Instructions to use unsloth/Qwen3.8-27B-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use unsloth/Qwen3.8-27B-GGUF 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 unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M
# Run inference directly in the terminal:
llama cli -hf unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M
# Run inference directly in the terminal:
llama cli -hf unsloth/Qwen3.8-27B-GGUF:UD-Q4_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 unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf unsloth/Qwen3.8-27B-GGUF:UD-Q4_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 unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M

Use Docker

docker model run hf.co/unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M

LM Studio
Jan
Ollama
How to use unsloth/Qwen3.8-27B-GGUF with Ollama:
```
ollama run hf.co/unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M
```
Unsloth Desktop

How to use unsloth/Qwen3.8-27B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M

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": "unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use unsloth/Qwen3.8-27B-GGUF with Docker Model Runner:
```
docker model run hf.co/unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M
```

Lemonade

How to use unsloth/Qwen3.8-27B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M

Run and chat with the model

lemonade run user.Qwen3.8-27B-GGUF-UD-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use unsloth/Qwen3.8-27B-GGUF with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf unsloth/Qwen3.8-27B-GGUF:UD-Q4_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 unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use unsloth/Qwen3.8-27B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M

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 "unsloth/Qwen3.8-27B-GGUF:UD-Q4_K_M" \
  --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"

UD-Q5_K_XL: "model has unused tensor"

#102

by calisti - opened Aug 22

Discussion

calisti

Aug 22

Unusual llama.cpp output with UD-Q5_K_XL – what is that about?

0.00.931.145 W model has unused tensor blk.64.attn_norm.weight (size = 20480 bytes) -- ignoring
0.00.931.162 W model has unused tensor blk.64.post_attention_norm.weight (size = 20480 bytes) -- ignoring
0.00.931.166 W model has unused tensor blk.64.attn_q.weight (size = 51609600 bytes) -- ignoring
0.00.931.168 W model has unused tensor blk.64.attn_k.weight (size = 5570560 bytes) -- ignoring
0.00.931.170 W model has unused tensor blk.64.attn_v.weight (size = 5570560 bytes) -- ignoring
0.00.931.175 W model has unused tensor blk.64.attn_output.weight (size = 25804800 bytes) -- ignoring
0.00.931.177 W model has unused tensor blk.64.attn_q_norm.weight (size = 1024 bytes) -- ignoring
0.00.931.178 W model has unused tensor blk.64.attn_k_norm.weight (size = 1024 bytes) -- ignoring
0.00.931.180 W model has unused tensor blk.64.ffn_gate.weight (size = 73113600 bytes) -- ignoring
0.00.931.182 W model has unused tensor blk.64.ffn_down.weight (size = 73113600 bytes) -- ignoring
0.00.931.183 W model has unused tensor blk.64.ffn_up.weight (size = 73113600 bytes) -- ignoring
0.00.931.187 W model has unused tensor blk.64.nextn.eh_proj.weight (size = 43008000 bytes) -- ignoring
0.00.931.189 W model has unused tensor blk.64.nextn.enorm.weight (size = 20480 bytes) -- ignoring
0.00.931.190 W model has unused tensor blk.64.nextn.hnorm.weight (size = 20480 bytes) -- ignoring
0.00.931.195 W model has unused tensor blk.64.nextn.shared_head_norm.weight (size = 20480 bytes) -- ignoring

Losiki

Aug 22

•

edited Aug 22

It's MTP. Use --spec-type draft-mtp --spec-draft-n-max 2 in llama.cpp to enable it. It will use more RAM, but accelerate the generation process.
If you do not intend to use MTP, you can ignore this notice.

Also see:
https://unsloth.ai/docs/models/mtp
https://github.com/ggml-org/llama.cpp/pull/26296

calisti

about 1 month ago

Ah, right. I had commented out some of the lines to try the larger quant – unaware of the fact that Bash does not continue \-ending-escaped lines across comment blocks.

    ./llama-server \
      -hf unsloth/Qwen3.8-27B-GGUF:UD-Q5_K_XL \
      -ngl 99 -c $(( 118 * 1024 )) -fa on -np 1 \
      # -hf unsloth/Qwen3.8-27B-GGUF:UD-Q5_K_M \ # <-- whoops, rest of the invocation isn't part of the command any more.
      # -ngl 99 -c $(( 143 * 1024 )) -fa on -np 1 \
      --cache-type-k q8_0 \
      --cache-type-v q5_1 \
      --spec-type draft-mtp --spec-draft-n-max 2 \
      --jinja \
      --reasoning on \
      --reasoning-budget -1 \
      --reasoning-preserve \
      --chat-template-kwargs '{"preserve_thinking":true}' \
      --temp 0.6 --top-p 0.95 --top-k 20 --min-p 0.0 --presence-penalty 0 \
      --alias "Qwen3.8-27B UD-Q5_K_XL (121K)" \
      --host 0.0.0.0 --port 8080

calisti changed discussion status to closed about 1 month ago

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment