Difference between revisions of "ML links"

From Interaction Station Wiki
Jump to navigation Jump to search
Line 2: Line 2:
 
*https://tangzhenyu.github.io/deep_learning/2015/01/09/fun-with-deep-learning.html
 
*https://tangzhenyu.github.io/deep_learning/2015/01/09/fun-with-deep-learning.html
  
== Installation ==
 
 
*We need to install our modified version of Darknet:
 
*[[Darknet Installation]]
 
 
== RNNs with Darknet ==
 
 
*Getting some calculated weights from several writers:
 
wget https://pjreddie.com/media/files/shakespeare.weights
 
wget https://pjreddie.com/media/files/kant.weights
 
wget https://pjreddie.com/media/files/shakespeare.weights
 
wget https://pjreddie.com/media/files/shakespeare.weights
 
 
*Generating new text
 
cd darknet
 
./darknet rnn generate cfg/rnn.cfg shakespeare.weights -srand 0
 
 
*Parameters:
 
-srand N -> N determines the random seed
 
-
 
 
 
*Training with other corpus of text
 
*Modify file cfg/rnn.train.cfg and set:
 
batch = 50
 
 
 
*Some online corpus of text:
 
George Meredith
 
http://www.gutenberg.org/ebooks/4500
 
Mark Twain
 
http://www.gutenberg.org/ebooks/3200
 
More
 
http://www.gutenberg.org/ebooks/search/?query=The+Complete+Works+of
 
 
 
 
*More info:
 
*https://pjreddie.com/darknet/rnns-in-darknet/
 
  
 
== Tutorial generative models ==
 
== Tutorial generative models ==

Revision as of 22:35, 4 January 2019