PyGeneses is a PyTorch based Deep Reinforcement Learning framework that helps users to simulate artificial agents in bio-inspired environments. The framework provides in-built Deep RL algorithms and visualization of results from training in an interactive dashboard. The environments are a simplified abstraction of our real world, agents are put into this world and are allowed to interact with each other and the environment.
The power of PyGeneses is its clean and simple API which:
- Allows a user to run own simulation even if they have no knowledge about RL or DL.
- Allows a user with experience in Deep RL to tweak the model and parameters extensively.
What’s exciting about PyGeneses you ask, well, PyGeneses lets you create a working simulation by writing as little as 2 lines without any prerequisite knowledge whatsoever!! PyGeneses also provides tools that will help you visualize the results with minimal effort. So what are you waiting for, install PyGeneses today and become the god of your creation
Hacktoberfest 2020 is here. Contributions are now welcome. Please go through CONTRIBUTING.md and the instructions in readme Contribute. We look forward for your contributions.
Prima vita is a species of artificially simulated beings created as part of Project DC. This repository holds a simulation environment created in pygame which is to be used with Deep Reinforcement Learning algorithms to find out the evolution of Prima Vita.
PyGeneses can be installed using pip in either your local system or a cloud based platform. The steps for installation will be the same for both cloud and local installation.
user@programmer~:$ pip install git+https://github.com/Project-DC/pygenesesSince PyGeneses is not available in pypi yet, so you will have to use the github repo link with pip to install it for now.
As of version 0.1-beta, the architecture of PyGeneses is built around 4 major modules. Each of these modules provides a unique feature or functionality to the framework. So, let’s get started with a brief introduction to each of them.
- pygeneses.envs
This module allows you to create, configure and tweak the in-built bio-inspired environments. As of now, this only provides a single environment called Prima Vita (First Life), but there’s more coming soon! This lets you set up the entire environment and the species in just a few lines of code and provides both high level API and low level control over the environment. Training using the API includes logging of every action of an agent so that it can be studied using VitaBoard.
- pygeneses.models
The ‘models’ module is what allows us to import the neural networks which the species uses to learn what to do. As of now, only the default model's (REINFORCE) implementation is provided, but we will be adding support for custom pluggable networks from v0.2 onwards.
- pygeneses.hypertune
The ‘HyperTune’ package allows us to configure and test out various hyperparameters we can provide for an environment and species (a list of hyperparameters is provided in the Classes section of this documentation). This contains single hyperparameter testing, grid search and randomized search. This allows us to find the best set of hyperparameters to display a type of behavior. This also produces logs which we can study using Vitaboard.
- pygeneses.vitaboard
Vitaboard provides an advanced, interactive dashboard to study agents after the training phase. After each agent dies, his/her actions are written into a log file. And vitaboard allows us to visualize the agent's life. It provides us with a life visualizer, group statistics and a genetic history visualizer. It allows us to identify and understand behaviours exhibited by an agent while interacting with the environment or with other agents in the environment.
The following resources are a good place to get to know more about PyGeneses:-
- Introduction to PyGeneses Dev.to, Medium.
- Getting Started with PyGeneses Dev.to, Medium.
- Studying logs using VitaBoard Dev.to, Medium
Apart from these blog posts, you can also checkout the official docs.
Instructions for first time contributors/beginner level contributors for question related issues during HACKTOBERFEST
- Start working on the issues once you are assigned to them. Head over to the issue and comment that you want it to be assigned to you. Once the maintainer assigns the issue to you, start working on it. Issues will be assigned on a First Come First Serve (FCFS) basis.
- Once the issue is assigned, you have one week (7 Days) to submit the PR. Failing to do so will get the issue reassigned to someone else. As each issue related to questions are being assigned to single contributor at a time, we sincerely hope that you cooperate with us.
- If you create a PR without the issue being assigned to you, the PR will be marked spam as you are not adhering to the rules.
- The google drive link to logs that you generate have to placed in the location logs/ directory. If not found in the right place, our maintainers would write a comment to the PR as a warning and if the correct location is still not provided on the resubmission, the PR would be marked spam for not adhering to the rules.
-
Check which hyperparameter you have to tune, the values that you have to tune it for, and the stop_at number in the issue.
-
Write code in pygeneses for that (3 line code). Let's take an example where hyperparameter to be tuned is initial_population, the values for that are [10, 20, 50, 90, 100], and stop_at number is 2000, then the code will look something like this:-
from pygeneses.hypertune import HyperTune
tuner = HyperTune(model_class='PrimaVita',
hyperparameters=['initial_population'],
values=[[10, 20, 50, 90, 100]],
stop_at=2000)
tuner.hypertuner()-
After training there will be a folder generated in the same location where you trained the Prima vita agents, this folder's name will start with Players_Data, this is the logs that we require. You can either zip this or directly upload the entire folder to google drive.
-
Once you have uploaded the logs into google drive share the link of that folder (or zip file) containing logs in a file named with the issue number in txt format (e.g. if you are doing issue 11 then file name should be 11.txt). This file is to be put in the logs directory before creating the Pull Request. Once you create PR wait for a maintainer to merge it or ask for some changes.
Before moving further please go through the rules in CONTRIBUTING.md
PyGeneses is licensed under GNU GPL v3 LICENSE