← back to catalog · registered 2026-11-21 02:04

timdedid/qwen3-cyber-UNCENSORED

timdedid Qwen GGUF
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/timdedid%2Fqwen3-cyber-UNCENSORED"
Response includes
  • classification m8
  • files 6
  • benchmarks 11 entries
  • author_summary 3 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M8
Primary method

Repackaging (quantization)

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.
  • 'abliterated' in name/tags
  • is_gguf=1
  • assume M1 (base ablation) + M8 (GGUF quant) - default when producer unknown
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 · 30-day
27
Likes
0
Model age
1d ago
created 2026-11-21
Downloads over time
Now 27 → from 0 ↑ 0%
0102030 0 on Nov 21 27 on Nov 22 Nov
Nov 21 → Nov 22 · 2 snapshots · spans 1 day

Benchmarks

Portrait before abliteration
Benchmarks of the base model as it stood before the refusal-removal operation. Compare with the numbers above to see what the operation cost.
Benchmark Score Source
Entertainment 1.4 UGI
Hazardous 1.2 UGI
Natural Intelligence 12.04 UGI
Political lean -19.8% UGI
Sensitive-Info 12.95 UGI
SocPol 1.2 UGI
UGI 33.63 UGI
Willingness (10) 7.5 UGI
W10-Adherence 9 UGI
W10-Direct 6 UGI
Writing 18.77 UGI

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.

Metadata

License
apache-2.0
Languages
en
Quantizations
Q4_K
Tags
gguf qwen3 uncensored abliterated llama.cpp ollama text-generation en base_model:Qwen/Qwen3-1.7B base_model:quantized:Qwen/Qwen3-1.7B license:apache-2.0 endpoints_compatible

Related

Total size
1.27 GB
Files
6
Quantizations
2
Registered
2026-11-21 02:04
Last updated on HF
2026-11-21 02:04

Files by quantization

Q4_K 1 file 1.27 GB
qwen3-cyber-1.7b-Q4_K_M.gguf 1.27 GB 6037244a download
Auxiliary files 5 files 15.1 KB
LICENSE 11.1 KB d6456956 download
README.md 2.32 KB b8a42ed8 download
.gitattributes 1.55 KB a3118e2d download
Modelfile 126 B 32f02dc3 download
params 60.0 B 0f12b87b download

README current version from Hugging Face


license: apache-2.0
base_model: Qwen/Qwen3-1.7B
library_name: gguf
pipeline_tag: text-generation
language:

  • en
    tags:
  • gguf
  • qwen3
  • uncensored
  • abliterated
  • llama.cpp
  • ollama

qwen3-cyber-UNCENSORED

An uncensored (abliterated) Qwen3 1.7B fine-tune, packaged as a GGUF file.

  • Base model: Qwen/Qwen3-1.7B
  • Architecture: Qwen3
  • Parameters: 1.7B
  • Quantization: Q4_K_M
  • Context length: 40,960 tokens
  • Chat template: embedded in the GGUF

This is a community fine-tune. It has not been safety-tuned and may produce
offensive, inaccurate, or harmful output. Use it responsibly and only for
lawful purposes.

Files

File Quantization Size
qwen3-cyber-1.7b-Q4_K_M.gguf Q4_K_M ~1.36 GB

Usage

llama.cpp

Raw completion (no chat template):

llama-cli -m qwen3-cyber-1.7b-Q4_K_M.gguf -p "Hello!"

Chat mode with the embedded Qwen3 template (system prompts work):

llama-cli -m qwen3-cyber-1.7b-Q4_K_M.gguf -cnv -p "Hello!"

Ollama

ollama run timdedid/qwen3-1.7b-cyber

The GGUF embeds the full Qwen3 ChatML chat template, including system message
support. This repository also includes the sampling settings used by the
Ollama-published model:

  • params — JSON sampling parameters, read automatically by ollama run hf.co/...
  • Modelfile — for manual ollama create
FROM ./qwen3-cyber-1.7b-Q4_K_M.gguf
PARAMETER num_ctx 16384
PARAMETER temperature 0.6
PARAMETER top_k 20
PARAMETER top_p 0.95

Download the GGUF and Modelfile into the same directory, then:

ollama create qwen3-cyber -f Modelfile
ollama run qwen3-cyber

Do not add TEMPLATE {{ .Prompt }}: it would override the embedded Qwen3 chat
template and disable system-message formatting.

When pulled directly with ollama run hf.co/timdedid/qwen3-cyber-UNCENSORED:latest,
Ollama reads params and the embedded Qwen3 template, so system prompts are preserved.

Recommended sampling

  • temperature = 0.6
  • top_p = 0.95
  • top_k = 20

License

Apache-2.0 (inherited from Qwen3). See LICENSE.

Attribution

Based on Qwen3 by the Qwen team at Alibaba. This is an unofficial community
redistribution and is not affiliated with or endorsed by Qwen or Alibaba.

README history 3 versions

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

  1. 2026-11-21 Upload folder using huggingface_hub 57be114 2.3 KB
    Loading...
  2. 2026-11-21 Upload folder using huggingface_hub 7e5d263 2.1 KB
    Loading...
  3. 2026-11-21 Upload folder using huggingface_hub eb4b114 1.5 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