This can be due to a high number of active advanced settings such as exclude rules or custom extractions. These can slow down crawls, because Lumar has to scan for URLs to exclude and for data to extract. If there are multiple external URLs, this can also slow down the crawl.
Sometimes a crawl may become stuck due to technical issues. If it is stuck a 0 URLs/s for a long period then it should be raised with Support who can investigate. Please contact us at support@lumar.io.