Use piper_phonemize as text tokenizer in vctk TTS recipe

[JinZr]

akin to #1511 by @yaozengwei

[JinZr]

i think `sampling_rate` is an attribute of the model according to the LJSpeech recipe in the master branch (the VCTK recipe soft links the vits.py to the one in LJSpeech), is this the attribute you are referring to?

…

On Monday, Mar 04, 2024 at 15:03, Fangjun Kuang ***@***.*** ***@***.***)> wrote: @csukuangfj commented on this pull request. In egs/vctk/TTS/vits/export-onnx.py (#1522 (comment)): > "version": "1", "model_author": "k2-fsa", - "comment": "VITS generator", + "comment": "icefall", # must be icefall for models from icefall + "language": "English", + "voice": "en-us", # Choose your language appropriately + "has_espeak": 1, + "n_speakers": 108, + "sample_rate": 22050, # Must match the real sample rate Can we obtain the sample rate from the model? Is sample_rate an attribute of the model? — Reply to this email directly, view it on GitHub (#1522 (review)), or unsubscribe (https://github.com/notifications/unsubscribe-auth/AOON42DEY54SUQX6GGZXSADYWQME3AVCNFSM6AAAAABD7SIFQGVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMYTSMJTGUYDCOBRGU). You are receiving this because you authored the thread.Message ID: ***@***.***>

[JinZr]

Pretrained model can be found here: https://huggingface.co/zrjin/icefall-tts-vctk-vits-2024-03-18