Conversation
…pu] extra Consumers that only need dataset definitions (bitmind-subnet's validator cache imports gasbench.dataset.config and nothing else) currently drag in the entire inference stack — onnxruntime-gpu, torch, timm, fvcore, decord, modelscope, etc. Invert the packaging: - Base deps are now just pyyaml; the full benchmark stack moves to the gpu extra (the name bitmind-subnet already requests). The dev extra includes gasbench[gpu]. - gasbench/__init__.py and gasbench/dataset/__init__.py resolve their benchmark-stack exports lazily (PEP 562) so importing the registry never touches torch/onnxruntime. Missing-dep access raises a guided ImportError pointing at gasbench[gpu]; the CLI prints the same hint. - New base-install CI job asserts the contract: registry importable with only pyyaml installed, and no heavy dist leaks into base. - README install section updated (also removes the reference to the long-gone [cpu] extra). No behavior change for gasbench[gpu] installs — all public exports resolve to the same objects.
refactor: make base install registry-only, move benchmark stack to [gpu] extra
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Consumers that only need dataset definitions (bitmind-subnet's validator cache imports gasbench.dataset.config and nothing else) currently drag in the entire inference stack — onnxruntime-gpu, torch, timm, fvcore, decord, modelscope, etc. Invert the packaging:
No behavior change for gasbench[gpu] installs — all public exports resolve to the same objects.