David J Wu

Okay, but where is the data that you are trying to train on?

Great, yes, npz files are the raw data, and assuming you're using the current training code that uses Tensorflow, the shuffled data should be tfrecord files. So where is your...

I have no idea, it depends on what file paths you're using. Are you shuffling the data at all? https://github.com/lightvector/KataGo/blob/master/python/selfplay/shuffle.sh

And: https://github.com/lightvector/KataGo/blob/master/python/selfplay/shuffle_loop.sh

If you have the raw data already, then you need to run a shuffle script that will take the raw data and shuffle it and convert it to tfrecord files...

Fascinating. How extensively have you tested this? I have to think there's a chance of some weird side effects - for example suppose that the opponent has just played a...

OpenCL is the software that tries to use your GPU that KataGo relies on. This indicates that OpenCL has an unknown problem using your GPU. Try updating your GPU drivers,...

No, I think changing the sorting is not a good idea. This is the order that Leela returns the moves, which as far as I can tell is the order...

Are you overclocking your GPU or have you noticed any hardware issues? Once before I've seen a case where someone's hardware was configured such that the GPU was overloaded, but...

What data are you training it on? If you're starting from very early in training from scratch, then it will of course be hard to achieve any high prediction accuracy...