Instructions to use AtomicChat/Qwen3-8B-DFlash-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 AtomicChat/Qwen3-8B-DFlash-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 AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0
# Run inference directly in the terminal:
llama cli -hf AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0
# Run inference directly in the terminal:
llama cli -hf AtomicChat/Qwen3-8B-DFlash-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 AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf AtomicChat/Qwen3-8B-DFlash-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 AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0

Use Docker

docker model run hf.co/AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0

LM Studio
Jan

vLLM

How to use AtomicChat/Qwen3-8B-DFlash-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "AtomicChat/Qwen3-8B-DFlash-GGUF"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "AtomicChat/Qwen3-8B-DFlash-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0

Ollama
How to use AtomicChat/Qwen3-8B-DFlash-GGUF with Ollama:
```
ollama run hf.co/AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0
```
Unsloth Desktop

How to use AtomicChat/Qwen3-8B-DFlash-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0

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": "AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use AtomicChat/Qwen3-8B-DFlash-GGUF with Docker Model Runner:
```
docker model run hf.co/AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0
```

Lemonade

How to use AtomicChat/Qwen3-8B-DFlash-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0

Run and chat with the model

lemonade run user.Qwen3-8B-DFlash-GGUF-Q8_0

List all available models

lemonade list

Hermes Agent

How to use AtomicChat/Qwen3-8B-DFlash-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 AtomicChat/Qwen3-8B-DFlash-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 AtomicChat/Qwen3-8B-DFlash-GGUF:Q8_0

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use AtomicChat/Qwen3-8B-DFlash-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf AtomicChat/Qwen3-8B-DFlash-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 "AtomicChat/Qwen3-8B-DFlash-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"

Qwen3-8B-DFlash-GGUF / README.md

AlexAtomic

Upload README.md with huggingface_hub

9017c28 verified 2 months ago

4.31 kB

license: mit
base_model:
  - z-lab/Qwen3-8B-DFlash-b16
base_model_relation: quantized
quantized_by: AlexAtomic
pipeline_tag: text-generation
library_name: gguf
tags:
  - atomic-chat
  - dflash
  - speculative-decoding
  - draft-model
  - qwen
  - gguf
  - llama.cpp

Draft: z-lab/Qwen3-8B-DFlash-b16 · Target: Qwen/Qwen3-8B

Qwen3 8B DFlash, the DFlash speculative-decoding draft converted to GGUF by Atomic Chat. Built straight from z-lab's original weights. Runs fully offline.

What this is

DFlash is a speculative-decoding method that drafts a whole block of candidate tokens in a single forward pass using a lightweight block-diffusion model, instead of one token at a time. This repo is the draft component only — it does nothing on its own. You run it alongside the target model Qwen/Qwen3-8B, which verifies the drafted block and keeps the longest correct prefix. Output is identical to running the target alone, just faster.

These GGUFs are converted from z-lab's original weights, not a repack of someone else's GGUF. Q8_0 and bf16 give the same draft acceptance, so Q8_0 is the pick.

Run in llama.cpp

Needs a build of llama.cpp with DFlash speculative decoding (PR #22105). You supply the target as -m and this draft as -md:

./llama-server \
    -m   Qwen3-8B.gguf \
    -md  Qwen3-8B-DFlash.Q8_0.gguf \
    --spec-type draft-dflash --spec-draft-n-max 15 \
    -ngl 99 -fa on --jinja -c 8192

DFlash is trained for non-thinking generation — pass enable_thinking=false in the chat template for best acceptance.

Choosing a quant

Quant	Size	Notes
`Q8_0`	1.12 GB	Recommended. Same acceptance as bf16, half the size and slightly faster drafting.
`bf16`	2.10 GB	Full-precision draft (reference). No acceptance gain over Q8_0.

Performance

z-lab report up to 6.17x lossless acceleration for Qwen3-8B on their reference stack (vLLM / SGLang / Transformers). In llama.cpp today the DFlash port is newer: on our RTX 4090 test (Q8_0 target + Q8_0 draft, code generation) it delivered about 2.3x end-to-end at roughly 26% draft acceptance. Acceptance is set by the implementation and the content, not by the quantization (bf16 and Q8_0 measure the same). Speedups grow on structured/code output and shrink on free-form prose.

How this was made

Download the DFlash draft z-lab/Qwen3-8B-DFlash-b16 (original weights).
Convert to GGUF with llama.cpp convert_hf_to_gguf.py --target-model-dir (the target supplies the tokenizer; its weights are not needed).
Quantize the draft to Q8_0.

License

Released by z-lab under the MIT license. Converted to GGUF by Atomic Chat. See the DFlash paper and project page.