print log

app.py

@@ -19,6 +19,8 @@ def text_to_speech(text):
         "acoustic_latest_ckpt.pickle",
         "duration_latest_ckpt.pickle",
     )
+    print("mel")
+    print(mel)
     wave = mel2wave(mel, "config.json", "hk_hifi.pickle")
     print("wave")
     print(wave)

vietTTS/nat/text2mel.py

@@ -89,6 +89,7 @@ def text2mel(
     duration_ckpt=FLAGS.ckpt_dir / "duration_latest_ckpt.pickle",
 ):
     tokens = text2tokens(text, lexicon_fn)
+    print("text2mel-token : " + tokens)
     durations = predict_duration(tokens, duration_ckpt)
     durations = jnp.where(
         np.array(tokens)[None, :] == FLAGS.sil_index,