auto-round
05ad2e16
- Merge branch 'copilot/optimize-int4-moe-performance' into W4A8 MoE work
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
11 days ago
Merge branch 'copilot/optimize-int4-moe-performance' into W4A8 MoE work Co-authored-by: a32543254 <53296245+a32543254@users.noreply.github.com>
References
#2143 - feat: W4A8 ARK XPU MoE kernel (int4 weight / int8 compute) with prefill + decode
Author
Copilot
Parents
f17d9cd4
42949b74
Loading