pythontensorflowkerastf.keras

How to get the layer details in tensorflow?

I have a layer which is defined as following

import tensorflow.keras.layers as layers
avg_pool = layers.AveragePooling1D(pool_size=2)

Is there a way to print the details of this layer? (in this case it would be printing the pool_size)? I cannot use summary on this layer as it gives me AttributeError: 'AveragePooling1D' object has no attribute 'summary'

Solution

  • The .summary is available for model class and not layer. That's why you can't do avg_pool.summary(). But you can wrap it into model class.

    avg_pool = layers.AveragePooling1D(pool_size=2)
avg_pool.pool_size
(2,)

# init model 
model = keras.Sequential([avg_pool])

# run on some dummy tensor to build the model
x = np.array([1., 2., 3., 4., 5.])
x = np.reshape(x, [1, 5, 1])
_ = model(x)

# print summary
model.summary() # OK