Support for manual CLIP loading in StableDiffusionPipeline - txt2img. #3832
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
patrickvonplaten
approved these changes
Jun 21, 2023
|
The documentation is not available anymore as the PR was closed or merged. |
sayakpaul
reviewed
Jun 22, 2023
src/diffusers/loaders.py
Outdated
Comment on lines
1342
to
1345
| clip_text_model (`transformers.models.clip.modeling_clip.CLIPTextModel`, *optional*, defaults to `None`): | ||
| An instance of `CLIPTextModel` to use. If this parameter is `None`, the function will load a new instance of `CLIPTextModel`, if needed. | ||
| clip_tokenizer (`transformers.models.clip.tokenization_clip.CLIPTokenizer`, *optional*, defaults to `None`): | ||
| An instance of `CLIPTokenizer` to use. If this parameter is `None`, the function will load a new instance of `CLIPTokenizer`, if needed. |
There was a problem hiding this comment.
Could we maybe follow something similar to for the docstrings here?
sayakpaul
reviewed
Jun 22, 2023
Comment on lines
1076
to
1079
| clip_text_model (`transformers.models.clip.modeling_clip.CLIPTextModel`, *optional*, defaults to `None`): | ||
| An instance of `CLIPTextModel` to use. If this parameter is `None`, the function will load a new instance of `CLIPTextModel`, if needed. | ||
| clip_tokenizer (`transformers.models.clip.tokenization_clip.CLIPTokenizer`, *optional*, defaults to `None`): | ||
| An instance of `CLIPTokenizer` to use. If this parameter is `None`, the function will load a new instance of `CLIPTokenizer`, if needed. |
sayakpaul
approved these changes
Jun 22, 2023
|
Regarding the failing test here, could you run |
|
@patrickvonplaten |
yoonseokjin
pushed a commit
to yoonseokjin/diffusers
that referenced
this pull request
Dec 25, 2023
…huggingface#3832) * Support for manual CLIP loading in StableDiffusionPipeline - txt2img. * Update src/diffusers/pipelines/stable_diffusion/convert_from_ckpt.py * Update variables & according docs to match previous style. * Updated to match style & quality of 'diffusers' --------- Co-authored-by: Patrick von Platen <[email protected]>
AmericanPresidentJimmyCarter
pushed a commit
to AmericanPresidentJimmyCarter/diffusers
that referenced
this pull request
Apr 26, 2024
…huggingface#3832) * Support for manual CLIP loading in StableDiffusionPipeline - txt2img. * Update src/diffusers/pipelines/stable_diffusion/convert_from_ckpt.py * Update variables & according docs to match previous style. * Updated to match style & quality of 'diffusers' --------- Co-authored-by: Patrick von Platen <[email protected]>
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.
What does this PR do?
Fixes #3822
This pull request introduces a feature that enhances the loading mechanism of the
CLIPmodel when used in conjunction withStableDiffusionPipeline.from_ckpt(). This PR affects only thetxt2imgpart of the mentioned pipeline. Users now have the flexibility to manually load the CLIP model and tokenizer, thereby bypassing the force loading behavior. This PR resolves the challenge of creating a fully portable solution, as the CLIP model and tokenizer would previously end up in the cache.With this enhancement, users can now specify their desired CLIP model and tokenizer location as follows:
It is important to note that this pull request does not impact the behavior if users choose not to provide any parameters for
clip_text_modelorclip_tokenizer. In such cases, the code will behave as it did before the PR.Before submitting
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
@patrickvonplaten
@sayakpaul