onnxruntime
d21946e6
- Add compact QSA selection and specialized INT8 attention dispatch
Go
Login via GitHub
Home
Pricing
FAQ
Install
Login
via GitHub
Commit
View On
GitHub
Commit
17 days ago
Add compact QSA selection and specialized INT8 attention dispatch Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
References
#33472 - [CUDA] Extend DynamicSparseAttention INT8 KV cache and PackedSparseAttentionIndexer QSA
Author
apsonawane
Parents
93565558
Loading