lmdeploy icon indicating copy to clipboard operation
lmdeploy copied to clipboard
Published 2 months ago • verified publisher icon InternLM

[Feature] Aarch64 Support

Open alexc-l opened this issue 2 months ago • 1 comments

Motivation

Does it support the aarch64 platform with NVIDIA GB10 Grace Blackwell (DGX Spark)?

Related resources

No response

Additional context

No response

alexc-l avatar Nov 23 '25 10:11 alexc-l

We do not have this kind of devices to verify, but you can use the similar way to build LMDeploy in Jetson. You need to use NVIDIA SBSA base image instead of L4t image.

windreamer avatar Nov 23 '25 10:11 windreamer

I tried building from source with TuboMind disabled. But there are some issues with Triton (#8539). Still working on it.

alexc-l avatar Nov 25 '25 08:11 alexc-l

@alexc-l were you able to find a solution for it?

I am also using Nvidia DGX Spark and I am unable to install fast-attention and lmdeploy.

karanl1 avatar Dec 16 '25 05:12 karanl1