CudaGeluOp.Dispatch.ixx File Reference

Implementation of the CUDA GELU kernel dispatch mechanism. More...

#include <memory>
#include <cuda_fp16.h>
#include <type_traits>
#include "Kernels/Gelu.cuh"
import Compute.CudaDevice;
import Compute.MemoryResource;
import Compute.OperationType;
import Dnn.Components.GeluConfig;
import Compute.UnaryOperation;
import Dnn.ApproximationMethod;
import Compute.ExecutionContext;
import Dnn.Tensor;
import Dnn.ITensor;
import Compute.OperationRegistry;
import Compute.CudaDeviceMemoryResource;
import Compute.DeviceType;
import Compute.CudaTensorDataType;
import Compute.ExecutionContextTemplate;
import Dnn.TensorDataTypeTraits;
import Dnn.TensorDataType;
import Dnn.ComponentConfig;
import Compute.OperationBase;
import Compute.IExecutionContext;

Classes
struct	Mila::Dnn::Compute::Cuda::Gelu::Detail::cuda_gelu_impl< float >
struct	Mila::Dnn::Compute::Cuda::Gelu::Detail::cuda_gelu_impl< half >

Namespaces
namespace	Mila
	Mila main API namespace.
namespace	Mila::Dnn
namespace	Mila::Dnn::Compute
namespace	Mila::Dnn::Compute::Cuda
namespace	Mila::Dnn::Compute::Cuda::Gelu
namespace	Mila::Dnn::Compute::Cuda::Gelu::Detail

Typedefs
using	Mila::Dnn::Compute::Cuda::Gelu::Detail::BackwardFp16Func = void (*)(half*, const half*, const half*, int, cudaStream_t)
using	Mila::Dnn::Compute::Cuda::Gelu::Detail::BackwardFp32Func = void (*)(float*, const float*, const float*, int, cudaStream_t)
using	Mila::Dnn::Compute::Cuda::Gelu::Detail::ForwardFp16Func = void (*)(half*, const half*, int, cudaStream_t)
using	Mila::Dnn::Compute::Cuda::Gelu::Detail::ForwardFp32Func = void (*)(float*, const float*, int, cudaStream_t)

Detailed Description

Implementation of the CUDA GELU kernel dispatch mechanism.