https://github.com/tstellar approved this pull request. LGTM. If the speed is still an issue, the other solution would be to have separate workflow files for the clang and llvm docs. https://github.com/llvm/llvm-project/pull/69763