clks-wzz

I found that this tf code uses "feed_dict" to feed the data for training. However, it's slow in tensorflow. I remember that "tfrecords" will speed up _io_ while training.

I want to utilize the pre-trained model for testing my data. However, I don't know the pre-processing method. For example, what's the normalization stype, ie, sub rgb average of images,...

Wish to copy the repo and change it for another topic (face anti-spoofing).

2

Hello, sir. I wish to copy the repo and implement a function for depth image in the face anti-spoofing topic and make it public. Does this violate the license? Thanks.

❓ [Question] Expected isITensor() to be true but got false Requested ITensor from Var, however Var type is c10::IValue

4

I try to use python trtorch==0.4.1 to compile my own pytorch jit traced model, and I find that it goes wrong with the following information: ` Traceback (most recent call...

Some confusion about the data preprocessing.

1

I found that the "preocessing.py" codes use a simple transforms.RandomResizedCrop(224) to preprocess the input training data. Why don't you utilize aspect preserving resize before cropping? Thanks.

I found that there is no moving_average_mean/var, which is necessary in BatchNorm. Is it ok?

2

Hello, sir. I found that there is no moving_average_mean/var, which is necessary in BatchNorm. Is it ok for GroupNorm?

The real Params in my experiment

2

There are 33.6M when I use the 'saved_model' pattern to save the mobilenetv3_large_1.0 models. Is there anything wrong that I can deal with?

Does your shufflenet implement random shuffling?

1

Sir, I saw that you only split the channels and implement group convolution. Shufflenet needs random shuffling the channels. But I can't find that. Would you mind helping me for...