Search Unity

Search Results

Search Again
  1. Print_Hello_World
    Thread

    Rewarding an agent when it "sees" something

    Hello :) I am trying to teach an agent to find a target but once it has "seen" the target, I want to give it a reward. I am using ray cast...
    Thread by: Print_Hello_World, Aug 20, 2020, 4 replies, in forum: ML-Agents
  2. Print_Hello_World
    Post

    Is there a way using own algorithm on training agents?

    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...
    Post by: Print_Hello_World, Apr 3, 2020 in forum: ML-Agents
  3. Print_Hello_World
    Post

    Does AgentReset() gets called when Max steps from the inspector has been reached?

    Hello, yeah I am on 0.14.1. That explains why only the deactivated agents don't get reset after they reach their max step. its because their...
    Post by: Print_Hello_World, Mar 19, 2020 in forum: ML-Agents
  4. Print_Hello_World
    Post

    Does AgentReset() gets called when Max steps from the inspector has been reached?

    ok so i managed to solve the issue but including this loop which checks at every fixedupdate() if each agent has reached its max step and if i has...
    Post by: Print_Hello_World, Mar 18, 2020 in forum: ML-Agents
  5. Print_Hello_World
    Thread

    Does AgentReset() gets called when Max steps from the inspector has been reached?

    Hello, I am a bit confused as to how are the agents and academy are being reset. I have two conditions for resetting in the game I'm making using...
    Thread by: Print_Hello_World, Mar 18, 2020, 4 replies, in forum: ML-Agents
  6. Print_Hello_World
    Post

    Using Collect Observations in Self-Play leads to issues

    Hi, I managed to fix the issue by removing the vector observations, therefore the agents only use raycasting to observe the environment. thanks!
    Post by: Print_Hello_World, Mar 18, 2020 in forum: ML-Agents
  7. Print_Hello_World
    Thread

    Using Collect Observations in Self-Play leads to issues

    Hi everyone, I am trying to make a shooter game using the Self-Play feature. I followed closely with the Soccer example except that I added a...
    Thread by: Print_Hello_World, Mar 16, 2020, 2 replies, in forum: ML-Agents
  8. Print_Hello_World
    Post

    Agent Subclass not working

    Hello, So after more and more tinkering and debugging. I managed to fix it. Not too sure what I did but the base.methods work and it is possible...
    Post by: Print_Hello_World, Feb 26, 2020 in forum: ML-Agents
  9. Print_Hello_World
    Post

    Agent Subclass not working

    So i tried debugging for a while more, it turns out that almost all methods such as collectobservations, agentaction inside my subclass has the...
    Post by: Print_Hello_World, Feb 25, 2020 in forum: ML-Agents
  10. Print_Hello_World
    Thread

    Agent Subclass not working

    Hello I am trying to use a subclass to train a class which derives from Agent. However, the subclass keeps calling AgentReset numerous times in...
    Thread by: Print_Hello_World, Feb 25, 2020, 2 replies, in forum: ML-Agents
  11. Print_Hello_World
    Post

    Can't seem to normalize mean reward

    Oh ok I had my suspicion that that was the case. So if I normalize my rewards per time step according to the best practices it should be ok...
    Post by: Print_Hello_World, Feb 20, 2020 in forum: ML-Agents
  12. Print_Hello_World
    Thread

    Can't seem to normalize mean reward

    Hi guys, I am trying to train a drone to fly towards a given direction, I have a reward function that uses AddReward(totalreward) where i made...
    Thread by: Print_Hello_World, Feb 20, 2020, 3 replies, in forum: ML-Agents
Showing results 1 to 12 of 12
Search Again