the minhash generation process is very slow

I test generate_minhashes.py with a 64 kernels server. It is very slow. Even for a 19G datasets, it requires 1 day for computing. Is it normal?

Thanks!

The report identifies generate_minhashes.py and a 19G dataset processed on a 64-kernel server. Start by reproducing or measuring the reported one-day runtime, then determine whether it is expected and identify what evidence is needed to define a fix; done means clarifying the expected performance or addressing a confirmed bottleneck.

Written by the indexing model from the issue text.