Google Dopamine (LIVE)
Siraj Raval:
Google recently released a research framework for fast prototyping of reinforcement learning algorithms called "Dopamine". They say that it aims to fill the need for a small, easily grokked codebase in which users can freely experiment with wild ideas (speculative research). In this live stream i'll test it out, try out a few basic reinforcement learning algorithms and compare it to similar frameworks. You can code along with me using CoLab. Get hype!
Code for this video:
https://github.com/llSourcell/Google_Dopamine_LIVE
Google recently released a research framework for fast prototyping of reinforcement learning algorithms called "Dopamine". They say that it aims to fill the need for a small, easily grokked codebase in which users can freely experiment with wild ideas (speculative research). In this live stream i'll test it out, try out a few basic reinforcement learning algorithms and compare it to similar frameworks. You can code along with me using CoLab. Get hype!
Code for this video:
https://github.com/llSourcell/Google_Dopamine_LIVE
Click to view