Align timeout settings of Code Generation request
Problem
Currently, Code Generation requests times out in 30 seconds. This is not aligned with our GitLab-Rails/Puma settings.
Proposal
We set 60 sec to the timeout settings in the following places:
- Workhorse
/api/v4/code_suggestions/completions
https://log.gprd.gitlab.net/app/r/s/2hjiL - AI Gateway Anthropic client https://gitlab.com/gitlab-org/modelops/applied-ml/code-suggestions/ai-assist/-/blob/main/ai_gateway/models/anthropic.py#L71