spinaotey / snl_tf

Sequential Neural Likelihood - Tensorflow Implementation
Other
2 stars 1 forks source link

Sequential Neural Likelihood estimation - Tensorflow implementation

This repository's aim is to adapt G. Papamakarios' Sequantial Neural Likelihood method from Theano to Tensorflow.

The method heavily depends on the Masked Autoregressive Flow, which has been already translated into Tensorflow in my other repository.

How to setup the snl_tf Python library.

This is a Python 3 library (current version used 3.6+). In order to install this library first install the required packages listed in requirements.txt (pip install -r requirements.txt)). Then just use pip install -e . in the main folder of the repository (where the setup.py is situated). After that, you should be able to import the package as usual with import snl_tf.