hongshaorou/Tianshou
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

debug: we should estimate our own win rate

Browse Source
This commit is contained in:
rtz19970824 2018-01-08 16:19:59 +08:00
parent 8b7b4b6c6b
commit 32b7b33ed5
1 changed files with 1 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
AlphaGo/model.py
View File

@ -284,7 +284,7 @@ class ResNet(object):
history.append(board) history.append(board)
states.append(self._history2state(history, color)) states.append(self._history2state(history, color))
probs.append(np.array(prob).reshape(1, self.board_size ** 2 + 1)) probs.append(np.array(prob).reshape(1, self.board_size ** 2 + 1))
winner.append(np.array(data.winner).reshape(1, 1)) winner.append(np.array(data.winner * color).reshape(1, 1))
color *= -1 color *= -1
states = np.concatenate(states, axis=0) states = np.concatenate(states, axis=0)
probs = np.concatenate(probs, axis=0) probs = np.concatenate(probs, axis=0)