merge sv4d changes: 1. reduce memory consumption (40G -> 20G) and speed up (500s -> 200s) 2. add gradio demo #223

	name: Test inference

	on:
	pull_request:
	push:
	branches:
	- main

	jobs:
	test:
	name: "Test inference"
	# This action is designed only to run on the Stability research cluster at this time, so many assumptions are made about the environment
	if: github.repository == 'stability-ai/generative-models'
	runs-on: [self-hosted, slurm, g40]
	steps:
	- uses: actions/checkout@v3
	- name: "Symlink checkpoints"
	run: ln -s ${{vars.SGM_CHECKPOINTS_PATH}} checkpoints
	- name: "Setup python"
	uses: actions/setup-python@v4
	with:
	python-version: "3.10"
	- name: "Install Hatch"
	run: pip install hatch
	- name: "Run inference tests"
	run: hatch run ci:test-inference --junit-xml test-results.xml
	- name: Surface failing tests
	if: always()
	uses: pmeier/pytest-results-action@main
	with:
	path: test-results.xml
	summary: true
	display-options: fEX
	fail-on-empty: true

Provide feedback