← back to catalog · registered 2026-11-20 00:04

xenoautomaton/Qwen3.8-27B-abliterated-cyber-preview-MTP-GGUF

xenoautomaton 27B GGUF 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/xenoautomaton%2FQwen3.8-27B-abliterated-cyber-preview-MTP-GGUF"
Response includes
  • classification m8
  • files 15
  • author_summary 2 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
405
Likes
1
Model age
2d ago
created 2026-11-19
Downloads over time
Now 405 → from 0 ↑ 0%
0149297446 0 on Nov 20 405 on Nov 22 Nov
Nov 20 → Nov 22 · 3 snapshots · spans 2 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
Quantizations
Q4_K Q6_K
Tags
llama.cpp gguf qwen3_5 quantized text-generation conversational base_model:hotdogs/Qwen3.8-27B-abliterated-cyber-preview base_model:quantized:hotdogs/Qwen3.8-27B-abliterated-cyber-preview license:apache-2.0 region:us

Related

Total size
36.7 GB
Files
15
Quantizations
4
Registered
2026-11-20 00:04
Last updated on HF
2026-11-19 23:01

Files by quantization

Q6_K 1 file 20.9 GB
Qwen3.8-27B-abliterated-cyber-preview-MTP-Q6_K.gguf 20.9 GB fb7c413e download
Q4_K 1 file 15.8 GB
Qwen3.8-27B-abliterated-cyber-preview-MTP-Q4_K_M.gguf 15.8 GB 1a3eb0de download
BF16 1 file 888 MB
mmproj-Qwen3.8-27B-Cyber-BF16.gguf 888 MB dd4710a5 download
Auxiliary files 12 files 19.1 MB
tokenizer.json 19.1 MB 87a7830d download
LICENSE 11.1 KB d6456956 download
chat_template.jinja 8.74 KB c0c686f9 download
tokenizer_config.json 6.99 KB 773a6327 download
config.json 3.71 KB 3f2a4d7c download
config.original-vlm.json 3.71 KB a0130033 download
README.md 2.82 KB 00b936ce download
.gitattributes 1.78 KB 666804fa download
processor_config.json 1.16 KB 33818c7f download
generation_config.json 213 B 79c8cce3 download
Qwen3.8-27B-abliterated-cyber-preview-MTP-Q4_K_M.gguf.sha256 120 B fc055c30 download
Qwen3.8-27B-abliterated-cyber-preview-MTP-Q6_K.gguf.sha256 118 B c5927015 download

README current version from Hugging Face


library_name: llama.cpp
license: apache-2.0
base_model: hotdogs/Qwen3.8-27B-abliterated-cyber-preview
base_model_relation: quantized
pipeline_tag: text-generation
tags: [gguf, qwen3_5, quantized]

Qwen3.8-27B Cyber Preview — self-quantized GGUF

Generated directly from the original merged BF16 checkpoint at revision
2fbed4d2a8cbf450797dfdcd65bb5ccef74569df, not copied from another quantized repo.
The redundant key wrappers were canonicalized without changing tensor payload
bytes. llama.cpp commit c406f4367e19bea7930e7f74e1984c9acc6d44bf produced the
BF16 intermediate, then independently quantized Q6_K and Q4_K_M from it.

File Size (decimal GB) GPU memory observed at 32K context
Q6_K 22.43 about 23.2 GiB
Q4_K_M 17.01 about 18.4 GiB

Both retain all 770 GGUF tensors, including the 15 embedded MTP tensors. Tensor
names and shapes were checked against the BF16 GGUF intermediate. Checksums and
source provenance are included. The separate BF16 vision projector and original
VLM configuration are retained; text inference is the validated workload.

Run with llama.cpp

llama-server -m Qwen3.8-27B-abliterated-cyber-preview-MTP-Q6_K.gguf \
  -ngl 999 -c 32768 -np 1 -fa on --jinja \
  --spec-type draft-mtp --spec-draft-n-max 1

For vLLM, use the tested W8A16 companion.
Direct GGUF/vLLM was not accepted: the tested GGUF plugin lacked mappings for
the fused attn_qkv_gate tensor emitted by this llama.cpp conversion. The
llama.cpp speeds below must not be presented as vLLM speeds.

Variant / engine Prompt tokens Output tokens End-to-end tok/s Observed stream tok/s
Q6_K / llama.cpp MTP=1 4832 256 58.3 60.2
Q6_K / llama.cpp MTP=1 32 256 60.4 62.2
Q4_K_M / llama.cpp MTP=1 4832 256 66.1 68.4
Q4_K_M / llama.cpp MTP=1 32 256 68.5 71.4

Validation scope

Measured on one NVIDIA A100-SXM4-40GB with a 32,768-token serving context,
one sequence, temperature 0 and thinking disabled per request. Tables show
medians excluding retained warmups. Stream rates are observed from SSE events,
not isolated GPU kernel measurements. Cached and cold TTFT differ; raw records
retain that distinction. Six functional checks cover arithmetic, sorting, Chinese
translation, executing a small generated Python function, tool-call generation
and retrieval. This is not a comprehensive accuracy benchmark. Quantization is
lossy. No RL or further fine-tuning was performed. Vision inference was not tested.

The source's native context remains 262,144 tokens. The 32,768-token runtime
setting does not extend RoPE or imply that the entire native context was tested
on this 40GB card. All upstream credits and Apache-2.0 terms remain applicable.

README history 2 versions

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

  1. 2026-11-19 Document verified A100 runtimes, MTP performance and validation limits 4723a13 2.8 KB
    Loading...
  2. 2026-11-19 Add self-quantized Q6_K with pinned source provenance b80c42a 1.8 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.

Open in Abliteration