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

add random preprocess, modify the uniform sample from training data

Browse Source
This commit is contained in:
rtz19970824 2017-12-23 16:27:09 +08:00
parent c50ee8f029
commit a787f73cf6
1 changed files with 65 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

72
AlphaGo/model.py
View File

@ -1,7 +1,6 @@
import os import os
import time import time
import random import copy
import sys
import cPickle import cPickle
from collections import deque from collections import deque
@ -224,11 +223,21 @@ class ResNet(object):
else: else:
start_time = time.time() start_time = time.time()
for i in range(batch_size): for i in range(batch_size):
game_num = random.randint(0, self.window_length-1) priority = self.training_data['length'] / sum(self.training_data['length'])
state_num = random.randint(0, self.training_data['length'][game_num]-1) game_num = np.random.choice(self.window_length, 1, p=priority)
training_data['states'].append(np.expand_dims(self.training_data['states'][game_num][state_num], 0)) state_num = np.random.randint(self.training_data['length'][game_num])
training_data['probs'].append(np.expand_dims(self.training_data['probs'][game_num][state_num], 0)) rotate_times = np.random.randint(4)
training_data['winner'].append(np.expand_dims(self.training_data['winner'][game_num][state_num], 0)) reflect_times = np.random.randint(2)
reflect_orientation = np.random.randint(2)
training_data['states'].append(
self._preprocession(self.training_data['states'][game_num][state_num], reflect_times,
reflect_orientation, rotate_times))
training_data['probs'].append(
self._preprocession(self.training_data['probs'][game_num][state_num], reflect_times,
reflect_orientation, rotate_times))
training_data['winner'].append(
self._preprocession(self.training_data['winner'][game_num][state_num], reflect_times,
reflect_orientation, rotate_times))
value_loss, policy_loss, reg, _ = self.sess.run( value_loss, policy_loss, reg, _ = self.sess.run(
[self.value_loss, self.policy_loss, self.reg, self.train_op], [self.value_loss, self.policy_loss, self.reg, self.train_op],
feed_dict={self.x: np.concatenate(training_data['states'], axis=0), feed_dict={self.x: np.concatenate(training_data['states'], axis=0),
@ -280,6 +289,55 @@ class ResNet(object):
winner = np.concatenate(winner, axis=0) winner = np.concatenate(winner, axis=0)
return states, probs, winner return states, probs, winner
def _preprocession(self, board, reflect_times=0, reflect_orientation=0, rotate_times=0):
"""
preprocessing for augmentation
:param board: a ndarray, board to process
:param reflect_times: an integer, how many times to reflect
:param reflect_orientation: an integer, which orientation to reflect
:param rotate_times: an integer, how many times to rotate
:return:
"""
new_board = copy.copy(board)
if new_board.ndim == 3:
np.expand_dims(new_board, axis=0)
new_board = self._board_reflection(new_board, reflect_times, reflect_orientation)
new_board = self._board_rotation(new_board, rotate_times)
return new_board
def _board_rotation(self, board, times):
"""
rotate the board for augmentation
note that board's shape should be [batch_size, board_size, board_size, channels]
:param board: a ndarray, shape [batch_size, board_size, board_size, channels]
:param times: an integer, how many times to rotate
:return:
"""
return np.rot90(board, times, (1, 2))
def _board_reflection(self, board, times, orientation):
"""
reflect the board for augmentation
note that board's shape should be [batch_size, board_size, board_size, channels]
:param board: a ndarray, shape [batch_size, board_size, board_size, channels]
:param times: an integer, how many times to reflect
:param orientation: an integer, which orientation to reflect
:return:
"""
new_board = copy.copy(board)
for _ in range(times):
if orientation == 0:
new_board = new_board[:, ::-1]
if orientation == 1:
new_board = new_board[:, :, ::-1]
return new_board
if __name__ == "__main__": if __name__ == "__main__":
model = ResNet(board_size=9, action_num=82, history_length=8) model = ResNet(board_size=9, action_num=82, history_length=8)