This project extends the idea of the innovative architecture of Kolmogorov-Arnold Networks (KAN) to the Convolutional Layers, changing the classic linear transformation of the convolution to learnable non linear activations in each pixel.
Great work!
I want to use KAN conv instead of ResNet conv, how can i do it?
First, from kan_convolutional.KANConv import KAN_Convolutional_Layer from https://github.com/AntonioTepsich/Convolutional-KANs.
Second, change the ResNet conv
But there is a problem with the code, can you help me solve it?
import torch.nn as nn
import torch
from kan_convolutional.KANConv import KAN_Convolutional_Layer
def conv3x3(in_planes, out_planes, stride=1):
"""3x3 convolution with padding"""
return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
padding=1, bias=False)
def conv1x1(in_planes, out_planes, stride= 1):
"""1x1 convolution"""
return nn.Conv2d(in_planes, out_planes, kernel_size=1, stride=stride, bias=False)
class BasicBlock(nn.Module):
expansion = 1
def __init__(self, in_channel, out_channel, stride=(1,1), downsample=None, device: str = 'cuda:0'):
super(BasicBlock, self).__init__()
# use the kan convolutional
self.conv1 = KAN_Convolutional_Layer(
n_convs = 5,
kernel_size= (3,3),
stride=stride,
device = device
)
self.conv2 = KAN_Convolutional_Layer(
n_convs = 5,
kernel_size= (3,3),
device = device
)
# self.conv1 = conv3x3(in_channel, out_channel, stride)
self.bn1 = nn.BatchNorm2d(15)
self.relu = nn.ReLU(inplace=True)
# self.conv2 = conv3x3(out_channel, out_channel)
self.bn2 = nn.BatchNorm2d(15)
self.downsample = downsample
self.stride = stride
def forward(self, x):
residual = x
print(f'Input shape: {x.shape}')
out = self.conv1(x) # 3x3conv，s=1
print(f'After conv1 shape: {out.shape}')
out = self.bn1(out)
out = self.relu(out)
out = self.conv2(out)
out = self.bn2(out)
if self.downsample is not None:
residual = self.downsample(x)
out += residual
out = self.relu(out)
return out
class Bottleneck(nn.Module):
expansion = 4
def __init__(self, in_channel, out_channel, stride=1, downsample=None):
super(Bottleneck, self).__init__()
self.conv1 = nn.Conv2d(in_channels=in_channel, out_channels=out_channel,
kernel_size=1, stride=1, bias=False) # squeeze channels
self.bn1 = nn.BatchNorm2d(out_channel)
# -----------------------------------------
self.conv2 = nn.Conv2d(in_channels=out_channel, out_channels=out_channel,
kernel_size=3, stride=stride, bias=False, padding=1)
self.bn2 = nn.BatchNorm2d(out_channel)
# -----------------------------------------
self.conv3 = nn.Conv2d(in_channels=out_channel, out_channels=out_channel*self.expansion,
kernel_size=1, stride=1, bias=False) # unsqueeze channels
self.bn3 = nn.BatchNorm2d(out_channel*self.expansion)
self.relu = nn.ReLU(inplace=True)
self.downsample = downsample
self.stride = stride
def forward(self, x):
residual = x
out = self.conv1(x)
out = self.bn1(out)
out = self.relu(out)
out = self.conv2(out)
out = self.bn2(out)
out = self.relu(out)
out = self.conv3(out)
out = self.bn3(out)
if self.downsample is not None:
residual = self.downsample(x)
out += residual
out = self.relu(out)
return out
class ResNet(nn.Module):
def __init__(self, block, blocks_num, num_classes=1000, include_top=True, device: str = 'cuda:0'):
super(ResNet, self).__init__()
self.include_top = include_top
self.in_channel = 64
# self.conv1 = nn.Conv2d(3, self.in_channel, kernel_size=7, stride=2,
# padding=3, bias=False)
self.conv1 = KAN_Convolutional_Layer(
n_convs = 5,
kernel_size= (7,7),
stride=(2,2),
padding=(3,3),
device = device
)
self.bn1 = nn.BatchNorm2d(15)
self.relu = nn.ReLU(inplace=True)
self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
self.layer1 = self._make_layer(block, 64, blocks_num[0])
self.layer2 = self._make_layer(block, 128, blocks_num[1], stride=2)
self.layer3 = self._make_layer(block, 256, blocks_num[2], stride=2)
self.layer4 = self._make_layer(block, 512, blocks_num[3], stride=2)
if self.include_top:
self.avgpool = nn.AdaptiveAvgPool2d((1, 1)) # output size = (1, 1)
self.fc = nn.Linear(512 * block.expansion, num_classes)
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
def _make_layer(self, block, channel, block_num, stride=1, device: str = 'cuda:0'):
downsample = None
if stride != 1 or self.in_channel != channel * block.expansion:
downsample = nn.Sequential(
KAN_Convolutional_Layer(n_convs = 5, kernel_size= (1,1), stride=(stride,stride), device = device),
# nn.Conv2d(self.in_channel, channel * block.expansion, kernel_size=1, stride=stride, bias=False),
# nn.BatchNorm2d(KAN_Convolutional_Layer(n_convs = 5, kernel_size= (1,1), stride=(stride,stride), device = device).convs[0].conv.in_features)
nn.BatchNorm2d(15)
)
layers = []
layers.append(block(self.in_channel, channel, downsample=downsample, stride=stride))
self.in_channel = channel * block.expansion
for _ in range(1, block_num):
layers.append(block(self.in_channel, channel))
return nn.Sequential(*layers)
def forward(self, x):
x = self.conv1(x)
x = self.bn1(x)
x = self.relu(x)
x = self.maxpool(x)
x = self.layer1(x)
x = self.layer2(x)
x = self.layer3(x)
x = self.layer4(x)
if self.include_top:
x = self.avgpool(x)
x = torch.flatten(x, 1)
x = self.fc(x)
return x
def resnet18(num_classes=1000, include_top=True):
return ResNet(BasicBlock, [2, 2, 2, 2], num_classes=num_classes, include_top=include_top)
Great work! I want to use KAN conv instead of ResNet conv, how can i do it?
First,
from kan_convolutional.KANConv import KAN_Convolutional_Layerfrom https://github.com/AntonioTepsich/Convolutional-KANs. Second, change the ResNet conv But there is a problem with the code, can you help me solve it?