Cap max_tokens and n to prevent local API DoS

[arditbe]

My Changes

Added server-side upper bounds for max_tokens and n in BaseCompletionRequest::parseImpl.
Requests exceeding these limits now return 400 via InvalidRequestError, preventing memory and CPU exhaustion in
/v1/completions and /v1/chat/completions.

Issue ticket number and link

Fixes #3635

Checklist before requesting a review

• I have performed a self-review of my code.
• If it is a core feature, I have added thorough tests.
• I have added thorough documentation for my code.
• I have tagged PR with relevant project labels.
• If this PR addresses a bug, I have provided both a screenshot/video of the original bug and the working solution.

Demo

N/A (validation change)

Steps to Reproduce

1. Start GPT4All with API server enabled.
2. Send a request with very large max_tokens or n.
3. Observe the server now responds with 400 instead of consuming excessive resources.

Notes

Limits used: max_tokens <= 4096, n <= 8. Happy to adjust per maintainer preference.

[arditbe]

Hi! This PR adds server-side caps for max_tokens and n in BaseCompletionRequest to prevent local API DoS on /v1/completions and /v1/chat/completions. Fixes #3635. Happy to adjust limits or add tests if you prefer.