runway: Increase memory per container from 4GB to 8GB, max_instances from 100 to 200
What does this merge request do and why?
Following https://gitlab.com/gitlab-org/modelops/applied-ml/code-suggestions/ai-assist/-/issues/374, we're still seeing elevated memory utilization and request concurrency:
We bumped the limits here: !489 (merged). This MR further raises those limits so that we have more room to breathe.
Edited by Igor