From 467daef993fe29cc4319058200b7ad797398e4b0 Mon Sep 17 00:00:00 2001 From: Omar Al Khatib Date: Thu, 13 Apr 2023 14:56:23 +0100 Subject: Implement CL kernel for a native batched matmul Quantized - LHS transposed, RHS transposed Resolves: [COMPMID-5924] Signed-off-by: Omar Al Khatib Change-Id: I9ba657737eb1e3a096c8341ad4ad311571f8edeb Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/9454 Benchmark: Arm Jenkins Tested-by: Arm Jenkins Reviewed-by: SiCong Li Comments-Addressed: Arm Jenkins --- src/gpu/cl/ClKernelLibrary.cpp | 1 + 1 file changed, 1 insertion(+) (limited to 'src/gpu/cl') diff --git a/src/gpu/cl/ClKernelLibrary.cpp b/src/gpu/cl/ClKernelLibrary.cpp index e657687887..4612ca35b8 100644 --- a/src/gpu/cl/ClKernelLibrary.cpp +++ b/src/gpu/cl/ClKernelLibrary.cpp @@ -325,6 +325,7 @@ const std::map ClKernelLibrary::_kernel_program_map = { "mat_mul_native_t_t", "common/mat_mul.cl" }, { "mat_mul_native_quantized_nt_nt", "common/mat_mul_quantized.cl" }, { "mat_mul_native_quantized_t_nt", "common/mat_mul_quantized.cl" }, + { "mat_mul_native_quantized_t_t", "common/mat_mul_quantized.cl" }, { "max_unpooling_layer_2", "common/unpooling_layer.cl" }, { "mean_stddev_normalization", "common/mean_stddev_normalization.cl" }, { "memset", "common/memset.cl" }, -- cgit v1.2.1