Transformers documentation

Muse Spark Open

Transformers

You are viewing main version, which requires installation from source. If you'd like regular pip install, checkout the latest stable version (v5.22.1).

Join the Hugging Face community

and get access to the augmented documentation experience

Collaborate on models, datasets and Spaces

Faster examples with accelerated inference

Switch between documentation themes

to get started

Эта модель была добавлена в Hugging Face Transformers 14.10.2026.

Muse Spark Open

Обзор

Muse Spark Open — это мультимодальная модель с архитектурой mixture-of-experts (MoE), разработанная Meta. Она принимает на вход текст, изображения, аудио и видео и генерирует текст. Модель содержит 816 миллиардов параметров в общей сложности, из которых 42 миллиарда активны на каждый токен, и поддерживает окно контекста до 1 048 576 токенов, распределенных между всеми входными модальностями. Модель обучена для агентных задач программирования, включая редактирование кода в нескольких файлах, отладку и использование инструментов.

Языковая основа состоит из 62 трансформерных слоев, каждый из которых маршрутизирует токены к 8 из 256 экспертов. Используется групповое запросно-ключевое внимание с нормализацией запросов и ключей, гейтинг и нормализация выходного внимания, а также обучаемые остаточные гейты. Общий перцептивный энкодер обрабатывает изображения и кадры видео, тогда как сверточный аудиоэнкодер формирует аудиопредставления со скоростью 12,5 токенов в секунду. Muse Spark Open разделяет токенизатор и словарь с Muse Glimmer, с расширениями шаблона чата для мультимодальных входов. Чекпоинты доступны в форматах BF16 и NVFP4.

Ресурсы

AutoModel

AutoModel Image

AutoModel Video

AutoModel Audio

import os

from transformers import AutoModelForMultimodalLM, AutoProcessor, DistributedConfig

distributed_config = DistributedConfig(
    tp_size=int(os.environ["WORLD_SIZE"]), # Number of devices across which the dense layers are sharded
    ep_size=int(os.environ["WORLD_SIZE"]), # Number of devices across which the experts are distributed
)

processor = AutoProcessor.from_pretrained("meta-models/Muse-Spark-Open-816B-A42B")
model = AutoModelForMultimodalLM.from_pretrained(
    "meta-models/Muse-Spark-Open-816B-A42B",
    distributed_config=distributed_config,
)

messages = [
    {"role": "user", "content": "Explain in one sentence why the sky is usually blue."},
]
inputs = processor.apply_chat_template(
    messages,
    reasoning_effort="low",  # default: xhigh, options: low/medium/high/xhigh
    add_generation_prompt=True,
    tokenize=True,
    return_dict=True,
    return_tensors="pt",
).to(model.device).to(model.dtype)
input_len = inputs["input_ids"].shape[-1]

outputs = model.generate(**inputs, max_new_tokens=512)
print(processor.decode(outputs[0][input_len:], skip_special_tokens=False))
#  to=self<|message|>Explain in one sentence why the sky is usually blue.
# ...
# <|eom|><|start|>assistant to=user<|message|>The sky is usually blue because molecules in Earth's atmosphere scatter shorter-wavelength blue sunlight in all directions much more strongly than longer-wavelength red light.<|eot|>

import os

from transformers import AutoModelForMultimodalLM, AutoProcessor, DistributedConfig

distributed_config = DistributedConfig(
    tp_size=int(os.environ["WORLD_SIZE"]),
    ep_size=int(os.environ["WORLD_SIZE"]),
)

processor = AutoProcessor.from_pretrained("meta-models/Muse-Spark-Open-816B-A42B")
model = AutoModelForMultimodalLM.from_pretrained(
    "meta-models/Muse-Spark-Open-816B-A42B",
    distributed_config=distributed_config,
)

messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/pipeline-cat-chonk.jpeg"},
            {"type": "text", "text": "What is shown in this image?"},
        ],
    },
]
inputs = processor.apply_chat_template(
    messages,
    reasoning_effort="medium", # default: xhigh, options: low/medium/high/xhigh
    add_generation_prompt=True,
    tokenize=True,
    return_dict=True,
    return_tensors="pt",
).to(model.device).to(model.dtype)
input_len = inputs["input_ids"].shape[-1]

outputs = model.generate(**inputs, max_new_tokens=512)
response = processor.decode(outputs[0][input_len:], skip_special_tokens=False)
print(response)
#  to=self<|message|>What is shown in this image?
# ...
# <|eom|><|start|>assistant to=user<|message|>The image shows a **Pallas's cat — also known as a manul, *Otocolobus manul*** — walking through snow.
#
# It is a small, stocky wild cat recognizable by:
#
# * **Its extremely dense, long, greyish-brown winter coat** with faint dark stripes and a dusting of snow on its back
# * **Its flat, broad face** with low-set ears, a short muzzle, and dark stripes on the cheeks
# * **Its short legs and thick, plump body** adapted for cold climates
#
# The cat is captured mid-stride on a snow-covered surface, with one front paw lifted. In the background are white birch tree trunks with black markings and a chain-link fence on the left, indicating the animal is in a zoo or wildlife enclosure in winter.<|eot|>

import os

from transformers import AutoModelForMultimodalLM, AutoProcessor, DistributedConfig

distributed_config = DistributedConfig(
    tp_size=int(os.environ["WORLD_SIZE"]),
    ep_size=int(os.environ["WORLD_SIZE"]),
)

processor = AutoProcessor.from_pretrained("meta-models/Muse-Spark-Open-816B-A42B")
model = AutoModelForMultimodalLM.from_pretrained(
    "meta-models/Muse-Spark-Open-816B-A42B",
    distributed_config=distributed_config,
)

messages = [
    {
        "role": "user",
        "content": [
            {"type": "video", "url": "https://huggingface.co/datasets/hf-internal-testing/fixtures_videos/resolve/main/tennis.mp4"},
            {"type": "text", "text": "What is happening in this video?"},
        ],
    },
]
inputs = processor.apply_chat_template(
    messages,
    reasoning_effort="medium",  # default: xhigh, options: low/medium/high/xhigh
    processor_kwargs={
        "fps": 2,   # default: 2, number of frames sampled per second
        "num_frames": 8, # default: 94, maximum number of frames sampled from the video
    },
    add_generation_prompt=True,
    tokenize=True,
    return_dict=True,
    return_tensors="pt",
).to(model.device).to(model.dtype)
input_len = inputs["input_ids"].shape[-1]

outputs = model.generate(**inputs, max_new_tokens=512)
print(processor.decode(outputs[0][input_len:], skip_special_tokens=False))
#  to=self<|message|>What is happening in this video?
# ...
# <|eom|><|start|>assistant to=user<|message|>In this video two people are playing tennis on an indoor court.
#
# The shot is taken from behind one of the players - a man in a white t-shirt, black shorts and white shoes and socks. He is standing near the baseline on an orange hard court with white lines, in a ready position with his knees bent and his racket held in front of him.
#
# Across the net, on the far side of the court, is his opponent, also dressed in light-colored clothing.
#
# The sequence shows a short rally:
#
# *   The opponent hits the ball over the net.
# *   The man in the foreground shuffles to his left, turns his body, and prepares for a shot.
# *   He executes a two-handed backhand, swinging the racket across his body and following through as he hits the ball back over the net.
#
# The setting is a large indoor tennis hall with a high, arched ceiling with exposed wooden beams and overhead fluorescent lights. The court is surrounded by a grey run-off area, green windscreening on the walls, and netting that separates it from adjacent courts. On the right side you can see an umpire's chair and a few yellow and black benches. An "Adobe Stock" watermark is visible in the center of the frame.<|eot|>

Веса NVFP4

NVFP4 квантованные веса с меньшими требованиями к памяти доступны на хаб. Модель можно загрузить на любое устройство, совместимое с NVFP4, без изменений в коде. Конфигурация квантования применяется автоматически при загрузке.

Для поддержки NVFP4 библиотека transformers должна быть установлена из этой ветки.
pip install git+https://github.com/huggingface/transformers.git@finegrained-per-scheme-demo

import os

from transformers import AutoModelForMultimodalLM, AutoProcessor, DistributedConfig

distributed_config = DistributedConfig(
    tp_size=int(os.environ["WORLD_SIZE"]),
    ep_size=int(os.environ["WORLD_SIZE"]),
)

processor = AutoProcessor.from_pretrained("meta-models/Muse-Spark-Open-816B-A42B-NVFP4")
model = AutoModelForMultimodalLM.from_pretrained(
    "meta-models/Muse-Spark-Open-816B-A42B-NVFP4",
    distributed_config=distributed_config,
)

# ...

Transformers

Muse Spark Open

Обзор

Ресурсы

Веса NVFP4

MuseSparkConfig

класс transformers.MuseSparkConfig

MuseSparkTextConfig

класс transformers.MuseSparkTextConfig

MuseSparkVisionConfig

класс transformers.MuseSparkVisionConfig

MuseSparkAudioConfig

класс transformers.MuseSparkAudioConfig

MuseSparkFeatureExtractor

класс transformers.MuseSparkFeatureExtractor

MuseSparkProcessor

класс transformers.MuseSparkProcessor

MuseSparkPreTrainedModel

класс transformers.MuseSparkPreTrainedModel

MuseSparkTextModel

класс transformers.MuseSparkTextModel

forward

MuseSparkVisionModel

класс transformers.MuseSparkVisionModel

forward

MuseSparkAudioModel

класс transformers.MuseSparkAudioModel

forward

MuseSparkModel

класс transformers.MuseSparkModel

forward

MuseSparkForConditionalGeneration

класс transformers.MuseSparkForConditionalGeneration

forward