Change image_processor_type to "Qwen2VLImageProcessor"

by ljoana - opened 14 days ago

←

ljoana

MLX Community org 14 days ago

Based on this issue https://github.com/huggingface/transformers/issues/36193 it seems that HF had updated the hub files and forced the model to use the Qwen2VL slow processor.

pcuenq

MLX Community org 12 days ago

Yes, the transformers release expects that class name, so this PR fixes pre-processing. Thank you, @ljoana !

pcuenq changed pull request status to merged 12 days ago

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment