facebookresearch / facebookresearch/SlowFast
The configuration of MViT-B-24
Open
- Dominant language
- Python
- Stars
- 7.4k
- Forks
- 1.3k
- PR merge metrics
- No merged PRs in 30d
Description
Hi, thank you for sharing this fine work.
I want to run MViT-B-24, however, I can not find the configuration of this.
Referring to the paper,
> the deeper variants are simply created by repeating the number of blocks N∗ in each scale stage.
The MViT-B has 16 blocks by default (1, 2, 11, 2). To me, it's a little bit confusing, since repeating it means the deeper MViT-B having 32 blocks. So, what is the configuration of MViT-B-24? Thanks in advance!
Contributor guide
Assessment
This issue has not been assessed yet.