[V1][VLM] Proper memory profiling for image language models (#11210)

Signed-off-by: Roger Wang <ywang@roblox.com>
Co-authored-by: ywang96 <ywang@example.com>
This commit is contained in:
Roger Wang
2024-12-16 22:10:57 -08:00
committed by GitHub
co-authored by ywang96
parent 66d4b16724
commit 59c9b6ebeb
6 changed files with 98 additions and 13 deletions
+1
View File
@@ -54,6 +54,7 @@ class MMInputMapperClient:
logger.debug("MMInputMapper: cache_hit_ratio = %.2f ",
self.mm_cache_hits / self.mm_cache_total)
# TODO: Support modalities beyond image.
def process_inputs(
self,
mm_data: MultiModalDataDict,