The memory of GPU just keep increasing with iterations. I just post the reproducible code.
environment:
pytorch: 1.5.1
GPU: 1080Ti
import torch.nn as nn
import torch
from collections import OrderedDict
import torch.nn.functional as F
class _DenseLayer(nn.Sequential):
def __init__(self, num_input_features, growth_rate, bn_size, drop_rate):
super(_DenseLayer, self).__init__()
self.add_module('norm1', nn.BatchNorm2d(num_input_features))
self.add_module('relu1', nn.ReLU(inplace=True))
self.add_module('conv1', nn.Conv2d(num_input_features, bn_size * growth_rate,
kernel_size=1, stride=1, bias=False))
self.add_module('norm2', nn.BatchNorm2d(bn_size * growth_rate))
self.add_module('relu2', nn.ReLU(inplace=True))
self.add_module('conv2', nn.Conv2d(bn_size * growth_rate, growth_rate,
kernel_size=3, stride=1, padding=1, bias=False))
self.drop_rate = drop_rate
def forward(self, x):
# print(f"{self.__class__.__name__}: \nInput shape {x.shape}")
new_features = super(_DenseLayer, self).forward(x)
if self.drop_rate > 0:
new_features = F.dropout(new_features, p=self.drop_rate, training=self.training)
new_features = torch.cat([x, new_features], 1)
# print(f"output shape{output.shape}")
return new_features
class _DenseBlock(nn.Sequential):
def __init__(self, num_layers, num_input_features, bn_size, growth_rate, drop_rate):
'''
:param num_layers: number of layers in every block.
:param num_input_features: the channel of input data.
:param bn_size:
:param growth_rate:
:param drop_rate:
'''
super(_DenseBlock, self).__init__()
for i in range(num_layers):
layer = _DenseLayer(num_input_features + i * growth_rate, growth_rate, bn_size, drop_rate)
self.add_module('denselayer%d' % (i + 1), layer)
# Transition layer
class _Transition(nn.Sequential):
'''Transition layer between two adjacent DenseBlock'''
def __init__(self, num_input_features, num_output_features):
super(_Transition, self).__init__()
self.add_module('norm', nn.BatchNorm2d(num_input_features))
self.add_module('relu', nn.ReLU(inplace=True))
self.add_module('conv', nn.Conv2d(num_input_features, num_output_features,
kernel_size=1, stride=1, bias=False))
self.add_module('pool', nn.AvgPool2d(2, stride=2))
# densenet
class DenseNetTrail5D(nn.Module):
'''DenseNet-BC model'''
def __init__(self, init_channels=4, growth_rate=32, block_config=(6, 12, 24, 16), num_init_features=64,
bn_size=4, compression_rate=0.5, drop_rate=0, num_classes=1000):
"""
:param growth_rate: (int) number of filters used in DenseLayer, `k` in the paper
:param block_config: (list of 4 ints) number of layers in each DenseBlock
:param num_init_features: (int) number of filters in the first Conv2d
:param bn_size: (int) the factor using in the bottleneck layer
:param compression_rate: (float) the compression rate used in Transition Layer
:param drop_rate: (float) the drop rate after each DenseLayer
:param num_classes: (int) number of classes for classification
"""
super(DenseNetTrail5D, self).__init__()
# first conv2d
self.features = nn.Sequential(
OrderedDict([
('conv0', nn.Conv2d(init_channels, num_init_features, kernel_size=3, stride=1, padding=1, bias=False)),
('norm0', nn.BatchNorm2d(num_init_features)),
('relu0', nn.ReLU(inplace=True)),
('pool0', nn.MaxPool2d(3, stride=1, padding=1))
])
)
# Dense block
num_features = num_init_features
for i, num_layers in enumerate(block_config):
block = _DenseBlock(num_layers, num_features, bn_size, growth_rate, drop_rate)
self.features.add_module('denseblock%d' % (i + 1), block)
num_features += num_layers * growth_rate
if i != len(block_config) - 1:
transition = _Transition(num_features, int(num_features * compression_rate))
self.features.add_module('transition%d' % (i + 1), transition)
num_features = int(num_features * compression_rate)
# final bn + ReLU
self.features.add_module('norm5', nn.BatchNorm2d(num_features))
self.features.add_module('relu5', nn.ReLU(inplace=True))
# classification layer
# self.classfier = nn.Linear(num_features, num_classes)
self.linear = nn.Sequential(
nn.Linear(3200, 1),
)
# params initialization
for m in self.modules():
if isinstance(m, nn.Conv2d):
nn.init.kaiming_normal_(m.weight)
elif isinstance(m, nn.BatchNorm2d):
nn.init.constant_(m.bias, 0)
nn.init.constant_(m.weight, 1)
elif isinstance(m, nn.Linear):
nn.init.constant_(m.bias, 0)
def forward(self, x):
features = self.features(x.view(-1, 5, 13, 40, 40).permute(0, 2, 3, 1, 4).reshape(-1, 13, 40, 200))
out = F.avg_pool2d(features, (3, 9), padding=(1, 0), stride=(1, 4)).view(features.size(0), -1)
out = self.linear(out)
return out
if __name__ == '__main__':
from torchsummary import summary
import torch
sample = torch.randn((64, 5, 13*1600))
model = DenseNetTrail5D(13, block_config=(2, 2, 2, 2))
model = model.cuda()
# with torch.no_grad():
for i in range(5):
print(i)
y = model(sample.cuda()).cpu()
Any help is appreciated~
