|
|
--- |
|
|
library_name: transformers |
|
|
tags: |
|
|
- bangla |
|
|
- bangla-classifier |
|
|
- binary-classifier |
|
|
- text-classifier |
|
|
datasets: |
|
|
- SayedShaun/sentigold |
|
|
language: |
|
|
- bn |
|
|
metrics: |
|
|
- accuracy |
|
|
base_model: |
|
|
- csebuetnlp/banglabert |
|
|
pipeline_tag: text-classification |
|
|
--- |
|
|
|
|
|
# 🔍 Bangla Binary Text Classifier |
|
|
|
|
|
## 🧠 Model Description |
|
|
|
|
|
This is a **Bangla binary sentiment classification** model, fine-tuned on top of [`csebuetnlp/banglabert`](https://huggingface.co/csebuetnlp/banglabert). The model was trained using the [**SayedShaun/sentigold**](https://huggingface.co/datasets/SayedShaun/sentigold) dataset. |
|
|
|
|
|
--- |
|
|
|
|
|
## 📦 How to Use |
|
|
|
|
|
```python |
|
|
from transformers import pipeline |
|
|
|
|
|
pipe = pipeline("text-classification", model="SayedShaun/bangla-classifier-binary") |
|
|
|
|
|
response = pipe("এটা যে এত খারাপ আগে জানতাম না।") |
|
|
print(response) |
|
|
>>> [{'label': 'LABEL_0', 'score': 0.9765}] |
|
|
``` |
|
|
|
|
|
## Result |
|
|
| Training Loss | Validation Loss | Accuracy | Precision | Recall | F1 Score | |
|
|
|---------------|-----------------|-----------|-----------|----------|-----------| |
|
|
| 0.354600 | 0.396599 | 0.825143 | 0.812587 | 0.842483 | 0.827265 | |
|
|
|
|
|
|
|
|
# Source Code |
|
|
Source code can be found in `files and versions` as `finetune.py` |
