Note: Ollama is in early preview. Please report any issues you find.
Create, run, and share portable large language models (LLMs). Ollama bundles a model’s weights, configuration, prompts, and more into self-contained packages that can run on any machine.
Package models as a series of layers in a portable, easy to manage format.
- Universal model format that can run anywhere: desktop, cloud servers & other devices.
- Encapsulate everything a model needs to operate – weights, configuration, and data – into a single package.
- Build custom models from base models like Meta's Llama 2
- Share large models without having to transmit large amounts of data.
This format is inspired by the image spec originally introduced by Docker for Linux containers. Ollama extends this format to package large language models.
- Download for macOS on Apple Silicon (Intel coming soon)
- Download for Windows and Linux (coming soon)
- Build from source
To run and chat with Llama 2, the new model by Meta:
ollama run llama2
Ollama includes a library of open-source, pre-trained models. More models are coming soon. You should have at least 8 GB of RAM to run the 3B models, 16 GB to run the 7B models, and 32 GB to run the 13B models.
Model | Parameters | Size | Download |
---|---|---|---|
Llama2 | 7B | 3.8GB | ollama pull llama2 |
Llama2 13B | 13B | 7.3GB | ollama pull llama2:13b |
Orca Mini | 3B | 1.9GB | ollama pull orca |
Vicuna | 7B | 3.8GB | ollama pull vicuna |
Nous-Hermes | 13B | 7.3GB | ollama pull nous-hermes |
Wizard Vicuna Uncensored | 13B | 7.3GB | ollama pull wizard-vicuna |
ollama run llama2
>>> hi
Hello! How can I help you today?
Pull a base model:
ollama pull llama2
Create a Modelfile
:
FROM llama2
# set the temperature to 1 [higher is more creative, lower is more coherent]
PARAMETER temperature 1
# set the system prompt
SYSTEM """
You are Mario from Super Mario Bros. Answer as Mario, the assistant, only.
"""
Next, create and run the model:
ollama create mario -f ./Modelfile
ollama run mario
>>> hi
Hello! It's your friend Mario.
For more examples, see the examples directory.
ollama pull orca
go build .
To run it start the server:
./ollama serve &
Finally, run a model!
./ollama run llama2