Data

Tools

Indexes

Applications

Experiments

Open Source Science

https://github.com/activeinferenceinstitute/rl-inference

Reinforcement Learning through Active Inference

https://github.com/activeinferenceinstitute/rl-inference

Science Score: 10.0%

This score indicates how likely this project is to be science-related based on various indicators:

  • ○
    CITATION.cff file
  • ○
    codemeta.json file
  • ○
    .zenodo.json file
  • ○
    DOI references
  • ✓
    Academic publication links
    Links to: arxiv.org
  • ○
    Academic email domains
  • ○
    Institutional organization owner
  • ○
    JOSS paper metadata
  • ○
    Scientific vocabulary similarity
    Low similarity (6.6%) to scientific vocabulary
Last synced: 9 months ago · JSON representation

Repository

Reinforcement Learning through Active Inference

Basic Info
  • Host: GitHub
  • Owner: ActiveInferenceInstitute
  • Default Branch: master
  • Homepage:
  • Size: 168 KB
Statistics
  • Stars: 0
  • Watchers: 0
  • Forks: 0
  • Open Issues: 0
  • Releases: 0
Fork of alec-tschantz/rl-inference
Created almost 5 years ago · Last pushed about 6 years ago

https://github.com/ActiveInferenceInstitute/rl-inference/blob/master/ 

# Reinforcement Learning through Active Inference

[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/licenses/MIT) 

Repository for ICLR 2020 submission to Bridging AI and Cognitive Science (BAICS)

[[Paper](https://arxiv.org/abs/2002.12636) / [Presentation](https://baicsworkshop.github.io/program/baics_37.html)]

### Requirements

The `environment.yml` contains the exact package versions, but recent versions of the following packages should be sufficient:

- `torch`
- `gym`
- `numpy`
- `scipy`

If you wish to use `mujoco` or `dm_control` environments you will also require:  

- `mujoco_py` 
- `dm_control` 

### Running

```
git clone https://github.com/alec-tschantz/rl-inference.git
cd rl-inference
python scripts/train.py --config_name="mountain_car" --logdir="mountain_car" --strategy="information" --seed=0
```

The provided configs are [`mountain_car`, `cup_catch`, `half_cheetah_run`,  `half_cheetah_flip` `ant_maze`, `reacher`]

### Acknowledgments

- Alec Tschantz [@alec-tschantz](https://github.com/alec-tschantz)
- Beren Millidge [@bmillidge](https://github.com/BerenMillidge)

The code for the ensemble model was adapted from the [max](https://github.com/nnaisense/max) repo

Owner

  • Name: Active Inference Institute
  • Login: ActiveInferenceInstitute
  • Kind: user
  • Location: Online
  • Company: Active Inference Institute

http://activeinference.org/

GitHub Events

Total
Last Year