unify configs into a single one and add detailed comments providing s…

…upported candidates. Signed-off-by: Xuesong Yang <[email protected]>
NVIDIA · Jul 31, 2023 · bb3e8fd · bb3e8fd
1 parent 38606c1
commit bb3e8fd
Show file tree

Hide file tree

Showing 4 changed files with 6 additions and 759 deletions.
diff --git a/...astpitch_align_24finals_22050_pinyin.yaml → ...es/tts/conf/zh/fastpitch_align_22050.yaml b/...astpitch_align_24finals_22050_pinyin.yaml → ...es/tts/conf/zh/fastpitch_align_22050.yaml
@@ -28,6 +28,12 @@ lowfreq: 0
 highfreq: null
 window: hann
 
+# There are four candidates of `phoneme_dict_path` provided for Chinese as shown below,
+#     1) (default) 24-final Pinyin: "scripts/tts_dataset_files/zh/24finals/pinyin_dict_nv_22.10.txt",
+#     2) IPA converted from 24-final Pinyin: "scripts/tts_dataset_files/zh/24finals/ipa_dict_nv23.05.txt",
+#     3) 36-final Pinyin: "scripts/tts_dataset_files/zh/36finals/pinyin_dict_nv23.05.txt",
+#     4) IPA converted from 36-final Pinyin: "scripts/tts_dataset_files/zh/36finals/ipa_dict_nv23.05.txt"
+# Suggest to choose IPA symbol set converted from 36-final Pinyin because better audio quality were observed.
 phoneme_dict_path: "scripts/tts_dataset_files/zh/24finals/pinyin_dict_nv_22.10.txt"
 
 model:

diff --git a/examples/tts/conf/zh/fastpitch_align_24finals_22050_ipa.yaml b/examples/tts/conf/zh/fastpitch_align_24finals_22050_ipa.yaml