Обновить Xenith/core.cpp

Xenith/core.cpp

@@ -2,6 +2,19 @@
 #include <cmath>
 #include <cstdlib>
 
+#define USE_PARALLEL
+#define MAX_THREADS 0
+
+#ifdef USE_PARALLEL
+#include <omp.h>
+#define OMP_PARALLEL _Pragma("omp parallel for")
+#define OMP_SET_THREADS() { if (MAX_THREADS > 0) omp_set_num_threads(MAX_THREADS); }
+#else
+#define OMP_PARALLEL 
+#define OMP_SET_THREADS() 
+#endif
+
+
 NeuralNetwork::NeuralNetwork(LayerStructure_t layers[], int count) : numLayers(count) {
     for (int i = 0; i < count; i++) sizes.push_back(layers[i].size);
     for (int i = 0; i < count - 1; i++) {
@@ -19,45 +32,71 @@ NeuralNetwork::NeuralNetwork(LayerStructure_t layers[], int count) : numLayers(c
 }
 
 std::vector<double> NeuralNetwork::feedForward(const std::vector<double>& input) {
+    OMP_SET_THREADS();
+
     outputs.clear();
     outputs.push_back(input);
+
     std::vector<double> curr = input;
+
     for (int i = 0; i < numLayers - 1; i++) {
-        std::vector<double> next;
+        std::vector<double> next(sizes[i + 1]);
-        for (int j = 0; j < sizes[i+1]; j++) {
+
+        OMP_PARALLEL
+            for (int j = 0; j < sizes[i + 1]; j++) {
                 double sum = biases[i][j];
-            for (int k = 0; k < (int)curr.size(); k++) sum += curr[k] * weights[i][j][k];
+                for (int k = 0; k < (int)curr.size(); k++) {
-            next.push_back(1.0 / (1.0 + exp(-sum)));
+                    sum += curr[k] * weights[i][j][k];
                 }
+                next[j] = 1.0 / (1.0 + exp(-sum));
+            }
+
         curr = next;
         outputs.push_back(curr);
     }
     return curr;
 }
 
+
 double NeuralNetwork::train(const std::vector<double>& input, const std::vector<double>& target, double lr) {
+    OMP_SET_THREADS();
+
     std::vector<double> pred = feedForward(input);
+
     std::vector<std::vector<double>> errors(numLayers);
-    errors[numLayers-1].resize(sizes[numLayers-1]);
+    errors[numLayers - 1].resize(sizes[numLayers - 1]);
+
     double totalErr = 0;
-    for (int i = 0; i < sizes[numLayers-1]; i++) {
+
+    for (int i = 0; i < sizes[numLayers - 1]; i++) {
         double e = target[i] - pred[i];
-        errors[numLayers-1][i] = e * pred[i] * (1.0 - pred[i]);
+        errors[numLayers - 1][i] = e * pred[i] * (1.0 - pred[i]);
         totalErr += e * e;
     }
+
     for (int i = numLayers - 2; i > 0; i--) {
         errors[i].resize(sizes[i]);
+
+        OMP_PARALLEL
             for (int j = 0; j < sizes[i]; j++) {
                 double e = 0;
-            for (int k = 0; k < sizes[i+1]; k++) e += errors[i+1][k] * weights[i][k][j];
+                for (int k = 0; k < sizes[i + 1]; k++) {
+                    e += errors[i + 1][k] * weights[i][k][j];
+                }
                 errors[i][j] = e * outputs[i][j] * (1.0 - outputs[i][j]);
             }
     }
+
     for (int i = 0; i < numLayers - 1; i++) {
-        for (int j = 0; j < sizes[i+1]; j++) {
+
-            for (int k = 0; k < sizes[i]; k++) weights[i][j][k] += lr * errors[i+1][j] * outputs[i][k];
+        OMP_PARALLEL
-            biases[i][j] += lr * errors[i+1][j];
+            for (int j = 0; j < sizes[i + 1]; j++) {
+                for (int k = 0; k < sizes[i]; k++) {
+                    weights[i][j][k] += lr * errors[i + 1][j] * outputs[i][k];
+                }
+                biases[i][j] += lr * errors[i + 1][j];
             }
     }
+
     return totalErr;
 }