Skip to content

Experiments using TF, Torch and Ray with Distributed RL

Notifications You must be signed in to change notification settings

seanjparker/distributed-rl

Repository files navigation

Distributed Reinforcement Learning

Requirements

  • Python 3 (>=3.7)
  • Pipenv

Setup

As long as you have Python3 and Pipenv installed, run the following command:

$ pipenv install

Pipenv will automatically create and install all the required dependencies for the project.

Training

Running the following command launches the trainer in the selected framework using a list of worker sizes. By default it runs for 2, 4, 8, 16 and 32 workers.

$ pipenv shell # Activate the virtualenv created by Pipenv
$ ./launch -f [tf|torch|ray] -e [number of epochs]

About

Experiments using TF, Torch and Ray with Distributed RL

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published