Sparse Matrices 4: Design is my passion
Read OriginalThis fourth post in a series details the design challenges of integrating sparse matrices with automatic differentiation in JAX for PyMC models. It analyzes different implementation strategies for efficiently computing log-determinants and linear solves, specifically discussing how to avoid costly extra Cholesky decompositions when evaluating and differentiating a log-posterior function.
Comments
No comments yet
Be the first to share your thoughts!
Browser Extension
Get instant access to AllDevBlogs from your browser
Top of the Week
No top articles yet