Instructions to use XHToken/Spark-X2.5-4B-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 XHToken/Spark-X2.5-4B-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 XHToken/Spark-X2.5-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf XHToken/Spark-X2.5-4B-GGUF:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf XHToken/Spark-X2.5-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf XHToken/Spark-X2.5-4B-GGUF: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 XHToken/Spark-X2.5-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf XHToken/Spark-X2.5-4B-GGUF: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 XHToken/Spark-X2.5-4B-GGUF:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf XHToken/Spark-X2.5-4B-GGUF:Q4_K_M

Use Docker

docker model run hf.co/XHToken/Spark-X2.5-4B-GGUF:Q4_K_M

LM Studio
Jan

vLLM

How to use XHToken/Spark-X2.5-4B-GGUF with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "XHToken/Spark-X2.5-4B-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": "XHToken/Spark-X2.5-4B-GGUF",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/XHToken/Spark-X2.5-4B-GGUF:Q4_K_M

Ollama
How to use XHToken/Spark-X2.5-4B-GGUF with Ollama:
```
ollama run hf.co/XHToken/Spark-X2.5-4B-GGUF:Q4_K_M
```
Unsloth Desktop

How to use XHToken/Spark-X2.5-4B-GGUF with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf XHToken/Spark-X2.5-4B-GGUF: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": "XHToken/Spark-X2.5-4B-GGUF:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use XHToken/Spark-X2.5-4B-GGUF with Docker Model Runner:
```
docker model run hf.co/XHToken/Spark-X2.5-4B-GGUF:Q4_K_M
```

Lemonade

How to use XHToken/Spark-X2.5-4B-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull XHToken/Spark-X2.5-4B-GGUF:Q4_K_M

Run and chat with the model

lemonade run user.Spark-X2.5-4B-GGUF-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use XHToken/Spark-X2.5-4B-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 XHToken/Spark-X2.5-4B-GGUF: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 XHToken/Spark-X2.5-4B-GGUF:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use XHToken/Spark-X2.5-4B-GGUF with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf XHToken/Spark-X2.5-4B-GGUF: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 "XHToken/Spark-X2.5-4B-GGUF: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"

Request: Build Compatible llama.cpp Runtime for Spark-X2.5-4B-GGUF (Windows, CUDA)

by KiberAndy - opened 26 days ago

Discussion

KiberAndy

26 days ago

Please provide a prebuilt llama.cpp binary with CUDA support for Windows that can run this GGUF (Spark-X2.5-4B). Standard builds don't work. If one exists, give a link; if not, please build it.

lyang01 changed discussion status to closed 26 days ago

lyang01

SparkLLM org 26 days ago

Which llama.cpp version/build are you using? Could you also share the error message you’re getting?

KiberAndy

26 days ago

Which llama.cpp version/build are you using? Could you also share the error message you’re getting?

[LM Studio]
I even downloaded a fairly recent version - b10834 from here: https://github.com/ggml-org/llama.cpp/releases

Still the error:
Failed to load model.
error loading model: unknown model architecture: 'spark2_5'

More detailed logs:

2026-09-07 20:58:40 [DEBUG]
0.00.034.483 I srv init: The UI is disabled
0.00.034.488 I srv init: Use --ui/--no-ui (or deprecated --webui/--no-webui) to enable/disable
2026-09-07 20:58:40 [DEBUG]
0.00.049.271 I srv load_model: loading model 'G:\Ai\LM\lmstudio-community\Spark-X2.5-4B\Spark-X2.5-4B.gguf'
2026-09-07 20:58:40 [DEBUG]
0.00.132.121 E llama_model_load: error loading model: unknown model architecture: 'spark2_5'
0.00.132.131 E llama_model_load_from_file_impl: failed to load model
0.00.132.158 E common_fit_params: encountered an error while trying to fit params to free device memory: error loading model: unknown model architecture: 'spark2_5'
2026-09-07 20:58:40 [DEBUG]
0.00.215.145 E llama_model_load: error loading model: unknown model architecture: 'spark2_5'
0.00.215.155 E llama_model_load_from_file_impl: failed to load model
0.00.215.156 LMSTUDIO_STARTUP_ERROR:{"schemaVersion":1,"code":"model_load_failed","message":"error loading model: unknown model architecture: 'spark2_5'"}
E cmn common_init_: failed to load model 'G:\Ai\LM\lmstudio-community\Spark-X2.5-4B\Spark-X2.5-4B.gguf'
0.00.215.164 E srv load_model: failed to load model, 'G:\Ai\LM\lmstudio-community\Spark-X2.5-4B\Spark-X2.5-4B.gguf': error loading model: unknown model architecture: 'spark2_5'
0.00.215.187 I srv operator(): operator(): cleaning up before exit...
2026-09-07 20:58:40 [DEBUG]
0.00.215.846 E srv llama_server: exiting due to model loading error

KiberAndy

25 days ago

I managed to compile it for CUDA. The model is recognized and loads specifically onto the GPU, as I wanted. The problem is solved.

Can you open the Discussion to everyone, and may I share my compiled archive here?

lyang01 changed discussion status to open 25 days ago

lyang01

SparkLLM org 25 days ago

Thank you so much, @KiberAndy ! Really appreciate your support!

KiberAndy

25 days ago

Prebuild the Compatible llama.cpp Runtime (CUDA, Windows)
https://mega.nz/file/HIJRgLZQ#NRu566nhEKbhCCTmTjKKm0Lghbf-8Qjlh4IEj6jqz3c

tftranslate

23 days ago

Would you care to elaborate whether you applied any patches to make llama.cpp recognize the model type, or did you just pull the latest code from github main/master? I need to custom-build for a different architecture.

KiberAndy

23 days ago

•

edited 23 days ago

Would you care to elaborate whether you applied any patches to make llama.cpp recognize the model type, or did you just pull the latest code from github main/master? I need to custom-build for a different architecture.

No patches. I simply cloned it following the instructions here: https://huggingface.co/XHToken/Spark-X2.5-4B-GGUF
But the commands were different:

cmake -S . -B build-cuda `
-G "Visual Studio 17 2022" `
-A x64 `
-T "cuda=C:\Users\User\scoop\apps\cuda\current" `
-DGGML_CUDA=ON `
-DGGML_NATIVE=OFF `
-DGGML_CCACHE=OFF `
-DLLAMA_BUILD_TESTS=OFF `
-DLLAMA_BUILD_EXAMPLES=ON

cmake --build build-cuda --config Release --target llama-server --parallel 8

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