|
--- |
|
license: mit |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
- precision |
|
- recall |
|
- f1 |
|
model-index: |
|
- name: roberta-hateful-meme |
|
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. --> |
|
|
|
# roberta-hateful-meme |
|
|
|
This model is a fine-tuned version of [FacebookAI/roberta-base](https://huggingface.co/FacebookAI/roberta-base) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 2.4643 |
|
- Accuracy: 0.631 |
|
- Precision: 0.6057 |
|
- Recall: 0.6081 |
|
- F1: 0.6066 |
|
|
|
## 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: 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: 30 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:| |
|
| No log | 1.0 | 282 | 0.5978 | 0.696 | 0.6714 | 0.6698 | 0.6705 | |
|
| 0.5742 | 2.0 | 564 | 0.6133 | 0.696 | 0.6691 | 0.6558 | 0.6599 | |
|
| 0.5742 | 3.0 | 846 | 0.6326 | 0.694 | 0.6681 | 0.6385 | 0.6430 | |
|
| 0.4682 | 4.0 | 1128 | 0.7503 | 0.662 | 0.6473 | 0.6564 | 0.6482 | |
|
| 0.4682 | 5.0 | 1410 | 0.8875 | 0.687 | 0.6805 | 0.6004 | 0.5918 | |
|
| 0.3907 | 6.0 | 1692 | 0.8072 | 0.651 | 0.6393 | 0.6489 | 0.6389 | |
|
| 0.3907 | 7.0 | 1974 | 1.0383 | 0.639 | 0.6205 | 0.6266 | 0.6215 | |
|
| 0.3445 | 8.0 | 2256 | 1.0678 | 0.643 | 0.6212 | 0.6257 | 0.6225 | |
|
| 0.2996 | 9.0 | 2538 | 1.1221 | 0.649 | 0.6131 | 0.6025 | 0.6046 | |
|
| 0.2996 | 10.0 | 2820 | 1.3531 | 0.646 | 0.6315 | 0.6397 | 0.6319 | |
|
| 0.2749 | 11.0 | 3102 | 1.1258 | 0.655 | 0.6186 | 0.6031 | 0.6053 | |
|
| 0.2749 | 12.0 | 3384 | 1.3594 | 0.641 | 0.6149 | 0.6166 | 0.6156 | |
|
| 0.2635 | 13.0 | 3666 | 1.6264 | 0.637 | 0.5983 | 0.5884 | 0.5899 | |
|
| 0.2635 | 14.0 | 3948 | 1.7278 | 0.636 | 0.6101 | 0.6120 | 0.6109 | |
|
| 0.2382 | 15.0 | 4230 | 1.5670 | 0.635 | 0.6125 | 0.6165 | 0.6137 | |
|
| 0.2387 | 16.0 | 4512 | 1.6334 | 0.63 | 0.6098 | 0.6149 | 0.6108 | |
|
| 0.2387 | 17.0 | 4794 | 1.8990 | 0.632 | 0.5963 | 0.5908 | 0.5923 | |
|
| 0.2214 | 18.0 | 5076 | 1.5244 | 0.644 | 0.6116 | 0.6073 | 0.6088 | |
|
| 0.2214 | 19.0 | 5358 | 1.7948 | 0.639 | 0.6017 | 0.5928 | 0.5946 | |
|
| 0.2138 | 20.0 | 5640 | 1.7595 | 0.638 | 0.6040 | 0.5991 | 0.6006 | |
|
| 0.2138 | 21.0 | 5922 | 1.9552 | 0.633 | 0.6086 | 0.6114 | 0.6096 | |
|
| 0.2038 | 22.0 | 6204 | 1.9481 | 0.636 | 0.6160 | 0.6214 | 0.6171 | |
|
| 0.2038 | 23.0 | 6486 | 1.7951 | 0.648 | 0.6216 | 0.6227 | 0.6221 | |
|
| 0.2003 | 24.0 | 6768 | 2.0499 | 0.636 | 0.6042 | 0.6016 | 0.6026 | |
|
| 0.188 | 25.0 | 7050 | 2.3175 | 0.64 | 0.6147 | 0.6169 | 0.6156 | |
|
| 0.188 | 26.0 | 7332 | 2.3284 | 0.634 | 0.6109 | 0.6145 | 0.6120 | |
|
| 0.1838 | 27.0 | 7614 | 2.4239 | 0.634 | 0.6049 | 0.6046 | 0.6048 | |
|
| 0.1838 | 28.0 | 7896 | 2.4116 | 0.639 | 0.6118 | 0.6127 | 0.6122 | |
|
| 0.177 | 29.0 | 8178 | 2.4789 | 0.63 | 0.6052 | 0.6079 | 0.6062 | |
|
| 0.177 | 30.0 | 8460 | 2.4643 | 0.631 | 0.6057 | 0.6081 | 0.6066 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.30.2 |
|
- Pytorch 1.13.1+cu117 |
|
- Datasets 2.13.2 |
|
- Tokenizers 0.11.0 |
|
|
