facebookresearch / facebookresearch/memory
Approximate FLOPs of Memory Layers
Open
- Dominant language
- Python
- Stars
- 379
- Forks
- 32
- PR merge metrics
- No merged PRs in 30d
Description
In the paper you state that Memory Layers achieve faster converge (i.e., "with 10x fewer FLOPS"). How are FLOPs calculated approximately for Memory Layers?
Contributor guide
Assessment
This issue has not been assessed yet.