Fast experiment on T4 GPU. Self play training on Dark Hex (Colab notebook) [P]

Last week I run a fun experiment on Dark Hex. Here's a visualization of two iterations (1800 vs 1900) of agent playing agains each other 😃 Here's my colab notebook if you like to run it yourself https://colab.research.google.com/drive/1-rm_Bh8CNaM861We97ZoicfgKxz0xOSi?usp=sharing