speech-text / app.py
Ra-Is's picture
Update app.py
ab8e629 verified
import gradio as gr
import requests
import os
def send_audio(audio):
url = os.getenv("TRANSLATION_URL")
if audio is None:
return {"error": "No audio file provided"}
try:
if isinstance(audio, bytes):
# If audio is a bytes object, save it to a temporary file
with open("temp_audio.wav", "wb") as f:
f.write(audio)
audio_path = "temp_audio.wav"
else:
# If audio is already a file path
audio_path = audio
# Prepare form data for request
files = {
'file': open(audio_path, 'rb'),
}
data = {
'lang': 'english-twi'
}
response = requests.post(url, files=files, data=data)
response_json = response.json()
#return response_json
# Extract the translation part
translation = response_json.get('data', 'No translation available')
return translation
except Exception as e:
return {"error": str(e)}
# Gradio interface for recording speech
iface = gr.Interface(
fn=send_audio,
inputs=gr.Audio(type="filepath"), # Expect a file path or file-like object
outputs="text",
title="English-Twi Speech Translation",
description="Record speech and send for processing.",
)
iface.launch()