bluehawana/Qwen3.8-27B-8bit-MTP-MLX
Same weights, repacked vs mlx-community/Qwen3.8-27B-8bit. Same weights in different files (re-saved, re-sharded or new metadata), plus 31 new tensors.
Tensors changed: Embeddings 0/6, Norms 0/319, Attention 0/1164, MLP 0/688, lm_head 0/3