feat(vllm)!: upgrade vllm backend and refactor deployment #1311

	# Catch all the things we ignore in the e2e workflow
	name: pytest Skip Shim
	on:
	pull_request:
	paths:
	# Catch updates to the .github directory, unless it is the pytest related files
	- ".github/**"
	- "!.github/workflows/pytest.yaml"
	- "!.github/workflows/pytest-shim.yaml"

	# Catch docs and website things
	- "**.md"
	- "adr/**"
	- "docs/**"
	- "website/**"
	- "netlify.toml"

	# Catch generic github metadata files
	- "CODEOWNERS"
	- ".gitignore"
	- "LICENSE"

	# Catch LFAI-UI things
	- "src/leapfrogai_ui/**"
	- "packages/ui/**"

	permissions:
	contents: read

	# This is here to act as a shim for branch protection rules to work correctly.
	# This is ugly but this seems to be the best way to do this since:
	# - Job names in a workflow must be unique
	# - When paths are ignored not all jobs are reported to the branch protection rules
	# - Multiple jobs of the same name are still required by branch protection rules

	# For more info see below:
	# https://github.com/orgs/community/discussions/54877
	# https://docs.github.com/en/pull-requests/collaborating-with-pull-requests/collaborating-on-repositories-with-code-quality-features/troubleshooting-required-status-checks#handling-skipped-but-required-checks

	# Abort prior jobs in the same workflow / PR
	concurrency:
	group: pytest-skip-${{ github.ref }}
	cancel-in-progress: true


	jobs:
	pytest:
	runs-on: ubuntu-latest
	steps:
	- name: Skipped
	run: \|
	echo skipped

Provide feedback