Downloads · 30 days
43
1% of all-time downloads
lemuralabs/Qwen3.6-27B-Claude-Opus-Reasoning-Distill-v2-abliterated-8bit-mlx
Qwen3.6-27B-Claude-Opus-Reasoning-Distill-v2-abliterated-8bit-mlx is a text generation model from lemuralabs. Use it when you need the model to write or continue text. It is set up for mlx. The card lists the license as apache-2.0.
<p align="center" <img src="logo.png" alt="Lemura Labs" width="110"/ </p
Downloads · 30 days
43
1% of all-time downloads
All-time downloads
4K
Public
Parameters
26.9B
28.6 GB on disk
Likes
1
Public
Click a slice to open those files.
.safetensors28.6 GB · 100%
How the weights are stored.
U3226.9B · 100%
From the Hugging Face model README
THIS IS A TEXT-ONLY MODEL — NO VISION
The upstream abliteration pass stripped the vision tower. For vision-capable Qwen 3.6 27B Opus-Distill MLX quants, see our parallel repos at huggingface.co/lemuralabs (look for repos without
-abliteratedin the name).
8-bit affine MLX quantization of an abliterated Qwen 3.6 27B Claude-Opus reasoning distill, by the Lemura Labs team.
Indistinguishable from BF16 on every benchmark we measured (NLL drift < 0.005). Use this if you have the RAM and want zero quantization drift.
| Disk size | ~27 GB |
| Effective BPW | 8.0 |
| Scheme | Affine 8-bit, group size 64 |
| Recommended RAM | 48 GB Apple Silicon (M4 Pro 48 GB, M4 Max, Studio Ultra) |
| Vision | No — text-only (the upstream abliteration step stripped the ViT) |
| Made by | Lemura Labs |
Qwen/Qwen3.6-27B (Qwen Team — base pretrain)
│
▼
TeichAI/Qwen3.6-27B-Claude-Opus-Reasoning-Distill-v2 (TeichAI — Claude-Opus reasoning distill)
│
▼
abliterated (refusal-ablated) via OBLITERATUS v0.1.2 (multi-direction SVD, BF16)
│
▼
this repo — 8-bit affine, MLX format (Lemura Labs team — quantization)
Direct upstream links:
mlx-lm (recommended)pip install mlx-lm
from mlx_lm import load, generate
model, tokenizer = load("lemuralabs/Qwen3.6-27B-Claude-Opus-Reasoning-Distill-v2-abliterated-8bit-mlx")
prompt = "Explain the difference between SSM and softmax attention in three sentences."
out = generate(model, tokenizer, prompt=prompt, max_tokens=400)
print(out)
messages = [
{"role": "system", "content": "You are a helpful, candid reasoning assistant."},
{"role": "user", "content": "Plan a 3-day Tokyo itinerary for a foodie."},
]
prompt = tokenizer.apply_chat_template(messages, add_generation_prompt=True, tokenize=False)
print(generate(model, tokenizer, prompt=prompt, max_tokens=600))
mlx_lm.generate --model lemuralabs/Qwen3.6-27B-Claude-Opus-Reasoning-Distill-v2-abliterated-8bit-mlx --prompt "Hello" --max-tokens 256
mlx-lm calibration_v5 (~427 KB English text, used for OptiQ sensitivity ranking; uniform/affine variants do not require calibration).The Qwen 3.6 27B family uses a hybrid attention stack — 4 GatedDeltaNet (linear-attention/SSM) layers followed by 1 full-softmax-attention layer, repeated 16× for 64 total layers, 5120 hidden, 248K vocab, 262K context. The SSM kernels lack a VJP path in MLX, so backward-pass-based quant methods (DWQ, dynamic quant) cannot be applied here — OptiQ's forward-only sensitivity approach is the only calibration-aware option that works on this architecture. That's why the OptiQ variants exist.
-abliterated in the name).| Quantization & release | Lemura Labs |
| Reasoning distill | TeichAI (Claude-Opus 4.5/4.6 high-reasoning datasets) |
| Foundation model | Qwen Team |
| Abliteration toolkit | OBLITERATUS by elder-plinius |
| Quant toolkit | mlx-lm, mlx-optiq |
Apache-2.0, inherited from the foundation and distill upstream.
<sub>Need a hosted endpoint, custom quant, or larger-scale inference? — multi-provider LLM routing for the Indian developer ecosystem.</sub>
This MLX build supports lossless block-diffusion speculative decoding via DFlash in mlx_vlm — no requantization, no model changes. On an Apple M4 Max we measured 3.38× (8-bit) and 3.67× (bf16) decode speedups with byte-identical output; other MLX quants of this model should see a similar ~3×.
python3 -m mlx_vlm generate \
--model lemuralabs/Qwen3.6-27B-Claude-Opus-Reasoning-Distill-v2-abliterated-8bit-mlx \
--draft-model z-lab/Qwen3.6-27B-DFlash --draft-kind dflash \
--prompt "Write a merge function for two sorted lists in Python." --max-tokens 256
mlx_vlm ≥ 0.5.0 and access to the gated drafter z-lab/Qwen3.6-27B-DFlash (one-click "Agree and access").DFLASH_SPECULATIVE_DECODING.md.