search

Unity-Technologies / ml-agents

The Unity Machine Learning Agents Toolkit (ML-Agents) is an open-source project that enables games and simulations to serve as environments for training intelligent agents using deep reinforcement learning and imitation learning.
https://unity.com/products/machine-learning-agents
Other
17.19k stars 4.16k forks source link

[Feature Request] Can you perform training multiple times within a build? #1299

Closed ConorZAM closed 5 years ago

ConorZAM commented 6 years ago

I am looking to combine a genetic algorithm with the RL implemented by ML agents.

I was wondering if you are (or will be) working on a way of initiating training in game? I would like to be able to create an agent, train it and evaluate it and then alter the agent over generations to simultaneously find optimal policy and topology for a given task.

If this isn't something you're looking at I'll get cracking, I just don't want to spend the time developing this only to see you release it anyway!

Thanks!

bjmolitor commented 6 years ago

You may use mlagents-learn with parameter --load to continue training and you might want to take a look at the curriculum learning guide in the docs folder of this project.

awjuliani commented 5 years ago

Hi @ConorZAM

We currently do not support training on-devices for running games.

awjuliani commented 5 years ago

Due to inactivity, I am closing this issue for now. Please feel free to re-open if you deem it necessary.

github-actions[bot] commented 3 years ago

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.