Adding usage instructions

[thundergolfer]

Hey,

You have done some really good and interesting work here, but I would love a README to be added. Just some basic information on how to:

1. Train the network
2. Evaluate the network
3. Find the produced modules
4. Run the train models on new data

Thanks mate

[amueller]

There is not network.
For better instructions, buy my book ;)

[thundergolfer]

Ah true of course. Does your book tell me how to get this code working? I am trying to detect insults in comments specifically, and not learn about Kaggle and Data Science generally.

[amueller]

Well it explains every method used in the code. This repository is not meant as a general solution to detect insults in comments. I would highly recommend learning about machine learning before trying to use it.

[chrisspradling1980]

I am fearful and grateful to read this. 9 years and not hardly any productivity to expound about.

Sir, your book sounds nice but, machine learning is the methodology in practice which organized this repository. In most cases, us inexperienced freelance people move through subjects while source extracting it is difficult to "a whole book" apply only a couple commonalities to the function usage.

Could you please add the words shortened in all the txt files which have bad words organized? The. Words. D1c, dic, dic-out, d1c-out, dic_0ut and d1c_0ut

I would be grateful if you could. Fearful because the world has become evoked into higher conscious exampling and as this event has progressed, the period of time these particular phrases have ridden, they have been hidden within this github. Keep in mind of the funneling funniest financial et. seq America ever seen in political crowd funding. Your crowd funding resulted in cross crowd sourcing and created Emmerich. With all you can eat Merks.. Nawmean? Little help please?
I'm only connected to 2 repositories, one is yours and the other is DHS CISA Vulnerability Policy.
3188203111
You should call so we can communicate directly.

Congrats and Respectfully Thanking you