Mila
Deep Neural Network Library
|
Implementation of the CUDA-based fused softmax and cross entropy operation for neural networks. More...
#include <vector>
#include <iostream>
#include <memory>
#include "Kernels/CudaOps.h"
#include <type_traits>
import Compute.CudaDevice;
import Compute.OperationAttributes;
import Compute.CudaMemoryResource;
import Compute.OperationType;
import Compute.DeviceContext;
import Compute.OperationRegistry;
import Compute.MemoryResource;
import Compute.BinaryOperation;
import Compute.OperationBase;
import Compute.DeviceType;
import Dnn.Tensor;
Classes | |
struct | Mila::Dnn::Compute::Detail::cuda_softmax_crossentropy_impl< float > |
struct | Mila::Dnn::Compute::Detail::cuda_softmax_crossentropy_impl< half > |
class | Mila::Dnn::Compute::FusedSoftmaxCrossEntropyOp< TPrecision > |
CUDA implementation of the fused softmax and cross entropy operation for neural networks. More... | |
class | Mila::Dnn::Compute::FusedSoftmaxCrossEntropyOpRegistrar |
Class responsible for registering the FusedSoftmaxCrossEntropyOp operation. More... | |
Namespaces | |
namespace | Mila |
namespace | Mila::Dnn |
namespace | Mila::Dnn::Compute |
namespace | Mila::Dnn::Compute::Detail |
Namespace for CUDA layer normalization implementation details. | |
Implementation of the CUDA-based fused softmax and cross entropy operation for neural networks.