nnlib
GPU-accelerated, C/C++ neural network library.
|
Source file defining methods of the Layer class. More...
#include "../../include/layer.h"
#include "../gpu/allocation_gpu.cuh"
#include "runtime.h"
#include "verify.cuh"
#include <utility>
Functions | |
float | getRandom () |
Generate a random float value. More... | |
sTensor | initializeBiases (size_t outSize) |
Initialize the biases of the layer. More... | |
sTensor | initializeWeights (size_t inSize, size_t outSize) |
Initialize the weights of the layer. More... | |
Source file defining methods of the Layer class.
float getRandom | ( | ) |
Generate a random float value.
Currently the method only returns uniformly distributed numbers in the range (-0.2, 0.2).
sTensor initializeBiases | ( | size_t | outSize | ) |
Initialize the biases of the layer.
This creates a vector of random numbers using the getRandomValue() method.
outSize | The size of the vector to generate. It is also the output size of the layer. |
sTensor initializeWeights | ( | size_t | inSize, |
size_t | outSize | ||
) |
Initialize the weights of the layer.
This creates a matrix of random numbers using the getRandomValue() method.
inSize | The number of rows of the matrix. It is also the input size to the layer. |
outSize | The number of columns of the matrix. It is also the output size of the layer. |