Dopamineisaresearchframeworkforfastprototypingofreinforcementlearningalgorithms.