Unity3D+ML-Agents Custom Algorithms and Environments

[@indrajitpal63]

This was so informative ! Thanks a ton for making this 🤩. Just wish I came across this a few months earlier so that I wouldn't have to struggle to figure stuff out on my own - would have saved a ton of time & energy. Thanks once again ! 🎉

[@damnsec]

Thank you for taking the time to create & share these videos, they are very useful!

[@mykewl5406]

I am so grateful that this video exists! Thank you!

[@MyungjinPark-y5f]

this is really helpful for me. thx a lot.

[@ShortVine]

thank you so much for 2 days i am trying to find this solution, glad i found your vid

[@guesswho1705]

Awesome explanation, wonderful video, thank you

[@AlMgAgape]

can i using Deep Q Network for training the agents? example for create a agent AI Car for racing game

[@momomo2298]

thank you so much! helps a lot!

[@omarsinno2774]

Thank you!

[@juleswombat5309]

That kinda useful and informative. The ML-Agents examples tend to use Multiple parallel Environments for faster Training. Will this Build and execute process against a Unity project Environment which has multiple copies of the Environment/Agent instances For faster Training ? I presume that it would recognise that ?

[@rachelstclair9897]

Hello Jules, yes you can absolutely use multiple training areas in the same build execution. I didn't show that here to keep it simple.

[@АлексеевА-л8н]

Thank you very much! You're cool! ))

[@893uma3]

Is it possible to use ML-Agents to create strategic AI that targets specific individuals, such as detecting the habits of players and exploiting their weaknesses in 3D action games?

[@rachelstclair9897]

Hi, I'm sure this is possible if you put your mind to you. You might want to look more at using tags on agents and objects. You can also look into side-channels in mlagents-envs if you need some more advanced methods. Good luck!

[@ShortVine]

one request can you please make vid using any RL algo because i am trying myself but there are so many values which i am not able to find in this env like env.action_space.high[0], low, reward_range etc

[@rachelstclair9897]

Am working on some videos for this kind of stuff, will post them soon. Thanks for watching!

[@ShortVine]

@@rachelstclair9897 ok nice am waiting.

[@rachelstclair9897]

kzbin.info/www/bejne/mKmcgGuLeddpnac

[@zezooalzezoo5200]

Thank , can i use it for enemies ai ?

[@rachelstclair9897]

Absolutely! If you get it working, make sure to drop a link so we can see the results!

[@zezooalzezoo5200]

@@rachelstclair9897 I will of course, still under processing, need to learn more about AI behavior & logic