Add Alexnet README

examples/alexnet/README.md

@@ -0,0 +1,28 @@
+# AlexNet
+
+AlexNet Inference on CUDANet
+
+## Usage
+
+1. Export pytorch AlexNet weight pretrained on ImageNet (requires pytorch and torchvision):
+
+```sh
+python alexnet.py
+```
+
+2. Follow the instructions from repository root to build CUDANet library.
+
+3. Build AlexNet
+
+```sh
+mkdir build
+cd build
+cmake -S ..
+make
+```
+
+4. Run AlexNet inference
+
+```sh
+alexnet ../alexnet_weights.bin ../image.jpg
+```

examples/alexnet/main.cpp

@@ -1,14 +1,14 @@
+#include <conv2d.cuh>
+#include <dense.cuh>
 #include <iostream>
+#include <max_pooling.cuh>
+#include <model.hpp>
+#include <opencv2/opencv.hpp>
 #include <string>
 #include <vector>
-#include <opencv2/opencv.hpp>
 
-#include <model.hpp>
-#include <conv2d.cuh>
-#include <max_pooling.cuh>
-#include <dense.cuh>
-
-std::vector<float> readAndNormalizeImage(const std::string& imagePath, int width, int height) {
+std::vector<float>
+readAndNormalizeImage(const std::string &imagePath, int width, int height) {
     // Read the image using OpenCV
     cv::Mat image = cv::imread(imagePath, cv::IMREAD_COLOR);
 
@@ -35,13 +35,18 @@ std::vector<float> readAndNormalizeImage(const std::string& imagePath, int width
     return imageData;
 }
 
-CUDANet::Model* createModel(const int inputSize, const int inputChannels, const int outputSize) {
+CUDANet::Model *createModel(
+    const int inputSize,
+    const int inputChannels,
+    const int outputSize
+) {
     CUDANet::Model *model =
         new CUDANet::Model(inputSize, inputChannels, outputSize);
 
     // Block 1
     CUDANet::Layers::Conv2d *conv1 = new CUDANet::Layers::Conv2d(
-        inputSize, inputChannels, 11, 4, 64, 2, CUDANet::Layers::ActivationType::RELU
+        inputSize, inputChannels, 11, 4, 64, 2,
+        CUDANet::Layers::ActivationType::RELU
     );
     model->addLayer("features.0", conv1);  // Match pytorch naming
     CUDANet::Layers::MaxPooling2D *pool1 = new CUDANet::Layers::MaxPooling2D(
@@ -101,9 +106,9 @@ CUDANet::Model* createModel(const int inputSize, const int inputChannels, const
 }
 
 int main(int argc, const char *const argv[]) {
-
     if (argc != 3) {
-        std::cerr << "Usage: " << argv[0] << "<model_weights_path> <image_path>" << std::endl;
+        std::cerr << "Usage: " << argv[0] << "<model_weights_path> <image_path>"
+                  << std::endl;
         return 1;  // Return error code indicating incorrect usage
     }
 
@@ -122,7 +127,8 @@ int main(int argc, const char* const argv[]) {
     model->loadWeights(modelWeightsPath);
 
     // Read and normalize the image
-    std::vector<float> imageData = readAndNormalizeImage(imagePath, inputSize, inputSize);
+    std::vector<float> imageData =
+        readAndNormalizeImage(imagePath, inputSize, inputSize);
 
     // Print the size of the image data
     float *output = model->predict(imageData.data());