v0.3.0

It's been a long journey, but a good one. More training would have been nice, but here goes... Late demo tmr!

What's Changed

• Maximise model perf by @jacobhq in #3
• Maximise agent score above hits by @jacobhq in #4
• Migrate to next-auth v5 by @jacobhq in #5
• Implement live game into pong by @jacobhq in #6

Full Changelog: v0.2.1...v0.3.0

v0.2.1

Model training

generations: 50
approx_time: 2hr15min
target_dimensions: 700x500