UpCoder

So, the weight which named as up.weight is useless? In the inference state, could be the ConvTranspose2d operation replaced with the resize operation with the bilinear kernel?

Hi, May i ask a question? Is the selective region dependency with caffe? I use tensorflow, can I use it? If yes, how to install it? Thanks

Me too, Why did the error occur? Because the divided number is zero?

Thanks, I have solved the problem. I try to set the width and height attribute of Echart. It works. `chart = Echart('Loss', 'Here show the loss values', width=200, height=200)` Hope...

maybe you can try modify gpu_options = tf.GPUOptions(per_process_gpu_memory_fraction=0.95) in train.py to gpu_options = tf.GPUOptions() It works for me

can I use it with react-bootstrap(version > 0.30.0 )?

Thank you for replying, I download the new .so file, the error solved, but there is a new error as followed ` Traceback (most recent call last): File "inference1.py", line...

Thank you for replying. What’s the version of opencv? I switch to python2.7, but there is also error... AttributeError: 'module' object has no attribute 'boxPoints'

Yeah! It works, Thank you very much!

@SchroeterJulien Please see https://github.com/hosang/gossipnet/issues/14, I successfully run the code (TF 1.12.0)