HiDream: Unlock max length for some text encoders (t5 and llama) #7685
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
|
There is no effective limit to the llama model currently and your 77 limit on the clips is wrong. |
|
Apologize. I got something wrong. |
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Outdated suggestions cannot be applied.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.
Improvement of the last PR: #7676
Currently, T5 and Llama text encoders have restrictive max length limits (77 tokens), which can hinder performance for long-context tasks. This PR removes those constraints, allowing full utilization of the models' capabilities.
Key changes:
Note:
The DiT model cannot generate high-quality images despite having long contexts. This may be because DiT wasn't trained for long tokens.