yehu3d

> Addition:xingzhen can find this rather quickly  xingzheng English name is galaxy

maybe you have 1 gpu but you set use 2 gpus

i cant upload sgf format ,so i upload txt ,you can download it and rename it from 2.txt to 2.sfg [2.txt](https://github.com/user-attachments/files/17881982/2.txt)

i also collected winrate information from player, white got 90+winrate during this game under 20m visits,but finally they lost.

@michito744 Chinese rules komi 7.0 button=true,AI competition dont use chinese rules or japanese rules.

@michito744 Yes, starting from this position, if you continue to calculate, you can arrive at the correct answer. However, if you start calculating from an earlier position (for example, 10...

maybe features.bin_input_shape should be [22,features.pos_len,features.pos_len] features.bin_input_shape should be [19] now

File "D:\KataGo\python\play.py", line 407, in dummy_outputs = gs.get_model_outputs(model) ^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "D:\KataGo\python\katago\game\gamestate.py", line 172, in get_model_outputs qwinloss = torch.tanh(policy_logits[6,:]).cpu().numpy() ~~~~~~~~~~~~~^^^^^ IndexError: index 6 is out of bounds for dimension 0 with...

print(policy_logits.size())：torch.Size([6, 362]) maybe should:qwinloss = torch.tanh(policy_logits[5,:]).cpu().numpy() instead?