DeepSpeed
79b81eda
- Enable memory efficient linear when ZeRO 3 model is initialized in Stage 3 initialize instead of using deepspeed.init
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
5 years ago
Enable memory efficient linear when ZeRO 3 model is initialized in Stage 3 initialize instead of using deepspeed.init
References
#921 - Samyamr/fix for fragmented linear inputs
Author
samyam
Parents
e193901a
Loading