Haojun Jiang(蒋昊峻)

I encountered the same problem as @Maryames . And I solved it with the help of @uahsan3 . If anyone else encounter this problem, please look at the trainval.prototxt generate...

It's a mismatch between batchsize( set by -bs ) and the first dimension in input layer (set in yournetwork.prototxt).

The basicblock is not implemented as a sk block, it is just a normal resnet basicblock.

I found another two mistakes: - **self.conv2 is not implemented as groupconv.** ``` self.conv2 = conv3x3(planes, planes, stride) self.bn2 = nn.BatchNorm2d(planes) self.conv2g = conv3x3(planes, planes, stride, padding=2, dilation=2, groups =...

I set the batchsize=256/lr=0.1, but the training result(top1-acc: 77.64) is much lower than paper reported(top1-acc: 78.24)! More details about hyperparameters are listed as below. The epoch setting is converted from...

Dose anyone solve it? I encounter the same error with torch-1.9.0.

I only used the model file which worked well on torch1.3.0/torchvision0.4.1a0+d94043a/python3.7.7.

I apologize for the unclear description. Let me make it more clear. - **./data/image_data/data/xxx/**: Take the Flickr30K dataset as an example, ./data/image_data/data/flickr/ shoud contain files about the dataset's validation/test annotations(bbox-query...

> [[email protected]](mailto:[email protected]) Hi, I send an email to you. Please leave a STAR ⭐ if this project help your research!

> [[email protected]](mailto:[email protected]) , thanks Hi, I send an email to you. Please leave a STAR ⭐ if this project help your research!