Get the app
← back to catalog · registered 2026-08-22 13:56

Spellman66/Huihui-ai-Qwen3.5-2B-Abliterated

Spellman66 Qwen 2B GGUF multimodal 262K ctx
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/Spellman66%2FHuihui-ai-Qwen3.5-2B-Abliterated"
Response includes
  • classification m8
  • files 10
  • hub_downloads_all_time 4,763
  • author_summary 1 models
  • readme_text full
Try next
GET/models?family=qwen&method=m8
Sibling qwen models · same method
 GET/methods/m8
What "m8" means · taxonomy
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
M1 Direct removal →
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
How we label → About M8 →
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
5K
868 last 30d - stable
Likes
0
Model age
3mo ago
created 2026-06-24
Downloads over time
Now5.1K→from655↑683%
4312.1K3.9K5.6K655 on Jun 245.1K on Oct 11JunJulAugSepOct
Jun 24 → Oct 11 · 55 snapshots · spans 109 days

Metadata

Quantizations
BF16 F16 Q4_K Q5_K Q6_K Q8_0
Tags
gguf qwen3_5 llama.cpp unsloth vision-language-model endpoints_compatible region:us conversational
Total size
12.9 GB
Files
10
Quantizations
8
Registered
2026-08-22 13:56
Last updated on HF
2026-06-24 23:54
Source huggingface.co/Spellman66/Huihui-ai-Qwen3.5-2B-Abliterated ↗

Files by quantization

BF16 1 file 3.52 GB
Huihui-Qwen3.5-2B-abliterated.BF16.gguf 3.52 GB 8fbcf293 download
F16 1 file 3.52 GB
Huihui-Qwen3.5-2B-abliterated.F16.gguf 3.52 GB 6ec234e9 download
Q8_0 1 file 1.87 GB
Huihui-Qwen3.5-2B-abliterated.Q8_0.gguf 1.87 GB b4915764 download
Q6_K 1 file 1.45 GB
Huihui-Qwen3.5-2B-abliterated.Q6_K.gguf 1.45 GB fd7a2879 download
Q5_K 1 file 1.31 GB
Huihui-Qwen3.5-2B-abliterated.Q5_K_M.gguf 1.31 GB 2c9ab62c download
Q4_K 1 file 1.19 GB
Huihui-Qwen3.5-2B-abliterated.Q4_K_M.gguf 1.19 GB 974879b9 download
mmproj 1 file 640 MB
Huihui-Qwen3.5-2B-abliterated.BF16-mmproj.gguf 640 MB d377af60 download
Auxiliary files 3 files 5.55 KB
config.json 2.74 KB 45132e81 download
.gitattributes 2.01 KB 8ec343d6 download
README.md 818 B 0e17f80a download

README current version from Hugging Face

tags:

  • gguf
  • llama.cpp
  • unsloth
  • vision-language-model

Huihui-ai-Qwen3.5-2B-Abliterated : GGUF

This model was finetuned and converted to GGUF format using Unsloth.

Example usage:

  • For text only LLMs: llama-cli -hf Spellman66/Huihui-ai-Qwen3.5-2B-Abliterated --jinja
  • For multimodal models: llama-mtmd-cli -hf Spellman66/Huihui-ai-Qwen3.5-2B-Abliterated --jinja

Available Model files:

  • Huihui-Qwen3.5-2B-abliterated.Q4_K_M.gguf
  • Huihui-Qwen3.5-2B-abliterated.BF16-mmproj.gguf
    This was trained 2x faster with Unsloth

README history 6 versions

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

  1. 2026-06-24Add README9c1eca7797 B
    Loading...
  2. 2026-06-24Add README74bd34a797 B
    Loading...
  3. 2026-06-24Add README3170a72795 B
    Loading...
  4. 2026-06-24Add README39dfd3b795 B
    Loading...
  5. 2026-06-24Add README4b5d07d795 B
    Loading...
  6. 2026-06-24Add README6c66e57794 B
    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.

Get the app Browse the catalog
Open in Abliteration