search

TensorSpeech / TensorFlowTTS

:stuck_out_tongue_closed_eyes: TensorFlowTTS: Real-Time State-of-the-art Speech Synthesis for Tensorflow 2 (supported including English, French, Korean, Chinese, German and Easy to adapt for other languages)
https://tensorspeech.github.io/TensorFlowTTS/
Apache License 2.0
3.84k stars 815 forks source link

Quick question regarding data structure. #267

Closed jun-danieloh closed 4 years ago

jun-danieloh commented 4 years ago 
|   |- stats.npy
|   |- stats_f0.npy
|   |- stats_energy.npy
|   |- train_utt_ids.npy
|   |- valid_utt_ids.npy

I am looking at code but not able to find these files usages. Are these files are used while training?

dathudeptrai commented 4 years ago

@jun-danieloh stats is mel_stat and it will use to normalize the mel, you can see a folder named norm-feats. stats_f0 and stats_energy will use to normalize f0/energy in FastSpeech2.

jun-danieloh commented 4 years ago

@dathudeptrai stats.npy is used for normalizing *-raw-feats.npy to *-norm-feats.npy, right?

dathudeptrai commented 4 years ago

Yes

Vào Th 6, 18 thg 9, 2020 lúc 12:12 Daniel Oh notifications@github.com đã viết:

@dathudeptrai https://github.com/dathudeptrai stats.npy is used for normalizing -raw-feats.npy to -norm-feats.npy, right?

— You are receiving this because you were mentioned.

Reply to this email directly, view it on GitHub https://github.com/TensorSpeech/TensorFlowTTS/issues/267#issuecomment-694654958, or unsubscribe https://github.com/notifications/unsubscribe-auth/AKOWD5YKI4XQKJZAGDZNVFTSGLT33ANCNFSM4RQLOH4A .