leela-chess
leela-chess copied to clipboard
glinscott
**MOVED TO https://github.com/LeelaChessZero/leela-chess ** A chess adaption of GCP's Leela Zero
This patch can be used to disable specific lc0 backends using the meson command line. The original intention was to disable the TensorFlow backend, which can be done by specifying...
This idea is based on a test Aloril made, where he trained 5 new nets on the same games, but in random feed order. Then he tested the nets in...
I know there has been talk of using syzygy, but the context of it was never clear, so I apologize if this is a rehash. First I want to be...
I had a problem to run client.exe on my laptop, the black window closed immediately after downloading. Now I downloaded the latest network and weights.txt is in the folder. Client...
Currently the training code sets batch normalization center parameter to False which disables the learnable beta bias parameters. This forces every batch norm output plane to have zero mean which...
I am trying to run Alexander's lc0 cudnn on 3 GPUs (1 Titan V + 2 x 1080). I use the command lc0.exe -w weights.txt --no-smart-pruning --backend=multiplexing "--backend-opts=x(backend=cudnn,gpu=0,max_batch=512),y(backend=cudnn,gpu=1,max_batch=256),z(backend=cudnn,gpu=2,max_batch=256)" --threads=4 and...
slightly more than 2x speedup (for large batch sizes) on supported hardware, without much loss of precision.
Observing the Discord channel, I have noticed that the most common question is:"How long for the next network to be out/tested?" And I must admit that it is a question...
