-
Notifications
You must be signed in to change notification settings - Fork 26.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Fix Llava conversion for LlavaQwen2ForCausalLM with Clip vision tower #33613
base: main
Are you sure you want to change the base?
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hi @Isotr0py, could you show the running, output and sucessful conversion of a previous Qwen checkpoint with these changes?
Otherwise it looks OK to me, but @zucchini-nlp will know which checkpoints we should verify are still compatible with these updates
I printed the converted model and config for llava-interleaveThe convert command for python transformers/src/transformers/models/llava/convert_llava_weights_to_hf.py --text_model_id Qwen/Qwen1.5-7B-Chat --vision_model_id google/siglip-so400m-patch14-384 --old_state_dict_id lmms-lab/llava-next-interleave-qwen-7b The converted model and config:
|
What does this PR do?
Fixes # (issue)
LlavaQwen2ForCausalLM
architecture, but they useClip
instead ofSiglip
as vision_tower.Qwen
withSiglip
, which caused the converted model with wrong vision tower configBefore submitting
Pull Request section?
to it if that's the case.
documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
@amyeroberts @zucchini-nlp
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.