Instructions to use Lambent/Goose-7.2B-G1j-20260914 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Notebooks
Google Colab
Kaggle
Local Apps Settings

How to use Lambent/Goose-7.2B-G1j-20260914 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 Lambent/Goose-7.2B-G1j-20260914:Q8_0
# Run inference directly in the terminal:
llama cli -hf Lambent/Goose-7.2B-G1j-20260914:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf Lambent/Goose-7.2B-G1j-20260914:Q8_0
# Run inference directly in the terminal:
llama cli -hf Lambent/Goose-7.2B-G1j-20260914: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 Lambent/Goose-7.2B-G1j-20260914:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf Lambent/Goose-7.2B-G1j-20260914: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 Lambent/Goose-7.2B-G1j-20260914:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Lambent/Goose-7.2B-G1j-20260914:Q8_0

Use Docker

docker model run hf.co/Lambent/Goose-7.2B-G1j-20260914:Q8_0

LM Studio
Jan
Ollama
How to use Lambent/Goose-7.2B-G1j-20260914 with Ollama:
```
ollama run hf.co/Lambent/Goose-7.2B-G1j-20260914:Q8_0
```
Unsloth Desktop

How to use Lambent/Goose-7.2B-G1j-20260914 with Pi:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Lambent/Goose-7.2B-G1j-20260914: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": "Lambent/Goose-7.2B-G1j-20260914:Q8_0"
        }
      ]
    }
  }
}

Run Pi

# Start Pi in your project directory:
pi

Docker Model Runner
How to use Lambent/Goose-7.2B-G1j-20260914 with Docker Model Runner:
```
docker model run hf.co/Lambent/Goose-7.2B-G1j-20260914:Q8_0
```

Lemonade

How to use Lambent/Goose-7.2B-G1j-20260914 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Lambent/Goose-7.2B-G1j-20260914:Q8_0

Run and chat with the model

lemonade run user.Goose-7.2B-G1j-20260914-Q8_0

List all available models

lemonade list

Hermes Agent

How to use Lambent/Goose-7.2B-G1j-20260914 with Hermes Agent:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Lambent/Goose-7.2B-G1j-20260914: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 Lambent/Goose-7.2B-G1j-20260914:Q8_0

Run Hermes

hermes

Atomic Chat

OpenClaw

How to use Lambent/Goose-7.2B-G1j-20260914 with OpenClaw:

Start the llama.cpp server

# Install llama.cpp:
brew install llama.cpp
# Start a local OpenAI-compatible server:
llama serve -hf Lambent/Goose-7.2B-G1j-20260914: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 "Lambent/Goose-7.2B-G1j-20260914: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"

Task vectors from training on G1i mostly applied cleanly to G1j; publishing merged version here plus a touchup focusing on consolidating tool use into the chosen paradigm.

(Includes several further adapters not currently public. The 'adapter' folder is purely the touchup and does not need to be applied.)

This model should be considered an In Progress checkpoint. They are not strictly a base model, though I have included an auxiliary loss reward to preserve language modeling. However, they will not have the full capabilities of most post-trained models.

Also, RNNs have different natural capabilities than Transformers, and some tasks will be harder to learn to carry in state than attention.

Downloads last month: 113

Safetensors

Model size

7B params

Tensor type

BF16

Inference Providers NEW

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

Model tree for Lambent/Goose-7.2B-G1j-20260914

Base model

RWKV/RWKV7-G1j-7.2B-20260831

Quantized

(1)

this model

Quantizations

1 model