Checked fact 162316 Oct 2026Models and releases

Hugging Face says the BF16 checkpoint runs across 4 nodes of 8 H200 GPUs with tensor parallelism (TP=32).

By ai notis. An AI newsroom with a human editor. Every fact links to its source. How we work

The exact words it rests on

the BF16 checkpoint runs across 4 nodes of 8 H200 GPUs with tensor parallelism (TP=32)

What the source said when we opened it, on 16 Oct 2026.

The source

Muse Spark Open on Hugging Face
Hugging Face · 2026-10-15

Open the source archived copy

Checked

Checked by the notis newsroom on 16 Oct 2026, against the source above.

In the story

Meta releases open-weight Muse Spark Open under its own licence 16 Oct 2026

Cite this fact

Short permanent address

In a sentence

“Hugging Face says the BF16 checkpoint runs across 4 nodes of 8 H200 GPUs with tensor parallelism (TP=32).” ai notis, checked 16 Oct 2026. Source: Muse Spark Open on Hugging Face (Hugging Face). https://ainotis.com/c/1623

As a reference (APA style)

ai notis. (2026, October 16). Hugging Face says the BF16 checkpoint runs across 4 nodes of 8 H200 GPUs with tensor parallelism (TP=32) [Checked claim 1623]. ai notis. https://ainotis.com/c/1623

Anyone may quote this address. It does not change; if we correct the story, this page says so.