temporarily disable zero on device memory during kv_cache alloc

microsoft · Feb 20, 2025 · 15b251d · 15b251d
1 parent 16fb079
commit 15b251d
Showing 1 changed file with 4 additions and 1 deletion.
diff --git a/src/models/kv_cache.cpp b/src/models/kv_cache.cpp
@@ -187,7 +187,10 @@ DefaultKeyValueCache::DefaultKeyValueCache(State& state)
           sb_kv_caches_.empty() ? OrtValue::CreateTensor(Allocator(), shape_, type_)
                                 : sb_kv_caches_[i]->CreateTensorOnStaticBuffer(shape_, type_));
       // Zero the memory so we don't leak any data from the previous run
-      ByteWrapTensor(Device(), *presents_.back()).Zero();
+      if (Device().GetType() != DeviceType::WEBGPU) {
+        // ort c api does have a method to update device memory - temporarily disable.
+        ByteWrapTensor(Device(), *presents_.back()).Zero();
+      }
     }
   } catch (const Ort::Exception&) {
     std::ostringstream oss;