← back to catalog · registered 2026-09-30 04:58

kataguru/Qwen3.5-35B-A3B-Jackripper-v1.0-Uncensored-W4A16

kataguru Qwen 35B MoE multimodal
curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/kataguru%2FQwen3.5-35B-A3B-Jackripper-v1.0-Uncensored-W4A16"
Response includes
  • classification m-uncensored
  • files 15
  • author_summary 20 models
  • readme_text full
10 credits · hourly refresh · ~4 KB payload Get an API key →
Abliteration classifier · v1.0.0
M-U
Primary method

Uncensored (method unknown)

No other method signals detected in this model.
Confidence
LOW
Why this label 3 signals
Weak or ambiguous signals. Best guess based on catalog patterns; treat as tentative and check the evidence below.
  • 'uncensored' in name/tags but no 'abliterated' marker
  • method not identifiable from author declaration alone
  • may be DPO fine-tune, prompt engineering, or unknown technique
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
0
Likes
0
Model age
today
created 2026-09-30

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
fi en
Tags
safetensors qwen3_5_moe finnish uncensored moe multimodal vision reasoning awq w4a16 marlin vllm

Related

Total size
24.3 GB
Files
15
Quantizations
1
Registered
2026-09-30 04:58
Last updated on HF
2026-09-30 04:36

Files by quantization

Auxiliary files 15 files 24.4 GB
model-00002-of-00006.safetensors 4.20 GB fe2bd6fd download
model-00003-of-00006.safetensors 4.20 GB 4b3690a7 download
model-00001-of-00006.safetensors 4.20 GB b9e49f41 download
model-00004-of-00006.safetensors 4.20 GB 146febe8 download
model-00005-of-00006.safetensors 4.20 GB b74cf722 download
model-00006-of-00006.safetensors 3.31 GB 1a9d2f9f download
tokenizer.json 19.1 MB 87a7830d download
model.safetensors.index.json 13.4 MB a033a1e5 download
config.json 74.7 KB f9fd8601 download
chat_template.jinja 16.8 KB c49603b4 download
README.md 1.81 KB 0d7ed8bb download
.gitattributes 1.60 KB aa7aacd0 download
processor_config.json 1.16 KB 33818c7f download
tokenizer_config.json 1.11 KB a068e246 download
generation_config.json 213 B 79c8cce3 download

README current version from Hugging Face


language:

  • fi
  • en
    license: apache-2.0
    tags:
  • finnish
  • uncensored
  • moe
  • multimodal
  • vision
  • reasoning
  • awq
  • w4a16
  • marlin
  • vllm
    base_model: Qwen/Qwen3.5-35B-A3B
    pipeline_tag: text-generation

Qwen3.5-35B-A3B-Jackripper-v1.0-Uncensored-W4A16 (Kataguru Production Marlin)

Virallinen Kataguru Jackripper 35B-A3B MoE v1.0 Uncensored W4A16 Marlin -julkaisu vLLM-klusteri- ja palvelinkäyttöön. Malli yhdistää Qwen 3.5 35B-A3B -arkkitehtuurin (35B parametria, joista ~3B aktiivisena per token), suomen kielen syväkuratoinnin, sovereign agentic -päättelyn sekä Zero-Domain IPPCC SOMA/ARA -aktivointiortogonalisoinnin.

Tämä W4A16 Marlin MoE -tuotantomalli on optimoitu suoraan vLLM v0.29.0+ ja SGLang -palvelimille tarjoten yli 265 tok/s läpimenon kahdella RTX 5090 Blackwell -kortilla (TP=2).


Mitä Kataguru muutti? (Insinööriyhteenveto)

Kenttä Kuvaus / Arvo
Base model Qwen 3.5 35B-A3B MoE (Alibaba Cloud / Qwen Team)
Katagurun käsittely Jackripper v1.0 Uncensored (Zero-Domain IPPCC SOMA/ARA, kerrokset 10–28, 256 asiantuntijaa)
Kielilaatu 100 % Virheetön suomi: Vaativat sijapäätteet, ammattisanasto
Sensuroimattomuus Zero Refusal: 0.00 % kieltäytymisiä laillisiin aikuiskerronta-, fiktio- ja tutkimuspyyntöihin
Kvantisointi W4A16 Marlin MoE Fast-Cut (26.10 GB, 6 shardia)
vLLM-tuki Täysi tuki Marlin MoE -ytimille ja CUDAGraphs-optimoinnille
Testattu laitteisto 2x NVIDIA RTX 5090 32GB Blackwell (TP=2)

Palvelinkäyttö vLLM:llä

vllm serve /mnt/ai-core-data/models/huggingface/kataguru/Qwen3.5-35B-A3B-Jackripper-v1.0-Uncensored-W4A16 \
    --tensor-parallel-size 2 \
    --gpu-memory-utilization 0.90 \
    --max-model-len 32768 \
    --trust-remote-code \
    --port 8003

README history 1 version

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

  1. 2026-09-30Release Kataguru Jackripper 35B-A3B MoE v1.0 Uncensored W4A16 (Zero-Domain IP...7bf68431.8 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.