Implementation of iterative inference in deep latent variable models
Switch branches/tags
Nothing to show
Clone or download
Type Name Latest commit message Commit time
Failed to load latest commit information.
cfg Updating MNIST configs with log_var instead of log_variance Aug 13, 2018
lib Minor fixes Jul 5, 2018
util Minor fixes Jul 5, 2018
.gitignore allow for non-averaging of gradients, added .gitignore Sep 10, 2017 updating readme Jul 5, 2018 Updating configs. Adding argparse. Jul 4, 2018

Iterative Amortized Inference

Code to accompany the paper Iterative Amortized Inference by Marino et al., ICML 2018.

Installation & Set-Up

First, clone the repository by opening a terminal and running:

$ git clone

The code uses PyTorch version 0.3.0.post4 and visdom version 0.1.7. To avoid conflicts with more recent versions of these packages, you may wish to create a conda environment:

$ conda create --name it_inf python=2.7

To enter the environment, run:

$ source activate it_inf

Within the environment, install PyTorch by visiting the list of versions here, and grabbing version 0.3.0.post4 for your version of CUDA (8.0, 9.0, 9.1, etc.). Note that the code requires CUDA. Be sure to also install torchvision.

To install vidsom, run

(it_inf) $ pip install visdom==0.1.7

You will also need to install dill, a serialization package, and scipy:

(it_inf) $ pip install dill scipy

To exit the environment, run:

(it_inf) $ source deactivate

Running the Code

To use visdom for plotting, open an terminal and run

python -m visdom.server

Note that if you created a conda environment to use visdom version 0.1.7, you will need to enter that environment before activating visdom (see above).

The code can be run from the terminal using command line arguments. The arguments are

  • dataset,
  • model_type,
  • inference_type,
  • data_path, and
  • log_path.

For instance, to run a single-level model on MNIST with iterative inference, run:

python --dataset 'mnist' --model_type 'single_level' --inference_type 'iterative' --data_path '/path/to/data/' --log_path '/path/to/logs'

Be sure to replace the paths for data_path and log_path with valid paths to where the data and logs should be saved, respectively.

You can watch the training progress by opening a browser window and navigating to http://localhost:8097, and selecting the visdom environment corresponding to the experiment.