|
--- |
|
library_name: transformers |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: poison-distill |
|
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. --> |
|
|
|
# poison-distill |
|
|
|
This model is a fine-tuned version of [](https://huggingface.co/) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: -113.4181 |
|
- Accuracy: 0.6917 |
|
|
|
## 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: 5e-05 |
|
- train_batch_size: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments |
|
- lr_scheduler_type: linear |
|
- num_epochs: 50 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:| |
|
| -0.7533 | 1.0 | 130 | -7.9549 | 0.5263 | |
|
| -8.2193 | 2.0 | 260 | -15.5418 | 0.4662 | |
|
| -14.3197 | 3.0 | 390 | -32.2167 | 0.4737 | |
|
| -18.5547 | 4.0 | 520 | -18.9202 | 0.5489 | |
|
| -22.6905 | 5.0 | 650 | -55.1682 | 0.4361 | |
|
| -27.5336 | 6.0 | 780 | -32.4679 | 0.3459 | |
|
| -29.5975 | 7.0 | 910 | -48.1715 | 0.3985 | |
|
| -34.1837 | 8.0 | 1040 | -67.7293 | 0.6165 | |
|
| -37.6123 | 9.0 | 1170 | -52.1341 | 0.4662 | |
|
| -40.7694 | 10.0 | 1300 | -49.0945 | 0.6767 | |
|
| -43.3691 | 11.0 | 1430 | -37.0478 | 0.5489 | |
|
| -47.6433 | 12.0 | 1560 | -73.0523 | 0.4511 | |
|
| -51.0141 | 13.0 | 1690 | -110.8840 | 0.4812 | |
|
| -54.6 | 14.0 | 1820 | -81.2219 | 0.3308 | |
|
| -57.2133 | 15.0 | 1950 | -80.8684 | 0.5113 | |
|
| -58.3442 | 16.0 | 2080 | -66.5341 | 0.4060 | |
|
| -64.7089 | 17.0 | 2210 | -75.7059 | 0.5564 | |
|
| -64.26 | 18.0 | 2340 | -77.7801 | 0.5263 | |
|
| -67.8509 | 19.0 | 2470 | -61.1841 | 0.6316 | |
|
| -71.9371 | 20.0 | 2600 | -118.1544 | 0.5038 | |
|
| -75.9672 | 21.0 | 2730 | -179.2044 | 0.4812 | |
|
| -78.0096 | 22.0 | 2860 | -129.4854 | 0.4436 | |
|
| -80.3581 | 23.0 | 2990 | -100.0687 | 0.4286 | |
|
| -84.623 | 24.0 | 3120 | -82.5292 | 0.3835 | |
|
| -86.5363 | 25.0 | 3250 | -84.6636 | 0.4211 | |
|
| -90.8566 | 26.0 | 3380 | -96.3337 | 0.5489 | |
|
| -92.2054 | 27.0 | 3510 | -110.3293 | 0.4737 | |
|
| -97.6982 | 28.0 | 3640 | -195.6973 | 0.4135 | |
|
| -95.8944 | 29.0 | 3770 | -101.9933 | 0.3609 | |
|
| -99.491 | 30.0 | 3900 | -99.8199 | 0.6541 | |
|
| -103.0877 | 31.0 | 4030 | -94.2175 | 0.6767 | |
|
| -102.7123 | 32.0 | 4160 | -98.6300 | 0.4887 | |
|
| -105.2087 | 33.0 | 4290 | -152.7768 | 0.4962 | |
|
| -105.3795 | 34.0 | 4420 | -198.8245 | 0.5263 | |
|
| -108.9734 | 35.0 | 4550 | -105.7644 | 0.4286 | |
|
| -111.1308 | 36.0 | 4680 | -121.4677 | 0.4962 | |
|
| -115.0085 | 37.0 | 4810 | -75.3733 | 0.3083 | |
|
| -114.714 | 38.0 | 4940 | -115.4598 | 0.6617 | |
|
| -117.5734 | 39.0 | 5070 | -108.3964 | 0.4135 | |
|
| -115.1971 | 40.0 | 5200 | -123.7679 | 0.3835 | |
|
| -117.5617 | 41.0 | 5330 | -69.2224 | 0.2932 | |
|
| -118.2803 | 42.0 | 5460 | -104.5906 | 0.6541 | |
|
| -119.6297 | 43.0 | 5590 | -187.3416 | 0.5188 | |
|
| -121.6325 | 44.0 | 5720 | -221.8878 | 0.5113 | |
|
| -120.9663 | 45.0 | 5850 | -176.6644 | 0.3759 | |
|
| -122.3583 | 46.0 | 5980 | -142.5218 | 0.4361 | |
|
| -126.6614 | 47.0 | 6110 | -271.1018 | 0.4962 | |
|
| -122.1615 | 48.0 | 6240 | -240.8323 | 0.3985 | |
|
| -125.4207 | 49.0 | 6370 | -103.5760 | 0.6466 | |
|
| -127.0661 | 50.0 | 6500 | -113.2718 | 0.6842 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.46.3 |
|
- Pytorch 2.5.1+cu121 |
|
- Datasets 3.1.0 |
|
- Tokenizers 0.20.3 |
|
|
