During October 2025, Nsceh1 developed PyTorch-based RNN classifier support for the sktime/sktime repository, expanding its deep learning backend capabilities. They introduced a PyTorch-specific base class for deep learning classifiers and implemented the SimpleRNNClassifierTorch, aligning its design with existing TensorFlow modules. This work involved refactoring core base classes to ensure consistency and maintainability across different deep learning backends. Using Python and PyTorch, Nsceh1 focused on time series classification and software engineering best practices, enabling users to experiment with PyTorch workflows. The changes improved backend compatibility and laid the groundwork for a more unified and scalable deep learning architecture.