How do I get the weights of a layer in Keras?

Retrieving the weights of a specific layer in Keras can be accomplished through a few straightforward steps. First, ensure you have a trained model. Then, use the model's get_layer() method to access the desired layer, followed by the get_weights() method to retrieve the layer's weights. Here is a specific example:

Assume you have built and trained a simple neural network model named model, and now you want to retrieve the weights of the first hidden layer in this model.

python
from keras.models import Sequential
from keras.layers import Dense

# Build the model
model = Sequential([
    Dense(32, input_shape=(10,), activation='relu'),  # First hidden layer
    Dense(1, activation='sigmoid')
])

# Compile the model
model.compile(optimizer='sgd', loss='binary_crossentropy')

# Assume you have trained the model

# Retrieve the weights of a specific layer
layer = model.get_layer(index=0)  # Or model.get_layer(name='dense')
weights = layer.get_weights()     # Returns a list where weights[0] is the weight matrix and weights[1] is the bias term

In this example, the get_layer() method specifies the target layer using either its name or index. The get_weights() method returns a list containing the weight matrix and bias term. Additionally, you can examine the weights of different layers to aid in analyzing and understanding the model's operation.