9.884215354919434, | |
| "learning_rate": 4.101678183613031e-05, | |
| "loss": 0.3674, | |
| "step": 910 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 167.9898223876953, | |
| "learning_rate": 4.091806515301086e-05, | |
| "loss": 0.4516, | |
| "step": 920 | |
| }, | |
| { | |
| "epoch": 0.92, | |
| "grad_norm": 34.41691207885742, | |
| "learning_rate": 4.0819348469891414e-05, | |
| "loss": 0.504, | |
| "step": 930 | |
| }, | |
| { | |
| "epoch": 0.93, | |
| "grad_norm": 10.135024070739746, | |
| "learning_rate": 4.072063178677197e-05, | |
| "loss": 0.2834, | |
| "step": 940 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 1.0688509941101074, | |
| "learning_rate": 4.062191510365252e-05, | |
| "loss": 0.3188, | |
| "step": 950 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 5.052711009979248, | |
| "learning_rate": 4.052319842053307e-05, | |
| "loss": 0.3693, | |
| "step": 960 | |
| }, | |
| { | |
| "epoch": 0.96, | |
| "grad_norm": 0.37648436427116394, | |
| "learning_rate": 4.042448173741363e-05, | |
| "loss": 0.1054, | |
| "step": 970 | |
| }, | |
| { | |
| "epoch": 0.97, | |
| "grad_norm": 18.3348445892334, | |
| "learning_rate": 4.032576505429418e-05, | |
| "loss": 0.3397, | |
| "step": 980 | |
| }, | |
| { | |
| "epoch": 0.98, | |
| "grad_norm": 10.808074951171875, | |
| "learning_rate": 4.022704837117473e-05, | |
| "loss": 0.3628, | |
| "step": 990 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 141.88064575195312, | |
| "learning_rate": 4.012833168805528e-05, | |
| "loss": 0.9269, | |
| "step": 1000 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "grad_norm": 0.555182695388794, | |
| "learning_rate": 4.002961500493584e-05, | |
| "loss": 0.1197, | |
| "step": 1010 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_balanced accuracy": 0.917760474601409, | |
| "eval_f1": 0.9176981176842771, | |
| "eval_loss": 0.40740078687667847, | |
| "eval_precision": 0.9176448492816227, | |
| "eval_recall": 0.917760474601409, | |
| "eval_runtime": 5.5647, | |
| "eval_samples_per_second": 161.733, | |
| "eval_steps_per_second": 10.243, | |
| "step": 1013 | |
| }, | |
| { | |
| "epoch": 1.01, | |
| "grad_norm": 4.549361228942871, | |
| "learning_rate": 3.993089832181639e-05, | |
| "loss": 0.5231, | |
| "step": 1020 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 5.699501991271973, | |
| "learning_rate": 3.983218163869694e-05, | |
| "loss": 0.4139, | |
| "step": 1030 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 2.1153147220611572, | |
| "learning_rate": 3.973346495557749e-05, | |
| "loss": 0.2718, | |
| "step": 1040 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 5.258866310119629, | |
| "learning_rate": 3.963474827245805e-05, | |
| "loss": 0.3115, | |
| "step": 1050 | |
| }, | |
| { | |
| "epoch": 1.05, | |
| "grad_norm": 13.351494789123535, | |
| "learning_rate": 3.95360315893386e-05, | |
| "loss": 0.3992, | |
| "step": 1060 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 9.7189359664917, | |
| "learning_rate": 3.943731490621915e-05, | |
| "loss": 0.1346, | |
| "step": 1070 | |
| }, | |
| { | |
| "epoch": 1.07, | |
| "grad_norm": 5.006288051605225, | |
| "learning_rate": 3.933859822309971e-05, | |
| "loss": 0.3118, | |
| "step": 1080 | |
| }, | |
| { | |
| "epoch": 1.08, | |
| "grad_norm": 7.094489574432373, | |
| "learning_rate": 3.923988153998026e-05, | |
| "loss": 0.1807, | |
| "step": 1090 | |
| }, | |
| { | |
| "epoch": 1.09, | |
| "grad_norm": 4.784492492675781, | |
| "learning_rate": 3.914116485686081e-05, | |
| "loss": 0.3839, | |
| "step": 1100 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.5643423795700073, | |
| "learning_rate": 3.904244817374136e-05, | |
| "loss": 0.1729, | |
| "step": 1110 | |
| }, | |
| { | |
| "epoch": 1.11, | |
| "grad_norm": 33.595703125, | |
| "learning_rate": 3.8943731490621916e-05, | |
| "loss": 0.1749, | |
| "step": 1120 | |
| }, | |
| { | |
| "epoch": 1.12, | |
| "grad_norm": 0.5887395143508911, | |
| "learning_rate": 3.884501480750247e-05, | |
| "loss": 0.2513, | |
| "step": 1130 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 22.53057289123535, | |
| "learning_rate": 3.874629812438302e-05, | |
| "loss": 0.2858, | |
| "step": 1140 | |
| }, | |
| { | |
| "epoch": 1.14, | |
| "grad_norm": 52.66212463378906, | |
| "learning_rate": 3.8647581441263576e-05, | |
| "loss": 0.1328, | |
| "step": 1150 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 5.8826117515563965, | |
| "learning_rate": 3.8548864758144125e-05, | |
| "loss": 0.3296, | |
| "step": 1160 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 10.208854675292969, | |
| "learning_rate": 3.845014807502468e-05, | |
| "loss": 0.1743, | |
| "step": 1170 | |
| }, | |
| { | |
| "epoch": 1.16, | |
| "grad_norm": 5.222922325134277, | |
| "learning_rate": 3.8351431391905236e-05, | |
| "loss": 0.2482, | |
| "step": 1180 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 0.3885471224784851, | |
| "learning_rate": 3.8252714708785785e-05, | |
| "loss": 0.3651, | |
| "step": 1190 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 68.36416625976562, | |
| "learning_rate": 3.815399802566634e-05, | |
| "loss": 0.5256, | |
| "step": 1200 | |
| }, | |
| { | |
| "epoch": 1.19, | |
| "grad_norm": 103.91950988769531, | |
| "learning_rate": 3.8055281342546896e-05, | |
| "loss": 0.2199, | |
| "step": 1210 | |
| }, | |
| { | |
| "epoch": 1.2, | |
| "grad_norm": 0.17333897948265076, | |
| "learning_rate": 3.7956564659427445e-05, | |
| "loss": 0.126, | |
| "step": 1220 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 41.487117767333984, | |
| "learning_rate": 3.7857847976308e-05, | |
| "loss": 0.2293, | |
| "step": 1230 | |
| }, | |
| { | |
| "epoch": 1.22, | |
| "grad_norm": 0.1527445763349533, | |
| "learning_rate": 3.775913129318855e-05, | |
| "loss": 0.2754, | |
| "step": 1240 | |
| }, | |
| { | |
| "epoch": 1.23, | |
| "grad_norm": 0.3720811605453491, | |
| "learning_rate": 3.7660414610069105e-05, | |
| "loss": 0.1904, | |
| "step": 1250 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 0.2801426947116852, | |
| "learning_rate": 3.756169792694966e-05, | |
| "loss": 0.2894, | |
| "step": 1260 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 0.912218451499939, | |
| "learning_rate": 3.746298124383021e-05, | |
| "loss": 0.4345, | |
| "step": 1270 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 0.25501587986946106, | |
| "learning_rate": 3.736426456071076e-05, | |
| "loss": 0.2249, | |
| "step": 1280 | |
| }, | |
| { | |
| "epoch": 1.27, | |
| "grad_norm": 19.25888442993164, | |
| "learning_rate": 3.7265547877591314e-05, | |
| "loss": 0.4532, | |
| "step": 1290 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 7.447415351867676, | |
| "learning_rate": 3.716683119447187e-05, | |
| "loss": 0.419, | |
| "step": 1300 | |
| }, | |
| { | |
| "epoch": 1.29, | |
| "grad_norm": 1.2623952627182007, | |
| "learning_rate": 3.706811451135242e-05, | |
| "loss": 0.3596, | |
| "step": 1310 | |
| }, | |
| { | |
| "epoch": 1.3, | |
| "grad_norm": 49.27845001220703, | |
| "learning_rate": 3.6969397828232974e-05, | |
| "loss": 0.1807, | |
| "step": 1320 | |
| }, | |
| { | |
| "epoch": 1.31, | |
| "grad_norm": 8.055280685424805, | |
| "learning_rate": 3.687068114511353e-05, | |
| "loss": 0.1877, | |
| "step": 1330 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 0.24801558256149292, | |
| "learning_rate": 3.677196446199408e-05, | |
| "loss": 0.1906, | |
| "step": 1340 | |
| }, | |
| { | |
| "epoch": 1.33, | |
| "grad_norm": 0.37148603796958923, | |
| "learning_rate": 3.6673247778874634e-05, | |
| "loss": 0.6613, | |
| "step": 1350 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 2.0603933334350586, | |
| "learning_rate": 3.657453109575518e-05, | |
| "loss": 0.1717, | |
| "step": 1360 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 1.4730746746063232, | |
| "learning_rate": 3.647581441263574e-05, | |
| "loss": 0.3606, | |
| "step": 1370 | |
| }, | |
| { | |
| "epoch": 1.36, | |
| "grad_norm": 11.129170417785645, | |
| "learning_rate": 3.6377097729516294e-05, | |
| "loss": 0.4668, | |
| "step": 1380 | |
| }, | |
| { | |
| "epoch": 1.37, | |
| "grad_norm": 107.76866912841797, | |
| "learning_rate": 3.627838104639684e-05, | |
| "loss": 0.4248, | |
| "step": 1390 | |
| }, | |
| { | |
| "epoch": 1.38, | |
| "grad_norm": 0.4574478566646576, | |
| "learning_rate": 3.617966436327739e-05, | |
| "loss": 0.2463, | |
| "step": 1400 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 9.523133277893066, | |
| "learning_rate": 3.6080947680157954e-05, | |
| "loss": 0.2986, | |
| "step": 1410 | |
| }, | |
| { | |
| "epoch": 1.4, | |
| "grad_norm": 724.2791137695312, | |
| "learning_rate": 3.59822309970385e-05, | |
| "loss": 0.1994, | |
| "step": 1420 | |
| }, | |
| { | |
| "epoch": 1.41, | |
| "grad_norm": 0.495822012424469, | |
| "learning_rate": 3.588351431391905e-05, | |
| "loss": 0.405, | |
| "step": 1430 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.7077971696853638, | |
| "learning_rate": 3.578479763079961e-05, | |
| "loss": 0.3258, | |
| "step": 1440 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 0.471545934677124, | |
| "learning_rate": 3.568608094768016e-05, | |
| "loss": 0.3381, | |
| "step": 1450 | |
| }, | |
| { | |
| "epoch": 1.44, | |
| "grad_norm": 160.64279174804688, | |
| "learning_rate": 3.558736426456071e-05, | |
| "loss": 0.4319, | |
| "step": 1460 | |
| }, | |
| { | |
| "epoch": 1.45, | |
| "grad_norm": 213.93475341796875, | |
| "learning_rate": 3.548864758144127e-05, | |
| "loss": 0.3506, | |
| "step": 1470 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 0.5124903917312622, | |
| "learning_rate": 3.5389930898321816e-05, | |
| "loss": 0.261, | |
| "step": 1480 | |
| }, | |
| { | |
| "epoch": 1.47, | |
| "grad_norm": 0.2033979296684265, | |
| "learning_rate": 3.529121421520237e-05, | |
| "loss": 0.3329, | |
| "step": 1490 | |
| }, | |
| { | |
| "epoch": 1.48, | |
| "grad_norm": 0.14042626321315765, | |
| "learning_rate": 3.519249753208293e-05, | |
| "loss": 0.198, | |
| "step": 1500 | |
| }, | |
| { | |
| "epoch": 1.49, | |
| "grad_norm": 0.052474942058324814, | |
| "learning_rate": 3.5093780848963476e-05, | |
| "loss": 0.3291, | |
| "step": 1510 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 0.7498096823692322, | |
| "learning_rate": 3.4995064165844024e-05, | |
| "loss": 0.5893, | |
| "step": 1520 | |
| }, | |
| { | |
| "epoch": 1.51, | |
| "grad_norm": 56.467071533203125, | |
| "learning_rate": 3.489634748272459e-05, | |
| "loss": 0.22, | |
| "step": 1530 | |
| }, | |
| { | |
| "epoch": 1.52, | |
| "grad_norm": 5.047154903411865, | |
| "learning_rate": 3.4797630799605136e-05, | |
| "loss": 0.3128, | |
| "step": 1540 | |
| }, | |
| { | |
| "epoch": 1.53, | |
| "grad_norm": 0.24173791706562042, | |
| "learning_rate": 3.4698914116485684e-05, | |
| "loss": 0.2632, | |
| "step": 1550 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 0.23745213449001312, | |
| "learning_rate": 3.460019743336624e-05, | |
| "loss": 0.1316, | |
| "step": 1560 | |
| }, | |
| { | |
| "epoch": 1.55, | |
| "grad_norm": 0.3697431683540344, | |
| "learning_rate": 3.4501480750246796e-05, | |
| "loss": 0.3162, | |
| "step": 1570 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": 125.36990356445312, | |
| "learning_rate": 3.4402764067127344e-05, | |
| "loss": 0.7252, | |
| "step": 1580 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 30.01531410217285, | |
| "learning_rate": 3.43040473840079e-05, | |
| "loss": 0.567, | |
| "step": 1590 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 44.524818420410156, | |
| "learning_rate": 3.420533070088845e-05, | |
| "loss": 0.4531, | |
| "step": 1600 | |
| }, | |
| { | |
| "epoch": 1.59, | |
| "grad_norm": 133.4363555908203, | |
| "learning_rate": 3.4106614017769004e-05, | |
| "loss": 0.4438, | |
| "step": 1610 | |
| }, | |
| { | |
| "epoch": 1.6, | |
| "grad_norm": 1119.47509765625, | |
| "learning_rate": 3.400789733464956e-05, | |
| "loss": 0.3973, | |
| "step": 1620 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 4.369329929351807, | |
| "learning_rate": 3.390918065153011e-05, | |
| "loss": 0.482, | |
| "step": 1630 | |
| }, | |
| { | |
| "epoch": 1.62, | |
| "grad_norm": 28.413909912109375, | |
| "learning_rate": 3.381046396841066e-05, | |
| "loss": 0.3454, | |
| "step": 1640 | |
| }, | |
| { | |
| "epoch": 1.63, | |
| "grad_norm": 76.58002471923828, | |
| "learning_rate": 3.371174728529122e-05, | |
| "loss": 0.2663, | |
| "step": 1650 | |
| }, | |
| { | |
| "epoch": 1.64, | |
| "grad_norm": 597.3102416992188, | |
| "learning_rate": 3.361303060217177e-05, | |
| "loss": 0.155, | |
| "step": 1660 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 24.984447479248047, | |
| "learning_rate": 3.351431391905232e-05, | |
| "loss": 0.2535, | |
| "step": 1670 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 30.53813934326172, | |
| "learning_rate": 3.341559723593287e-05, | |
| "loss": 0.315, | |
| "step": 1680 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 1.5513701438903809, | |
| "learning_rate": 3.331688055281343e-05, | |
| "loss": 0.3617, | |
| "step": 1690 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 3.676360845565796, | |
| "learning_rate": 3.321816386969398e-05, | |
| "loss": 0.6472, | |
| "step": 1700 | |
| }, | |
| { | |
| "epoch": 1.69, | |
| "grad_norm": 23.96689796447754, | |
| "learning_rate": 3.311944718657453e-05, | |
| "loss": 0.5382, | |
| "step": 1710 | |
| }, | |
| { | |
| "epoch": 1.7, | |
| "grad_norm": 18.116992950439453, | |
| "learning_rate": 3.302073050345508e-05, | |
| "loss": 0.345, | |
| "step": 1720 | |
| }, | |
| { | |
| "epoch": 1.71, | |
| "grad_norm": 4.786412239074707, | |
| "learning_rate": 3.292201382033564e-05, | |
| "loss": 0.3599, | |
| "step": 1730 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 2.5227644443511963, | |
| "learning_rate": 3.282329713721619e-05, | |
| "loss": 0.4313, | |
| "step": 1740 | |
| }, | |
| { | |
| "epoch": 1.73, | |
| "grad_norm": 4.462274074554443, | |
| "learning_rate": 3.272458045409674e-05, | |
| "loss": 0.5479, | |
| "step": 1750 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 49.19129180908203, | |
| "learning_rate": 3.26258637709773e-05, | |
| "loss": 0.5215, | |
| "step": 1760 | |
| }, | |
| { | |
| "epoch": 1.75, | |
| "grad_norm": 89.65460968017578, | |
| "learning_rate": 3.252714708785785e-05, | |
| "loss": 0.7555, | |
| "step": 1770 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 0.9293081760406494, | |
| "learning_rate": 3.24284304047384e-05, | |
| "loss": 0.3071, | |
| "step": 1780 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 11.949310302734375, | |
| "learning_rate": 3.232971372161895e-05, | |
| "loss": 0.2182, | |
| "step": 1790 | |
| }, | |
| { | |
| "epoch": 1.78, | |
| "grad_norm": 15.446320533752441, | |
| "learning_rate": 3.2230997038499506e-05, | |
| "loss": 0.2696, | |
| "step": 1800 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 5.7437567710876465, | |
| "learning_rate": 3.213228035538006e-05, | |
| "loss": 0.3771, | |
| "step": 1810 | |
| }, | |
| { | |
| "epoch": 1.8, | |
| "grad_norm": 209.53298950195312, | |
| "learning_rate": 3.203356367226061e-05, | |
| "loss": 0.3023, | |
| "step": 1820 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 1.2472151517868042, | |
| "learning_rate": 3.1934846989141166e-05, | |
| "loss": 0.29, | |
| "step": 1830 | |
| }, | |
| { | |
| "epoch": 1.82, | |
| "grad_norm": 370.38800048828125, | |
| "learning_rate": 3.1836130306021715e-05, | |
| "loss": 0.3409, | |
| "step": 1840 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 145.07717895507812, | |
| "learning_rate": 3.173741362290227e-05, | |
| "loss": 0.3839, | |
| "step": 1850 | |
| }, | |
| { | |
| "epoch": 1.84, | |
| "grad_norm": 48.441585540771484, | |
| "learning_rate": 3.1638696939782826e-05, | |
| "loss": 0.2765, | |
| "step": 1860 | |
| }, | |
| { | |
| "epoch": 1.85, | |
| "grad_norm": 8.114079475402832, | |
| "learning_rate": 3.1539980256663375e-05, | |
| "loss": 0.4797, | |
| "step": 1870 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 2.0335161685943604, | |
| "learning_rate": 3.144126357354393e-05, | |
| "loss": 0.3283, | |
| "step": 1880 | |
| }, | |
| { | |
| "epoch": 1.87, | |
| "grad_norm": 571.5001831054688, | |
| "learning_rate": 3.1342546890424486e-05, | |
| "loss": 0.3749, | |
| "step": 1890 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 26.4891414642334, | |
| "learning_rate": 3.1243830207305035e-05, | |
| "loss": 0.5855, | |
| "step": 1900 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 373.5781555175781, | |
| "learning_rate": 3.114511352418559e-05, | |
| "loss": 0.5779, | |
| "step": 1910 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 12.153056144714355, | |
| "learning_rate": 3.1046396841066146e-05, | |
| "loss": 0.5083, | |
| "step": 1920 | |
| }, | |
| { | |
| "epoch": 1.91, | |
| "grad_norm": 246.8365936279297, | |
| "learning_rate": 3.0947680157946695e-05, | |
| "loss": 0.2794, | |
| "step": 1930 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 16.38204002380371, | |
| "learning_rate": 3.084896347482725e-05, | |
| "loss": 0.1689, | |
| "step": 1940 | |
| }, | |
| { | |
| "epoch": 1.92, | |
| "grad_norm": 86.90618896484375, | |
| "learning_rate": 3.07502467917078e-05, | |
| "loss": 0.124, | |
| "step": 1950 | |
| }, | |
| { | |
| "epoch": 1.93, | |
| "grad_norm": 13.767565727233887, | |
| "learning_rate": 3.0651530108588355e-05, | |
| "loss": 0.4286, | |
| "step": 1960 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 25.554912567138672, | |
| "learning_rate": 3.0552813425468904e-05, | |
| "loss": 0.342, | |
| "step": 1970 | |
| }, | |
| { | |
| "epoch": 1.95, | |
| "grad_norm": 0.8774542212486267, | |
| "learning_rate": 3.045409674234946e-05, | |
| "loss": 0.2024, | |
| "step": 1980 | |
| }, | |
| { | |
| "epoch": 1.96, | |
| "grad_norm": 53.33576583862305, | |
| "learning_rate": 3.0355380059230008e-05, | |
| "loss": 0.3032, | |
| "step": 1990 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 6.252757549285889, | |
| "learning_rate": 3.0256663376110567e-05, | |
| "loss": 0.3633, | |
| "step": 2000 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 0.8560687899589539, | |
| "learning_rate": 3.015794669299112e-05, | |
| "loss": 0.2662, | |
| "step": 2010 | |
| }, | |
| { | |
| "epoch": 1.99, | |
| "grad_norm": 5.560061454772949, | |
| "learning_rate": 3.0059230009871668e-05, | |
| "loss": 0.3057, | |
| "step": 2020 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_balanced accuracy": 0.9113459399332592, | |
| "eval_f1": 0.9119394500117044, | |
| "eval_loss": 0.29854172468185425, | |
| "eval_precision": 0.913535516192521, | |
| "eval_recall": 0.9113459399332592, | |
| "eval_runtime": 5.6113, | |
| "eval_samples_per_second": 160.392, | |
| "eval_steps_per_second": 10.158, | |
| "step": 2026 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 5065, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 5, | |
| "save_steps": 500, | |
| "total_flos": 96787312128000.0, | |
| "train_batch_size": 8, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |