| { | |
| "best_metric": 1.0, | |
| "best_model_checkpoint": "swin-tiny-patch4-window7-224-finetuned-car0001/checkpoint-126", | |
| "epoch": 6.958579881656805, | |
| "eval_steps": 500, | |
| "global_step": 882, | |
| "is_hyper_param_search": false, | |
| "is_local_process_zero": true, | |
| "is_world_process_zero": true, | |
| "log_history": [ | |
| { | |
| "epoch": 0.08, | |
| "grad_norm": 14.615551948547363, | |
| "learning_rate": 5.617977528089888e-06, | |
| "loss": 2.2752, | |
| "step": 10 | |
| }, | |
| { | |
| "epoch": 0.16, | |
| "grad_norm": 12.17358684539795, | |
| "learning_rate": 1.1235955056179776e-05, | |
| "loss": 2.1418, | |
| "step": 20 | |
| }, | |
| { | |
| "epoch": 0.24, | |
| "grad_norm": 24.358016967773438, | |
| "learning_rate": 1.6853932584269665e-05, | |
| "loss": 1.9448, | |
| "step": 30 | |
| }, | |
| { | |
| "epoch": 0.32, | |
| "grad_norm": 22.546871185302734, | |
| "learning_rate": 2.2471910112359552e-05, | |
| "loss": 1.6261, | |
| "step": 40 | |
| }, | |
| { | |
| "epoch": 0.39, | |
| "grad_norm": 22.2501220703125, | |
| "learning_rate": 2.8089887640449443e-05, | |
| "loss": 1.2392, | |
| "step": 50 | |
| }, | |
| { | |
| "epoch": 0.47, | |
| "grad_norm": 27.82279396057129, | |
| "learning_rate": 3.370786516853933e-05, | |
| "loss": 0.8038, | |
| "step": 60 | |
| }, | |
| { | |
| "epoch": 0.55, | |
| "grad_norm": 21.99744987487793, | |
| "learning_rate": 3.9325842696629214e-05, | |
| "loss": 0.4458, | |
| "step": 70 | |
| }, | |
| { | |
| "epoch": 0.63, | |
| "grad_norm": 7.488814830780029, | |
| "learning_rate": 4.4943820224719104e-05, | |
| "loss": 0.1849, | |
| "step": 80 | |
| }, | |
| { | |
| "epoch": 0.71, | |
| "grad_norm": 23.27031898498535, | |
| "learning_rate": 4.9936948297604034e-05, | |
| "loss": 0.1282, | |
| "step": 90 | |
| }, | |
| { | |
| "epoch": 0.79, | |
| "grad_norm": 4.6876349449157715, | |
| "learning_rate": 4.930643127364439e-05, | |
| "loss": 0.077, | |
| "step": 100 | |
| }, | |
| { | |
| "epoch": 0.87, | |
| "grad_norm": 8.07778549194336, | |
| "learning_rate": 4.8675914249684744e-05, | |
| "loss": 0.0453, | |
| "step": 110 | |
| }, | |
| { | |
| "epoch": 0.95, | |
| "grad_norm": 8.182723045349121, | |
| "learning_rate": 4.8045397225725096e-05, | |
| "loss": 0.0328, | |
| "step": 120 | |
| }, | |
| { | |
| "epoch": 0.99, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.0029437176417559385, | |
| "eval_runtime": 32.9924, | |
| "eval_samples_per_second": 54.558, | |
| "eval_steps_per_second": 1.728, | |
| "step": 126 | |
| }, | |
| { | |
| "epoch": 1.03, | |
| "grad_norm": 7.431449890136719, | |
| "learning_rate": 4.741488020176545e-05, | |
| "loss": 0.0312, | |
| "step": 130 | |
| }, | |
| { | |
| "epoch": 1.1, | |
| "grad_norm": 1.8369030952453613, | |
| "learning_rate": 4.67843631778058e-05, | |
| "loss": 0.029, | |
| "step": 140 | |
| }, | |
| { | |
| "epoch": 1.18, | |
| "grad_norm": 3.7118313312530518, | |
| "learning_rate": 4.615384615384616e-05, | |
| "loss": 0.0334, | |
| "step": 150 | |
| }, | |
| { | |
| "epoch": 1.26, | |
| "grad_norm": 2.428013324737549, | |
| "learning_rate": 4.552332912988651e-05, | |
| "loss": 0.021, | |
| "step": 160 | |
| }, | |
| { | |
| "epoch": 1.34, | |
| "grad_norm": 0.6982976794242859, | |
| "learning_rate": 4.489281210592686e-05, | |
| "loss": 0.0186, | |
| "step": 170 | |
| }, | |
| { | |
| "epoch": 1.42, | |
| "grad_norm": 0.09051941335201263, | |
| "learning_rate": 4.426229508196721e-05, | |
| "loss": 0.0234, | |
| "step": 180 | |
| }, | |
| { | |
| "epoch": 1.5, | |
| "grad_norm": 1.7318867444992065, | |
| "learning_rate": 4.3631778058007565e-05, | |
| "loss": 0.0125, | |
| "step": 190 | |
| }, | |
| { | |
| "epoch": 1.58, | |
| "grad_norm": 5.148034572601318, | |
| "learning_rate": 4.3001261034047924e-05, | |
| "loss": 0.0354, | |
| "step": 200 | |
| }, | |
| { | |
| "epoch": 1.66, | |
| "grad_norm": 5.717312812805176, | |
| "learning_rate": 4.2370744010088275e-05, | |
| "loss": 0.0255, | |
| "step": 210 | |
| }, | |
| { | |
| "epoch": 1.74, | |
| "grad_norm": 5.518301010131836, | |
| "learning_rate": 4.174022698612863e-05, | |
| "loss": 0.0191, | |
| "step": 220 | |
| }, | |
| { | |
| "epoch": 1.81, | |
| "grad_norm": 1.6274638175964355, | |
| "learning_rate": 4.110970996216898e-05, | |
| "loss": 0.0316, | |
| "step": 230 | |
| }, | |
| { | |
| "epoch": 1.89, | |
| "grad_norm": 0.7395884394645691, | |
| "learning_rate": 4.047919293820933e-05, | |
| "loss": 0.034, | |
| "step": 240 | |
| }, | |
| { | |
| "epoch": 1.97, | |
| "grad_norm": 7.8381757736206055, | |
| "learning_rate": 3.984867591424969e-05, | |
| "loss": 0.0107, | |
| "step": 250 | |
| }, | |
| { | |
| "epoch": 2.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.00012798496754840016, | |
| "eval_runtime": 32.2135, | |
| "eval_samples_per_second": 55.877, | |
| "eval_steps_per_second": 1.769, | |
| "step": 253 | |
| }, | |
| { | |
| "epoch": 2.05, | |
| "grad_norm": 2.6944656372070312, | |
| "learning_rate": 3.921815889029004e-05, | |
| "loss": 0.0197, | |
| "step": 260 | |
| }, | |
| { | |
| "epoch": 2.13, | |
| "grad_norm": 0.7401177883148193, | |
| "learning_rate": 3.858764186633039e-05, | |
| "loss": 0.0136, | |
| "step": 270 | |
| }, | |
| { | |
| "epoch": 2.21, | |
| "grad_norm": 0.7982012033462524, | |
| "learning_rate": 3.7957124842370744e-05, | |
| "loss": 0.0153, | |
| "step": 280 | |
| }, | |
| { | |
| "epoch": 2.29, | |
| "grad_norm": 2.5111560821533203, | |
| "learning_rate": 3.7326607818411096e-05, | |
| "loss": 0.0138, | |
| "step": 290 | |
| }, | |
| { | |
| "epoch": 2.37, | |
| "grad_norm": 21.724336624145508, | |
| "learning_rate": 3.6696090794451454e-05, | |
| "loss": 0.0231, | |
| "step": 300 | |
| }, | |
| { | |
| "epoch": 2.45, | |
| "grad_norm": 1.90717351436615, | |
| "learning_rate": 3.6065573770491806e-05, | |
| "loss": 0.0087, | |
| "step": 310 | |
| }, | |
| { | |
| "epoch": 2.52, | |
| "grad_norm": 8.899871826171875, | |
| "learning_rate": 3.543505674653216e-05, | |
| "loss": 0.0222, | |
| "step": 320 | |
| }, | |
| { | |
| "epoch": 2.6, | |
| "grad_norm": 0.35501202940940857, | |
| "learning_rate": 3.480453972257251e-05, | |
| "loss": 0.0101, | |
| "step": 330 | |
| }, | |
| { | |
| "epoch": 2.68, | |
| "grad_norm": 0.01468953862786293, | |
| "learning_rate": 3.417402269861286e-05, | |
| "loss": 0.0064, | |
| "step": 340 | |
| }, | |
| { | |
| "epoch": 2.76, | |
| "grad_norm": 2.1835994720458984, | |
| "learning_rate": 3.354350567465322e-05, | |
| "loss": 0.0081, | |
| "step": 350 | |
| }, | |
| { | |
| "epoch": 2.84, | |
| "grad_norm": 0.027698852121829987, | |
| "learning_rate": 3.291298865069357e-05, | |
| "loss": 0.0207, | |
| "step": 360 | |
| }, | |
| { | |
| "epoch": 2.92, | |
| "grad_norm": 1.9383903741836548, | |
| "learning_rate": 3.2282471626733923e-05, | |
| "loss": 0.007, | |
| "step": 370 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "grad_norm": 0.31910964846611023, | |
| "learning_rate": 3.1651954602774275e-05, | |
| "loss": 0.015, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 0.00020072748884558678, | |
| "eval_runtime": 32.1614, | |
| "eval_samples_per_second": 55.968, | |
| "eval_steps_per_second": 1.772, | |
| "step": 380 | |
| }, | |
| { | |
| "epoch": 3.08, | |
| "grad_norm": 0.042351339012384415, | |
| "learning_rate": 3.102143757881463e-05, | |
| "loss": 0.0072, | |
| "step": 390 | |
| }, | |
| { | |
| "epoch": 3.16, | |
| "grad_norm": 3.833139657974243, | |
| "learning_rate": 3.0390920554854985e-05, | |
| "loss": 0.0142, | |
| "step": 400 | |
| }, | |
| { | |
| "epoch": 3.23, | |
| "grad_norm": 9.079660415649414, | |
| "learning_rate": 2.9760403530895337e-05, | |
| "loss": 0.0151, | |
| "step": 410 | |
| }, | |
| { | |
| "epoch": 3.31, | |
| "grad_norm": 1.185364007949829, | |
| "learning_rate": 2.912988650693569e-05, | |
| "loss": 0.0094, | |
| "step": 420 | |
| }, | |
| { | |
| "epoch": 3.39, | |
| "grad_norm": 0.005675105378031731, | |
| "learning_rate": 2.849936948297604e-05, | |
| "loss": 0.0069, | |
| "step": 430 | |
| }, | |
| { | |
| "epoch": 3.47, | |
| "grad_norm": 0.7052080035209656, | |
| "learning_rate": 2.7868852459016392e-05, | |
| "loss": 0.0066, | |
| "step": 440 | |
| }, | |
| { | |
| "epoch": 3.55, | |
| "grad_norm": 0.4254097044467926, | |
| "learning_rate": 2.723833543505675e-05, | |
| "loss": 0.0189, | |
| "step": 450 | |
| }, | |
| { | |
| "epoch": 3.63, | |
| "grad_norm": 0.8299044370651245, | |
| "learning_rate": 2.6607818411097103e-05, | |
| "loss": 0.0045, | |
| "step": 460 | |
| }, | |
| { | |
| "epoch": 3.71, | |
| "grad_norm": 1.3662859201431274, | |
| "learning_rate": 2.5977301387137454e-05, | |
| "loss": 0.0128, | |
| "step": 470 | |
| }, | |
| { | |
| "epoch": 3.79, | |
| "grad_norm": 0.009952099993824959, | |
| "learning_rate": 2.5346784363177806e-05, | |
| "loss": 0.0086, | |
| "step": 480 | |
| }, | |
| { | |
| "epoch": 3.87, | |
| "grad_norm": 1.3820610046386719, | |
| "learning_rate": 2.471626733921816e-05, | |
| "loss": 0.0144, | |
| "step": 490 | |
| }, | |
| { | |
| "epoch": 3.94, | |
| "grad_norm": 0.028887655586004257, | |
| "learning_rate": 2.4085750315258513e-05, | |
| "loss": 0.0048, | |
| "step": 500 | |
| }, | |
| { | |
| "epoch": 4.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 5.466423317557201e-06, | |
| "eval_runtime": 32.5873, | |
| "eval_samples_per_second": 55.236, | |
| "eval_steps_per_second": 1.749, | |
| "step": 507 | |
| }, | |
| { | |
| "epoch": 4.02, | |
| "grad_norm": 0.012447195127606392, | |
| "learning_rate": 2.3455233291298868e-05, | |
| "loss": 0.0123, | |
| "step": 510 | |
| }, | |
| { | |
| "epoch": 4.1, | |
| "grad_norm": 5.536173343658447, | |
| "learning_rate": 2.282471626733922e-05, | |
| "loss": 0.0118, | |
| "step": 520 | |
| }, | |
| { | |
| "epoch": 4.18, | |
| "grad_norm": 0.02774740196764469, | |
| "learning_rate": 2.2194199243379575e-05, | |
| "loss": 0.0059, | |
| "step": 530 | |
| }, | |
| { | |
| "epoch": 4.26, | |
| "grad_norm": 0.2927396297454834, | |
| "learning_rate": 2.1563682219419927e-05, | |
| "loss": 0.0068, | |
| "step": 540 | |
| }, | |
| { | |
| "epoch": 4.34, | |
| "grad_norm": 0.0665479302406311, | |
| "learning_rate": 2.093316519546028e-05, | |
| "loss": 0.0044, | |
| "step": 550 | |
| }, | |
| { | |
| "epoch": 4.42, | |
| "grad_norm": 0.009304400533437729, | |
| "learning_rate": 2.0302648171500633e-05, | |
| "loss": 0.0142, | |
| "step": 560 | |
| }, | |
| { | |
| "epoch": 4.5, | |
| "grad_norm": 0.6083464026451111, | |
| "learning_rate": 1.9672131147540985e-05, | |
| "loss": 0.0148, | |
| "step": 570 | |
| }, | |
| { | |
| "epoch": 4.58, | |
| "grad_norm": 0.3781495690345764, | |
| "learning_rate": 1.904161412358134e-05, | |
| "loss": 0.0065, | |
| "step": 580 | |
| }, | |
| { | |
| "epoch": 4.65, | |
| "grad_norm": 0.01165520865470171, | |
| "learning_rate": 1.8411097099621692e-05, | |
| "loss": 0.0144, | |
| "step": 590 | |
| }, | |
| { | |
| "epoch": 4.73, | |
| "grad_norm": 0.023280547931790352, | |
| "learning_rate": 1.7780580075662044e-05, | |
| "loss": 0.0038, | |
| "step": 600 | |
| }, | |
| { | |
| "epoch": 4.81, | |
| "grad_norm": 0.9535312652587891, | |
| "learning_rate": 1.71500630517024e-05, | |
| "loss": 0.012, | |
| "step": 610 | |
| }, | |
| { | |
| "epoch": 4.89, | |
| "grad_norm": 0.9200620055198669, | |
| "learning_rate": 1.651954602774275e-05, | |
| "loss": 0.0102, | |
| "step": 620 | |
| }, | |
| { | |
| "epoch": 4.97, | |
| "grad_norm": 1.7359685897827148, | |
| "learning_rate": 1.5889029003783106e-05, | |
| "loss": 0.0133, | |
| "step": 630 | |
| }, | |
| { | |
| "epoch": 4.99, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 7.2136931521527e-06, | |
| "eval_runtime": 32.8061, | |
| "eval_samples_per_second": 54.868, | |
| "eval_steps_per_second": 1.737, | |
| "step": 633 | |
| }, | |
| { | |
| "epoch": 5.05, | |
| "grad_norm": 0.00256229005753994, | |
| "learning_rate": 1.5258511979823456e-05, | |
| "loss": 0.0038, | |
| "step": 640 | |
| }, | |
| { | |
| "epoch": 5.13, | |
| "grad_norm": 0.02973257750272751, | |
| "learning_rate": 1.4627994955863808e-05, | |
| "loss": 0.0085, | |
| "step": 650 | |
| }, | |
| { | |
| "epoch": 5.21, | |
| "grad_norm": 0.014365817420184612, | |
| "learning_rate": 1.3997477931904163e-05, | |
| "loss": 0.0047, | |
| "step": 660 | |
| }, | |
| { | |
| "epoch": 5.29, | |
| "grad_norm": 5.261212348937988, | |
| "learning_rate": 1.3366960907944514e-05, | |
| "loss": 0.0024, | |
| "step": 670 | |
| }, | |
| { | |
| "epoch": 5.36, | |
| "grad_norm": 1.2331159114837646, | |
| "learning_rate": 1.273644388398487e-05, | |
| "loss": 0.0078, | |
| "step": 680 | |
| }, | |
| { | |
| "epoch": 5.44, | |
| "grad_norm": 0.011407382786273956, | |
| "learning_rate": 1.2105926860025221e-05, | |
| "loss": 0.0037, | |
| "step": 690 | |
| }, | |
| { | |
| "epoch": 5.52, | |
| "grad_norm": 0.9521317481994629, | |
| "learning_rate": 1.1475409836065575e-05, | |
| "loss": 0.0099, | |
| "step": 700 | |
| }, | |
| { | |
| "epoch": 5.6, | |
| "grad_norm": 3.931731939315796, | |
| "learning_rate": 1.0844892812105928e-05, | |
| "loss": 0.0046, | |
| "step": 710 | |
| }, | |
| { | |
| "epoch": 5.68, | |
| "grad_norm": 6.321459770202637, | |
| "learning_rate": 1.021437578814628e-05, | |
| "loss": 0.006, | |
| "step": 720 | |
| }, | |
| { | |
| "epoch": 5.76, | |
| "grad_norm": 0.41888654232025146, | |
| "learning_rate": 9.583858764186633e-06, | |
| "loss": 0.0013, | |
| "step": 730 | |
| }, | |
| { | |
| "epoch": 5.84, | |
| "grad_norm": 0.03879016637802124, | |
| "learning_rate": 8.953341740226987e-06, | |
| "loss": 0.0064, | |
| "step": 740 | |
| }, | |
| { | |
| "epoch": 5.92, | |
| "grad_norm": 0.14714492857456207, | |
| "learning_rate": 8.32282471626734e-06, | |
| "loss": 0.0072, | |
| "step": 750 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "grad_norm": 0.005957466084510088, | |
| "learning_rate": 7.692307692307694e-06, | |
| "loss": 0.0032, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 6.0, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 1.4342664371724823e-06, | |
| "eval_runtime": 32.4522, | |
| "eval_samples_per_second": 55.466, | |
| "eval_steps_per_second": 1.756, | |
| "step": 760 | |
| }, | |
| { | |
| "epoch": 6.07, | |
| "grad_norm": 0.0026133994106203318, | |
| "learning_rate": 7.061790668348046e-06, | |
| "loss": 0.0069, | |
| "step": 770 | |
| }, | |
| { | |
| "epoch": 6.15, | |
| "grad_norm": 21.232322692871094, | |
| "learning_rate": 6.431273644388399e-06, | |
| "loss": 0.013, | |
| "step": 780 | |
| }, | |
| { | |
| "epoch": 6.23, | |
| "grad_norm": 0.010045124217867851, | |
| "learning_rate": 5.800756620428751e-06, | |
| "loss": 0.0047, | |
| "step": 790 | |
| }, | |
| { | |
| "epoch": 6.31, | |
| "grad_norm": 6.4604997634887695, | |
| "learning_rate": 5.170239596469105e-06, | |
| "loss": 0.0203, | |
| "step": 800 | |
| }, | |
| { | |
| "epoch": 6.39, | |
| "grad_norm": 0.6475911736488342, | |
| "learning_rate": 4.539722572509458e-06, | |
| "loss": 0.0028, | |
| "step": 810 | |
| }, | |
| { | |
| "epoch": 6.47, | |
| "grad_norm": 2.049196720123291, | |
| "learning_rate": 3.909205548549811e-06, | |
| "loss": 0.0046, | |
| "step": 820 | |
| }, | |
| { | |
| "epoch": 6.55, | |
| "grad_norm": 0.4927035868167877, | |
| "learning_rate": 3.278688524590164e-06, | |
| "loss": 0.0076, | |
| "step": 830 | |
| }, | |
| { | |
| "epoch": 6.63, | |
| "grad_norm": 3.101295232772827, | |
| "learning_rate": 2.648171500630517e-06, | |
| "loss": 0.0061, | |
| "step": 840 | |
| }, | |
| { | |
| "epoch": 6.71, | |
| "grad_norm": 0.005894433241337538, | |
| "learning_rate": 2.0176544766708702e-06, | |
| "loss": 0.0062, | |
| "step": 850 | |
| }, | |
| { | |
| "epoch": 6.79, | |
| "grad_norm": 2.8131444454193115, | |
| "learning_rate": 1.3871374527112234e-06, | |
| "loss": 0.0042, | |
| "step": 860 | |
| }, | |
| { | |
| "epoch": 6.86, | |
| "grad_norm": 0.9489508867263794, | |
| "learning_rate": 7.566204287515763e-07, | |
| "loss": 0.0061, | |
| "step": 870 | |
| }, | |
| { | |
| "epoch": 6.94, | |
| "grad_norm": 0.0012047379277646542, | |
| "learning_rate": 1.261034047919294e-07, | |
| "loss": 0.0097, | |
| "step": 880 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "eval_accuracy": 1.0, | |
| "eval_loss": 1.1116125051557901e-06, | |
| "eval_runtime": 33.6881, | |
| "eval_samples_per_second": 53.431, | |
| "eval_steps_per_second": 1.692, | |
| "step": 882 | |
| }, | |
| { | |
| "epoch": 6.96, | |
| "step": 882, | |
| "total_flos": 2.8025139821285376e+18, | |
| "train_loss": 0.13459565003810117, | |
| "train_runtime": 2248.9328, | |
| "train_samples_per_second": 50.411, | |
| "train_steps_per_second": 0.392 | |
| } | |
| ], | |
| "logging_steps": 10, | |
| "max_steps": 882, | |
| "num_input_tokens_seen": 0, | |
| "num_train_epochs": 7, | |
| "save_steps": 500, | |
| "total_flos": 2.8025139821285376e+18, | |
| "train_batch_size": 32, | |
| "trial_name": null, | |
| "trial_params": null | |
| } | |