Skip to content
Stacy edited this page Jan 23, 2023 · 3 revisions

Epoch 1/10 29/29 [==============================] - 302s 10s/step - loss: 0.0084 - val_loss: 0.0085 Epoch 2/10 29/29 [==============================] - 302s 10s/step - loss: 0.0083 - val_loss: 0.0075 Epoch 3/10 29/29 [==============================] - 290s 10s/step - loss: 0.0070 - val_loss: 0.0064 Epoch 4/10 29/29 [==============================] - 287s 10s/step - loss: 0.0058 - val_loss: 0.0058 Epoch 5/10 29/29 [==============================] - 288s 10s/step - loss: 0.0053 - val_loss: 0.0054 Epoch 6/10 29/29 [==============================] - 294s 10s/step - loss: 0.0050 - val_loss: 0.0051 Epoch 7/10 29/29 [==============================] - 290s 10s/step - loss: 0.0049 - val_loss: 0.0051 Epoch 8/10 29/29 [==============================] - 287s 10s/step - loss: 0.0047 - val_loss: 0.0047 Epoch 9/10 29/29 [==============================] - 286s 10s/step - loss: 0.0046 - val_loss: 0.0046 Epoch 10/10 29/29 [==============================] - 297s 10s/step - loss: 0.0046 - val_loss: 0.0047 (152, 200, 3) Model: "sequential"


Layer (type) Output Shape Param #

batch_normalization (BatchN (None, 152, 200, 3) 12 ormalization)

Conv1 (Conv2D) (None, 150, 198, 8) 224

Conv2 (Conv2D) (None, 148, 196, 16) 1168

max_pooling2d (MaxPooling2D (None, 74, 98, 16) 0 )

Conv3 (Conv2D) (None, 72, 96, 16) 2320

dropout (Dropout) (None, 72, 96, 16) 0

Conv4 (Conv2D) (None, 70, 94, 32) 4640

dropout_1 (Dropout) (None, 70, 94, 32) 0

Conv5 (Conv2D) (None, 68, 92, 32) 9248

dropout_2 (Dropout) (None, 68, 92, 32) 0

max_pooling2d_1 (MaxPooling (None, 34, 46, 32) 0 2D)

Conv6 (Conv2D) (None, 32, 44, 64) 18496

dropout_3 (Dropout) (None, 32, 44, 64) 0

Conv7 (Conv2D) (None, 30, 42, 64) 36928

dropout_4 (Dropout) (None, 30, 42, 64) 0

max_pooling2d_2 (MaxPooling (None, 15, 21, 64) 0 2D)

up_sampling2d (UpSampling2D (None, 30, 42, 64) 0 )

Deconv1 (Conv2DTranspose) (None, 32, 44, 64) 36928

dropout_5 (Dropout) (None, 32, 44, 64) 0

Deconv2 (Conv2DTranspose) (None, 34, 46, 64) 36928

dropout_6 (Dropout) (None, 34, 46, 64) 0

up_sampling2d_1 (UpSampling (None, 68, 92, 64) 0 2D)

Deconv3 (Conv2DTranspose) (None, 70, 94, 32) 18464

dropout_7 (Dropout) (None, 70, 94, 32) 0

Deconv4 (Conv2DTranspose) (None, 72, 96, 32) 9248

dropout_8 (Dropout) (None, 72, 96, 32) 0

Deconv5 (Conv2DTranspose) (None, 74, 98, 16) 4624

dropout_9 (Dropout) (None, 74, 98, 16) 0

up_sampling2d_2 (UpSampling (None, 148, 196, 16) 0 2D)

Deconv6 (Conv2DTranspose) (None, 150, 198, 16) 2320

Final (Conv2DTranspose) (None, 152, 200, 1) 145

================================================================= Total params: 181,693 Trainable params: 0 Non-trainable params: 181,693

With GPU

Epoch 1/10 120/119 [==============================] - 67s 560ms/step - loss: 0.0078 - val_loss: 0.0058 Epoch 2/10 120/119 [==============================] - 62s 520ms/step - loss: 0.0053 - val_loss: 0.0048 Epoch 3/10 120/119 [==============================] - 83s 695ms/step - loss: 0.0048 - val_loss: 0.0045 Epoch 4/10 120/119 [==============================] - 96s 803ms/step - loss: 0.0046 - val_loss: 0.0045 Epoch 5/10 120/119 [==============================] - 94s 787ms/step - loss: 0.0044 - val_loss: 0.0044 Epoch 6/10 120/119 [==============================] - 91s 759ms/step - loss: 0.0043 - val_loss: 0.0042 Epoch 7/10 120/119 [==============================] - 100s 836ms/step - loss: 0.0042 - val_loss: 0.0042 Epoch 8/10 120/119 [==============================] - 100s 834ms/step - loss: 0.0041 - val_loss: 0.0040 Epoch 9/10 120/119 [==============================] - 106s 880ms/step - loss: 0.0041 - val_loss: 0.0041 Epoch 10/10 120/119 [==============================] - 124s 1s/step - loss: 0.0040 - val_loss: 0.0039 Model: "sequential"


Layer (type) Output Shape Param #

batch_normalization (BatchNo (None, 152, 200, 3) 12


Conv1 (Conv2D) (None, 150, 198, 8) 224


Conv2 (Conv2D) (None, 148, 196, 16) 1168


max_pooling2d (MaxPooling2D) (None, 74, 98, 16) 0


Conv3 (Conv2D) (None, 72, 96, 16) 2320


dropout (Dropout) (None, 72, 96, 16) 0


Conv4 (Conv2D) (None, 70, 94, 32) 4640


dropout_1 (Dropout) (None, 70, 94, 32) 0


Conv5 (Conv2D) (None, 68, 92, 32) 9248


dropout_2 (Dropout) (None, 68, 92, 32) 0


max_pooling2d_1 (MaxPooling2 (None, 34, 46, 32) 0


Conv6 (Conv2D) (None, 32, 44, 64) 18496


dropout_3 (Dropout) (None, 32, 44, 64) 0


Conv7 (Conv2D) (None, 30, 42, 64) 36928


dropout_4 (Dropout) (None, 30, 42, 64) 0


max_pooling2d_2 (MaxPooling2 (None, 15, 21, 64) 0


up_sampling2d (UpSampling2D) (None, 30, 42, 64) 0


Deconv1 (Conv2DTranspose) (None, 32, 44, 64) 36928


dropout_5 (Dropout) (None, 32, 44, 64) 0


Deconv2 (Conv2DTranspose) (None, 34, 46, 64) 36928


dropout_6 (Dropout) (None, 34, 46, 64) 0


up_sampling2d_1 (UpSampling2 (None, 68, 92, 64) 0


Deconv3 (Conv2DTranspose) (None, 70, 94, 32) 18464


dropout_7 (Dropout) (None, 70, 94, 32) 0


Deconv4 (Conv2DTranspose) (None, 72, 96, 32) 9248


dropout_8 (Dropout) (None, 72, 96, 32) 0


Deconv5 (Conv2DTranspose) (None, 74, 98, 16) 4624


dropout_9 (Dropout) (None, 74, 98, 16) 0


up_sampling2d_2 (UpSampling2 (None, 148, 196, 16) 0


Deconv6 (Conv2DTranspose) (None, 150, 198, 16) 2320


Final (Conv2DTranspose) (None, 152, 200, 1) 145

Total params: 181,693 Trainable params: 0 Non-trainable params: 181,693