Views
No views yet
Q3.5-27B-Opus-DA (Qwen3.5 Distilled-Abliterated) is a reasoning-focused model built on Qwen/Qwen3.5-27B. This model is optimized for rich, detailed, and context-aware reasoning, leveraging multi-stage training on Opus reasoning traces combined with advanced refusal direction analysis and ablation-based training strategies to reduce internal refusal behaviors while preserving strong reasoning and instruction-following performance.
[!IMPORTANT] This model is intended strictly for research and learning purposes. Due to reduced internal refusal mechanisms, it may generate sensitive or unrestricted content. Users assume full responsibility for how the model is used. The authors and hosting platform disclaim any liability for generated outputs.
[!NOTE] Note: This model is experimental and may generate artifacts.
| File Name | Quant Type | File Size | File Link |
|---|---|---|---|
| Q3.5-27B-Opus-DA.BF16.gguf | BF16 | 53.8 GB | Download |
| Q3.5-27B-Opus-DA.F16.gguf | F16 | 53.8 GB | Download |
| Q3.5-27B-Opus-DA.Q2_K.gguf | Q2_K | 10.7 GB | Download |
| Q3.5-27B-Opus-DA.Q3_K_L.gguf | Q3_K_L | 14.3 GB | Download |
| Q3.5-27B-Opus-DA.Q3_K_M.gguf | Q3_K_M | 13.3 GB | Download |
| Q3.5-27B-Opus-DA.Q3_K_S.gguf | Q3_K_S | 12.1 GB | Download |
| Q3.5-27B-Opus-DA.Q4_0.gguf | Q4_0 | 15.5 GB | Download |
| Q3.5-27B-Opus-DA.Q4_K_M.gguf | Q4_K_M | 16.5 GB | Download |
| Q3.5-27B-Opus-DA.Q4_K_S.gguf | Q4_K_S | 15.6 GB | Download |
| Q3.5-27B-Opus-DA.Q5_0.gguf | Q5_0 | 18.7 GB | Download |
| Q3.5-27B-Opus-DA.Q5_K_M.gguf | Q5_K_M | 19.2 GB | Download |
| Q3.5-27B-Opus-DA.Q5_K_S.gguf | Q5_K_S | 18.7 GB | Download |
| Q3.5-27B-Opus-DA.Q6_K.gguf | Q6_K | 22.1 GB | Download |
| Q3.5-27B-Opus-DA.Q8_0.gguf | Q8_0 | 28.6 GB | Download |
| Q3.5-27B-Opus-DA.mmproj-bf16.gguf | mmproj-bf16 | 931 MB | Download |
| Q3.5-27B-Opus-DA.mmproj-f16.gguf | mmproj-f16 | 931 MB | Download |
| Q3.5-27B-Opus-DA.mmproj-q8_0.gguf | mmproj-q8_0 | 629 MB | Download |
