app.py

import gradio as gr
from bark import SAMPLE_RATE, generate_audio, preload_models
from scipy.io.wavfile import write as write_wav
import tempfile
import os

# Download and load all models
preload_models()

# Define a function to generate audio from text
def generate_audio_from_text(text_prompt, language_prompt, gender_prompt):
    if language_prompt == "english":
        if gender_prompt == "male":
            history_prompt = "v2/en_speaker_6"
        else:
            history_prompt = "v2/en_speaker_9"
    elif language_prompt == "french":
        if gender_prompt == "male":
            history_prompt = "v2/fr_speaker_0"
        else:
            history_prompt = "v2/fr_speaker_1"
    elif language_prompt =="german":
        if gender_prompt=="male":
            history_prompt = "v2/de_speaker_2"
        else:
            history_prompt="v2/de_speaker_3"
    elif language_prompt =="hindi":
        if gender_prompt=="male":
            history_prompt = "v2/hi_speaker_8"
        else:
            history_prompt="v2/hi_speaker_3"
    elif language_prompt =="chinese":
        if gender_prompt=="male":
            history_prompt = "v2/zh_speaker_1"
        else:
            history_prompt="v2/zh_speaker_4"
    elif language_prompt =="italian":
        if gender_prompt=="male":
            history_prompt = "v2/it_speaker_4"
        else:
            history_prompt="v2/it_speaker_7"
    elif language_prompt =="japanese":
        if gender_prompt=="male":
            history_prompt = "v2/ja_speaker_2"
        else:
            history_prompt="v2/ja_speaker_0"
    else:
        raise ValueError("Invalid language or gender selection")

    audio_array = generate_audio(text_prompt, history_prompt)

    # Create a temporary WAV file
    with tempfile.NamedTemporaryFile(suffix=".wav", delete=False) as temp_wav:
        temp_wav_path = temp_wav.name
        write_wav(temp_wav_path, SAMPLE_RATE, audio_array)

    return temp_wav_path

# Define lists of language and gender options
language_options = [
    "english",
    "french",
    "german",
    "hindi",
    "chinese",
    "italian",
    "japanese",


]

gender_options = [
    "male",
    "female",
]

# Create a Gradio interface with text input and dropdown menus for language and gender
iface = gr.Interface(
    fn=generate_audio_from_text,
    inputs=[
        gr.Textbox(text="Enter text to convert to speech:"),
        gr.Dropdown(choices=language_options, label="Select language:"),
        gr.Dropdown(choices=gender_options, label="Select gender:"),
    ],
    outputs=gr.outputs.File(label="Download WAV File"),
    title="Text-to-Speech App Vertical Solution", 
)

# Launch the Gradio app with sharing enabled
iface.launch(debug=True)