fix(training): lower default learning rate to 0.01 to stabilize convergence

Author: Keshav-writes-code

try1 (OOP Approach)/cpp/src/main.cpp

@@ -1,75 +1,63 @@
-#include <iostream>
 #include "../include/NeuralNet/classes.h"
+#include <iostream>
 
 using namespace std;
 
-int *genRandomInts(int arrLen, int min, int max)
-{
+int *genRandomInts(int arrLen, int min, int max) {
   int *arr = (int *)malloc(arrLen * sizeof(int));
-  for (int i = 0; i < arrLen; i++)
-  {
+  for (int i = 0; i < arrLen; i++) {
     int randVal = rand() % (max - min + 1) + min;
     arr[i] = randVal;
   }
   return arr;
 }
-float *genRandomFloats(int arrLen, float min, float max)
-{
+float *genRandomFloats(int arrLen, float min, float max) {
   float *arr = new float[arrLen];
-  for (int i = 0; i < arrLen; i++)
-  {
+  for (int i = 0; i < arrLen; i++) {
     float randVal = ((float)rand() / RAND_MAX) * (max - min) + min;
     arr[i] = randVal;
   }
   return arr;
 }
-float **genRandom2DFloats(int rows, int cols, float min, float max)
-{
+float **genRandom2DFloats(int rows, int cols, float min, float max) {
   // Allocate memory for the array of row pointers
   float **array2D = new float *[rows];
   // Generate each row using genRandomFloats
-  for (int i = 0; i < rows; i++)
-  {
+  for (int i = 0; i < rows; i++) {
     array2D[i] = genRandomFloats(cols, min, max);
   }
   return array2D;
 }
 
-float **getIdentityMatrix(int rows, int cols)
-{
+float **getIdentityMatrix(int rows, int cols) {
   float **matrix = new float *[rows];
-  for (int i = 0; i < rows; i++)
-  {
+  for (int i = 0; i < rows; i++) {
     matrix[i] = new float[cols];
-    for (int j = 0; j < cols; j++)
-    {
+    for (int j = 0; j < cols; j++) {
       matrix[i][j] = (i == j) ? 1.0f : 0.0f;
     }
   }
   return matrix;
 }
 
-void print2DArray(float **matrix, int rows, int cols)
-{
+void print2DArray(float **matrix, int rows, int cols) {
   cout << "2D Array : " << endl;
-  for (int i = 0; i < rows; i++)
-  {
-    for (int j = 0; j < cols; j++)
-    {
+  for (int i = 0; i < rows; i++) {
+    for (int j = 0; j < cols; j++) {
       cout << matrix[i][j] << " ";
     }
     cout << endl;
   }
 }
 
-int main()
-{
+int main() {
   srand(time(NULL));
   const int inputLayerSize = 10;
   const int hidOutLayerCount = 2;
   const int outputLayerSize = 10;
   int *hidOutLayerSizes = genRandomInts(hidOutLayerCount, 16, 20);
-  NeuralNet::MLP *MLP = new NeuralNet::MLP(inputLayerSize, hidOutLayerCount, hidOutLayerSizes, outputLayerSize, 0.3);
+  NeuralNet::MLP *MLP = new NeuralNet::MLP(
+      inputLayerSize, hidOutLayerCount, hidOutLayerSizes, outputLayerSize, 0.01);
 
   // Dataset with labels
   const int samples_count = 10;
@@ -85,4 +73,4 @@ int main()
   // Predict
   MLP->predict(inputs, inputLayerSize, targets, outputLayerSize, samples_count);
   cout << "Ended" << endl;
-}
+}