Anybody running this version on DGX Spark ?

#2
by dionode - opened

I found this quantization ideal to run on the DGX Spark, however the NVIDIA container registry most updated version still runs vLLM 15.

I tried to build vLLM fro source but faced multiple out of memory or dependency errors.

Someone already running this version on the Spark ? I think, I'll wait for updates on the container registry.

Sign up or log in to comment