Fix setting fp16 dtype in AnimateDiff convert script. #7127
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
sayakpaul
reviewed
Feb 28, 2024
| @@ -48,4 +48,4 @@ def get_args(): | |||
| # skip loading position embeddings | |||
| adapter.load_state_dict(conv_state_dict, strict=False) | |||
| adapter.save_pretrained(args.output_path) | |||
| adapter.save_pretrained(args.output_path, variant="fp16", torch_dtype=torch.float16) | |||
| adapter.to(torch.float16).save_pretrained(args.output_path, variant="fp16") | |||
There was a problem hiding this comment.
Shouldn't this be made configurable? Folks might not want to serialize the model in fp16.
|
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. |
sayakpaul
approved these changes
Feb 28, 2024
yiyixuxu
reviewed
Mar 5, 2024
| @@ -48,4 +49,6 @@ def get_args(): | |||
| # skip loading position embeddings | |||
| adapter.load_state_dict(conv_state_dict, strict=False) | |||
| adapter.save_pretrained(args.output_path) | |||
| adapter.save_pretrained(args.output_path, variant="fp16", torch_dtype=torch.float16) | |||
|
|
|||
| if args.save_fp16: | |||
There was a problem hiding this comment.
are we missing an else statement here?
There was a problem hiding this comment.
IIUC the scrtipt always serializes in the full precision. I think the script additionally serializes in FP16 when args.save_fp16 is True.
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Outdated suggestions cannot be applied.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?
save_pretraineddoesn't actually usetorch_dtypeto convert the checkpoint tofp16before saving. This PR adds a fix to the convert script so thatfp16weights are saved correctly for AnimateDiff modelsFixes # (issue)
Before submitting
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.