My fault, I took the requirement literally even if it was clearly a waste of CPU time
Thank you

lol, after getting rejected many times for lack of speed I came up with this monstruosity
It's faster than most,but don't think it's worth the extra complexity. Also I think it can be made even faster and simpler with ST

Love this one from donaldsebleung, taught me lots about maps: https://www.codewars.com/kata/reviews/5a086e50d87df7e3ca000083/groups/5a7ecbb7dc8ae467d1000647

I guess you run out of digits. Floating point breaks pretty soon with factorial

My haskell solution works for the test, but always fails submission.

Thought of a performance issue, but doubt it: on my (beefy) laptop does 1000 in about 4 ms, 10000 in about 340 (10000! has 35660 digits)

Wonder if the property based testing is longer than the platform allows

It's a math precision issue. This is about integer arithmetics, floating point doesn't work.

2^31 * 17^2 * 969743 * 25772783 = 15511210043330986055303168
25! = 15511210043330985984000000 = 2^22 * 3^10 * 5^6 * 7^3 * 11^2 * 13 * 17 * 19 * 23