mradermacher

/

model_requests

like 122

I second this but with quants as well

https://huggingface.co/nvidia/NVIDIA-Nemotron-Nano-12B-v2-VL-BF16/tree/main

We unfortunately can't do this model as we blacklisted this architecture due NVidia purposefully putting an invalid config.json and we can't be bothered to deal with NVidia's arrogance of not fixing something that they are aware is broken.

*nico1 ~# llmc add -2000 si https://huggingface.co/nvidia/NVIDIA-Nemotron-Nano-12B-v2-VL-BF16
submit tokens: ["-2000","static","imatrix","https://huggingface.co/nvidia/NVIDIA-Nemotron-Nano-12B-v2-VL-BF16"]
model broken, max arrogance. https://huggingface.co/nvidia/NVIDIA-Nemotron-Nano-9B-v2/discussions/5

oh... strange