[Core] Harmonize single file ckpt model loading #6971
Merged
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
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
DN6
reviewed
Feb 14, 2024
| for pat in vae._keys_to_ignore_on_load_unexpected: | ||
| unexpected_keys = [k for k in unexpected_keys if re.search(pat, k) is None] | ||
|
|
||
| if len(unexpected_keys) > 0: |
There was a problem hiding this comment.
Small nit. Since unexpected_keys is always of type list we can just use
if unexpected_keys:
...There was a problem hiding this comment.
Oh, I followed what's done in modeling_utils.py to be consistent.
DN6
reviewed
Feb 14, 2024
| @@ -48,6 +48,7 @@ def build_sub_model_components( | |||
| load_safety_checker=False, | |||
| model_type=None, | |||
| image_size=None, | |||
| torch_dtype=None, | |||
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Outdated suggestions cannot be applied.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?
Currently, we make use of
set_module_tensor_to_devicein the single file loading utilities to speed up model loading time. However, insrc/models/modeling_utils.py, we already have a utility calledload_model_dict_into_meta()that abstracts the boilerplate code.diffusers/src/diffusers/models/modeling_utils.py
Line 133 in 0ca7b68
This PR updates the single file utilities to use
load_model_dict_into_meta()as well.I have run the single-file SLOW tests, but @DN6, feel free to do so on your end, too, just to be sure.