Tutorial: Ollama Chat Example (E02)

This tutorial corresponds to the example file examples/E02_ollama_chat_example.py.

It demonstrates the simplest way to interact with a Large Language Model using Genie. It shows how to: - Use FeatureSettings to quickly configure Genie to use an Ollama provider (llm="ollama"). - Specify the model to use (e.g., mistral:latest). - Use genie.llm.chat() to send a message to the LLM and receive a response.

Prerequisite: Ensure you have an Ollama instance running and the specified model pulled (e.g., ollama pull mistral).

Example Code

examples/E02_ollama_chat_example.py

""" Example: Ollama Chat with Genie Facade

---

This example demonstrates how to use the Genie facade to chat with an Ollama-hosted LLM (e.g., Mistral).

To Run: 1. Ensure Genie Tooling is installed (poetry install --all-extras). 2. Ensure Ollama is running and the model 'mistral:latest' (or your chosen model) is pulled: ollama serve ollama pull mistral 3. Run from the root of the project: poetry run python examples/E02_ollama_chat_example.py """ import asyncio import logging from typing import List, Optional

from genie_tooling.config.features import FeatureSettings from genie_tooling.config.models import MiddlewareConfig from genie_tooling.genie import Genie from genie_tooling.llm_providers.types import ChatMessage

async def run_ollama_chat_demo(): print("--- Ollama Chat Example ---")

app_config = MiddlewareConfig(
    features=FeatureSettings(
        llm="ollama",
        llm_ollama_model_name="mistral:latest" # Or any other model you have pulled
    )
)

genie: Optional[Genie] = None
try:
    print("\nInitializing Genie with Ollama...")
    genie = await Genie.create(config=app_config)
    print("Genie initialized!")

    messages: List[ChatMessage] = [
        {"role": "user", "content": "Hello, Ollama! Tell me a fun fact about llamas."}
    ]
    print(f"\nSending to Ollama: {messages[0]['content']}")

    response = await genie.llm.chat(messages)
    # Default provider is Ollama as per FeatureSettings

    assistant_response = response.get("message", {}).get("content", "No content received.")
    print(f"\nOllama says: {assistant_response}")

except Exception as e:
    print(f"\nAn error occurred: {e}")
    logging.exception("Ollama chat error details:")
finally:
    if genie:
        await genie.close()
        print("\nGenie torn down.")

if name == "main": logging.basicConfig(level=logging.INFO) # logging.getLogger("genie_tooling").setLevel(logging.DEBUG) asyncio.run(run_ollama_chat_demo())