Instructions to use XpressAI/Qwen3.6-27B-RYS-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use XpressAI/Qwen3.6-27B-RYS-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="XpressAI/Qwen3.6-27B-RYS-GGUF",
	filename="Qwen3.6-27B-DFlash-Q8_0-rys.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 XpressAI/Qwen3.6-27B-RYS-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 XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0
# Run inference directly in the terminal:
llama cli -hf XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0
# Run inference directly in the terminal:
llama cli -hf XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0

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 XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0

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 XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0

Use Docker

docker model run hf.co/XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0

LM Studio
Jan
Ollama
How to use XpressAI/Qwen3.6-27B-RYS-GGUF with Ollama:
```
ollama run hf.co/XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0
```

Unsloth Studio

How to use XpressAI/Qwen3.6-27B-RYS-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 XpressAI/Qwen3.6-27B-RYS-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 XpressAI/Qwen3.6-27B-RYS-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for XpressAI/Qwen3.6-27B-RYS-GGUF to start chatting

How to use XpressAI/Qwen3.6-27B-RYS-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0

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": "XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use XpressAI/Qwen3.6-27B-RYS-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 XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0

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 XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use XpressAI/Qwen3.6-27B-RYS-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0

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 "XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0" \
  --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"

Docker Model Runner
How to use XpressAI/Qwen3.6-27B-RYS-GGUF with Docker Model Runner:
```
docker model run hf.co/XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0
```

Lemonade

How to use XpressAI/Qwen3.6-27B-RYS-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull XpressAI/Qwen3.6-27B-RYS-GGUF:Q8_0

Run and chat with the model

lemonade run user.Qwen3.6-27B-RYS-GGUF-Q8_0

List all available models

lemonade list

wmeddie commited on Jun 18

Commit

3ac3fac

verified ·

1 Parent(s): 6c66129

Document DFlash speculative-decoding draft model

Browse files

Files changed (1) hide show

README.md +41 -4

README.md CHANGED Viewed

@@ -9,6 +9,8 @@ tags:
   - reasoning
   - bfcl
   - function-calling
 language:
   - en
 ---
@@ -46,10 +48,11 @@ variant slightly underperforms base.
 ## Files
-| File | Quant | Layers | Size |
-|------|-------|--------|------|
-| `Qwen3.6-27B-rys_33-36-UD-Q4_K_XL.gguf` | Q4_K_XL | 68 | 18 GiB |
-| `Qwen3.6-27B-rys_33-36-Q8_0.gguf` | Q8_0 | 68 | 29 GiB |
 The base GGUF (no surgery) is at
 [unsloth/Qwen3.6-27B-GGUF](https://huggingface.co/unsloth/Qwen3.6-27B-GGUF).
@@ -238,6 +241,40 @@ truncating mid-thought.
 About 22 GiB at Q4_K_XL with 32k context and Q8 KV cache. Fits comfortably on
 a single A100 40 GB.
 ---
 ## When to use this

   - reasoning
   - bfcl
   - function-calling
+  - speculative-decoding
+  - dflash
 language:
   - en
 ---
 ## Files
+| File | Quant | Layers | Size | Role |
+|------|-------|--------|------|------|
+| `Qwen3.6-27B-rys_33-36-UD-Q4_K_XL.gguf` | Q4_K_XL | 68 | 18 GiB | target model |
+| `Qwen3.6-27B-rys_33-36-Q8_0.gguf` | Q8_0 | 68 | 29 GiB | target model |
+| `Qwen3.6-27B-DFlash-Q8_0-rys.gguf` | Q8_0 | 5 | 1.8 GiB | DFlash draft (speculative decoding) |
 The base GGUF (no surgery) is at
 [unsloth/Qwen3.6-27B-GGUF](https://huggingface.co/unsloth/Qwen3.6-27B-GGUF).
 About 22 GiB at Q4_K_XL with 32k context and Q8 KV cache. Fits comfortably on
 a single A100 40 GB.
+### Speculative decoding with DFlash (faster inference)
+RYS makes the model deeper (68 vs 64 layers), so each token costs a little more
+compute. You can win that back — and then some — with **DFlash speculative
+decoding**, a diffusion-style draft head that proposes a block of tokens per
+step which the target model verifies in one pass.
+`Qwen3.6-27B-DFlash-Q8_0-rys.gguf` is a DFlash draft **trained against the
+68-layer RYS layout** — its hidden-state taps (`target_layer_ids = [1, 16, 31,
+50, 65]`) index into the duplicated-block arrangement, so it must be paired with
+the RYS target, not the stock base model.
+DFlash support is not in upstream llama.cpp yet; use
+[BeeLlama](https://github.com/Anbeeld/beellama.cpp), a llama.cpp fork that adds
+the `dflash` spec type:
+```bash
+./build/bin/llama-server \
+  -m ~/models/Qwen3.6-27B-rys_33-36-UD-Q4_K_XL.gguf \
+  --spec-draft-model ~/models/Qwen3.6-27B-DFlash-Q8_0-rys.gguf \
+  --spec-type copyspec,dflash \
+  --spec-dflash-cross-ctx 1024 \
+  --port 9999 -np 1 --kv-unified \
+  -ngl all --spec-draft-ngl all \
+  -b 2048 -ub 1024 --flash-attn on \
+  --jinja --no-host --reasoning on \
+  --chat-template-kwargs '{"preserve_thinking":true}' \
+  --temp 0.6 --top-k 20 --top-p 1.0 --min-p 0.0
+```
+The draft adds ~1.8 GiB of VRAM. Acceptance is highest on the long, structured
+thinking chains this model produces, which is exactly where the RYS variant
+spends its tokens.
 ---
 ## When to use this