Changes

From CDOT Wiki
Jump to: navigation, search

A-Team

1,521 bytes added, 1 April
Assignment 2
This actually costs us an additional 20 minutes when profiling could be done.
 
====The next steps====
Well firstly we had to engage in research as to understand how the actual neural network was learning; for example why they used relu() function, how back-propagation worked and so much more.
 
=====After that and many coffees!=====
__global__ void train(float* d_W1, float* d_W2, float* d_W3, float* d_b_X, float* d_b_Y, float* d_a2, float* d_a1, float* d_dyhat, float* d_dW3, float* d_dW2, float* d_dW1, float* d_dz2, float* d_dz1) {
int BATCH_SIZE = 256;
float lr = .01 / BATCH_SIZE;
 
kdot<<< 50,51>>>(ktranspose(d_a2, BATCH_SIZE, 64), d_dyhat, 64, BATCH_SIZE, 10, d_dW3);
 
 
kdot << <80,32>> >(d_dyhat, ktranspose(d_W3, 64, 10), BATCH_SIZE, 10, 64, d_dz2);
kreluPrime(d_a2, 128 * 64);
for (int i = 0; i < BATCH_SIZE * 10; i++) {
d_dz2[i] = d_dz2[i] * d_a2[i];
}
 
kdot << <1024, 32>> >(ktranspose(d_a1, BATCH_SIZE, 128), d_dz2, 128, BATCH_SIZE, 64, d_dW2);
 
kdot << <512,32>> >(d_dz2, ktranspose(d_W2, 128, 64), BATCH_SIZE, 64, 128, d_dz1);
kreluPrime(d_a1, BATCH_SIZE * 784);
for (int i = 0; i < 256 * 64; i++) {
d_dz1[i] = d_dz1[i] * d_a1[i];
}
 
kdot <<<512,512,32 >>>(ktranspose(d_b_X, BATCH_SIZE, 784), d_dz1, 784, BATCH_SIZE, 128, d_dW1);
// Updating the parameters
//W3 = W3 - lr * dW3;
for (int i = 0; i < (64*10); i++) {
d_W3[i] = d_W3[i] - lr * d_dW3[i];
}
//W2 = W2 - lr * dW2;
for (int i = 0; i < (128*64); i++) {
d_W2[i] = d_W2[i] - lr * d_dW2[i];
}
//W1 = W1 - lr * dW1;
for (int i = 0; i < (784*128); i++) {
d_W1[i] = d_W1[i] - lr * d_dW1[i];
}
 
}
=== Assignment 3 ===
113
edits

Navigation menu