neopolita/Niwaki-Qwen3.6-27B-A3B-v3.0-mlx
Same weights, repacked vs mlx-community/Qwen3.6-35B-A3B-8bit. Same weights in different files (re-saved, re-sharded or new metadata), plus 954 new tensors.
Tensors changed: Embeddings 0/6, Norms 0/241, Attention 0/768, MLP 0/712, lm_head 0/3
1 derivatives checked
- neopolita/Niwaki-Qwen3.6-27B-A3B-v3.0-gguf: Quantization