Instructions to use ddh0/rocket-3B-GGUF-fp16 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use ddh0/rocket-3B-GGUF-fp16 with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="ddh0/rocket-3B-GGUF-fp16",
	filename="rocket-3B-fp16.gguf",
)

output = llm(
	"Once upon a time,",
	max_tokens=512,
	echo=True
)
print(output)

Notebooks
Google Colab
Kaggle
Local Apps

llama.cpp

How to use ddh0/rocket-3B-GGUF-fp16 with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf ddh0/rocket-3B-GGUF-fp16:Q8_0
# Run inference directly in the terminal:
llama-cli -hf ddh0/rocket-3B-GGUF-fp16:Q8_0

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf ddh0/rocket-3B-GGUF-fp16:Q8_0
# Run inference directly in the terminal:
llama-cli -hf ddh0/rocket-3B-GGUF-fp16: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 ddh0/rocket-3B-GGUF-fp16:Q8_0
# Run inference directly in the terminal:
./llama-cli -hf ddh0/rocket-3B-GGUF-fp16: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 ddh0/rocket-3B-GGUF-fp16:Q8_0
# Run inference directly in the terminal:
./build/bin/llama-cli -hf ddh0/rocket-3B-GGUF-fp16:Q8_0

Use Docker

docker model run hf.co/ddh0/rocket-3B-GGUF-fp16:Q8_0

LM Studio
Jan

vLLM

How to use ddh0/rocket-3B-GGUF-fp16 with vLLM:

Install from pip and serve model

# Install vLLM from pip:
pip install vllm
# Start the vLLM server:
vllm serve "ddh0/rocket-3B-GGUF-fp16"
# Call the server using curl (OpenAI-compatible API):
curl -X POST "http://localhost:8000/v1/completions" \
	-H "Content-Type: application/json" \
	--data '{
		"model": "ddh0/rocket-3B-GGUF-fp16",
		"prompt": "Once upon a time,",
		"max_tokens": 512,
		"temperature": 0.5
	}'

Use Docker

docker model run hf.co/ddh0/rocket-3B-GGUF-fp16:Q8_0

Ollama
How to use ddh0/rocket-3B-GGUF-fp16 with Ollama:
```
ollama run hf.co/ddh0/rocket-3B-GGUF-fp16:Q8_0
```

Unsloth Studio new

How to use ddh0/rocket-3B-GGUF-fp16 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 ddh0/rocket-3B-GGUF-fp16 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 ddh0/rocket-3B-GGUF-fp16 to start chatting

Using HuggingFace Spaces for Unsloth

# No setup required
# Open https://huggingface.co/spaces/unsloth/studio in your browser
# Search for ddh0/rocket-3B-GGUF-fp16 to start chatting

Docker Model Runner
How to use ddh0/rocket-3B-GGUF-fp16 with Docker Model Runner:
```
docker model run hf.co/ddh0/rocket-3B-GGUF-fp16:Q8_0
```

Lemonade

How to use ddh0/rocket-3B-GGUF-fp16 with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull ddh0/rocket-3B-GGUF-fp16:Q8_0

Run and chat with the model

lemonade run user.rocket-3B-GGUF-fp16-Q8_0

List all available models

lemonade list

This is pansophic's rocket-3B, converted to GGUF. No other changes were made.

Two files are avaliable here:

rocket-3B-fp16.gguf: the original model converted to GGUF without quantization
rocket-3B-q8_0-LOT.gguf: the original model converted to GGUF with q8_0 quantization using the --leave-output-tensor command-line option

From llama.cpp/quantize --help:

--leave-output-tensor: Will leave output.weight un(re)quantized. Increases model size but may also increase quality, especially when requantizing

The model was converted using convert-hf-to-gguf.py from Georgi Gerganov's llama.cpp repo, commit #8e672ef.

All credit belongs to pansophic for training and releasing this model. Thank you!

Downloads last month: 8

GGUF

Model size

3B params

Architecture

stablelm

Hardware compatibility

8-bit

View +1 variant