Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
import streamlit as st
|
| 2 |
-
import fitz # PyMuPDF
|
| 3 |
-
from TTS.api import TTS
|
| 4 |
import os
|
| 5 |
|
| 6 |
# Title of the app
|
|
@@ -27,21 +27,39 @@ if uploaded_file is not None:
|
|
| 27 |
if text.strip() == "":
|
| 28 |
st.warning("No text found in the PDF. Please upload a valid document.")
|
| 29 |
else:
|
| 30 |
-
#
|
| 31 |
st.subheader("Generate Speech")
|
| 32 |
st.info("Using TTS single-speaker model: 'tts_models/en/ljspeech/tacotron2-DDC'")
|
|
|
|
| 33 |
model_name = "tts_models/en/ljspeech/tacotron2-DDC" # Single-speaker model
|
| 34 |
tts = TTS(model_name)
|
| 35 |
-
|
| 36 |
-
# Specify output file
|
| 37 |
audio_path = "output.wav"
|
| 38 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 39 |
try:
|
| 40 |
-
|
| 41 |
-
|
|
|
|
|
|
|
|
|
|
| 42 |
st.success("Speech generation complete!")
|
| 43 |
-
|
| 44 |
-
# Audio playback
|
| 45 |
st.audio(audio_path, format="audio/wav", start_time=0)
|
| 46 |
|
| 47 |
# Cleanup button
|
|
|
|
| 1 |
import streamlit as st
|
| 2 |
+
import fitz # PyMuPDF
|
| 3 |
+
from TTS.api import TTS
|
| 4 |
import os
|
| 5 |
|
| 6 |
# Title of the app
|
|
|
|
| 27 |
if text.strip() == "":
|
| 28 |
st.warning("No text found in the PDF. Please upload a valid document.")
|
| 29 |
else:
|
| 30 |
+
# Text-to-Speech Conversion
|
| 31 |
st.subheader("Generate Speech")
|
| 32 |
st.info("Using TTS single-speaker model: 'tts_models/en/ljspeech/tacotron2-DDC'")
|
| 33 |
+
|
| 34 |
model_name = "tts_models/en/ljspeech/tacotron2-DDC" # Single-speaker model
|
| 35 |
tts = TTS(model_name)
|
|
|
|
|
|
|
| 36 |
audio_path = "output.wav"
|
| 37 |
|
| 38 |
+
# Split text into smaller chunks (e.g., 500 characters)
|
| 39 |
+
def split_text(text, max_length=500):
|
| 40 |
+
sentences = text.split(". ")
|
| 41 |
+
chunks = []
|
| 42 |
+
chunk = ""
|
| 43 |
+
for sentence in sentences:
|
| 44 |
+
if len(chunk) + len(sentence) < max_length:
|
| 45 |
+
chunk += sentence + ". "
|
| 46 |
+
else:
|
| 47 |
+
chunks.append(chunk.strip())
|
| 48 |
+
chunk = sentence + ". "
|
| 49 |
+
if chunk:
|
| 50 |
+
chunks.append(chunk.strip())
|
| 51 |
+
return chunks
|
| 52 |
+
|
| 53 |
+
chunks = split_text(text)
|
| 54 |
+
|
| 55 |
+
# Generate audio for each chunk
|
| 56 |
try:
|
| 57 |
+
with open(audio_path, "wb") as audio_file:
|
| 58 |
+
for i, chunk in enumerate(chunks):
|
| 59 |
+
st.write(f"Processing chunk {i + 1} of {len(chunks)}...")
|
| 60 |
+
audio_chunk = tts.tts(chunk)
|
| 61 |
+
audio_file.write(audio_chunk)
|
| 62 |
st.success("Speech generation complete!")
|
|
|
|
|
|
|
| 63 |
st.audio(audio_path, format="audio/wav", start_time=0)
|
| 64 |
|
| 65 |
# Cleanup button
|