SIGN IN SIGN UP

kv-cache : pad the cache size to 256 for performance (#17046)

* kv-cache : pad the size of the small SWA cache for performance

* context : pad the total context to 256

* cont : future-proof the swa pad

* server : adjust test params to new logic
G
Georgi Gerganov committed
16bcc1259d311d0fd37fe00fefcc7900324d38cb
Parent: 9eb9a13
Committed by GitHub <noreply@github.com> on 11/7/2025, 6:03:25 PM