Instructions to use SurgeFF/AriannaV3 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use SurgeFF/AriannaV3 with Transformers:

# Use a pipeline as a high-level helper
from transformers import pipeline

pipe = pipeline("image-text-to-text", model="SurgeFF/AriannaV3")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
pipe(text=messages)

# Load model directly
from transformers import AutoProcessor, AutoModelForMultimodalLM

processor = AutoProcessor.from_pretrained("SurgeFF/AriannaV3")
model = AutoModelForMultimodalLM.from_pretrained("SurgeFF/AriannaV3", device_map="auto")
messages = [
    {
        "role": "user",
        "content": [
            {"type": "image", "url": "https://huggingface.co/datasets/huggingface/documentation-images/resolve/main/p-blog/candy.JPG"},
            {"type": "text", "text": "What animal is on the candy?"}
        ]
    },
]
inputs = processor.apply_chat_template(
	messages,
	add_generation_prompt=True,
	tokenize=True,
	return_dict=True,
	return_tensors="pt",
).to(model.device)

outputs = model.generate(**inputs, max_new_tokens=40)
print(processor.decode(outputs[0][inputs["input_ids"].shape[-1]:]))

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use SurgeFF/AriannaV3 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 SurgeFF/AriannaV3:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf SurgeFF/AriannaV3:Q4_K_M

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf SurgeFF/AriannaV3:Q4_K_M
# Run inference directly in the terminal:
llama cli -hf SurgeFF/AriannaV3: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 SurgeFF/AriannaV3:Q4_K_M
# Run inference directly in the terminal:
./llama-cli -hf SurgeFF/AriannaV3: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 SurgeFF/AriannaV3:Q4_K_M
# Run inference directly in the terminal:
./build/bin/llama-cli -hf SurgeFF/AriannaV3:Q4_K_M

Use Docker

docker model run hf.co/SurgeFF/AriannaV3:Q4_K_M

LM Studio
Jan

vLLM

How to use SurgeFF/AriannaV3 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "SurgeFF/AriannaV3"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SurgeFF/AriannaV3",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker

docker model run hf.co/SurgeFF/AriannaV3:Q4_K_M

SGLang

How to use SurgeFF/AriannaV3 with SGLang:

Install from pip and serve model

# Install SGLang from pip:
pip install sglang
# Start the SGLang server:
python3 -m sglang.launch_server \
    --model-path "SurgeFF/AriannaV3" \
    --host 0.0.0.0 \
    --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SurgeFF/AriannaV3",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Use Docker images

docker run --gpus all \
    --shm-size 32g \
    -p 30000:30000 \
    -v ~/.cache/huggingface:/root/.cache/huggingface \
    --env "HF_TOKEN=<secret>" \
    --ipc=host \
    lmsysorg/sglang:latest \
    python3 -m sglang.launch_server \
        --model-path "SurgeFF/AriannaV3" \
        --host 0.0.0.0 \
        --port 30000
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:30000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "SurgeFF/AriannaV3",
		"messages": [
			{
				"role": "user",
				"content": [
					{
						"type": "text",
						"text": "Describe this image in one sentence."
					},
					{
						"type": "image_url",
						"image_url": {
							"url": "https://cdn.britannica.com/61/93061-050-99147DCE/Statue-of-Liberty-Island-New-York-Bay.jpg"
						}
					}
				]
			}
		]
	}'

Ollama
How to use SurgeFF/AriannaV3 with Ollama:
```
ollama run hf.co/SurgeFF/AriannaV3:Q4_K_M
```

Unsloth Studio

How to use SurgeFF/AriannaV3 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 SurgeFF/AriannaV3 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 SurgeFF/AriannaV3 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for SurgeFF/AriannaV3 to start chatting

How to use SurgeFF/AriannaV3 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf SurgeFF/AriannaV3:Q4_K_M

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": "SurgeFF/AriannaV3:Q4_K_M"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use SurgeFF/AriannaV3 with Docker Model Runner:
```
docker model run hf.co/SurgeFF/AriannaV3:Q4_K_M
```

Lemonade

How to use SurgeFF/AriannaV3 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull SurgeFF/AriannaV3:Q4_K_M

Run and chat with the model

lemonade run user.AriannaV3-Q4_K_M

List all available models

lemonade list

Hermes Agent

How to use SurgeFF/AriannaV3 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf SurgeFF/AriannaV3: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 SurgeFF/AriannaV3:Q4_K_M

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use SurgeFF/AriannaV3 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf SurgeFF/AriannaV3: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 "SurgeFF/AriannaV3: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"

You need to agree to share your contact information to access this model

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

AriannaV3

Production-tested all-in-one local assistant for Surge's home AI fleet. AriannaV3 = gemma-4-12b-it + the Aria adapter-v17 QLoRA merged into the weights — one standalone, multimodal, encoder-free core. The published gguf/ quant ladder and the original adapter at adapter/ are provided for re-stacking.

What's new in V3

V3 ships the same v17 weights as V1 and V2. The difference is the surrounding runtime — AgentOS is now in production on the same fleet and is what gates every change.

AgentOS runtime (/home/surge/aria-v9/agentos) — append-only event log, ReAct executive, gated self-improvement, semantic memory with Hebbian strengthening, toolsmith + LoRA pipeline, repair module, realtime interruptible chat.
Per-capability regression floor — promotions are gated in code (harness.gate()), not by a hand-applied mean.
Suite fingerprint — stale comparisons are structurally impossible.
8-modules integration complete — AgentOS Runtime Core, Self-Scaffolding Loop, Endogenous Memory, Continuous Thought, Neuroevolution Layer, Self-Healing, Curiosity-Driven Exploration, Eval-Gated Pipeline (all shipped, verified live).
Active incumbent: adapter-v17, mean 0.864 on the gated set (math 0.90, tools 1.00, identity 0.90, memory 0.85).
Disjoint 150-problem math confirmation: 93.3%.

Architecture

The weights carry the core: text, reasoning, identity, memory, math, code, grammar, storytelling, tool-selection, safety, and vision-/audio-understanding. Every other modality is a sidecar the core orchestrates — realtime voice (Whisper STT / Piper TTS), video generation, image generation, retrieval embeddings — never baked into the weights.

The AgentOS runtime is the self-modifying brain that wraps this core: every change is reversible (LoRA-as-version), gated (eval-harness rule in code), and lineaged (append-only event log).

Files

Repo root — merged standalone weights (AutoModelForImageTextToText.from_pretrained).
gguf/ — GGUF quant ladder (F16 + Q2_K … Q8_0) for llama.cpp / Ollama / LM Studio.
adapter/ — the original Aria adapter-v17 LoRA (r=32, alpha=32, q/k/v/o/gate/up/down + lm_head + embed_tokens), to stack on the base yourself.

License

Gemma license, inherited from the base model. Built by Sergio Williams / Surge.

Downloads last month: 1

Safetensors

Model size

12B params

Tensor type

BF16

Model tree for SurgeFF/AriannaV3

Base model

google/gemma-4-12B

Finetuned

google/gemma-4-12B-it

Finetuned

unsloth/gemma-4-12b-it

Quantized

(11)

this model