[Bugfix][Pooling] Handle multimodal cache misses without crashing (#58975)

Signed-off-by: linchaoying6 <linchaoying6@jd.com>
Co-authored-by: linchaoying6 <linchaoying6@jd.com>
Co-authored-by: OpenAI Codex <codex@openai.com>
This commit is contained in:
Chaoying
2026-10-02 10:14:33 -04:00
committed by GitHub
co-authored by linchaoying6 OpenAI Codex
parent 443dccd3c9
commit 592c6f3fbc
21 changed files with 525 additions and 17 deletions
+7
View File
@@ -143,6 +143,13 @@ class WorkerBase:
if callable(reset_fn):
reset_fn()
def release_late_interaction_query_cache(self, query_keys: list[str]) -> None:
release_fn = getattr(
self.model_runner, "release_late_interaction_query_cache", None
)
if callable(release_fn):
release_fn(query_keys)
def get_model(self) -> nn.Module:
raise NotImplementedError