Instructions to use deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF",
	filename="Qwen3-30B-A3B-Cerebellum-v1-Q3_K_M.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M
# Run inference directly in the terminal:
llama-cli -hf deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M
# Run inference directly in the terminal:
llama-cli -hf deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M

Use pre-built binary

# Download pre-built binary from:
# https://github.com/ggerganov/llama.cpp/releases
# Start a local OpenAI-compatible server with a web UI:
./llama-server -hf deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M
# Run inference directly in the terminal:
./llama-cli -hf deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M

Build from source code

git clone https://github.com/ggerganov/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-server llama-cli
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama-server -hf deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M

Use Docker

docker model run hf.co/deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M

LM Studio
Jan
Ollama
How to use deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF with Ollama:
```
ollama run hf.co/deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M
```

Unsloth Studio

How to use deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF with Unsloth Studio:

Install Unsloth Studio (macOS, Linux, WSL)

curl -fsSL https://unsloth.ai/install.sh | sh
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF to start chatting

Install Unsloth Studio (Windows)

irm https://unsloth.ai/install.ps1 | iex
# Run unsloth studio
unsloth studio -H 0.0.0.0 -p 8888
# Then open http://localhost:8888 in your browser
# Search for deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF to start chatting

How to use deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M

Configure the model in Pi

# Install Pi:
npm install -g @mariozechner/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama-server -hf deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M

Configure Hermes

# Install Hermes:
curl -fsSL https://hermes-agent.nousresearch.com/install.sh | bash
hermes setup
# Point Hermes at the local server:
hermes config set model.provider custom
hermes config set model.base_url http://127.0.0.1:8080/v1
hermes config set model.default deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M

Run Hermes

hermes

Atomic Chat new
Docker Model Runner
How to use deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF with Docker Model Runner:
```
docker model run hf.co/deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M
```

Lemonade

How to use deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF:Q3_K_M

Run and chat with the model

lemonade run user.Qwen3-30B-A3B-Cerebellum-GGUF-Q3_K_M

List all available models

lemonade list

Cerebellum

Qwen3-30B-A3B — Cerebellum GGUF

Ablation-guided mixed-precision quantization of Qwen/Qwen3-30B-A3B. 30B total parameters, 3B active (MoE with 128 experts, 8 active per token).

What is Cerebellum?

Instead of uniform quantization, we measure which weight groups survive aggressive compression and which don't. Groups that tolerate Q2_K get demoted; groups that don't stay at Q3_K_M or higher. The result: smaller files with less quality loss than uniform quants of the same size.

Files

File	Size	Description
`Qwen3-30B-A3B-Cerebellum-v3-Q3_K_M.gguf`	12 GB	Recommended — coder-optimized imatrix, improved allocation
`Qwen3-30B-A3B-Cerebellum-v2-Q3_K_M.gguf`	12 GB	Previous release
`Qwen3-30B-A3B-Cerebellum-v1-Q3_K_M.gguf`	9.6 GB	Maximum compression — all expert groups at Q2_K

Benchmarks

Evaluated using our standardized benchmark suite (ARC-Challenge, HellaSwag, MMLU-Redux, HumanEval+) with temperature=0, no thinking mode.

The model-index metadata in this card's frontmatter mirrors the recommended v3 build, measured with the local llama.cpp benchmark harness on RTX 3090. Full per-question artifacts are in benchmark_results/.

Cerebellum v3 (12 GB) — Recommended

Benchmark	Score	Questions
ARC-Challenge	92.7%	1,172
HellaSwag	83.8%	10,042
MMLU-Redux	66.6%	2,400
HumanEval+ (base)	75.0%	164
HumanEval+ (plus)	70.7%	164

Size vs Quality

Model	Size	ARC	HellaSwag	MMLU	HumanEval
Q3_K_M (baseline)	14 GB	—	—	—	—
Cerebellum v3	12 GB	92.7%	83.8%	66.6%†	75.0%‡
Cerebellum v2	12 GB	90.5%	80.3%	69.9%*	72.0%
Cerebellum v1	9.6 GB	—	—	—	—

* v2 MMLU = full MMLU (11,643 questions); v3 = MMLU-Redux (2,400 questions, harder subset)
‡ v3 HumanEval+ uses EvalPlus (stricter test cases than original HumanEval)

v3 improves +2.2 to +3.5 points over v2 on directly comparable benchmarks (ARC, HellaSwag).

Methodology

Group ablation: Demote each of 7 expert weight groups (attn_k, attn_q, attn_v, ffn_down_exps, ffn_gate_exps, ffn_up_exps, output) to Q2_K individually. Measure PPL impact.
Reverse ablation: From an all-Q2_K baseline, promote one group back to Q3_K_M. Measure PPL recovery.
Build optimal mix: Groups with the best recovery-per-byte get promoted. Groups that survive Q2_K stay demoted.

Override Maps

v3 (coder imatrix) — Sacred (kept at Q3_K_M): attn_q, attn_v, ffn_down_exps
v3 — Demoted (Q2_K): attn_k, attn_output, ffn_gate_exps, ffn_up_exps

v2 (wiki imatrix) — Sacred (kept at Q3_K_M): attn_v, ffn_down_exps, output
v2 — Demoted (Q2_K): attn_k, attn_q, ffn_gate_exps, ffn_up_exps

All 48 layers treated uniformly (no per-layer variation needed for this model).

The key difference: v3 keeps attn_q at higher precision (swap with attn_output vs v2), driven by the coder-focused imatrix finding query projections more sensitive to quantization in code-focused inference.

Usage

Works with any llama.cpp-compatible tool:

# llama.cpp
./llama-server --model Qwen3-30B-A3B-Cerebellum-v3-Q3_K_M.gguf -ngl 99 --ctx-size 4096

# Ollama (create Modelfile pointing to the GGUF)
# LM Studio (drag and drop)
# koboldcpp, text-generation-webui, etc.

Hardware Requirements

v3 / v2 (12 GB): Fits in 16 GB VRAM with room for context. RTX 4060 Ti 16GB, RTX 3090, etc.
v1 (9.6 GB): Fits in 12 GB VRAM. RTX 4070, RTX 3060 12GB, etc.

Credits

Quantized with Cerebellum — ablation-guided mixed-precision quantization by deucebucket.
Base model by Qwen.

Downloads last month: 838

GGUF

Model size

31B params

Architecture

qwen3moe

Hardware compatibility

3-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for deucebucket/Qwen3-30B-A3B-Cerebellum-GGUF

Base model

Qwen/Qwen3-30B-A3B-Base

Finetuned

Qwen/Qwen3-30B-A3B

Quantized

(119)

this model

Evaluation results

normalized accuracy on AI2 Reasoning Challenge
test set Local benchmark run (RTX 3090, llama.cpp)

0.927
accuracy on HellaSwag
validation set Local benchmark run (RTX 3090, llama.cpp)

0.838
accuracy on MMLU-Redux
test set Local benchmark run (RTX 3090, llama.cpp)

0.666
pass@1 on HumanEval+ (pass@1)
test set Local benchmark run (RTX 3090, llama.cpp)

0.707