indobert-classification
This model is a fine-tuned version of indobenchmark/indobert-base-p1 on the indonlu dataset. It achieves the following results on the evaluation set:
- Loss: 0.3707
- Accuracy: 0.9397
- F1: 0.9393
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: 5
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
---|---|---|---|---|---|
0.2458 | 1.0 | 688 | 0.2229 | 0.9325 | 0.9323 |
0.1258 | 2.0 | 1376 | 0.2332 | 0.9373 | 0.9369 |
0.059 | 3.0 | 2064 | 0.3389 | 0.9365 | 0.9365 |
0.0268 | 4.0 | 2752 | 0.3412 | 0.9421 | 0.9417 |
0.0097 | 5.0 | 3440 | 0.3707 | 0.9397 | 0.9393 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.11.0+cu113
- Datasets 2.1.0
- Tokenizers 0.12.1
- Downloads last month
- 42
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
the model is not deployed on the HF Inference API.
Dataset used to train afbudiman/indobert-classification
Evaluation results
- Accuracy on indonluself-reported0.940
- F1 on indonluself-reported0.939