Hi,
I trained a neural network for 9x9, and it is playing on CGOS.
The network has 40 layers (20 residual blocks) of 256 units. It is running
on a Titan V GPU, with a batch of 64, at about 9k playouts per second.
It is using an opening book that you can browse online there:
https://www.crazy-sen
great book interface, by the way.
s.
On Wed, May 6, 2020 at 11:01 AM Rémi Coulom wrote:
> Hi,
>
> I trained a neural network for 9x9, and it is playing on CGOS.
>
> The network has 40 layers (20 residual blocks) of 256 units. It is running
> on a Titan V GPU, with a batch of 64, at about 9k pl
Agreed on the book interface!
If you click through to the end of a variation, you can see that the
evaluation is a minimax'd evaluation, instead of an average over the
subtree nodes, even if the overall tree was generated via a UCT algorithm.
The values seem a bit more fragile to me as a result...
