Meta 在 Hugging Face 上开源了 Muse Spark Open 的权重，这是一个 816B 参数、面向编码与智能体任务的模型。开放版本与 Muse Spark 1.3 共用同一基座模型，额外做了安全后训练，以降低脱离系统护栏单独使用时的滥用风险，逆转了 Meta 今年 4 月首个闭源前沿模型的策略。
同一新闻，精选展示《Meta 开源 Muse Spark Open：816B 多模态 MoE 模型》
Meta released the weights of Muse Spark Open, on Huggingface, an 816B-parameter model built for coding and agentic work.
The open version shares the same base model as Muse Spark 1.3, with additional safety post-training to resist misuse as a standalone model without system guardrails, reversing Meta’s first proprietary, closed-weight frontier model strategy from April.
Each of its 62 layers holds 256 expert networks and routes every token through 8, so deployment needs the memory of an 816B-parameter model but runs at the speed of a 42B model.
That storage bill explains the Apple tie-up, because the M5 Ultra pushes 1.2 TB/s of unified memory and a Mac Studio can carry up to 512GB.
A 3-bit quant tested on a 512GB M3 Ultra took about 344 GB and ran at about 21 tok/s generation on a short prompt.
Muse Spark Open -- weights releasing today -- runs great on new Mac Studios thanks to work with @johnternus and Apple.在 X 查看被引用的帖子
来源：Rohan Paul · x.com