| { | |
| "best_metric": 0.9064, | |
| "best_model_checkpoint": "resnet-18-finetuned-cifar10/checkpoint-819", | |
| "epoch": 2.9945155393053016, | |
| "eval_steps": 500, | |
| "global_step": 819, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.04, | |
| "grad_norm": 16.46918296813965, | |
| "learning_rate": 6.0975609756097564e-06, | |
| "loss": 3.2109, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.07, | |
| "grad_norm": 22.162967681884766, | |
| "learning_rate": 1.2195121951219513e-05, | |
| "loss": 2.929, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.11, | |
| "grad_norm": 18.4061336517334, | |
| "learning_rate": 1.8292682926829268e-05, | |
| "loss": 2.6339, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.15, | |
| "grad_norm": 14.292740821838379, | |
| "learning_rate": 2.4390243902439026e-05, | |
| "loss": 2.2668, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.18, | |
| "grad_norm": 15.263444900512695, | |
| "learning_rate": 3.048780487804878e-05, | |
| "loss": 1.947, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.22, | |
| "grad_norm": 18.372751235961914, | |
| "learning_rate": 3.6585365853658535e-05, | |
| "loss": 1.7001, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.26, | |
| "grad_norm": 14.107364654541016, | |
| "learning_rate": 4.26829268292683e-05, | |
| "loss": 1.4365, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.29, | |
| "grad_norm": 11.996474266052246, | |
| "learning_rate": 4.878048780487805e-05, | |
| "loss": 1.3653, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.33, | |
| "grad_norm": 13.601606369018555, | |
| "learning_rate": 4.94572591587517e-05, | |
| "loss": 1.2311, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.37, | |
| "grad_norm": 12.406356811523438, | |
| "learning_rate": 4.877883310719132e-05, | |
| "loss": 1.2409, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.4, | |
| "grad_norm": 10.42392349243164, | |
| "learning_rate": 4.810040705563094e-05, | |
| "loss": 1.1009, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.44, | |
| "grad_norm": 11.649271011352539, | |
| "learning_rate": 4.742198100407056e-05, | |
| "loss": 1.1132, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.48, | |
| "grad_norm": 10.590466499328613, | |
| "learning_rate": 4.674355495251018e-05, | |
| "loss": 1.0677, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 0.51, | |
| "grad_norm": 10.409879684448242, | |
| "learning_rate": 4.60651289009498e-05, | |
| "loss": 1.0262, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 11.766679763793945, | |
| "learning_rate": 4.5386702849389416e-05, | |
| "loss": 1.0399, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 0.59, | |
| "grad_norm": 10.652596473693848, | |
| "learning_rate": 4.470827679782904e-05, | |
| "loss": 1.0203, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 0.62, | |
| "grad_norm": 10.809771537780762, | |
| "learning_rate": 4.402985074626866e-05, | |
| "loss": 0.9633, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 0.66, | |
| "grad_norm": 12.764290809631348, | |
| "learning_rate": 4.335142469470828e-05, | |
| "loss": 0.9355, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 0.69, | |
| "grad_norm": 10.757213592529297, | |
| "learning_rate": 4.26729986431479e-05, | |
| "loss": 0.9337, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 0.73, | |
| "grad_norm": 11.284212112426758, | |
| "learning_rate": 4.199457259158752e-05, | |
| "loss": 0.9314, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 0.77, | |
| "grad_norm": 11.635641098022461, | |
| "learning_rate": 4.131614654002714e-05, | |
| "loss": 0.8976, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 0.8, | |
| "grad_norm": 10.869036674499512, | |
| "learning_rate": 4.063772048846676e-05, | |
| "loss": 0.9065, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 0.84, | |
| "grad_norm": 11.703560829162598, | |
| "learning_rate": 3.995929443690638e-05, | |
| "loss": 0.8444, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 0.88, | |
| "grad_norm": 10.33545207977295, | |
| "learning_rate": 3.9280868385345995e-05, | |
| "loss": 0.825, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 0.91, | |
| "grad_norm": 7.844155311584473, | |
| "learning_rate": 3.860244233378562e-05, | |
| "loss": 0.8369, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 9.158990859985352, | |
| "learning_rate": 3.792401628222524e-05, | |
| "loss": 0.8422, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "grad_norm": 9.740525245666504, | |
| "learning_rate": 3.724559023066486e-05, | |
| "loss": 0.8502, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 1.0, | |
| "eval_accuracy": 0.8632, | |
| "eval_loss": 0.4063829481601715, | |
| "eval_runtime": 123.7138, | |
| "eval_samples_per_second": 121.248, | |
| "eval_steps_per_second": 3.791, | |
| "step": 273 | |
| }, | |
| { | |
| "epoch": 1.02, | |
| "grad_norm": 7.927298069000244, | |
| "learning_rate": 3.656716417910448e-05, | |
| "loss": 0.847, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 1.06, | |
| "grad_norm": 9.951263427734375, | |
| "learning_rate": 3.58887381275441e-05, | |
| "loss": 0.8189, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 7.741189479827881, | |
| "learning_rate": 3.521031207598372e-05, | |
| "loss": 0.804, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 1.13, | |
| "grad_norm": 8.110347747802734, | |
| "learning_rate": 3.453188602442334e-05, | |
| "loss": 0.7829, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 1.17, | |
| "grad_norm": 8.17768669128418, | |
| "learning_rate": 3.385345997286296e-05, | |
| "loss": 0.8124, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 1.21, | |
| "grad_norm": 8.95181655883789, | |
| "learning_rate": 3.3175033921302575e-05, | |
| "loss": 0.7392, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 1.24, | |
| "grad_norm": 9.440756797790527, | |
| "learning_rate": 3.24966078697422e-05, | |
| "loss": 0.8488, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 1.28, | |
| "grad_norm": 8.526937484741211, | |
| "learning_rate": 3.181818181818182e-05, | |
| "loss": 0.7525, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 1.32, | |
| "grad_norm": 8.15047836303711, | |
| "learning_rate": 3.113975576662144e-05, | |
| "loss": 0.7966, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 1.35, | |
| "grad_norm": 8.909910202026367, | |
| "learning_rate": 3.046132971506106e-05, | |
| "loss": 0.7682, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 1.39, | |
| "grad_norm": 9.869742393493652, | |
| "learning_rate": 2.9782903663500678e-05, | |
| "loss": 0.7143, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 1.43, | |
| "grad_norm": 9.056320190429688, | |
| "learning_rate": 2.91044776119403e-05, | |
| "loss": 0.7467, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 1.46, | |
| "grad_norm": 8.215534210205078, | |
| "learning_rate": 2.842605156037992e-05, | |
| "loss": 0.7589, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 8.56048583984375, | |
| "learning_rate": 2.7747625508819542e-05, | |
| "loss": 0.786, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 1.54, | |
| "grad_norm": 7.839065074920654, | |
| "learning_rate": 2.7069199457259158e-05, | |
| "loss": 0.7033, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 1.57, | |
| "grad_norm": 8.194784164428711, | |
| "learning_rate": 2.639077340569878e-05, | |
| "loss": 0.6867, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 1.61, | |
| "grad_norm": 7.5709547996521, | |
| "learning_rate": 2.57123473541384e-05, | |
| "loss": 0.7555, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 1.65, | |
| "grad_norm": 8.456563949584961, | |
| "learning_rate": 2.5033921302578023e-05, | |
| "loss": 0.6939, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 1.68, | |
| "grad_norm": 8.453351020812988, | |
| "learning_rate": 2.4355495251017642e-05, | |
| "loss": 0.6965, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 1.72, | |
| "grad_norm": 7.400697231292725, | |
| "learning_rate": 2.367706919945726e-05, | |
| "loss": 0.7067, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 1.76, | |
| "grad_norm": 8.554298400878906, | |
| "learning_rate": 2.299864314789688e-05, | |
| "loss": 0.712, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 1.79, | |
| "grad_norm": 8.703104972839355, | |
| "learning_rate": 2.2320217096336503e-05, | |
| "loss": 0.7176, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 1.83, | |
| "grad_norm": 8.274211883544922, | |
| "learning_rate": 2.164179104477612e-05, | |
| "loss": 0.7214, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 1.86, | |
| "grad_norm": 10.009110450744629, | |
| "learning_rate": 2.0963364993215738e-05, | |
| "loss": 0.7518, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 1.9, | |
| "grad_norm": 9.217292785644531, | |
| "learning_rate": 2.028493894165536e-05, | |
| "loss": 0.7125, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 1.94, | |
| "grad_norm": 7.580327987670898, | |
| "learning_rate": 1.960651289009498e-05, | |
| "loss": 0.6731, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 7.33704137802124, | |
| "learning_rate": 1.89280868385346e-05, | |
| "loss": 0.6924, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 0.8956, | |
| "eval_loss": 0.3149263858795166, | |
| "eval_runtime": 128.5203, | |
| "eval_samples_per_second": 116.713, | |
| "eval_steps_per_second": 3.649, | |
| "step": 547 | |
| }, | |
| { | |
| "epoch": 2.01, | |
| "grad_norm": 8.008723258972168, | |
| "learning_rate": 1.824966078697422e-05, | |
| "loss": 0.7104, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 6.283710479736328, | |
| "learning_rate": 1.757123473541384e-05, | |
| "loss": 0.6671, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 2.08, | |
| "grad_norm": 7.36352014541626, | |
| "learning_rate": 1.689280868385346e-05, | |
| "loss": 0.6464, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 2.12, | |
| "grad_norm": 8.875202178955078, | |
| "learning_rate": 1.6214382632293083e-05, | |
| "loss": 0.6495, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 2.16, | |
| "grad_norm": 8.145440101623535, | |
| "learning_rate": 1.55359565807327e-05, | |
| "loss": 0.6609, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 2.19, | |
| "grad_norm": 8.108186721801758, | |
| "learning_rate": 1.485753052917232e-05, | |
| "loss": 0.7062, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 2.23, | |
| "grad_norm": 7.363587856292725, | |
| "learning_rate": 1.417910447761194e-05, | |
| "loss": 0.6395, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 2.27, | |
| "grad_norm": 7.521917343139648, | |
| "learning_rate": 1.3500678426051561e-05, | |
| "loss": 0.657, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 2.3, | |
| "grad_norm": 7.513711452484131, | |
| "learning_rate": 1.282225237449118e-05, | |
| "loss": 0.6638, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 2.34, | |
| "grad_norm": 8.112990379333496, | |
| "learning_rate": 1.2143826322930801e-05, | |
| "loss": 0.6968, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 2.38, | |
| "grad_norm": 7.811177730560303, | |
| "learning_rate": 1.1465400271370422e-05, | |
| "loss": 0.7007, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 2.41, | |
| "grad_norm": 7.438766956329346, | |
| "learning_rate": 1.0786974219810041e-05, | |
| "loss": 0.6458, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 9.478456497192383, | |
| "learning_rate": 1.010854816824966e-05, | |
| "loss": 0.6534, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 2.49, | |
| "grad_norm": 7.470586776733398, | |
| "learning_rate": 9.430122116689281e-06, | |
| "loss": 0.6625, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 7.3036956787109375, | |
| "learning_rate": 8.751696065128902e-06, | |
| "loss": 0.6581, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 2.56, | |
| "grad_norm": 7.530010223388672, | |
| "learning_rate": 8.073270013568522e-06, | |
| "loss": 0.6214, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 6.2767333984375, | |
| "learning_rate": 7.394843962008141e-06, | |
| "loss": 0.6136, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 2.63, | |
| "grad_norm": 8.302584648132324, | |
| "learning_rate": 6.716417910447762e-06, | |
| "loss": 0.6215, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 2.67, | |
| "grad_norm": 7.744215965270996, | |
| "learning_rate": 6.037991858887382e-06, | |
| "loss": 0.6382, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 2.71, | |
| "grad_norm": 7.538488864898682, | |
| "learning_rate": 5.359565807327002e-06, | |
| "loss": 0.6239, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 2.74, | |
| "grad_norm": 7.9057488441467285, | |
| "learning_rate": 4.681139755766622e-06, | |
| "loss": 0.6413, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 2.78, | |
| "grad_norm": 6.213993549346924, | |
| "learning_rate": 4.002713704206242e-06, | |
| "loss": 0.6153, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 2.82, | |
| "grad_norm": 6.704429626464844, | |
| "learning_rate": 3.324287652645862e-06, | |
| "loss": 0.6502, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 2.85, | |
| "grad_norm": 10.968541145324707, | |
| "learning_rate": 2.645861601085482e-06, | |
| "loss": 0.6439, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 2.89, | |
| "grad_norm": 8.878872871398926, | |
| "learning_rate": 1.967435549525102e-06, | |
| "loss": 0.6751, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 2.93, | |
| "grad_norm": 8.374876022338867, | |
| "learning_rate": 1.289009497964722e-06, | |
| "loss": 0.6586, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 2.96, | |
| "grad_norm": 8.15294075012207, | |
| "learning_rate": 6.10583446404342e-07, | |
| "loss": 0.6714, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "eval_accuracy": 0.9064, | |
| "eval_loss": 0.2777732014656067, | |
| "eval_runtime": 127.5667, | |
| "eval_samples_per_second": 117.586, | |
| "eval_steps_per_second": 3.677, | |
| "step": 819 | |
| }, | |
| { | |
| "epoch": 2.99, | |
| "step": 819, | |
| "total_flos": 1.0585264325663785e+18, | |
| "train_loss": 0.9103804560371371, | |
| "train_runtime": 753.233, | |
| "train_samples_per_second": 139.399, | |
| "train_steps_per_second": 1.087 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 819, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 3, | |
| "save_steps": 500, | |
| "total_flos": 1.0585264325663785e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |