---
title: GPT
emoji: 💬
colorFrom: yellow
colorTo: purple
sdk: gradio
sdk_version: 5.0.1
app_file: app.py
pinned: false
short_description: AI Chatbot with File Upload Support
---

# 🤖 AI Chatbot with File Upload Support

A ChatGPT-like interface built with Gradio that supports file uploads and uses a local transformer model for text generation.

## ✨ Features

- **🤖 Chat Interface**: Modern, clean chat interface similar to ChatGPT
- **📁 File Upload Support**: Upload and analyze multiple file types:
  - PDF documents (.pdf)
  - Word documents (.docx)
  - Text files (.txt)
  - CSV files (.csv)
  - Excel spreadsheets (.xlsx, .xls)
- **⚙️ Configurable Settings**:
  - Custom system messages
  - Temperature control for response creativity
  - Maximum token limits
- **🔒 Local Model**: Uses your specified transformer model for privacy and control
- **💬 Conversation History**: Maintains chat context throughout the session
- **🛡️ Error Handling**: Graceful fallback to smaller models if the main model fails
- **📱 Responsive Design**: Clean, user-friendly interface

## 🚀 How to Run the Chatbot

### ⚠️ **IMPORTANT: Virtual Environment Required**

This project uses a virtual environment with all dependencies installed. Follow the **exact sequence** below to run the chatbot successfully.

### **✅ Recommended Method (Step-by-Step)**

1. **Open Terminal** in the project folder:

   ```bash
   cd "C:\Users\Cosmo\Desktop\GPT\GPT"
   ```

2. **Activate Virtual Environment**:

   ```bash
   .\env\Scripts\activate.bat
   ```

3. **Run with Full Python Path** (this ensures it works even if activation doesn't fully work):

   ```bash
   C:/Users/Cosmo/Desktop/GPT/GPT/env/Scripts/python.exe app.py
   ```

4. **Success! You should see**:

   ```
   * Running on local URL:  http://127.0.0.1:7860
   ```

5. **Open your browser** and go to: **http://127.0.0.1:7860**

### **Alternative Methods**

### **Method 1: Using the Batch File (if available)**

1. **Double-click** the `start_chatbot.bat` file in the project folder
2. This should automatically handle everything

### **Method 2: PowerShell (if execution policy allows)**

1. **Open PowerShell** in the project folder
2. **Fix execution policy** (one-time setup):
   ```powershell
   Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser
   ```
3. **Activate and run**:
   ```powershell
   .\env\Scripts\Activate.ps1
   python app.py
   ```

### **Method 3: From VS Code**

1. **Open the project folder** in VS Code
2. **Open integrated terminal** (Ctrl + `)
3. **Use the recommended sequence**:
   ```bash
   .\env\Scripts\activate.bat
   C:/Users/Cosmo/Desktop/GPT/GPT/env/Scripts/python.exe app.py
   ```

### **✅ Verify Virtual Environment is Active**

When the virtual environment is active, you should see `(env)` at the beginning of your command prompt:

```
(env) C:\Users\Cosmo\Desktop\GPT\GPT>
```

**⚠️ Important**: Even with `(env)` showing, you may still need to use the full Python path for the app to work correctly.

### **❌ Common Errors & Solutions**

#### **Error: "ModuleNotFoundError: No module named 'gradio'"**

**Even if you see `(env)` in your prompt**, this error means the Python path isn't correct.

**✅ Solution**: Always use the full Python path:

```bash
C:/Users/Cosmo/Desktop/GPT/GPT/env/Scripts/python.exe app.py
```

#### **Error: ".\env\Scripts\Activate.ps1 cannot be loaded"**

This is a PowerShell execution policy issue.

**✅ Solutions** (choose one):

1. **Use Command Prompt activation instead**:

   ```bash
   .\env\Scripts\activate.bat
   ```

2. **Fix PowerShell execution policy**:

   ```powershell
   Set-ExecutionPolicy -ExecutionPolicy RemoteSigned -Scope CurrentUser
   ```

3. **Use the batch file**: Double-click `start_chatbot.bat`

#### **Error: Missing packages**

If packages aren't installed in the virtual environment:

```bash
# Activate first
.\env\Scripts\activate.bat

# Then install
pip install -r requirements.txt
```

## 🌐 Using the Chatbot

1. **Start the application** using one of the methods above
2. **Open your browser** to the displayed URL (usually http://127.0.0.1:7860)
3. **Upload files** (optional) using the file upload area
4. **Type your message** in the text box
5. **Click Send** or press Enter to get AI responses
6. **Adjust settings** in the right panel as needed

## 📄 File Processing Capabilities

The chatbot can extract and analyze content from:

- **PDF files**: Extracts text from all pages using PyPDF2
- **Word documents**: Extracts text from paragraphs using python-docx
- **Text files**: Reads content directly with UTF-8 encoding
- **CSV/Excel files**: Converts data to readable format using pandas
- **Multi-file Analysis**: Upload multiple files simultaneously for comprehensive analysis

## 🔧 Model Information

- **Primary Model**: `openai/gpt-oss-120b`
- **Fallback Model**: `microsoft/DialoGPT-medium` (if primary model fails)
- **Framework**: Hugging Face Transformers
- **Interface**: Gradio web interface
- **Local Processing**: All inference runs locally for privacy

## ⚙️ Customization

### Model Configuration

Edit `model.py` to customize:

- Change the `model_id` variable to use different models
- Adjust model parameters like `torch_dtype` and `device_map`
- Modify fallback behavior

### Interface Customization

Edit `app.py` to modify:

- Chat interface appearance and behavior
- File upload restrictions and processing
- System message defaults
- Parameter ranges and defaults

## 📋 Requirements

- **Python**: 3.13+ (configured in virtual environment)
- **GPU**: CUDA-compatible GPU recommended for larger models
- **Memory**: Sufficient RAM for model loading (varies by model size)
- **Storage**: Space for model files (can be several GB)
- **Internet**: Required for initial model download

## 📦 Dependencies

All dependencies are managed in the virtual environment:

- `gradio` - Web interface framework
- `transformers` - Hugging Face model library
- `torch` - PyTorch deep learning framework
- `PyPDF2` - PDF text extraction
- `python-docx` - Word document processing
- `pandas` - Data analysis and CSV/Excel handling
- `openpyxl` - Excel file support

## 🛠️ Troubleshooting

### Common Issues:

1. **Model Loading Errors**:

   - Check internet connection for initial download
   - Verify sufficient disk space
   - The app will automatically fall back to a smaller model

2. **File Upload Issues**:

   - Ensure file types are supported
   - Check file permissions and accessibility
   - Large files may take time to process

3. **Performance Issues**:

   - Consider using a smaller model for faster responses
   - Reduce max_tokens for quicker generation
   - Ensure adequate GPU memory

4. **Environment Issues**:
   - Verify the virtual environment is properly activated
   - Check that all dependencies are installed
   - Review terminal output for specific error messages

### Getting Help:

- Check the terminal output for detailed error messages
- Verify file formats are supported
- Ensure the virtual environment is activated
- Try the fallback model if the primary model fails

## 🎯 Usage Tips

- **File Analysis**: Upload relevant documents before asking questions about them
- **Context**: The AI maintains conversation history, so you can ask follow-up questions
- **Settings**: Adjust temperature for more creative (higher) or focused (lower) responses
- **Batch Processing**: Upload multiple files at once for comprehensive analysis
- **Privacy**: All processing happens locally - your files and conversations stay private

Your ChatGPT-like chatbot is ready to use! 🎊PT
emoji: 💬
colorFrom: yellow
colorTo: purple
sdk: gradio
sdk_version: 5.0.1
app_file: app.py
pinned: false
short_description: My own gpt

---

An example chatbot using [Gradio](https://gradio.app), [`huggingface_hub`](https://huggingface.co/docs/huggingface_hub/v0.22.2/en/index), and the [Hugging Face Inference API](https://huggingface.co/docs/api-inference/index).