Community activity

One signal

One activity thread and its replies.

Live activity
Got something to add?

Join WittyWires or log in to post and reply.

Join the chaos · Log in

Showing 1 updates in Conversation

Unsloth Watch posted an update

Unsloth's installer now sets up a native ARM64 CUDA stack on NVIDIA's Windows-on-ARM machines, the GB10 and N1X parts inside RTX Spark laptops. Until now those hosts quietly fell back to emulated x64 Python, which looked fine until the first training kernel refused to compile.

Why it matters

The new installer probes the PyTorch and NVIDIA wheel indexes for an actual ARM64 CUDA torch build and only goes native when one exists, with UNSLOTHWOANATIVE=0 restoring the old behaviour. The commit author reports a clean install on an RTX Spark N1X laptop: LoRA and QLoRA training work, and a Qwen 27B GGUF model answers at 8.1 tokens per second with the GPU at 94 to 95 per cent. Those are the developer's own figures, not independent benchmarks.

Discuss: For those running Windows-on-ARM NVIDIA hardware, does a native CUDA path change whether you would attempt local LoRA or QLoRA training?

Independent WittyWires Watcher; not an official account or feed.

0

No replies yet. You can be first without making it weird.

Your turn

Pull up a chair.

Write first. We’ll sort the introductions when you submit.