← back to catalog · registered 2026-09-27 05:57

tangles/Qwen3.8-27B-Uncensored-NVFP4-1Cat

tangles 17B multimodal 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/tangles%2FQwen3.8-27B-Uncensored-NVFP4-1Cat"
Response includes
  • classification m-uncensored
  • files 17
  • hub_downloads_all_time 69
  • author_summary 2 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 · lifetime
69
Likes
2
Model age
8w ago
created 2026-09-27
Downloads over time
Now 340 → from 0 ↑ 0%
0125249374 0 on Sep 27 340 on Nov 22 340 on Nov 21 SepOctNov
Sep 27 → Nov 22 · 57 snapshots · spans 56 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.

Metadata

License
apache-2.0
Tags
transformers safetensors qwen3_5 image-text-to-text Qwen3.8 nvfp4 fp4 w4a4 compressed-tensors 1cat-vllm multimodal mtp

Related

Total size
19.1 GB
Files
17
Quantizations
1
Registered
2026-09-27 05:57
Last updated on HF
2026-09-27 11:03

Files by quantization

Auxiliary files 17 files 19.2 GB
model-00001-of-00004.safetensors 4.66 GB 726ae4cf download
model-00003-of-00004.safetensors 4.63 GB 91398bb0 download
model-00002-of-00004.safetensors 4.62 GB 9d32b8d6 download
model-00004-of-00004.safetensors 4.45 GB 8f905ed8 download
model_mtp.safetensors 810 MB 7a9fd4ee download
tokenizer.json 19.1 MB 06b95093 download
model.safetensors.index.json 272 KB f1d9127b download
chat_template.jinja 8.74 KB c0c686f9 download
config.json 5.16 KB 5713df7f download
.gitattributes 1.53 KB 52373fe2 download
README.md 1.30 KB d518981a download
processor_config.json 1.19 KB 43c4343e download
tokenizer_config.json 1.14 KB 1d134cd2 download
quantization-audit.json 710 B d83a9438 download
calibration.json 334 B 4888af77 download
recipe.yaml 280 B 993fed19 download
generation_config.json 214 B c53835dc download

README current version from Hugging Face


library_name: transformers
pipeline_tag: image-text-to-text
license: apache-2.0
base_model: orcarouter/Qwen3.8-27B-Uncensored
base_model_relation: quantized
tags:

  • Qwen3.8
  • nvfp4
  • fp4
  • w4a4
  • compressed-tensors
  • 1cat-vllm
  • multimodal
  • mtp

Qwen3.8-27B-Uncensored NVFP4 for 1Cat-vLLM

An all-transformer-linear NVFP4 W4A4 PTQ build of
orcarouter/Qwen3.8-27B-Uncensored,
exported in the compressed-tensors nvfp4-pack-quantized layout used by
QUASAR-QAT/Qwen3.8-27B-QUASAR-NVFP4
and supported by 1CatAI/1Cat-vLLM.

  • 496/496 transformer linears: NVFP4 W4A4, group size 16
  • lm_head, token embeddings, vision tower, and MTP tensors: BF16
  • Calibration: 256 UltraChat samples, maximum sequence length 2,048
  • LLM Compressor revision: c8676edd2f6cddd0be2427f034825ca0d476fa72

This is post-training quantization, not QUASAR quantization-aware training.
It matches the runtime/checkpoint layout but does not claim QUASAR's benchmark
quality. See quantization-audit.json and calibration.json.

The source model has substantially reduced safety alignment. Use it only for
lawful, controlled purposes and add suitable safeguards before deployment.

README history 2 versions

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

  1. 2026-09-27 Replace with MLP-only NVFP4 PTQ; keep attention and GDN in BF16 f1e503c 1.2 KB
    Loading...
  2. 2026-09-27 Add all-linear compressed-tensors NVFP4 PTQ checkpoint 4087b4d 1.3 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