🔥 r1.3.0 release note

[yt605155624]

T2S

• Add seek for BytesIO. BytesIO类型时，要保证切到初始位置，这样多次读取才能够正常。比如__call__函数。 #2484 by @ZapBird
• Add mix finetune. [tts] add mix finetune #2525 [tts ] Chmod run_mix.sh #2647 by @lym0302
• Add streaming TTS fastdeploy serving. Add TTS fastdeploy serving #2528 by @HexToString
• Add SSML for Chinese Text Frontend. [TTS]Add SSML for Chinese Text Frontend #2531 by @david-95
• Add end-to-end Prosody Prediction pipeline (including using prosody labels in Acoustic Model). [Text]Add Rhythm Prediction Function #2548 Add rhythm tags for MFA, test=tts #2615 Add prosody prediction in synthesize_e2e, test=tts #2693 by @WongLaw
• Add Adversarial Loss for Chinese English mixed TTS. [tts] add adversarial loss #2588 by @lym0302
• Fix frontend bugs. [TTS]fix g2p #2539 fix frontend bug, test=tts #2606 by @yt605155624
• Add TN for English unit. Revised TN qualifier for measure notation, test=tts #2629 by @WongLaw
• Add male voice for TTS. [tts] Add male voice for tts #2660 by @lym0302
• Add double byte char for zh normalization. add double byte char for zh normalization #2661 by @david-95
• Fix badcase of g2pW. fix badcase #2664 by @kFoodie
• Add TTS Paddle-Lite x86 inference. [TTS]Add export2lite, test=tts #2636 [TTS]Add TTS Paddle-Lite x86 inference #2667 by @yt605155624
• Add greek char and fix [TTS]特殊的句子及标点导致报错 #2571. add greek char and fix issue2571 #2683 by @david-95
• Add Slim for TTS. [TTS]Add slim for TTS #2729 by @yt605155624

S2T

• Add whisper. [s2t] add whisper asr large model #2640 [ASR] update whisper model source, test=doc #2704 by @zxcd
• Fix gpu training hang. [ASR] Chang memory allocator strategy to fix gpu training hang #2478 by @Zth9730
• Support u2++ based cli and server. [ASR] support u2pp based cli and server, optimiz code of u2pp decode (reversed_weight parameter) #2489 [s2t] use reverse_weight in decode.yaml #2510 by @Zth9730
• Add wav2vec2-en. [ASR] wav2vec2 ASR, pre-trained wav2vec2 based CTC for librispeech #2518 [doc] release wav2vec2ASR and wav2vec2.0 model, update Recent Update #2527 [ASR] wav2vec2_en, test=asr #2637 by @Zth9730
• Add wav2vec2-zh cli. [ASR] support wav2vec2-zh cli, test=asr #2697 by @Zth9730

Text

• Fix bug of Punctuation Restoration 标点恢复代码更新，test=asr #2554 by @dahu1

Audio

• Move paddlespeech/audio to paddleaudio. [audio] mv paddlespeech/audio to paddleaudio #2706 by @SmileGoat
• Fix bug of paddleaudio.load. [audio] mv paddlespeech/audio to paddleaudio #2706 by @SmileGoat

Demo

• Add TTSAndroid demo. [TTS]add TTSAndroid demo #2703 by @yt605155624
• Add whisper. add all whisper model size support, test=asr #2677 by @zxcd
• Add wav2vec2. [doc] update wav2vec2 demos README.md, test=doc #2674 @Zth9730

Documentation

• Update install.md. Update install.md #2666 by @michael-skynorth
• Update docs. update docs test=doc #2688 by @heyudage

Other

• Fixed paddlenlp version to paddlenlp >=2.4.3 in setup.py. [ASR] support whisper cli, test=doc #2701 by @zxcd
• Fixed the error when dim of tensor is 0. 支持0维Tensor需要的修改 #2621 by @Zth9730

Acknowledgements

Special thanks to @SmileGoat @yt605155624 @Zth9730 @zxcd @WongLaw @lym0302 @ZapBird
@HexToString @david-95 @kFoodie @michael-skynort @heyudage