[Demo]Add st demo.

[Demo]Update asr and audio tagging demo.

[setup.py]remove the py3.6 in setup.py

[ctcdecoders] Separate the ctcdecoders

[ST] Add st cli

[CLI]Update tags of pretrained_models.

[CLI][CLS]Add audio classification cli.

update_bs

[setup.py]change the required py_version

[demos][docs]update demos readme

[asr] update librispeech asr1 transformer result