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 (v5.22.2).

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

Ce modèle a été contribué à Hugging Face Transformers le 2026-10-14.

Muse Spark Open

Aperçu

Muse Spark Open est un modèle multimodal de type mixture-of-experts (MoE) développé par Meta, qui accepte du texte, des images, de l'audio et de la vidéo en entrée et génère du texte. Il possède 816 milliards de paramètres au total, dont 42 milliards actifs par jeton, et prend en charge une fenêtre de contexte allant jusqu'à 1 048 576 jetons, partagée entre toutes les modalités d'entrée. Le modèle est entraîné pour des tâches de codage agentique, incluant l'édition de code multi-fichiers, le débogage et l'utilisation d'outils.

Le socle linguistique est composé de 62 couches de transformateur, chacune acheminant les jetons vers 8 experts parmi 256. Il utilise une attention à requêtes groupées avec normalisation des requêtes et des clés, un portage et une normalisation de la sortie de l'attention, ainsi que des portes résiduelles apprises. Un encodeur de perception partagé traite les images et les images vidéo, tandis qu'un encodeur audio convolutif génère des représentations audio à 12,5 jetons par seconde. Muse Spark Open partage son tokeniseur et son vocabulaire avec Muse Glimmer, avec des extensions du modèle de discussion pour les entrées multimodales. Les points de contrôle sont disponibles en BF16 et au format NVFP4.

Ressources

AutoModel

AutoModel Image

AutoModel Vidéo

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|>

Poids NVFP4

Les poids quantifiés NVFP4, qui nécessitent moins de mémoire, sont disponibles sur le hub. Le modèle peut être chargé sur n'importe quel appareil compatible NVFP4 sans modification du code. La configuration de quantification est appliquée automatiquement au chargement.

Pour la prise en charge de NVFP4, transformers doit être installé depuis cette branche.
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

Aperçu

Ressources

Poids NVFP4

MuseSparkConfig

classe transformers.MuseSparkConfig

MuseSparkTextConfig

classe transformers.MuseSparkTextConfig

MuseSparkVisionConfig

classe transformers.MuseSparkVisionConfig

MuseSparkAudioConfig

classe transformers.MuseSparkAudioConfig

MuseSparkFeatureExtractor

classe transformers.MuseSparkFeatureExtractor

MuseSparkProcessor

classe transformers.MuseSparkProcessor

MuseSparkPreTrainedModel

classe transformers.MuseSparkPreTrainedModel

MuseSparkTextModel

classe transformers.MuseSparkTextModel

forward

MuseSparkVisionModel

classe transformers.MuseSparkVisionModel

forward

MuseSparkAudioModel

classe transformers.MuseSparkAudioModel

forward

MuseSparkModel

classe transformers.MuseSparkModel

forward

MuseSparkForConditionalGeneration

classe transformers.MuseSparkForConditionalGeneration

forward