<!-- 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-base_rvl_tobacco_crl_allv2
This model is a fine-tuned version of jordyvl/vit-base_rvl-cdip on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4914
- Accuracy: 0.905
- Brier Loss: 0.1511
- Nll: 0.7302
- F1 Micro: 0.905
- F1 Macro: 0.9056
- Ece: 0.1839
- Aurc: 0.0184
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: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- gradient_accumulation_steps: 16
- total_train_batch_size: 256
- 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 | 0.96 | 3 | 2.3674 | 0.04 | 0.9050 | 9.6086 | 0.04 | 0.0438 | 0.1520 | 0.9678 |
No log | 1.96 | 6 | 2.3484 | 0.05 | 0.9004 | 8.5656 | 0.0500 | 0.0549 | 0.1567 | 0.9600 |
No log | 2.96 | 9 | 2.3106 | 0.095 | 0.8923 | 6.8919 | 0.095 | 0.0853 | 0.1832 | 0.9122 |
No log | 3.96 | 12 | 2.2538 | 0.27 | 0.8801 | 5.5263 | 0.27 | 0.1666 | 0.2921 | 0.7417 |
No log | 4.96 | 15 | 2.1898 | 0.39 | 0.8630 | 3.9890 | 0.39 | 0.2376 | 0.3516 | 0.4066 |
No log | 5.96 | 18 | 2.1124 | 0.49 | 0.8400 | 2.4772 | 0.49 | 0.3221 | 0.4366 | 0.2273 |
No log | 6.96 | 21 | 2.0178 | 0.65 | 0.8115 | 2.0300 | 0.65 | 0.5251 | 0.5447 | 0.1261 |
No log | 7.96 | 24 | 1.9053 | 0.74 | 0.7769 | 1.7062 | 0.74 | 0.6400 | 0.5939 | 0.0731 |
No log | 8.96 | 27 | 1.7938 | 0.78 | 0.7383 | 1.3250 | 0.78 | 0.6962 | 0.6183 | 0.0523 |
No log | 9.96 | 30 | 1.6845 | 0.825 | 0.6966 | 1.1128 | 0.825 | 0.7679 | 0.6216 | 0.0419 |
No log | 10.96 | 33 | 1.5726 | 0.835 | 0.6507 | 0.9279 | 0.835 | 0.7902 | 0.5892 | 0.0413 |
No log | 11.96 | 36 | 1.4703 | 0.855 | 0.6058 | 0.8610 | 0.855 | 0.8168 | 0.5789 | 0.0388 |
No log | 12.96 | 39 | 1.3783 | 0.88 | 0.5647 | 0.7009 | 0.88 | 0.8462 | 0.5745 | 0.0333 |
No log | 13.96 | 42 | 1.2986 | 0.885 | 0.5277 | 0.6657 | 0.885 | 0.8617 | 0.5458 | 0.0319 |
No log | 14.96 | 45 | 1.2260 | 0.91 | 0.4930 | 0.6487 | 0.91 | 0.8999 | 0.5400 | 0.0284 |
No log | 15.96 | 48 | 1.1579 | 0.91 | 0.4596 | 0.6307 | 0.91 | 0.8999 | 0.5268 | 0.0261 |
No log | 16.96 | 51 | 1.0942 | 0.915 | 0.4281 | 0.5573 | 0.915 | 0.9011 | 0.4938 | 0.0242 |
No log | 17.96 | 54 | 1.0383 | 0.915 | 0.4004 | 0.6216 | 0.915 | 0.9011 | 0.4780 | 0.0228 |
No log | 18.96 | 57 | 0.9877 | 0.92 | 0.3742 | 0.6119 | 0.92 | 0.9077 | 0.4502 | 0.0212 |
No log | 19.96 | 60 | 0.9380 | 0.925 | 0.3495 | 0.5968 | 0.925 | 0.9167 | 0.4332 | 0.0195 |
No log | 20.96 | 63 | 0.8961 | 0.92 | 0.3280 | 0.5971 | 0.92 | 0.9130 | 0.4097 | 0.0188 |
No log | 21.96 | 66 | 0.8581 | 0.925 | 0.3086 | 0.5969 | 0.925 | 0.9169 | 0.4032 | 0.0190 |
No log | 22.96 | 69 | 0.8218 | 0.92 | 0.2901 | 0.5901 | 0.92 | 0.9130 | 0.3764 | 0.0186 |
No log | 23.96 | 72 | 0.7899 | 0.92 | 0.2741 | 0.5890 | 0.92 | 0.9130 | 0.3736 | 0.0181 |
No log | 24.96 | 75 | 0.7627 | 0.93 | 0.2603 | 0.5946 | 0.93 | 0.9245 | 0.3535 | 0.0175 |
No log | 25.96 | 78 | 0.7346 | 0.93 | 0.2470 | 0.5839 | 0.93 | 0.9245 | 0.3440 | 0.0165 |
No log | 26.96 | 81 | 0.7114 | 0.93 | 0.2361 | 0.5823 | 0.93 | 0.9245 | 0.3261 | 0.0169 |
No log | 27.96 | 84 | 0.6914 | 0.93 | 0.2265 | 0.5842 | 0.93 | 0.9245 | 0.3151 | 0.0170 |
No log | 28.96 | 87 | 0.6730 | 0.925 | 0.2179 | 0.5840 | 0.925 | 0.9216 | 0.3048 | 0.0170 |
No log | 29.96 | 90 | 0.6564 | 0.92 | 0.2107 | 0.6412 | 0.92 | 0.9159 | 0.2932 | 0.0171 |
No log | 30.96 | 93 | 0.6416 | 0.92 | 0.2043 | 0.6918 | 0.92 | 0.9159 | 0.2884 | 0.0172 |
No log | 31.96 | 96 | 0.6287 | 0.92 | 0.1986 | 0.6890 | 0.92 | 0.9140 | 0.2801 | 0.0169 |
No log | 32.96 | 99 | 0.6165 | 0.92 | 0.1934 | 0.6360 | 0.92 | 0.9150 | 0.2650 | 0.0170 |
No log | 33.96 | 102 | 0.6064 | 0.92 | 0.1892 | 0.6819 | 0.92 | 0.9150 | 0.2585 | 0.0170 |
No log | 34.96 | 105 | 0.5970 | 0.92 | 0.1855 | 0.6806 | 0.92 | 0.9150 | 0.2525 | 0.0169 |
No log | 35.96 | 108 | 0.5880 | 0.915 | 0.1818 | 0.6778 | 0.915 | 0.9121 | 0.2494 | 0.0171 |
No log | 36.96 | 111 | 0.5801 | 0.91 | 0.1789 | 0.6766 | 0.91 | 0.9089 | 0.2427 | 0.0170 |
No log | 37.96 | 114 | 0.5732 | 0.91 | 0.1764 | 0.6757 | 0.91 | 0.9089 | 0.2376 | 0.0166 |
No log | 38.96 | 117 | 0.5672 | 0.91 | 0.1739 | 0.6741 | 0.91 | 0.9089 | 0.2337 | 0.0165 |
No log | 39.96 | 120 | 0.5617 | 0.91 | 0.1717 | 0.6732 | 0.91 | 0.9089 | 0.2305 | 0.0165 |
No log | 40.96 | 123 | 0.5562 | 0.91 | 0.1698 | 0.6723 | 0.91 | 0.9089 | 0.2266 | 0.0164 |
No log | 41.96 | 126 | 0.5516 | 0.91 | 0.1682 | 0.6724 | 0.91 | 0.9089 | 0.2230 | 0.0165 |
No log | 42.96 | 129 | 0.5470 | 0.91 | 0.1666 | 0.6719 | 0.91 | 0.9089 | 0.2201 | 0.0165 |
No log | 43.96 | 132 | 0.5428 | 0.91 | 0.1653 | 0.6718 | 0.91 | 0.9089 | 0.2174 | 0.0167 |
No log | 44.96 | 135 | 0.5400 | 0.91 | 0.1642 | 0.6726 | 0.91 | 0.9089 | 0.2145 | 0.0167 |
No log | 45.96 | 138 | 0.5373 | 0.91 | 0.1632 | 0.6738 | 0.91 | 0.9089 | 0.2216 | 0.0168 |
No log | 46.96 | 141 | 0.5347 | 0.91 | 0.1621 | 0.6744 | 0.91 | 0.9089 | 0.2095 | 0.0167 |
No log | 47.96 | 144 | 0.5318 | 0.91 | 0.1613 | 0.6775 | 0.91 | 0.9089 | 0.2074 | 0.0166 |
No log | 48.96 | 147 | 0.5294 | 0.91 | 0.1606 | 0.7357 | 0.91 | 0.9089 | 0.2054 | 0.0166 |
No log | 49.96 | 150 | 0.5270 | 0.91 | 0.1598 | 0.7350 | 0.91 | 0.9089 | 0.2029 | 0.0166 |
No log | 50.96 | 153 | 0.5240 | 0.91 | 0.1593 | 0.7347 | 0.91 | 0.9089 | 0.1920 | 0.0166 |
No log | 51.96 | 156 | 0.5218 | 0.91 | 0.1586 | 0.7340 | 0.91 | 0.9089 | 0.1902 | 0.0167 |
No log | 52.96 | 159 | 0.5200 | 0.91 | 0.1581 | 0.7336 | 0.91 | 0.9089 | 0.1925 | 0.0168 |
No log | 53.96 | 162 | 0.5178 | 0.905 | 0.1576 | 0.7333 | 0.905 | 0.9056 | 0.1903 | 0.0171 |
No log | 54.96 | 165 | 0.5156 | 0.905 | 0.1571 | 0.7323 | 0.905 | 0.9056 | 0.1888 | 0.0172 |
No log | 55.96 | 168 | 0.5138 | 0.905 | 0.1565 | 0.7318 | 0.905 | 0.9056 | 0.1971 | 0.0172 |
No log | 56.96 | 171 | 0.5122 | 0.905 | 0.1561 | 0.7313 | 0.905 | 0.9056 | 0.1771 | 0.0173 |
No log | 57.96 | 174 | 0.5106 | 0.905 | 0.1557 | 0.7308 | 0.905 | 0.9056 | 0.1947 | 0.0175 |
No log | 58.96 | 177 | 0.5093 | 0.905 | 0.1554 | 0.7308 | 0.905 | 0.9056 | 0.2033 | 0.0176 |
No log | 59.96 | 180 | 0.5080 | 0.905 | 0.1550 | 0.7307 | 0.905 | 0.9056 | 0.2021 | 0.0175 |
No log | 60.96 | 183 | 0.5068 | 0.905 | 0.1547 | 0.7305 | 0.905 | 0.9056 | 0.1914 | 0.0176 |
No log | 61.96 | 186 | 0.5055 | 0.905 | 0.1545 | 0.7301 | 0.905 | 0.9056 | 0.1813 | 0.0178 |
No log | 62.96 | 189 | 0.5039 | 0.905 | 0.1541 | 0.7292 | 0.905 | 0.9056 | 0.1803 | 0.0178 |
No log | 63.96 | 192 | 0.5029 | 0.905 | 0.1539 | 0.7297 | 0.905 | 0.9056 | 0.1893 | 0.0179 |
No log | 64.96 | 195 | 0.5018 | 0.905 | 0.1537 | 0.7292 | 0.905 | 0.9056 | 0.1883 | 0.0178 |
No log | 65.96 | 198 | 0.5007 | 0.905 | 0.1534 | 0.7286 | 0.905 | 0.9056 | 0.1873 | 0.0178 |
No log | 66.96 | 201 | 0.5002 | 0.905 | 0.1534 | 0.7292 | 0.905 | 0.9056 | 0.1869 | 0.0179 |
No log | 67.96 | 204 | 0.4993 | 0.905 | 0.1532 | 0.7287 | 0.905 | 0.9056 | 0.1771 | 0.0179 |
No log | 68.96 | 207 | 0.4989 | 0.905 | 0.1530 | 0.7287 | 0.905 | 0.9056 | 0.1668 | 0.0179 |
No log | 69.96 | 210 | 0.4988 | 0.905 | 0.1530 | 0.7290 | 0.905 | 0.9056 | 0.1664 | 0.0179 |
No log | 70.96 | 213 | 0.4983 | 0.905 | 0.1528 | 0.7289 | 0.905 | 0.9056 | 0.1658 | 0.0179 |
No log | 71.96 | 216 | 0.4977 | 0.905 | 0.1526 | 0.7290 | 0.905 | 0.9056 | 0.1653 | 0.0180 |
No log | 72.96 | 219 | 0.4971 | 0.905 | 0.1525 | 0.7289 | 0.905 | 0.9056 | 0.1732 | 0.0181 |
No log | 73.96 | 222 | 0.4966 | 0.905 | 0.1525 | 0.7291 | 0.905 | 0.9056 | 0.1731 | 0.0181 |
No log | 74.96 | 225 | 0.4961 | 0.905 | 0.1523 | 0.7291 | 0.905 | 0.9056 | 0.1726 | 0.0182 |
No log | 75.96 | 228 | 0.4955 | 0.905 | 0.1522 | 0.7292 | 0.905 | 0.9056 | 0.1718 | 0.0181 |
No log | 76.96 | 231 | 0.4950 | 0.905 | 0.1521 | 0.7290 | 0.905 | 0.9056 | 0.1714 | 0.0181 |
No log | 77.96 | 234 | 0.4945 | 0.905 | 0.1520 | 0.7289 | 0.905 | 0.9056 | 0.1712 | 0.0181 |
No log | 78.96 | 237 | 0.4946 | 0.905 | 0.1519 | 0.7290 | 0.905 | 0.9056 | 0.1709 | 0.0181 |
No log | 79.96 | 240 | 0.4944 | 0.905 | 0.1518 | 0.7292 | 0.905 | 0.9056 | 0.1705 | 0.0181 |
No log | 80.96 | 243 | 0.4942 | 0.905 | 0.1518 | 0.7293 | 0.905 | 0.9056 | 0.1732 | 0.0182 |
No log | 81.96 | 246 | 0.4939 | 0.905 | 0.1517 | 0.7293 | 0.905 | 0.9056 | 0.1698 | 0.0182 |
No log | 82.96 | 249 | 0.4937 | 0.905 | 0.1517 | 0.7299 | 0.905 | 0.9056 | 0.1695 | 0.0182 |
No log | 83.96 | 252 | 0.4932 | 0.905 | 0.1515 | 0.7294 | 0.905 | 0.9056 | 0.1698 | 0.0182 |
No log | 84.96 | 255 | 0.4929 | 0.905 | 0.1514 | 0.7292 | 0.905 | 0.9056 | 0.1695 | 0.0182 |
No log | 85.96 | 258 | 0.4928 | 0.905 | 0.1515 | 0.7297 | 0.905 | 0.9056 | 0.1693 | 0.0182 |
No log | 86.96 | 261 | 0.4925 | 0.905 | 0.1514 | 0.7296 | 0.905 | 0.9056 | 0.1690 | 0.0182 |
No log | 87.96 | 264 | 0.4923 | 0.905 | 0.1513 | 0.7294 | 0.905 | 0.9056 | 0.1688 | 0.0183 |
No log | 88.96 | 267 | 0.4921 | 0.905 | 0.1512 | 0.7296 | 0.905 | 0.9056 | 0.1687 | 0.0182 |
No log | 89.96 | 270 | 0.4920 | 0.905 | 0.1512 | 0.7299 | 0.905 | 0.9056 | 0.1685 | 0.0182 |
No log | 90.96 | 273 | 0.4918 | 0.905 | 0.1512 | 0.7298 | 0.905 | 0.9056 | 0.1683 | 0.0182 |
No log | 91.96 | 276 | 0.4917 | 0.905 | 0.1512 | 0.7298 | 0.905 | 0.9056 | 0.1775 | 0.0183 |
No log | 92.96 | 279 | 0.4916 | 0.905 | 0.1512 | 0.7299 | 0.905 | 0.9056 | 0.1773 | 0.0183 |
No log | 93.96 | 282 | 0.4915 | 0.905 | 0.1511 | 0.7300 | 0.905 | 0.9056 | 0.1843 | 0.0184 |
No log | 94.96 | 285 | 0.4915 | 0.905 | 0.1511 | 0.7300 | 0.905 | 0.9056 | 0.1842 | 0.0184 |
No log | 95.96 | 288 | 0.4915 | 0.905 | 0.1511 | 0.7302 | 0.905 | 0.9056 | 0.1841 | 0.0184 |
No log | 96.96 | 291 | 0.4914 | 0.905 | 0.1511 | 0.7302 | 0.905 | 0.9056 | 0.1840 | 0.0184 |
No log | 97.96 | 294 | 0.4914 | 0.905 | 0.1511 | 0.7302 | 0.905 | 0.9056 | 0.1840 | 0.0184 |
No log | 98.96 | 297 | 0.4914 | 0.905 | 0.1511 | 0.7302 | 0.905 | 0.9056 | 0.1840 | 0.0184 |
No log | 99.96 | 300 | 0.4914 | 0.905 | 0.1511 | 0.7302 | 0.905 | 0.9056 | 0.1839 | 0.0184 |
Framework versions
- Transformers 4.26.1
- Pytorch 1.13.1.post200
- Datasets 2.9.0
- Tokenizers 0.13.2