Skip to content

srk97/Reinforce

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

11 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Reinforce

PPO agent for RL environments. This code is an adaptation of FOR.ai's RL repository

Environments

  • Gym environments (Cartpole, MountainCar, Pong etc)
  • Gibson 3D Photo-realistic environment

Training

Environment

  • Cartpole: python train.py --hparams ppo_cartpole --sys local
  • Pong: python train.py --hparams ppo_pong --sys local
  • MountainCar: python train.py --hparams ppo_mountaincar --sys local
  • Gibson: python train.py --hparams ppo_gibson --sys local

To train asynchronously, set --num_workers flag to the number of worker threads.

A complete list of hyperparameters used for each environment can be found here for gym envs and here for Gibson

Installation

pip install -r requirements.txt

To be able to train on the Gibson environment, you would need to install the following packages

  • Install habitat_sim from here
  • Install habitat_api from here
  • Download the data from here and follow the structure mentioned on the repository
  • Place the data in a folder named data under Habitat-TF/

Structure

rl/
├── agents
│   ├── agent.py
│   ├── algos
│   │   ├── action_function
│   │   │   ├── basic.py
│   │   │   ├── __init__.py
│   │   │   └── registry.py
│   │   ├── compute_gradient
│   │   │   ├── basic.py
│   │   │   ├── __init__.py
│   │   │   ├── registry.py
│   │   │   └── utils.py
│   │   ├── gibson_ppo.py
│   │   ├── __init__.py
│   │   ├── ppo.py
│   │   ├── registry.py
│   │   └── utils.py
│   ├── __init__.py
│   └── registry.py
├── envs
│   ├── configs
│   │   ├── baselines
│   │   ├── datasets
│   │   │   └── pointnav
│   │   ├── tasks
│   │   └── test
│   ├── env.py
│   ├── gibson.py
│   ├── gym_env.py
│   ├── habitat_env
│   │   └── gibson.py
│   ├── __init__.py
│   ├── registry.py
│   ├── reward_augmentation.py
│   └── utils.py
├── hparams
│   ├── defaults.py
│   ├── gibson_ppo.py
│   ├── __init__.py
│   ├── ppo.py
│   ├── registry.py
│   └── utils.py
├── __init__.py
├── memory
│   ├── __init__.py
│   ├── memory.py
│   ├── registry.py
│   └── simple.py
├── models
│   ├── basic
│   │   ├── basic.py
│   │   └── __init__.py
│   ├── gibson_model
│   │   ├── gibson_model.py
│   │   └── __init__.py
│   ├── __init__.py
│   ├── model.py
│   └── registry.py
└── utils
    ├── checkpoint.py
    ├── flags.py
    ├── __init__.py
    ├── logger.py
    ├── lr_schemes.py
    ├── rand.py
    ├── sys.py
    └── utils.py


Acknowledgments

References

Releases

No releases published

Packages

No packages published

Languages