pyro-ppl/pyro

Implement Nonparametric Hamiltonian Monte Carlo

Open

#2.903 geöffnet am 15. Juli 2021

Auf GitHub ansehen
 (0 Kommentare) (0 Reaktionen) (0 zugewiesene Personen)Python (981 Forks)batch import
discussionenhancementhelp wanted

Repository-Metriken

Stars
 (8.211 Stars)
PR-Merge-Metriken
 (Durchschn. Merge 10T 19h) (1 gemergte PR in 30 T)

Beschreibung

This issue proposes adding an implementation of nonparametric Hamiltonian Monte Carlo, a modified version of HMC that provably correctly handles models with stochastic control flow, unlike vanilla HMC.

@fzaiser (one of the paper's authors) has open-sourced a PyTorch reference implementation, and turning this into a Pyro inference algorithm would be a great project for contributors with some preexisting interest in and knowledge of the line of research (on extensions of gradient-based approximate inference algorithms to discontinuous models) embodied in the paper.

Some questions for discussion:

  1. What extra information about the model is necessary to support the algorithms used in the paper's experiments?
  2. Can this information be extracted automatically, or would it require additional user-supplied annotations?
  3. Would it make more sense to implement this functionality in a new MCMC kernel or to modify our existing HMC kernels?
  4. Would NumPyro be a better substrate for this algorithm since Jax does not compile away control flow?

Contributor Guide