PyTorch 의 NCHW universe 에서 conv layer
PyTorch 는 image data 를 (N, C, H, W) 로: batch, channel, height, width. (다른 주요 convention 은 NHWC, TensorFlow 와 CoreML 사용 — framework 넘을 때 permute 준비.)
nn.Conv2d(in_channels, out_channels, kernel_size, stride=1, padding=0) 가 (kernel_size, kernel_size, in_channels) shape 의 out_channels filter 학습, 각각 input 위 slide, out_channels feature map 생산. output spatial size 표준 공식: out = (in + 2*padding - kernel) / stride + 1.
실제 쓰게 될 variant
Conv2dConv1dConvTranspose2dDepthwise separable conv — Conv2d(groups=in_channels) + Conv2d(1x1) 결합으로 짓기. MobileNet / EfficientNet 을 효율적으로 만드는 trick.
padding shorthand
PyTorch 1.10+ 에 padding='same' 을 string 으로 추가, stride=1 에 spatial dim 유지. 모두가 손으로 쓰던 'padding 직접 계산' 의 편한 버전. spatial dim 떨어뜨릴 specific 이유 없으면 사용.
Code Conv2d 기본· python import torch
import torch.nn as nn
# 3 RGB channels in, 16 feature maps out, 3x3 kernel
conv = nn.Conv2d(in_channels=3, out_channels=16, kernel_size=3,
stride=1, padding=1)
x = torch.randn(8, 3, 32, 32) # batch=8, RGB, 32x32 image
y = conv(x)
print(y.shape) # torch.Size([8, 16, 32, 32]) — same spatial
print(conv.weight.shape) # torch.Size([16, 3, 3, 3]) — out, in, kH, kW
print(conv.bias.shape) # torch.Size([16])
# stride=2 halves spatial dims
conv_down = nn.Conv2d(3, 16, 3, stride=2, padding=1)
print(conv_down(x).shape) # torch.Size([8, 16, 16, 16])간단 CNN — canonical pattern· python import torch
import torch.nn as nn
class SimpleCNN(nn.Module):
def __init__(self, num_classes=10):
super().__init__()
self.features = nn.Sequential(
nn.Conv2d(3, 32, 3, padding=1),
nn.BatchNorm2d(32),
nn.ReLU(),
nn.MaxPool2d(2), # 32x32 → 16x16
nn.Conv2d(32, 64, 3, padding=1),
nn.BatchNorm2d(64),
nn.ReLU(),
nn.MaxPool2d(2), # 16x16 → 8x8
nn.Conv2d(64, 128, 3, padding=1),
nn.BatchNorm2d(128),
nn.ReLU(),
nn.AdaptiveAvgPool2d(1), # global avg pool → 1x1
)
self.classifier = nn.Linear(128, num_classes)
def forward(self, x):
x = self.features(x)
x = x.flatten(1) # (B, 128, 1, 1) → (B, 128)
return self.classifier(x)
model = SimpleCNN()
print(model(torch.randn(4, 3, 32, 32)).shape) # torch.Size([4, 10])Depthwise separable conv — MobileNet trick· python import torch.nn as nn
class DepthwiseSeparable(nn.Module):
"""Replace a regular conv with depthwise + pointwise — far fewer params."""
def __init__(self, in_ch, out_ch, kernel=3):
super().__init__()
# Depthwise: each input channel gets its own kernel
self.depthwise = nn.Conv2d(in_ch, in_ch, kernel,
padding=kernel // 2, groups=in_ch)
# Pointwise: 1x1 conv to mix channels
self.pointwise = nn.Conv2d(in_ch, out_ch, kernel_size=1)
def forward(self, x):
return self.pointwise(self.depthwise(x))
# Compare param counts
regular = nn.Conv2d(64, 128, 3, padding=1)
sep = DepthwiseSeparable(64, 128, 3)
regular_params = sum(p.numel() for p in regular.parameters())
sep_params = sum(p.numel() for p in sep.parameters())
print(f"Regular: {regular_params:,}") # 73,856
print(f"Separable: {sep_params:,}") # 8,896 — about 8x fewerExercise 작은 ResNet-style block 짓기: Conv2d(64, 64, 3, padding=1) → BatchNorm → ReLU → Conv2d(64, 64, 3, padding=1) → BatchNorm, 그 다음 input 다시 더하기 (skip connection) 하고 final ReLU 전. spatial dim 과 channel 수 보존 검증.