search

PaddlePaddle / Parakeet

PAddle PARAllel text-to-speech toolKIT (supporting Tacotron2, Transformer TTS, FastSpeech2/FastPitch, SpeedySpeech, WaveFlow and Parallel WaveGAN)
Other
599 stars 83 forks source link

Refactor of examples, add fastspeech2_ljspeech, pwg_ljspeech, update … #190

Closed yt605155624 closed 3 years ago

yt605155624 commented 3 years ago

1.Refactor of examples, 1)move examples/train.py to Parakeet/utils/_train.py 2)move examples/synthesize.py to Parakeet/utils/_syn.py 3)move examples/_updater.py to parakeet/training/_updater.py 4)move examples/batch_fn.py to parakeet/datasets am_batch_fn.py and vocoder_batch_fn.py

  1. update speedyspeech preprocess（use MFA）and frontend (same with fastspecch2)
  2. update pwg baker preprocess (use MFA to cut sil)
  3. add fastspeech2 ljspeech
  4. add pwg ljspeech
  5. update pwg_baker benchmark
  6. update speedyspeech_baker tests(全链条)