Thanks, I am looking forward to that. TensorFlow is the most complicated deep learning library I have ever tried, almost every operations have to crawl through tons of documents, search the tutorial on web, for me there are many designs looks weird and always give me “surprise”, I study it because I cannot find a better choice(api are ultra complicated, Google engineers must be super smart because they do not think these api are hard to use).

If there are another feature riches, well maintain deep learning design for production use, I would definitely give it a try.

One clap, two clap, three clap, forty?

By clapping more or less, you can signal to us which stories really stand out.