Spaces:
Sleeping
Sleeping
File size: 1,370 Bytes
f41d3a4 c05c224 0bcf9d7 88a9a72 4f47fc6 c05c224 3cf914c 55d6b51 3cf914c 55d6b51 3cf914c 55d6b51 0bcf9d7 c05c224 4f47fc6 f41d3a4 f86b0c1 774ebda f41d3a4 774ebda f41d3a4 f6fc20c f41d3a4 774ebda f41d3a4 774ebda f41d3a4 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 |
import os
import torch
import gradio as gr
from TTS.api import TTS
# ✅ Accept Coqui License Automatically
os.environ["COQUI_TOS_AGREED"] = "1"
# ✅ Ensure CPU-only execution (disable GPU)
device = "cpu"
# ✅ Manually Load XTTS Model (Without `XttsConfig`)
model_name = "tts_models/multilingual/multi-dataset/xtts_v2"
model_path = f"{os.getenv('HOME')}/.local/share/tts/{model_name.replace('/', '--')}" # Custom model path
tts = TTS(model_name=model_name).to(device)
def generate_cloned_voice(text, reference_audio, language):
output_path = "output.wav"
# Generate cloned speech with language specification
tts.tts_to_file(text=text, speaker_wav=reference_audio, file_path=output_path, language=language)
return output_path
print(tts.languages)
# Create the Gradio interface
interface = gr.Interface(
fn=generate_cloned_voice,
inputs=[
gr.Textbox(label="Enter Translated Text"),
gr.Audio(label="Upload Reference Audio", type="filepath"),
gr.Dropdown(["en", "fr", "de", "es", "it"], label="Select Target Language", value="en")
],
outputs=gr.Audio(label="Generated Cloned Voice"),
title="Free Voice Cloning API",
description="Upload a sample voice and input text. Select a language, and the system will generate the text in the same voice."
)
# Launch the Gradio app
interface.launch()
|