こんな感じでどうでしょうか?
単純に>>1をgrepすると>>10や>>100も引っかかることに注意
#!/usr/bin/perl
use feature qw(say);
open (FH,"input.txt") or die $!;
@list=<FH>;
close FH;
@list = grep {/\>\>\d+/} @list;
for (1..1000){
$num=$_;
$num='>>'.$num;
$coun{$num} = grep {/\>\>\d+/ and ($num eq $&)} @list;
}
for (sort {$coun{$b} <=> $coun{$a}} keys %coun){
printf ("%-5s => %03d\n",$_,$coun{$_});
last if $i++>=10;
}