This issue is related to the latest HTML report feature. Currently, some of the NBI strings are represented as bytes string literals with a ‘b’ prefix in the HTML report (see screenshot). One solution could be to use chardet to detect encoding and then use the decoded strings in the report.