Language Model Management Example

This guide provides a professional, step-by-step walkthrough for registering, setting, and using a language model with the LLMManager in the agenticaiframework package. It is intended for developers integrating custom or third-party LLMs into their applications.

18+ LLM Providers Supported

Part of 400+ modules supporting OpenAI, Anthropic, Gemini, Azure, AWS Bedrock, Ollama, and more. See LLM Documentation.

Prerequisites & Configuration

• Installation: Ensure agenticaiframework is installed and accessible in your Python environment.
• No additional configuration is required for this example.
• Python Version: Compatible with Python 3.10+.

Code

import logging

logger = logging.getLogger(__name__)

from agenticaiframework.llms import LLMManager

if __name__ == "__main__":
    llm_manager = LLMManager()

    # Register a demo LLM model
    llm_manager.register_model("demo-llm", lambda prompt: f"[Demo LLM Response to: {prompt}]")

    # Set the active model
    llm_manager.set_active_model("demo-llm")

    # Generate text using the active model
    logger.info("Generated Text:", llm_manager.generate("Explain the concept of machine learning in simple terms."))

    # List available models
    logger.info("Available Models:", list(llm_manager.models.keys()))

Step-by-Step Execution

1. Import the Class Import LLMManager from agenticaiframework.llms.

2. Instantiate the Manager Create an instance of LLMManager to handle model registration and usage.

3. Register a Model Use register_model with:

4. name: Unique identifier for the model.

5. callable: A function or lambda that takes a prompt and returns generated text.

6. Set the Active Model Use set_active_model to select which model will be used for generation.

7. Generate Text Call generate with a prompt to produce output from the active model.

8. List Available Models Access the models dictionary to see all registered models.

Best Practice: Wrap third-party LLM APIs in a callable function to standardize the interface for LLMManager.

Expected Input

No user input is required; the script uses hardcoded values for demonstration purposes. In production, prompts could be dynamically generated from user input, workflows, or other runtime data.

Expected Output

[YYYY-MM-DD HH:MM:SS] [LLMManager] Registered LLM model 'demo-llm'
[YYYY-MM-DD HH:MM:SS] [LLMManager] Active LLM model set to 'demo-llm'
Generated Text: [Demo LLM Response to: Explain the concept of machine learning in simple terms.]
Available Models: ['demo-llm']

How to Run

Run the example from the project root:

python examples/llms_example.py

If installed as a package, you can also run it from anywhere:

python -m examples.llms_example

Tip: Use LLMManager to manage multiple models and switch between them dynamically based on task requirements.