shanglianlm0525 / pytorch-networks Goto Github PK

h

@shanglianlm0525 请问大佬您能否提供一下LPN的预训练模型吗

Hey Author,

the original version in your repo:
def forward(self, x):
# [N, C, H, W]
b, c, h, w = x.size()
# [N, C/2, H * W]
x_phi = self.conv_phi(x).view(b, c, -1)

I guess it should be revised to:
x_phi = self.conv_phi(x).view(b, c//2, -1)

Best,
Zewei

Hello and so happy to see you use Pytorch-Lightning! 🎉
Just wondering if you already heard about quite the new Pytorch Lightning (PL) ecosystem CI where we would like to invite you to... You can check out our blog post about it: Stay Ahead of Breaking Changes with the New Lightning Ecosystem CI ⚡
As you use PL framework for your cool project, we would like to enhance your experience and offer you safe updates to our future releases. At this moment, you run tests with a particular PL version, but it may accidentally happen that the next version will be incompatible with your project... 😕 We do not intend to change anything on our project side, but still here we have a solution - ecosystem CI with testing both - your and our latest development head we can find it very early and prevent releasing eventually bad version... 👍

What is needed to do?

• have some tests, including PL integration
• add config to ecosystem CI - https://github.com/PyTorchLightning/ecosystem-ci

What will you get?

• scheduled nightly testing configured for development/stable versions
• slack notification if something went wrong to investigate
• testing also on multi-GPU machine as our gift to you 🐰

Is there any pre-trained model on ImageNet dataset?

SolveMixnet:MDConv class self.layers add to the network by self.layers = nn.Sequential(*self.layers【Better change the name or something】)
Otherwise an error:RuntimeError: Expected object of device type cuda but got device type cpu for argument #2 'weight' in call to _thnn_conv_depthwise2d_forward

大佬你好，有怎么训练LFFD代码吗，或者预训练权重，谢谢啊

You forgot to put the ECA_modul into Bottleneck

As title, do you have the project of YOLT?
Thank you a lot!

厉害了，老哥~

I find your InceptionV4 code is not finished, so could you tell me which are finished and which are not?

class LWbottleneck(nn.Module):
def init(self, in_channels,out_channels,stride):
super(LWbottleneck, self).init()
self.stride = stride
self.pyramid_list = nn.ModuleList()
self.pyramid_list.append(ConvBNReLU(in_channels, in_channels, kernel_size=[5,1], stride=stride, padding=[2,0]))
self.pyramid_list.append(ConvBNReLU(in_channels, in_channels, kernel_size=[1,5], stride=stride, padding=[0,2]))
self.pyramid_list.append(ConvBNReLU(in_channels, in_channels, kernel_size=[3,1], stride=stride, padding=[1,0]))
self.pyramid_list.append(ConvBNReLU(in_channels, in_channels, kernel_size=[1,3], stride=stride, padding=[0,1]))
self.pyramid_list.append(ConvBNReLU(in_channels, in_channels, kernel_size=[2,1], stride=stride, padding=[1,0]))
self.pyramid_list.append(ConvBNReLU(in_channels, in_channels, kernel_size=[1,2], stride=stride, padding=[0,1]))
self.pyramid_list.append(ConvBNReLU(in_channels, in_channels, kernel_size=2, stride=stride, padding=1))
self.pyramid_list.append(ConvBNReLU(in_channels, in_channels, kernel_size=3, stride=stride, padding=1))

    self.shrink = Conv1x1BN(in_channels*8,out_channels)

def forward(self, x):
    b,c,w,h = x.shape
    if self.stride>1:
        w, h = w//self.stride,h//self.stride
    outputs = []
    for pyconv in self.pyramid_list:
        pyconv_x = pyconv(x)
        if x.shape[2:] != pyconv_x.shape[2:]:
            pyconv_x = pyconv_x[:,:,:w,:h]
        outputs.append(pyconv_x)
    out = torch.cat(outputs, 1)
    return self.shrink(out)

In the original paper, They used shotcut and add, which seems not to be in your code

您好，我是从csdn上的IA YOLO文章找到您的仓库的。但是我没有找到IA YOLO的代码，请问是在哪里呢？

Darknet19 is for YOLOv2, YOLOv3 needs Darknet53 and upsample.

Alexnet模型的各个conv层的通道数是不是错了呢？看论文中各个conv层的通道数应该分别是96, 256, 384, 384, 256啊。

大佬您好，我在您的仓库里面没有找到LFFD的训练代码，不知道您方不方便上传一下。感谢！！！

你好，关于MobileNetOne代码没找到，是在哪个路径下的呢？

Hello, may I ask if there are weights for the MobileneTxt network pre trained on the ImageNet dataset?

没有DCN的代码？