{ "best_metric": 0.99, "best_model_checkpoint": "./ViT-LungCancerbeit2e-5/checkpoint-1256", "epoch": 30.0, "global_step": 4710, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.06, "learning_rate": 4.246284501061571e-07, "loss": 1.4025, "step": 10 }, { "epoch": 0.13, "learning_rate": 8.492569002123142e-07, "loss": 1.3924, "step": 20 }, { "epoch": 0.19, "learning_rate": 1.2738853503184715e-06, "loss": 1.3456, "step": 30 }, { "epoch": 0.25, "learning_rate": 1.6985138004246284e-06, "loss": 1.2978, "step": 40 }, { "epoch": 0.32, "learning_rate": 2.1231422505307858e-06, "loss": 1.1964, "step": 50 }, { "epoch": 0.38, "learning_rate": 2.547770700636943e-06, "loss": 1.0992, "step": 60 }, { "epoch": 0.45, "learning_rate": 2.9723991507431e-06, "loss": 1.0569, "step": 70 }, { "epoch": 0.51, "learning_rate": 3.397027600849257e-06, "loss": 0.9881, "step": 80 }, { "epoch": 0.57, "learning_rate": 3.821656050955415e-06, "loss": 0.9427, "step": 90 }, { "epoch": 0.64, "learning_rate": 4.2462845010615716e-06, "loss": 0.886, "step": 100 }, { "epoch": 0.7, "learning_rate": 4.670912951167728e-06, "loss": 0.8384, "step": 110 }, { "epoch": 0.76, "learning_rate": 5.095541401273886e-06, "loss": 0.7751, "step": 120 }, { "epoch": 0.83, "learning_rate": 5.520169851380043e-06, "loss": 0.7307, "step": 130 }, { "epoch": 0.89, "learning_rate": 5.9447983014862e-06, "loss": 0.6992, "step": 140 }, { "epoch": 0.96, "learning_rate": 6.369426751592357e-06, "loss": 0.7137, "step": 150 }, { "epoch": 1.0, "eval_accuracy": 0.545, "eval_loss": 0.8226455450057983, "eval_runtime": 124.7601, "eval_samples_per_second": 3.206, "eval_steps_per_second": 0.401, "step": 157 }, { "epoch": 1.02, "learning_rate": 6.794055201698514e-06, "loss": 0.6937, "step": 160 }, { "epoch": 1.08, "learning_rate": 7.218683651804671e-06, "loss": 0.6486, "step": 170 }, { "epoch": 1.15, "learning_rate": 7.64331210191083e-06, "loss": 0.5992, "step": 180 }, { "epoch": 1.21, "learning_rate": 8.067940552016986e-06, "loss": 0.6029, "step": 190 }, { "epoch": 1.27, "learning_rate": 8.492569002123143e-06, "loss": 0.5752, "step": 200 }, { "epoch": 1.34, "learning_rate": 8.874734607218685e-06, "loss": 0.5625, "step": 210 }, { "epoch": 1.4, "learning_rate": 9.299363057324842e-06, "loss": 0.5727, "step": 220 }, { "epoch": 1.46, "learning_rate": 9.723991507430999e-06, "loss": 0.5125, "step": 230 }, { "epoch": 1.53, "learning_rate": 1.0148619957537155e-05, "loss": 0.4857, "step": 240 }, { "epoch": 1.59, "learning_rate": 1.0573248407643314e-05, "loss": 0.494, "step": 250 }, { "epoch": 1.66, "learning_rate": 1.099787685774947e-05, "loss": 0.4601, "step": 260 }, { "epoch": 1.72, "learning_rate": 1.1422505307855627e-05, "loss": 0.4402, "step": 270 }, { "epoch": 1.78, "learning_rate": 1.1847133757961784e-05, "loss": 0.4221, "step": 280 }, { "epoch": 1.85, "learning_rate": 1.227176220806794e-05, "loss": 0.3983, "step": 290 }, { "epoch": 1.91, "learning_rate": 1.2696390658174097e-05, "loss": 0.3952, "step": 300 }, { "epoch": 1.97, "learning_rate": 1.3121019108280256e-05, "loss": 0.4095, "step": 310 }, { "epoch": 2.0, "eval_accuracy": 0.8175, "eval_loss": 0.4449588656425476, "eval_runtime": 3.4033, "eval_samples_per_second": 117.532, "eval_steps_per_second": 14.691, "step": 314 }, { "epoch": 2.04, "learning_rate": 1.3545647558386413e-05, "loss": 0.3653, "step": 320 }, { "epoch": 2.1, "learning_rate": 1.397027600849257e-05, "loss": 0.2847, "step": 330 }, { "epoch": 2.17, "learning_rate": 1.4394904458598726e-05, "loss": 0.2728, "step": 340 }, { "epoch": 2.23, "learning_rate": 1.4819532908704883e-05, "loss": 0.2814, "step": 350 }, { "epoch": 2.29, "learning_rate": 1.5244161358811043e-05, "loss": 0.2642, "step": 360 }, { "epoch": 2.36, "learning_rate": 1.56687898089172e-05, "loss": 0.2365, "step": 370 }, { "epoch": 2.42, "learning_rate": 1.6093418259023356e-05, "loss": 0.2209, "step": 380 }, { "epoch": 2.48, "learning_rate": 1.6475583864118898e-05, "loss": 0.1913, "step": 390 }, { "epoch": 2.55, "learning_rate": 1.6900212314225057e-05, "loss": 0.2052, "step": 400 }, { "epoch": 2.61, "learning_rate": 1.732484076433121e-05, "loss": 0.2289, "step": 410 }, { "epoch": 2.68, "learning_rate": 1.774946921443737e-05, "loss": 0.1969, "step": 420 }, { "epoch": 2.74, "learning_rate": 1.8174097664543525e-05, "loss": 0.1984, "step": 430 }, { "epoch": 2.8, "learning_rate": 1.8598726114649684e-05, "loss": 0.1795, "step": 440 }, { "epoch": 2.87, "learning_rate": 1.9023354564755842e-05, "loss": 0.1749, "step": 450 }, { "epoch": 2.93, "learning_rate": 1.9447983014861997e-05, "loss": 0.1832, "step": 460 }, { "epoch": 2.99, "learning_rate": 1.9872611464968155e-05, "loss": 0.1446, "step": 470 }, { "epoch": 3.0, "eval_accuracy": 0.9525, "eval_loss": 0.17200100421905518, "eval_runtime": 3.8995, "eval_samples_per_second": 102.577, "eval_steps_per_second": 12.822, "step": 471 }, { "epoch": 3.06, "learning_rate": 1.996697334276952e-05, "loss": 0.108, "step": 480 }, { "epoch": 3.12, "learning_rate": 1.991979240386884e-05, "loss": 0.1203, "step": 490 }, { "epoch": 3.18, "learning_rate": 1.9872611464968155e-05, "loss": 0.0864, "step": 500 }, { "epoch": 3.25, "learning_rate": 1.982543052606747e-05, "loss": 0.1001, "step": 510 }, { "epoch": 3.31, "learning_rate": 1.9778249587166787e-05, "loss": 0.0915, "step": 520 }, { "epoch": 3.38, "learning_rate": 1.9731068648266103e-05, "loss": 0.0715, "step": 530 }, { "epoch": 3.44, "learning_rate": 1.968388770936542e-05, "loss": 0.0675, "step": 540 }, { "epoch": 3.5, "learning_rate": 1.9636706770464734e-05, "loss": 0.0877, "step": 550 }, { "epoch": 3.57, "learning_rate": 1.958952583156405e-05, "loss": 0.09, "step": 560 }, { "epoch": 3.63, "learning_rate": 1.9542344892663366e-05, "loss": 0.0647, "step": 570 }, { "epoch": 3.69, "learning_rate": 1.949516395376268e-05, "loss": 0.0583, "step": 580 }, { "epoch": 3.76, "learning_rate": 1.9447983014861997e-05, "loss": 0.061, "step": 590 }, { "epoch": 3.82, "learning_rate": 1.9400802075961313e-05, "loss": 0.0621, "step": 600 }, { "epoch": 3.89, "learning_rate": 1.935362113706063e-05, "loss": 0.0551, "step": 610 }, { "epoch": 3.95, "learning_rate": 1.9306440198159944e-05, "loss": 0.0883, "step": 620 }, { "epoch": 4.0, "eval_accuracy": 0.965, "eval_loss": 0.11082524061203003, "eval_runtime": 3.4167, "eval_samples_per_second": 117.073, "eval_steps_per_second": 14.634, "step": 628 }, { "epoch": 4.01, "learning_rate": 1.925925925925926e-05, "loss": 0.0584, "step": 630 }, { "epoch": 4.08, "learning_rate": 1.9212078320358576e-05, "loss": 0.0288, "step": 640 }, { "epoch": 4.14, "learning_rate": 1.916489738145789e-05, "loss": 0.0283, "step": 650 }, { "epoch": 4.2, "learning_rate": 1.9117716442557207e-05, "loss": 0.03, "step": 660 }, { "epoch": 4.27, "learning_rate": 1.9070535503656523e-05, "loss": 0.0209, "step": 670 }, { "epoch": 4.33, "learning_rate": 1.9023354564755842e-05, "loss": 0.0266, "step": 680 }, { "epoch": 4.39, "learning_rate": 1.8976173625855158e-05, "loss": 0.024, "step": 690 }, { "epoch": 4.46, "learning_rate": 1.8928992686954473e-05, "loss": 0.0403, "step": 700 }, { "epoch": 4.52, "learning_rate": 1.888181174805379e-05, "loss": 0.0115, "step": 710 }, { "epoch": 4.59, "learning_rate": 1.88346308091531e-05, "loss": 0.0479, "step": 720 }, { "epoch": 4.65, "learning_rate": 1.8787449870252417e-05, "loss": 0.0259, "step": 730 }, { "epoch": 4.71, "learning_rate": 1.8740268931351736e-05, "loss": 0.0125, "step": 740 }, { "epoch": 4.78, "learning_rate": 1.8693087992451052e-05, "loss": 0.0366, "step": 750 }, { "epoch": 4.84, "learning_rate": 1.8645907053550368e-05, "loss": 0.0225, "step": 760 }, { "epoch": 4.9, "learning_rate": 1.8598726114649684e-05, "loss": 0.0173, "step": 770 }, { "epoch": 4.97, "learning_rate": 1.8551545175749e-05, "loss": 0.0124, "step": 780 }, { "epoch": 5.0, "eval_accuracy": 0.9775, "eval_loss": 0.06490612775087357, "eval_runtime": 3.938, "eval_samples_per_second": 101.575, "eval_steps_per_second": 12.697, "step": 785 }, { "epoch": 5.03, "learning_rate": 1.8504364236848315e-05, "loss": 0.0124, "step": 790 }, { "epoch": 5.1, "learning_rate": 1.845718329794763e-05, "loss": 0.0073, "step": 800 }, { "epoch": 5.16, "learning_rate": 1.8410002359046946e-05, "loss": 0.0114, "step": 810 }, { "epoch": 5.22, "learning_rate": 1.8362821420146262e-05, "loss": 0.0064, "step": 820 }, { "epoch": 5.29, "learning_rate": 1.8315640481245578e-05, "loss": 0.0144, "step": 830 }, { "epoch": 5.35, "learning_rate": 1.8268459542344894e-05, "loss": 0.0037, "step": 840 }, { "epoch": 5.41, "learning_rate": 1.822127860344421e-05, "loss": 0.0028, "step": 850 }, { "epoch": 5.48, "learning_rate": 1.8174097664543525e-05, "loss": 0.0031, "step": 860 }, { "epoch": 5.54, "learning_rate": 1.812691672564284e-05, "loss": 0.0042, "step": 870 }, { "epoch": 5.61, "learning_rate": 1.807973578674216e-05, "loss": 0.0037, "step": 880 }, { "epoch": 5.67, "learning_rate": 1.8032554847841472e-05, "loss": 0.0037, "step": 890 }, { "epoch": 5.73, "learning_rate": 1.7985373908940788e-05, "loss": 0.0038, "step": 900 }, { "epoch": 5.8, "learning_rate": 1.7938192970040104e-05, "loss": 0.0073, "step": 910 }, { "epoch": 5.86, "learning_rate": 1.789101203113942e-05, "loss": 0.0147, "step": 920 }, { "epoch": 5.92, "learning_rate": 1.784383109223874e-05, "loss": 0.014, "step": 930 }, { "epoch": 5.99, "learning_rate": 1.7796650153338054e-05, "loss": 0.0036, "step": 940 }, { "epoch": 6.0, "eval_accuracy": 0.9825, "eval_loss": 0.05782594531774521, "eval_runtime": 3.5364, "eval_samples_per_second": 113.109, "eval_steps_per_second": 14.139, "step": 942 }, { "epoch": 6.05, "learning_rate": 1.774946921443737e-05, "loss": 0.0026, "step": 950 }, { "epoch": 6.11, "learning_rate": 1.7702288275536686e-05, "loss": 0.0029, "step": 960 }, { "epoch": 6.18, "learning_rate": 1.7655107336636e-05, "loss": 0.0024, "step": 970 }, { "epoch": 6.24, "learning_rate": 1.7607926397735314e-05, "loss": 0.0041, "step": 980 }, { "epoch": 6.31, "learning_rate": 1.7560745458834633e-05, "loss": 0.0034, "step": 990 }, { "epoch": 6.37, "learning_rate": 1.751356451993395e-05, "loss": 0.0029, "step": 1000 }, { "epoch": 6.43, "learning_rate": 1.7466383581033264e-05, "loss": 0.0013, "step": 1010 }, { "epoch": 6.5, "learning_rate": 1.741920264213258e-05, "loss": 0.0041, "step": 1020 }, { "epoch": 6.56, "learning_rate": 1.7372021703231896e-05, "loss": 0.0146, "step": 1030 }, { "epoch": 6.62, "learning_rate": 1.732484076433121e-05, "loss": 0.0202, "step": 1040 }, { "epoch": 6.69, "learning_rate": 1.7277659825430527e-05, "loss": 0.0089, "step": 1050 }, { "epoch": 6.75, "learning_rate": 1.7230478886529843e-05, "loss": 0.0038, "step": 1060 }, { "epoch": 6.82, "learning_rate": 1.718329794762916e-05, "loss": 0.0012, "step": 1070 }, { "epoch": 6.88, "learning_rate": 1.7136117008728475e-05, "loss": 0.0027, "step": 1080 }, { "epoch": 6.94, "learning_rate": 1.708893606982779e-05, "loss": 0.0056, "step": 1090 }, { "epoch": 7.0, "eval_accuracy": 0.9625, "eval_loss": 0.13798828423023224, "eval_runtime": 3.3996, "eval_samples_per_second": 117.662, "eval_steps_per_second": 14.708, "step": 1099 }, { "epoch": 7.01, "learning_rate": 1.7041755130927106e-05, "loss": 0.0033, "step": 1100 }, { "epoch": 7.07, "learning_rate": 1.6994574192026422e-05, "loss": 0.0079, "step": 1110 }, { "epoch": 7.13, "learning_rate": 1.6947393253125738e-05, "loss": 0.0587, "step": 1120 }, { "epoch": 7.2, "learning_rate": 1.6900212314225057e-05, "loss": 0.021, "step": 1130 }, { "epoch": 7.26, "learning_rate": 1.6853031375324372e-05, "loss": 0.0333, "step": 1140 }, { "epoch": 7.32, "learning_rate": 1.6805850436423685e-05, "loss": 0.0163, "step": 1150 }, { "epoch": 7.39, "learning_rate": 1.6758669497523e-05, "loss": 0.0094, "step": 1160 }, { "epoch": 7.45, "learning_rate": 1.6711488558622316e-05, "loss": 0.0014, "step": 1170 }, { "epoch": 7.52, "learning_rate": 1.6664307619721632e-05, "loss": 0.0048, "step": 1180 }, { "epoch": 7.58, "learning_rate": 1.661712668082095e-05, "loss": 0.0015, "step": 1190 }, { "epoch": 7.64, "learning_rate": 1.6569945741920267e-05, "loss": 0.0095, "step": 1200 }, { "epoch": 7.71, "learning_rate": 1.6522764803019582e-05, "loss": 0.0016, "step": 1210 }, { "epoch": 7.77, "learning_rate": 1.6475583864118898e-05, "loss": 0.0015, "step": 1220 }, { "epoch": 7.83, "learning_rate": 1.6428402925218214e-05, "loss": 0.0008, "step": 1230 }, { "epoch": 7.9, "learning_rate": 1.638122198631753e-05, "loss": 0.0029, "step": 1240 }, { "epoch": 7.96, "learning_rate": 1.6334041047416845e-05, "loss": 0.0015, "step": 1250 }, { "epoch": 8.0, "eval_accuracy": 0.99, "eval_loss": 0.05614970996975899, "eval_runtime": 3.8659, "eval_samples_per_second": 103.469, "eval_steps_per_second": 12.934, "step": 1256 }, { "epoch": 8.03, "learning_rate": 1.628686010851616e-05, "loss": 0.0057, "step": 1260 }, { "epoch": 8.09, "learning_rate": 1.6239679169615477e-05, "loss": 0.0012, "step": 1270 }, { "epoch": 8.15, "learning_rate": 1.6192498230714793e-05, "loss": 0.0013, "step": 1280 }, { "epoch": 8.22, "learning_rate": 1.614531729181411e-05, "loss": 0.0046, "step": 1290 }, { "epoch": 8.28, "learning_rate": 1.6098136352913424e-05, "loss": 0.0139, "step": 1300 }, { "epoch": 8.34, "learning_rate": 1.605095541401274e-05, "loss": 0.0014, "step": 1310 }, { "epoch": 8.41, "learning_rate": 1.6003774475112056e-05, "loss": 0.0076, "step": 1320 }, { "epoch": 8.47, "learning_rate": 1.595659353621137e-05, "loss": 0.0007, "step": 1330 }, { "epoch": 8.54, "learning_rate": 1.5909412597310687e-05, "loss": 0.0011, "step": 1340 }, { "epoch": 8.6, "learning_rate": 1.5862231658410003e-05, "loss": 0.0026, "step": 1350 }, { "epoch": 8.66, "learning_rate": 1.581505071950932e-05, "loss": 0.0008, "step": 1360 }, { "epoch": 8.73, "learning_rate": 1.5767869780608634e-05, "loss": 0.0044, "step": 1370 }, { "epoch": 8.79, "learning_rate": 1.5720688841707953e-05, "loss": 0.0079, "step": 1380 }, { "epoch": 8.85, "learning_rate": 1.567350790280727e-05, "loss": 0.0025, "step": 1390 }, { "epoch": 8.92, "learning_rate": 1.5626326963906585e-05, "loss": 0.0012, "step": 1400 }, { "epoch": 8.98, "learning_rate": 1.5579146025005897e-05, "loss": 0.0008, "step": 1410 }, { "epoch": 9.0, "eval_accuracy": 0.98, "eval_loss": 0.0886029601097107, "eval_runtime": 3.3874, "eval_samples_per_second": 118.085, "eval_steps_per_second": 14.761, "step": 1413 }, { "epoch": 9.04, "learning_rate": 1.5531965086105213e-05, "loss": 0.001, "step": 1420 }, { "epoch": 9.11, "learning_rate": 1.548478414720453e-05, "loss": 0.0065, "step": 1430 }, { "epoch": 9.17, "learning_rate": 1.5437603208303848e-05, "loss": 0.0005, "step": 1440 }, { "epoch": 9.24, "learning_rate": 1.5390422269403163e-05, "loss": 0.0007, "step": 1450 }, { "epoch": 9.3, "learning_rate": 1.534324133050248e-05, "loss": 0.0008, "step": 1460 }, { "epoch": 9.36, "learning_rate": 1.5296060391601795e-05, "loss": 0.0015, "step": 1470 }, { "epoch": 9.43, "learning_rate": 1.524887945270111e-05, "loss": 0.0009, "step": 1480 }, { "epoch": 9.49, "learning_rate": 1.5201698513800426e-05, "loss": 0.0005, "step": 1490 }, { "epoch": 9.55, "learning_rate": 1.515451757489974e-05, "loss": 0.0015, "step": 1500 }, { "epoch": 9.62, "learning_rate": 1.5107336635999058e-05, "loss": 0.0017, "step": 1510 }, { "epoch": 9.68, "learning_rate": 1.5060155697098374e-05, "loss": 0.0086, "step": 1520 }, { "epoch": 9.75, "learning_rate": 1.501297475819769e-05, "loss": 0.0034, "step": 1530 }, { "epoch": 9.81, "learning_rate": 1.4965793819297005e-05, "loss": 0.0023, "step": 1540 }, { "epoch": 9.87, "learning_rate": 1.491861288039632e-05, "loss": 0.0036, "step": 1550 }, { "epoch": 9.94, "learning_rate": 1.4871431941495638e-05, "loss": 0.0016, "step": 1560 }, { "epoch": 10.0, "learning_rate": 1.4824251002594954e-05, "loss": 0.0005, "step": 1570 }, { "epoch": 10.0, "eval_accuracy": 0.9825, "eval_loss": 0.09263397008180618, "eval_runtime": 3.4873, "eval_samples_per_second": 114.7, "eval_steps_per_second": 14.338, "step": 1570 }, { "epoch": 10.06, "learning_rate": 1.4777070063694268e-05, "loss": 0.0004, "step": 1580 }, { "epoch": 10.13, "learning_rate": 1.4729889124793584e-05, "loss": 0.0004, "step": 1590 }, { "epoch": 10.19, "learning_rate": 1.46827081858929e-05, "loss": 0.0008, "step": 1600 }, { "epoch": 10.25, "learning_rate": 1.4635527246992217e-05, "loss": 0.0005, "step": 1610 }, { "epoch": 10.32, "learning_rate": 1.4588346308091532e-05, "loss": 0.0009, "step": 1620 }, { "epoch": 10.38, "learning_rate": 1.4541165369190848e-05, "loss": 0.0008, "step": 1630 }, { "epoch": 10.45, "learning_rate": 1.4493984430290164e-05, "loss": 0.0003, "step": 1640 }, { "epoch": 10.51, "learning_rate": 1.4446803491389481e-05, "loss": 0.0004, "step": 1650 }, { "epoch": 10.57, "learning_rate": 1.4399622552488797e-05, "loss": 0.0006, "step": 1660 }, { "epoch": 10.64, "learning_rate": 1.4352441613588111e-05, "loss": 0.0003, "step": 1670 }, { "epoch": 10.7, "learning_rate": 1.4305260674687427e-05, "loss": 0.0003, "step": 1680 }, { "epoch": 10.76, "learning_rate": 1.4258079735786743e-05, "loss": 0.0002, "step": 1690 }, { "epoch": 10.83, "learning_rate": 1.4210898796886058e-05, "loss": 0.0004, "step": 1700 }, { "epoch": 10.89, "learning_rate": 1.4163717857985376e-05, "loss": 0.0002, "step": 1710 }, { "epoch": 10.96, "learning_rate": 1.4116536919084691e-05, "loss": 0.0002, "step": 1720 }, { "epoch": 11.0, "eval_accuracy": 0.98, "eval_loss": 0.08138495683670044, "eval_runtime": 3.8806, "eval_samples_per_second": 103.076, "eval_steps_per_second": 12.884, "step": 1727 }, { "epoch": 11.02, "learning_rate": 1.4069355980184007e-05, "loss": 0.0002, "step": 1730 }, { "epoch": 11.08, "learning_rate": 1.4022175041283323e-05, "loss": 0.0002, "step": 1740 }, { "epoch": 11.15, "learning_rate": 1.397499410238264e-05, "loss": 0.0002, "step": 1750 }, { "epoch": 11.21, "learning_rate": 1.3927813163481953e-05, "loss": 0.0002, "step": 1760 }, { "epoch": 11.27, "learning_rate": 1.388063222458127e-05, "loss": 0.0002, "step": 1770 }, { "epoch": 11.34, "learning_rate": 1.3833451285680586e-05, "loss": 0.0002, "step": 1780 }, { "epoch": 11.4, "learning_rate": 1.3786270346779902e-05, "loss": 0.0002, "step": 1790 }, { "epoch": 11.46, "learning_rate": 1.3739089407879217e-05, "loss": 0.0002, "step": 1800 }, { "epoch": 11.53, "learning_rate": 1.3691908468978535e-05, "loss": 0.0001, "step": 1810 }, { "epoch": 11.59, "learning_rate": 1.364472753007785e-05, "loss": 0.0002, "step": 1820 }, { "epoch": 11.66, "learning_rate": 1.3597546591177166e-05, "loss": 0.0004, "step": 1830 }, { "epoch": 11.72, "learning_rate": 1.355036565227648e-05, "loss": 0.0003, "step": 1840 }, { "epoch": 11.78, "learning_rate": 1.3503184713375796e-05, "loss": 0.0003, "step": 1850 }, { "epoch": 11.85, "learning_rate": 1.3456003774475113e-05, "loss": 0.0002, "step": 1860 }, { "epoch": 11.91, "learning_rate": 1.3408822835574429e-05, "loss": 0.0002, "step": 1870 }, { "epoch": 11.97, "learning_rate": 1.3361641896673745e-05, "loss": 0.0002, "step": 1880 }, { "epoch": 12.0, "eval_accuracy": 0.9825, "eval_loss": 0.0756840705871582, "eval_runtime": 3.4057, "eval_samples_per_second": 117.449, "eval_steps_per_second": 14.681, "step": 1884 }, { "epoch": 12.04, "learning_rate": 1.331446095777306e-05, "loss": 0.0002, "step": 1890 }, { "epoch": 12.1, "learning_rate": 1.3267280018872376e-05, "loss": 0.0001, "step": 1900 }, { "epoch": 12.17, "learning_rate": 1.3220099079971694e-05, "loss": 0.0001, "step": 1910 }, { "epoch": 12.23, "learning_rate": 1.317291814107101e-05, "loss": 0.0001, "step": 1920 }, { "epoch": 12.29, "learning_rate": 1.3125737202170324e-05, "loss": 0.0001, "step": 1930 }, { "epoch": 12.36, "learning_rate": 1.307855626326964e-05, "loss": 0.0001, "step": 1940 }, { "epoch": 12.42, "learning_rate": 1.3031375324368955e-05, "loss": 0.0002, "step": 1950 }, { "epoch": 12.48, "learning_rate": 1.2984194385468272e-05, "loss": 0.0001, "step": 1960 }, { "epoch": 12.55, "learning_rate": 1.2937013446567588e-05, "loss": 0.0001, "step": 1970 }, { "epoch": 12.61, "learning_rate": 1.2889832507666904e-05, "loss": 0.0001, "step": 1980 }, { "epoch": 12.68, "learning_rate": 1.284265156876622e-05, "loss": 0.0001, "step": 1990 }, { "epoch": 12.74, "learning_rate": 1.2795470629865537e-05, "loss": 0.0001, "step": 2000 }, { "epoch": 12.8, "learning_rate": 1.274828969096485e-05, "loss": 0.0001, "step": 2010 }, { "epoch": 12.87, "learning_rate": 1.2701108752064167e-05, "loss": 0.0001, "step": 2020 }, { "epoch": 12.93, "learning_rate": 1.2653927813163483e-05, "loss": 0.0001, "step": 2030 }, { "epoch": 12.99, "learning_rate": 1.2606746874262798e-05, "loss": 0.0001, "step": 2040 }, { "epoch": 13.0, "eval_accuracy": 0.9875, "eval_loss": 0.06277866661548615, "eval_runtime": 3.4996, "eval_samples_per_second": 114.3, "eval_steps_per_second": 14.287, "step": 2041 }, { "epoch": 13.06, "learning_rate": 1.2559565935362114e-05, "loss": 0.0001, "step": 2050 }, { "epoch": 13.12, "learning_rate": 1.2512384996461431e-05, "loss": 0.0001, "step": 2060 }, { "epoch": 13.18, "learning_rate": 1.2465204057560747e-05, "loss": 0.0001, "step": 2070 }, { "epoch": 13.25, "learning_rate": 1.2418023118660063e-05, "loss": 0.0001, "step": 2080 }, { "epoch": 13.31, "learning_rate": 1.2370842179759379e-05, "loss": 0.0001, "step": 2090 }, { "epoch": 13.38, "learning_rate": 1.2323661240858693e-05, "loss": 0.0001, "step": 2100 }, { "epoch": 13.44, "learning_rate": 1.227648030195801e-05, "loss": 0.0001, "step": 2110 }, { "epoch": 13.5, "learning_rate": 1.2229299363057326e-05, "loss": 0.0001, "step": 2120 }, { "epoch": 13.57, "learning_rate": 1.2182118424156642e-05, "loss": 0.0001, "step": 2130 }, { "epoch": 13.63, "learning_rate": 1.2134937485255957e-05, "loss": 0.0001, "step": 2140 }, { "epoch": 13.69, "learning_rate": 1.2087756546355273e-05, "loss": 0.0002, "step": 2150 }, { "epoch": 13.76, "learning_rate": 1.204057560745459e-05, "loss": 0.0001, "step": 2160 }, { "epoch": 13.82, "learning_rate": 1.1993394668553906e-05, "loss": 0.0001, "step": 2170 }, { "epoch": 13.89, "learning_rate": 1.1946213729653222e-05, "loss": 0.0001, "step": 2180 }, { "epoch": 13.95, "learning_rate": 1.1899032790752536e-05, "loss": 0.0001, "step": 2190 }, { "epoch": 14.0, "eval_accuracy": 0.9875, "eval_loss": 0.064232736825943, "eval_runtime": 3.911, "eval_samples_per_second": 102.275, "eval_steps_per_second": 12.784, "step": 2198 }, { "epoch": 14.01, "learning_rate": 1.1851851851851852e-05, "loss": 0.0001, "step": 2200 }, { "epoch": 14.08, "learning_rate": 1.1804670912951169e-05, "loss": 0.0001, "step": 2210 }, { "epoch": 14.14, "learning_rate": 1.1757489974050485e-05, "loss": 0.0001, "step": 2220 }, { "epoch": 14.2, "learning_rate": 1.17103090351498e-05, "loss": 0.0001, "step": 2230 }, { "epoch": 14.27, "learning_rate": 1.1663128096249116e-05, "loss": 0.0001, "step": 2240 }, { "epoch": 14.33, "learning_rate": 1.1615947157348432e-05, "loss": 0.0001, "step": 2250 }, { "epoch": 14.39, "learning_rate": 1.156876621844775e-05, "loss": 0.0001, "step": 2260 }, { "epoch": 14.46, "learning_rate": 1.1521585279547063e-05, "loss": 0.0001, "step": 2270 }, { "epoch": 14.52, "learning_rate": 1.147440434064638e-05, "loss": 0.0001, "step": 2280 }, { "epoch": 14.59, "learning_rate": 1.1427223401745695e-05, "loss": 0.0001, "step": 2290 }, { "epoch": 14.65, "learning_rate": 1.138004246284501e-05, "loss": 0.0001, "step": 2300 }, { "epoch": 14.71, "learning_rate": 1.1332861523944328e-05, "loss": 0.0001, "step": 2310 }, { "epoch": 14.78, "learning_rate": 1.1285680585043644e-05, "loss": 0.0001, "step": 2320 }, { "epoch": 14.84, "learning_rate": 1.123849964614296e-05, "loss": 0.0001, "step": 2330 }, { "epoch": 14.9, "learning_rate": 1.1191318707242275e-05, "loss": 0.0001, "step": 2340 }, { "epoch": 14.97, "learning_rate": 1.1144137768341593e-05, "loss": 0.0001, "step": 2350 }, { "epoch": 15.0, "eval_accuracy": 0.9875, "eval_loss": 0.0644247755408287, "eval_runtime": 3.4455, "eval_samples_per_second": 116.093, "eval_steps_per_second": 14.512, "step": 2355 }, { "epoch": 15.03, "learning_rate": 1.1096956829440905e-05, "loss": 0.0001, "step": 2360 }, { "epoch": 15.1, "learning_rate": 1.1049775890540222e-05, "loss": 0.0001, "step": 2370 }, { "epoch": 15.16, "learning_rate": 1.1002594951639538e-05, "loss": 0.0001, "step": 2380 }, { "epoch": 15.22, "learning_rate": 1.0955414012738854e-05, "loss": 0.0001, "step": 2390 }, { "epoch": 15.29, "learning_rate": 1.090823307383817e-05, "loss": 0.0001, "step": 2400 }, { "epoch": 15.35, "learning_rate": 1.0861052134937487e-05, "loss": 0.0001, "step": 2410 }, { "epoch": 15.41, "learning_rate": 1.0813871196036803e-05, "loss": 0.0001, "step": 2420 }, { "epoch": 15.48, "learning_rate": 1.0766690257136119e-05, "loss": 0.0001, "step": 2430 }, { "epoch": 15.54, "learning_rate": 1.0719509318235434e-05, "loss": 0.0001, "step": 2440 }, { "epoch": 15.61, "learning_rate": 1.0672328379334748e-05, "loss": 0.0001, "step": 2450 }, { "epoch": 15.67, "learning_rate": 1.0625147440434066e-05, "loss": 0.0001, "step": 2460 }, { "epoch": 15.73, "learning_rate": 1.0577966501533381e-05, "loss": 0.0001, "step": 2470 }, { "epoch": 15.8, "learning_rate": 1.0530785562632697e-05, "loss": 0.0001, "step": 2480 }, { "epoch": 15.86, "learning_rate": 1.0483604623732013e-05, "loss": 0.0001, "step": 2490 }, { "epoch": 15.92, "learning_rate": 1.0436423684831329e-05, "loss": 0.0001, "step": 2500 }, { "epoch": 15.99, "learning_rate": 1.0389242745930646e-05, "loss": 0.0001, "step": 2510 }, { "epoch": 16.0, "eval_accuracy": 0.9875, "eval_loss": 0.06587812304496765, "eval_runtime": 3.5758, "eval_samples_per_second": 111.862, "eval_steps_per_second": 13.983, "step": 2512 }, { "epoch": 16.05, "learning_rate": 1.0342061807029962e-05, "loss": 0.0001, "step": 2520 }, { "epoch": 16.11, "learning_rate": 1.0294880868129276e-05, "loss": 0.0001, "step": 2530 }, { "epoch": 16.18, "learning_rate": 1.0247699929228592e-05, "loss": 0.0001, "step": 2540 }, { "epoch": 16.24, "learning_rate": 1.0200518990327907e-05, "loss": 0.0001, "step": 2550 }, { "epoch": 16.31, "learning_rate": 1.0153338051427225e-05, "loss": 0.0001, "step": 2560 }, { "epoch": 16.37, "learning_rate": 1.010615711252654e-05, "loss": 0.0001, "step": 2570 }, { "epoch": 16.43, "learning_rate": 1.0058976173625856e-05, "loss": 0.0001, "step": 2580 }, { "epoch": 16.5, "learning_rate": 1.0011795234725172e-05, "loss": 0.0001, "step": 2590 }, { "epoch": 16.56, "learning_rate": 9.964614295824488e-06, "loss": 0.0001, "step": 2600 }, { "epoch": 16.62, "learning_rate": 9.917433356923803e-06, "loss": 0.0001, "step": 2610 }, { "epoch": 16.69, "learning_rate": 9.870252418023119e-06, "loss": 0.0001, "step": 2620 }, { "epoch": 16.75, "learning_rate": 9.823071479122437e-06, "loss": 0.0001, "step": 2630 }, { "epoch": 16.82, "learning_rate": 9.77589054022175e-06, "loss": 0.0001, "step": 2640 }, { "epoch": 16.88, "learning_rate": 9.728709601321066e-06, "loss": 0.0001, "step": 2650 }, { "epoch": 16.94, "learning_rate": 9.681528662420384e-06, "loss": 0.0001, "step": 2660 }, { "epoch": 17.0, "eval_accuracy": 0.9875, "eval_loss": 0.06573827564716339, "eval_runtime": 3.9163, "eval_samples_per_second": 102.137, "eval_steps_per_second": 12.767, "step": 2669 }, { "epoch": 17.01, "learning_rate": 9.6343477235197e-06, "loss": 0.0001, "step": 2670 }, { "epoch": 17.07, "learning_rate": 9.587166784619013e-06, "loss": 0.0001, "step": 2680 }, { "epoch": 17.13, "learning_rate": 9.539985845718331e-06, "loss": 0.0001, "step": 2690 }, { "epoch": 17.2, "learning_rate": 9.492804906817647e-06, "loss": 0.0001, "step": 2700 }, { "epoch": 17.26, "learning_rate": 9.445623967916962e-06, "loss": 0.0001, "step": 2710 }, { "epoch": 17.32, "learning_rate": 9.398443029016278e-06, "loss": 0.0001, "step": 2720 }, { "epoch": 17.39, "learning_rate": 9.351262090115594e-06, "loss": 0.0001, "step": 2730 }, { "epoch": 17.45, "learning_rate": 9.30408115121491e-06, "loss": 0.0001, "step": 2740 }, { "epoch": 17.52, "learning_rate": 9.256900212314225e-06, "loss": 0.0001, "step": 2750 }, { "epoch": 17.58, "learning_rate": 9.209719273413543e-06, "loss": 0.0001, "step": 2760 }, { "epoch": 17.64, "learning_rate": 9.162538334512857e-06, "loss": 0.0001, "step": 2770 }, { "epoch": 17.71, "learning_rate": 9.115357395612172e-06, "loss": 0.0001, "step": 2780 }, { "epoch": 17.77, "learning_rate": 9.06817645671149e-06, "loss": 0.0001, "step": 2790 }, { "epoch": 17.83, "learning_rate": 9.020995517810806e-06, "loss": 0.0001, "step": 2800 }, { "epoch": 17.9, "learning_rate": 8.973814578910121e-06, "loss": 0.0001, "step": 2810 }, { "epoch": 17.96, "learning_rate": 8.926633640009437e-06, "loss": 0.0001, "step": 2820 }, { "epoch": 18.0, "eval_accuracy": 0.9875, "eval_loss": 0.06650857627391815, "eval_runtime": 3.418, "eval_samples_per_second": 117.026, "eval_steps_per_second": 14.628, "step": 2826 }, { "epoch": 18.03, "learning_rate": 8.879452701108753e-06, "loss": 0.0001, "step": 2830 }, { "epoch": 18.09, "learning_rate": 8.832271762208069e-06, "loss": 0.0001, "step": 2840 }, { "epoch": 18.15, "learning_rate": 8.785090823307384e-06, "loss": 0.0001, "step": 2850 }, { "epoch": 18.22, "learning_rate": 8.7379098844067e-06, "loss": 0.0001, "step": 2860 }, { "epoch": 18.28, "learning_rate": 8.690728945506016e-06, "loss": 0.0001, "step": 2870 }, { "epoch": 18.34, "learning_rate": 8.643548006605333e-06, "loss": 0.0001, "step": 2880 }, { "epoch": 18.41, "learning_rate": 8.596367067704649e-06, "loss": 0.0001, "step": 2890 }, { "epoch": 18.47, "learning_rate": 8.549186128803963e-06, "loss": 0.0001, "step": 2900 }, { "epoch": 18.54, "learning_rate": 8.50200518990328e-06, "loss": 0.0001, "step": 2910 }, { "epoch": 18.6, "learning_rate": 8.454824251002596e-06, "loss": 0.0001, "step": 2920 }, { "epoch": 18.66, "learning_rate": 8.407643312101912e-06, "loss": 0.0001, "step": 2930 }, { "epoch": 18.73, "learning_rate": 8.360462373201228e-06, "loss": 0.0001, "step": 2940 }, { "epoch": 18.79, "learning_rate": 8.313281434300543e-06, "loss": 0.0001, "step": 2950 }, { "epoch": 18.85, "learning_rate": 8.266100495399859e-06, "loss": 0.0001, "step": 2960 }, { "epoch": 18.92, "learning_rate": 8.218919556499175e-06, "loss": 0.0001, "step": 2970 }, { "epoch": 18.98, "learning_rate": 8.171738617598492e-06, "loss": 0.0001, "step": 2980 }, { "epoch": 19.0, "eval_accuracy": 0.9875, "eval_loss": 0.06798232346773148, "eval_runtime": 3.8011, "eval_samples_per_second": 105.232, "eval_steps_per_second": 13.154, "step": 2983 }, { "epoch": 19.04, "learning_rate": 8.124557678697806e-06, "loss": 0.0001, "step": 2990 }, { "epoch": 19.11, "learning_rate": 8.077376739797122e-06, "loss": 0.0001, "step": 3000 }, { "epoch": 19.17, "learning_rate": 8.03019580089644e-06, "loss": 0.0001, "step": 3010 }, { "epoch": 19.24, "learning_rate": 7.983014861995755e-06, "loss": 0.0001, "step": 3020 }, { "epoch": 19.3, "learning_rate": 7.935833923095069e-06, "loss": 0.0001, "step": 3030 }, { "epoch": 19.36, "learning_rate": 7.888652984194387e-06, "loss": 0.0001, "step": 3040 }, { "epoch": 19.43, "learning_rate": 7.841472045293702e-06, "loss": 0.0001, "step": 3050 }, { "epoch": 19.49, "learning_rate": 7.794291106393018e-06, "loss": 0.0001, "step": 3060 }, { "epoch": 19.55, "learning_rate": 7.747110167492334e-06, "loss": 0.0001, "step": 3070 }, { "epoch": 19.62, "learning_rate": 7.69992922859165e-06, "loss": 0.0001, "step": 3080 }, { "epoch": 19.68, "learning_rate": 7.652748289690965e-06, "loss": 0.0001, "step": 3090 }, { "epoch": 19.75, "learning_rate": 7.605567350790282e-06, "loss": 0.0001, "step": 3100 }, { "epoch": 19.81, "learning_rate": 7.5583864118895975e-06, "loss": 0.0001, "step": 3110 }, { "epoch": 19.87, "learning_rate": 7.511205472988912e-06, "loss": 0.0, "step": 3120 }, { "epoch": 19.94, "learning_rate": 7.464024534088229e-06, "loss": 0.0001, "step": 3130 }, { "epoch": 20.0, "learning_rate": 7.416843595187545e-06, "loss": 0.0001, "step": 3140 }, { "epoch": 20.0, "eval_accuracy": 0.9875, "eval_loss": 0.06838066875934601, "eval_runtime": 3.3858, "eval_samples_per_second": 118.139, "eval_steps_per_second": 14.767, "step": 3140 }, { "epoch": 20.06, "learning_rate": 7.369662656286861e-06, "loss": 0.0001, "step": 3150 }, { "epoch": 20.13, "learning_rate": 7.322481717386176e-06, "loss": 0.0, "step": 3160 }, { "epoch": 20.19, "learning_rate": 7.275300778485493e-06, "loss": 0.0001, "step": 3170 }, { "epoch": 20.25, "learning_rate": 7.2281198395848084e-06, "loss": 0.0, "step": 3180 }, { "epoch": 20.32, "learning_rate": 7.180938900684124e-06, "loss": 0.0001, "step": 3190 }, { "epoch": 20.38, "learning_rate": 7.13375796178344e-06, "loss": 0.0001, "step": 3200 }, { "epoch": 20.45, "learning_rate": 7.086577022882756e-06, "loss": 0.0001, "step": 3210 }, { "epoch": 20.51, "learning_rate": 7.039396083982072e-06, "loss": 0.0, "step": 3220 }, { "epoch": 20.57, "learning_rate": 6.992215145081388e-06, "loss": 0.0, "step": 3230 }, { "epoch": 20.64, "learning_rate": 6.9450342061807045e-06, "loss": 0.0, "step": 3240 }, { "epoch": 20.7, "learning_rate": 6.897853267280019e-06, "loss": 0.0001, "step": 3250 }, { "epoch": 20.76, "learning_rate": 6.850672328379335e-06, "loss": 0.0001, "step": 3260 }, { "epoch": 20.83, "learning_rate": 6.803491389478652e-06, "loss": 0.0, "step": 3270 }, { "epoch": 20.89, "learning_rate": 6.7563104505779674e-06, "loss": 0.0, "step": 3280 }, { "epoch": 20.96, "learning_rate": 6.709129511677282e-06, "loss": 0.0001, "step": 3290 }, { "epoch": 21.0, "eval_accuracy": 0.9875, "eval_loss": 0.06882834434509277, "eval_runtime": 3.4429, "eval_samples_per_second": 116.181, "eval_steps_per_second": 14.523, "step": 3297 }, { "epoch": 21.02, "learning_rate": 6.661948572776599e-06, "loss": 0.0001, "step": 3300 }, { "epoch": 21.08, "learning_rate": 6.614767633875915e-06, "loss": 0.0001, "step": 3310 }, { "epoch": 21.15, "learning_rate": 6.567586694975231e-06, "loss": 0.0001, "step": 3320 }, { "epoch": 21.21, "learning_rate": 6.520405756074546e-06, "loss": 0.0, "step": 3330 }, { "epoch": 21.27, "learning_rate": 6.473224817173862e-06, "loss": 0.0, "step": 3340 }, { "epoch": 21.34, "learning_rate": 6.426043878273178e-06, "loss": 0.0, "step": 3350 }, { "epoch": 21.4, "learning_rate": 6.378862939372494e-06, "loss": 0.0, "step": 3360 }, { "epoch": 21.46, "learning_rate": 6.331682000471809e-06, "loss": 0.0, "step": 3370 }, { "epoch": 21.53, "learning_rate": 6.284501061571126e-06, "loss": 0.0, "step": 3380 }, { "epoch": 21.59, "learning_rate": 6.237320122670441e-06, "loss": 0.0, "step": 3390 }, { "epoch": 21.66, "learning_rate": 6.190139183769758e-06, "loss": 0.0, "step": 3400 }, { "epoch": 21.72, "learning_rate": 6.142958244869074e-06, "loss": 0.0001, "step": 3410 }, { "epoch": 21.78, "learning_rate": 6.0957773059683885e-06, "loss": 0.0001, "step": 3420 }, { "epoch": 21.85, "learning_rate": 6.048596367067705e-06, "loss": 0.0, "step": 3430 }, { "epoch": 21.91, "learning_rate": 6.001415428167021e-06, "loss": 0.0001, "step": 3440 }, { "epoch": 21.97, "learning_rate": 5.954234489266337e-06, "loss": 0.0, "step": 3450 }, { "epoch": 22.0, "eval_accuracy": 0.9875, "eval_loss": 0.06946733593940735, "eval_runtime": 3.9313, "eval_samples_per_second": 101.749, "eval_steps_per_second": 12.719, "step": 3454 }, { "epoch": 22.04, "learning_rate": 5.907053550365652e-06, "loss": 0.0, "step": 3460 }, { "epoch": 22.1, "learning_rate": 5.859872611464969e-06, "loss": 0.0, "step": 3470 }, { "epoch": 22.17, "learning_rate": 5.812691672564285e-06, "loss": 0.0001, "step": 3480 }, { "epoch": 22.23, "learning_rate": 5.7655107336636e-06, "loss": 0.0, "step": 3490 }, { "epoch": 22.29, "learning_rate": 5.718329794762916e-06, "loss": 0.0, "step": 3500 }, { "epoch": 22.36, "learning_rate": 5.671148855862232e-06, "loss": 0.0, "step": 3510 }, { "epoch": 22.42, "learning_rate": 5.623967916961548e-06, "loss": 0.0, "step": 3520 }, { "epoch": 22.48, "learning_rate": 5.576786978060864e-06, "loss": 0.0, "step": 3530 }, { "epoch": 22.55, "learning_rate": 5.529606039160181e-06, "loss": 0.0, "step": 3540 }, { "epoch": 22.61, "learning_rate": 5.4824251002594955e-06, "loss": 0.0, "step": 3550 }, { "epoch": 22.68, "learning_rate": 5.435244161358811e-06, "loss": 0.0001, "step": 3560 }, { "epoch": 22.74, "learning_rate": 5.388063222458128e-06, "loss": 0.0, "step": 3570 }, { "epoch": 22.8, "learning_rate": 5.340882283557444e-06, "loss": 0.0, "step": 3580 }, { "epoch": 22.87, "learning_rate": 5.2937013446567585e-06, "loss": 0.0, "step": 3590 }, { "epoch": 22.93, "learning_rate": 5.246520405756075e-06, "loss": 0.0, "step": 3600 }, { "epoch": 22.99, "learning_rate": 5.199339466855391e-06, "loss": 0.0, "step": 3610 }, { "epoch": 23.0, "eval_accuracy": 0.99, "eval_loss": 0.06991100311279297, "eval_runtime": 3.4089, "eval_samples_per_second": 117.34, "eval_steps_per_second": 14.667, "step": 3611 }, { "epoch": 23.06, "learning_rate": 5.152158527954707e-06, "loss": 0.0001, "step": 3620 }, { "epoch": 23.12, "learning_rate": 5.104977589054022e-06, "loss": 0.0, "step": 3630 }, { "epoch": 23.18, "learning_rate": 5.057796650153338e-06, "loss": 0.0, "step": 3640 }, { "epoch": 23.25, "learning_rate": 5.0106157112526545e-06, "loss": 0.0, "step": 3650 }, { "epoch": 23.31, "learning_rate": 4.96343477235197e-06, "loss": 0.0, "step": 3660 }, { "epoch": 23.38, "learning_rate": 4.916253833451286e-06, "loss": 0.0, "step": 3670 }, { "epoch": 23.44, "learning_rate": 4.869072894550603e-06, "loss": 0.0001, "step": 3680 }, { "epoch": 23.5, "learning_rate": 4.8218919556499175e-06, "loss": 0.0, "step": 3690 }, { "epoch": 23.57, "learning_rate": 4.774711016749234e-06, "loss": 0.0, "step": 3700 }, { "epoch": 23.63, "learning_rate": 4.72753007784855e-06, "loss": 0.0, "step": 3710 }, { "epoch": 23.69, "learning_rate": 4.6803491389478655e-06, "loss": 0.0, "step": 3720 }, { "epoch": 23.76, "learning_rate": 4.633168200047181e-06, "loss": 0.0, "step": 3730 }, { "epoch": 23.82, "learning_rate": 4.585987261146497e-06, "loss": 0.0, "step": 3740 }, { "epoch": 23.89, "learning_rate": 4.538806322245813e-06, "loss": 0.0001, "step": 3750 }, { "epoch": 23.95, "learning_rate": 4.491625383345129e-06, "loss": 0.0, "step": 3760 }, { "epoch": 24.0, "eval_accuracy": 0.99, "eval_loss": 0.0699341669678688, "eval_runtime": 3.5673, "eval_samples_per_second": 112.131, "eval_steps_per_second": 14.016, "step": 3768 }, { "epoch": 24.01, "learning_rate": 4.444444444444444e-06, "loss": 0.0, "step": 3770 }, { "epoch": 24.08, "learning_rate": 4.397263505543761e-06, "loss": 0.0001, "step": 3780 }, { "epoch": 24.14, "learning_rate": 4.3500825666430765e-06, "loss": 0.0, "step": 3790 }, { "epoch": 24.2, "learning_rate": 4.302901627742392e-06, "loss": 0.0, "step": 3800 }, { "epoch": 24.27, "learning_rate": 4.255720688841709e-06, "loss": 0.0, "step": 3810 }, { "epoch": 24.33, "learning_rate": 4.2085397499410245e-06, "loss": 0.0, "step": 3820 }, { "epoch": 24.39, "learning_rate": 4.16135881104034e-06, "loss": 0.0, "step": 3830 }, { "epoch": 24.46, "learning_rate": 4.114177872139656e-06, "loss": 0.0, "step": 3840 }, { "epoch": 24.52, "learning_rate": 4.066996933238972e-06, "loss": 0.0, "step": 3850 }, { "epoch": 24.59, "learning_rate": 4.019815994338287e-06, "loss": 0.0, "step": 3860 }, { "epoch": 24.65, "learning_rate": 3.972635055437604e-06, "loss": 0.0, "step": 3870 }, { "epoch": 24.71, "learning_rate": 3.925454116536919e-06, "loss": 0.0001, "step": 3880 }, { "epoch": 24.78, "learning_rate": 3.8782731776362355e-06, "loss": 0.0, "step": 3890 }, { "epoch": 24.84, "learning_rate": 3.831092238735551e-06, "loss": 0.0, "step": 3900 }, { "epoch": 24.9, "learning_rate": 3.783911299834867e-06, "loss": 0.0, "step": 3910 }, { "epoch": 24.97, "learning_rate": 3.7367303609341826e-06, "loss": 0.0, "step": 3920 }, { "epoch": 25.0, "eval_accuracy": 0.99, "eval_loss": 0.07017427682876587, "eval_runtime": 3.7177, "eval_samples_per_second": 107.593, "eval_steps_per_second": 13.449, "step": 3925 }, { "epoch": 25.03, "learning_rate": 3.689549422033499e-06, "loss": 0.0, "step": 3930 }, { "epoch": 25.1, "learning_rate": 3.642368483132815e-06, "loss": 0.0001, "step": 3940 }, { "epoch": 25.16, "learning_rate": 3.5951875442321303e-06, "loss": 0.0, "step": 3950 }, { "epoch": 25.22, "learning_rate": 3.5480066053314464e-06, "loss": 0.0, "step": 3960 }, { "epoch": 25.29, "learning_rate": 3.500825666430762e-06, "loss": 0.0, "step": 3970 }, { "epoch": 25.35, "learning_rate": 3.4536447275300783e-06, "loss": 0.0, "step": 3980 }, { "epoch": 25.41, "learning_rate": 3.406463788629394e-06, "loss": 0.0, "step": 3990 }, { "epoch": 25.48, "learning_rate": 3.35928284972871e-06, "loss": 0.0, "step": 4000 }, { "epoch": 25.54, "learning_rate": 3.3121019108280255e-06, "loss": 0.0, "step": 4010 }, { "epoch": 25.61, "learning_rate": 3.2649209719273416e-06, "loss": 0.0, "step": 4020 }, { "epoch": 25.67, "learning_rate": 3.2177400330266574e-06, "loss": 0.0, "step": 4030 }, { "epoch": 25.73, "learning_rate": 3.1705590941259735e-06, "loss": 0.0, "step": 4040 }, { "epoch": 25.8, "learning_rate": 3.123378155225289e-06, "loss": 0.0, "step": 4050 }, { "epoch": 25.86, "learning_rate": 3.076197216324605e-06, "loss": 0.0, "step": 4060 }, { "epoch": 25.92, "learning_rate": 3.029016277423921e-06, "loss": 0.0, "step": 4070 }, { "epoch": 25.99, "learning_rate": 2.981835338523237e-06, "loss": 0.0, "step": 4080 }, { "epoch": 26.0, "eval_accuracy": 0.99, "eval_loss": 0.07053221017122269, "eval_runtime": 3.3931, "eval_samples_per_second": 117.885, "eval_steps_per_second": 14.736, "step": 4082 }, { "epoch": 26.05, "learning_rate": 2.934654399622553e-06, "loss": 0.0, "step": 4090 }, { "epoch": 26.11, "learning_rate": 2.8874734607218683e-06, "loss": 0.0, "step": 4100 }, { "epoch": 26.18, "learning_rate": 2.8402925218211845e-06, "loss": 0.0, "step": 4110 }, { "epoch": 26.24, "learning_rate": 2.7931115829205002e-06, "loss": 0.0, "step": 4120 }, { "epoch": 26.31, "learning_rate": 2.7459306440198164e-06, "loss": 0.0, "step": 4130 }, { "epoch": 26.37, "learning_rate": 2.698749705119132e-06, "loss": 0.0, "step": 4140 }, { "epoch": 26.43, "learning_rate": 2.6515687662184483e-06, "loss": 0.0, "step": 4150 }, { "epoch": 26.5, "learning_rate": 2.6043878273177636e-06, "loss": 0.0, "step": 4160 }, { "epoch": 26.56, "learning_rate": 2.5572068884170797e-06, "loss": 0.0, "step": 4170 }, { "epoch": 26.62, "learning_rate": 2.5100259495163954e-06, "loss": 0.0, "step": 4180 }, { "epoch": 26.69, "learning_rate": 2.4628450106157116e-06, "loss": 0.0, "step": 4190 }, { "epoch": 26.75, "learning_rate": 2.4156640717150273e-06, "loss": 0.0, "step": 4200 }, { "epoch": 26.82, "learning_rate": 2.368483132814343e-06, "loss": 0.0, "step": 4210 }, { "epoch": 26.88, "learning_rate": 2.321302193913659e-06, "loss": 0.0, "step": 4220 }, { "epoch": 26.94, "learning_rate": 2.274121255012975e-06, "loss": 0.0, "step": 4230 }, { "epoch": 27.0, "eval_accuracy": 0.99, "eval_loss": 0.0709276795387268, "eval_runtime": 3.8712, "eval_samples_per_second": 103.328, "eval_steps_per_second": 12.916, "step": 4239 }, { "epoch": 27.01, "learning_rate": 2.2269403161122907e-06, "loss": 0.0, "step": 4240 }, { "epoch": 27.07, "learning_rate": 2.1797593772116064e-06, "loss": 0.0, "step": 4250 }, { "epoch": 27.13, "learning_rate": 2.1325784383109226e-06, "loss": 0.0, "step": 4260 }, { "epoch": 27.2, "learning_rate": 2.0853974994102387e-06, "loss": 0.0, "step": 4270 }, { "epoch": 27.26, "learning_rate": 2.0382165605095544e-06, "loss": 0.0, "step": 4280 }, { "epoch": 27.32, "learning_rate": 1.99103562160887e-06, "loss": 0.0, "step": 4290 }, { "epoch": 27.39, "learning_rate": 1.9438546827081863e-06, "loss": 0.0, "step": 4300 }, { "epoch": 27.45, "learning_rate": 1.896673743807502e-06, "loss": 0.0, "step": 4310 }, { "epoch": 27.52, "learning_rate": 1.8494928049068178e-06, "loss": 0.0, "step": 4320 }, { "epoch": 27.58, "learning_rate": 1.8023118660061337e-06, "loss": 0.0, "step": 4330 }, { "epoch": 27.64, "learning_rate": 1.7551309271054495e-06, "loss": 0.0, "step": 4340 }, { "epoch": 27.71, "learning_rate": 1.7079499882047654e-06, "loss": 0.0, "step": 4350 }, { "epoch": 27.77, "learning_rate": 1.6607690493040811e-06, "loss": 0.0, "step": 4360 }, { "epoch": 27.83, "learning_rate": 1.613588110403397e-06, "loss": 0.0, "step": 4370 }, { "epoch": 27.9, "learning_rate": 1.566407171502713e-06, "loss": 0.0, "step": 4380 }, { "epoch": 27.96, "learning_rate": 1.5192262326020287e-06, "loss": 0.0001, "step": 4390 }, { "epoch": 28.0, "eval_accuracy": 0.99, "eval_loss": 0.07099998742341995, "eval_runtime": 3.495, "eval_samples_per_second": 114.449, "eval_steps_per_second": 14.306, "step": 4396 }, { "epoch": 28.03, "learning_rate": 1.472045293701345e-06, "loss": 0.0, "step": 4400 }, { "epoch": 28.09, "learning_rate": 1.4248643548006608e-06, "loss": 0.0, "step": 4410 }, { "epoch": 28.15, "learning_rate": 1.3776834158999766e-06, "loss": 0.0, "step": 4420 }, { "epoch": 28.22, "learning_rate": 1.3305024769992925e-06, "loss": 0.0, "step": 4430 }, { "epoch": 28.28, "learning_rate": 1.2833215380986082e-06, "loss": 0.0, "step": 4440 }, { "epoch": 28.34, "learning_rate": 1.2361405991979242e-06, "loss": 0.0, "step": 4450 }, { "epoch": 28.41, "learning_rate": 1.1889596602972401e-06, "loss": 0.0, "step": 4460 }, { "epoch": 28.47, "learning_rate": 1.1417787213965559e-06, "loss": 0.0, "step": 4470 }, { "epoch": 28.54, "learning_rate": 1.0945977824958718e-06, "loss": 0.0, "step": 4480 }, { "epoch": 28.6, "learning_rate": 1.0474168435951875e-06, "loss": 0.0, "step": 4490 }, { "epoch": 28.66, "learning_rate": 1.0002359046945035e-06, "loss": 0.0, "step": 4500 }, { "epoch": 28.73, "learning_rate": 9.530549657938194e-07, "loss": 0.0, "step": 4510 }, { "epoch": 28.79, "learning_rate": 9.058740268931352e-07, "loss": 0.0, "step": 4520 }, { "epoch": 28.85, "learning_rate": 8.586930879924511e-07, "loss": 0.0, "step": 4530 }, { "epoch": 28.92, "learning_rate": 8.11512149091767e-07, "loss": 0.0, "step": 4540 }, { "epoch": 28.98, "learning_rate": 7.643312101910829e-07, "loss": 0.0, "step": 4550 }, { "epoch": 29.0, "eval_accuracy": 0.99, "eval_loss": 0.07124363631010056, "eval_runtime": 3.4536, "eval_samples_per_second": 115.823, "eval_steps_per_second": 14.478, "step": 4553 }, { "epoch": 29.04, "learning_rate": 7.171502712903987e-07, "loss": 0.0, "step": 4560 }, { "epoch": 29.11, "learning_rate": 6.699693323897146e-07, "loss": 0.0, "step": 4570 }, { "epoch": 29.17, "learning_rate": 6.227883934890305e-07, "loss": 0.0, "step": 4580 }, { "epoch": 29.24, "learning_rate": 5.756074545883464e-07, "loss": 0.0, "step": 4590 }, { "epoch": 29.3, "learning_rate": 5.284265156876623e-07, "loss": 0.0, "step": 4600 }, { "epoch": 29.36, "learning_rate": 4.812455767869781e-07, "loss": 0.0, "step": 4610 }, { "epoch": 29.43, "learning_rate": 4.3406463788629403e-07, "loss": 0.0, "step": 4620 }, { "epoch": 29.49, "learning_rate": 3.8688369898560987e-07, "loss": 0.0, "step": 4630 }, { "epoch": 29.55, "learning_rate": 3.397027600849257e-07, "loss": 0.0, "step": 4640 }, { "epoch": 29.62, "learning_rate": 2.925218211842416e-07, "loss": 0.0, "step": 4650 }, { "epoch": 29.68, "learning_rate": 2.4534088228355743e-07, "loss": 0.0, "step": 4660 }, { "epoch": 29.75, "learning_rate": 1.9815994338287335e-07, "loss": 0.0, "step": 4670 }, { "epoch": 29.81, "learning_rate": 1.509790044821892e-07, "loss": 0.0, "step": 4680 }, { "epoch": 29.87, "learning_rate": 1.0379806558150507e-07, "loss": 0.0, "step": 4690 }, { "epoch": 29.94, "learning_rate": 5.661712668082095e-08, "loss": 0.0, "step": 4700 }, { "epoch": 30.0, "learning_rate": 9.436187780136825e-09, "loss": 0.0, "step": 4710 }, { "epoch": 30.0, "eval_accuracy": 0.99, "eval_loss": 0.0712626725435257, "eval_runtime": 3.9175, "eval_samples_per_second": 102.106, "eval_steps_per_second": 12.763, "step": 4710 }, { "epoch": 30.0, "step": 4710, "total_flos": 2.32480136060928e+19, "train_loss": 0.06245271290856119, "train_runtime": 9025.7434, "train_samples_per_second": 33.238, "train_steps_per_second": 0.522 } ], "max_steps": 4710, "num_train_epochs": 30, "total_flos": 2.32480136060928e+19, "trial_name": null, "trial_params": null }