Instructions to use georgeanton/alice-m5-cortex-8b-6.3gb with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use georgeanton/alice-m5-cortex-8b-6.3gb 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 georgeanton/alice-m5-cortex-8b-6.3gb
# Run inference directly in the terminal:
llama cli -hf georgeanton/alice-m5-cortex-8b-6.3gb

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf georgeanton/alice-m5-cortex-8b-6.3gb
# Run inference directly in the terminal:
llama cli -hf georgeanton/alice-m5-cortex-8b-6.3gb

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 georgeanton/alice-m5-cortex-8b-6.3gb
# Run inference directly in the terminal:
./llama-cli -hf georgeanton/alice-m5-cortex-8b-6.3gb

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 georgeanton/alice-m5-cortex-8b-6.3gb
# Run inference directly in the terminal:
./build/bin/llama-cli -hf georgeanton/alice-m5-cortex-8b-6.3gb

Use Docker

docker model run hf.co/georgeanton/alice-m5-cortex-8b-6.3gb

LM Studio
Jan
Ollama
How to use georgeanton/alice-m5-cortex-8b-6.3gb with Ollama:
```
ollama run hf.co/georgeanton/alice-m5-cortex-8b-6.3gb
```
Unsloth Desktop

How to use georgeanton/alice-m5-cortex-8b-6.3gb with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf georgeanton/alice-m5-cortex-8b-6.3gb

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": "georgeanton/alice-m5-cortex-8b-6.3gb"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use georgeanton/alice-m5-cortex-8b-6.3gb with Docker Model Runner:
```
docker model run hf.co/georgeanton/alice-m5-cortex-8b-6.3gb
```

Lemonade

How to use georgeanton/alice-m5-cortex-8b-6.3gb with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull georgeanton/alice-m5-cortex-8b-6.3gb

Run and chat with the model

lemonade run user.alice-m5-cortex-8b-6.3gb-{{QUANT_TAG}}

List all available models

lemonade list

Hermes Agent

How to use georgeanton/alice-m5-cortex-8b-6.3gb with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf georgeanton/alice-m5-cortex-8b-6.3gb

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 georgeanton/alice-m5-cortex-8b-6.3gb

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use georgeanton/alice-m5-cortex-8b-6.3gb with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf georgeanton/alice-m5-cortex-8b-6.3gb

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 "georgeanton/alice-m5-cortex-8b-6.3gb" \
  --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"

alice-m5-cortex-8b-6.3gb

File size: 1,643 Bytes

---
license: apache-2.0
tags:
  - sifta
  - ollama
  - gemma4
  - multimodal
  - local-ai
library_name: ollama
---

# Alice M5 Cortex 8B 6.3GB

Primary M5 SIFTA cortex tag:

```bash
ollama create alice-m5-cortex-8b-6.3gb:latest -f Modelfile
```

This is the promoted M5 cortex used by Alice on the Foundry node. It is a
SIFTA-owned Ollama tag and does not use the retired LoRA candidate.

## Live Probe

Verified on 2026-05-09:

- architecture: Gemma4
- parameters: 8B
- context length: 131072
- runtime context: 8192
- capabilities reported by `ollama show`: completion, vision, audio, tools,
  thinking
- `/api/chat` with `think:false` answered both text and image prompts

## SIFTA Runtime Note

Alice's Talk path uses `/api/chat` with `think:false`. Raw calls that omit this
can spend the whole output budget in the thinking field and return blank
assistant content.

## SIFTA Field Breakthrough

The current SIFTA public repo includes a stigmergic field breakthrough brief:
https://github.com/antonpictures/ANTON-SIFTA/blob/main/Documents/CARLTON_STIGMERGIC_FIELD_BREAKTHROUGH_2026-05-11.md

The cortex is only one organ in that system. The field mechanism itself runs in
the Python body: Bell analogue simulator, kernel scheduler, and hippocampus.
Credit boundary: Bell/CHSH/Hall/pilot-wave/stigmergy literature grounds the
analogy; SIFTA claims a receipt-backed classical contextual analogue, not a
proof of the physical cause of quantum nonlocality.

## Local State Boundary

This model package is public species DNA. It must not include a node's raw
`.sifta_state/` memory, contacts, owner traces, camera frames, or private
receipts.