aboutsummaryrefslogtreecommitdiff
path: root/src/gpu/cl/kernels/ClMatMulNativeKernel.h
diff options
context:
space:
mode:
authorMohammed Suhail Munshi <MohammedSuhail.Munshi@arm.com>2023-05-25 16:48:43 +0100
committerMohmun02 <MohammedSuhail.Munshi@arm.com>2023-06-16 15:38:39 +0000
commit94abde4f4e98f6f1adb5c46b194527f34a8ea07d (patch)
treed6d717031788850d970fb44ff3f41de311cc5fc0 /src/gpu/cl/kernels/ClMatMulNativeKernel.h
parentdd8d7f4102653ef55d872c71ae5d5f2ca2ead0c1 (diff)
downloadComputeLibrary-94abde4f4e98f6f1adb5c46b194527f34a8ea07d.tar.gz
Add Fused Activation to OpenCL MatMul
- Added fused activation to MatMul function interface - Added fused activation to CL backend - Includes tests for supported Activation Functions in MatMul Resolves: [COMPMID-6192] Signed-off-by: Mohammed Suhail Munshi <MohammedSuhail.Munshi@arm.com> Change-Id: Ie103212b600b60699eaf6a6394d609e6e1f5aba6 Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/c/VisualCompute/ComputeLibrary/+/522465 Comments-Addressed: bsgcomp <bsgcomp@arm.com> Reviewed-by: Viet-Hoa Do <viet-hoa.do@arm.com> Tested-by: bsgcomp <bsgcomp@arm.com> Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/9714 Comments-Addressed: Arm Jenkins <bsgcomp@arm.com> Reviewed-by: Jakub Sujak <jakub.sujak@arm.com> Tested-by: Arm Jenkins <bsgcomp@arm.com> Benchmark: Arm Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/gpu/cl/kernels/ClMatMulNativeKernel.h')
-rw-r--r--src/gpu/cl/kernels/ClMatMulNativeKernel.h20
1 files changed, 11 insertions, 9 deletions
diff --git a/src/gpu/cl/kernels/ClMatMulNativeKernel.h b/src/gpu/cl/kernels/ClMatMulNativeKernel.h
index f706256e31..02d8ac3067 100644
--- a/src/gpu/cl/kernels/ClMatMulNativeKernel.h
+++ b/src/gpu/cl/kernels/ClMatMulNativeKernel.h
@@ -42,22 +42,24 @@ public:
ARM_COMPUTE_DISALLOW_COPY_ALLOW_MOVE(ClMatMulNativeKernel);
/** Initialise the kernel's input and output.
*
- * @param[in] compile_context The compile context to be used.
- * @param[in] lhs Input tensor for the LHS matrix. Data type supported: F32/F16.
- * Dimensions above 2 are collapsed onto dimension 2 and represent the batch.
- * @param[in] rhs Input tensor for the RHS matrix. Data type supported: same as @p lhs.
- * Dimensions above 2 are collapsed onto dimension 2 and represent the batch.
- * @param[out] dst Output tensor info. Data type supported: same as @p lhs
- * @param[in] matmul_info Attributes for Batch MatMul Kernel
+ * @param[in] compile_context The compile context to be used.
+ * @param[in] lhs Input tensor for the LHS matrix. Data type supported: F32/F16.
+ * Dimensions above 2 are collapsed onto dimension 2 and represent the batch.
+ * @param[in] rhs Input tensor for the RHS matrix. Data type supported: same as @p lhs.
+ * Dimensions above 2 are collapsed onto dimension 2 and represent the batch.
+ * @param[out] dst Output tensor info. Data type supported: same as @p lhs
+ * @param[in] matmul_kernel_info Attributes for Batch MatMul Kernel
+ * @param[in] act_info Specifies activation function to use after Matrix multiplication. Default is Identity function.
*/
- void configure(const ClCompileContext &compile_context, ITensorInfo *lhs, ITensorInfo *rhs, ITensorInfo *dst, const MatMulKernelInfo &matmul_info);
+ void configure(const ClCompileContext &compile_context, ITensorInfo *lhs, ITensorInfo *rhs, ITensorInfo *dst, const MatMulKernelInfo &matmul_kernel_info,
+ const ActivationLayerInfo &act_info = ActivationLayerInfo());
/** Static function to check if given info will lead to a valid configuration
*
* Similar to @ref ClMatMulNativeKernel::configure()
*
* @return a status
*/
- static Status validate(const ITensorInfo *lhs, const ITensorInfo *rhs, const ITensorInfo *dst, const MatMulKernelInfo &matmul_info);
+ static Status validate(const ITensorInfo *lhs, const ITensorInfo *rhs, const ITensorInfo *dst, const MatMulKernelInfo &matmul_kernel_info, const ActivationLayerInfo &act_info = ActivationLayerInfo());
// Inherited methods overridden:
void run_op(ITensorPack &tensors, const Window &window, cl::CommandQueue &queue) override;