Negation_Scope_Detection_NubEs_Spanish_mBERT_fine_tuned
This model is a fine-tuned version of bert-base-multilingual-cased on the nubes dataset. It achieves the following results on the evaluation set:
- Loss: 0.1624
- Precision: 0.9012
- Recall: 0.9184
- F1: 0.9098
- Accuracy: 0.9744
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: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 7
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.1802 | 1.0 | 1726 | 0.1849 | 0.7843 | 0.8526 | 0.8170 | 0.9509 |
| 0.1216 | 2.0 | 3452 | 0.1512 | 0.8706 | 0.8352 | 0.8525 | 0.9579 |
| 0.0817 | 3.0 | 5178 | 0.1083 | 0.8845 | 0.9038 | 0.8940 | 0.9710 |
| 0.0517 | 4.0 | 6904 | 0.1314 | 0.8858 | 0.8960 | 0.8909 | 0.9693 |
| 0.0265 | 5.0 | 8630 | 0.1514 | 0.8963 | 0.9079 | 0.9021 | 0.9721 |
| 0.0136 | 6.0 | 10356 | 0.1524 | 0.9045 | 0.9092 | 0.9068 | 0.9729 |
| 0.0045 | 7.0 | 12082 | 0.1624 | 0.9012 | 0.9184 | 0.9098 | 0.9744 |
Framework versions
- Transformers 4.21.2
- Pytorch 1.12.1+cu113
- Datasets 2.4.0
- Tokenizers 0.12.1