Skip to content

Commit

Permalink
Only mention docker image in ICLR19 branch. (#70)
Browse files Browse the repository at this point in the history
I'd like to avoid a frequent source of confusion, which is people
trying to use the docker image and demo dataset with code from
the master branch (which doesn't work). Instead, I propose we
have a link to the iclr19 branch on master, and only link to
the docker and demo dataset from the iclr19 branch.
  • Loading branch information
maximecb committed May 2, 2019
1 parent 312ffcf commit b1ab9d1
Show file tree
Hide file tree
Showing 2 changed files with 6 additions and 26 deletions.
25 changes: 6 additions & 19 deletions README.md
Expand Up @@ -2,13 +2,13 @@

[![Build Status](https://travis-ci.org/mila-iqia/babyai.svg?branch=master)](https://travis-ci.org/mila-iqia/babyai)

A platform for simulating language learning with a human in the loop. This is an ongoing research project based at [Mila](https://mila.quebec/en/).
A platform for simulating language learning with a human in the loop. This is an ongoing research project based at [Mila](https://mila.quebec/en/).

Contents:
- [Citation](#citation)
- [Replicating ICLR19 Results](#replicating-iclr19-results)
- [Installation](#installation)
- [Usage](#usage)
- [Demo Dataset](docs/demo-dataset.md)
- [Codebase Structure](docs/codebase.md)
- [Levels](#the-levels)
- [Training and Evaluation](docs/train-eval.md)
Expand All @@ -30,6 +30,10 @@ If you use this platform in your research, please cite:
}
```

## Replicating ICLR19 Results

The master branch of this repository is updated frequently. If you are looking to replicate or compare against the results from the [ICLR19 BabyAI paper](https://openreview.net/forum?id=rJeXCo0cYX), please use the docker image, demonstration dataset and source code from the [iclr19 branch](https://github.com/mila-iqia/babyai/tree/iclr19) of this repository.

## Installation

Requirements:
Expand Down Expand Up @@ -77,22 +81,6 @@ cd ../babyai
pip install --editable .
```

### Docker Image

A prebuilt docker image is available [on Docker Hub](https://hub.docker.com/r/maximecb/babyai/). You can download this image by executing:

```
docker pull maximecb/babyai
```

You should run the image with `nvidia-docker` (which allows you to use CUDA):

```
nvidia-docker run -it maximecb/babyai bash
```

Pretrained IL and RL models can be found in the `models` directory of the image.

### BabyAI Storage Path

Add this line to `.bashrc` (Linux), or `.bash_profile` (Mac).
Expand Down Expand Up @@ -126,7 +114,6 @@ Documentation for the ICLR19 levels can be found in
There are also older levels documented in
[docs/bonus_levels.md](docs/bonus_levels.md).


## About this Project

BabyAI is an open-ended grounded language acquisition effort at [Mila](https://mila.quebec/en/). The current BabyAI platform was designed to study data-effiency of existing methods under the assumption that a human provides all teaching signals
Expand Down
7 changes: 0 additions & 7 deletions docs/demo-dataset.md

This file was deleted.

0 comments on commit b1ab9d1

Please sign in to comment.