fix(training): apply DROID RLDS shuffle only when enabled#997
Open
Bartok9 wants to merge 1 commit into
Open
Conversation
Salvage of Physical-Intelligence#746 by @zavidnyi — rebased to main. The constructor exposes shuffle= but always shuffled the concatenated dataset. Gate dataset.shuffle on the flag so evaluation can be deterministic.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
final_dataset.shuffle(...)on the existingshuffleconstructor argument indroid_rlds_dataset.py.Motivation
DroidRldsDataset(shuffle=False)still shuffled after dataset mixing. That breaks deterministic evaluation / debugging.Salvage of #746 by @zavidnyi — rebased to current
main(original PR conflicted).Verification
shuffle_buffer_sizecallself.shuffle = shufflenow matches actual pipeline behaviorAttribution
Salvage of #746 by @zavidnyi — rebased to main