OpenMed/North-Micro-Vision-Instruct-bf16-mlx
Same weights, repacked vs CohereLabs/North-Micro-Vision-Instruct. Same weights in different files (re-saved, re-sharded or new metadata), plus 226 new tensors.
Tensors changed: Attention 0/108, MLP 0/124, Norms 0/116, Embeddings 0/2
1 other repos hold byte-for-byte identical weight files.