Repository navigation
Conversation
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes #1297. With PyTorch 2.6 the default of torch.load switched to weights_only=True, which breaks PolicyPersistence.restore whenever the policy was saved in the default Mode.POLICY. That mode deliberately pickles the whole policy object (so it can be reloaded without an environment), and weights_only=True refuses to unpickle arbitrary class globals like tianshou's policy classes, raising an UnpicklingError on load. The state-dict mode keeps working because it only contains tensors.
The fix opts out of weights_only for the POLICY-mode load only, since that file is by design a pickled object produced by tianshou itself; POLICY_STATE_DICT is left untouched. A round-trip regression test covering both modes was added.
Tested with pytest on the new test file: both modes pass after the change, while on the parent commit the POLICY case fails with the same Weights-only UnpicklingError reported in #1297.