find_all() has a while loop that loads 1000 more rows from find() until there is no more results to load, that is how it gets all rows and not just 1000.
The way it works is that all the rows of the data is divided into sections of 1000 rows in each, and what section to load is defined by the offset, and in the while loop it increases the offset by 1 every time that it runs the loop, and stops when there is no more sections to load.
!ALIVE
RE: ALIVE weekly tokenomics