Search Unity

  1. Welcome to the Unity Forums! Please take the time to read our Code of Conduct to familiarize yourself with the forum rules and how to post constructively.
  2. We have updated the language to the Editor Terms based on feedback from our employees and community. Learn more.
    Dismiss Notice
  3. Join us on November 16th, 2023, between 1 pm and 9 pm CET for Ask the Experts Online on Discord and on Unity Discussions.
    Dismiss Notice
New Forum User Notice Update to the Unity Editor Software Terms Unite 2023 - Ask the Experts Online

Is there a way using own algorithm on training agents?

Discussion in 'ML-Agents' started by Xinzz, Apr 1, 2020.

  1. Xinzz

    Xinzz

    Joined:
    Jun 28, 2015
    Posts:
    67
    For example, using my own implement of dqn as the training algorithm instead of the ppo or sac that have been pre-written? and of course using it in my own created game environment for RL?

    Thanks.
     
    Xinzz, Apr 1, 2020
    #1

  2. Print_Hello_World

    Print_Hello_World

    Joined:
    Jan 14, 2020
    Posts:
    12
    Think you may have to use the python API and manually integrate your own algo. Not so sure how you're gonna do it but look up unity ML-Agents python API
     
    Print_Hello_World, Apr 3, 2020
    #2

  3. Xinzz

    Xinzz

    Joined:
    Jun 28, 2015
    Posts:
    67
    Use the unity gym wrapper will do, but only one agent at a time, so i think i will just stick with ppo or sac that has been prewritten by unity.
     
    Xinzz, Apr 4, 2020
    #3