|
--- |
|
license: apache-2.0 |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: vit-tiny_rvl_cdip_100_examples_per_class_kd_CEKD_t5.0_a0.7 |
|
results: [] |
|
--- |
|
|
|
<!-- 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-tiny_rvl_cdip_100_examples_per_class_kd_CEKD_t5.0_a0.7 |
|
|
|
This model is a fine-tuned version of [WinKawaks/vit-tiny-patch16-224](https://huggingface.co/WinKawaks/vit-tiny-patch16-224) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.6224 |
|
- Accuracy: 0.555 |
|
- Brier Loss: 0.5813 |
|
- Nll: 2.4451 |
|
- F1 Micro: 0.555 |
|
- F1 Macro: 0.5481 |
|
- Ece: 0.1732 |
|
- Aurc: 0.2090 |
|
|
|
## 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: |
|
- learning_rate: 0.0001 |
|
- train_batch_size: 128 |
|
- eval_batch_size: 128 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_ratio: 0.1 |
|
- num_epochs: 100 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Brier Loss | Nll | F1 Micro | F1 Macro | Ece | Aurc | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:----------:|:------:|:--------:|:--------:|:------:|:------:| |
|
| No log | 1.0 | 7 | 4.2974 | 0.04 | 1.0725 | 7.3803 | 0.04 | 0.0315 | 0.2825 | 0.9545 | |
|
| No log | 2.0 | 14 | 3.3036 | 0.095 | 0.9476 | 5.7516 | 0.095 | 0.0792 | 0.1590 | 0.8944 | |
|
| No log | 3.0 | 21 | 2.9979 | 0.215 | 0.8918 | 5.3309 | 0.2150 | 0.1763 | 0.1514 | 0.6635 | |
|
| No log | 4.0 | 28 | 2.5846 | 0.2875 | 0.7979 | 3.5812 | 0.2875 | 0.2707 | 0.1619 | 0.4840 | |
|
| No log | 5.0 | 35 | 2.2908 | 0.3925 | 0.7162 | 3.1082 | 0.3925 | 0.3675 | 0.1724 | 0.3500 | |
|
| No log | 6.0 | 42 | 2.1582 | 0.4275 | 0.6903 | 3.3486 | 0.4275 | 0.3933 | 0.1723 | 0.3106 | |
|
| No log | 7.0 | 49 | 2.1910 | 0.445 | 0.7011 | 2.9994 | 0.445 | 0.4233 | 0.1889 | 0.3105 | |
|
| No log | 8.0 | 56 | 2.0287 | 0.485 | 0.6673 | 2.8482 | 0.485 | 0.4803 | 0.1781 | 0.2848 | |
|
| No log | 9.0 | 63 | 2.1037 | 0.4775 | 0.6684 | 2.8143 | 0.4775 | 0.4715 | 0.2010 | 0.2690 | |
|
| No log | 10.0 | 70 | 2.1168 | 0.4825 | 0.6846 | 2.8143 | 0.4825 | 0.4703 | 0.2082 | 0.2818 | |
|
| No log | 11.0 | 77 | 2.1094 | 0.495 | 0.6825 | 2.9020 | 0.495 | 0.4834 | 0.1990 | 0.2632 | |
|
| No log | 12.0 | 84 | 2.0835 | 0.48 | 0.6897 | 2.7522 | 0.48 | 0.4651 | 0.2261 | 0.2741 | |
|
| No log | 13.0 | 91 | 1.9606 | 0.505 | 0.6631 | 2.4794 | 0.505 | 0.4988 | 0.2034 | 0.2712 | |
|
| No log | 14.0 | 98 | 1.9519 | 0.4975 | 0.6567 | 2.7608 | 0.4975 | 0.4833 | 0.2042 | 0.2563 | |
|
| No log | 15.0 | 105 | 1.8794 | 0.52 | 0.6304 | 2.6588 | 0.52 | 0.5121 | 0.1814 | 0.2337 | |
|
| No log | 16.0 | 112 | 1.7934 | 0.5375 | 0.6191 | 2.5142 | 0.5375 | 0.5234 | 0.1853 | 0.2272 | |
|
| No log | 17.0 | 119 | 1.8110 | 0.5225 | 0.6242 | 2.5106 | 0.5225 | 0.5071 | 0.1918 | 0.2336 | |
|
| No log | 18.0 | 126 | 1.8027 | 0.515 | 0.6283 | 2.4142 | 0.515 | 0.4983 | 0.2020 | 0.2359 | |
|
| No log | 19.0 | 133 | 1.8123 | 0.5375 | 0.6318 | 2.5551 | 0.5375 | 0.5235 | 0.2132 | 0.2358 | |
|
| No log | 20.0 | 140 | 1.7937 | 0.5225 | 0.6292 | 2.5237 | 0.5225 | 0.5145 | 0.2070 | 0.2336 | |
|
| No log | 21.0 | 147 | 1.7272 | 0.5175 | 0.6046 | 2.4278 | 0.5175 | 0.5103 | 0.1806 | 0.2286 | |
|
| No log | 22.0 | 154 | 1.8337 | 0.5325 | 0.6396 | 2.5603 | 0.5325 | 0.5136 | 0.1884 | 0.2405 | |
|
| No log | 23.0 | 161 | 1.7416 | 0.5275 | 0.6102 | 2.5228 | 0.5275 | 0.5077 | 0.1784 | 0.2232 | |
|
| No log | 24.0 | 168 | 1.7036 | 0.55 | 0.6063 | 2.4933 | 0.55 | 0.5380 | 0.1776 | 0.2230 | |
|
| No log | 25.0 | 175 | 1.7330 | 0.545 | 0.6084 | 2.4943 | 0.545 | 0.5365 | 0.1989 | 0.2251 | |
|
| No log | 26.0 | 182 | 1.6911 | 0.55 | 0.5993 | 2.4401 | 0.55 | 0.5416 | 0.1792 | 0.2208 | |
|
| No log | 27.0 | 189 | 1.7329 | 0.5475 | 0.6162 | 2.4824 | 0.5475 | 0.5380 | 0.1830 | 0.2317 | |
|
| No log | 28.0 | 196 | 1.6890 | 0.5475 | 0.5992 | 2.4828 | 0.5475 | 0.5401 | 0.1725 | 0.2178 | |
|
| No log | 29.0 | 203 | 1.7256 | 0.5425 | 0.6124 | 2.5121 | 0.5425 | 0.5299 | 0.1765 | 0.2260 | |
|
| No log | 30.0 | 210 | 1.6854 | 0.5375 | 0.5952 | 2.4275 | 0.5375 | 0.5282 | 0.2015 | 0.2163 | |
|
| No log | 31.0 | 217 | 1.7010 | 0.5475 | 0.6030 | 2.4832 | 0.5475 | 0.5385 | 0.1862 | 0.2236 | |
|
| No log | 32.0 | 224 | 1.6840 | 0.535 | 0.5934 | 2.4512 | 0.535 | 0.5310 | 0.1794 | 0.2182 | |
|
| No log | 33.0 | 231 | 1.6808 | 0.545 | 0.6016 | 2.4424 | 0.545 | 0.5396 | 0.1828 | 0.2222 | |
|
| No log | 34.0 | 238 | 1.6965 | 0.535 | 0.6000 | 2.5453 | 0.535 | 0.5270 | 0.1846 | 0.2243 | |
|
| No log | 35.0 | 245 | 1.6650 | 0.545 | 0.5930 | 2.4901 | 0.545 | 0.5418 | 0.1706 | 0.2130 | |
|
| No log | 36.0 | 252 | 1.6494 | 0.54 | 0.5979 | 2.4011 | 0.54 | 0.5319 | 0.1731 | 0.2201 | |
|
| No log | 37.0 | 259 | 1.6738 | 0.54 | 0.5892 | 2.4632 | 0.54 | 0.5257 | 0.1781 | 0.2088 | |
|
| No log | 38.0 | 266 | 1.6502 | 0.55 | 0.5889 | 2.4733 | 0.55 | 0.5416 | 0.1957 | 0.2087 | |
|
| No log | 39.0 | 273 | 1.6539 | 0.55 | 0.5832 | 2.5105 | 0.55 | 0.5424 | 0.1584 | 0.2120 | |
|
| No log | 40.0 | 280 | 1.6399 | 0.545 | 0.5919 | 2.4450 | 0.545 | 0.5439 | 0.1877 | 0.2168 | |
|
| No log | 41.0 | 287 | 1.6968 | 0.535 | 0.6006 | 2.5310 | 0.535 | 0.5281 | 0.1895 | 0.2221 | |
|
| No log | 42.0 | 294 | 1.6430 | 0.5425 | 0.5955 | 2.4020 | 0.5425 | 0.5449 | 0.1722 | 0.2203 | |
|
| No log | 43.0 | 301 | 1.6743 | 0.535 | 0.5970 | 2.5138 | 0.535 | 0.5239 | 0.1869 | 0.2167 | |
|
| No log | 44.0 | 308 | 1.6544 | 0.5475 | 0.5949 | 2.4259 | 0.5475 | 0.5408 | 0.1651 | 0.2164 | |
|
| No log | 45.0 | 315 | 1.6763 | 0.535 | 0.5980 | 2.4379 | 0.535 | 0.5249 | 0.1868 | 0.2174 | |
|
| No log | 46.0 | 322 | 1.6509 | 0.525 | 0.5933 | 2.4351 | 0.525 | 0.5137 | 0.1861 | 0.2194 | |
|
| No log | 47.0 | 329 | 1.6530 | 0.5475 | 0.5929 | 2.4628 | 0.5475 | 0.5419 | 0.1859 | 0.2148 | |
|
| No log | 48.0 | 336 | 1.6410 | 0.555 | 0.5835 | 2.4992 | 0.555 | 0.5490 | 0.1805 | 0.2110 | |
|
| No log | 49.0 | 343 | 1.6398 | 0.5525 | 0.5857 | 2.5060 | 0.5525 | 0.5409 | 0.1706 | 0.2101 | |
|
| No log | 50.0 | 350 | 1.6343 | 0.5525 | 0.5814 | 2.4890 | 0.5525 | 0.5442 | 0.1608 | 0.2065 | |
|
| No log | 51.0 | 357 | 1.6335 | 0.5475 | 0.5846 | 2.4407 | 0.5475 | 0.5392 | 0.1720 | 0.2109 | |
|
| No log | 52.0 | 364 | 1.6309 | 0.555 | 0.5844 | 2.4944 | 0.555 | 0.5488 | 0.1697 | 0.2091 | |
|
| No log | 53.0 | 371 | 1.6308 | 0.5575 | 0.5826 | 2.4815 | 0.5575 | 0.5505 | 0.1704 | 0.2080 | |
|
| No log | 54.0 | 378 | 1.6279 | 0.56 | 0.5832 | 2.4741 | 0.56 | 0.5525 | 0.1724 | 0.2067 | |
|
| No log | 55.0 | 385 | 1.6226 | 0.55 | 0.5825 | 2.4048 | 0.55 | 0.5425 | 0.1656 | 0.2094 | |
|
| No log | 56.0 | 392 | 1.6141 | 0.555 | 0.5797 | 2.4716 | 0.555 | 0.5474 | 0.1813 | 0.2076 | |
|
| No log | 57.0 | 399 | 1.6179 | 0.56 | 0.5760 | 2.4682 | 0.56 | 0.5549 | 0.1474 | 0.2030 | |
|
| No log | 58.0 | 406 | 1.6278 | 0.56 | 0.5831 | 2.4758 | 0.56 | 0.5540 | 0.1681 | 0.2075 | |
|
| No log | 59.0 | 413 | 1.6257 | 0.5525 | 0.5817 | 2.4462 | 0.5525 | 0.5455 | 0.1648 | 0.2084 | |
|
| No log | 60.0 | 420 | 1.6306 | 0.5575 | 0.5861 | 2.5090 | 0.5575 | 0.5505 | 0.1687 | 0.2108 | |
|
| No log | 61.0 | 427 | 1.6314 | 0.555 | 0.5821 | 2.5024 | 0.555 | 0.5480 | 0.1725 | 0.2092 | |
|
| No log | 62.0 | 434 | 1.6322 | 0.545 | 0.5846 | 2.4694 | 0.545 | 0.5407 | 0.1848 | 0.2123 | |
|
| No log | 63.0 | 441 | 1.6229 | 0.5575 | 0.5829 | 2.4413 | 0.5575 | 0.5508 | 0.1698 | 0.2098 | |
|
| No log | 64.0 | 448 | 1.6187 | 0.56 | 0.5809 | 2.4420 | 0.56 | 0.5522 | 0.1848 | 0.2083 | |
|
| No log | 65.0 | 455 | 1.6160 | 0.555 | 0.5794 | 2.4349 | 0.555 | 0.5506 | 0.1761 | 0.2076 | |
|
| No log | 66.0 | 462 | 1.6254 | 0.55 | 0.5822 | 2.4752 | 0.55 | 0.5432 | 0.1629 | 0.2086 | |
|
| No log | 67.0 | 469 | 1.6259 | 0.55 | 0.5839 | 2.4425 | 0.55 | 0.5423 | 0.1769 | 0.2113 | |
|
| No log | 68.0 | 476 | 1.6245 | 0.55 | 0.5822 | 2.4382 | 0.55 | 0.5421 | 0.1736 | 0.2091 | |
|
| No log | 69.0 | 483 | 1.6243 | 0.5575 | 0.5830 | 2.4422 | 0.5575 | 0.5497 | 0.1808 | 0.2094 | |
|
| No log | 70.0 | 490 | 1.6223 | 0.5575 | 0.5810 | 2.4787 | 0.5575 | 0.5507 | 0.1556 | 0.2089 | |
|
| No log | 71.0 | 497 | 1.6204 | 0.5575 | 0.5809 | 2.4408 | 0.5575 | 0.5515 | 0.1555 | 0.2083 | |
|
| 0.3852 | 72.0 | 504 | 1.6225 | 0.55 | 0.5816 | 2.4404 | 0.55 | 0.5424 | 0.1750 | 0.2104 | |
|
| 0.3852 | 73.0 | 511 | 1.6237 | 0.55 | 0.5822 | 2.4403 | 0.55 | 0.5429 | 0.1772 | 0.2107 | |
|
| 0.3852 | 74.0 | 518 | 1.6220 | 0.55 | 0.5815 | 2.4441 | 0.55 | 0.5420 | 0.1649 | 0.2105 | |
|
| 0.3852 | 75.0 | 525 | 1.6228 | 0.5475 | 0.5818 | 2.4736 | 0.5475 | 0.5405 | 0.1882 | 0.2109 | |
|
| 0.3852 | 76.0 | 532 | 1.6224 | 0.5525 | 0.5814 | 2.4442 | 0.5525 | 0.5446 | 0.1817 | 0.2108 | |
|
| 0.3852 | 77.0 | 539 | 1.6225 | 0.5525 | 0.5815 | 2.4431 | 0.5525 | 0.5448 | 0.1798 | 0.2098 | |
|
| 0.3852 | 78.0 | 546 | 1.6213 | 0.555 | 0.5812 | 2.4417 | 0.555 | 0.5471 | 0.1680 | 0.2096 | |
|
| 0.3852 | 79.0 | 553 | 1.6208 | 0.5575 | 0.5808 | 2.4423 | 0.5575 | 0.5501 | 0.1784 | 0.2082 | |
|
| 0.3852 | 80.0 | 560 | 1.6218 | 0.5525 | 0.5811 | 2.4425 | 0.5525 | 0.5447 | 0.1683 | 0.2095 | |
|
| 0.3852 | 81.0 | 567 | 1.6225 | 0.5525 | 0.5814 | 2.4429 | 0.5525 | 0.5447 | 0.1856 | 0.2098 | |
|
| 0.3852 | 82.0 | 574 | 1.6222 | 0.5575 | 0.5812 | 2.4469 | 0.5575 | 0.5501 | 0.1953 | 0.2085 | |
|
| 0.3852 | 83.0 | 581 | 1.6219 | 0.555 | 0.5811 | 2.4442 | 0.555 | 0.5471 | 0.1940 | 0.2093 | |
|
| 0.3852 | 84.0 | 588 | 1.6220 | 0.555 | 0.5813 | 2.4443 | 0.555 | 0.5471 | 0.1867 | 0.2095 | |
|
| 0.3852 | 85.0 | 595 | 1.6223 | 0.555 | 0.5813 | 2.4446 | 0.555 | 0.5471 | 0.1885 | 0.2094 | |
|
| 0.3852 | 86.0 | 602 | 1.6222 | 0.5525 | 0.5812 | 2.4448 | 0.5525 | 0.5447 | 0.1749 | 0.2095 | |
|
| 0.3852 | 87.0 | 609 | 1.6222 | 0.555 | 0.5813 | 2.4454 | 0.555 | 0.5481 | 0.1745 | 0.2091 | |
|
| 0.3852 | 88.0 | 616 | 1.6222 | 0.5575 | 0.5813 | 2.4446 | 0.5575 | 0.5504 | 0.1767 | 0.2087 | |
|
| 0.3852 | 89.0 | 623 | 1.6222 | 0.5575 | 0.5813 | 2.4445 | 0.5575 | 0.5504 | 0.1839 | 0.2087 | |
|
| 0.3852 | 90.0 | 630 | 1.6221 | 0.555 | 0.5812 | 2.4447 | 0.555 | 0.5481 | 0.1814 | 0.2091 | |
|
| 0.3852 | 91.0 | 637 | 1.6222 | 0.5575 | 0.5813 | 2.4446 | 0.5575 | 0.5504 | 0.1790 | 0.2087 | |
|
| 0.3852 | 92.0 | 644 | 1.6222 | 0.555 | 0.5813 | 2.4447 | 0.555 | 0.5481 | 0.1755 | 0.2091 | |
|
| 0.3852 | 93.0 | 651 | 1.6223 | 0.5575 | 0.5813 | 2.4446 | 0.5575 | 0.5504 | 0.1747 | 0.2087 | |
|
| 0.3852 | 94.0 | 658 | 1.6223 | 0.5575 | 0.5813 | 2.4449 | 0.5575 | 0.5504 | 0.1747 | 0.2087 | |
|
| 0.3852 | 95.0 | 665 | 1.6223 | 0.555 | 0.5813 | 2.4448 | 0.555 | 0.5481 | 0.1732 | 0.2090 | |
|
| 0.3852 | 96.0 | 672 | 1.6224 | 0.555 | 0.5813 | 2.4451 | 0.555 | 0.5481 | 0.1720 | 0.2091 | |
|
| 0.3852 | 97.0 | 679 | 1.6223 | 0.5575 | 0.5813 | 2.4451 | 0.5575 | 0.5504 | 0.1735 | 0.2088 | |
|
| 0.3852 | 98.0 | 686 | 1.6223 | 0.5575 | 0.5813 | 2.4451 | 0.5575 | 0.5504 | 0.1747 | 0.2087 | |
|
| 0.3852 | 99.0 | 693 | 1.6224 | 0.555 | 0.5813 | 2.4451 | 0.555 | 0.5481 | 0.1732 | 0.2090 | |
|
| 0.3852 | 100.0 | 700 | 1.6224 | 0.555 | 0.5813 | 2.4451 | 0.555 | 0.5481 | 0.1732 | 0.2090 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.26.1 |
|
- Pytorch 1.13.1.post200 |
|
- Datasets 2.9.0 |
|
- Tokenizers 0.13.2 |
|
|
