Klazkin / player-zero

1 stars 0 forks source link

OpenAI model architecture research #61

Closed Klazkin closed 8 months ago

Klazkin commented 8 months ago

Reading and researching the paper published by OpenAI "Dota 2 with Large Scale Deep Reinforcement Learning". Specifically, I am interested in how they encoded various aspects related to gameplay such as positions of units (highly relevant in our case).

Time tracking

Time Estimate: undefined Time spent: 1 hours 45 minutes

Resources

OpenAI Five https://cdn.openai.com/dota-2.pdf