BUG: Invalid EMBEDDING_MAX_BATCH_SIZE Value Triggers ValueError in backend/config.py

[itzsudipta]

Is there an existing issue for this?

• I have searched the existing issues

What happened?

📌 Issue Overview

EMBEDDING_MAX_BATCH_SIZE Crashes Application if Set to Non-Integer Value

🔍 Steps to Reproduce

1. Set the environment variable EMBEDDING_MAX_BATCH_SIZE to a non-integer value (e.g., abc or leave it empty).
2. Run the application.
3. The application attempts to convert the value to int in backend/config.py.
4. Show the error.

🎯 Expected Behavior

The application should handle invalid or missing EMBEDDING_MAX_BATCH_SIZE values gracefully and fall back to a safe default without crashing.

🚨 Actual Behavior

If EMBEDDING_MAX_BATCH_SIZE is set to a non-integer value or is empty, the application crashes with a ValueError in backend/config.py.

📷 Screenshot

💡 Suggested Improvements

• Add error handling in backend/config.py to catch invalid EMBEDDING_MAX_BATCH_SIZE values.
• Use the default value if parsing fails.

Record

• I agree to follow this project's Code of Conduct
• I want to work on this issue

[bnyashwanth]

Hi maintainers! 👋

I would like to work on this issue.

It looks like the EMBEDDING_MAX_BATCH_SIZE in backend/config.py is currently setting a value that exceeds the expected limit, which triggers the ValueError during initialization.

I can fix this by:

1. Investigating the validator logic in config.py.
2. Adjusting the default value or adding a constraint check to ensure the batch size remains within valid limits.

Could you please assign this issue to me? I can submit a PR for this by tomorrow.

[itzsudipta]

@smokeyScraper Hi there! I created issue #244 and have just uploaded my fix. I noticed a concurrent PR was opened, but I would ask you to consider my PR because it includes error-handling logic that prevents the server from failing to start due to configuration typos.

[bnyashwanth]

Hi @itzsudipta, thanks for pointing this out and for working on the fix 👍
I’ve pushed PR #247 with a defensive fallback for EMBEDDING_MAX_BATCH_SIZE to prevent startup crashes when the env var is invalid or missing. Happy to adjust or improve it if needed, or even align with your approach if maintainers prefer a single consolidated solution.
Leaving it to the maintainers to decide which PR fits best — I’m happy to help either way 🙂