And predict the resulting loss of accuracy of each epoch, the drawing can be seen that the convergence of the training results and the visual pleasure.
Coding #:. 8 UTF- Import matplotlib.pyplot AS PLT data_dir = "D: \\ the result.txt" Train_Loss_list = [] Train_Accuracy_list = [] Valid_Loss_list = [] Valid_Accuracy_list = [] F1 = Open (data_dir, 'R & lt') Data = [] # outputs the result to the training result.txt, the stupid way, pick up the digital results byte bits for F1 in Line: IF (line.find ( 'train')> = 0): # IF (to string.find has (Line, 'Train') = -1!): Train_Loss_list.append (Line [12:18]) Train_Accuracy_list.append (Line [24:30]) IF (line.find ( 'Valid')> 0 =): # IF (to string.find has (Line, 'Valid') = -1):! Valid_Loss_list.append (Line [12:18]) Valid_Accuracy_list.append(line[24:30]) f1.close () # iterated 30 times, the range of x (0,30), and then each time and the corresponding loss of accuracy on the x attached X1 = Range (0, 30) X2 Range = (0, 30) Y1 = Train_Accuracy_list Y2 = Train_Loss_list Y3 = Valid_Accuracy_list Y4 = Valid_Loss_list plt.subplot (2,. 1,. 1) # plt.plot (X1, Y1, 'O -', Color = 'R & lt') PLT .plot (X1, Y1, 'O -', label = "Train_Accuracy") plt.plot (X1, Y3, 'O -', label = "Valid_Accuracy") plt.title ( 'the Test Accuracy FC vs. Epoches') PLT .ylabel ( 'the Test Accuracy') plt.legend (LOC = 'Best') plt.subplot (2,. 1, 2) plt.plot (X2, Y2, '.-', label = "Train_Loss") plt.plot (X2, Y4, '.-', label = "Valid_Loss") PLT.xlabel('Test loss vs. epoches') plt.ylabel('Test loss') plt.legend (loc = 'best') plt.show ()