There's no generalized tool for separating a huge file with many sequences into different files. ...well, actually there is; it's Perl. But you have to think through your problem and use programming to craft a solution.

So let's get into specifics; if you were to do this with pencil and paper, how would you identify where one sequence ends and another begins? Have you started a solution yet? What can you show us?