app.py

import gradio as gr
import json
import torch
import time
import wavio

from tqdm import tqdm
from huggingface_hub import snapshot_download
from models import AudioDiffusion, DDPMScheduler
from audioldm.audio.stft import TacotronSTFT
from audioldm.variational_autoencoder import AutoencoderKL

# Automatic device detection
if torch.cuda.is_available():
    device_type = "cuda"
    device_selection = "cuda:0"
else:
    device_type = "cpu"
    device_selection = "cpu"

class Tango:
    def __init__(self, name = "declare-lab/tango2", device = device_selection):
        
        path = snapshot_download(repo_id = name)
        
        vae_config = json.load(open("{}/vae_config.json".format(path)))
        stft_config = json.load(open("{}/stft_config.json".format(path)))
        main_config = json.load(open("{}/main_config.json".format(path)))
        
        self.vae = AutoencoderKL(**vae_config).to(device)
        self.stft = TacotronSTFT(**stft_config).to(device)
        self.model = AudioDiffusion(**main_config).to(device)
        
        vae_weights = torch.load("{}/pytorch_model_vae.bin".format(path), map_location = device)
        stft_weights = torch.load("{}/pytorch_model_stft.bin".format(path), map_location = device)
        main_weights = torch.load("{}/pytorch_model_main.bin".format(path), map_location = device)
        
        self.vae.load_state_dict(vae_weights)
        self.stft.load_state_dict(stft_weights)
        self.model.load_state_dict(main_weights)

        print ("Successfully loaded checkpoint from:", name)
        
        self.vae.eval()
        self.stft.eval()
        self.model.eval()
        
        self.scheduler = DDPMScheduler.from_pretrained(main_config["scheduler_name"], subfolder = "scheduler")
        
    def chunks(self, lst, n):
        # Yield successive n-sized chunks from a list
        for i in range(0, len(lst), n):
            yield lst[i:i + n]
        
    def generate(self, prompt, steps = 100, guidance = 3, samples = 1, disable_progress = True):
        # Generate audio for a single prompt string
        with torch.no_grad():
            latents = self.model.inference([prompt], self.scheduler, steps, guidance, samples, disable_progress = disable_progress)
            mel = self.vae.decode_first_stage(latents)
            wave = self.vae.decode_to_waveform(mel)
        return wave
    
    def generate_for_batch(self, prompts, steps = 200, guidance = 3, samples = 1, batch_size = 8, disable_progress = True):
        # Generate audio for a list of prompt strings
        outputs = []
        for k in tqdm(range(0, len(prompts), batch_size)):
            batch = prompts[k: k + batch_size]
            with torch.no_grad():
                latents = self.model.inference(batch, self.scheduler, steps, guidance, samples, disable_progress = disable_progress, length = 20)
                mel = self.vae.decode_first_stage(latents)
                wave = self.vae.decode_to_waveform(mel)
                outputs += [item for item in wave]
        if samples == 1:
            return outputs
        return list(self.chunks(outputs, samples))

# Initialize TANGO

tango = Tango(device = "cpu")
tango.vae.to(device_type)
tango.stft.to(device_type)
tango.model.to(device_type)

def check(
    prompt,
    output_format,
    output_number,
    steps,
    guidance
):
    if prompt is None or prompt == "":
        raise gr.Error("Please provide a prompt input.")
    if not output_format in ["wav", "mp3"]:
        raise gr.Error("Please choose an allowed output format (.wav or .mp3).")
    if not output_number in [1, 2, 3]:
        raise gr.Error("Please ask for 1, 2 or 3 output files.")

def update_output(output_format, output_number):
    return [
        gr.update(format = output_format),
        gr.update(format = output_format, visible = (2 <= output_number)),
        gr.update(format = output_format, visible = (output_number == 3))
    ]

def text2audio(
    prompt,
    output_format,
    output_number,
    steps,
    guidance
):
    start = time.time()
    output_wave = tango.generate(prompt, steps, guidance, output_number)

    output_filename_1 = "tmp1_.wav"
    wavio.write(output_filename, output_wave[0], rate = 16000, sampwidth = 2)

    if (2 <= output_number):
        output_filename_2 = "tmp2_.wav"
        wavio.write(output_filename, output_wave[1], rate = 16000, sampwidth = 2)
    else:
        output_filename_2 = None

    if (output_number == 3):
        output_filename_3 = "tmp3_.wav"
        wavio.write(output_filename, output_wave[2], rate = 16000, sampwidth = 2)
    else:
        output_filename_3 = None

    end = time.time()
    secondes = int(end - start)
    minutes = secondes // 60
    secondes = secondes - (minutes * 60)
    hours = minutes // 60
    minutes = minutes - (hours * 60)
    return [
        output_filename_1,
        output_filename_2,
        output_filename_3,
        "Start again to get a different result. The output have been generated in " + ((str(hours) + " h, ") if hours != 0 else "") + ((str(minutes) + " min, ") if hours != 0 or minutes != 0 else "") + str(secondes) + " sec."
    ]

# Gradio interface
with gr.Blocks() as interface:
    gr.Markdown("""
        <p style="text-align: center;">
        <b><big><big><big>Text-to-Audio</big></big></big></b>
        <br/>Generates 10 seconds of sound effects from description, freely, without account, without watermark, that you can download.
        </p>
        <br/>
        <br/>
        ✨ Powered by <i>Tango 2</i> AI.
        <br/>
        <ul>
        <li>If you need to generate <b>music</b>, I recommend to use <i>MusicGen</i>,</li>
        </ul>
        <br/>
        🐌 Slow process... ~2 hours. Your computer must <b><u>not</u></b> enter into standby mode.<br/>You can duplicate this space on a free account, it works on CPU.<br/>
        <a href='https://huggingface.co/spaces/Fabrice-TIERCELIN/Text-to-Audio?duplicate=true'><img src='https://img.shields.io/badge/-Duplicate%20Space-blue?labelColor=white&style=flat&logo=data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAAAXNSR0IArs4c6QAAAP5JREFUOE+lk7FqAkEURY+ltunEgFXS2sZGIbXfEPdLlnxJyDdYB62sbbUKpLbVNhyYFzbrrA74YJlh9r079973psed0cvUD4A+4HoCjsA85X0Dfn/RBLBgBDxnQPfAEJgBY+A9gALA4tcbamSzS4xq4FOQAJgCDwV2CPKV8tZAJcAjMMkUe1vX+U+SMhfAJEHasQIWmXNN3abzDwHUrgcRGmYcgKe0bxrblHEB4E/pndMazNpSZGcsZdBlYJcEL9Afo75molJyM2FxmPgmgPqlWNLGfwZGG6UiyEvLzHYDmoPkDDiNm9JR9uboiONcBXrpY1qmgs21x1QwyZcpvxt9NS09PlsPAAAAAElFTkSuQmCC&logoWidth=14'></a>
        <br/>
        ⚖️ You can use, modify and share the generated sounds but not for commercial uses.
        """
    )
    input_text = gr.Textbox(label = "Prompt", value = "Snort of a horse", lines = 2, autofocus = True)
    output_format = gr.Radio(label = "Output format", info = "The file you can dowload", choices = ["mp3", "wav"], value = "wav")
    with gr.Accordion("Advanced options", open = False):
        output_number = gr.Slider(label = "Number of generations", info = "1, 2 or 3 output files", minimum = 1, maximum = 3, value = 3, step = 1, interactive = True)
        denoising_steps = gr.Slider(label = "Steps", info = "lower=faster & variant, higher=audio quality & similar", minimum = 100, maximum = 200, value = 100, step = 1, interactive = True)
        guidance_scale = gr.Slider(label = "Guidance Scale", info = "lower=audio quality, higher=follow the prompt", minimum = 1, maximum = 10, value = 3, step = 0.1, interactive = True)

    submit = gr.Button("🚀 Generate", variant = "primary")

    output_audio_1 = gr.Audio(label = "Generated Audio #1/3", format = "wav", type="filepath", autoplay = True)
    output_audio_2 = gr.Audio(label = "Generated Audio #2/3", format = "wav", type="filepath")
    output_audio_3 = gr.Audio(label = "Generated Audio #3/3", format = "wav", type="filepath")
    information = gr.Label(label = "Information")

    submit.click(fn = check, inputs = [
        input_text,
        output_format,
        output_number,
        denoising_steps,
        guidance_scale
    ], outputs = [], queue = False, show_progress = False).success(fn = update_output, inputs = [
        output_format,
        output_number
    ], outputs = [
        output_audio_1,
        output_audio_2,
        output_audio_3
    ], queue = False, show_progress = False).success(fn = text2audio, inputs = [
        input_text,
        output_format,
        output_number,
        denoising_steps,
        guidance_scale
    ], outputs = [
        output_audio_1,
        output_audio_2,
        output_audio_3,
        information
    ], scroll_to_output = True)

    gr.Examples(
        fn = text2audio,
	    inputs = [
            input_text,
            output_format,
            output_number,
            denoising_steps,
            guidance_scale
        ],
	    outputs = [
            output_audio_1,
            output_audio_2,
            output_audio_3,
            information
        ],
        examples = [
                ["A hammer is hitting a wooden surface", "mp3", 3, 100, 3],
                ["Peaceful and calming ambient music with singing bowl and other instruments.", "wav", 3, 100, 3],
                ["A man is speaking in a small room.", "mp3", 2, 100, 3],
                ["A female is speaking followed by footstep sound", "mp3", 1, 100, 3],
                ["Wooden table tapping sound followed by water pouring sound.", "mp3", 3, 200, 3],
            ],
        cache_examples = "lazy",
    )
    
    gr.Markdown(
        """
        ## How to prompt your sound
        You can use round brackets to increase the importance of a part:
        ```
        Peaceful and (calming) ambient music with singing bowl and other instruments
        ```
        You can use several levels of round brackets to even more increase the importance of a part:
        ```
        (Peaceful) and ((calming)) ambient music with singing bowl and other instruments
        ```
        You can use number instead of several round brackets:
        ```
        (Peaceful:1.5) and ((calming)) ambient music with singing bowl and other instruments
        ```
        You can do the same thing with square brackets to decrease the importance of a part:
        ```
        (Peaceful:1.5) and ((calming)) ambient music with [singing:2] bowl and other instruments
        """
    )
        
    interface.queue(10).launch()