| { | |
| "best_metric": 0.7865168539325843, | |
| "best_model_checkpoint": "distilhubert-finetuned-not-a-word2/run-12/checkpoint-240", | |
| "epoch": 9.0, | |
| "eval_steps": 500, | |
| "global_step": 432, | |
| "is_hyper_param_search": true, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.1, | |
| "grad_norm": 1.2605946063995361, | |
| "learning_rate": 5.868670588038625e-06, | |
| "loss": 0.7019, | |
| "step": 5 | |
| }, | |
| { | |
| "epoch": 0.21, | |
| "grad_norm": 1.1554573774337769, | |
| "learning_rate": 1.173734117607725e-05, | |
| "loss": 0.6822, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.31, | |
| "grad_norm": 1.2631292343139648, | |
| "learning_rate": 1.7606011764115876e-05, | |
| "loss": 0.6823, | |
| "step": 15 | |
| }, | |
| { | |
| "epoch": 0.42, | |
| "grad_norm": 0.8986643552780151, | |
| "learning_rate": 2.34746823521545e-05, | |
| "loss": 0.6573, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.52, | |
| "grad_norm": 0.8833038210868835, | |
| "learning_rate": 2.934335294019313e-05, | |
| "loss": 0.6302, | |
| "step": 25 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 0.6974115967750549, | |
| "learning_rate": 3.521202352823175e-05, | |
| "loss": 0.6181, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 2.8611502647399902, | |
| "learning_rate": 4.1080694116270374e-05, | |
| "loss": 0.5431, | |
| "step": 35 | |
| }, | |
| { | |
| "epoch": 0.83, | |
| "grad_norm": 1.6229726076126099, | |
| "learning_rate": 4.6949364704309e-05, | |
| "loss": 0.6722, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.94, | |
| "grad_norm": 3.6481218338012695, | |
| "learning_rate": 5.281803529234763e-05, | |
| "loss": 0.6342, | |
| "step": 45 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_f1": 0.7326732673267327, | |
| "eval_loss": 0.7596015930175781, | |
| "eval_runtime": 1.3819, | |
| "eval_samples_per_second": 46.313, | |
| "eval_steps_per_second": 5.789, | |
| "step": 48 | |
| }, | |
| { | |
| "epoch": 1.04, | |
| "grad_norm": 4.841191291809082, | |
| "learning_rate": 5.607840784125798e-05, | |
| "loss": 0.7106, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 1.15, | |
| "grad_norm": 0.9242434501647949, | |
| "learning_rate": 5.542633333147591e-05, | |
| "loss": 0.703, | |
| "step": 55 | |
| }, | |
| { | |
| "epoch": 1.25, | |
| "grad_norm": 1.9776262044906616, | |
| "learning_rate": 5.4774258821693835e-05, | |
| "loss": 0.4098, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 0.7563148140907288, | |
| "learning_rate": 5.412218431191177e-05, | |
| "loss": 0.5599, | |
| "step": 65 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 1.44520103931427, | |
| "learning_rate": 5.34701098021297e-05, | |
| "loss": 0.4823, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 1.56, | |
| "grad_norm": Infinity, | |
| "learning_rate": 5.294845019430404e-05, | |
| "loss": 0.7782, | |
| "step": 75 | |
| }, | |
| { | |
| "epoch": 1.67, | |
| "grad_norm": 2.0363686084747314, | |
| "learning_rate": 5.229637568452197e-05, | |
| "loss": 0.6186, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 1.77, | |
| "grad_norm": 1.6177257299423218, | |
| "learning_rate": 5.16443011747399e-05, | |
| "loss": 0.5907, | |
| "step": 85 | |
| }, | |
| { | |
| "epoch": 1.88, | |
| "grad_norm": 1.8490890264511108, | |
| "learning_rate": 5.0992226664957835e-05, | |
| "loss": 0.5909, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 1.98, | |
| "grad_norm": 6.961443901062012, | |
| "learning_rate": 5.034015215517576e-05, | |
| "loss": 0.603, | |
| "step": 95 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_f1": 0.74, | |
| "eval_loss": 0.6603889465332031, | |
| "eval_runtime": 1.4385, | |
| "eval_samples_per_second": 44.492, | |
| "eval_steps_per_second": 5.561, | |
| "step": 96 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 1.232475757598877, | |
| "learning_rate": 4.981849254735011e-05, | |
| "loss": 0.572, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 1.0792688131332397, | |
| "learning_rate": 4.9166418037568046e-05, | |
| "loss": 0.4763, | |
| "step": 105 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 1.4945162534713745, | |
| "learning_rate": 4.8514343527785974e-05, | |
| "loss": 0.5616, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 2.4, | |
| "grad_norm": 1.323120355606079, | |
| "learning_rate": 4.78622690180039e-05, | |
| "loss": 0.5489, | |
| "step": 115 | |
| }, | |
| { | |
| "epoch": 2.5, | |
| "grad_norm": 5.674253463745117, | |
| "learning_rate": 4.721019450822183e-05, | |
| "loss": 0.5401, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 4.727626323699951, | |
| "learning_rate": 4.668853490039618e-05, | |
| "loss": 0.4278, | |
| "step": 125 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 8.496376037597656, | |
| "learning_rate": 4.6036460390614105e-05, | |
| "loss": 0.6216, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 2.81, | |
| "grad_norm": 2.796064615249634, | |
| "learning_rate": 4.538438588083204e-05, | |
| "loss": 0.3347, | |
| "step": 135 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 7.401052951812744, | |
| "learning_rate": 4.486272627300638e-05, | |
| "loss": 0.3109, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_f1": 0.72, | |
| "eval_loss": 0.8806066513061523, | |
| "eval_runtime": 1.3788, | |
| "eval_samples_per_second": 46.418, | |
| "eval_steps_per_second": 5.802, | |
| "step": 144 | |
| }, | |
| { | |
| "epoch": 3.02, | |
| "grad_norm": 3.1919455528259277, | |
| "learning_rate": 4.421065176322431e-05, | |
| "loss": 0.485, | |
| "step": 145 | |
| }, | |
| { | |
| "epoch": 3.12, | |
| "grad_norm": 56.47049331665039, | |
| "learning_rate": 4.355857725344224e-05, | |
| "loss": 0.3607, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 4.917665481567383, | |
| "learning_rate": 4.290650274366017e-05, | |
| "loss": 0.2077, | |
| "step": 155 | |
| }, | |
| { | |
| "epoch": 3.33, | |
| "grad_norm": 8.88060188293457, | |
| "learning_rate": 4.238484313583451e-05, | |
| "loss": 0.6496, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 3.44, | |
| "grad_norm": 1.9647566080093384, | |
| "learning_rate": 4.1732768626052447e-05, | |
| "loss": 0.1394, | |
| "step": 165 | |
| }, | |
| { | |
| "epoch": 3.54, | |
| "grad_norm": 50.96482849121094, | |
| "learning_rate": 4.1080694116270374e-05, | |
| "loss": 0.2834, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 3.65, | |
| "grad_norm": 132.5663299560547, | |
| "learning_rate": 4.042861960648831e-05, | |
| "loss": 0.485, | |
| "step": 175 | |
| }, | |
| { | |
| "epoch": 3.75, | |
| "grad_norm": 3.1624388694763184, | |
| "learning_rate": 3.977654509670624e-05, | |
| "loss": 0.2353, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 3.85, | |
| "grad_norm": 0.5671400427818298, | |
| "learning_rate": 3.912447058692417e-05, | |
| "loss": 0.3058, | |
| "step": 185 | |
| }, | |
| { | |
| "epoch": 3.96, | |
| "grad_norm": 23.86330223083496, | |
| "learning_rate": 3.84723960771421e-05, | |
| "loss": 0.3764, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_f1": 0.7500000000000001, | |
| "eval_loss": 1.1437296867370605, | |
| "eval_runtime": 1.4019, | |
| "eval_samples_per_second": 45.651, | |
| "eval_steps_per_second": 5.706, | |
| "step": 192 | |
| }, | |
| { | |
| "epoch": 4.06, | |
| "grad_norm": 0.443531334400177, | |
| "learning_rate": 3.782032156736003e-05, | |
| "loss": 0.2652, | |
| "step": 195 | |
| }, | |
| { | |
| "epoch": 4.17, | |
| "grad_norm": 5.118373394012451, | |
| "learning_rate": 3.716824705757796e-05, | |
| "loss": 0.3585, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 4.27, | |
| "grad_norm": 0.7104140520095825, | |
| "learning_rate": 3.651617254779589e-05, | |
| "loss": 0.2533, | |
| "step": 205 | |
| }, | |
| { | |
| "epoch": 4.38, | |
| "grad_norm": Infinity, | |
| "learning_rate": 3.599451293997023e-05, | |
| "loss": 0.1963, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 4.48, | |
| "grad_norm": 1.0030689239501953, | |
| "learning_rate": 3.5342438430188165e-05, | |
| "loss": 0.0602, | |
| "step": 215 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 193.33819580078125, | |
| "learning_rate": 3.469036392040609e-05, | |
| "loss": 0.2754, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 4.69, | |
| "grad_norm": 1.3048127889633179, | |
| "learning_rate": 3.403828941062403e-05, | |
| "loss": 0.1878, | |
| "step": 225 | |
| }, | |
| { | |
| "epoch": 4.79, | |
| "grad_norm": 0.18795377016067505, | |
| "learning_rate": 3.3386214900841956e-05, | |
| "loss": 0.0245, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 4.9, | |
| "grad_norm": 0.7579631805419922, | |
| "learning_rate": 3.2864555293016304e-05, | |
| "loss": 0.3502, | |
| "step": 235 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "grad_norm": 0.39707404375076294, | |
| "learning_rate": 3.221248078323424e-05, | |
| "loss": 0.2378, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.0, | |
| "eval_f1": 0.7865168539325843, | |
| "eval_loss": 1.2380468845367432, | |
| "eval_runtime": 1.3627, | |
| "eval_samples_per_second": 46.967, | |
| "eval_steps_per_second": 5.871, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 5.1, | |
| "grad_norm": 0.18084272742271423, | |
| "learning_rate": 3.1560406273452166e-05, | |
| "loss": 0.0143, | |
| "step": 245 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "grad_norm": 0.6576858758926392, | |
| "learning_rate": 3.0908331763670094e-05, | |
| "loss": 0.0151, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 5.31, | |
| "grad_norm": 1.1509562730789185, | |
| "learning_rate": 3.025625725388803e-05, | |
| "loss": 0.3467, | |
| "step": 255 | |
| }, | |
| { | |
| "epoch": 5.42, | |
| "grad_norm": 29.21381187438965, | |
| "learning_rate": 2.9604182744105953e-05, | |
| "loss": 0.0524, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "grad_norm": 0.1710425168275833, | |
| "learning_rate": 2.8952108234323884e-05, | |
| "loss": 0.0106, | |
| "step": 265 | |
| }, | |
| { | |
| "epoch": 5.62, | |
| "grad_norm": 0.14679908752441406, | |
| "learning_rate": 2.8300033724541812e-05, | |
| "loss": 0.0411, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 5.73, | |
| "grad_norm": 85.6775894165039, | |
| "learning_rate": 2.7647959214759747e-05, | |
| "loss": 0.4682, | |
| "step": 275 | |
| }, | |
| { | |
| "epoch": 5.83, | |
| "grad_norm": 0.12999875843524933, | |
| "learning_rate": 2.6995884704977678e-05, | |
| "loss": 0.2671, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 5.94, | |
| "grad_norm": 23.26946258544922, | |
| "learning_rate": 2.634381019519561e-05, | |
| "loss": 0.2224, | |
| "step": 285 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_f1": 0.7741935483870968, | |
| "eval_loss": 1.5522700548171997, | |
| "eval_runtime": 1.3653, | |
| "eval_samples_per_second": 46.878, | |
| "eval_steps_per_second": 5.86, | |
| "step": 288 | |
| }, | |
| { | |
| "epoch": 6.04, | |
| "grad_norm": 0.12312161922454834, | |
| "learning_rate": 2.569173568541354e-05, | |
| "loss": 0.2473, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "grad_norm": 0.21050110459327698, | |
| "learning_rate": 2.503966117563147e-05, | |
| "loss": 0.0108, | |
| "step": 295 | |
| }, | |
| { | |
| "epoch": 6.25, | |
| "grad_norm": 0.1071925163269043, | |
| "learning_rate": 2.43875866658494e-05, | |
| "loss": 0.0104, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 6.35, | |
| "grad_norm": 0.5196150541305542, | |
| "learning_rate": 2.3735512156067328e-05, | |
| "loss": 0.0197, | |
| "step": 305 | |
| }, | |
| { | |
| "epoch": 6.46, | |
| "grad_norm": 0.13510240614414215, | |
| "learning_rate": 2.308343764628526e-05, | |
| "loss": 0.023, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 6.56, | |
| "grad_norm": 0.20159611105918884, | |
| "learning_rate": 2.243136313650319e-05, | |
| "loss": 0.2273, | |
| "step": 315 | |
| }, | |
| { | |
| "epoch": 6.67, | |
| "grad_norm": 106.47545623779297, | |
| "learning_rate": 2.177928862672112e-05, | |
| "loss": 0.3449, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 6.77, | |
| "grad_norm": 0.2443103939294815, | |
| "learning_rate": 2.1127214116939053e-05, | |
| "loss": 0.3164, | |
| "step": 325 | |
| }, | |
| { | |
| "epoch": 6.88, | |
| "grad_norm": 19.54550552368164, | |
| "learning_rate": 2.0475139607156984e-05, | |
| "loss": 0.0099, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 6.98, | |
| "grad_norm": 0.08734071254730225, | |
| "learning_rate": 1.9823065097374912e-05, | |
| "loss": 0.2767, | |
| "step": 335 | |
| }, | |
| { | |
| "epoch": 7.0, | |
| "eval_f1": 0.7578947368421053, | |
| "eval_loss": 1.8833692073822021, | |
| "eval_runtime": 1.3822, | |
| "eval_samples_per_second": 46.303, | |
| "eval_steps_per_second": 5.788, | |
| "step": 336 | |
| }, | |
| { | |
| "epoch": 7.08, | |
| "grad_norm": 19.562559127807617, | |
| "learning_rate": 1.9170990587592843e-05, | |
| "loss": 0.2792, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 7.19, | |
| "grad_norm": 0.08157272636890411, | |
| "learning_rate": 1.8518916077810774e-05, | |
| "loss": 0.031, | |
| "step": 345 | |
| }, | |
| { | |
| "epoch": 7.29, | |
| "grad_norm": 0.14632558822631836, | |
| "learning_rate": 1.7866841568028706e-05, | |
| "loss": 0.0108, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 7.4, | |
| "grad_norm": 0.40814200043678284, | |
| "learning_rate": 1.7214767058246637e-05, | |
| "loss": 0.0057, | |
| "step": 355 | |
| }, | |
| { | |
| "epoch": 7.5, | |
| "grad_norm": 0.07355981320142746, | |
| "learning_rate": 1.6562692548464565e-05, | |
| "loss": 0.0056, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 7.6, | |
| "grad_norm": 0.10633735358715057, | |
| "learning_rate": 1.5910618038682496e-05, | |
| "loss": 0.005, | |
| "step": 365 | |
| }, | |
| { | |
| "epoch": 7.71, | |
| "grad_norm": 0.4227176308631897, | |
| "learning_rate": 1.5258543528900426e-05, | |
| "loss": 0.0337, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 7.81, | |
| "grad_norm": 0.06743155419826508, | |
| "learning_rate": 1.4606469019118355e-05, | |
| "loss": 0.2789, | |
| "step": 375 | |
| }, | |
| { | |
| "epoch": 7.92, | |
| "grad_norm": 0.07822620868682861, | |
| "learning_rate": 1.3954394509336287e-05, | |
| "loss": 0.0048, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 8.0, | |
| "eval_f1": 0.782608695652174, | |
| "eval_loss": 1.4861700534820557, | |
| "eval_runtime": 1.3792, | |
| "eval_samples_per_second": 46.405, | |
| "eval_steps_per_second": 5.801, | |
| "step": 384 | |
| }, | |
| { | |
| "epoch": 8.02, | |
| "grad_norm": 0.31482478976249695, | |
| "learning_rate": 1.3302319999554218e-05, | |
| "loss": 0.006, | |
| "step": 385 | |
| }, | |
| { | |
| "epoch": 8.12, | |
| "grad_norm": 0.08642608672380447, | |
| "learning_rate": 1.2650245489772149e-05, | |
| "loss": 0.0042, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 8.23, | |
| "grad_norm": 0.061254099011421204, | |
| "learning_rate": 1.1998170979990079e-05, | |
| "loss": 0.2253, | |
| "step": 395 | |
| }, | |
| { | |
| "epoch": 8.33, | |
| "grad_norm": 0.06673790514469147, | |
| "learning_rate": 1.134609647020801e-05, | |
| "loss": 0.273, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 8.44, | |
| "grad_norm": 0.09871594607830048, | |
| "learning_rate": 1.0694021960425941e-05, | |
| "loss": 0.0041, | |
| "step": 405 | |
| }, | |
| { | |
| "epoch": 8.54, | |
| "grad_norm": 0.10887040942907333, | |
| "learning_rate": 1.0041947450643869e-05, | |
| "loss": 0.0044, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 8.65, | |
| "grad_norm": 0.06360302865505219, | |
| "learning_rate": 9.3898729408618e-06, | |
| "loss": 0.0046, | |
| "step": 415 | |
| }, | |
| { | |
| "epoch": 8.75, | |
| "grad_norm": 0.13728582859039307, | |
| "learning_rate": 8.737798431079732e-06, | |
| "loss": 0.009, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 8.85, | |
| "grad_norm": 0.07806556671857834, | |
| "learning_rate": 8.085723921297661e-06, | |
| "loss": 0.0048, | |
| "step": 425 | |
| }, | |
| { | |
| "epoch": 8.96, | |
| "grad_norm": 0.07510797679424286, | |
| "learning_rate": 7.433649411515592e-06, | |
| "loss": 0.0049, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 9.0, | |
| "eval_f1": 0.7857142857142857, | |
| "eval_loss": 1.4449275732040405, | |
| "eval_runtime": 1.3668, | |
| "eval_samples_per_second": 46.824, | |
| "eval_steps_per_second": 5.853, | |
| "step": 432 | |
| } | |
| ], | |
| "logging_steps": 5, | |
| "max_steps": 480, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 10, | |
| "save_steps": 500, | |
| "total_flos": 6549027725117760.0, | |
| "train_batch_size": 4, | |
| "trial_name": null, | |
| "trial_params": { | |
| "learning_rate": 5.6339237645170805e-05, | |
| "per_device_train_batch_size": 4 | |
| } | |
| } | |