0xA50C1A1/Muse-Glimmer-30B-abliterix-gguf overview
GGUF quantizations of 0xA50C1A1/Muse Glimmer 30B abliterix https://huggingface.co/0xA50C1A1/Muse Glimmer 30B abliterix . Warning As of September 13, 2026, this…
Runs locally from ~1.91 GB disk (4 GB VRAM class GPUs with llama.cpp / guIDE).
Repository Files & Downloads
Model Details
Model README
---
license: apache-2.0
base_model: 0xA50C1A1/Muse-Glimmer-30B-abliterix
tags:
- abliterated
- uncensored
- muse_glimmer
- gguf
- llama-cpp
- nvfp4
---
GGUF quantizations of 0xA50C1A1/Muse-Glimmer-30B-abliterix.
> [!Warning]
> As of September 13, 2026, this NVFP4 GGUF may not work out of the box in llama.cpp. Track issue: https://github.com/ggml-org/llama.cpp/issues/27178
> [!Note]
> lm_head and embeddings are kept in Q8_0 instead of BF16 to reduce VRAM usage.
| Link | Type | Size |
|:-----|:-----|--------:|
| GGUF | Vision projector (Q8_0) | 2.1 GB |
| GGUF | NVFP4 | 17 GB |
Run 0xA50C1A1/Muse-Glimmer-30B-abliterix-gguf with guIDE
Download guIDE — the AI-native code editor with local LLM inference and 69 built-in tools.
Source: Hugging Face · Compare models