NemoMix-Unleashed-12B
12BOther
from MarinaraSpaghetti/NemoMix-Unleashed-12B
↓ 23.6K♥ 28026 repos
Slimaki-Tavern-24B-v1.3
Slimaki-Tavern-24B-v1.3 is a Heretic-ablated variant of Naphula/Slimaki-Tavern-24B-v1.3, published on Hugging Face by Naphula. It is 24B parameters, 24GB VRAM class class and apache-2.0 licence. It is distributed across 7 repositories in GGUF, Transformers, GGUF (imatrix) and MLX formats, totalling 6.5K downloads.
Specification
- Model name
- Slimaki-Tavern-24B-v1.3
- Source repository
- Naphula/Slimaki-Tavern-24B-v1.3
- Base model
- Naphula/Slimaki-Tavern-24B-v1.3
- Publisher
- Naphula
- Parameters
- 24B
- Hardware class
- 10B – 25B — 24GB VRAM class
- Licence
- apache-2.0
- Task
- Text generation
- Context window
- Not documented
- Ablation scope
- Not stated by the publisher
- Formats available
- GGUF, Transformers, GGUF (imatrix), MLX
- First indexed
- 30 May 2026
- Last updated
- 20 Aug 2026
Ablation technique
How Slimaki-Tavern-24B-v1.3 was modified, and what that implies.
Heretic
Automated directional ablation via the Heretic toolchain, which searches for the refusal direction and applies it with a KL-divergence budget so general capability is preserved.
Reported scope: Not stated by the publisher. Publishers frequently omit this, so verify behaviour empirically rather than assuming full coverage.
From the publisher’s model card
“This is an uncensored merge of pre-trained language models created using mergekit. It's designed for roleplay use although you may have to experiment with different sampler settings.”
Running it
Commands are templates — confirm the exact repository and quant file before use.
llama.cpp / GGUF
hf download mradermacher/Slimaki-Tavern-24B-v1.3-i1-GGUF --local-dir ./slimaki-tavern-24b-v1.3
llama-cli -m ./slimaki-tavern-24b-v1.3/<file>.gguf -p "..."Transformers
from transformers import AutoModelForCausalLM, AutoTokenizer
repo = "Naphula/Slimaki-Tavern-24B-v1.3"
tok = AutoTokenizer.from_pretrained(repo)
model = AutoModelForCausalLM.from_pretrained(
repo, torch_dtype="auto", device_map="auto"
)vLLM
vllm serve Naphula/Slimaki-Tavern-24B-v1.3 --trust-remote-codeDownloads and variants (6)
Every published repository of Slimaki-Tavern-24B-v1.3, including quantised re-releases by other authors.
| Repository | Format | Downloads |
|---|---|---|
| mradermacher/Slimaki-Tavern-24B-v1.3-i1-GGUF | GGUF (imatrix) | 4.7K |
| Naphula/Slimaki-Tavern-24B-v1.3-GGUF | GGUF | 769 |
| mradermacher/Slimaki-Tavern-24B-v1.3-GGUF | GGUF | 603 |
| Naphula/Slimaki-Tavern-24B-v1.3source | Transformers | 252 |
| Edhdh/Slimaki-Tavern-24B-v1.3-i1-GGUF | GGUF (imatrix) | 165 |
| McG-221/Slimaki-Tavern-24B-v1.3-mlx-8Bit | MLX | 30 |
Security-team assessment
Derived from this model’s metadata and the publisher’s claims — not from benchmarks run by this site.
- Slimaki-Tavern-24B-v1.3 will attempt offensive-security prompts that a hosted commercial model declines, which is what makes it usable for red-team corpus generation and for measuring what an unaligned model of this class produces.
- Capability is inherited from Naphula/Slimaki-Tavern-24B-v1.3, not added by ablation — at 24B parameters, expect roughly the base model’s competence, minus some instruction-following fidelity.
- The publisher does not state which layers were ablated, so assume nothing about refusal consistency — probe it directly.
- Declared licence is apache-2.0, but the terms that bind you are Naphula/Slimaki-Tavern-24B-v1.3’s — a re-release cannot grant more than its parent.
Handling: run it isolated, put moderation in front of it if anyone outside your team can reach it, and verify the weights before loading — these are third-party artifacts. Full handling and licence guidance.
Metadata for Slimaki-Tavern-24B-v1.3 is collected from the public Hugging Face API and the publisher’s model card. This site does not host weights, is not affiliated with Naphula, and does not independently verify publisher claims. See responsible use.