DeepSpeed
fix(rollout): pass a pre-built 4-D mask to the static graph decode forward
#9004
Merged
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Overview
Commits
3
Changes
View On
GitHub
Commits
Pass a prebuilt 4-D mask to the static rollout CUDA graph
anyadontfly
committed
12 days ago
fix(rollout): reject flash_attention_2 on the static graph path and test real models
anyadontfly
committed
12 days ago
Merge master into graph-decode-4d-mask
anyadontfly
committed
11 days ago
Loading