bzamecnik

"""
When classifying upon a sequence usually we stack some LSTM returning sequences,
then one LSTM returning a point, then Dense with softmax activation.

Is it possible instead to give the last non-sequential LSTM a softmax activation?

The answer is yes.

In this example we have 3 sequential layers and one layer producing the final result.

simoncoulton

======================================
Setting up Nginx, uWSGI and Python3
======================================

First off, I'm traditionally a PHP developer, but am looking at moving across to Python. I really struggled to find decent documentation on how to get a server up and running for deploying Python web applications from the point of view of someone coming from PHP. The main problems I came across with documentation were:

1) Only showed you how to run the server for a single web application.
2) Only showed you how to configure the app, not the server it was running on.

My preferred workflow for development is by setting up a new VM in VMware Fusion and then forwarding through all requests to that VM via /etc/hosts. This might not be the optimal way to get things up and running, but it works for me.