Skip to main content

Advanced Topics

Deep dive into LLMBoost's advanced capabilities for production deployments and custom integrations.

What's in This Section​

This section covers advanced use cases and deployment patterns beyond the core features. These topics are designed for users who:

  • Want to integrate LLMBoost directly into Python applications
  • Need to combine LLMBoost Hub with advanced Docker workflows
  • Require custom UI integrations
  • Need fine-grained performance tuning

Topics Overview​

LLMBoost Hub Advanced Usage​

Combine LLMBoost Hub (lbh) with advanced Docker workflows and custom configurations.

When to use:

  • Need custom volume mounts or network configurations
  • Managing models in shared storage (e.g., /lustre1/$USER/llm_models)
  • Combining lbh convenience with manual Docker control
  • Advanced multi-container setups

You'll learn:

  • Using lbh run and lbh attach for container management
  • Environment variables for custom configurations
  • Integration with existing Docker workflows
  • Advanced storage and networking options

In-Process Python SDK​

Integrate LLMBoost directly into your Python applications without a separate server.

When to use:

  • Embedding inference in existing Python apps
  • Batch processing workflows
  • Research and experimentation
  • Maximum control over inference pipeline

You'll learn:

  • Direct SDK usage patterns
  • Async and sync interfaces
  • Custom prompt formatting
  • Performance tuning parameters

OpenWebUI Integration​

Deploy LLMBoost with OpenWebUI for a chat interface.

When to use:

  • Internal team chatbot deployments
  • Testing and demonstration
  • User-friendly inference interface
  • Rapid prototyping

You'll learn:

  • OpenWebUI setup and configuration
  • Connecting to LLMBoost backend
  • Custom UI configurations

Prerequisites​

Before diving into advanced topics, ensure you're familiar with:

Getting Help​

For advanced deployment assistance:

  • Documentation: Refer to specific topic pages for detailed guides
  • Support: Contact contact@mangoboost.io

Ready to explore? Choose a topic from the sidebar to get started!