Hey @junpenglao it worked!
with basic_model:
# draw 500 posterior samples
trace = pm.sample(500, cores=1)
OUTPUT:
Auto-assigning NUTS sampler... Initializing NUTS using jitter+adapt_diag... Sequential sampling (2 chains in 1 job) NUTS: [sigma, beta, alpha] 100%|██████████████████████████████████████████████████████████████████████████████| 1000/1000 [05:57<00:00, 4.05it/s] 100%|██████████████████████████████████████████████████████████████████████████████| 1000/1000 [05:27<00:00, 3.05it/s]
If I understand this correctly, it means that the code isn’t fully utilizing the processor. Any idea how I can remedy this?
Either way thanks a lot. 