mlx-community/Qwen2-VL-2B-mlx
Same weights, repacked vs Qwen/Qwen2-VL-2B. Same weights in different files (re-saved, re-sharded or new metadata), plus 338 new tensors.
Tensors changed: Attention 0/128, MLP 0/132, Norms 0/130
Same weights, repacked vs Qwen/Qwen2-VL-2B. Same weights in different files (re-saved, re-sharded or new metadata), plus 338 new tensors.
Tensors changed: Attention 0/128, MLP 0/132, Norms 0/130