Open
Conversation
Replace the DictWrapper class with a plain Python dict in
LLM.create_feature_dict(). The DictWrapper existed solely to prevent
input/output features from being registered as nn.Module submodules
(which would confuse DeepSpeed). A plain dict achieves the same goal
more simply -- it is not an nn.Module, so PyTorch and DeepSpeed never
see it.
Key changes:
- Remove the DictWrapper class (42 lines)
- Remove unused LudwigFeatureDict import
- create_feature_dict() now returns {}
- Fix items()[0] indexing (list->dict_items view) with next(iter(...))
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
DictWrapperclass fromludwig/models/llm.pyand replace it with a plain PythondictDictWrapperexisted to prevent input/output features from being registered asnn.Modulesubmodules (which would confuse DeepSpeed). A plaindictachieves the same goal more simply since it is not annn.Moduleitems()[0]indexing (which relied onLudwigFeatureDict.items()returning a list) to usenext(iter(...))compatible withdict_itemsviewsTest plan