Unsloth says it now runs natively on ARM64 Windows with CUDA on NVIDIA RTX Spark, covering diffusion, multimodal, language, and decision models. The author reports training is 2.7x faster and gives winget or the Unsloth download page as install options.
Unsloth now works with ARM64 + Windows + CUDA on RTX Spark! We made [running, training] X [diffusion, multimodal, language, decision] models all run natively + we optimized it to make it 2.7x faster!
Get it via winget or https://unsloth.ai/download/windows
You can now run and train 120B+ parameter models locally on a laptop! NVIDIA RTX Spark is here. 💚 With Unsloth and RTX Spark, train diffusion, decision, multimodal models ~2.7× faster with optimized kernels and binaries for CUDA on ARM. All local with up to 128GB unified memory
Source: Daniel Han · x.comPublished · added here
