With improvements in AI, it has become easier to field large language models out to communities and organizations, like Llama 3. These models have long consumed substantial computational resources and technical acumen, but Ollama abstracts away much of that complexity for a much more widespread audience. Ollama is a free and open-source local deployment package for various large language models, like Llama 3, that you can install on your own machine. So without further ado, this article will guide you on how to make Llama 3 run easily with Ollama to let you use next generation AI language models with ease.
First of all, download Ollama from their official website. Ollama supports MacOS and Windows (currently in preview) and Ubuntu. Go to the official Ollama website > click on the download link for your OS and then install it accordingly. For installation there is a standard installation procedure that is generally to download installer and run it and to follow instructions on screen.
After installing Ollama, the second step is open your terminal. The terminal is a text based interface that allows you to interact with your computer using commands given in text form. If you are using MacOS you can get this in your Applications directory under a folder called Utilites. If you are using Ubuntu, you can open the terminal with by pressing Ctrl + Alt + T, or you can use any other terminal tool if you are on windows, such as Command Prompt, PowerShell, Windows Subsystem for Linux (WSL) for a Unix-like terminal environment.
Using the terminal you can now download and run the Llama 3 Model using a simple command. Running “ollama run llama3” in your terminal would begin the download of the default 8B Ollama instruct model. This command will pick up the Ollama repository, download the model to your local machine, and make it ready to use. Downloading, in the beginning, might take a few minutes depending on your network speed since the models are a little big.
You can optionally add the tags if you want to use a specific version of Llama 3 models. The command to actually run an instruct model, such as the 70B model, using Ollama would be: “ollama run llama3:70b-instruct”. Ollama comes in different versions— select one that suits your requirements
Once you have downloaded and loaded the model, Ollama gives a terminal-like user interface to operate Llama 3. With the text input, you can enter text and see the returned text from the model in order for you to converse with the AI in real time. This simplification in interface means you can begin to use the capabilities of Llama 3 almost immediately, without having to browse through difficult setups or configurations.
While the terminal interface provided by Ollama is functional, some users would likely appreciate a more user-friendly, graphical interface, with a less cumbersome integration. To do this, Ollama and Open WebUI integration were done. Open WebUI is a self-hosted user interface run on docker that brings WebUI to the Docker environment, making the user-friendly environment similar to well-known AI chat platforms like ChatGPT.
Open WebUI needs Docker to be setup in your PC. Docker is a tool designed to make it easier to create, deploy, and run applications by using containers. After you install Docker, you can fetch and set up Open WebUI by referring to Open WebUI GitHub page or will read its official documentation.
Once Open WebUI is up and running, you can train and configure it to use the Llama 3 model provided by Ollama This integration helps you in interacting with the model with a front end browser-based GUI, resulting in ease and comfort in using the software, especially for those who are not using terminals that efficiently.
With a few simple steps, you now have Llama 3 running on your computer with Ollama! It also reduces the requirement of heavy hardware and tech skills, and hence helping all to get the benefit of top-notch AI models. Whether you use the basic terminal interface or bump up the experience with Open WebUI, Ollama offers a solid and easy-to-use deployment & interaction solution for large language models out of the box. Take advantage of AI and start to explore what Llamas and Ollama really can do.