Skip to content
Starter Kit for the NeurIPS 2019 Disentanglement Challenge
Python Shell
Branch: master
Clone or download
Latest commit 17ade89 Aug 6, 2019


NeurIPS 2019 : Disentanglement Challenge Starter Kit


Instructions to make submissions to the NeurIPS 2019 : Disentanglement Challenge.

Participants will have to submit their code, with packaging specifications, and the evaluator will automatically build a docker image and execute their code against a series of secret datasets.


  • Anaconda (By following instructions here) At least version 4.5.11 is required to correctly populate environment.yml.
  • Create your new conda environment
conda create python=3.6 --name disentanglement_challenge
conda activate disentanglement_challenge
  • Your code specific dependencies
# If say you want to install PyTorch
conda install pytorch torchvision -c pytorch

Clone repository

git clone
cd neurips2019_disentanglement_challenge_starter_kit
pip install -r requirements.txt

Download and Prepare Dataset

Follow the instructions here to download the publicly available datasets, and store them inside ./scratch/datasets folder. Please ensure that the datasets are not checked into the repository.

Train and Test Submission locally with Tensorflow / Pytorch / Numpy


The first three steps apply to all frameworks.

# Step 1:
cd neurips2019_disentanglement_challenge_starter_kit

# Step 2: 
# Make sure that your datasets live in "./scatch/dataset". 
# If they do not, either make sure they do (recommended)
# or change accordingly. Finally:

# Step 3:
# Enter the name of your experiment
# Enter the name of the dataset


# Step 4 (PyTorch): 
# Train a VAE with Pytorch
python pytorch/

Take a look at the python file for available commandline arguments.


# Step 4 (TensorFlow)
# Train a VAE with TF and disentanglement_lib 
python tensorflow/

Note: This is just a test submission which will only train for 10 steps (to keep training time short). To attain reasonable performance, about 300'000 training steps are recommended. This setting can be changed in tensorflow/model.gin.


# Step 4 (NumPy): 
# This saves a hard-coded representation function to file.
python numpy/


Finally, it's time to run the evaluation locally. This script will detect what framework you used, provided you used the right utility functions for PyTorch or NumPy (links lead to examples).

Step 5:
# Run the local evaluation

How do I specify my software runtime ?

The software runtime is specified by exporting your conda env to the root of your repository by doing :

conda env export --no-build > environment.yml

# It might make sense here to remove the requirements.txt here to avoid confusion as environment.yml takes precedence over requirements.txt

This environment.yml file will be used to recreate the conda environment. This repository includes an example environment.yml

What should my code structure be like ?

Please follow the structure documented in the included to adapt your already existing code to the required structure for this round.

Important Concepts

Repository Structure

  • aicrowd.json Each repository should have a aicrowd.json with the following content :
  "challenge_id": "aicrowd-neurips-2019-disentanglement-challenge",
  "grader_id": "aicrowd-neurips-2019-disentanglement-challenge",
  "authors": ["your-aicrowd-username"],
  "description": "sample description about your awesome agent",
  "license": "MIT",
  "gpu": true

This is used to map your submission to the said challenge, so please remember to use the correct challenge_id and grader_id as specified above.

Please specify if your code will a GPU or not for the evaluation of your model. If you specify true for the GPU, a NVIDIA Tesla K80 GPU will be provided and used for the evaluation.

Packaging of your software environment

The recommended way is to use Anaconda configuration files using environment.yml files.

# The included environment.yml is generated by the command below, and you do not need to run it again
# if you did not add any custom dependencies

conda env export --no-build > environment.yml

# Note the `--no-build` flag, which is important if you want your anaconda env to be replicable across all
# Or if you have a rather simple softwar runtime, then even a :
# pip freeze > requirements.txt
# could work.

Code Entrypoint

The evaluator will use /home/aicrowd/ as the entrypoint, so please remember to have a at the root, which can instantitate any necessary environment variables, and also start executing your actual code. This repository includes a sample file. If you are using a Dockerfile to specify your software environment, please remember to create a aicrowd user, and place the entrypoint code at

Frequently Asked Questions (FAQs)

Please find a list of frequently asked questions here, and please feel free to send a pull request with more questions that you think can help other participants.


To make a submission, you will have to create a private repository on

You will have to add your SSH Keys to your GitLab account by following the instructions here. If you do not have SSH Keys, you will first need to generate one.

Then you can create a submission by making a tag push to your repository on Any tag push (where the tag name begins with "submission-") to your private repository is considered as a submission
Then you can add the correct git remote, and finally submit by doing :

cd neurips2019_disentanglement_challenge_starter_kit
# Add AIcrowd git remote endpoint
git remote add aicrowd<YOUR_AICROWD_USER_NAME>/neurips2019_disentanglement_challenge_starter_kit.git
git push aicrowd master

# Create a tag for your submission and push
git tag -am "submission-v0.1" submission-v0.1
git push aicrowd master
git push aicrowd submission-v0.1

# Note : If the contents of your repository (latest commit hash) does not change,
# then pushing a new tag will **not** trigger a new evaluation.

You now should be able to see the details of your submission at :<YOUR_AICROWD_USER_NAME>/neurips2019_disentanglement_challenge_starter_kit/issues

NOTE: Remember to update your username in the link above 😉

In the link above, you should start seeing something like this take shape (each of the steps can take a bit of time, so please be patient too 😉 ) :

and if everything works out correctly, then you should be able to see the final scores like this :

Best of Luck 🎉 🎉


Sharada Mohanty

You can’t perform that action at this time.