mnist_classifier_example.hpp File Reference

Real MNIST classifier: Conv2DModule -> ReluModule -> FlattenModule -> LinearModule -> CrossEntropyLoss, mirroring XorNetwork's training-loop pattern and grad_cam_mnist_demo.cpp's network shape exactly. More...

#include <vector>
#include "pulsatrix/adam_optimizer.hpp"
#include "pulsatrix/conv2d_module.hpp"
#include "pulsatrix/cross_entropy_loss.hpp"
#include "pulsatrix/flatten_module.hpp"
#include "pulsatrix/linear_module.hpp"
#include "pulsatrix/metrics_sink.hpp"
#include "pulsatrix/relu_module.hpp"

Include dependency graph for mnist_classifier_example.hpp:

Go to the source code of this file.

Classes
class	pulsatrix::MnistConvNet
	Conv2D(1,8,5,5) -> ReLU -> Flatten -> Linear(4608,10), trained via CrossEntropyLoss + Adam, one real MNIST image at a time (this library has no batch dimension anywhere, same constraint XorNetwork already works under). More...

Namespaces
namespace	pulsatrix

Detailed Description

Real MNIST classifier: Conv2DModule -> ReluModule -> FlattenModule -> LinearModule -> CrossEntropyLoss, mirroring XorNetwork's training-loop pattern and grad_cam_mnist_demo.cpp's network shape exactly.