Instructions to use KarlSage/oraculum_alpha_08b with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use KarlSage/oraculum_alpha_08b 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 KarlSage/oraculum_alpha_08b:Q8_0
# Run inference directly in the terminal:
llama cli -hf KarlSage/oraculum_alpha_08b:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf KarlSage/oraculum_alpha_08b:Q8_0
# Run inference directly in the terminal:
llama cli -hf KarlSage/oraculum_alpha_08b:Q8_0

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 KarlSage/oraculum_alpha_08b:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf KarlSage/oraculum_alpha_08b:Q8_0

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 KarlSage/oraculum_alpha_08b:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf KarlSage/oraculum_alpha_08b:Q8_0

Use Docker

docker model run hf.co/KarlSage/oraculum_alpha_08b:Q8_0

LM Studio
Jan

vLLM

How to use KarlSage/oraculum_alpha_08b with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "KarlSage/oraculum_alpha_08b"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/chat/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "KarlSage/oraculum_alpha_08b",
		"messages": [
			{
				"role": "user",
				"content": "What is the capital of France?"
			}
		]
	}'

Use Docker

docker model run hf.co/KarlSage/oraculum_alpha_08b:Q8_0

Ollama
How to use KarlSage/oraculum_alpha_08b with Ollama:
```
ollama run hf.co/KarlSage/oraculum_alpha_08b:Q8_0
```

Unsloth Studio

How to use KarlSage/oraculum_alpha_08b 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 KarlSage/oraculum_alpha_08b 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 KarlSage/oraculum_alpha_08b to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for KarlSage/oraculum_alpha_08b to start chatting

How to use KarlSage/oraculum_alpha_08b with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf KarlSage/oraculum_alpha_08b: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": "KarlSage/oraculum_alpha_08b:Q8_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use KarlSage/oraculum_alpha_08b with Docker Model Runner:
```
docker model run hf.co/KarlSage/oraculum_alpha_08b:Q8_0
```

Lemonade

How to use KarlSage/oraculum_alpha_08b with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull KarlSage/oraculum_alpha_08b:Q8_0

Run and chat with the model

lemonade run user.oraculum_alpha_08b-Q8_0

List all available models

lemonade list

Hermes Agent

How to use KarlSage/oraculum_alpha_08b with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf KarlSage/oraculum_alpha_08b: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 KarlSage/oraculum_alpha_08b:Q8_0

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use KarlSage/oraculum_alpha_08b with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf KarlSage/oraculum_alpha_08b: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 "KarlSage/oraculum_alpha_08b: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"

oraculum_alpha_08b / README.md

KarlSage

Super-squash branch 'main' using huggingface_hub

ffa9e09 6 days ago

preview code

Raw

History Blame Contribute Delete

1.83 kB

metadata

license: apache-2.0
base_model: Qwen/Qwen3.5-0.8B
pipeline_tag: text-generation
language:
  - en
  - pt
tags:
  - reasoning
  - structured-reasoning
  - gguf
  - experimental

Oraculum Alpha 0.8B

Experimental public preview. A 0.8B-parameter reasoning model fine-tuned from Qwen3.5-0.8B.

Instead of free-form chain-of-thought, Oraculum writes its reasoning as a structured XML trace, discrete steps with explicit operations and dependencies, inside its thinking channel, and only then gives the final answer. The structured trace is the research subject of this release.

<thoughtml>
  <step id="s1" op="observe">The user asks for the sum of 2 and 2.</step>
  <step id="s2" op="deduce" depends_on="s1">2 + 2 = 4.</step>
  <step id="s3" op="answer" depends_on="s2">The answer is 4.</step>
</thoughtml>
The answer is 4.

Files

file	format	use
`model.safetensors`	bf16	transformers / fine-tuning
`oraculum_alpha_08b-bf16.gguf`	GGUF bf16	llama.cpp, full precision
`oraculum_alpha_08b-Q8_0.gguf`	GGUF Q8_0	llama.cpp, recommended

Run it (llama.cpp)

llama-server -m oraculum_alpha_08b-Q8_0.gguf -c 8192
# or
llama-cli -m oraculum_alpha_08b-Q8_0.gguf -cnv

Use the built-in chat template. The reasoning trace appears in the thinking channel; the visible reply comes after it.

Notes and limitations

Alpha. This is a quick public test of the format, not a polished assistant. Expect mistakes, it is a 0.8B model.
Reasoning is strongest in math, logic and step-by-step analysis; Portuguese and English are the primary languages.
The XML trace is machine-parseable by design; malformed traces can still occur under long generations.

License

Apache 2.0, same as the base model.