JoaoZaokk/Qwen-Image-Edit-2511-W4A8-ConvRot
Same weights, repacked vs Qwen/Qwen-Image-Edit-2511. Same weights in different files (re-saved, re-sharded or new metadata), plus 2521 new tensors.
Tensors changed: other 0/84, Embeddings 0/4, Attention 0/325, Norms 0/166, MLP 0/160