xlm-roberta-base-twitter-indonesia-sarcastic
This model is a fine-tuned version of xlm-roberta-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4359
- Accuracy: 0.8513
- F1: 0.7386
- Precision: 0.6570
- Recall: 0.8433
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: 1e-05
- train_batch_size: 32
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- num_epochs: 100.0
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
|---|---|---|---|---|---|---|---|
| 0.5641 | 1.0 | 59 | 0.5260 | 0.75 | 0.0 | 0.0 | 0.0 |
| 0.5317 | 2.0 | 118 | 0.5030 | 0.75 | 0.0 | 0.0 | 0.0 |
| 0.4995 | 3.0 | 177 | 0.4656 | 0.75 | 0.0 | 0.0 | 0.0 |
| 0.4599 | 4.0 | 236 | 0.4503 | 0.7687 | 0.6026 | 0.5281 | 0.7015 |
| 0.4082 | 5.0 | 295 | 0.3785 | 0.8470 | 0.6435 | 0.7708 | 0.5522 |
| 0.3274 | 6.0 | 354 | 0.3605 | 0.8619 | 0.6992 | 0.7679 | 0.6418 |
| 0.2621 | 7.0 | 413 | 0.3765 | 0.8619 | 0.6838 | 0.8 | 0.5970 |
| 0.2332 | 8.0 | 472 | 0.3408 | 0.8769 | 0.7591 | 0.7429 | 0.7761 |
| 0.1579 | 9.0 | 531 | 0.4382 | 0.8731 | 0.7213 | 0.8 | 0.6567 |
| 0.1467 | 10.0 | 590 | 0.3855 | 0.8806 | 0.7895 | 0.7059 | 0.8955 |
| 0.098 | 11.0 | 649 | 0.4693 | 0.8806 | 0.7500 | 0.7869 | 0.7164 |
| 0.0929 | 12.0 | 708 | 0.6206 | 0.8806 | 0.7333 | 0.8302 | 0.6567 |
| 0.0555 | 13.0 | 767 | 0.7134 | 0.8843 | 0.7634 | 0.7812 | 0.7463 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.1+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0