201 ml streaming endpoint

[grillazz]

No description provided.

[Copilot]

Pull Request Overview

This PR introduces a new streaming endpoint for chat-based interactions with a local large language model (LLM) and updates related dependencies, Docker compose settings, and documentation.

• Added a new asynchronous chat client test in tests/chat.py.
• Introduced a new LLM streaming service via app/services/llm.py and its corresponding API endpoint in app/api/ml.py, with integration updates in app/main.py and README.md.
• Upgraded project dependencies in pyproject.toml and adjusted Docker networking for app, db, and redis containers in compose.yml.

Reviewed Changes

Copilot reviewed 8 out of 9 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
tests/chat.py	Added a basic async chat client for testing the streaming endpoint.
pyproject.toml	Updated project version and bumped several dependency versions.
compose.yml	Configured host networking for containers.
app/services/llm.py	Added a new LLM service for streaming chat responses.
app/main.py	Introduced the new ml router and updated API version.
app/api/ml.py	Created a StreamingResponse endpoint for LLM chat.
README.md	Documented the new LLM integration and usage instructions.

Files not reviewed (1)

• .env: Language not supported

Comments suppressed due to low confidence (1)

compose.yml:4

• Using host networking may reduce container isolation and could expose additional security risks. Please confirm that this configuration is intentional and acceptable for the deployment environment.

network_mode: host

[Copilot]

Avoid silently passing exceptions using a bare 'except Exception:' block. Consider logging the error details or handling the exception explicitly to aid in debugging.

Suggested change

	except Exception:
	pass
	except Exception as e:
	logging.exception("Error processing streamed line: %s", line)

[Copilot]

Using the synchronous input() call inside an async function may block the event loop. Consider using an asynchronous input strategy or executing the blocking call in a separate thread to avoid potential performance issues.

Suggested change

	prompt = input("\nYou: ")
	prompt = await anyio.to_thread.run_sync(input, "\nYou: ")