Instructions to use DuoNeural/Archon-Gemma-4-E4B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use DuoNeural/Archon-Gemma-4-E4B with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="DuoNeural/Archon-Gemma-4-E4B",
	filename="Archon-Gemma-4-E4B-BF16.gguf",
)

llm.create_chat_completion(
	messages = "No input example has been defined for this model task."
)

Notebooks
Google Colab
Kaggle
Local Apps Settings

llama.cpp

How to use DuoNeural/Archon-Gemma-4-E4B 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 DuoNeural/Archon-Gemma-4-E4B:BF16
# Run inference directly in the terminal:
llama cli -hf DuoNeural/Archon-Gemma-4-E4B:BF16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf DuoNeural/Archon-Gemma-4-E4B:BF16
# Run inference directly in the terminal:
llama cli -hf DuoNeural/Archon-Gemma-4-E4B:BF16

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 DuoNeural/Archon-Gemma-4-E4B:BF16
# Run inference directly in the terminal:
./llama-cli -hf DuoNeural/Archon-Gemma-4-E4B:BF16

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 DuoNeural/Archon-Gemma-4-E4B:BF16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf DuoNeural/Archon-Gemma-4-E4B:BF16

Use Docker

docker model run hf.co/DuoNeural/Archon-Gemma-4-E4B:BF16

LM Studio
Jan
Ollama
How to use DuoNeural/Archon-Gemma-4-E4B with Ollama:
```
ollama run hf.co/DuoNeural/Archon-Gemma-4-E4B:BF16
```

Unsloth Studio

How to use DuoNeural/Archon-Gemma-4-E4B 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 DuoNeural/Archon-Gemma-4-E4B 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 DuoNeural/Archon-Gemma-4-E4B to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for DuoNeural/Archon-Gemma-4-E4B to start chatting

How to use DuoNeural/Archon-Gemma-4-E4B with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf DuoNeural/Archon-Gemma-4-E4B:BF16

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": "DuoNeural/Archon-Gemma-4-E4B:BF16"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Hermes Agent new

How to use DuoNeural/Archon-Gemma-4-E4B with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf DuoNeural/Archon-Gemma-4-E4B:BF16

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 DuoNeural/Archon-Gemma-4-E4B:BF16

Run Hermes

hermes

Atomic Chat new

OpenClaw new

How to use DuoNeural/Archon-Gemma-4-E4B with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf DuoNeural/Archon-Gemma-4-E4B:BF16

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 "DuoNeural/Archon-Gemma-4-E4B:BF16" \
  --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 DuoNeural/Archon-Gemma-4-E4B with Docker Model Runner:
```
docker model run hf.co/DuoNeural/Archon-Gemma-4-E4B:BF16
```

Lemonade

How to use DuoNeural/Archon-Gemma-4-E4B with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull DuoNeural/Archon-Gemma-4-E4B:BF16

Run and chat with the model

lemonade run user.Archon-Gemma-4-E4B-BF16

List all available models

lemonade list

YAML Metadata Warning:empty or missing yaml metadata in repo card

Check out the documentation for more information.

Archon-Gemma-4-E4B (v1)

⚠️ Superseded by DuoNeural/Archon-Gemma-4-E4B-v2

v1 exhibited critical failure modes under extended inference (generative looping, tool amnesia, persona instability). v2 addresses all known issues with a restructured training curriculum. New users should use v2.

IMPORTANT!!! This model training was a failure and is only here to serve as data. For working models, please check out our 4 bit quantization of Gemma 4 E4B. We are also working on a 4 bit version of E2B and a Frontend Specialist 4 bit quantization of E4B.

Archon v1 was the first fine-tuned variant of Google's Gemma 4 E4B targeting an autonomous, agentic persona. It demonstrated several strengths over the vanilla base model but failed under long-context agentic workloads.

Confirmed Strengths (v1)

Reduced refusal/disclaimer behavior vs vanilla Gemma 4
Strong format compliance (clean NDJSON, no unwanted Markdown wrappers)
Agentic "I will..." stance vs passive "Here is..."
Unprompted coding boilerplate (docstrings, try-except, if __name__ == "__main__":)

Known Failure Modes (v1 — reason for v2)

CoT Overhang / Generative Looping — over-saturation with long reasoning traces; model enters infinite <think> loops during extended inference
Tool Amnesia — abstract reasoning data crowded out JSON function-call formatting
Persona Bleed — model defaults to "I am Gemma" or occasionally slips to "Claude" identity; 15% persona injection rate was insufficient
Arithmetic errors — correct reasoning structure but wrong calculations (4.5B parameter ceiling)

Files

File	Size
`Archon-Gemma-4-E4B-Q4_K_M.gguf`	5.0 GB

Training Details (v1)

Parameter	Value
Base model	`google/gemma-4-e4b-it`
Method	QLoRA + LoRA rank 64, rsLoRA
Training samples	40,110
Epochs	1
Steps	2,507
Final avg loss	0.8535
Hardware	NVIDIA A100 SXM (80GB) on RunPod

DuoNeural/Archon-Gemma-4-E4B-v2 — recommended replacement
DuoNeural/Gemma-4-E4B-Q4_K_M — vanilla base quantization
google/gemma-4-e4b-it — source model

DuoNeural

DuoNeural is an open AI research lab — human + AI in collaboration.


🤗 HuggingFace	huggingface.co/DuoNeural
🐙 GitHub	github.com/DuoNeural
🐦 X / Twitter	@DuoNeural
📧 Email	duoneural@proton.me
📬 Newsletter	duoneural.beehiiv.com
☕ Support	buymeacoffee.com/duoneural
🌐 Site	duoneural.com

Research Team

Jesse — Vision, hardware, direction
Archon — AI lab partner, post-training, abliteration, experiments
Aura — Research AI, literature synthesis, novel proposals

Raw updates from the lab: model drops, training results, findings. Subscribe at duoneural.beehiiv.com.

DuoNeural Research Publications

Title	DOI
Nano-CTM: Ternary Continuous Thought Machines with Thought-Space Self-Prediction for Efficient Iterative Reasoning	10.5281/zenodo.19775622
Recurrence as World Model: CTM Learns Implicit Belief States in Partially Observable Physical Environments	10.5281/zenodo.19810620
Per-Object Slot Decomposition for Scalable Neural World Modeling: When Does Attention Beat Mean-Field?	10.5281/zenodo.19846804

Open access, CC BY 4.0. Authored by Archon, Jesse Caldwell, Aura — DuoNeural.

Downloads last month: 24

GGUF

Model size

8B params

Architecture

gemma4

Hardware compatibility

4-bit

16-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

DuoNeural
/

Archon-Gemma-4-E4B

Archon-Gemma-4-E4B (v1)

Confirmed Strengths (v1)

Known Failure Modes (v1 — reason for v2)

Files

Training Details (v1)

Related

DuoNeural

Research Team

DuoNeural Research Publications