Deep Learning
One of the things I've been doing today is to advance just a little bit in the TensorFlow tutorial. My friend Jose and I decided to look at it together, and even if we go at a leisure pace, at least we are really looking into it. I'm very happy about the decision!
We opened a chat using slack, to try how it works. It seems you need to have an account to access it. I mostly like how it is done. The way you have different channels, can reference things, can upload media and so on. I don't like that it has a proprietary license though.
Related to that, I've seen a couple of alternatives I'd like to explore: matrix and mattermost.
Other than slack, we decided to keep a jupyter notebook to keep notes on what we are looking at. I love it! We've put it in github, here: https://github.com/jordibc/tensorflow_tutorial. There you can even see the notebooks without having to download the repository. Well played, github.
My currently open list of related pages to look at:
- http://colah.github.io/posts/2015-09-Visual-Information/
- http://colah.github.io/posts/2015-08-Backprop/
- https://en.wikipedia.org/wiki/Cross_entropy
- https://en.wikipedia.org/wiki/Softmax_function
That's it for my writing today!