# 🧪 Pankaj-TestCaseGeneration-Mistral

## 🚀 Model Overview
This project uses a fine-tuned version of **[mistralai/Mistral-7B-Instruct-v0.3](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3)**, specialized for **automatically generating software test cases** directly from user stories.  

Given a user story as input, the model generates a comprehensive set of:
- ✅ Positive Test Cases
- ❌ Negative Test Cases
- ⚠️ Edge Case Scenarios  

This helps QA teams **accelerate test case design** by automating repetitive analysis work.

---

## 🧪 Model Details

| Property         | Value |
|------------------|-----|
| **Base Model**   | mistralai/Mistral-7B-Instruct-v0.3 |
| **Task**         | Test Case Generation (QA Automation) |
| **Language**     | English |
| **Input**        | User Story (free text) |
| **Output**       | Structured Test Cases |

---

## 👅 How to Use

You can use this model directly via the Hugging Face Transformers library in Python.

### Example Code
```python
from transformers import AutoModelForCausalLM, AutoTokenizer

model_id = "Pankaj/TestCaseGeneration-Mistral"

tokenizer = AutoTokenizer.from_pretrained(model_id)
model = AutoModelForCausalLM.from_pretrained(model_id)

prompt = """
User Story:
As a user, I want to reset my password so that I can regain access to my account if I forget my password.

Test Cases:
"""

inputs = tokenizer(prompt, return_tensors="pt")
outputs = model.generate(**inputs, max_new_tokens=500)

print(tokenizer.decode(outputs[0], skip_special_tokens=True))
```

---

## ⚒️ Training Process

The model was **fine-tuned on a custom dataset** containing:

✅ Real-world user stories collected from projects  
✅ Manually curated test cases written by experienced QA engineers  

The fine-tuning was performed using **Hugging Face Transformers** in Python.

---

## 📚 Intended Use

This model is designed to:

✅ Automate the first draft of test cases for QA teams  
✅ Serve as an **assistive tool** for junior QA engineers or developers  
✅ Save time when starting test design from scratch  

---

## ✅ Sample Input & Output

### Input
```text
User Story:
As a user, I want to search for products by name so that I can quickly find items I am interested in.

Test Cases:
```

### Output
```text
1. [Positive] Search for existing product
   Pre-condition: Product exists in catalog
   Steps:
   1. Navigate to the search page
   2. Enter product name in search bar
   3. Click "Search"
   Expected Result: Product is shown in search results

2. [Negative] Search with invalid characters
   Pre-condition: None
   Steps:
   1. Enter "@@@@" in search bar
   2. Click "Search"
   Expected Result: "No results found" message is shown

3. [Edge] Case-insensitive search
   Pre-condition: Product exists with lowercase name
   Steps:
   1. Enter product name in uppercase
   2. Click "Search"
   Expected Result: Product still appears in search results
```

---

## ⚠️ Limitations

- Output quality may degrade with **very complex or ambiguous user stories**.
- Domain-specific terminology (finance, healthcare, etc.) may require **further fine-tuning**.

---

## 🔐 Handling API Keys Safely

To run this project locally, you need a Hugging Face API token.  

**Recommended Approach:**

1. Create a `.env` file and add:
    ```
    HF_TOKEN=your_hugging_face_token_here
    ```
2. Add `.env` to `.gitignore` to avoid accidental exposure.
3. Use this code to read the token:
    ```python
    from dotenv import load_dotenv
    import os

    load_dotenv()
    hf_token = os.getenv("HF_TOKEN")
    ```

---

## 💬 Feedback and Contributions

Contributions are welcome — whether it's improving prompts, enhancing the fine-tuning dataset, or adding new features like **Jira integration** for fetching user stories automatically.

Feel free to raise issues, create pull requests, or simply share feedback. Let's make test automation smarter together! ✨

---

### 🎉 Happy Testing!