Modeling human response time data with an exponential model: model comparison issues

Hi all,

ArviZ is not saving the log_likelihood when more than one observed variable is present. We will try to fix this ASAP.