Skip to content

LLM inference in C/C++

C++ ggml

ggml : add MUL_MAT to the list of ops that may need additional memory (for WebGPU) (#28071)

fairydreaming557614e0296f
diff --git a/ggml/src/ggml-backend.cpp b/ggml/src/ggml-backend.cpp
index 8f40fb2f7b..ffe20b9d05 100644
--- a/ggml/src/ggml-backend.cpp
+++ b/ggml/src/ggml-backend.cpp
@@ -2115,6 +2115,7 @@ ggml_backend_t ggml_backend_sched_get_tensor_backend(ggml_backend_sched_t sched,
bool ggml_backend_op_alloc_size_may_expand(enum ggml_op op) {
switch (op) {
case GGML_OP_FLASH_ATTN_EXT:
+ case GGML_OP_MUL_MAT:
case GGML_OP_MUL_MAT_ID:
case GGML_OP_CUMSUM:
case GGML_OP_ARGSORT: