|
--- |
|
license: mit |
|
base_model: xlm-roberta-large |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
- f1 |
|
- precision |
|
- recall |
|
model-index: |
|
- name: xlm-roberta-large-twitter-indonesia-sarcastic |
|
results: [] |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# xlm-roberta-large-twitter-indonesia-sarcastic |
|
|
|
This model is a fine-tuned version of [xlm-roberta-large](https://huggingface.co./xlm-roberta-large) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.4322 |
|
- Accuracy: 0.8885 |
|
- F1: 0.7692 |
|
- Precision: 0.7937 |
|
- Recall: 0.7463 |
|
|
|
## 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 | |
|
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:| |
|
| 0.5862 | 1.0 | 59 | 0.5304 | 0.75 | 0.0 | 0.0 | 0.0 | |
|
| 0.5168 | 2.0 | 118 | 0.4897 | 0.75 | 0.0 | 0.0 | 0.0 | |
|
| 0.4771 | 3.0 | 177 | 0.4535 | 0.7948 | 0.3373 | 0.875 | 0.2090 | |
|
| 0.4101 | 4.0 | 236 | 0.4235 | 0.7910 | 0.6585 | 0.5567 | 0.8060 | |
|
| 0.3225 | 5.0 | 295 | 0.4733 | 0.8507 | 0.5918 | 0.9355 | 0.4328 | |
|
| 0.2246 | 6.0 | 354 | 0.3362 | 0.8694 | 0.7009 | 0.82 | 0.6119 | |
|
| 0.166 | 7.0 | 413 | 0.3672 | 0.8769 | 0.7227 | 0.8269 | 0.6418 | |
|
| 0.0989 | 8.0 | 472 | 0.3835 | 0.8769 | 0.7626 | 0.7361 | 0.7910 | |
|
| 0.0797 | 9.0 | 531 | 0.4379 | 0.8993 | 0.7939 | 0.8125 | 0.7761 | |
|
| 0.08 | 10.0 | 590 | 0.7677 | 0.8545 | 0.7451 | 0.6628 | 0.8507 | |
|
| 0.0505 | 11.0 | 649 | 0.7316 | 0.8806 | 0.7288 | 0.8431 | 0.6418 | |
|
| 0.073 | 12.0 | 708 | 0.4796 | 0.9104 | 0.8182 | 0.8308 | 0.8060 | |
|
| 0.05 | 13.0 | 767 | 0.8469 | 0.8694 | 0.7059 | 0.8077 | 0.6269 | |
|
| 0.0583 | 14.0 | 826 | 0.7266 | 0.8918 | 0.7563 | 0.8654 | 0.6716 | |
|
| 0.0275 | 15.0 | 885 | 0.8974 | 0.8918 | 0.7387 | 0.9318 | 0.6119 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.36.2 |
|
- Pytorch 2.1.1+cu121 |
|
- Datasets 2.15.0 |
|
- Tokenizers 0.15.0 |
|
|