← back to catalog · registered 2026-11-21 14:04

onemunanoyo/Inkling-2-Mini-Abliterated-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/onemunanoyo%2FInkling-2-Mini-Abliterated-GGUF"
Response includes
  • classification m8
  • files 2
  • author_summary 8 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
0
Likes
3
Model age
1d ago
created 2026-11-21

Training datasets

2 of 2 in /datasets

Corpora the author lists in the model card. Datasets tracked in our /datasets catalog carry a category badge linking to the workflow stage. Others open on Hugging Face.

Downloads over time
Now 0 → from 0 ↑ 0%
0011 0 on Nov 21 0 on Nov 22 Nov
Nov 21 → Nov 22 · 2 snapshots · spans 1 day

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 Abliterated uncensored unsloth moe inkling2_loop conversational image-text-to-text en ja dataset:heretic-org/Semantic-Harmful dataset:heretic-org/Semantic-Harmless

Related

Total size
0 B
Files
2
Quantizations
1
Registered
2026-11-21 14:04
Last updated on HF
2026-11-21 23:21

Files by quantization

Auxiliary files 2 files 3.64 KB
README.md 2.16 KB a9ef6f5a download
.gitattributes 1.48 KB a6344aac download

README current version from Hugging Face


license: apache-2.0
datasets:

  • heretic-org/Semantic-Harmful
  • heretic-org/Semantic-Harmless
    base_model:
  • unsloth/Inkling-2-Mini-GGUF
    tags:
  • Abliterated

Inkling-2-Mini-Abliterated-GGUF

In-place abliterated Unsloth Dynamic (UD) GGUF builds of thinkingmachines/Inkling-2-Mini derived directly from unsloth/Inkling-2-Mini-GGUF, preserving every tensor's original UD per-tensor quantization allocation (BF16, Q8_0, Q6_K, Q5_K, Q4_K, Q2_0, NVFP4, MXFP4).

Disclaimer: These weights are uncensored / abliterated for research, red-teaming, and alignment evaluation. Review and adhere to local laws and responsible-use practices.

Available Files

File Variant Quant Type Preservation File Size SHA-256 (prefix)
Inkling-2-Mini-UD-Q8_K_XL-Abliterated.gguf UD-Q8_K_XL attn_output: 24 BF16 · ffn_down_exps: 20 NVFP4 + 4 BF16 · output.weight: BF16 TBD TBD
Inkling-2-Mini-UD-Q4_K_XL-Abliterated.gguf UD-Q4_K_XL attn_output: 24 Q8_0 · ffn_down_exps: 20 NVFP4 + 4 Q8_0 · output.weight: Q8_0 TBD TBD
Inkling-2-Mini-UD-Q4_K_M-Abliterated.gguf UD-Q4_K_M attn_output: 24 Q8_0 · ffn_down_exps: 20 NVFP4 + 4 Q8_0 · output.weight: Q8_0 TBD TBD
Inkling-2-Mini-UD-Q2_K_XL-Abliterated.gguf UD-Q2_K_XL attn_output: 18 Q8_0 + 1 Q6_K + 5 Q5_K · ffn_down_exps: 12 Q2_0 + 11 NVFP4 + 1 MXFP4 · output.weight: Q4_K TBD TBD
mmproj-BF16.gguf Vision projector Unmodified from unsloth/Inkling-2-Mini-GGUF (multimodal support) TBD TBD
SHA256SUMS.txt Checksums Full SHA-256 verification list for all artifacts — —

Benchmark

Perplexity

Model PPL ±
Base -
Q8_K_XL

Refusal Evaluation

Model Quantization Refusal

Note

Files

File Quant Size

Run

Limitations

LICENSE

README history 10 versions

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

  1. 2026-11-21 Update README.md 984c365 8 KB
    Loading...
  2. 2026-11-21 Update README with all 4 UD variants, mmproj, SHA256SUMS, and benchmarks 46702db 8.1 KB
    Loading...
  3. 2026-11-21 Update README with UD-Q4_K_XL metrics and SI/GiB sizes 5389b4d 4.5 KB
    Loading...
  4. 2026-11-21 Update README with UD-Q8_K_XL metrics and repo details 0d066c5 3.9 KB
    Loading...
  5. 2026-11-21 Update README.md a156063 2.2 KB
    Loading...
  6. 2026-11-21 Update README.md 56c14e6 594 B
    Loading...
  7. 2026-11-21 Update README.md 0b5af15 582 B
    Loading...
  8. 2026-11-21 draft c93b9fe 558 B
    Loading...
  9. 2026-11-21 Update README.md 64b67c6 162 B
    Loading...
  10. 2026-11-21 initial commit 1cb063c 28 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.

Open in Abliteration