transformers
a6f5824a
- Merge branch 'main' into fsdp2-ep-2d-mesh
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
22 days ago
Merge branch 'main' into fsdp2-ep-2d-mesh
References
#48204 - [Umbrella] FSDP2 + expert-parallel token dispatch, with sharded-loading prefetch
#48516 - Enable FSDP2 + expert parallelism via a 2-D (fsdp, tp) device mesh
#48792 - Add Expert parallel
Author
qgallouedec
Parents
1be3a7f5
ff4bb5ee
Loading