[GH-ISSUE #69] Input size #63

New issue

Closed

opened 2026-10-31 21:10:23 +00:00 by bsakaguchi · 2 comments

bsakaguchi commented

2026-10-31 21:10:23 +00:00

Owner

Originally created by @qwertyforce on GitHub (Mar 23, 2021).
Original GitHub issue: https://github.com/openai/CLIP/issues/69

Is it possible to use this neural network with images larger than 224x224 without cropping/resizing?

Originally created by @qwertyforce on GitHub (Mar 23, 2021). Original GitHub issue: https://github.com/openai/CLIP/issues/69 Is it possible to use this neural network with images larger than 224x224 without cropping/resizing?

bsakaguchi closed this issue

2026-10-31 21:10:24 +00:00

bsakaguchi commented

2026-10-31 21:10:24 +00:00

Author

Owner

@jongwook commented on GitHub (Mar 24, 2021):

Not out-of-the-box, but you might want to try the 2-D interpolation technique similar to what Vision Transformer (Section 3.2) did for fine-tuning on higher resolution. This is how we trained our largest model (ViT-L/14-336px).

@jongwook commented on GitHub (Mar 24, 2021): Not out-of-the-box, but you might want to try the 2-D interpolation technique similar to what [Vision Transformer (Section 3.2)](https://arxiv.org/pdf/2010.11929.pdf) did for fine-tuning on higher resolution. This is how we trained our largest model (ViT-L/14-336px).

bsakaguchi commented

2026-10-31 21:10:24 +00:00

Author

Owner

@qwertyforce commented on GitHub (Mar 24, 2021):

Not out-of-the-box, but you might want to try the 2-D interpolation technique similar to what Vision Transformer (Section 3.2) did for fine-tuning on higher resolution. This is how we trained our largest model (ViT-L/14-336px).

Thanks!

@qwertyforce commented on GitHub (Mar 24, 2021): > Not out-of-the-box, but you might want to try the 2-D interpolation technique similar to what [Vision Transformer (Section 3.2)](https://arxiv.org/pdf/2010.11929.pdf) did for fine-tuning on higher resolution. This is how we trained our largest model (ViT-L/14-336px). Thanks!

bsakaguchi referenced this issue

2026-10-31 21:11:30 +00:00

[GH-ISSUE #146] Why the results of cosine similarities between the corresponding image and text features may appear negative numbers and NaN? #128

bsakaguchi referenced this issue

2026-10-31 21:13:31 +00:00

[GH-ISSUE #340] RN50x16, RN50x4, ViT-B/32 output nan #294