← back to catalog · registered 2026-08-22 13:56

alexgusevski/L3.2-Rogue-Creative-Instruct-Uncensored-Abliterated-7B-q8-mlx

alexgusevski Llama 7.1B second-order
Your rig guess connected
? Why do I need an app?
Reading your rig…

This is a rough estimate. Install the free app - we'll show exact numbers.

Reading real hardware from your app right now. Numbers below are exact.

Below is the per-quantization compatibility for this model.

curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/alexgusevski%2FL3.2-Rogue-Creative-Instruct-Uncensored-Abliterated-7B-q8-mlx"
Response includes
  • classification m5
  • files 9
  • hub_downloads_all_time 692
  • author_summary 64 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M5
Primary method

Mergekit merge

Applied on top of direct removal inherited from the base model.
Confidence
MEDIUM
Inherited from base model
Why this label 3 signals
Method inferred from partial signals - repository name, related files, or tag patterns. Producer identity not confirmed; label may sharpen or shift as we gather more evidence.
  • merge tag / mergekit / dare-ties in tags or name
  • no unusual architecture pattern (regular merge)
  • abliterated marker present
Refusal direction extraction

No specific extraction method could be identified for this model. The producer either did not document it or used a proprietary pipeline.

What is a refusal direction? →
Downloads · lifetime
692
33 last 30d - cooling
Likes
0
Model age
8mo ago
created 2026-01-15
Downloads over time
Now703→from36↑1,853%
325851477036 on Jan 14703 on Oct 11703 on Oct 9JanMarMayJulSep
Jan 14 → Oct 11 · 78 snapshots · spans 270 days

Genealogy 0 direct forks

Full fork graph →

This model's place in the market. Above: what it was derived from. Below: the tree of everything derived from it.

Variants by this author 6 formats · 212 downloads combined

The same weights this author released in different packaging. Pick the format that matches your runtime.

Metadata

Tags
mlx safetensors llama mergekit merge text-generation conversational base_model:DavidAU/L3.2-Rogue-Creative-Instruct-Uncensored-Abliterated-7B base_model:quantized:DavidAU/L3.2-Rogue-Creative-Instruct-Uncensored-Abliterated-7B 8-bit region:us

Related

Total size
7.06 GB
Files
9
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-01-15 14:05

Files by quantization

Auxiliary files 9 files 7.08 GB
model-00001-of-00002.safetensors 5.00 GB 6b210be5 download
model-00002-of-00002.safetensors 2.07 GB 970f0903 download
tokenizer.json 16.4 MB 65ff5472 download
model.safetensors.index.json 129 KB 1ea80351 download
chat_template.jinja 3.74 KB 1bad6a0f download
.gitattributes 1.53 KB 52373fe2 download
config.json 1.14 KB 56ab51ef download
README.md 1.12 KB 54d2159c download
tokenizer_config.json 465 B cfe021b2 download

README current version from Hugging Face


library_name: mlx
tags:

  • mergekit
  • merge
  • mlx
    base_model: DavidAU/L3.2-Rogue-Creative-Instruct-Uncensored-Abliterated-7B
    pipeline_tag: text-generation

alexgusevski/L3.2-Rogue-Creative-Instruct-Uncensored-Abliterated-7B-q8-mlx

This model alexgusevski/L3.2-Rogue-Creative-Instruct-Uncensored-Abliterated-7B-q8-mlx was
converted to MLX format from DavidAU/L3.2-Rogue-Creative-Instruct-Uncensored-Abliterated-7B
using mlx-lm version 0.30.2.

Use with mlx

pip install mlx-lm
from mlx_lm import load, generate

model, tokenizer = load("alexgusevski/L3.2-Rogue-Creative-Instruct-Uncensored-Abliterated-7B-q8-mlx")

prompt = "hello"

if tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, add_generation_prompt=True, return_dict=False,
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)

README history 1 version

The author's README evolved over time. Click a version to see its content at that point.

  1. 2026-01-15Add files using upload-large-folder tool0b268eb1.1 KB
    Loading...
Catalog is the map. Apps are the tools.

Run models on your own machine, not in the cloud.

Every model page has an "Open in Abliteration" button that hands the model directly to the first-party desktop client, at the quantization your rig can actually run. No API keys, no subscription, no prompt leakage.

Open in Abliteration