When using brms without invoking within-chain parallelization, I would see something like the following for each chain:
Chain #: Gradient evaluation took 0.025388 seconds
Chain #: 1000 transitions using 10 leapfrog steps per transition would take 253.88 seconds.
Chain #: Adjust your expectations accordingly!
I understand that the timing information cannot be interpreted in an absolute sense, but it does provide some benchmark when running and comparing different models or datasets.
However, when running brms with within-chain parallelization, the timing information is absent. Any reason such information is not available?