Fix config
#4
by
Xenova
HF staff
- opened
from transformers import CLIPVisionModelWithProjection
m = CLIPVisionModelWithProjection.from_pretrained('hf-internal-testing/tiny-random-CLIPModel', revision='main') # FAILS
m = CLIPVisionModelWithProjection.from_pretrained('hf-internal-testing/tiny-random-CLIPModel', revision='refs/pr/4') # WORKS
Xenova
changed pull request status to
merged