Unfortunately, you can't recover a random password that is that long within a reasonable amount of time.
Just do the math:
?u?l = 26+26 = 52 characters
?d = 10 characters
=> 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 10 = (52 ^ 15) * 10 =
549604341280186671227207680 password candidates
There would be no way to recover a password with a complexity like this for NTML/MD5 (these are among the fastest hashing algorithms that hashcat supports), let alone for a hashing algorithm as slow as ethereum (the scrypt variant or pbkdf2? It doesn't matter).
Unfortunately, either you have stored the password somewhere or remember at least parts of it, or you should forget about recovering this random extremely long password.
Update: I just noticed that the examples that you listed above do not correspond to the mask that you have specified (my math example was just about the mask you have posted above)
It seems that the character set is actually -2 ?l?u?d ?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2
This means, that the keyspace is even worse i.e. (26+26+10) ^ 16 = 47672401706823533450263330816
I think even the fact that you used "pronouncable passwords" (-a 0) doesn't help much (except if there is some really bad weakness of the random output produced by apg -a 0 or the seed that was used to feed the random number generator). Anyway, length 16 will still be infeasible for a slow hashing algorithm like this (ethereum).
Just do the math:
?u?l = 26+26 = 52 characters
?d = 10 characters
=> 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 52 * 10 = (52 ^ 15) * 10 =
549604341280186671227207680 password candidates
There would be no way to recover a password with a complexity like this for NTML/MD5 (these are among the fastest hashing algorithms that hashcat supports), let alone for a hashing algorithm as slow as ethereum (the scrypt variant or pbkdf2? It doesn't matter).
Unfortunately, either you have stored the password somewhere or remember at least parts of it, or you should forget about recovering this random extremely long password.
Update: I just noticed that the examples that you listed above do not correspond to the mask that you have specified (my math example was just about the mask you have posted above)
It seems that the character set is actually -2 ?l?u?d ?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2?2
This means, that the keyspace is even worse i.e. (26+26+10) ^ 16 = 47672401706823533450263330816
I think even the fact that you used "pronouncable passwords" (-a 0) doesn't help much (except if there is some really bad weakness of the random output produced by apg -a 0 or the seed that was used to feed the random number generator). Anyway, length 16 will still be infeasible for a slow hashing algorithm like this (ethereum).