- [ ] GPU linux runners with caching - [ ] windows GHA #1918 - [ ] windows GPU PRs #1560 - [ ] windows GPU anvil reporting in GHA (e.g. catching bugs like #1714 earlier) - [x] linux GPU #1729 - [x] unskip tests where appropriate #1731 - [ ] docker GPU - [ ] macos - [ ] all supported Python & numpy combinations #1914 (or just get on `conda-forge`) Probably closes #1205
conda-forge)Probably closes #1205