Fix Minibatch Alignment in Bayesian Neural Network Example

• Notebook follows style guidehttps://docs.pymc.io/en/latest/contributing/jupyter_style.html
• PR description contains a link to the relevant issue:
  • Fixes issueBug in "Variational Inference: Bayesian Neural Networks" notebook #654
• Check the notebook is not excluded from any pre-commit check:https://github.com/pymc-devs/pymc-examples/blob/main/.pre-commit-config.yaml

Issue

Fixes#654.

Description

This PR addresses the issue whereMinibatch variables forX_train andY_train were created separately, causing potential misalignment in slices and thus random pairing ofX with unrelatedy.

Changes Made

• Updated the code to create a singleMinibatch forX_train andY_train together, ensuring identical slices.
• Modified the construction ofann_input andann_output using the unifiedMinibatch variables.
• Removed redundant parameters from theconstruct_nn function.
• Confirmed model architecture and initialization remain consistent.

Impact

This change ensures that the model training process is more robust and prevents issues related to incorrect data alignment in minibatches.

Helpful links

• https://github.com/pymc-devs/pymc-examples/blob/main/CONTRIBUTING.md

📚 Documentation preview 📚:https://pymc-examples--689.org.readthedocs.build/en/689/