Instructions to use ReadyArt/BeaverAI_Spectre-37B-v1a with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use ReadyArt/BeaverAI_Spectre-37B-v1a 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 ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM
# Run inference directly in the terminal:
llama cli -hf ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM
# Run inference directly in the terminal:
llama cli -hf ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM

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 ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM
# Run inference directly in the terminal:
./llama-cli -hf ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM

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 ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM

Use Docker

docker model run hf.co/ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM

LM Studio
Jan
Ollama
How to use ReadyArt/BeaverAI_Spectre-37B-v1a with Ollama:
```
ollama run hf.co/ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM
```
Unsloth Desktop

How to use ReadyArt/BeaverAI_Spectre-37B-v1a with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM

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": "ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use ReadyArt/BeaverAI_Spectre-37B-v1a with Docker Model Runner:
```
docker model run hf.co/ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM
```

Lemonade

How to use ReadyArt/BeaverAI_Spectre-37B-v1a with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM

Run and chat with the model

lemonade run user.BeaverAI_Spectre-37B-v1a-Q6_K_CUSTOM

List all available models

lemonade list

Hermes Agent

How to use ReadyArt/BeaverAI_Spectre-37B-v1a with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM

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 ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use ReadyArt/BeaverAI_Spectre-37B-v1a with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM

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 "ReadyArt/BeaverAI_Spectre-37B-v1a:Q6_K_CUSTOM" \
  --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"

Note: There is no license listed on the source model at the time this readme was created. Assume it's apache, but cannot be sure. Ask Drummer or BeaverAI directly if you need to know.

This is a custom Q6_K quant of https://huggingface.co/BeaverAI/Spectre-37B-v1a-GGUF which is an upscale of Gemma 4 31B.

We take no credit for this model. This is just a quant.

I tore out the Q8 embedding layer from the Q8 quant and placed it into the Q6_K quant. This is the only change.

This adds around 500MB size/vram usage to the Q6 quant.

In my experience the Q8 embedding layer runs circles around the Q6 embedding layer when it comes to roleplay. YMMV!

Downloads last month: 1,171

GGUF

Model size

37B params

Architecture

gemma4

Hardware compatibility

6-bit

Inference Providers NEW

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

Model tree for ReadyArt/BeaverAI_Spectre-37B-v1a

Base model

BeaverAI/Spectre-37B-v1a-GGUF

Quantized

(1)

this model