app.py

import numpy as np
import torch
import gradio as gr
import spaces
from queue import Queue
from threading import Thread
from typing import Optional
from transformers import MusicgenForConditionalGeneration, MusicgenProcessor, set_seed
from transformers.generation.streamers import BaseStreamer

model = MusicgenForConditionalGeneration.from_pretrained("facebook/musicgen-small")
processor = MusicgenProcessor.from_pretrained("facebook/musicgen-small")

title = "9🌍MusicHub - Text to Music Stream Generator"
description = """  Facebook MusicGen-Small Model - Generate and stream music with model https://huggingface.co/facebook/musicgen-small  """
article = """
## How It Works:
MusicGen is an auto-regressive transformer-based model, meaning generates audio codes (tokens) in a causal fashion.
At each decoding step, the model generates a new set of audio codes, conditional on the text input and all previous audio codes. From the 
frame rate of the [EnCodec model](https://huggingface.co/facebook/encodec_32khz) used to decode the generated codes to audio waveform.
"""


class MusicgenStreamer(BaseStreamer):
    def __init__(
        self,
        model: MusicgenForConditionalGeneration,
        device: Optional[str] = None,
        play_steps: Optional[int] = 10,
        stride: Optional[int] = None,
        timeout: Optional[float] = None,
    ):
        """
        Streamer that stores playback-ready audio in a queue, to be used by a downstream application as an iterator. This is
        useful for applications that benefit from acessing the generated audio in a non-blocking way (e.g. in an interactive
        Gradio demo).
        Parameters:
            model (`MusicgenForConditionalGeneration`):
                The MusicGen model used to generate the audio waveform.
            device (`str`, *optional*):
                The torch device on which to run the computation. If `None`, will default to the device of the model.
            play_steps (`int`, *optional*, defaults to 10):
                The number of generation steps with which to return the generated audio array. Using fewer steps will 
                mean the first chunk is ready faster, but will require more codec decoding steps overall. This value 
                should be tuned to your device and latency requirements.
            stride (`int`, *optional*):
                The window (stride) between adjacent audio samples. Using a stride between adjacent audio samples reduces
                the hard boundary between them, giving smoother playback. If `None`, will default to a value equivalent to 
                play_steps // 6 in the audio space.
            timeout (`int`, *optional*):
                The timeout for the audio queue. If `None`, the queue will block indefinitely. Useful to handle exceptions
                in `.generate()`, when it is called in a separate thread.
        """
        self.decoder = model.decoder
        self.audio_encoder = model.audio_encoder
        self.generation_config = model.generation_config
        self.device = device if device is not None else model.device

        # variables used in the streaming process
        self.play_steps = play_steps
        if stride is not None:
            self.stride = stride
        else:
            hop_length = np.prod(self.audio_encoder.config.upsampling_ratios)
            self.stride = hop_length * (play_steps - self.decoder.num_codebooks) // 6
        self.token_cache = None
        self.to_yield = 0

        # varibles used in the thread process
        self.audio_queue = Queue()
        self.stop_signal = None
        self.timeout = timeout

    def apply_delay_pattern_mask(self, input_ids):
        # build the delay pattern mask for offsetting each codebook prediction by 1 (this behaviour is specific to MusicGen)
        _, decoder_delay_pattern_mask = self.decoder.build_delay_pattern_mask(
            input_ids[:, :1],
            pad_token_id=self.generation_config.decoder_start_token_id,
            max_length=input_ids.shape[-1],
        )
        # apply the pattern mask to the input ids
        input_ids = self.decoder.apply_delay_pattern_mask(input_ids, decoder_delay_pattern_mask)

        # revert the pattern delay mask by filtering the pad token id
        input_ids = input_ids[input_ids != self.generation_config.pad_token_id].reshape(
            1, self.decoder.num_codebooks, -1
        )

        # append the frame dimension back to the audio codes
        input_ids = input_ids[None, ...]

        # send the input_ids to the correct device
        input_ids = input_ids.to(self.audio_encoder.device)

        output_values = self.audio_encoder.decode(
            input_ids,
            audio_scales=[None],
        )
        audio_values = output_values.audio_values[0, 0]
        return audio_values.cpu().float().numpy()

    def put(self, value):
        batch_size = value.shape[0] // self.decoder.num_codebooks
        if batch_size > 1:
            raise ValueError("MusicgenStreamer only supports batch size 1")

        if self.token_cache is None:
            self.token_cache = value
        else:
            self.token_cache = torch.concatenate([self.token_cache, value[:, None]], dim=-1)

        if self.token_cache.shape[-1] % self.play_steps == 0:
            audio_values = self.apply_delay_pattern_mask(self.token_cache)
            self.on_finalized_audio(audio_values[self.to_yield : -self.stride])
            self.to_yield += len(audio_values) - self.to_yield - self.stride

    def end(self):
        """Flushes any remaining cache and appends the stop symbol."""
        if self.token_cache is not None:
            audio_values = self.apply_delay_pattern_mask(self.token_cache)
        else:
            audio_values = np.zeros(self.to_yield)

        self.on_finalized_audio(audio_values[self.to_yield :], stream_end=True)

    def on_finalized_audio(self, audio: np.ndarray, stream_end: bool = False):
        """Put the new audio in the queue. If the stream is ending, also put a stop signal in the queue."""
        self.audio_queue.put(audio, timeout=self.timeout)
        if stream_end:
            self.audio_queue.put(self.stop_signal, timeout=self.timeout)

    def __iter__(self):
        return self

    def __next__(self):
        value = self.audio_queue.get(timeout=self.timeout)
        if not isinstance(value, np.ndarray) and value == self.stop_signal:
            raise StopIteration()
        else:
            return value


sampling_rate = model.audio_encoder.config.sampling_rate
frame_rate = model.audio_encoder.config.frame_rate

target_dtype = np.int16
max_range = np.iinfo(target_dtype).max


@spaces.GPU
def generate_audio(text_prompt, audio_length_in_s=10.0, play_steps_in_s=2.0, seed=0):
    max_new_tokens = int(frame_rate * audio_length_in_s)
    play_steps = int(frame_rate * play_steps_in_s)

    device = "cuda:0" if torch.cuda.is_available() else "cpu"
    if device != model.device:
        model.to(device)
        if device == "cuda:0":
            model.half()

    inputs = processor(
        text=text_prompt,
        padding=True,
        return_tensors="pt",
    )

    streamer = MusicgenStreamer(model, device=device, play_steps=play_steps)

    generation_kwargs = dict(
        **inputs.to(device),
        streamer=streamer,
        max_new_tokens=max_new_tokens,
    )
    thread = Thread(target=model.generate, kwargs=generation_kwargs)
    thread.start()

    set_seed(seed)
    for new_audio in streamer:
        print(f"Sample of length: {round(new_audio.shape[0] / sampling_rate, 2)} seconds")
        new_audio = (new_audio * max_range).astype(np.int16)
        yield (sampling_rate, new_audio)


demo = gr.Interface(
    fn=generate_audio,
    inputs=[
        gr.Text(label="Prompt", value="80s pop track with synth and instrumentals"),
        gr.Slider(10, 30, value=15, step=5, label="Audio length in seconds"),
        gr.Slider(0.5, 2.5, value=0.5, step=0.5, label="Streaming interval in seconds", info="Lower = shorter chunks, lower latency, more codec steps"),
        gr.Slider(0, 10, value=5, step=1, label="Seed for random generations"),
    ],
    outputs=[
        gr.Audio(label="Generated Music", streaming=True, autoplay=True)
    ],
    #examples = [
    #    ["Country acoustic guitar fast line dance singer like  Kenny Chesney and Garth brooks and Luke Combs and Chris Stapleton. bpm: 100", 30, 0.5, 5],
    #    ["Electronic Dance track with pulsating bass and high energy synths. bpm: 126", 30, 0.5, 5],
    #    ["Rap Beats with deep bass and snappy snares. bpm: 80", 30, 0.5, 5],
    #    ["Lo-Fi track with smooth beats and chill vibes. bpm: 100", 30, 0.5, 5],
    #    ["Global Groove track with international instruments and dance rhythms. bpm: 128", 30, 0.5, 5],
    #    ["Relaxing Meditation music with ambient pads and soothing melodies. bpm: 80", 30, 0.5, 5],
    #    ["Rave Dance track with hard-hitting beats and euphoric synths. bpm: 128", 30, 0.5, 5]
    #],

    examples = [
    ["🧘 Yoga, pilates, and other low-intensity activities. bpm: 60-90", 30, 0.5, 5],
    ["🌟 Power yoga. bpm: 100-140", 30, 0.5, 5],
    ["💪 CrossFit, indoor cycling, or other HIIT forms. bpm: 140-180+", 30, 0.5, 5],
    ["💃 Zumba and dance. bpm: 130-170", 30, 0.5, 5],
    ["🏃 Steady-state cardio, such as jogging. bpm: 120-140", 30, 0.5, 5],
    ["🏃‍♂️ Runners. bpm: 150-190", 30, 0.5, 5],
    ["🚶 Walking or cycling. bpm: 80-110", 30, 0.5, 5],
    ["🏃‍♀️ Long-distance runs. bpm: 120-140", 30, 0.5, 5],
    ["🏃‍♂️ Shorter, more intense runs. bpm: 147-160", 30, 0.5, 5],
    ["🏋️ Weightlifting. bpm: 130-140", 30, 0.5, 5],
    ["🤸 Low impact aerobics. bpm: 133-148", 30, 0.5, 5],
    ["🎸 Ballad / Slow. bpm: 50-85", 30, 0.5, 5],
    ["🎹 Mid-Tempo. bpm: 90-105", 30, 0.5, 5],
    ["🚀 Up-Tempo. bpm: 110-125", 30, 0.5, 5],
    ["🔥 Fast. bpm: 130+", 30, 0.5, 5],
    ["🎵 Blues. bpm: 50+", 30, 0.5, 5],
    ["🎬 Ambient/Movie Score. bpm: 80", 30, 0.5, 5],
    ["🌌 Down Tempo. bpm: 65-95", 30, 0.5, 5],
    ["🌴 Reggae. bpm: 60-90", 30, 0.5, 5],
    ["🎤 Hip-Hop. bpm: 85-110", 30, 0.5, 5],
    ["🎸 Rock. bpm: 90-100", 30, 0.5, 5],
    ["🎸 Alternative Rock. bpm: 120", 30, 0.5, 5],
    ["💖 RnB/Motown. bpm: 75-100", 30, 0.5, 5],
    ["🕺 Dance/House. bpm: 110-130", 30, 0.5, 5],
    ["✨ Trance. bpm: 120-140", 30, 0.5, 5],
    ["🎛️ Techno. bpm: 130-150", 30, 0.5, 5],
    ["🔊 Dubstep. bpm: 130-145", 30, 0.5, 5],
    ["🥁 Drum n' Bass. bpm: 150-170", 30, 0.5, 5],
    ["🤘 Punk Rock. bpm: 140-200", 30, 0.5, 5],
    ["🌾 Bluegrass. bpm: 120-240", 30, 0.5, 5]
    ],

    
    title=title,
    description=description,
    article=article,
    cache_examples=False,
)

demo.queue().launch()