← back to catalog · registered 2026-09-25 21:57

r3lax/Huihui-Ornith-1.5-35B-A3B-abliterated-MTPLX-mixed8-4bit-FP16

r3lax 35B MoE second-order
curl -H "Authorization: Bearer $ABL_KEY" \
     "https://abliteration.org/api/v1/models/r3lax%2FHuihui-Ornith-1.5-35B-A3B-abliterated-MTPLX-mixed8-4bit-FP16"
Response includes
  • classification m1
  • files 19
  • author_summary 1 models
  • readme_text full
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)
Refusal direction extracted via
Extraction technique

Difference-of-means

Confidence
MEDIUM
Why we say so
primary_method=M1; difference-of-means is the reference extraction for M1/M3 (Arditi 2024)
Downloads · 30-day
0
Likes
1
Model age
today
created 2026-09-25

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
mit
Tags
mlx safetensors qwen3_5_moe mtplx qwen3-moe ornith abliterated uncensored fp16 apple-silicon text-generation conversational

Related

Total size
23.5 GB
Files
19
Quantizations
1
Registered
2026-09-25 21:57
Last updated on HF
2026-09-25 21:36

Files by quantization

Auxiliary files 19 files 23.6 GB
model-00001-of-00005.safetensors 5.00 GB b5f9d6de download
model-00003-of-00005.safetensors 4.99 GB 71969b69 download
model-00004-of-00005.safetensors 4.91 GB f12c460a download
model-00002-of-00005.safetensors 4.91 GB d2cf751c download
mtp.safetensors 1.57 GB ecb67550 download
model-00005-of-00005.safetensors 1.32 GB 340e4aea download
model-vision.safetensors 852 MB 0b8aec60 download
tokenizer.json 26.8 MB 697dcad3 download
model.safetensors.index.json 200 KB a7455305 download
config.json 73.1 KB 697d438a download
chat_template.jinja 7.36 KB b07660cc download
mtplx_runtime.json 5.66 KB 5507add8 download
.gitattributes 1.53 KB 52373fe2 download
README.md 1.42 KB 4bd65a77 download
processor_config.json 1.16 KB 33818c7f download
tokenizer_config.json 1.13 KB 8d610519 download
preprocessor_config.json 390 B 2ea84a43 download
video_preprocessor_config.json 385 B 3ba673a5 download
generation_config.json 202 B 023756cf download

README current version from Hugging Face


license: mit
base_model: huihui-ai/Huihui-Ornith-1.5-35B-A3B-abliterated
tags:

  • mlx
  • mtplx
  • qwen3-moe
  • ornith
  • abliterated
  • uncensored
  • fp16
  • apple-silicon
    pipeline_tag: text-generation

Huihui Ornith 1.5 35B A3B Abliterated — MTPLX mixed 8/4-bit FP16

MTPLX (native MTP speculative decoding on Apple Silicon) forge of
huihui-ai/Huihui-Ornith-1.5-35B-A3B-abliterated,
itself an abliterated (refusal-removed, layers 11–29) uncensored build of
ornith-ai/Ornith-1.5-35B-A3B.

Quant recipe (stolen from the attention8 Ornith forge, fp16 lane)

  • body: 4-bit affine, group 32 — MoE experts (bulk, sparse-active)
  • 8-bit affine, group 64: all linear_attn projections, all self_attn q/k/v/o,
    mlp.gate, mlp.shared_expert_gate, lm_head, embed_tokens
  • trunk dtype fp16 + MTP sidecar fp16 — the fast native path on M1/M2 Macs,
    which have no native BF16. Runs fine on M3+ too.

Measured on M1 Max 64GB

mode tok/s vs AR
AR 51.1 1.00x
D1 68.2 1.34x
D2 74.8 1.46x
D3 73.3 1.44x

Tuned best: D2 (74.0 t/s, 1.49x). Profile: sustained.

Use

mtplx serve --model r3lax/Huihui-Ornith-1.5-35B-A3B-abliterated-MTPLX-mixed8-4bit-FP16 --download

License: MIT (matches source). Forged with MTPLX 2.11.2.

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 app" button that hands off directly to a local runtime of your choice - Abliteration, LM Studio, or Ollama. No API keys, no subscription, no prompt leakage.