|
--- |
|
library_name: transformers |
|
base_model: airesearch/wangchanberta-base-att-spm-uncased |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- accuracy |
|
model-index: |
|
- name: wangchan-sentiment-thai-text-model |
|
results: [] |
|
datasets: |
|
- Wongnai/wongnai_reviews |
|
- pythainlp/wisesight_sentiment |
|
language: |
|
- th |
|
pipeline_tag: text-classification |
|
--- |
|
|
|
<!-- 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. --> |
|
|
|
# wangchan-sentiment-thai-text-model |
|
|
|
This model is a fine-tuned version of [airesearch/wangchanberta-base-att-spm-uncased](https://huggingface.co/airesearch/wangchanberta-base-att-spm-uncased) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.5849 |
|
- Accuracy: 0.7535 |
|
|
|
## 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: 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 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Accuracy | |
|
|:-------------:|:-----:|:-----:|:---------------:|:--------:| |
|
| 0.6445 | 1.0 | 3822 | 0.6295 | 0.7309 | |
|
| 0.5865 | 2.0 | 7644 | 0.5855 | 0.7430 | |
|
| 0.5285 | 3.0 | 11466 | 0.5754 | 0.7455 | |
|
| 0.5127 | 4.0 | 15288 | 0.5816 | 0.7492 | |
|
| 0.4861 | 5.0 | 19110 | 0.5849 | 0.7535 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.44.2 |
|
- Pytorch 2.4.1+cu121 |
|
- Datasets 3.0.1 |
|
- Tokenizers 0.19.1 |
