README.md

metadata

title: Petite LLM 3
emoji: 💃🏻
colorFrom: green
colorTo: purple
sdk: gradio
sdk_version: 5.38.2
app_file: app.py
pinned: true
license: mit
short_description: Smollm3 for French Understanding

🤖 Petite Elle L'Aime 3 - Chat Interface

A complete Gradio application for the Petite Elle L'Aime 3 model, featuring the full fine-tuned version for maximum performance and quality.

🚀 Features

• Multilingual Support: English, French, Italian, Portuguese, Chinese, Arabic
• Full Fine-Tuned Model: Maximum performance and quality with full precision
• Interactive Chat Interface: Real-time conversation with the model
• Customizable System Prompt: Define the assistant's personality and behavior
• Thinking Mode: Enable reasoning mode with thinking tags
• Responsive Design: Modern UI following the reference layout
• Chat Template Integration: Proper Jinja template formatting
• Automatic Model Download: Downloads full model at build time

📋 Model Information

• Base Model: HuggingFaceTB/SmolLM3-3B
• Parameters: ~3B
• Context Length: 128k
• Precision: Full fine-tuned model (float16/float32)
• Performance: Maximum quality and accuracy
• Languages: English, French, Italian, Portuguese, Chinese, Arabic

🛠️ Installation

1. Clone this repository:

git clone <repository-url>
cd Petite-LLM-3

2. Install dependencies:

pip install -r requirements.txt

🚀 Usage

Local Development

Run the application locally:

python app.py

The application will be available at http://localhost:7860

Hugging Face Spaces

This application is configured for deployment on Hugging Face Spaces with automatic model download:

1. Build Process: The build.py script automatically downloads the int4 model during Space build
2. Model Loading: Uses local model files when available, falls back to Hugging Face download
3. Caching: Model files are cached for faster subsequent runs

🎛️ Interface Features

Layout Structure

The interface follows the reference layout with:

• Title Section: Main heading and description
• Information Panels: Features and model information
• Input Section: Context and user input areas
• Advanced Settings: Collapsible parameter controls
• Chat Interface: Real-time conversation display

System Prompt

• Default: "Tu es TonicIA, un assistant francophone rigoureux et bienveillant."
• Editable: Users can customize the system prompt to define the assistant's personality
• Real-time: Changes take effect immediately for new conversations

Generation Parameters

• Max Length: Maximum number of tokens to generate (64-2048)
• Temperature: Controls randomness in generation (0.01-1.0)
• Top-p: Nucleus sampling parameter (0.1-1.0)
• Enable Thinking: Enable reasoning mode with thinking tags
• Advanced Settings: Collapsible panel for fine-tuning

🔧 Technical Details

Model Loading Strategy

The application uses a smart loading strategy:

1. Local Check: First checks if full model files exist locally
2. Local Loading: If available, loads from ./model folder
3. Fallback Download: If not available, downloads from Hugging Face
4. Tokenizer: Always uses main repo for chat template and configuration

Build Process

For Hugging Face Spaces deployment:

1. Build Script: build.py runs during Space build
2. Model Download: download_model.py downloads full model files
3. Local Storage: Model files stored in ./model directory
4. Fast Loading: Subsequent runs use local files

Chat Template Integration

The application uses the custom chat template from the model, which supports:

• System prompt integration
• User and assistant message formatting
• Thinking mode with <think> tags
• Proper conversation flow management

Memory Optimization

• Uses full fine-tuned model for maximum quality
• Automatic device detection (CUDA/CPU)
• Efficient tokenization and generation
• Float16 precision on GPU for optimal performance

📝 Example Usage

1. Basic Conversation:

   • Add context in the system prompt area
   • Type your message in the user input box
   • Click the generate button to start chatting

2. Customizing System Prompt:

   • Edit the context in the dedicated text area
   • Changes apply to new messages immediately
   • Example: "Tu es un expert en programmation Python."

3. Advanced Settings:

   • Check the "Advanced Settings" checkbox
   • Adjust generation parameters as needed
   • Enable/disable thinking mode

4. Real-time Chat:

   • Messages appear in the chat interface
   • Conversation history is maintained
   • Responses are generated using the model's chat template

🐛 Troubleshooting

Common Issues

1. Model Loading Errors:

   • Ensure you have sufficient RAM (8GB+ recommended)
   • Check your internet connection for model download
   • Verify all dependencies are installed

2. Generation Errors:

   • Try reducing the "Max Length" parameter
   • Adjust temperature and top-p values
   • Check the console for detailed error messages

3. Performance Issues:

   • The full model provides maximum quality but requires more memory
   • GPU acceleration recommended for optimal performance
   • Consider reducing model parameters if memory is limited

4. System Prompt Issues:

   • Ensure the system prompt is not too long (max 1000 characters)
   • Check that the prompt follows the expected format

5. Build Process Issues:

   • Check that download_model.py runs successfully
   • Verify that model files are downloaded to ./int4 directory
   • Ensure sufficient storage space for model files

📄 License

This project is licensed under the MIT License. The underlying model is licensed under Apache 2.0.

🙏 Acknowledgments

• Model: Tonic/petite-elle-L-aime-3-sft
• Base Model: SmolLM3-3B by HuggingFaceTB
• Training Data: legmlai/openhermes-fr
• Framework: Gradio, Transformers, PyTorch
• Layout Reference: Tonic/Nvidia-OpenReasoning

🔗 Links

• Model on Hugging Face
• Chat Template
• Original App Reference

---

Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference