2 ms·
Look at the “dataset” column: CLIP was trained on 400m images, UForm on 4m.
by brookst 4y ago
Look at the “dataset” column: CLIP was trained on 400m images, UForm on 4m.
- vov_or 4y agoThere are also dataset sizes for Albef and ViCHA.
loading thread…