← back to catalog · registered 2026-11-17 03:03

jakeatx/qwen3.8-27b-Swift1.5-Uncensored-ATX-AI3

jakeatx Qwen 27B 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/jakeatx%2Fqwen3.8-27b-Swift1.5-Uncensored-ATX-AI3"
Response includes
  • classification m-uncensored
  • files 4
  • author_summary 12 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
1K
Likes
6
Model age
5d ago
created 2026-11-17
Downloads over time
Now 1.1K → from 0 ↑ 0%
04158301.2K 0 on Nov 17 1.1K on Nov 22 Nov
Nov 17 → Nov 22 · 6 snapshots · spans 5 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
gguf qwen3_5 mtp speculative-decoding llamampere ampere base_model:Qwen/Qwen3.8-27B base_model:quantized:Qwen/Qwen3.8-27B license:apache-2.0 region:us

Related

Control vectors

Total size
10.5 GB
Files
4
Quantizations
2
Registered
2026-11-17 03:03
Last updated on HF
2026-11-17 13:46

Files by quantization

Drafts and companions 1 file 8.00 MB
lsa-qwen38-27b-d2048.gguf 8.00 MB 5489a4f2 download
Auxiliary files 3 files 10.5 GB
qwen3.8-27b-Swift1.5-Uncensored-ATX-AI3.gguf 10.5 GB d28a3c00 download
README.md 3.37 KB f0d5a92b download
.gitattributes 1.78 KB d77bcb99 download

README current version from Hugging Face


license: apache-2.0
base_model: Qwen/Qwen3.8-27B
tags:

  • gguf
  • qwen3_5
  • mtp
  • speculative-decoding
  • llamampere
  • ampere

Swift 1.5 Qwen3.8-27B Uncensored, ATX AI3 (3.0 bits per weight on card, with MTP, 12 GB cards)

Qwen3.8-27B Swift 1.5, compressed by llamAmpere to fit a 12 GB GPU with long context and MTP speculative decoding on.

Files in this repo: qwen3.8-27b-Swift1.5-Uncensored-ATX-AI3.gguf plus lsa-qwen38-27b-d2048.gguf. Keep them all in one folder.

Small cards: 8 / 10 / 12 GB

Card File Context MTP Peak VRAM (measured)
12 GB qwen3.8-27b-Swift1.5-Uncensored-ATX-AI3.gguf 262,144 On 11,544 MiB
10 GB qwen3.8-27b-Swift1.5-Uncensored-ATX-AI2 (2.3 bpw) 212,992 On —
8 GB qwen3.8-27b-Swift1.5-Uncensored-ATX-AI2.1.gguf 204,800 On 7,506 MiB

Quality: drift from full precision (KLD, lower is better). All rows use the same 16,200 positions (coding, agentic and RAG at ~12K ctx), scored against Swift 1.5 BF16. The Unsloth rows are Unsloth's exact UD per-tensor recipes and imatrix applied to Swift 1.5, so every row shares one teacher.

File Size KLD Top-1
ATX 12 GB 11.24 GB 0.0270 95.98%
Unsloth UD-IQ3_XXS recipe 10.93 GB 0.0396 95.33%
Unsloth UD-Q2_K_XL recipe 9.95 GB 0.0591 94.00%
ATX AI2 2.3 bpw (10 GB) 9.14 GB 0.0593 94.06%
Mirai S — 0.0745 —

The 12 GB file has 54% less drift than UD-Q2_K_XL. The 2.3 bpw file matches UD-Q2_K_XL in a file 19% smaller.

Shorter answers. On LiveCodeBench v6 (the same 100 problems), the 12 GB file used 33% fewer tokens than UD-Q2_K_XL: median 4,608 vs 8,932 per answer, 95% range 22–44% fewer. It was shorter on 74 of 100 problems.

Coding score, one seed (directional, not statistically significant). LiveCodeBench v6, strict grader: ATX 12 GB 89%, UD-Q2_K_XL 87% (run on llamAmpere), Qwen BF16 published 90.3%.

Fit. With standard attention, UD-Q2_K_XL cannot run full context with MTP on a 12 GB card (estimate). The ATX 12 GB file runs the full 262K context with MTP.

Running the small-card files requires llamAmpere 0.7 (coming, not yet released): LSA attention + host-offloaded KV. The --preset commands below will work once 0.7 ships. Download the sidecar files into the same folder as the model:

# 12 GB
./llama-server -m qwen3.8-27b-Swift1.5-Uncensored-ATX-AI3.gguf --preset 12dense
# 8 GB
./llama-server -m qwen3.8-27b-Swift1.5-Uncensored-ATX-AI2.1.gguf --preset 8dense

Sidecars: lsa-qwen38-27b-d2048.gguf (LSA index map, both sizes). For 8 GB only: mtp-dm.gguf, head-ait2.gguf (draft overlays) and head-exact-f16.gguf (exact rescoring head).

SHA-256:

  • qwen3.8-27b-Swift1.5-Uncensored-ATX-AI3.gguf d28a3c00f53ce95566df6478b9733d570e8143d5cca669ff057465f1e26812f5
  • qwen3.8-27b-Swift1.5-Uncensored-ATX-AI2.1.gguf cfb2ece5c79ba94cb3880f5edaab023c655e99c1797c4438e9e97c7b8e74aad7
  • lsa-qwen38-27b-d2048.gguf 5489a4f2f6e9169c57db7344afdebdca243c3d438d8e70197c52a6c95244aa68
  • mtp-dm.gguf 4a8e9a5cfb6b1395088bd88b26400b504e47a10fad5a109e4cbee3276f9d5cae
  • head-ait2.gguf 673687fe741bf65c34ebdc73c53248de60f82184cd3a0138b320acfe3d811e5d
  • head-exact-f16.gguf f812690e5925605edafb2f364e1b40ef89756856c7b6e16ba1e4515bc80d1227

README history 4 versions

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

  1. 2026-11-17 Mark 8 GB file withdrawn 000dd3c 3.3 KB
    Loading...
  2. 2026-11-17 Rename to AI+ format name 248c155 3.4 KB
    Loading...
  3. 2026-11-17 Retitle to on-card bpw (3.0 bpw) cc50019 3.4 KB
    Loading...
  4. 2026-11-17 Add README.md f13553c 3.2 KB
    Loading...

Discussions 1 thread

  1. 2026-11-18 Thanks!! open 3 💬 #1
    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