DeepSpeed
970015bb
- efficient communication layout for training MoE architecture at large scale
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
2 years ago
efficient communication layout for training MoE architecture at large scale
References
#4695 - Communication Optimization for Large-Scale Training
Author
Reza Yazdani
Parents
6d0b4c2c
Loading