Modification of V13

Notes.txt

@@ -8,7 +8,7 @@ V9: 3 layer, doppelte depth Conv pro layer, norm. Daten,kernel 5, phase only
 V10: 3 layer, eine Conv pro layer, norm Daten, phase only
 V11: 3 layer, doppel Conv, normDaten, phase + angle
 V12: 3 layer, doppel Conv, normDataen,phase 64 
-V13: 4 layer, doppel Conv, normDataen,phase 64
+V13: 4 layer, doppel Conv, normDataen,angles 64
 V14: 4 layer, single conv, normDataen,phase + angle 64
 V15: 3 layer, doppelte depth Conv pro layer, norm. Daten,kernel 7, phase only, dropout 0.3, 32
 V16: 3 layer, doppelte depth Conv pro layer, norm. Daten,kernel 7, angelsonly, dropout 0.5, 32

UNet/UNet_V13.py

@@ -134,7 +134,7 @@ def accuracy(outputs, labels,normalization, threshold = 0.05):
     return percentage
     
 class UNet(UNetBase):
-    def __init__(self,kernel_size = 5, enc_chs=((6,6,16), (16,16,32), (32,32,64), (64,128,128)), dec_chs_up=(128, 128, 64, 32), dec_chs_conv=((192,128,128),(160,64,64),(80,32,32),(38,16,1)),normalization=np.array([0,1])):
+    def __init__(self,kernel_size = 7, enc_chs=((6,16,32), (32,32,64), (64,64,128), (128,128,256)), dec_chs_up=(256,256, 128, 64), dec_chs_conv=((384,256,256),(320,128,128),(160,64,64),(70,32,32)),normalization=np.array([0,1])):
         super().__init__()
         self.encoder     = Encoder(kernel_size = kernel_size, chs = enc_chs)
         self.decoder     = Decoder(kernel_size = kernel_size, chs_upsampling = dec_chs_up, chs_conv = dec_chs_conv)
@@ -225,8 +225,8 @@ if __name__ == '__main__':
     path_to_rep = '/home/yk138599/Hiwi/damask3'
     use_seeds = False
     seed = 373686838
-    num_epochs = 300
-    b_size = 10
+    num_epochs = 200
+    b_size = 16
     opt_func = torch.optim.Adam
     lr = 0.00003
     kernel = 7