README.md

---
license: mit
language:
- en
metrics:
- accuracy
base_model:
- distilbert/distilbert-base-uncased
pipeline_tag: text-classification
tags:
- Bug Detection
- Review Classification
- NLP
- Deep Learning
- App Reviews
- Game Reviews
- BlaikHole
---

# 🚀 DistilBert Reviews Bug Classifier by BlaikHole

<p align="center">
  <a href="https://huggingface.co/blaikhole/distilbert-review-bug-classifier" target="_blank" rel="noopener noreferrer">
    <img src="https://img.shields.io/badge/HuggingFace-Model-yellow?logo=huggingface" height="35">
  </a>
  <a href="https://huggingface.co/spaces/blaikhole/review-bug-classifier" target="_blank" rel="noopener noreferrer">
    <img src="https://img.shields.io/badge/Demo-Space-blue?style=flat-square" height="35">
  </a>
</p>

## 📌 Overview

This repository provides a **fine-tuned** model trained on our **private Playstore reviews data** using **quick still efficient DistilBert architecture**. It can be used for **Reviews classification with 3 classes (with 4th class - No bug)**.

---

## 🎨 Model Outputs & Labels

The model identifies the following labels:

| Label Name | Description |
|------------|-------------|
| 🟥 **LABEL_0 > Graphics Issue** | Screen touch controls issue, graphics flickering, rendering issues. |
| 🟩 **LABEL_1 > Network Issue** | Login/signup, account issues, wi-fi/data or ping problems etc. |
| 🟦 **LABEL_2 > No Bug** | No bug discussion found. |
| 🟨 **LABEL_3 > Performance Issue** | Overheating mobile, lag, crash, stuck game and so on. |

---

## 🚀 Quick Usage

You can easily load and use this model with `transformers`:

### 🔹 Named Entity Recognition (NER)
```python
from transformers import pipeline

# Label Mapping
LABEL_MAP = {
    "LABEL_0": "Graphics issue", 
    "LABEL_1": "Network issue", 
    "LABEL_2": "No Bug", 
    "LABEL_3": "Performance issue"
}

# Load Text Classification Model
MODEL_NAME = "blaikhole/distilbert-review-bug-classifier"
classifier = pipeline("text-classification", model=MODEL_NAME, tokenizer=MODEL_NAME)

def classify_text(text):
    result = classifier(text)[0]
    label = LABEL_MAP.get(result["label"], "Unknown")
    return f"Predicted Label: {label} (Confidence: {result['score']:.2f})"

# Example Usage
if __name__ == "__main__":
    sample_text = "The game keeps lagging and frame rates drop frequently."
    print(classify_text(sample_text))

```
---

## 📦 Installation

To use this model, install the required dependencies:

```bash
pip install transformers torch
```
---

## 📜 License

MIT