Lei Mao

That is because the discriminator became too strong while the generator was not strong enough. We want to have generator loss to decrease all the time, but some times it...

Thanks for the contribution, again. I will start to review shortly.

This problem is also reproducible on my side. ``` # make time_two_test 2020-09-29 23:28:14.648088: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1 2020-09-29 23:28:16.572897: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1...

I got the same problem. This `cuda_fp16.h` file is missing (I installed TensorFlow via pip). I also checked the TensorFlow master branch, there is no such file in the `third_party`...

> export GPG_TTY=$(tty) This trick works on my Ubuntu 18.04 LTS.

> With a different config structure it works correctly now: > > ``` > deploy: > type: git > repo: > github: > url: https://github.com//.git > branch: gh-pages > token:...

I think this should be the correct way of deploying the website using token, although the token unfortunately has to be concretely typed in the yaml file. https://github.com/hexojs/hexo/issues/2084#issuecomment-466350089

> For a library that is built strictly to support the Python bindings, I decided the most appropriate solution was to build it as a static library and link it...

You did not provide a command to reproduce. I have updated the repo recently and please use the Docker container to reproduce the examples. I don't help troubleshooting without using...

You could figure out a way to load checkpoint models as a Keras model, then you could try to export a frozen model using the methods I described.