|
--- |
|
license: apache-2.0 |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- f1 |
|
- accuracy |
|
model-index: |
|
- name: roberta-finetuned-CPV_Spanish |
|
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-finetuned-CPV_Spanish |
|
|
|
This model is a fine-tuned version of [PlanTL-GOB-ES/roberta-base-bne](https://huggingface.co/PlanTL-GOB-ES/roberta-base-bne) on a dataset derived from Spanish Public Procurement documents from 2019. The whole fine-tuning process is available in the following [Kaggle notebook](https://www.kaggle.com/code/marianavasloro/fine-tuned-roberta-for-spanish-cpv-codes). |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0152 |
|
- F1: 0.9462 |
|
- Roc Auc: 0.9698 |
|
- Accuracy: 0.9297 |
|
- Coverage Error: 3.6573 |
|
- Label Ranking Average Precision Score: 0.9451 |
|
|
|
## Intended uses & limitations |
|
|
|
This model only predicts the first two digits of the CPV codes. |
|
|
|
## Training procedure |
|
|
|
### Training hyperparameters |
|
|
|
The following hyperparameters were used during training: |
|
- learning_rate: 2e-05 |
|
- train_batch_size: 8 |
|
- eval_batch_size: 8 |
|
- seed: 42 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- num_epochs: 10 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | F1 | Roc Auc | Accuracy | Coverage Error | Label Ranking Average Precision Score | |
|
|:-------------:|:-----:|:------:|:---------------:|:------:|:-------:|:--------:|:--------------:|:-------------------------------------:| |
|
| 0.0287 | 1.0 | 20385 | 0.0270 | 0.8235 | 0.8815 | 0.7695 | 10.4603 | 0.8167 | |
|
| 0.0172 | 2.0 | 40770 | 0.0199 | 0.8773 | 0.9210 | 0.8306 | 7.5943 | 0.8768 | |
|
| 0.01 | 3.0 | 61155 | 0.0168 | 0.9028 | 0.9364 | 0.8639 | 6.2111 | 0.9045 | |
|
| 0.0062 | 4.0 | 81540 | 0.0152 | 0.9207 | 0.9520 | 0.8871 | 5.1353 | 0.9213 | |
|
| 0.0037 | 5.0 | 101925 | 0.0151 | 0.9300 | 0.9569 | 0.9026 | 4.7350 | 0.9295 | |
|
| 0.0021 | 6.0 | 122310 | 0.0147 | 0.9365 | 0.9625 | 0.9123 | 4.2946 | 0.9355 | |
|
| 0.0013 | 7.0 | 142695 | 0.0148 | 0.9396 | 0.9659 | 0.9184 | 3.9912 | 0.9387 | |
|
| 0.001 | 8.0 | 163080 | 0.0150 | 0.9426 | 0.9680 | 0.9243 | 3.8065 | 0.9422 | |
|
| 0.0006 | 9.0 | 183465 | 0.0152 | 0.9445 | 0.9693 | 0.9274 | 3.7064 | 0.9438 | |
|
| 0.0003 | 10.0 | 203850 | 0.0152 | 0.9462 | 0.9698 | 0.9297 | 3.6573 | 0.9451 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.16.2 |
|
- Pytorch 1.9.1 |
|
- Datasets 1.18.4 |
|
- Tokenizers 0.11.6 |
|
|
