How GluonTs handles non-continuous time stamps

My data is hourly stock price. There is no missing values but everyday the data is only available from 10am to 4pm hourly. So there is a gap in the time stamps between 2 days (from 4pm until 10am the next day). . Plus data is only available 5 days every week (except Saturday and Sunday) instead of 7 days.

Just curious how do models in GluonTs handle this type of data? Do they consider the gap in time stamps missing values? I don’t want to just consider them hourly data starting from the ‘start’ time because I think the exact time stamp affects how DeepAR generates derived features.