Hugging Face's logo

Qwen
/
Qwen2.5-VL-72B-Instruct-AWQ

Image-Text-to-Text
Transformers
Safetensors
English
qwen2_5_vl
multimodal
conversational
text-generation-inference
Inference Endpoints
4-bit precision
awq
Model card Files Community
11

Update preprocessor_config.json

#7
by Benasd - opened
base: refs/heads/main
←
from: refs/pr/7
Discussion Files changed
+1
-1
Benasd

Change Qwen2_5_VLImageProcessor to Qwen2VLImageProcessor.
due to the merged PR in Transformers https://github.com/huggingface/transformers/pull/36164.

WIthout his PR, the model with cause an error in the latest transformers.

Update preprocessor_config.jsonb2140f7c
Benasd

This was fixed in a959e37.
And was accidentally reverted in 666b8f5.

Benasd
β€’
edited 4 days ago

This fixes #5 and #8

Benasd

For anyone encountering the same issue, I've created a fork of the model with the fix applied.
This is a temporary workaround. The forked model will be deleted once this PR is merged.
https://huggingface.co/Benasd/Qwen2.5-VL-72B-Instruct-AWQ-fix

yunusserhat
β€’
edited 4 days ago

thanks @Benasd

Ready to merge
This branch is ready to get merged automatically.

Sign up or log in to comment