Hey!
I recently have struggles with models I run in a conda environment in which I installed pymc v4 from source. The sampler is unbearably slow. Which is not the case, when I install a pymc release via conda.
Is there an up-to-date guide on installing pymc from source (Intel Linux) somewhere?
How I did it:
install (with python 3.10) env from pymc/environment-dev.yml at main · pymc-devs/pymc · GitHub
and then install pymc with pip install -e
I also installed the requirements-dev.txt packages.
Sampling is not only very slow, but the
“Auto-assigning NUTS sampler…
Initializing NUTS using jitter+adapt_diag…
Multiprocess sampling (4 chains in 4 jobs)
NUTS: [i_t, i_s, ms_mz, ms_s, alpha, me]”
doesn’t show up at all.
However, I am not getting any aesara related warnings e.g. BLAS warnings …
Thank you!