Instructions to use singulared/Qwen3.8-27B-ROCmFP4-MTP-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 singulared/Qwen3.8-27B-ROCmFP4-MTP-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 singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF
# Run inference directly in the terminal:
llama cli -hf singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF
# Run inference directly in the terminal:
llama cli -hf singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

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 singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF
# Run inference directly in the terminal:
./llama-cli -hf singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

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 singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF
# Run inference directly in the terminal:
./build/bin/llama-cli -hf singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

Use Docker

docker model run hf.co/singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

LM Studio
Jan

vLLM

How to use singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "singulared/Qwen3.8-27B-ROCmFP4-MTP-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": "singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

Ollama
How to use singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF with Ollama:
```
ollama run hf.co/singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF
```

Unsloth Studio

How to use singulared/Qwen3.8-27B-ROCmFP4-MTP-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 singulared/Qwen3.8-27B-ROCmFP4-MTP-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 singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF to start chatting

How to use singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

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": "singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

OpenClaw new

How to use singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

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 "singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF" \
  --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 singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF with Docker Model Runner:
```
docker model run hf.co/singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF
```

Lemonade

How to use singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

Run and chat with the model

lemonade run user.Qwen3.8-27B-ROCmFP4-MTP-GGUF-{{QUANT_TAG}}

List all available models

lemonade list

Hermes Agent

How to use singulared/Qwen3.8-27B-ROCmFP4-MTP-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 singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

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 singulared/Qwen3.8-27B-ROCmFP4-MTP-GGUF

Run Hermes

hermes

Atomic Chat

Qwen3.8-27B-ROCmFP4-MTP-GGUF / README.md

singulared

Model card: ROCmFP4 builds + MTP drafter ladder with acceptance data

0a95e53 verified about 18 hours ago

preview code

Raw

History Blame Contribute Delete

5.08 kB

	---
	license: apache-2.0
	base_model:
	- Qwen/Qwen3.8-27B
	- ggml-org/Qwen3.8-27B-GGUF
	base_model_relation: quantized
	library_name: llama.cpp
	pipeline_tag: text-generation
	tags:
	- gguf
	- rocm
	- rocmfpx
	- rocmfp4
	- amd
	- strix-halo
	- gfx1151
	- mtp
	- speculative-decoding
	- qwen3.5
	---

	# Qwen3.8-27B — ROCmFP4 + MTP drafter ladder (Strix Halo)

	ROCmFP4 builds of [Qwen3.8-27B](https://huggingface.co/Qwen/Qwen3.8-27B), quantised from
	`ggml-org/Qwen3.8-27B-GGUF`'s BF16 (sha256 `5a3eedc837bcbd13…`, verified), **plus an MTP drafter
	at five precisions** so the speculative-decoding numbers below can be reproduced rather than
	taken on trust.

	Other ROCmFP4 builds of this model already exist. What is here and (at publication) not
	elsewhere: draft-acceptance rates, a per-backend `n-max` sweep, a `-ub` sweep, and a
	ROCm 7.2 vs 10.1 comparison that reverses the preset ranking.

	> ⚠ Read this before choosing FP4. On the same machine, **mainline llama.cpp on Vulkan with a
	> plain `Q4_K_M` and the official MTP drafter is faster** — 330 vs 227 prefill (+43%), decode tied.
	> These FP4 files are worth it for the 1.7 GiB smaller footprint (15.6 vs 17.3 GiB resident),
	> which matters when co-residing two models. They are not the throughput winner.

	## Files

	\| file \| preset \| size \|
	\| --- \| --- \| ---: \|
	\| `Qwen3.8-27B-ROCMFP4-STRIX.gguf` \| `Q4_0_ROCMFP4_STRIX` \| 13.75 GiB \|
	\| `Qwen3.8-27B-ROCMFP4-COHERENT.gguf` \| `Q4_0_ROCMFP4_COHERENT` \| 14.41 GiB \|
	\| `mtp-Qwen3.8-27B-ROCMFP4-STRIX.gguf` \| FP4 drafter \| 1.85 GiB \|
	\| `mtp-Qwen3.8-27B-ROCMFP3.gguf` \| FP3 drafter \| 1.55 GiB \|
	\| `mtp-Qwen3.8-27B-ROCMFP6.gguf` \| FP6 drafter \| 2.27 GiB \|
	\| `mtp-Qwen3.8-27B-ROCMFP8.gguf` \| FP8 drafter \| 2.86 GiB \|
	\| `mtp-Qwen3.8-27B-ROCMFP2.gguf` \| FP2 drafter — broken, see below \| 1.48 GiB \|

	Requires a [ROCmFPX](https://github.com/charlie12345/ROCmFPX) build; mainline llama.cpp does not
	know the `Q4_0_ROCMFP4_*` tensor types.

	## Hardware / method

	AMD Ryzen AI MAX+ 395, Radeon 8060S (`gfx1151`, RDNA 3.5), 124 GiB GTT, Debian sid, kernel 7.2.
	Server-measured (`llama-server` + probe), ~8K-token prompt, temperature 0, one job at a time.
	ROCm nightly pinned to `therock-dist-linux-gfx1151-10.1.0a20260815`.

	## 1. Draft depth (`--spec-draft-n-max`) is per-backend

	\| n-max \| Vulkan Q4_K_M decode \| acc \| FP4 decode \| acc \|
	\| ---: \| ---: \| ---: \| ---: \| ---: \|
	\| 3 \| — \| — \| 33.42 \| 100.0% \|
	\| 4 \| 35.80 \| 88.1% \| 35.16 \| 98.7% \|
	\| 5 \| 38.94 \| 91.6% \| 38.67 \| 98.1% \|
	\| 6 \| 38.47 \| 86.5% \| 38.04 \| 97.5% \|
	\| 7 \| 37.84 \| 82.0% \| 39.26 \| 94.7% \|
	\| 8 \| 28.56 \| 78.1% \| 32.36 \| 95.2% \|
	\| 10 \| 25.47 \| 59.4% \| — \| — \|

	Acceptance decays monotonically with depth; the knee is where verifying rejected drafts costs
	more than the accepted ones save. FP4 holds higher acceptance, so its knee sits deeper and
	flatter. This is not DeepSeek-V4's n=2 — draft depth does not transfer between models.

	## 2. Drafter precision is a bandwidth lever, not a quality one

	Target fixed, drafter varied, Vulkan, n=5:

	\| drafter \| size \| decode \| acceptance \|
	\| --- \| ---: \| ---: \| ---: \|
	\| Q4_K_M \| 1.89 GiB \| 39.16 \| 91.6% \|
	\| Q6_K \| 2.28 GiB \| 38.28 \| 92.1% \|
	\| Q5_K_M \| 2.08 GiB \| 36.93 \| 89.0% \|
	\| Q8_0 \| 2.95 GiB \| 34.74 \| 89.0% \|

	Acceptance is flat (89–92%) while decode spans 13% — so shrinking the drafter buys bandwidth
	and costs nothing in draft quality. Advice to keep drafters at ≥Q8 does not hold here.

	## 3. FP2 destroys a drafter

	FPX ladder, STRIX target, ROCm 10.1, n=5:

	\| drafter \| decode \| acceptance \|
	\| --- \| ---: \| ---: \|
	\| FP4-STRIX \| 37.03 \| 97.4% \|
	\| FP3 \| 36.07 \| 98.1% \|
	\| FP6 \| 30.92 \| 96.6% \|
	\| FP8 \| 29.76 \| 96.6% \|
	\| FP2 \| 22.07 \| 64.0% \|

	FP2's codebook has no exact zero. This drafter is BF16-sourced — the case usually assumed safe —
	and acceptance still collapses. Do not use FP2 for a draft model.

	## 4. The preset ranking flips with the ROCm version

	`llama-bench`, pp2048:

	\| preset \| ROCm 7.2.4 \| ROCm 10.1 nightly \|
	\| --- \| ---: \| ---: \|
	\| COHERENT \| 205.7 \| 208.6 (+1%) \|
	\| STRIX \| 151.8 \| 272.0 (+79%) \|

	COHERENT wins on 7.2; STRIX wins on 10.1. Any "preset X is best" claim — including ones in other
	repos — is conditional on a ROCm version that usually goes unstated.

	## 5. `-ub 256`, not the default

	\| `-ub` \| Vulkan pp2048 \| FP4 pp2048 (10.1) \|
	\| ---: \| ---: \| ---: \|
	\| 256 \| 370.6 \| 300.3 \|
	\| 512 \| 360.9 \| 267.3 \|
	\| 1024 \| 343.7 \| 236.5 \|
	\| 2048 \| 332.0 \| 235.8 \|

	Both backends prefer a small micro-batch; `-ub 2048` costs FP4 27% of its prefill. Batch size
	(`-b` 512→4096) changes nothing.

	## Usage

	```bash
	llama-server \
	-m Qwen3.8-27B-ROCMFP4-STRIX.gguf \
	-md mtp-Qwen3.8-27B-ROCMFP4-STRIX.gguf \
	--spec-type draft-mtp --spec-draft-n-max 5 \
	-ngl 99 -ngld 99 -fa on -ub 256
	```

	## Provenance

	Every file derives from `ggml-org/Qwen3.8-27B-GGUF`, sha256-verified before quantisation:
	`Qwen3.8-27B-BF16.gguf` = `5a3eedc837bcbd13…`, `mtp-Qwen3.8-27B-BF16.gguf` = `5723e551c4ee2b8c…`.