feat: implement session-managed persistent KV cache architecture with slot persistence and management API

2026-09-04 02:53:54 +00:00
parent 5daafe24e8
commit 9ab0583cab
4 changed files with 244 additions and 3 deletions
@@ -19,7 +19,7 @@ TimeoutStopSec=60
 # Basic hardening (server only listens on :8080, reads models read-only)
 ProtectSystem=strict
 ProtectHome=read-only
-ReadWritePaths=
+ReadWritePaths=/var/cache/llama-slots /tmp
 NoNewPrivileges=true
 PrivateTmp=true
 # Do not cap memory/CPU; inference needs all of it