I feel a little bad about linking to this because it’s such a clear advertisement for TF. But…it’s a good post and a quick read. I actually didn’t realize they had added an eager execution environment! From the docs
TensorFlow’s eager execution is an imperative programming environment that evaluates operations immediately, without building graphs: operations return concrete values instead of constructing a computational graph to run later.
Seems like a really significant ease-of-use improvement. Many other useful tidbits in the post.