Stuck In First Epoch When Training CNN model in google Colab - tensorflow

Question

I created a model to identify plant diseases. I expected to identify 10 diseases. in jupyter notebook, it worked fine but it was slow due to GPU constraints. Then I decided to run that model in google colab but it did not run. it stuck at the first epoch.
The code I use to construct the model is given below
BATCH_SIZE = 64
IMAGE_SIZE = 256
CHANNELS=3
EPOCHS=10
dataset = tf.keras.preprocessing.image_dataset_from_directory(
"/content/drive/MyDrive/google-colab-files/PlantVillage",
seed=123,
shuffle=True,
image_size=(IMAGE_SIZE,IMAGE_SIZE),
batch_size=BATCH_SIZE
)
def get_dataset_partisions_tf(ds,trains_split=0.8,val_split=0.1,test_split=0.1,shuffle=True,shuffle_size=10000):
ds_size = len(ds)
if shuffle:
ds = ds.shuffle(shuffle_size,seed=12)
train_size = int(trains_split * ds_size)
val_size = int(val_split * ds_size)
train_ds = ds.take(train_size)
val_ds = ds.skip(train_size).take(val_size)
test_ds = ds.skip(train_size).skip(val_size)
return train_ds,val_ds,test_ds
train_ds,val_ds,test_ds = get_dataset_partisions_tf(dataset)
train_ds = train_ds.cache().shuffle(1000).prefetch(buffer_size = tf.data.AUTOTUNE)
val_ds = val_ds.cache().shuffle(1000).prefetch(buffer_size = tf.data.AUTOTUNE)
test_ds = test_ds.cache().shuffle(1000).prefetch(buffer_size = tf.data.AUTOTUNE)
resize_and_rescales = Sequential([
layers.experimental.preprocessing.Resizing(IMAGE_SIZE,IMAGE_SIZE),
layers.experimental.preprocessing.Rescaling(1.0/255)
])
data_agmetation = Sequential([
layers.experimental.preprocessing.RandomFlip('horizontal_and_vertical'),
layers.experimental.preprocessing.RandomRotation(0.2),
])
input_shape = (BATCH_SIZE,IMAGE_SIZE,IMAGE_SIZE,CHANNELS)
n_classes = 10
model = Sequential([
resize_and_rescales,
data_agmetation,
layers.Conv2D(32,(3,3), activation='relu',input_shape = input_shape),
layers.MaxPooling2D((2,2)),
layers.Conv2D(64,kernel_size = (3,3), activation='relu'),
layers.MaxPooling2D((2,2)),
layers.Conv2D(64,kernel_size = (3,3), activation='relu'),
layers.MaxPooling2D((2,2)),
layers.Conv2D(64,(3,3), activation='relu'),
layers.MaxPooling2D((2,2)),
layers.Conv2D(64,(3,3), activation='relu'),
layers.MaxPooling2D((2,2)),
layers.Conv2D(64,(3,3), activation='relu'),
layers.MaxPooling2D((2,2)),
layers.Flatten(),
layers.Dense(64,activation='relu'),
layers.Dense(n_classes, activation='softmax'),
])
model.build(input_shape = input_shape)
model.summary()
A screenshot of the model summary is:
model.compile(
optimizer='adam',
loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=False),
metrics=['accuracy']
)
When I use the following code to train data:
model.fit(
train_ds,
epochs=EPOCHS,
batch_size=BATCH_SIZE,
verbose=2,
validation_data=val_ds
)
it keeps stuck in the first epoch

Answer 1

Check if TensorFlow is using a GPU or not. You can try reducing batch size.

Answer 2

my assumption is that this is because of your verbose, you should set verbose to 1 to see the step of the epoch you are in.