← back to catalog · registered 2026-09-17 19:56

ajgazin/Swift-Qwen3.8-27B-Uncensored-NVFP4

ajgazin 27B multimodal second-order
Abliteration classifier · v1.0.0
M1
Primary method

Direct removal

No other method signals detected in this model.
Confidence
MEDIUM
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=0 (base model)
  • no specific method indicators - defaulting to M1 (most common)
Refusal direction extracted via
Extraction technique

Difference-of-means

Confidence
MEDIUM
Why we say so
primary_method=M1; difference-of-means is the reference extraction for M1/M3 (Arditi 2024)
Downloads · 30-day
0
Likes
2
Model age
today
created 2026-09-17

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
other
Tags
vllm safetensors qwen3_5 nvfp4 fp4 fp8 modelopt blackwell qwen3_8 abliterated uncensored mtp

Related

Total size
20.4 GB
Files
18
Quantizations
1
Registered
2026-09-17 19:56
Last updated on HF
2026-09-17 19:24

Files by quantization

Auxiliary files 18 files 20.4 GB
model-00001-of-00005.safetensors 4.65 GB b62a77a0 download
model-00004-of-00005.safetensors 4.65 GB 25e09b77 download
model-00002-of-00005.safetensors 4.63 GB 7eace8b4 download
model-00003-of-00005.safetensors 4.62 GB dde98f6a download
model-00005-of-00005.safetensors 1.87 GB 048c5040 download
tokenizer.json 12.2 MB 0997f410 download
vocab.json 6.41 MB 0aa0ce06 download
merges.txt 3.20 MB a494e019 download
model.safetensors.index.json 210 KB 69fbba31 download
config.json 85.4 KB cb705f48 download
hf_quant_config.json 52.5 KB 4b9b154c download
tokenizer_config.json 17.5 KB 5de744b3 download
chat_template.jinja 8.74 KB c0c686f9 download
README.md 6.65 KB 480e5bda download
.gitattributes 1.53 KB 52373fe2 download
preprocessor_config.json 390 B 2ea84a43 download
video_preprocessor_config.json 385 B 3ba673a5 download
generation_config.json 221 B 1adb5e77 download

README current version from Hugging Face


license: other
license_name: swift-open-license-1.0
license_link: https://huggingface.co/ukisai/Swift-Qwen3.8-27b
library_name: vllm
pipeline_tag: image-text-to-text
base_model:

  • ajgazin/Swift-Qwen3.8-27B-Uncensored-MTP
    base_model_relation: quantized
    tags:
  • nvfp4
  • fp4
  • fp8
  • modelopt
  • blackwell
  • qwen3_8
  • abliterated
  • uncensored
  • mtp

Swift-Qwen3.8-27B-Uncensored-NVFP4

NVFP4 checkpoint of
ajgazin/Swift-Qwen3.8-27B-Uncensored-MTP,
an abliterated Swift-Qwen3.8-27B
(UkisAI's reasoning-efficient fine-tune of Qwen3.8-27B).
For vLLM and SGLang. GGUFs for llama.cpp:
Swift-Qwen3.8-27B-Uncensored-Dynamic-MTP-GGUF.

  • Uncensored: 15/100 refusals against Swift's 98/100, KL divergence 0.0634 (measured on the BF16
    source).
  • Swift's own NVFP4 recipe, unmodified, from
    ukisai/Swift-Qwen3.8-27B-NVFP4, calibrated
    with NVIDIA ModelOpt.
  • W4A4: runs on native FP4 kernels on Blackwell.
  • MTP head and vision tower in BF16, bit-identical to the source.

Precision

Component Precision
192 MLP projections and lm_head NVFP4 (block size 16, weights and activations)
144 Gated DeltaNet projections (in_proj_qkv, in_proj_z, out_proj) FP8 E4M3
64 attention projections (q_proj, k_proj, v_proj, o_proj) FP8 E4M3
Embeddings, norms, small DeltaNet tensors, vision tower, MTP head BF16
KV cache not quantized

21.9 GB, NVIDIA ModelOpt mixed-precision format.

Usage

vllm serve ajgazin/Swift-Qwen3.8-27B-Uncensored-NVFP4 \
  --max-model-len 32768 \
  --reasoning-parser qwen3 \
  --enable-auto-tool-choice \
  --tool-call-parser qwen3_coder

# self-speculative decoding with the built-in MTP head: add
  --speculative-config '{"method":"mtp","num_speculative_tokens":2}'

Needs a vLLM with ModelOpt mixed-precision support (tested on 0.29.0). No --quantization flag.

Sampling, as for Swift and Qwen: temperature 1.0, top_p 0.95, top_k 20, min_p 0. The model thinks
before answering by default.

Tested on an RTX 5090 (32 GB) with vLLM 0.29.0: NVFP4 layers on FlashInfer's CUTLASS FP4 kernel
(W4A4), FP8 layers on FlashInfer's FP8 kernel, text, MTP decoding and image input all working.

--gpu-memory-utilization 0.88 Weights KV cache
Without MTP 18.9 GiB 91,629 tokens
With MTP 19.7 GiB 43,194 tokens

The model

The source applies the refusal direction of
orcarouter/Qwen3.8-27B-Uncensored
(Arditi et al. 2024, one direction) to Swift's weights. Method, direction and scripts are on the
source model's card.

Model Refusals KL divergence
BF16 source (against Swift) 15/100 0.0634
Swift-Qwen3.8-27B 98/100 0

Measured with Heretic on the BF16 source. Not re-measured on this
checkpoint: Heretic cannot load a ModelOpt export.

Quantization

  • ModelOpt 0.47.0rc0 at commit
    022767c,
    the one Swift used. One RTX PRO 6000 Blackwell.
  • Recipe: modelopt-recipe.json, Swift's. Local-Hessian
    calibration with the FP8 scale sweep for NVFP4, max calibration for FP8.
  • Calibration: 2,048 samples (6,450,474 tokens, up to 4,096 each) from NVIDIA's seven Nemotron
    post-training v3 datasets, rendered with the chat template, seed 42. Swift's sample selection
    script is not published, so the samples differ from theirs;
    calib-manifest.json records ours.

Checks

Held-out NLL (64 samples, disjoint from calibration)
BF16 2.93182
Quantized (ModelOpt simulation, weights and activations) 2.93752 (+0.19%)
  • 193 NVFP4 and 208 FP8 modules, exactly the recipe's; tensor and scale counts equal Swift's checkpoint.
  • All 798 BF16 tensors, the 15 MTP tensors included, are bit-identical to the source
    (precision-inventory.csv).

Reproduce

quantization/ has the recipe, manifests, scripts, package list and run log.

git clone https://github.com/NVIDIA/Model-Optimizer && cd Model-Optimizer
git checkout 022767c7ab3d7d36211affd85e5c496770cde768
pip install torch==2.13.0 --index-url https://download.pytorch.org/whl/cu130
pip install ".[hf]" transformers==5.14.1 datasets==5.0.1 accelerate==1.14.0 safetensors==0.8.0 \
  flash-linear-attention==0.5.2
cd ..

# from a download of quantization/; <source dir> is the source repo at revision f39e20d
python scripts/check_recipe.py --recipe modelopt-recipe.json \
  --index <source dir>/model.safetensors.index.json
python scripts/build_calib.py --manifest calibration-manifest.json --tokenizer <source dir> --out calib
python scripts/quantize_nvfp4.py --model <source dir> --recipe modelopt-recipe.json \
  --calib calib/calib.jsonl --heldout calib/heldout.jsonl --out out \
  --modelopt-examples Model-Optimizer/examples/hf_ptq
python scripts/check_export.py --out out --source <source dir> --recipe modelopt-recipe.json

About 60 GB of GPU memory, about an hour.

Not evaluated

Refusals and KL divergence of the quantized model, general benchmarks, and FP8 KV cache (the
checkpoint has no KV scales).

License

Derivative of Swift-Qwen3.8-27B, under the Swift Open License v1.0
(Swift model card): free for individuals and
organizations up to US$1,000,000 annual recurring revenue, above that commercial use needs a Swift
Enterprise License from UkisAI. Qwen3.8-27B and orcarouter/Qwen3.8-27B-Uncensored are Apache 2.0.

Intended use

The model answers requests the original declines. You are responsible for how you use it and for
complying with applicable law and the license.

Credits

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 app" button that hands off directly to a local runtime of your choice - Infrahuman, LM Studio, or Ollama. No API keys, no subscription, no prompt leakage.