Reduce refresh time in incremental refresh

I have setup an incremental refresh policy which archives past 24 months of data and refreshes past 1 day with an offset on 1 day. While doing the monthly partitions refreshes for the fact table which contains around 156 million rows for one month it takes over four hours. Is there anything that can be done to reduce the refresh time of the partition? Executing the same query on databricks for one months data takes less than 5 mins. Any help would be appreciated.