A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 9, 16:00 UTC
[EncoderCacheManager] Remove unnecessary copy (#32800)
Signed-off-by: Lukas Geiger <lukas.geiger94@gmail.com>
Lukas Geiger committed
Jan 24, 2026 at 14:28 UTC
5fa0f6efa9f8778b61a586fdf86b218e1d19395f
1 file changed
+1
-1
vllm/v1/core/encoder_cache_manager.py
+1
-1
@@ -237,7 +237,7 @@ class EncoderCacheManager:
237
238
Typically called when a request is finished, cancelled, or aborted.
239
"""
240
- input_ids = self.get_cached_input_ids(request).copy()
240
+ input_ids = self.get_cached_input_ids(request)
241
for input_id in input_ids:
242
self.free_encoder_input(request, input_id)
243