generated_from_trainer

<!-- This model card has been generated automatically according to the information the Trainer had access to. You should probably proofread and complete it, then remove this comment. -->

vit-small_rvl_cdip_100_examples_per_class_kd_CEKD_t1.5_a0.7

This model is a fine-tuned version of WinKawaks/vit-small-patch16-224 on the None dataset. It achieves the following results on the evaluation set:

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

Training results

Training Loss Epoch Step Validation Loss Accuracy Brier Loss Nll F1 Micro F1 Macro Ece Aurc
No log 1.0 25 3.2176 0.1275 0.9297 15.5568 0.1275 0.1255 0.1544 0.8595
No log 2.0 50 2.4392 0.405 0.7503 9.6083 0.405 0.3723 0.1816 0.3640
No log 3.0 75 1.9211 0.5025 0.6287 5.6023 0.5025 0.4930 0.1991 0.2451
No log 4.0 100 1.7474 0.5375 0.5956 4.5712 0.5375 0.5387 0.1677 0.2244
No log 5.0 125 1.7107 0.535 0.6051 4.3431 0.535 0.5180 0.1796 0.2269
No log 6.0 150 1.7144 0.545 0.5988 3.6699 0.545 0.5455 0.1918 0.2253
No log 7.0 175 1.9096 0.5625 0.6262 4.6856 0.5625 0.5459 0.1966 0.2362
No log 8.0 200 1.6325 0.575 0.5815 3.9279 0.575 0.5705 0.1893 0.2026
No log 9.0 225 1.8268 0.56 0.6088 4.5140 0.56 0.5482 0.1976 0.2213
No log 10.0 250 1.9253 0.5575 0.6493 4.2860 0.5575 0.5427 0.2286 0.2445
No log 11.0 275 1.6941 0.5725 0.5940 3.9317 0.5725 0.5827 0.2019 0.2232
No log 12.0 300 1.8197 0.5575 0.6138 4.7928 0.5575 0.5476 0.2079 0.2240
No log 13.0 325 1.8958 0.54 0.6508 4.2978 0.54 0.5338 0.2379 0.2357
No log 14.0 350 1.8939 0.535 0.6522 4.5557 0.535 0.5143 0.2324 0.2350
No log 15.0 375 1.8018 0.585 0.6042 4.4728 0.585 0.5829 0.2205 0.2182
No log 16.0 400 1.7645 0.5975 0.5978 3.9939 0.5975 0.5992 0.2130 0.1927
No log 17.0 425 1.6392 0.5925 0.5842 3.6783 0.5925 0.6039 0.1986 0.2017
No log 18.0 450 1.6124 0.5875 0.5761 4.0535 0.5875 0.5721 0.2060 0.1792
No log 19.0 475 1.7517 0.585 0.6102 3.9076 0.585 0.5786 0.2082 0.2071
0.6436 20.0 500 1.7467 0.5575 0.6166 3.5052 0.5575 0.5476 0.2252 0.2247
0.6436 21.0 525 1.6719 0.5825 0.5745 4.1235 0.5825 0.5877 0.1831 0.1723
0.6436 22.0 550 1.4222 0.605 0.5237 3.2051 0.605 0.6083 0.1813 0.1559
0.6436 23.0 575 1.6436 0.595 0.5701 4.3949 0.595 0.5834 0.1921 0.1901
0.6436 24.0 600 1.4244 0.6075 0.5197 3.3207 0.6075 0.6100 0.1548 0.1616
0.6436 25.0 625 1.4567 0.6075 0.5356 3.5288 0.6075 0.6107 0.1768 0.1652
0.6436 26.0 650 1.5889 0.595 0.5587 4.1521 0.595 0.5907 0.1943 0.1768
0.6436 27.0 675 1.4828 0.5725 0.5532 3.4259 0.5725 0.5720 0.2125 0.1803
0.6436 28.0 700 1.4671 0.5975 0.5509 3.2612 0.5975 0.6006 0.1983 0.1797
0.6436 29.0 725 1.4049 0.6225 0.5273 3.3136 0.6225 0.6237 0.1995 0.1600
0.6436 30.0 750 1.4039 0.6175 0.5208 3.2588 0.6175 0.6063 0.1770 0.1534
0.6436 31.0 775 1.4333 0.6 0.5378 3.6417 0.6 0.5995 0.1899 0.1632
0.6436 32.0 800 1.3311 0.64 0.5032 3.0056 0.64 0.6394 0.1699 0.1476
0.6436 33.0 825 1.3361 0.61 0.5079 3.2304 0.61 0.6123 0.1536 0.1517
0.6436 34.0 850 1.2984 0.64 0.4982 3.1446 0.64 0.6444 0.1636 0.1424
0.6436 35.0 875 1.3153 0.6275 0.4995 3.0722 0.6275 0.6288 0.1634 0.1486
0.6436 36.0 900 1.2773 0.6375 0.4880 2.7136 0.6375 0.6422 0.1606 0.1411
0.6436 37.0 925 1.2881 0.64 0.4946 3.0452 0.64 0.6437 0.1732 0.1440
0.6436 38.0 950 1.2609 0.64 0.4824 2.7407 0.64 0.6430 0.1485 0.1424
0.6436 39.0 975 1.2685 0.645 0.4869 2.7203 0.645 0.6484 0.1680 0.1398
0.0861 40.0 1000 1.2546 0.635 0.4808 2.7042 0.635 0.6356 0.1669 0.1416
0.0861 41.0 1025 1.2599 0.6425 0.4858 2.6880 0.6425 0.6457 0.1539 0.1387
0.0861 42.0 1050 1.2413 0.635 0.4783 2.8343 0.635 0.6361 0.1679 0.1369
0.0861 43.0 1075 1.2670 0.6325 0.4901 2.8366 0.6325 0.6337 0.1501 0.1399
0.0861 44.0 1100 1.2793 0.63 0.4919 3.1711 0.63 0.6309 0.1672 0.1465
0.0861 45.0 1125 1.2555 0.635 0.4844 2.9284 0.635 0.6379 0.1791 0.1401
0.0861 46.0 1150 1.2491 0.635 0.4806 2.8475 0.635 0.6358 0.1611 0.1392
0.0861 47.0 1175 1.2533 0.6325 0.4837 2.8229 0.6325 0.6352 0.1623 0.1378
0.0861 48.0 1200 1.2602 0.635 0.4857 2.9963 0.635 0.6368 0.1535 0.1426
0.0861 49.0 1225 1.2598 0.635 0.4848 2.8569 0.635 0.6370 0.1718 0.1389
0.0861 50.0 1250 1.2577 0.6225 0.4839 2.8645 0.6225 0.6237 0.1678 0.1420
0.0861 51.0 1275 1.2547 0.63 0.4817 2.8344 0.63 0.6314 0.1721 0.1399
0.0861 52.0 1300 1.2525 0.64 0.4819 2.7720 0.64 0.6411 0.1567 0.1378
0.0861 53.0 1325 1.2627 0.6325 0.4854 2.9202 0.6325 0.6337 0.1688 0.1406
0.0861 54.0 1350 1.2565 0.63 0.4836 2.8392 0.63 0.6320 0.1612 0.1404
0.0861 55.0 1375 1.2514 0.6325 0.4813 2.9887 0.6325 0.6343 0.1652 0.1386
0.0861 56.0 1400 1.2541 0.6275 0.4822 2.9067 0.6275 0.6296 0.1649 0.1401
0.0861 57.0 1425 1.2529 0.64 0.4810 2.9166 0.64 0.6432 0.1765 0.1372
0.0861 58.0 1450 1.2464 0.6275 0.4799 2.9713 0.6275 0.6291 0.1653 0.1401
0.0861 59.0 1475 1.2576 0.63 0.4826 2.9124 0.63 0.6323 0.1557 0.1397
0.0496 60.0 1500 1.2494 0.63 0.4804 2.8355 0.63 0.6317 0.1672 0.1390
0.0496 61.0 1525 1.2496 0.6325 0.4803 2.9091 0.6325 0.6352 0.1510 0.1383
0.0496 62.0 1550 1.2592 0.6375 0.4838 2.8980 0.6375 0.6384 0.1758 0.1398
0.0496 63.0 1575 1.2504 0.63 0.4806 2.9843 0.63 0.6316 0.1691 0.1391
0.0496 64.0 1600 1.2528 0.6325 0.4810 2.9045 0.6325 0.6349 0.1737 0.1388
0.0496 65.0 1625 1.2589 0.6425 0.4833 2.9817 0.6425 0.6447 0.1719 0.1380
0.0496 66.0 1650 1.2531 0.63 0.4811 2.9027 0.63 0.6321 0.1751 0.1391
0.0496 67.0 1675 1.2520 0.635 0.4808 2.9794 0.635 0.6379 0.1715 0.1378
0.0496 68.0 1700 1.2543 0.64 0.4815 2.9771 0.64 0.6420 0.1562 0.1380
0.0496 69.0 1725 1.2538 0.6325 0.4808 2.9080 0.6325 0.6345 0.1681 0.1385
0.0496 70.0 1750 1.2543 0.6325 0.4813 2.9102 0.6325 0.6347 0.1725 0.1390
0.0496 71.0 1775 1.2534 0.6325 0.4809 2.9778 0.6325 0.6353 0.1495 0.1385
0.0496 72.0 1800 1.2539 0.6375 0.4809 2.9024 0.6375 0.6394 0.1588 0.1381
0.0496 73.0 1825 1.2531 0.635 0.4806 2.9812 0.635 0.6378 0.1552 0.1380
0.0496 74.0 1850 1.2531 0.635 0.4805 2.9783 0.635 0.6377 0.1700 0.1380
0.0496 75.0 1875 1.2533 0.6375 0.4809 2.9772 0.6375 0.6400 0.1645 0.1372
0.0496 76.0 1900 1.2539 0.6375 0.4808 2.9777 0.6375 0.6393 0.1675 0.1376
0.0496 77.0 1925 1.2537 0.635 0.4808 2.9832 0.635 0.6375 0.1648 0.1381
0.0496 78.0 1950 1.2539 0.6375 0.4807 2.9769 0.6375 0.6394 0.1636 0.1374
0.0496 79.0 1975 1.2534 0.6375 0.4805 2.9796 0.6375 0.6399 0.1599 0.1375
0.048 80.0 2000 1.2537 0.6375 0.4806 3.0539 0.6375 0.6399 0.1657 0.1375
0.048 81.0 2025 1.2535 0.6375 0.4805 3.0534 0.6375 0.6399 0.1728 0.1375
0.048 82.0 2050 1.2539 0.6375 0.4806 2.9831 0.6375 0.6393 0.1674 0.1375
0.048 83.0 2075 1.2542 0.6375 0.4807 3.0538 0.6375 0.6399 0.1674 0.1375
0.048 84.0 2100 1.2539 0.6375 0.4805 3.0531 0.6375 0.6394 0.1564 0.1375
0.048 85.0 2125 1.2542 0.6375 0.4806 3.0531 0.6375 0.6393 0.1676 0.1376
0.048 86.0 2150 1.2541 0.6375 0.4806 3.0527 0.6375 0.6399 0.1691 0.1375
0.048 87.0 2175 1.2542 0.6375 0.4805 3.0525 0.6375 0.6394 0.1677 0.1376
0.048 88.0 2200 1.2542 0.6375 0.4806 3.0525 0.6375 0.6393 0.1651 0.1375
0.048 89.0 2225 1.2543 0.6375 0.4805 3.0525 0.6375 0.6394 0.1601 0.1375
0.048 90.0 2250 1.2543 0.6375 0.4805 3.0521 0.6375 0.6394 0.1661 0.1375
0.048 91.0 2275 1.2541 0.6375 0.4805 3.0521 0.6375 0.6394 0.1665 0.1376
0.048 92.0 2300 1.2542 0.6375 0.4805 3.0521 0.6375 0.6394 0.1638 0.1375
0.048 93.0 2325 1.2544 0.6375 0.4805 3.0518 0.6375 0.6394 0.1671 0.1376
0.048 94.0 2350 1.2543 0.6375 0.4805 3.0519 0.6375 0.6394 0.1601 0.1376
0.048 95.0 2375 1.2544 0.6375 0.4805 3.0518 0.6375 0.6394 0.1638 0.1376
0.048 96.0 2400 1.2544 0.6375 0.4805 3.0518 0.6375 0.6394 0.1638 0.1376
0.048 97.0 2425 1.2544 0.6375 0.4805 3.0517 0.6375 0.6394 0.1655 0.1376
0.048 98.0 2450 1.2544 0.6375 0.4805 3.0517 0.6375 0.6394 0.1638 0.1376
0.048 99.0 2475 1.2544 0.6375 0.4805 3.0517 0.6375 0.6394 0.1654 0.1376
0.0478 100.0 2500 1.2544 0.6375 0.4805 3.0517 0.6375 0.6394 0.1654 0.1376

Framework versions