[TTS] updated config with a German IPA phoneme tokenizer (#4756)

* [TTS] added a German IPA phoneme tokenizer * [TTS][ASR] enabled customized arguments for trimming the leading and trailing silence. * [TTS] disabled spline interpolation for beta-binomial distribution. Let it generate align prior and save to disks. Use a new phoneme tokenizer. * [TTS] use consistent spline interpolation with fastpitch checkpoint when generating mel-spectrograms for hifigan finetune. Signed-off-by: Xuesong Yang <1646669+XuesongYang@users.noreply.github.com>
NVIDIA · Sep 10, 2022 · 052751e · 052751e
1 parent b317f2c
commit 052751e
Show file tree

Hide file tree

Showing 3 changed files with 587 additions and 13 deletions.
diff --git a/examples/tts/conf/de/fastpitch_align_44100.yaml b/examples/tts/conf/de/fastpitch_align_44100.yaml
@@ -67,7 +67,7 @@ model:
     punct_post_process: true
 
   text_tokenizer:
-    _target_: nemo.collections.common.tokenizers.text_to_speech.tts_tokenizers.GermanCharsTokenizer
+    _target_: nemo.collections.tts.torch.tts_tokenizers.GermanPhonemesTokenizer
     punct: true
     apostrophe: true
     pad_with_space: true