|
|
--- |
|
|
library_name: transformers |
|
|
license: apache-2.0 |
|
|
base_model: distilbert-base-uncased |
|
|
tags: |
|
|
- generated_from_trainer |
|
|
datasets: |
|
|
- lener_br |
|
|
metrics: |
|
|
- precision |
|
|
- recall |
|
|
- f1 |
|
|
- accuracy |
|
|
model-index: |
|
|
- name: ner_bert_model |
|
|
results: |
|
|
- task: |
|
|
name: Token Classification |
|
|
type: token-classification |
|
|
dataset: |
|
|
name: lener_br |
|
|
type: lener_br |
|
|
config: lener_br |
|
|
split: test |
|
|
args: lener_br |
|
|
metrics: |
|
|
- name: Precision |
|
|
type: precision |
|
|
value: 0.828094932649134 |
|
|
- name: Recall |
|
|
type: recall |
|
|
value: 0.8532716457369465 |
|
|
- name: F1 |
|
|
type: f1 |
|
|
value: 0.8404947916666666 |
|
|
- name: Accuracy |
|
|
type: accuracy |
|
|
value: 0.9840912469998513 |
|
|
--- |
|
|
|
|
|
<!-- 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. --> |
|
|
|
|
|
# ner_bert_model |
|
|
|
|
|
This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the lener_br dataset. |
|
|
It achieves the following results on the evaluation set: |
|
|
- Loss: 0.0922 |
|
|
- Precision: 0.8281 |
|
|
- Recall: 0.8533 |
|
|
- F1: 0.8405 |
|
|
- Accuracy: 0.9841 |
|
|
|
|
|
## 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: 16 |
|
|
- eval_batch_size: 16 |
|
|
- seed: 42 |
|
|
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments |
|
|
- lr_scheduler_type: linear |
|
|
- num_epochs: 20 |
|
|
|
|
|
### Training results |
|
|
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy | |
|
|
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:| |
|
|
| No log | 1.0 | 490 | 0.0971 | 0.6373 | 0.7607 | 0.6936 | 0.9706 | |
|
|
| 0.2449 | 2.0 | 980 | 0.0820 | 0.6916 | 0.8063 | 0.7446 | 0.9760 | |
|
|
| 0.0634 | 3.0 | 1470 | 0.0750 | 0.7106 | 0.8473 | 0.7730 | 0.9778 | |
|
|
| 0.0352 | 4.0 | 1960 | 0.0707 | 0.7690 | 0.8361 | 0.8011 | 0.9799 | |
|
|
| 0.0226 | 5.0 | 2450 | 0.0812 | 0.8063 | 0.8394 | 0.8225 | 0.9821 | |
|
|
| 0.0157 | 6.0 | 2940 | 0.0779 | 0.7931 | 0.8486 | 0.8199 | 0.9826 | |
|
|
| 0.0105 | 7.0 | 3430 | 0.0958 | 0.7314 | 0.8586 | 0.7899 | 0.9779 | |
|
|
| 0.0082 | 8.0 | 3920 | 0.0810 | 0.8158 | 0.8460 | 0.8306 | 0.9829 | |
|
|
| 0.0067 | 9.0 | 4410 | 0.0830 | 0.8190 | 0.8526 | 0.8355 | 0.9832 | |
|
|
| 0.0054 | 10.0 | 4900 | 0.0810 | 0.8165 | 0.8500 | 0.8329 | 0.9833 | |
|
|
| 0.0051 | 11.0 | 5390 | 0.0855 | 0.8180 | 0.8493 | 0.8333 | 0.9832 | |
|
|
| 0.0037 | 12.0 | 5880 | 0.0862 | 0.8195 | 0.8519 | 0.8354 | 0.9841 | |
|
|
| 0.0034 | 13.0 | 6370 | 0.0867 | 0.8165 | 0.8586 | 0.8370 | 0.9833 | |
|
|
| 0.0027 | 14.0 | 6860 | 0.0922 | 0.8214 | 0.8420 | 0.8316 | 0.9832 | |
|
|
| 0.0024 | 15.0 | 7350 | 0.0910 | 0.8147 | 0.8486 | 0.8313 | 0.9836 | |
|
|
| 0.002 | 16.0 | 7840 | 0.0928 | 0.8191 | 0.8559 | 0.8371 | 0.9840 | |
|
|
| 0.0018 | 17.0 | 8330 | 0.0928 | 0.8119 | 0.8559 | 0.8333 | 0.9834 | |
|
|
| 0.0017 | 18.0 | 8820 | 0.0920 | 0.8228 | 0.8592 | 0.8406 | 0.9838 | |
|
|
| 0.0015 | 19.0 | 9310 | 0.0919 | 0.8242 | 0.8553 | 0.8394 | 0.9837 | |
|
|
| 0.0011 | 20.0 | 9800 | 0.0922 | 0.8281 | 0.8533 | 0.8405 | 0.9841 | |
|
|
|
|
|
|
|
|
### Framework versions |
|
|
|
|
|
- Transformers 4.52.4 |
|
|
- Pytorch 2.6.0+cu124 |
|
|
- Datasets 3.6.0 |
|
|
- Tokenizers 0.21.1 |
|
|
|
