Spaces:
Sleeping
Sleeping
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,43 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import streamlit as st
|
| 2 |
+
from TTS.api import TTS
|
| 3 |
+
import torch
|
| 4 |
+
from io import BytesIO
|
| 5 |
+
|
| 6 |
+
# Initialize the TTS model
|
| 7 |
+
device = "cuda" if torch.cuda.is_available() else "cpu"
|
| 8 |
+
tts = TTS(model_name="tts_models/en/ljspeech/tacotron2-DDC", progress_bar=True, gpu=device == "cuda")
|
| 9 |
+
|
| 10 |
+
# App title
|
| 11 |
+
st.title("🐸 Coqui TTS - Text to Speech App")
|
| 12 |
+
|
| 13 |
+
# Text input
|
| 14 |
+
text = st.text_area("Enter the text to convert to speech", "Imagine being able to offer your clients curated travel insights at their fingertips.")
|
| 15 |
+
|
| 16 |
+
# File upload for voice cloning (optional)
|
| 17 |
+
speaker_wav = st.file_uploader("Upload a speaker audio file for voice cloning (optional)", type=["wav"])
|
| 18 |
+
|
| 19 |
+
# Submit button
|
| 20 |
+
if st.button("Generate Speech"):
|
| 21 |
+
if text.strip() == "":
|
| 22 |
+
st.warning("Please enter some text.")
|
| 23 |
+
else:
|
| 24 |
+
# Run TTS
|
| 25 |
+
st.text("Generating speech...")
|
| 26 |
+
try:
|
| 27 |
+
if speaker_wav is not None:
|
| 28 |
+
# Use voice cloning
|
| 29 |
+
wav = tts.tts(text=text, speaker_wav=speaker_wav)
|
| 30 |
+
else:
|
| 31 |
+
# Default voice
|
| 32 |
+
wav = tts.tts(text=text)
|
| 33 |
+
|
| 34 |
+
# Display audio player
|
| 35 |
+
st.audio(wav, format="audio/wav")
|
| 36 |
+
|
| 37 |
+
# Option to download the audio file
|
| 38 |
+
buffer = BytesIO()
|
| 39 |
+
buffer.write(wav)
|
| 40 |
+
buffer.seek(0)
|
| 41 |
+
st.download_button("Download Audio", buffer, file_name="output.wav", mime="audio/wav")
|
| 42 |
+
except Exception as e:
|
| 43 |
+
st.error(f"An error occurred: {e}")
|