Refusal directions computed from mlabonne/harmful_behaviors and mlabonne/harmless_alpaca
500 Optuna trials optimizing refusal vs. KL divergence
Best trial (Trial 445, 6.5% refusals, KL=0.076) applied via LoRA adapters
LoRA weights merged, then converted to GGUF with llama.cpp

GGUF Details

Format: Q8_0
File size: ~28 GB, near-lossless quality
Converted with: llama.cpp convert_hf_to_gguf.py
Quantized with: llama.cpp llama-quantize

Usage

llama.cpp

./llama-cli -m Muse-Glimmer-30B-Heretic-Abliterated-Q8_0.gguf -p "Your prompt here"

Ollama

Create a Modelfile:

FROM ./Muse-Glimmer-30B-Heretic-Abliterated-Q8_0.gguf

Then:

bash

ollama create muse-glimmer-30b-heretic-q8_0
ollama run muse-glimmer-30b-heretic-q8_0

Hardware Requirements

RAM: ~28 GB, near-lossless quality
VRAM offloading: 12-24 GB recommended

Vision (Multimodal)

This model accepts image input when paired with a vision projector (mmproj). Abliteration only modified the language backbone — the vision encoder is untouched — so the standard Meta projector works directly with this repo.

This repository bundles mmproj-Muse-Glimmer-30B-Q4_K_M.gguf (~1.4 GB), Meta's official vision encoder

projector for Muse Glimmer 30B.

Usage (llama.cpp)

bash

huggingface-cli download mlasli/Muse-Glimmer-30B-Heretic-Abliterated-Q8_0-GGUF \
  --include "Muse-Glimmer-30B-Heretic-Abliterated-Q8_0.gguf" \
  --include "mmproj-Muse-Glimmer-30B-Q4_K_M.gguf" \
  --local-dir ./models

./build/bin/llama-mtmd-cli \
  -m ./models/Muse-Glimmer-30B-Heretic-Abliterated-Q8_0.gguf \
  --mmproj ./models/mmproj-Muse-Glimmer-30B-Q4_K_M.gguf \
  --image photo.png \
  -p "Describe this image."

Ollama note: Ollama does not currently support separate mmproj files for this architecture. For image input, use llama.cpp (llama-mtmd-cli or llama-server --mmproj).

License

Apache 2.0 (same as base model)

Changelog

v1.1.0 — vision (multimodal) support (2026-08-16)

Added mmproj-Muse-Glimmer-30B-Q4_K_M.gguf (~1.4 GB), Meta's official vision encoder + projector, enabling image input via llama.cpp.
The vision tower is untouched by abliteration, so this projector matches the base model (meta-models/Muse-Glimmer-30B).
v1.0.0 was the initial (unversioned) text-only upload.