Instructions to use erokhins/Qwen3.8-test-Q4Q2 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use erokhins/Qwen3.8-test-Q4Q2 with llama.cpp:

Install (macOS, Linux)

curl -LsSf https://llama.app/install.sh | sh
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf erokhins/Qwen3.8-test-Q4Q2
# Run inference directly in the terminal:
llama cli -hf erokhins/Qwen3.8-test-Q4Q2

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf erokhins/Qwen3.8-test-Q4Q2
# Run inference directly in the terminal:
llama cli -hf erokhins/Qwen3.8-test-Q4Q2

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 erokhins/Qwen3.8-test-Q4Q2
# Run inference directly in the terminal:
./llama-cli -hf erokhins/Qwen3.8-test-Q4Q2

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 erokhins/Qwen3.8-test-Q4Q2
# Run inference directly in the terminal:
./build/bin/llama-cli -hf erokhins/Qwen3.8-test-Q4Q2

Use Docker

docker model run hf.co/erokhins/Qwen3.8-test-Q4Q2

LM Studio
Jan
Ollama
How to use erokhins/Qwen3.8-test-Q4Q2 with Ollama:
```
ollama run hf.co/erokhins/Qwen3.8-test-Q4Q2
```
Unsloth Desktop

How to use erokhins/Qwen3.8-test-Q4Q2 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf erokhins/Qwen3.8-test-Q4Q2

Configure the model in Pi

# Install Pi:
npm install -g @earendil-works/pi-coding-agent
# Add to ~/.pi/agent/models.json:
{
  "providers": {
    "llama-cpp": {
      "baseUrl": "http://localhost:8080/v1",
      "api": "openai-completions",
      "apiKey": "none",
      "models": [
        {
          "id": "erokhins/Qwen3.8-test-Q4Q2"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use erokhins/Qwen3.8-test-Q4Q2 with Docker Model Runner:
```
docker model run hf.co/erokhins/Qwen3.8-test-Q4Q2
```

Lemonade

How to use erokhins/Qwen3.8-test-Q4Q2 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull erokhins/Qwen3.8-test-Q4Q2

Run and chat with the model

lemonade run user.Qwen3.8-test-Q4Q2-{{QUANT_TAG}}

List all available models

lemonade list

Hermes Agent

How to use erokhins/Qwen3.8-test-Q4Q2 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf erokhins/Qwen3.8-test-Q4Q2

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 erokhins/Qwen3.8-test-Q4Q2

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use erokhins/Qwen3.8-test-Q4Q2 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf erokhins/Qwen3.8-test-Q4Q2

Configure OpenClaw

# Install OpenClaw:
npm install -g openclaw@latest
# Register the local server and set it as the default model:
openclaw onboard --non-interactive --mode local \
  --auth-choice custom-api-key \
  --custom-base-url http://127.0.0.1:8080/v1 \
  --custom-model-id "erokhins/Qwen3.8-test-Q4Q2" \
  --custom-provider-id llama-cpp \
  --custom-compatibility openai \
  --custom-text-input \
  --accept-risk \
  --skip-health

Run OpenClaw

openclaw agent --local --agent main --message "Hello from Hugging Face"

Qwen3.8-test-Q4Q2 — mixed Q4/Q2 FFN quantization (emulated, GGUF)

Experimental. A 50/50 weight merge of Qwen3.6-27B and Qwen3.8-27B, quantized to llama.cpp Q4_0, with a mixed-precision emulation applied to the FFN:

per layer, the 25% most drift-stable FFN channels (by min cosine of gate/up/down rows between Qwen3.6 and Qwen3.8) keep their full 4-bit codes;
the remaining 75% of channels are reduced to 2 effective bits: their Q4_0 codes are rounded to the nearest of {0, 4, 8, 12} (xy00), block scales unchanged.

This is an emulation of a Q4/Q2 mixed format: the file is a standard Q4_0 GGUF (same size, ~~15.5 GB); only the information content of the unstable channels is Q2. A native split-tensor implementation would land at ~3.6 bits/weight overall (~~12.3 GB).

The chat template is Qwen3.8's. The MTP head is included (nextn layer, kept at full Q4_0) — enable self-drafting speculative decoding with:

llama-server -m Qwen3.8-test-Q4Q2.gguf --spec-type draft-mtp

Observed on an Apple Silicon Mac (Metal): smoke tests indistinguishable from the plain Q4_0 baseline at T=0; 46–48 tok/s decode at 65–79% MTP draft acceptance (mean accepted chain ~3.3).

Give generation enough budget: the template enables thinking, which can consume several hundred tokens before the answer.

Downloads last month: -

GGUF

Model size

27B params

Architecture

qwen35

Hardware compatibility

We're not able to determine the quantization variants.

View all variants

Inference Providers NEW

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

Model tree for erokhins/Qwen3.8-test-Q4Q2

Base model

Qwen/Qwen3.6-27B

Quantized

(738)

this model