bert-base-multilingual-uncased-finetuned-ner-lenerBR
This model is a fine-tuned version of google-bert/bert-base-multilingual-uncased on the lener_br dataset. It achieves the following results on the evaluation set:
- Loss: 0.1568
- Precision: 0.8678
- Recall: 0.8758
- F1: 0.8718
- Accuracy: 0.9707
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: 2e-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: 10
Training results
Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
---|---|---|---|---|---|---|---|
No log | 1.0 | 245 | 0.1819 | 0.7691 | 0.8118 | 0.7899 | 0.9585 |
No log | 2.0 | 490 | 0.1487 | 0.7383 | 0.8098 | 0.7724 | 0.9586 |
0.1325 | 3.0 | 735 | 0.1532 | 0.8662 | 0.8777 | 0.8719 | 0.9683 |
0.1325 | 4.0 | 980 | 0.1470 | 0.8770 | 0.8800 | 0.8785 | 0.9698 |
0.0233 | 5.0 | 1225 | 0.1155 | 0.8493 | 0.8839 | 0.8663 | 0.9750 |
0.0233 | 6.0 | 1470 | 0.1727 | 0.8874 | 0.8822 | 0.8848 | 0.9701 |
0.0126 | 7.0 | 1715 | 0.1698 | 0.8890 | 0.8853 | 0.8871 | 0.9710 |
0.0126 | 8.0 | 1960 | 0.1687 | 0.8651 | 0.8783 | 0.8716 | 0.9702 |
0.0076 | 9.0 | 2205 | 0.1593 | 0.8077 | 0.8797 | 0.8422 | 0.9668 |
0.0076 | 10.0 | 2450 | 0.1568 | 0.8678 | 0.8758 | 0.8718 | 0.9707 |
Framework versions
- Transformers 4.41.1
- Pytorch 2.1.2
- Datasets 2.19.1
- Tokenizers 0.19.1
- Downloads last month
- 8
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.
Model tree for GuiTap/bert-base-multilingual-uncased-finetuned-ner-lenerBR
Base model
google-bert/bert-base-multilingual-uncasedDataset used to train GuiTap/bert-base-multilingual-uncased-finetuned-ner-lenerBR
Evaluation results
- Precision on lener_brvalidation set self-reported0.868
- Recall on lener_brvalidation set self-reported0.876
- F1 on lener_brvalidation set self-reported0.872
- Accuracy on lener_brvalidation set self-reported0.971