Meta’s Muse Spark Open now available in Amazon SageMaker JumpStart
Muse Spark Open, Meta's coding-optimized open-weight model, is now available in Amazon SageMaker JumpStart. Deploy it to a SageMaker AI endpoint in your own AWS account using SageMaker Studio or the SageMaker Python SDK, and run your first inference request.
Meta has made Muse Spark Open, its coding-optimized open-weight model, accessible through Amazon SageMaker JumpStart, the machine learning hub for Amazon SageMaker AI. This model was trained using real developer workflows, including multi-file refactors and extensive debugging sessions, and features a context window of 1 million tokens.
To deploy Muse Spark Open on SageMaker AI, users can choose between two variants: BF16 ( Muse-Spark-Open-816B-A42B ) and NVFP4 quantized ( Muse-Spark-Open-816B-A42B-NVFP4 ). The latter variant stores weights in NVIDIA's 4-bit floating point format, requiring less memory than the standard BF16 variant. Muse Spark Open was designed to excel in long-running coding tasks, allowing it to handle entire projects and dependencies without the need to split tasks.
The model can process various input modalities, such as text, images, documents, audio, and audio-video, and returns text output. It is built using a mixture-of-experts (MoE) architecture with 816 billion total parameters, with 42 billion active for each token. Users can deploy Muse Spark Open in their own AWS account using Amazon SageMaker Studio or the SageMaker Python SDK, without managing serving infrastructure.
The NVFP4 variant is recommended for cost-effective deployment, as it requires less GPU memory to hold all 816 billion parameters.
Written by urgent.news from AWS Machine Learning's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
Also reported by 6 other outlets
- Welcome Muse Spark Open: open-weight frontier multimodal mixture-of-experts model from Meta huggingface.co
- Meta and Apple Bring Muse Spark Open AI Model to Mac Studio macrumors.com
- Meta releases Muse Spark Open, saying it is the most capable US open-weight model, with 816B total and 42B active parameters and a 1M-token context window (Meta AI Research) research.meta.ai
- Meta and Apple Bring Muse Spark Open AI Model to Mac Studio apple.slashdot.org
- Meta and Apple team up to make Muse AI models run better on Macs theverge.com
- Apple and Meta team up to make Muse AI model run on Mac computers cnbc.com
