ParentFull threadbtobolaski·mpt-7b and the falcon models have longer contexts available. They aren’t trained on longer contexts so, the results might not be very good.View on HN