Upload 2 files

app.py

@@ -0,0 +1,35 @@
+import streamlit as st
+from tensorflow_tts.inference import AutoProcessor, TFAutoModel
+import tensorflow as tf
+
+import numpy as np
+import soundfile as sf
+import yaml
+
+processor = AutoProcessor.from_pretrained("MarcNg/fastspeech2-vi-infore")
+fastspeech2 = TFAutoModel.from_pretrained("MarcNg/fastspeech2-vi-infore")
+mb_melgan = TFAutoModel.from_pretrained("tensorspeech/tts-mb_melgan-ljspeech-en")
+
+output = "output.wav"
+
+st.header("MarcNg/fastspeech2-vi-infore Demo")
+
+def tts(text):
+  input_ids = processor.text_to_sequence(text)
+
+  mel_before, mel_after, duration_outputs, _, _ = fastspeech2.inference(
+      input_ids=tf.expand_dims(tf.convert_to_tensor(input_ids, dtype=tf.int32), 0),
+      speaker_ids=tf.convert_to_tensor([0], dtype=tf.int32),
+      speed_ratios=tf.convert_to_tensor([1.0], dtype=tf.float32),
+      f0_ratios =tf.convert_to_tensor([1.0], dtype=tf.float32),
+      energy_ratios =tf.convert_to_tensor([1.0], dtype=tf.float32),
+  )
+  return mel_after
+
+text = st.text_input("Text to process")
+
+if st.button("Speak"):
+  mel_after = tts(text)
+  audio_after = mb_melgan.inference(mel_after)[0, :, 0]
+  sf.write(output, audio_after, 22050, 'PCM_16')
+  st.audio(output, format='audio/wav')