Instructions to use MachineFi/Trio-Spark-DI-1.4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use MachineFi/Trio-Spark-DI-1.4 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 MachineFi/Trio-Spark-DI-1.4:Q8_0
# Run inference directly in the terminal:
llama cli -hf MachineFi/Trio-Spark-DI-1.4:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf MachineFi/Trio-Spark-DI-1.4:Q8_0
# Run inference directly in the terminal:
llama cli -hf MachineFi/Trio-Spark-DI-1.4:Q8_0

Build from source code

git clone https://github.com/ggml-org/llama.cpp.git
cd llama.cpp
cmake -B build
cmake --build build -j --target llama-app
# Start a local OpenAI-compatible server with a web UI:
./build/bin/llama serve -hf MachineFi/Trio-Spark-DI-1.4:Q8_0
# Run inference directly in the terminal:
./build/bin/llama cli -hf MachineFi/Trio-Spark-DI-1.4:Q8_0

Use Docker

docker model run hf.co/MachineFi/Trio-Spark-DI-1.4:Q8_0

Llama new
LM Studio
Jan

vLLM

How to use MachineFi/Trio-Spark-DI-1.4 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "MachineFi/Trio-Spark-DI-1.4"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "MachineFi/Trio-Spark-DI-1.4",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/MachineFi/Trio-Spark-DI-1.4:Q8_0

Ollama
How to use MachineFi/Trio-Spark-DI-1.4 with Ollama:
```
ollama run hf.co/MachineFi/Trio-Spark-DI-1.4:Q8_0
```
Unsloth Desktop

How to use MachineFi/Trio-Spark-DI-1.4 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf MachineFi/Trio-Spark-DI-1.4: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": "MachineFi/Trio-Spark-DI-1.4:Q8_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use MachineFi/Trio-Spark-DI-1.4 with Docker Model Runner:
```
docker model run hf.co/MachineFi/Trio-Spark-DI-1.4:Q8_0
```

Lemonade

How to use MachineFi/Trio-Spark-DI-1.4 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull MachineFi/Trio-Spark-DI-1.4:Q8_0

Run and chat with the model

lemonade run user.Trio-Spark-DI-1.4-Q8_0

List all available models

lemonade list

Hermes Agent

How to use MachineFi/Trio-Spark-DI-1.4 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf MachineFi/Trio-Spark-DI-1.4: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 MachineFi/Trio-Spark-DI-1.4:Q8_0

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use MachineFi/Trio-Spark-DI-1.4 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf MachineFi/Trio-Spark-DI-1.4: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 "MachineFi/Trio-Spark-DI-1.4: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"

Trio-Spark-DI-1.4 / README.md

Raullen

Same model the hosted API serves

6ea4336 verified 12 days ago

preview code

Raw History Blame Contribute Delete

4.19 kB

---
license: cc-by-nc-4.0
base_model: Qwen/Qwen3.5-9B
base_model_relation: finetune
library_name: gguf
pipeline_tag: text-generation
tags:
- decision-model
- decision-index
- gguf
extra_gated_heading: "Access Trio-Spark DI 1.4 (research, non-commercial)"
extra_gated_prompt: "These weights are released under CC BY-NC 4.0 for research and evaluation only. Commercial use requires a separate license from MachineFi (or use the hosted Trio-Spark API)."
extra_gated_fields:
  Name: text
  Organization: text
  Intended use: text
  I agree to use these weights for non-commercial purposes only: checkbox
---
# Trio-Spark DI 1.4 (research checkpoint)

Trio-Spark DI 1.4 is a research checkpoint of MachineFi's Trio-Spark decision model, trained for the
[Decision Index](https://github.com/apolinario/decision-index) typed-decision protocol:
given a `state` and typed questions with explicit options, it returns a probability for every option.

- **Base model:** [Qwen/Qwen3.5-9B](https://huggingface.co/Qwen/Qwen3.5-9B) (Apache-2.0).
- **Weights:** `trio-spark-di-1.4-q8_0.gguf` (Q8_0, about 9.5 GB, sha256 `1c3369fe…`), text only. This is the file evaluated below and submitted to the text board.
- This is the same Trio-Spark v1.4 model the hosted API serves as `trio-spark`, as a text-only Q8_0 build; the API also accepts images and video. The earlier research checkpoint is [MachineFi/Trio-Spark-DI](https://huggingface.co/MachineFi/Trio-Spark-DI).

## Decision Index 0.3 (public suite, self-run)

| | Decision Index | Knowledge | Language | Retrieval | Tools | Arts |
|---|---:|---:|---:|---:|---:|---:|
| **Trio-Spark DI 1.4 (Q8_0), full public suite** | **56.41** | 39.4 | 59.3 | 58.3 | 80.0 | 46.1 |
All 140,178 requests of the public 0.3 suite, run with the official kit against the server in `serve/` (18 requests longer than 8,192 tokens refused, none truncated), with four servers per GPU. With one server per GPU (the default) the server is deterministic: the same request always gets the same answer.
Submitted to the Decision Index leaderboard; the official score is the maintainers' private Full score and may differ.
## Running it
Trio-Spark DI 1.4 is not a chat model. It answers through its decision prompt and reads a probability for every option from the model's next-letter scores, so plain chat prompts will not reproduce the scores above.
The model reads the whole prompt with full (non-causal) attention, so it must be served with the settings in `serve/setup.sh` (llama.cpp `--override-kv qwen35.attention.causal=bool:false`, no prompt cache); a default llama.cpp server gives different, worse answers.
The `/v1/systemone` server used for the evaluation is in `serve/`: llama.cpp (release b11249, CUDA 12.8) and a small Python front end (standard library only) that sends the questions of a request to llama.cpp together, in a fixed order, and clears the slots between requests, so answers do not depend on what ran before. It is compatible with the Decision Index kit's `http` engine.
```sh
hf download MachineFi/Trio-Spark-DI-1.4 --local-dir trio-spark-di-1.4
bash trio-spark-di-1.4/serve/setup.sh $PWD/trio-spark-di-1.4/trio-spark-di-1.4-q8_0.gguf
# then, from the Decision Index kit:
python -m decision_index pipeline --engine http --option base_url=http://127.0.0.1:8800 --option model=trio-spark --out runs/trio-spark-di-1.4
```
Server limits: up to 8,192 tokens per decision prompt (longer requests are refused, never truncated); questions with more than 8 options are decided by an 8-way tournament (groups of 8, then a final over the group winners), so no option is dropped.
## License
The weights are released under [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/): free for research and evaluation with attribution,
**no commercial use**. For commercial use, use the [hosted Trio-Spark API](https://platform.machinefi.com/spark) or contact MachineFi.
The base model Qwen3.5-9B is Apache-2.0; its license and notice apply to the base weights (see `NOTICE`). Full license text: `LICENSE`.
The hosted API, which serves this model with vision, is documented at [MachineFi/Trio-Spark](https://huggingface.co/MachineFi/Trio-Spark).

	---
	license: cc-by-nc-4.0
	base_model: Qwen/Qwen3.5-9B
	base_model_relation: finetune
	library_name: gguf
	pipeline_tag: text-generation
	tags:
	- decision-model
	- decision-index
	- gguf
	extra_gated_heading: "Access Trio-Spark DI 1.4 (research, non-commercial)"
	extra_gated_prompt: "These weights are released under CC BY-NC 4.0 for research and evaluation only. Commercial use requires a separate license from MachineFi (or use the hosted Trio-Spark API)."
	extra_gated_fields:
	Name: text
	Organization: text
	Intended use: text
	I agree to use these weights for non-commercial purposes only: checkbox
	---

	# Trio-Spark DI 1.4 (research checkpoint)

	Trio-Spark DI 1.4 is a research checkpoint of MachineFi's Trio-Spark decision model, trained for the
	[Decision Index](https://github.com/apolinario/decision-index) typed-decision protocol:
	given a `state` and typed questions with explicit options, it returns a probability for every option.

	- Base model: [Qwen/Qwen3.5-9B](https://huggingface.co/Qwen/Qwen3.5-9B) (Apache-2.0).
	- Weights: `trio-spark-di-1.4-q8_0.gguf` (Q8_0, about 9.5 GB, sha256 `1c3369fe…`), text only. This is the file evaluated below and submitted to the text board.
	- This is the same Trio-Spark v1.4 model the hosted API serves as `trio-spark`, as a text-only Q8_0 build; the API also accepts images and video. The earlier research checkpoint is [MachineFi/Trio-Spark-DI](https://huggingface.co/MachineFi/Trio-Spark-DI).

	## Decision Index 0.3 (public suite, self-run)

	\| \| Decision Index \| Knowledge \| Language \| Retrieval \| Tools \| Arts \|
	\|---\|---:\|---:\|---:\|---:\|---:\|---:\|
	\| Trio-Spark DI 1.4 (Q8_0), full public suite \| 56.41 \| 39.4 \| 59.3 \| 58.3 \| 80.0 \| 46.1 \|

	All 140,178 requests of the public 0.3 suite, run with the official kit against the server in `serve/` (18 requests longer than 8,192 tokens refused, none truncated), with four servers per GPU. With one server per GPU (the default) the server is deterministic: the same request always gets the same answer.
	Submitted to the Decision Index leaderboard; the official score is the maintainers' private Full score and may differ.

	## Running it

	Trio-Spark DI 1.4 is not a chat model. It answers through its decision prompt and reads a probability for every option from the model's next-letter scores, so plain chat prompts will not reproduce the scores above.

	The model reads the whole prompt with full (non-causal) attention, so it must be served with the settings in `serve/setup.sh` (llama.cpp `--override-kv qwen35.attention.causal=bool:false`, no prompt cache); a default llama.cpp server gives different, worse answers.

	The `/v1/systemone` server used for the evaluation is in `serve/`: llama.cpp (release b11249, CUDA 12.8) and a small Python front end (standard library only) that sends the questions of a request to llama.cpp together, in a fixed order, and clears the slots between requests, so answers do not depend on what ran before. It is compatible with the Decision Index kit's `http` engine.

	```sh
	hf download MachineFi/Trio-Spark-DI-1.4 --local-dir trio-spark-di-1.4
	bash trio-spark-di-1.4/serve/setup.sh $PWD/trio-spark-di-1.4/trio-spark-di-1.4-q8_0.gguf
	# then, from the Decision Index kit:
	python -m decision_index pipeline --engine http --option base_url=http://127.0.0.1:8800 --option model=trio-spark --out runs/trio-spark-di-1.4
	```

	Server limits: up to 8,192 tokens per decision prompt (longer requests are refused, never truncated); questions with more than 8 options are decided by an 8-way tournament (groups of 8, then a final over the group winners), so no option is dropped.

	## License

	The weights are released under [CC BY-NC 4.0](https://creativecommons.org/licenses/by-nc/4.0/): free for research and evaluation with attribution,
	no commercial use. For commercial use, use the [hosted Trio-Spark API](https://platform.machinefi.com/spark) or contact MachineFi.
	The base model Qwen3.5-9B is Apache-2.0; its license and notice apply to the base weights (see `NOTICE`). Full license text: `LICENSE`.

	The hosted API, which serves this model with vision, is documented at [MachineFi/Trio-Spark](https://huggingface.co/MachineFi/Trio-Spark).