* replace kaldi-fbank with kaldi-native-fbank

* make kaldi-native-fbank work

* fix openfst download error

* add acknowledgments of openfst

* refactor directory

* clean ctc_decoders dir

* add nnet cache && make 2 thread work

* do not compile websocket

* rm ds2 && rm boost

* rm ds2 example

* add nnet cache && make 2 thread work

* do not compile websocket

* refactor directory

* fix openfst download error

* add acknowledgments of openfst

* fix tensor utils

* update paddle from 2.4.0rc to 2.4.1 in readme

* split paddlespeech/audio to paddleaudio.
* add sox io ,sox effect, kaldi native fbank to paddleaudio.

* rm paddlelite in setup.py

* fix setup.py

* update released_model.md, test=doc

* Update released_model.md

* add TTSAndroid demo

* doc for ds2 websocket

* support wav2vec2-zh cli, test=asr

* support wav2vec2-zh cli, test=asr

* support wav2vec2-zh cli, test=asr

* support wav2vec2-zh cli, test=asr

* more doc of speechx u2 onnx

Co-authored-by: TianYuan <white-sky@qq.com>

* add all whisper model size support

* add choices in parser.

* fix Whisper cli model download path error.

* fix resource download path.

* fix code style

* fix wav2vec2 demos, test=doc

* fix wav2vec2 demos, test=doc

* fix enc_dropout and nor.py, test=asr