diff --git a/ggml/src/ggml-vulkan/ggml-vulkan.cpp b/ggml/src/ggml-vulkan/ggml-vulkan.cpp
index c5e3cb19d..fa0c264f0 100644
--- a/ggml/src/ggml-vulkan/ggml-vulkan.cpp
+++ b/ggml/src/ggml-vulkan/ggml-vulkan.cpp
@@ -3090,7 +3090,7 @@ static void ggml_vk_mul_mat_vec_q_f16(ggml_backend_vk_context * ctx, vk_context&
         to_fp16_vk_1 = ggml_vk_get_to_fp16(ctx, src1->type);
     }
     vk_pipeline dmmv = ggml_vk_get_dequantize_mul_mat_vec(ctx, src0->type, src1->type);
-    GGML_ASSERT(dmmv != nullptr && (!qx_needs_dequant || to_fp16_vk_0 != nullptr || !qy_needs_dequant || to_fp16_vk_1 != nullptr));  // NOLINT
+    GGML_ASSERT((!qx_needs_dequant || to_fp16_vk_0 != nullptr || !qy_needs_dequant || to_fp16_vk_1 != nullptr) && dmmv != nullptr);  // NOLINT
 
     if (dryrun) {
         const uint64_t x_sz_upd = x_sz * ne02 * ne03;
@@ -3369,7 +3369,7 @@ static void ggml_vk_mul_mat_id_q_f16(ggml_backend_vk_context * ctx, vk_context&
     std::cerr << "), (" << src1 << ", name=" << src1->name << ", type=" << src1->type << ", ne0=" << src1->ne[0] << ", ne1=" << src1->ne[1] << ", ne2=" << src1->ne[2] << ", ne3=" << src1->ne[3] << ", nb0=" << src1->nb[0] << ", nb1=" << src1->nb[1] << ", nb2=" << src1->nb[2] << ", nb3=" << src1->nb[3];
     std::cerr << "), (" << ids << ", name=" << ids->name << ", type=" << ids->type << ", ne0=" << ids->ne[0] << ", ne1=" << ids->ne[1] << ", ne2=" << ids->ne[2] << ", ne3=" << ids->ne[3] << ", nb0=" << ids->nb[0] << ", nb1=" << ids->nb[1] << ", nb2=" << ids->nb[2] << ", nb3=" << ids->nb[3];
     std::cerr << "), (" << dst << ", name=" << dst->name << ", type=" << dst->type << ", ne0=" << dst->ne[0] << ", ne1=" << dst->ne[1] << ", ne2=" << dst->ne[2] << ", ne3=" << dst->ne[3] << ", nb0=" << dst->nb[0] << ", nb1=" << dst->nb[1] << ", nb2=" << dst->nb[2] << ", nb3=" << dst->nb[3] << "),)");
-    GGML_ASSERT(ids->type == GGML_TYPE_I32 && (ggml_vk_dim01_contiguous(src1) || src1->type == GGML_TYPE_F32 || src1->type == GGML_TYPE_F16));  // NOLINT
+    GGML_ASSERT((ggml_vk_dim01_contiguous(src1) || src1->type == GGML_TYPE_F32 || src1->type == GGML_TYPE_F16) && ids->type == GGML_TYPE_I32);  // NOLINT
 
     const uint64_t ne00 = src0->ne[0];
     const uint64_t ne01 = src0->ne[1];
@@ -3573,7 +3573,7 @@ static void ggml_vk_mul_mat_vec_id_q_f16(ggml_backend_vk_context * ctx, vk_conte
     std::cerr << "), (" << ids << ", name=" << ids->name << ", type=" << ids->type << ", ne0=" << ids->ne[0] << ", ne1=" << ids->ne[1] << ", ne2=" << ids->ne[2] << ", ne3=" << ids->ne[3] << ", nb0=" << ids->nb[0] << ", nb1=" << ids->nb[1] << ", nb2=" << ids->nb[2] << ", nb3=" << ids->nb[3];
     std::cerr << "), (" << dst << ", name=" << dst->name << ", type=" << dst->type << ", ne0=" << dst->ne[0] << ", ne1=" << dst->ne[1] << ", ne2=" << dst->ne[2] << ", ne3=" << dst->ne[3] << ", nb0=" << dst->nb[0] << ", nb1=" << dst->nb[1] << ", nb2=" << dst->nb[2] << ", nb3=" << dst->nb[3];
     std::cerr << "), " << (dryrun ? "dryrun" : "") << ")");
-    GGML_ASSERT(ids->type == GGML_TYPE_I32 && (ggml_vk_dim01_contiguous(src0) || src0->type == GGML_TYPE_F32 || src0->type == GGML_TYPE_F16 || ggml_vk_dim01_contiguous(src1) || src1->type == GGML_TYPE_F32 || src1->type == GGML_TYPE_F16));  // NOLINT
+    GGML_ASSERT((ggml_vk_dim01_contiguous(src0) || src0->type == GGML_TYPE_F32 || src0->type == GGML_TYPE_F16 || ggml_vk_dim01_contiguous(src1) || src1->type == GGML_TYPE_F32 || src1->type == GGML_TYPE_F16) && ids->type == GGML_TYPE_I32);  // NOLINT
 
     const uint64_t ne00 = src0->ne[0];
     const uint64_t ne01 = src0->ne[1];
@@ -3655,7 +3655,7 @@ static void ggml_vk_mul_mat_vec_id_q_f16(ggml_backend_vk_context * ctx, vk_conte
         to_fp16_vk_1 = ggml_vk_get_to_fp16(ctx, src1->type);
     }
     vk_pipeline dmmv = ggml_vk_get_dequantize_mul_mat_vec_id(ctx, src0->type, src1->type);
-    GGML_ASSERT(dmmv != nullptr && (!qx_needs_dequant || to_fp16_vk_0 != nullptr || !qy_needs_dequant || to_fp16_vk_1 != nullptr));  // NOLINT
+    GGML_ASSERT((!qx_needs_dequant || to_fp16_vk_0 != nullptr || !qy_needs_dequant || to_fp16_vk_1 != nullptr) && dmmv != nullptr);  // NOLINT
 
     if (dryrun) {
         const uint64_t x_sz_upd = x_sz * ne02 * ne03;