This repository contains codes for implementing a deep temporal neural network model, based on dilated CNN and bidirectional LSTM. This modular model was developed initially for micro-EDM process monitoring but can be tailored for other process monitoring purposes. Techniques of integrated Grad-CAM and t-SNE are provided to interpret the model predictions. Data for training and testing is only available on request.