This repository was archived by the owner on Sep 10, 2025. It is now read-only.
[Feature] Add ability to load HF checkpoints into T5 model #1918
Conversation
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
Nayef211
reviewed
Sep 30, 2022
joecummings
changed the title
Nayef211
approved these changes
Oct 4, 2022
Contributor
Nayef211
left a comment
Nayef211
left a comment
There was a problem hiding this comment.
LGTM! Thanks for adding this capability @joecummings! I would recommend pulling in these changes and ensuring it works with internal Meta infra before merging in the PR. 😄
| ) -> T5Model: | ||
| """Build T5Model model from a HuggingFace checkpoint. | ||
| Note: Only works with Huggingface models saved in the PyTorch format. Will not work \ |
Contributor
There was a problem hiding this comment.
nit: do we need the \ in the docstrings?
Member
Author
There was a problem hiding this comment.
Sorry - not entirely sure what you're referring to here?
Contributor
There was a problem hiding this comment.
Sorry updated the comment. Looks like backslash characters don't show up when you put quotes around them which was a noob mistake 😅
joecummings
force-pushed
the
hf-to-pt-t5
branch
from
fbd2af3 to
a624031
Compare
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
3 participants
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Outdated suggestions cannot be applied.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.
Add the ability to load a T5 model from pretrained HuggingFace weights
Changes
build_model_from_huggingface_ckptstatic method to theT5Bundler.Testing
Future considerations: This does not support loading from external URLs - planning on adding this in a follow-up Diff from the internal FB side so I can test with Manifold (+ possible GDrive, static public links, Github).