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

ApacheOne/Qwen2.5-VL-7B-Instruct-abliterated-nvfp4

ApacheOne Qwen 7B multimodal 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/ApacheOne%2FQwen2.5-VL-7B-Instruct-abliterated-nvfp4"
Response includes
  • classification m1
  • files 6
  • hub_downloads_all_time 358
  • author_summary 1 models
  • readme_text full
Try next
GET/models?family=qwen&method=m1
Sibling qwen models · same method
 GET/methods/m1
What "m1" means · taxonomy
10 credits · hourly refresh · ~4 KB payload Get an API key →
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)
How we label → About M1 →
Refusal direction extracted via
Extraction technique

Difference-of-means

FromArditi et al. (2024) ↗
Confidence
MEDIUM
Why we say so
primary_method=M1; difference-of-means is the reference extraction for M1/M3 (Arditi 2024)
What is a refusal direction? → Read the paper ↗
Downloads · lifetime
358
63 last 30d - stable
Likes
6
Model age
7mo ago
created 2026-02-23
Downloads over time
Now388→from115↑237%
101206311415115 on Feb 25388 on Oct 11FebAprJunAugOct
Feb 25 → Oct 11 · 72 snapshots · spans 228 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.

huihui-ai/Qwen2.5-VL-7B-Instruct-abliterated↓ 300,877 this lineage
ApacheOne/Qwen2.5-VL-7B-Instruct-abliterated-nvfp4↓ 63

Metadata

License
apache-2.0
Languages
en
Tags
qwen2_5_vl qwen qwen2.5 vision-language split nvfp4 fp16 quantization art en base_model:huihui-ai/Qwen2.5-VL-7B-Instruct-abliterated base_model:quantized:huihui-ai/Qwen2.5-VL-7B-Instruct-abliterated

Related

Same base model
mradermacher/Qwen2.5-VL-7B-Instruct-abliterated-GGUF 45K mradermacher/Qwen2.5-VL-7B-Instruct-abliterated-i1-GGUF 7K decerto/Qwen2.5-VL-7B-Instruct-abliterated-Q4_K_M-GGUF 1K BABAYAGA890/Qwen2.5-VL-7B-Instruct-abliterated-GGUF 604 dthryjdrk/Qwen2.5-VL-7B-Instruct-abliterated-Q8_0-GGUF 216
Total size
26.8 GB
Files
6
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-02-24 18:19
Source huggingface.co/ApacheOne/Qwen2.5-VL-7B-Instruct-abliterated-nvfp4 ↗ Base model huihui-ai/Qwen2.5-VL-7B-Instruct-abliterated →

Files by quantization

Auxiliary files 6 files 26.8 GB
Qwen2.5-VL-7B-Instruct-abliterated_merged.safetensors 15.4 GB d6368f80 download
Qwen25VL7B_nvfp4_fixedBF16.safetensors 6.43 GB 737a7330 download
Qwen25VL7B_nvfp4_simple_raw.safetensors 4.91 GB abbfe7ad download
config.json 2.32 KB 4f3d3544 download
.gitattributes 1.48 KB a6344aac download
README.md 986 B 0f2db078 download

README current version from Hugging Face

base_model:

  • huihui-ai/Qwen2.5-VL-7B-Instruct-abliterated
    base_model_relation: quantized
    tags:
  • qwen
  • qwen2.5
  • vision-language
  • split
  • nvfp4
  • fp16
  • quantization
  • art
    license: apache-2.0
    language:
  • en

Combined NVFP4 Encoder

  • Qwen25VL7B_nvfp4_fixedBF16.safetensors - Replaced uint8 (152064, 1792) (lm_head.weight)lm_head and (model.embed_tokens.weight)nn.Embedding with BF16 (152064, 3584). works in comfy

note: Havent done in depth testing

  • Qwen25VL7B_nvfp4_simple_raw.safetensors - has uint8 (152064, 1792) (lm_head.weight)lm_head and (model.embed_tokens.weight)nn.Embedding. comfy wont load this

Split NVFP4 Encoders

  • has uint8 (152064, 1792) lm_head and nn.Embedding. comfy wont load these
  • split_files/text_encoders/qwen_2.5_vl_7b_text_nvfp4_simple.safetensors - Text encoder
  • split_files/vision_encoders/qwen_2.5_vl_7b_vision_nvfp4_simple.safetensors - Vision encoder

note: Still working on mixed layers but

README history 9 versions

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

  1. 2026-02-24Update README.md70c4950986 B
    Loading...
  2. 2026-02-24Update README.md40d49bd955 B
    Loading...
  3. 2026-02-24Update README.md657ce86954 B
    Loading...
  4. 2026-02-24Update README.md5f29ef2933 B
    Loading...
  5. 2026-02-24Update README.md3f7b8c4828 B
    Loading...
  6. 2026-02-24Update README.md1b8cba0550 B
    Loading...
  7. 2026-02-24Update README.mdebe3e46442 B
    Loading...
  8. 2026-02-24Update README.md8c81ae0442 B
    Loading...
  9. 2026-02-24Create README.md0273fd078 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