model param fetch, text edit

granite 512 tokens again

changing max tokens for granite

granite debug

Revert "cleaning moderate section"

cleaning moderate section

custom eval fix

edits to custom installations

custom script support

added moving source eval

pyannote add

cohere fix

custom eval update

cleanup

beautification, custom scripts loader, artefacts json to csv

more requirement update

backend dependent requirements

frontend fixes

backend fix

multiple changes including the local eval

Visualization fixes

bug fixes

evaluation methods update

code fixes

error fix

added support for remote models

New leaderboard method updating script per class

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