[claudesquad] update from 'int-minference-1' on 08 Jan 26 23:22 CST

2026-01-08 23:22:38 +08:00
parent 0bfe1984ef
commit ea4e904de0
11 changed files with 853 additions and 533 deletions
--- a/nanovllm/kvcache/sparse/quest.py
+++ b/nanovllm/kvcache/sparse/quest.py
@@ -158,6 +158,7 @@ class QuestPolicy(SparsePolicy):
    # Quest is decode-only
    supports_prefill = False
    supports_decode = True
+    requires_block_selection = True  # Quest affects KV load strategy (selective block loading)

    def __init__(self, config: QuestConfig):
        """