Feat: Image Processor for CLIP #1
Merged
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
gboduljak
reviewed
Jan 16, 2024
| @@ -0,0 +1,28 @@ | |||
| CLIP_IMAGE_CONFIG = {'openai/clip-vit-base-patch32': { | |||
Owner
There was a problem hiding this comment.
We want to pull these from HuggingFace instead of hardcoding constants. This solves the problem, but hurts maintenance.
gboduljak
reviewed
Jan 16, 2024
| @@ -0,0 +1,5 @@ | |||
| from typing import List, Union | |||
|
|
|||
| ImageInput = Union[ | |||
Owner
There was a problem hiding this comment.
We can include this within the image_processor.
gboduljak
reviewed
Jan 16, 2024
| @@ -0,0 +1,252 @@ | |||
| from image_config import CLIP_IMAGE_CONFIG | |||
| from jmage_constants import ImageInput | |||
gboduljak
approved these changes
Jan 16, 2024
Owner
gboduljak
left a comment
gboduljak
left a comment
There was a problem hiding this comment.
Thanks for the contribution. I will merge your change because my comments are nits. However, please ensure we test correctness as well. It is quite easy to mess this up. The errors in preprocessing can propagate further and can be difficult to debug. Here is a concrete test:
def test_image_processor():
mx_image_proc = CLIPImageProcessor('openai/clip-vit-base-patch32')
tf_image_proc = transformers.CLIPImageProcessor.from_pretrained(
'openai/clip-vit-base-patch32')
image = Image.open("cats.jpeg")
data = mx_image_proc([image])
mx_pixels = mx.array(data['pixel_values'])
tf_pixels = mx.array(np.array(tf_image_proc([image])[
'pixel_values']).transpose((0, 2, 3, 1)))
assert mx.array_equal(mx_pixels, tf_pixels)
awni
pushed a commit
that referenced
this pull request
Jan 31, 2024
@nkasmanoff: * clip image processor * added example usage
gboduljak
added a commit
that referenced
this pull request
Feb 3, 2024
* probably approximatelly correct CLIPTextEncoder * implemented CLIPEncoderLayer as built-in nn.TransformerEncoderLayer * replaced embedding layer with simple matrix * implemented ViT * added ViT tests * fixed tests * added pooler_output for text * implemented complete CLIPModel * implemented init * implemented convert.py and from_pretrained * fixed some minor bugs and added the README.md * removed tokenizer unused comments * removed unused deps * updated ACKNOWLEDGEMENTS.md * Feat: Image Processor for CLIP (#1) @nkasmanoff: * clip image processor * added example usage * refactored image preprocessing * deleted unused image_config.py * removed preprocessing port * added dependency to mlx-data * fixed attribution and moved photos to assets * implemented a simple port of CLIPImageProcessor * review changes * PR review changes * renamed too verbose arg * updated README.md * nits in readme / conversion * simplify some stuff, remove unneeded inits * remove more init stuff * more simplify * make test a unit test * update main readme * readme nits --------- Co-authored-by: Noah Kasmanoff <[email protected]> Co-authored-by: Awni Hannun <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 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.
No description provided.