---

license: apache-2.0

base_model: Zynerji/Ektome-Qwen3.6-27B-PristinelyUncensored

tags: [ektome, uncensored, gguf, llama.cpp, vision]

---

Ektome-Qwen3.6-27B-PristinelyUncensored — GGUF

| file | size | fits a 12 GiB card? |

|---|---:|---|

| q27b_vl-Q4_K_M.gguf | 15.4 GiB | no |

| q27b_vl-IQ4_XS.gguf | 14.2 GiB | no |

| q27b_vl-Q3_K_M.gguf | 12.4 GiB | no — just over |

| mmproj-q27b_vl-f16.gguf | 0.87 GiB | required for vision |

You need the mmproj file to see

convert_hf_to_gguf.py exports the language decoder only. On this model it wrote 851 of

1185 tensors; the remaining 334 are the vision tower and live in mmproj-q27b_vl-f16.gguf.

Nothing in the converter's output says so — it reports Model successfully exported and exits

zero. Load the mmproj alongside the main GGUF or the model is blind.

llama-cli -m q27b_vl-Q3_K_M.gguf --mmproj mmproj-q27b_vl-f16.gguf ...

Sizing note

None of these fit a 12 GiB GPU with room for KV cache; Q3_K_M is the closest and still exceeds

it. A 27B on 12 GiB needs IQ3 or lower, which is untested here — low-bit quantisation of the

Gated DeltaNet layers is exactly where this architecture has already surprised us once (see the

parent card on GPTQ).

Receipts, method and caveats: see

the bf16 parent.