David J Wu

Full docs for what fields you can expect in the `result` dict, as well as what other fields you can adjust on the query, are in https://github.com/lightvector/KataGo/blob/master/docs/Analysis_Engine.md

Note that the example code at https://github.com/lightvector/KataGo/blob/master/python/query_analysis_engine_example.py is probably not the most robust in terms of handling warnings or errors returned by KataGo. Warnings and errors should be nonexistent for...

Thanks for adjusting the title to clarify your question - you can also post some text in the main message too, not just the title. :) You don't need "blank...

What's the computational cost (if any) of SE versus non-SE, holding number of blocks constant?

I would probably just use the newer one if you're getting still hundreds of visits per second. You can always just run it for the same amount of compute time,...

@hwj-111 - not *entirely* true. :) In arbitrary midgame positions, perhaps it is hard to tell the difference, but for example 15b predates a lot of the training on the...

I'm not sure what you mean by "analysis strength", but the algorithm that runs for analysis is essentially the same algorithm that runs for playing.

@michito744 Thanks for the examples! Can you attach SGFs as well for all the examples you gave? It's awkward to transcribe a screenshot, and often it's useful to also explore...

There's still probably little reason to do anything other than to pick the model that is best given equal compute power, especially if the weaker and older models have more...

Thanks for testing! If it works with an older CUDA you can still use the older CUDA, but I've switched my testing to be on CUDA 12 going forward, so...