JANGQ-AI/Nemotron-3-Super-120B-A12B-JANG_4M
Different architecture vs nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8. Different architecture: only 375 of 124941/1593 tensors match by name and shape.
Different architecture vs nvidia/NVIDIA-Nemotron-3-Super-120B-A12B-FP8. Different architecture: only 375 of 124941/1593 tensors match by name and shape.