lombardata's picture
Evaluation on the test set completed on 2024_09_05.
ac02534 verified
|
raw
history blame
13.1 kB
---
license: apache-2.0
base_model: facebook/dinov2-base
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: DinoVdeau-base-2024_09_03-batch-size32_epochs150_freeze
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. -->
# DinoVdeau-base-2024_09_03-batch-size32_epochs150_freeze
This model is a fine-tuned version of [facebook/dinov2-base](https://huggingface.co/facebook/dinov2-base) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1260
- F1 Micro: 0.8131
- F1 Macro: 0.6976
- Roc Auc: 0.8760
- Accuracy: 0.3014
- Learning Rate: 0.0000
## 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: 0.001
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 150
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss | F1 Micro | F1 Macro | Roc Auc | Accuracy | Rate |
|:-------------:|:-----:|:-----:|:---------------:|:--------:|:--------:|:-------:|:--------:|:------:|
| No log | 1.0 | 273 | 0.1752 | 0.7311 | 0.5105 | 0.8187 | 0.2079 | 0.001 |
| 0.2857 | 2.0 | 546 | 0.1578 | 0.7583 | 0.5498 | 0.8363 | 0.2349 | 0.001 |
| 0.2857 | 3.0 | 819 | 0.1516 | 0.7722 | 0.6037 | 0.8505 | 0.2315 | 0.001 |
| 0.1764 | 4.0 | 1092 | 0.1522 | 0.7650 | 0.6140 | 0.8387 | 0.2422 | 0.001 |
| 0.1764 | 5.0 | 1365 | 0.1484 | 0.7720 | 0.6162 | 0.8403 | 0.2422 | 0.001 |
| 0.1677 | 6.0 | 1638 | 0.1482 | 0.7750 | 0.6052 | 0.8435 | 0.2561 | 0.001 |
| 0.1677 | 7.0 | 1911 | 0.1486 | 0.7729 | 0.6177 | 0.8431 | 0.2419 | 0.001 |
| 0.1652 | 8.0 | 2184 | 0.1486 | 0.7767 | 0.6172 | 0.8485 | 0.2512 | 0.001 |
| 0.1652 | 9.0 | 2457 | 0.1483 | 0.7805 | 0.6366 | 0.8570 | 0.2512 | 0.001 |
| 0.1617 | 10.0 | 2730 | 0.1503 | 0.7683 | 0.6081 | 0.8352 | 0.2453 | 0.001 |
| 0.1615 | 11.0 | 3003 | 0.1441 | 0.7757 | 0.6200 | 0.8409 | 0.2609 | 0.001 |
| 0.1615 | 12.0 | 3276 | 0.1487 | 0.7815 | 0.6299 | 0.8543 | 0.2495 | 0.001 |
| 0.1614 | 13.0 | 3549 | 0.1490 | 0.7779 | 0.6242 | 0.8446 | 0.2519 | 0.001 |
| 0.1614 | 14.0 | 3822 | 0.1434 | 0.7826 | 0.6379 | 0.8475 | 0.2606 | 0.001 |
| 0.1599 | 15.0 | 4095 | 0.1435 | 0.7874 | 0.6397 | 0.8552 | 0.2554 | 0.001 |
| 0.1599 | 16.0 | 4368 | 0.1439 | 0.7793 | 0.6344 | 0.8464 | 0.2568 | 0.001 |
| 0.1589 | 17.0 | 4641 | 0.1448 | 0.7878 | 0.6422 | 0.8596 | 0.2543 | 0.001 |
| 0.1589 | 18.0 | 4914 | 0.1440 | 0.7865 | 0.6417 | 0.8552 | 0.2568 | 0.001 |
| 0.1604 | 19.0 | 5187 | 0.1420 | 0.7864 | 0.6318 | 0.8550 | 0.2540 | 0.001 |
| 0.1604 | 20.0 | 5460 | 0.1409 | 0.7869 | 0.6409 | 0.8522 | 0.2588 | 0.001 |
| 0.1586 | 21.0 | 5733 | 0.1425 | 0.7865 | 0.6413 | 0.8561 | 0.2620 | 0.001 |
| 0.1587 | 22.0 | 6006 | 0.1538 | 0.7854 | 0.6371 | 0.8608 | 0.2370 | 0.001 |
| 0.1587 | 23.0 | 6279 | 0.1419 | 0.7842 | 0.6390 | 0.8497 | 0.2557 | 0.001 |
| 0.1592 | 24.0 | 6552 | 0.1414 | 0.7870 | 0.6459 | 0.8561 | 0.2599 | 0.001 |
| 0.1592 | 25.0 | 6825 | 0.1399 | 0.7868 | 0.6263 | 0.8523 | 0.2685 | 0.001 |
| 0.1586 | 26.0 | 7098 | 0.1465 | 0.7847 | 0.6238 | 0.8561 | 0.2592 | 0.001 |
| 0.1586 | 27.0 | 7371 | 0.1551 | 0.7720 | 0.6344 | 0.8433 | 0.2380 | 0.001 |
| 0.16 | 28.0 | 7644 | 0.1443 | 0.7891 | 0.6430 | 0.8550 | 0.2616 | 0.001 |
| 0.16 | 29.0 | 7917 | 0.1428 | 0.7874 | 0.6416 | 0.8565 | 0.2568 | 0.001 |
| 0.1589 | 30.0 | 8190 | 0.1416 | 0.7799 | 0.6308 | 0.8425 | 0.2526 | 0.001 |
| 0.1589 | 31.0 | 8463 | 0.1398 | 0.7895 | 0.6431 | 0.8566 | 0.2689 | 0.001 |
| 0.1588 | 32.0 | 8736 | 0.1448 | 0.7891 | 0.6521 | 0.8601 | 0.2568 | 0.001 |
| 0.1581 | 33.0 | 9009 | 0.1404 | 0.7896 | 0.6497 | 0.8582 | 0.2640 | 0.001 |
| 0.1581 | 34.0 | 9282 | 0.1426 | 0.7871 | 0.6449 | 0.8537 | 0.2557 | 0.001 |
| 0.1578 | 35.0 | 9555 | 0.1414 | 0.7846 | 0.6428 | 0.8487 | 0.2630 | 0.001 |
| 0.1578 | 36.0 | 9828 | 0.1465 | 0.7834 | 0.6434 | 0.8484 | 0.2678 | 0.001 |
| 0.1576 | 37.0 | 10101 | 0.1380 | 0.7924 | 0.6438 | 0.8577 | 0.2668 | 0.001 |
| 0.1576 | 38.0 | 10374 | 0.1392 | 0.7892 | 0.6475 | 0.8555 | 0.2637 | 0.001 |
| 0.1556 | 39.0 | 10647 | 0.1458 | 0.7872 | 0.6592 | 0.8680 | 0.2460 | 0.001 |
| 0.1556 | 40.0 | 10920 | 0.1389 | 0.7946 | 0.6469 | 0.8660 | 0.2699 | 0.001 |
| 0.1577 | 41.0 | 11193 | 0.1402 | 0.7848 | 0.6510 | 0.8491 | 0.2616 | 0.001 |
| 0.1577 | 42.0 | 11466 | 0.1404 | 0.7928 | 0.6609 | 0.8625 | 0.2717 | 0.001 |
| 0.1576 | 43.0 | 11739 | 0.1394 | 0.7931 | 0.6427 | 0.8593 | 0.2696 | 0.001 |
| 0.1543 | 44.0 | 12012 | 0.1367 | 0.7989 | 0.6568 | 0.8632 | 0.2755 | 0.0001 |
| 0.1543 | 45.0 | 12285 | 0.1362 | 0.8018 | 0.6686 | 0.8652 | 0.2827 | 0.0001 |
| 0.1481 | 46.0 | 12558 | 0.1338 | 0.8022 | 0.6640 | 0.8656 | 0.2852 | 0.0001 |
| 0.1481 | 47.0 | 12831 | 0.1410 | 0.7999 | 0.6573 | 0.8621 | 0.2786 | 0.0001 |
| 0.1472 | 48.0 | 13104 | 0.1338 | 0.8044 | 0.6728 | 0.8675 | 0.2848 | 0.0001 |
| 0.1472 | 49.0 | 13377 | 0.1322 | 0.8058 | 0.6742 | 0.8724 | 0.2855 | 0.0001 |
| 0.1448 | 50.0 | 13650 | 0.1332 | 0.8063 | 0.6739 | 0.8703 | 0.2897 | 0.0001 |
| 0.1448 | 51.0 | 13923 | 0.1306 | 0.8063 | 0.6771 | 0.8702 | 0.2897 | 0.0001 |
| 0.1432 | 52.0 | 14196 | 0.1311 | 0.8044 | 0.6727 | 0.8654 | 0.2872 | 0.0001 |
| 0.1432 | 53.0 | 14469 | 0.1316 | 0.8071 | 0.6703 | 0.8713 | 0.2872 | 0.0001 |
| 0.1438 | 54.0 | 14742 | 0.1316 | 0.8064 | 0.6788 | 0.8688 | 0.2883 | 0.0001 |
| 0.1417 | 55.0 | 15015 | 0.1308 | 0.8061 | 0.6699 | 0.8686 | 0.2876 | 0.0001 |
| 0.1417 | 56.0 | 15288 | 0.1297 | 0.8094 | 0.6800 | 0.8744 | 0.2942 | 0.0001 |
| 0.1415 | 57.0 | 15561 | 0.1296 | 0.8087 | 0.6717 | 0.8711 | 0.2935 | 0.0001 |
| 0.1415 | 58.0 | 15834 | 0.1297 | 0.8069 | 0.6785 | 0.8708 | 0.2924 | 0.0001 |
| 0.1413 | 59.0 | 16107 | 0.1300 | 0.8087 | 0.6811 | 0.8707 | 0.2911 | 0.0001 |
| 0.1413 | 60.0 | 16380 | 0.1302 | 0.8056 | 0.6726 | 0.8658 | 0.2879 | 0.0001 |
| 0.1404 | 61.0 | 16653 | 0.1287 | 0.8096 | 0.6843 | 0.8721 | 0.2949 | 0.0001 |
| 0.1404 | 62.0 | 16926 | 0.1291 | 0.8080 | 0.6822 | 0.8690 | 0.2900 | 0.0001 |
| 0.1393 | 63.0 | 17199 | 0.1287 | 0.8076 | 0.6813 | 0.8685 | 0.2980 | 0.0001 |
| 0.1393 | 64.0 | 17472 | 0.1286 | 0.8091 | 0.6806 | 0.8722 | 0.2959 | 0.0001 |
| 0.1395 | 65.0 | 17745 | 0.1280 | 0.8093 | 0.6838 | 0.8704 | 0.2931 | 0.0001 |
| 0.1389 | 66.0 | 18018 | 0.1278 | 0.8108 | 0.6855 | 0.8744 | 0.2959 | 0.0001 |
| 0.1389 | 67.0 | 18291 | 0.1282 | 0.8098 | 0.6849 | 0.8746 | 0.2949 | 0.0001 |
| 0.1376 | 68.0 | 18564 | 0.1280 | 0.8123 | 0.6903 | 0.8771 | 0.2980 | 0.0001 |
| 0.1376 | 69.0 | 18837 | 0.1280 | 0.8105 | 0.6800 | 0.8711 | 0.2952 | 0.0001 |
| 0.1375 | 70.0 | 19110 | 0.1276 | 0.8096 | 0.6848 | 0.8709 | 0.2931 | 0.0001 |
| 0.1375 | 71.0 | 19383 | 0.1279 | 0.8073 | 0.6797 | 0.8675 | 0.2904 | 0.0001 |
| 0.1368 | 72.0 | 19656 | 0.1278 | 0.8103 | 0.6802 | 0.8719 | 0.2938 | 0.0001 |
| 0.1368 | 73.0 | 19929 | 0.1272 | 0.8091 | 0.6806 | 0.8683 | 0.2976 | 0.0001 |
| 0.137 | 74.0 | 20202 | 0.1280 | 0.8064 | 0.6777 | 0.8648 | 0.2935 | 0.0001 |
| 0.137 | 75.0 | 20475 | 0.1273 | 0.8110 | 0.6885 | 0.8731 | 0.2924 | 0.0001 |
| 0.1367 | 76.0 | 20748 | 0.1273 | 0.8089 | 0.6811 | 0.8696 | 0.2973 | 0.0001 |
| 0.1358 | 77.0 | 21021 | 0.1275 | 0.8102 | 0.6863 | 0.8739 | 0.2924 | 0.0001 |
| 0.1358 | 78.0 | 21294 | 0.1271 | 0.8122 | 0.6897 | 0.8765 | 0.2945 | 0.0001 |
| 0.1352 | 79.0 | 21567 | 0.1271 | 0.8098 | 0.6882 | 0.8697 | 0.2935 | 0.0001 |
| 0.1352 | 80.0 | 21840 | 0.1272 | 0.8124 | 0.6914 | 0.8773 | 0.2983 | 0.0001 |
| 0.1353 | 81.0 | 22113 | 0.1265 | 0.8104 | 0.6899 | 0.8716 | 0.2966 | 0.0001 |
| 0.1353 | 82.0 | 22386 | 0.1264 | 0.8105 | 0.6845 | 0.8694 | 0.2914 | 0.0001 |
| 0.1337 | 83.0 | 22659 | 0.1273 | 0.8100 | 0.6832 | 0.8701 | 0.2935 | 0.0001 |
| 0.1337 | 84.0 | 22932 | 0.1264 | 0.8124 | 0.6944 | 0.8756 | 0.2959 | 0.0001 |
| 0.1354 | 85.0 | 23205 | 0.1265 | 0.8127 | 0.6880 | 0.8750 | 0.2973 | 0.0001 |
| 0.1354 | 86.0 | 23478 | 0.1259 | 0.8136 | 0.6933 | 0.8746 | 0.2952 | 0.0001 |
| 0.1334 | 87.0 | 23751 | 0.1264 | 0.8111 | 0.6882 | 0.8738 | 0.2966 | 0.0001 |
| 0.1335 | 88.0 | 24024 | 0.1264 | 0.8127 | 0.6860 | 0.8754 | 0.2990 | 0.0001 |
| 0.1335 | 89.0 | 24297 | 0.1269 | 0.8140 | 0.6990 | 0.8792 | 0.2983 | 0.0001 |
| 0.1332 | 90.0 | 24570 | 0.1261 | 0.8155 | 0.6994 | 0.8798 | 0.2980 | 0.0001 |
| 0.1332 | 91.0 | 24843 | 0.1268 | 0.8109 | 0.6828 | 0.8728 | 0.2893 | 0.0001 |
| 0.1326 | 92.0 | 25116 | 0.1261 | 0.8124 | 0.6858 | 0.8724 | 0.2952 | 0.0001 |
| 0.1326 | 93.0 | 25389 | 0.1258 | 0.8138 | 0.6897 | 0.8759 | 0.2966 | 1e-05 |
| 0.132 | 94.0 | 25662 | 0.1268 | 0.8138 | 0.6941 | 0.8755 | 0.2976 | 1e-05 |
| 0.132 | 95.0 | 25935 | 0.1257 | 0.8134 | 0.6913 | 0.8750 | 0.2949 | 1e-05 |
| 0.1294 | 96.0 | 26208 | 0.1259 | 0.8147 | 0.6957 | 0.8763 | 0.2976 | 1e-05 |
| 0.1294 | 97.0 | 26481 | 0.1256 | 0.8126 | 0.6941 | 0.8720 | 0.2945 | 1e-05 |
| 0.1302 | 98.0 | 26754 | 0.1253 | 0.8159 | 0.6951 | 0.8785 | 0.2994 | 1e-05 |
| 0.1298 | 99.0 | 27027 | 0.1249 | 0.8142 | 0.6968 | 0.8752 | 0.2994 | 1e-05 |
| 0.1298 | 100.0 | 27300 | 0.1252 | 0.8135 | 0.6936 | 0.8732 | 0.2973 | 1e-05 |
| 0.1304 | 101.0 | 27573 | 0.1248 | 0.8149 | 0.6961 | 0.8765 | 0.2990 | 1e-05 |
| 0.1304 | 102.0 | 27846 | 0.1266 | 0.8137 | 0.6927 | 0.8738 | 0.2963 | 1e-05 |
| 0.1287 | 103.0 | 28119 | 0.1249 | 0.8146 | 0.6954 | 0.8754 | 0.2990 | 1e-05 |
| 0.1287 | 104.0 | 28392 | 0.1252 | 0.8149 | 0.6927 | 0.8770 | 0.2976 | 1e-05 |
| 0.1282 | 105.0 | 28665 | 0.1251 | 0.8152 | 0.6962 | 0.8773 | 0.2990 | 1e-05 |
| 0.1282 | 106.0 | 28938 | 0.1251 | 0.8147 | 0.6964 | 0.8770 | 0.2997 | 1e-05 |
| 0.1293 | 107.0 | 29211 | 0.1250 | 0.8145 | 0.6946 | 0.8759 | 0.2980 | 1e-05 |
| 0.1293 | 108.0 | 29484 | 0.1249 | 0.8145 | 0.6935 | 0.8751 | 0.2997 | 0.0000 |
| 0.129 | 109.0 | 29757 | 0.1253 | 0.8116 | 0.6901 | 0.8713 | 0.2952 | 0.0000 |
| 0.1293 | 110.0 | 30030 | 0.1252 | 0.8144 | 0.6949 | 0.8768 | 0.2980 | 0.0000 |
| 0.1293 | 111.0 | 30303 | 0.1250 | 0.8137 | 0.6932 | 0.8755 | 0.2983 | 0.0000 |
### Framework versions
- Transformers 4.41.1
- Pytorch 2.3.0+cu121
- Datasets 2.19.1
- Tokenizers 0.19.1