final checks before release

stripping and lowering

normalizer debug

Add all moving conditions to the leaderboard

Revert "cleaning moderate section"

cleaning moderate section

whisper normalizer

added moving source eval

selected dataset eval

new data eval

leaderboard csv fixes

new dataset update

update zerogpu fix

multiple changes including the local eval

updating overall features

updated plots, job queues etc

Refactor ASR evaluation into pluggable backends and add model family dropdown.