dukguu/alexnet-from-scratch
0
AlexNet — From-Scratch PyTorch Implementation
Paper: ImageNet Classification with Deep Convolutional Neural Networks Authors: Alex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton (NeurIPS 2012)
Architecture
Input (224×224×3)
▼ Conv1: 96 × 11×11 / stride 4 → ReLU → LRN → MaxPool
▼ Conv2: 256 × 5×5 / pad 2 → ReLU → LRN → MaxPool
▼ Conv3: 384 × 3×3 / pad 1 → ReLU
▼ Conv4: 384 × 3×3 / pad 1 → ReLU
▼ Conv5: 256 × 3×3 / pad 1 → ReLU → MaxPool
▼ FC6: 4096 → ReLU → Dropout(0.5)
▼ FC7: 4096 → ReLU → Dropout(0.5)
▼ FC8: 1000 (Softmax at inference)