|
--- |
|
license: mit |
|
base_model: microsoft/deberta-v3-base |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- precision |
|
- recall |
|
- f1 |
|
- accuracy |
|
model-index: |
|
- name: deberta-v3-base-A |
|
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. --> |
|
|
|
# deberta-v3-base-A |
|
|
|
This model is a fine-tuned version of [microsoft/deberta-v3-base](https://huggingface.co/microsoft/deberta-v3-base) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0554 |
|
- Precision: 0.9085 |
|
- Recall: 0.9353 |
|
- F1: 0.9217 |
|
- Accuracy: 0.9838 |
|
|
|
## 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: 16 |
|
- eval_batch_size: 16 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 8 |
|
- total_train_batch_size: 128 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 3 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:| |
|
| 0.0668 | 0.13 | 257 | 0.0703 | 0.8962 | 0.8377 | 0.8660 | 0.9767 | |
|
| 0.0511 | 0.25 | 514 | 0.0652 | 0.8348 | 0.9211 | 0.8758 | 0.9760 | |
|
| 0.0536 | 0.38 | 771 | 0.0541 | 0.8800 | 0.8998 | 0.8898 | 0.9802 | |
|
| 0.0392 | 0.5 | 1028 | 0.0552 | 0.8712 | 0.9226 | 0.8961 | 0.9797 | |
|
| 0.0433 | 0.63 | 1285 | 0.0538 | 0.8711 | 0.9242 | 0.8968 | 0.9799 | |
|
| 0.0411 | 0.75 | 1542 | 0.0502 | 0.8850 | 0.9258 | 0.9049 | 0.9807 | |
|
| 0.0341 | 0.88 | 1799 | 0.0473 | 0.9021 | 0.9166 | 0.9093 | 0.9828 | |
|
| 0.042 | 1.0 | 2056 | 0.0475 | 0.9111 | 0.9154 | 0.9133 | 0.9827 | |
|
| 0.0277 | 1.13 | 2313 | 0.0486 | 0.9132 | 0.9166 | 0.9149 | 0.9828 | |
|
| 0.026 | 1.25 | 2570 | 0.0484 | 0.9056 | 0.9250 | 0.9152 | 0.9831 | |
|
| 0.0259 | 1.38 | 2827 | 0.0504 | 0.8986 | 0.9291 | 0.9136 | 0.9824 | |
|
| 0.031 | 1.5 | 3084 | 0.0518 | 0.8889 | 0.9352 | 0.9115 | 0.9819 | |
|
| 0.0269 | 1.63 | 3341 | 0.0492 | 0.8993 | 0.9338 | 0.9162 | 0.9828 | |
|
| 0.022 | 1.75 | 3598 | 0.0496 | 0.9029 | 0.9307 | 0.9166 | 0.9831 | |
|
| 0.0228 | 1.88 | 3855 | 0.0494 | 0.9101 | 0.9296 | 0.9198 | 0.9835 | |
|
| 0.0166 | 2.0 | 4112 | 0.0514 | 0.9095 | 0.9316 | 0.9204 | 0.9835 | |
|
| 0.0162 | 2.13 | 4369 | 0.0533 | 0.9041 | 0.9329 | 0.9183 | 0.9833 | |
|
| 0.0144 | 2.26 | 4626 | 0.0545 | 0.9074 | 0.9319 | 0.9195 | 0.9835 | |
|
| 0.0126 | 2.38 | 4883 | 0.0538 | 0.9044 | 0.9360 | 0.9199 | 0.9836 | |
|
| 0.013 | 2.51 | 5140 | 0.0551 | 0.9085 | 0.9332 | 0.9207 | 0.9834 | |
|
| 0.0138 | 2.63 | 5397 | 0.0565 | 0.9054 | 0.9351 | 0.9200 | 0.9833 | |
|
| 0.0144 | 2.76 | 5654 | 0.0544 | 0.9065 | 0.9357 | 0.9209 | 0.9838 | |
|
| 0.0136 | 2.88 | 5911 | 0.0554 | 0.9085 | 0.9353 | 0.9217 | 0.9838 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.35.2 |
|
- Pytorch 2.1.0+cu118 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
|
|
