Open
Description
I am currently trying to implement a Bayesian Neural Network for image classification. However, two warnings are raised:
/usr/local/lib/python3.10/dist-packages/tensorflow_probability/python/layers/util.py:95: UserWarning: `layer.add_variable` is deprecated and will be removed in a future version. Please use the `layer.add_weight()` method instead.
loc = add_variable_fn(
/usr/local/lib/python3.10/dist-packages/tensorflow_probability/python/layers/util.py:105: UserWarning: `layer.add_variable` is deprecated and will be removed in a future version. Please use the `layer.add_weight()` method instead.
untransformed_scale = add_variable_fn(
I am currently using the following code:
bayesian_model = Sequential([
tfpl.Convolution2DReparameterization(input_shape=(512, 512, 1), filters=8, kernel_size=16, activation='relu',
kernel_prior_fn=tfpl.default_multivariate_normal_fn,
kernel_posterior_fn=tfpl.default_mean_field_normal_fn(is_singular=False),
kernel_divergence_fn=divergence_fn,
bias_prior_fn=tfpl.default_multivariate_normal_fn,
bias_posterior_fn=tfpl.default_mean_field_normal_fn(is_singular=False),
bias_divergence_fn=divergence_fn),
Conv2D(kernel_size=(5, 5), filters=8, activation='relu', padding='VALID'),
MaxPooling2D(pool_size=(6, 6)),
Flatten(),
Dropout(0.2),
tfpl.DenseReparameterization(units=tfpl.OneHotCategorical.params_size(3), activation=None,
kernel_prior_fn=tfpl.default_multivariate_normal_fn,
kernel_posterior_fn=tfpl.default_mean_field_normal_fn(is_singular=False),
kernel_divergence_fn=divergence_fn,
bias_prior_fn=tfpl.default_multivariate_normal_fn,
bias_posterior_fn=tfpl.default_mean_field_normal_fn(is_singular=False),
bias_divergence_fn=divergence_fn
),
tfpl.OneHotCategorical(3)
])
bayesian_model.compile(loss=negative_log_likelihood,
optimizer=Adam(learning_rate=0.005),
metrics=['accuracy'],
experimental_run_tf_function=False)Metadata
Metadata
Assignees
Labels
No labels