indobert-base-uncased-twitter-indonesia-sarcastic
This model is a fine-tuned version of indolem/indobert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4202
- Accuracy: 0.8030
- F1: 0.6467
- Precision: 0.5843
- Recall: 0.7239
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 |
|---|---|---|---|---|---|---|---|
| 0.5531 | 1.0 | 59 | 0.4977 | 0.7724 | 0.4078 | 0.5833 | 0.3134 |
| 0.4992 | 2.0 | 118 | 0.4785 | 0.7724 | 0.3441 | 0.6154 | 0.2388 |
| 0.44 | 3.0 | 177 | 0.4819 | 0.7799 | 0.3656 | 0.6538 | 0.2537 |
| 0.3815 | 4.0 | 236 | 0.4524 | 0.8097 | 0.6623 | 0.5952 | 0.7463 |
| 0.3104 | 5.0 | 295 | 0.4547 | 0.8172 | 0.5421 | 0.725 | 0.4328 |
| 0.2592 | 6.0 | 354 | 0.4058 | 0.8172 | 0.5664 | 0.6957 | 0.4776 |
| 0.2083 | 7.0 | 413 | 0.4358 | 0.8060 | 0.5738 | 0.6364 | 0.5224 |
Framework versions
- Transformers 4.36.2
- Pytorch 2.1.1+cu121
- Datasets 2.15.0
- Tokenizers 0.15.0