go

puzzle.py

@@ -8,8 +8,10 @@ from random import *
 from agent import *
 import numpy as np
 import pickle
+import time
 
-TRAIN = 5000
+
+TRAIN = 20000
 SIZE = 500
 GRID_LEN = 4
 GRID_PADDING = 10
@@ -120,9 +122,9 @@ class GameGrid(Frame):
         if (game_state(self.matrix)=='win' ) or  (game_state(self.matrix)=='lose'):
             # print(self.agent.W)
             if (self.count == TRAIN):
-                # f = open("train_result_after_"+str(TRAIN)+".txt",'wb')
-                # pickle.dump(self.agent.W ,f)
-                # f.close()
+                f = open("train_result_after_"+str(TRAIN)+".txt",'wb')
+                pickle.dump(self.agent.W ,f)
+                f.close()
                 f = open("train_history_after_"+str(TRAIN)+".txt",'wb')
                 np.savetxt(f, self.history)
                 f.close()
@@ -146,4 +148,6 @@ class GameGrid(Frame):
             index = (self.gen(), self.gen())
         self.matrix[index[0]][index[1]] = 2
 
+start_time = time.time()
 gamegrid = GameGrid()
+print("--- %s seconds ---" % (time.time() - start_time))

train_history_after_1.txt

+1.280000000000000000e+02