I'm currently developing a numerical solver to minimise the action of a classical path via gradient descent. This project is a personal exploration of how machine learning techniques can be applied to theoretical physics problems, particularly in the context of path integrals and action minimisation. The code is written in Python and uses JAX for automatic differentiation. The goal is to find the path that minimises the action scalar, which is a fundamental concept in classical mechanics. This project is ongoing and I plan to share more details as it progresses.