using sed, this solution is not consider the white spaces in header.

$ sed '/^>/s/.\{5\}$//' in.fa

for fasta and fastq file, bioawk https://github.com/lh3/bioawk is also good option, it can separate the $name and $comment in header.

$ bioawk -cfastx '{id=substr($name, 0, length($name) - 5); print ">"id"\n"$seq}'

What have you tried? This sort of problem has been addressed on the site multiple times

sed can match the first character of each line to pick lines where an operation is performed - you can use that to restrict the operation to just header lines. You can also capture the last five characters with the regex (.{5})$.

Please use these hints to get to the solution.