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

Lord-H4D3ZS/Qwen3.6-35B-A3B-abliterated-ROCmFP4

Lord-H4D3ZS Qwen 35B GGUF MoE second-order 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/Lord-H4D3ZS%2FQwen3.6-35B-A3B-abliterated-ROCmFP4"
Response includes
  • classification m8
  • files 3
  • hub_downloads_all_time 432
  • author_summary 2 models
  • readme_text full
Try next
GET/models?author=Lord-H4D3ZS
All 2 models by Lord-H4D3ZS
 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
432
126 last 30d - stable
Likes
0
Model age
8w ago
created 2026-08-13
Downloads over time
Now506→from294↑72%
283365446527294 on Aug 19506 on Oct 11AugSepOct
Aug 19 → Oct 11 · 48 snapshots · spans 53 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.

wangzhang/Qwen3.6-35B-A3B-abliterated-v2↓ 309 this lineage
Lord-H4D3ZS/Qwen3.6-35B-A3B-abliterated-ROCmFP4↓ 126

Metadata

License
apache-2.0
Tags
gguf rocmfpx rocmfp4 mtp text-generation conversational base_model:wangzhang/Qwen3.6-35B-A3B-abliterated-v2 base_model:quantized:wangzhang/Qwen3.6-35B-A3B-abliterated-v2 license:apache-2.0 endpoints_compatible region:us

Related

Other by Lord-H4D3ZS
Qwen3.8-Distill-35B-A3B-Coder-Abliterated 21K
Same base model
mradermacher/Qwen3.6-35B-A3B-abliterated-v2-i1-GGUF 2K mradermacher/Qwen3.6-35B-A3B-abliterated-v2-GGUF 1K zecanard/Qwen3.6-35B-A3B-uncensored-abliterix-v2-MLX-2bit-mixed_2_6 230 zecanard/Qwen3.6-35B-A3B-uncensored-abliterix-v2-MLX-3bit-mixed_3_6 205 zecanard/Qwen3.6-35B-A3B-uncensored-abliterix-v2-MLX-4bit-mixed_4_6 78
Total size
17.4 GB
Files
3
Quantizations
1
Registered
2026-08-22 13:56
Last updated on HF
2026-08-13 16:49
Source huggingface.co/Lord-H4D3ZS/Qwen3.6-35B-A3B-abliterated-ROCmFP4 ↗ Base model wangzhang/Qwen3.6-35B-A3B-abliterated-v2 →

Files by quantization

Auxiliary files 3 files 17.4 GB
qwen3.6-35b-v2-ROCMFP4.gguf 17.4 GB a073b894 download
.gitattributes 1.55 KB fe9ecabf download
README.md 704 B 1b8e6b21 download

README current version from Hugging Face

license: apache-2.0
base_model: wangzhang/Qwen3.6-35B-A3B-abliterated-v2
tags:

  • gguf
  • rocmfpx
  • rocmfp4
  • mtp
  • text-generation
  • conversational
    pipeline_tag: text-generation

Qwen3.6-35B-A3B Abliterated (ROCmFP4)

This is an abliterated (uncensored) Qwen 3.6 35B MoE model converted to GGUF and quantized to ROCmFP4 (Q4_0_ROCMFP4_FAST_COHERENT) using the ROCmFPX toolchain.

Optimized for AMD Radeon GPUs (e.g., RX 9060 XT 16GB / RX 7900 XT 20GB) via Vulkan or ROCm backends with Multi-Token Prediction (MTP) speculative decoding enabled.

How to Run with llama-server

llama-server.exe -m qwen3.6-35b-v2-ROCMFP4.gguf --spec-type draft-mtp -c 32768 --port 8080

README history 1 version

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

  1. 2026-08-13Create README.md74120b5704 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