Kevin Ferguson

@constant5 The generator version creates large temporary files on disk, so I suspect that's why it won't work with colab (just guessing though). As for the overfitting, more games is...

Hi Yamil, I have some examples in another repository that might help you. https://github.com/macfergus/badukai Here are some key files: https://github.com/macfergus/badukai/blob/master/build_index.py 👈 builds an index of your SGF collection https://github.com/macfergus/badukai/blob/master/train_from_sgf.py 👈...

Hey @computer-idol, I actually put together a C++ implementation of the `GoBoard` / `GameState` library (with Python wrapper). It's a LOT faster than the pure Python version in the book....

also, on the subject of parallel MCTS, it turns out to be fairly complicated, see this paper for an overview of the some of the options and tradeoffs: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.159.4373&rank=1 That...

Hi aletote, can you provide a full stack trace?

Aha, I believe this bug is fixed in the `chapter_14` branch. Can you check out that branch and try again? Sorry for the confusion-- we need to clean out the...

Hi @nutpen85, I've run into the same problem and eventually found a (partial) solution. It didn't make it into the book because I only learned about it recently. Your diagnosis...

I guess there are two separate problems here. One, passing is a perfectly valid move, so the demo server should allow the bot to pass, and the client should handle...

It's hard to tell precisely. The chunking definitely adds some overhead to encoding the game records. I would guess it's about 25% slower but I did not take any careful...

This is a good question. Currently the GoBoard class in goboard.py tracks only the current state of the board. Nothing tracks history, komi, captures, etc, not even whose turn it...