Repo

nvfp4-ft-spark

Full fine-tuning and QAT of NVFP4-quantized LLMs on DGX Spark / GB10: NVFP4 in, NVFP4 out, no BF16 original needed

Built with

Pythondgx-sparkfine-tuninggb10llmnvfp4

Comments · 0

No comments yet. Tried it? Tell Niklas what worked, what didn't, or what you'd want next.

See what other members are building and what each project is looking for.