Skip to content

Token chunking for CLIP in SD1.5 and SDXL#1405

Draft
FauxCodes wants to merge 3 commits intoNerogar:masterfrom
FauxCodes:feature/clip-token-chunking
Draft

Token chunking for CLIP in SD1.5 and SDXL#1405
FauxCodes wants to merge 3 commits intoNerogar:masterfrom
FauxCodes:feature/clip-token-chunking

Conversation

@FauxCodes
Copy link
Copy Markdown
Contributor

Implementation of #430

A hack that enables using more than 75 tokens for captions in SDXL and SD1.5.

@@ -0,0 +1,228 @@
from contextlib import nullcontext
Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This file exists for easier testing. To remove after merging https://github.com/Nerogar/mgds/pull/48/changes in mgds and before moving this PR out of draft.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant