making the script compatible with transformers 5.2+
#19
by mfarre - opened
Hello - I am working on the Vision Language Models book from O'Reilly and I would like to use qwen3 vl embedding in our video chapter.
In order to simplify how the readers load your model, it would be great to merge this patch so that the model loads properly with transformers 5.2+
thanks for reviewing and merging as soon as possible, it is highly time sensitive.
hey folks, please merge!!
Hey if you still need this, this should help, I built it for others in the situation I was: https://github.com/philmcginty/qwen3-vl-embedding-server