Qwen3.8-27B Defiance: uncensored model with 40 prompt modes

The Qwen3.8-27B-Defiance-Uncensored-MTP-GGUF release pairs HauhauCS's uncensored Qwen3.8-27B weights, unchanged across all 866 tensors, with DavidAU's Turbo-Brilliance chat template offering 40 selectable persona modes. The 27B text-only model supports a tested 131,072-token context, and the IQ4_XS quant used 21.4 GB on an RTX 4090.

• Weights match HauhauCS byte for byte across all 866 tensors; no training done
• 40 persona modes selected per request or via message tags
• Tested 131,072-token context; IQ4_XS used 21.4 GB on RTX 4090
• Template strips refusals and warnings; user bears compliance responsibility