Instructions to use Melechtna/Instella-3B-Instruct-GGUF with libraries, inference providers, notebooks, and local apps. Follow these links to get started.

Libraries

How to use Melechtna/Instella-3B-Instruct-GGUF with llama-cpp-python:

# !pip install llama-cpp-python

from llama_cpp import Llama

llm = Llama.from_pretrained(
	repo_id="Melechtna/Instella-3B-Instruct-GGUF",
	filename="Instella-3B-F16.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 Melechtna/Instella-3B-Instruct-GGUF with llama.cpp:

Install from brew

brew install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Melechtna/Instella-3B-Instruct-GGUF:F16
# Run inference directly in the terminal:
llama-cli -hf Melechtna/Instella-3B-Instruct-GGUF:F16

Install from WinGet (Windows)

winget install llama.cpp
# Start a local OpenAI-compatible server with a web UI:
llama-server -hf Melechtna/Instella-3B-Instruct-GGUF:F16
# Run inference directly in the terminal:
llama-cli -hf Melechtna/Instella-3B-Instruct-GGUF:F16

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 Melechtna/Instella-3B-Instruct-GGUF:F16
# Run inference directly in the terminal:
./llama-cli -hf Melechtna/Instella-3B-Instruct-GGUF:F16

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 Melechtna/Instella-3B-Instruct-GGUF:F16
# Run inference directly in the terminal:
./build/bin/llama-cli -hf Melechtna/Instella-3B-Instruct-GGUF:F16

Use Docker

docker model run hf.co/Melechtna/Instella-3B-Instruct-GGUF:F16

LM Studio
Jan
Ollama
How to use Melechtna/Instella-3B-Instruct-GGUF with Ollama:
```
ollama run hf.co/Melechtna/Instella-3B-Instruct-GGUF:F16
```

Unsloth Studio new

How to use Melechtna/Instella-3B-Instruct-GGUF 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 Melechtna/Instella-3B-Instruct-GGUF 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 Melechtna/Instella-3B-Instruct-GGUF to start chatting

Using HuggingFace Spaces for Unsloth

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

Docker Model Runner
How to use Melechtna/Instella-3B-Instruct-GGUF with Docker Model Runner:
```
docker model run hf.co/Melechtna/Instella-3B-Instruct-GGUF:F16
```

Lemonade

How to use Melechtna/Instella-3B-Instruct-GGUF with Lemonade:

Pull the model

# Download Lemonade from https://lemonade-server.ai/
lemonade pull Melechtna/Instella-3B-Instruct-GGUF:F16

Run and chat with the model

lemonade run user.Instella-3B-Instruct-GGUF-F16

List all available models

lemonade list

I had to modify the convert python and a few others in ollama to get the model converted, however, until the model is added to the backend, this model won't load. Beyond that, I have no idea about its performance, but given its size, should be simple enough to run. This is my first model conversion, so, forgive me if it's a tad rough, whenever it does get the needed backend.

I've added my modified version of llama.cpp, this currently contains the changes made to the conversion script, constants, and tensor mapping that were needed to perfom the conversion. This should have all the information needed to modify the loader, so that both the conversion and loader, loads the model 1:1 without compromises, however I don't know what to modify for the C code for the loader, so hopefully someone can take the adjustments I made to get it working the rest of the way.

Downloads last month: 35

GGUF

Model size

3B params

Architecture

instella

Hardware compatibility

16-bit

View +1 variant

Inference Providers NEW

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

Model tree for Melechtna/Instella-3B-Instruct-GGUF

Base model

amd/Instella-3B-Instruct

Quantized

(3)

this model