Instructions to use MoonieJeon/COSMOS-3B with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use MoonieJeon/COSMOS-3B with Cosmos:

# No code snippets available yet for this library.

# To use this model, check the repository files and the library's documentation.

# Want to help? PRs adding snippets are welcome at:
# https://github.com/huggingface/huggingface.js

llama-cpp-python

How to use MoonieJeon/COSMOS-3B with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="MoonieJeon/COSMOS-3B",
	filename="cosmos-3b.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 MoonieJeon/COSMOS-3B 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 MoonieJeon/COSMOS-3B
# Run inference directly in the terminal:
llama cli -hf MoonieJeon/COSMOS-3B

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama serve -hf MoonieJeon/COSMOS-3B
# Run inference directly in the terminal:
llama cli -hf MoonieJeon/COSMOS-3B

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 MoonieJeon/COSMOS-3B
# Run inference directly in the terminal:
./llama-cli -hf MoonieJeon/COSMOS-3B

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 MoonieJeon/COSMOS-3B
# Run inference directly in the terminal:
./build/bin/llama-cli -hf MoonieJeon/COSMOS-3B

Use Docker

docker model run hf.co/MoonieJeon/COSMOS-3B

LM Studio
Jan
Ollama
How to use MoonieJeon/COSMOS-3B with Ollama:
```
ollama run hf.co/MoonieJeon/COSMOS-3B
```

Unsloth Studio

How to use MoonieJeon/COSMOS-3B 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 MoonieJeon/COSMOS-3B 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 MoonieJeon/COSMOS-3B to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for MoonieJeon/COSMOS-3B to start chatting

Atomic Chat new
Docker Model Runner
How to use MoonieJeon/COSMOS-3B with Docker Model Runner:
```
docker model run hf.co/MoonieJeon/COSMOS-3B
```

Lemonade

How to use MoonieJeon/COSMOS-3B with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull MoonieJeon/COSMOS-3B

Run and chat with the model

lemonade run user.COSMOS-3B-{{QUANT_TAG}}

List all available models

lemonade list

COSMOS-3B

COSMOS-3B is a sovereign language model optimized for structured writing and professional assistance. It is designed for seamless offline execution, ensuring complete data privacy and sovereignty.

Technical Specifications

Base Model: Phi-3-Mini (3.8B)
Quantization: Q4_K_M
Format: GGUF (Optimized for LLamaSharp)
Primary Goal: Structured content generation (AIDA framework).

Why COSMOS-3B?

COSMOS-3B is the core inference engine of the Aether Project, built to address three critical needs:

Resilience: Fully operational without internet dependency.
Privacy: On-device processing ensures that user data never leaves the local environment.
Interoperability: Standardized GGUF format for cross-platform compatibility.

Compatibility

COSMOS-3B is provided in GGUF format. It is compatible with all standard GGUF-compliant inference engines, including (but not limited to) LLamaSharp (C#), llama.cpp (C++), and various PHP-based LLM frameworks.

License

This model is distributed under the MIT license. This ensures maximum flexibility and transparency for your projects.

Downloads last month: 180

GGUF

Model size

4B params

Architecture

phi3

Hardware compatibility

We're not able to determine the quantization variants.

View all variants

Inference Providers NEW

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

Model tree for MoonieJeon/COSMOS-3B

Base model

microsoft/Phi-3-mini-4k-instruct

Quantized

(176)

this model