bark_exemple.py

import numpy as np
import gradio as gr
from bark import SAMPLE_RATE, generate_audio, preload_models
from bark.generation import SUPPORTED_LANGS
from share_btn import community_icon_html, loading_icon_html, share_js

DEBUG_MODE = False

if not DEBUG_MODE:
    _ = preload_models()

AVAILABLE_PROMPTS = ["Unconditional", "Announcer"]
PROMPT_LOOKUP = {}
for _, lang in SUPPORTED_LANGS:
    for n in range(10):
        label = f"Speaker {n} ({lang})"
        AVAILABLE_PROMPTS.append(label)
        PROMPT_LOOKUP[label] = f"{lang}_speaker_{n}"
PROMPT_LOOKUP["Unconditional"] = None
PROMPT_LOOKUP["Announcer"] = "announcer"

default_text = "Hello, my name is Suno. And, uh — and I like pizza. [laughs]\nBut I also have other interests such as playing tic tac toe."

title = "# 🐶 Bark</div>"

description = """
"""

article = """
"""

examples = [
    ["Please surprise me and speak in whatever voice you enjoy. Vielen Dank und Gesundheit!",
        "Unconditional"],  # , 0.7, 0.7],
    ["Hello, my name is Suno. And, uh — and I like pizza. [laughs] But I also have other interests such as playing tic tac toe.",
        "Speaker 1 (en)"],  # , 0.7, 0.7],
    ["Buenos días Miguel. Tu colega piensa que tu alemán es extremadamente malo. But I suppose your english isn't terrible.",
        "Speaker 0 (es)"],  # , 0.7, 0.7],
]


def gen_tts(text, history_prompt):  # , temp_semantic, temp_waveform):
    history_prompt = PROMPT_LOOKUP[history_prompt]
    if DEBUG_MODE:
        audio_arr = np.zeros(SAMPLE_RATE)
    else:
        # , text_temp=temp_semantic, waveform_temp=temp_waveform)
        audio_arr = generate_audio(text, history_prompt=history_prompt)
    audio_arr = (audio_arr * 32767).astype(np.int16)
    return (SAMPLE_RATE, audio_arr)


css = """

"""
with gr.Blocks(css=css) as block:
    gr.Markdown(title)
    gr.Markdown(description)
    with gr.Row():
        with gr.Column():
            input_text = gr.Textbox(
                label="Input Text", lines=2, value=default_text, elem_id="input_text")
            options = gr.Dropdown(
                AVAILABLE_PROMPTS, value="Speaker 1 (en)", label="Acoustic Prompt", elem_id="speaker_option")
            run_button = gr.Button(text="Generate Audio", type="button")
        with gr.Column():
            audio_out = gr.Audio(label="Generated Audio",
                                 type="numpy", elem_id="audio_out")
            with gr.Row(visible=False) as share_row:
                with gr.Group(elem_id="share-btn-container"):
                    community_icon = gr.HTML(community_icon_html)
                    loading_icon = gr.HTML(loading_icon_html)
                    share_button = gr.Button(
                        "Share to community", elem_id="share-btn")
                    share_button.click(None, [], [], _js=share_js)
    inputs = [input_text, options]
    outputs = [audio_out]
    gr.Examples(examples=examples, fn=gen_tts, inputs=inputs,
                outputs=outputs, cache_examples=True)
    gr.Markdown(article)
    run_button.click(fn=lambda: gr.update(visible=False), inputs=None, outputs=share_row, queue=False).then(
        fn=gen_tts, inputs=inputs, outputs=outputs, queue=True).then(
        fn=lambda: gr.update(visible=True), inputs=None, outputs=share_row, queue=False)

block.queue()
block.launch()