LLM inference in C/C++
C++ ggml
vulkan : check src0 type in GGML_OP_SET_ROWS to avoid failures due to unimplemented f16 support (#25351)
fairydreaminga8cfdbb9e4c4
diff --git a/ggml/src/ggml-vulkan/ggml-vulkan.cpp b/ggml/src/ggml-vulkan/ggml-vulkan.cppindex 3e38c17ded..96d89943ae 100644--- a/ggml/src/ggml-vulkan/ggml-vulkan.cpp+++ b/ggml/src/ggml-vulkan/ggml-vulkan.cpp@@ -17370,21 +17370,24 @@ static bool ggml_backend_vk_device_supports_op(ggml_backend_dev_t dev, const ggmreturn op->type == GGML_TYPE_F32 && op->src[0]->type == GGML_TYPE_F32;case GGML_OP_SET_ROWS:{- switch (op->type) {- case GGML_TYPE_F32:- case GGML_TYPE_F16:- case GGML_TYPE_BF16:- case GGML_TYPE_Q1_0:- case GGML_TYPE_Q4_0:- case GGML_TYPE_Q4_1:- case GGML_TYPE_Q5_0:- case GGML_TYPE_Q5_1:- case GGML_TYPE_Q8_0:- case GGML_TYPE_IQ4_NL:- return true;- default:- return false;+ if (op->src[0]->type == GGML_TYPE_F32) {+ switch (op->type) {+ case GGML_TYPE_F32:+ case GGML_TYPE_F16:+ case GGML_TYPE_BF16:+ case GGML_TYPE_Q1_0:+ case GGML_TYPE_Q4_0:+ case GGML_TYPE_Q4_1:+ case GGML_TYPE_Q5_0:+ case GGML_TYPE_Q5_1:+ case GGML_TYPE_Q8_0:+ case GGML_TYPE_IQ4_NL:+ return true;+ default:+ return false;+ }}+ return false;}case GGML_OP_CONT:case GGML_OP_CPY: