|
import gradio as gr |
|
import os |
|
|
|
|
|
os.system('/usr/local/bin/python -m pip install --upgrade pip') |
|
os.system("pip install git+https://github.com/allthingssecurity/vall-e/") |
|
os.system("pip install pydub") |
|
from pydub import AudioSegment |
|
import tempfile |
|
|
|
def synthesis(audio_file: tempfile._TemporaryFileWrapper, text: str, ar_ckpt: str, nar_ckpt: str): |
|
|
|
audio = AudioSegment.from_file(audio_file) |
|
|
|
temp_file = "temp.wav" |
|
audio.export(temp_file, format="wav") |
|
|
|
output_file = "output.wav" |
|
vall_e.synthesis(text, temp_file, output_file, ar_ckpt=ar_ckpt, nar_ckpt=nar_ckpt) |
|
|
|
output_audio = AudioSegment.from_file(output_file) |
|
return output_audio |
|
|
|
synthesis_interface = gr.Interface(synthesis, |
|
inputs={"audio_file": gr.File(file_count="multiple"), "text": "text", "ar_ckpt": "text", "nar_ckpt": "text"}, |
|
live=True, |
|
capture_session=True,outputs="audio") |
|
synthesis_interface.launch() |