Add CLIP text model #643
Merged
Add CLIP text model #643
Conversation
This file contains 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
sogartar
force-pushed
the
clip-text-model
branch
2 times, most recently
from
d658a9f to
a6de6cb
Compare
rsuderman
previously requested changes
Dec 4, 2024
IanNod
reviewed
Dec 4, 2024
KyleHerndon
approved these changes
Dec 6, 2024
sogartar
force-pushed
the
clip-text-model
branch
3 times, most recently
from
526440f to
7bc98ef
Compare
sogartar
dismissed
rsuderman’s stale review
I will address model changes after adding the IREE tests so that we make refactoring while the code is under test.
Ports the CLIP text model from Hugging Face. Add numeric verification tests for the various components of the stack when executing in eager mode. Verifications are made for float32 and bfloat16. There are tests for toy-sized components and the whole model as well as the Large pretrained variant. These tests does not include testing with IREE. Functionalities for mask creation are not yet ported.
sogartar
force-pushed
the
clip-text-model
branch
from
7bc98ef to
0d30d62
Compare
monorimet
pushed a commit
that referenced
this pull request
Jan 8, 2025
Ports the CLIP text model from Hugging Face. This is the first iteration so not much is changed from the original model. Things like dropout and checkpointing are removed. Add numeric verification tests for the various components of the stack when executing in eager mode. Verifications are made for float32 and bfloat16. There are tests for toy-sized components and the whole model as well as the Large pretrained variant. These tests does not include testing with IREE. Functionalities for mask creation are not yet ported.
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.
Ports the CLIP text model from Hugging Face. This is the first iteration so not much is changed from the original model. Things like dropout and checkpointing are removed.
Add numeric verification tests for the various components of the stack when executing in eager mode. Verifications are made for float32 and bfloat16. There are tests for toy-sized components and the whole model as well as the Large pretrained variant.
These tests does not include testing with IREE.
Functionalities for mask creation are not yet ported.