Issues / #2716

#2716 [Feature] Request support for the NEW NVIDIA NVFP4 checkpoint

open · @Pragmatism0220 · 1 comments · View on GitHub

Models & quants

Description

Nvidia just released a new quantization checkpoint (https://huggingface.co/nvidia/Qwen3.8-Flash-Next-FP4-Local) that achieves nearly lossless performance using a new lossless compression format. It would be highly beneficial if Strata could manage to support it. Thank you! 🙏

Related on this site

Quick links to related pages.