[GH-ISSUE #303] Can we change CLIP Vision encoder ? #259

New issue

Open

opened 2026-10-31 21:13:05 +00:00 by bsakaguchi · 1 comment

bsakaguchi commented

2026-10-31 21:13:05 +00:00

Owner

Originally created by @TitanAxeSolutionsLimited on GitHub (Nov 21, 2022).
Original GitHub issue: https://github.com/openai/CLIP/issues/303

Good Day,
I was wondering if it's possible to change the vision encoder of CLIP Model? More specifically, can we use GAN as vision encoder and the extracted features are feed to the vision encoder of CLIP? I want to perform the image classification on the dataset which has very minor differences for each class.

Originally created by @TitanAxeSolutionsLimited on GitHub (Nov 21, 2022). Original GitHub issue: https://github.com/openai/CLIP/issues/303 Good Day, I was wondering if it's possible to change the vision encoder of CLIP Model? More specifically, can we use GAN as vision encoder and the extracted features are feed to the vision encoder of CLIP? I want to perform the image classification on the dataset which has very minor differences for each class.

bsakaguchi commented

2026-10-31 21:13:06 +00:00

Author

Owner

@lr19960813 commented on GitHub (Mar 27, 2023):

The same

@lr19960813 commented on GitHub (Mar 27, 2023): The same

bsakaguchi referenced this issue

2026-10-31 21:15:23 +00:00

[PR #259] [MERGED] Add PyTorch Hub configuration file #477