indonesian-roberta-base-nerp-tagger
This model is a fine-tuned version of flax-community/indonesian-roberta-base on the indonlu dataset. It achieves the following results on the evaluation set:
- Loss: 0.1180
- Precision: 0.8102
- Recall: 0.8107
- F1: 0.8105
- Accuracy: 0.9615
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: 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 | 420 | 0.1419 | 0.7491 | 0.8034 | 0.7753 | 0.9551 |
| 0.2261 | 2.0 | 840 | 0.1317 | 0.7889 | 0.7983 | 0.7936 | 0.9569 |
| 0.1081 | 3.0 | 1260 | 0.1430 | 0.7587 | 0.8300 | 0.7927 | 0.9546 |
| 0.0777 | 4.0 | 1680 | 0.1459 | 0.7848 | 0.8266 | 0.8052 | 0.9577 |
| 0.0563 | 5.0 | 2100 | 0.1525 | 0.7923 | 0.8125 | 0.8022 | 0.9579 |
| 0.0441 | 6.0 | 2520 | 0.1552 | 0.7986 | 0.8176 | 0.8080 | 0.9584 |
| 0.0441 | 7.0 | 2940 | 0.1692 | 0.7910 | 0.8232 | 0.8068 | 0.9584 |
| 0.0387 | 8.0 | 3360 | 0.1677 | 0.7894 | 0.8306 | 0.8095 | 0.9588 |
| 0.032 | 9.0 | 3780 | 0.1784 | 0.7939 | 0.8249 | 0.8091 | 0.9586 |
| 0.0284 | 10.0 | 4200 | 0.1817 | 0.7950 | 0.8261 | 0.8102 | 0.9588 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.2.0+cu118
- Datasets 2.16.1
- Tokenizers 0.15.1