Skip to content

Latest commit

 

History

History
36 lines (24 loc) · 1004 Bytes

step-02.01.02-Theano-Lasagne-Framework.md

File metadata and controls

36 lines (24 loc) · 1004 Bytes

Logistic Regression with Theano and Lasagne Framework [back]

At this point Theano is already installed in the system.

  1. Install lateset Lasagne framework:
sudo pip install Lasagne
  1. Run prepared emxaple with jupyter notebook:
$ cd ~/gitlab.altoros/776_DL_Libs_Benchmark.git/src/Step02/SubStep-01.02-Theano-Lasagne
$ jupyter notebook Theanoe_Lasagne_LogisticRegression.ipynb

Go to the URL: http://ec2-54-172-161-206.compute-1.amazonaws.com:9999/

To work with the MNIST dataset we will use a helper method from Tensorflow package:

from tensorflow.examples.tutorials.mnist import input_data
mnist = input_data.read_data_sets("MNIST_data/", one_hot=False)

Traing process:

Loss plot


Weights visualisation:

Weights