metadata
license: mit
tags:
- generated_from_trainer
datasets:
- lener_br
metrics:
- precision
- recall
- f1
- accuracy
model_index:
- name: bertimbau-large-lener_br
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: lener_br
type: lener_br
args: lener_br
metric:
name: Accuracy
type: accuracy
value: 0.9790177823152056
bertimbau-large-lener_br
This model is a fine-tuned version of neuralmind/bert-large-portuguese-cased on the lener_br dataset. It achieves the following results on the evaluation set:
- Loss: 0.1064
- Precision: 0.8676
- Recall: 0.9204
- F1: 0.8932
- Accuracy: 0.9790
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: 4
- eval_batch_size: 4
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.0601 | 1.0 | 1957 | 0.1097 | 0.8223 | 0.9004 | 0.8596 | 0.9700 |
| 0.0301 | 2.0 | 3914 | 0.1148 | 0.8070 | 0.9333 | 0.8656 | 0.9722 |
| 0.0122 | 3.0 | 5871 | 0.1064 | 0.8676 | 0.9204 | 0.8932 | 0.9790 |
Framework versions
- Transformers 4.8.2
- Pytorch 1.9.0+cu102
- Datasets 1.9.0
- Tokenizers 0.10.3