David J Wu

Thanks, do you have an SGF file? It's a bit tedious to test a result from a screenshot.

Training doesn't exist for that in the sense of training a neural net or anything like that, but there are parameters you can set. KataGo will obey time controls communicated...

Hi there, nice to meet you. Sure, you can find my contact info from the commit info of commits in this repo, or find me on discord at https://discord.gg/bqkZAz3 and...

@Chenvincentkevin - If you can proof of concept your idea, by showing that it leads to a strength improvement compared to spending the same amount of compute power on doing...

Yes, the npz and sgf files correspond exactly. I will be curious to see what your results are. I would expect that the training you are proposing to do has...

> since this will be a "very special positions it hasn't seen before" I think you misunderstand what I meant. The model has definitely seen lots of draw games, and...

You might have to give more detail about what you're trying to do, and study the log files produced by your training script for what it did. If you didn't...

These lines are wrong: ``` Updated training data: G:\Projects\KataGo\Training\BaseDir\shuffleddata\current Dropping 1/1 files in: G:\Projects\KataGo\Training\BaseDir\shuffleddata\current\train as already used ``` The training code has protection against reading the same training file twice...

If things are working correctly, the outputted log lines should be more like: ``` Updated training data: G:\Projects\KataGo\Training\BaseDir\shuffleddata\2025038-... ``` showing that it is resolving shuffleddata/current into real path with the...

Just so you know, the official training already centers the mean komi around the fairest value. `komiMean` is not even used, instead `komiAuto` is used, which automatically sets the mean...