Numpyro is fantastic, great syntax and great contributors as well. PyMC now supports jax and numba as well, and can use the numpyro samplers. https://www.pymc.

Numpyro is fantastic, great syntax and great contributors as well. PyMC now supports jax and numba as well, and can use the numpyro samplers. https://www.pymc.io/projects/examples/en/latest/samplers/fast_sampling_with_jax_and_numba.html https://www.pymc.io/projects/examples/en/latest/samplers/fas... Between all these tools we hope the modern bayesian can find the one that works best for them