Concentration inequality of maximum likelihood estimator

For a sequence of independent and identically distributed random variables (r.v.) valued in some metric space (E,d)(E,d), we obtain a sharp concentration inequality of the maximum likelihood estimator under some standard concavity condition.