I have this code:

# Build the PFN model
model = Sequential([   
    Dense(32, input_shape=(sequence_length, num_features)),
    tfp.layers.DenseVariational(
        units=32,
        make_posterior_fn=lambda posterior_tensor:     

tfp.layers.VariableLayer(tfp.layers.MultivariateNormalTriL.params_size(num_features))(posterior_tensor),
        make_prior_fn=lambda prior_tensor:     tfp.layers.VariableLayer(tfp.layers.MultivariateNormalTriL.params_size(num_features))(prior_tensor),
        kl_weight=1/X_train_scaled.shape[0],
        activation='relu'
    ),
    Dropout(0.1),
    tfp.layers.DenseVariational(
        units=32,
        make_posterior_fn=lambda posterior_tensor:     tfp.layers.VariableLayer(tfp.layers.MultivariateNormalTriL.params_size(num_features))(posterior_tensor),
        make_prior_fn=lambda prior_tensor:     tfp.layers.VariableLayer(tfp.layers.MultivariateNormalTriL.params_size(num_features))(prior_tensor),
        kl_weight=1/X_train_scaled.shape[0],
        activation='relu'
    ),
    Dropout(0.1),    
    tfp.layers.DenseVariational(
        units=1,
        make_posterior_fn=lambda posterior_tensor: tfp.layers.VariableLayer(tfp.layers.MultivariateNormalTriL.params_size(1))(posterior_tensor),
        make_prior_fn=lambda prior_tensor: tfp.layers.VariableLayer(tfp.layers.MultivariateNormalTriL.params_size(1))(prior_tensor),
        kl_weight=1/X_train_scaled.shape[0]
    )
])`

But I get this error:

TypeError: <lambda>() takes 1 positional argument but 3 were given

I tried adding the variables needed. Still I am getting the error.

I added the missing variables needed. Restructured the code if that was the problem, gave values to the variables in case that was why the error was coming up.