How to install LLaMA 3 locally on Windows 11

Most recently, Meta introduced Llama 3-a giant language model. It has many applications, such as answering your questions and helping with schoolwork and projects, among others. You can access it at any time, whether online or offline, by installing Llama 3 locally on your Windows 11 PC. Here is how you can install Llama 3 locally on a Windows 11/10 PC.

Approaches to Installation of Llama 3 on Windows 11 Desktop Installing Llama 3 using Python in a Windows 11 or 10 computer requires technical expertise. However, there are several workarounds that will actually enable you to install Llama 3 locally on a Windows 11 computer. Look, let me illustrate these approaches.

To install and run Llama 3, you have to run a couple of commands via Command Prompt on a Windows 11 PC. That said, you will be restricted to the command line version of it. Using its web UI takes more steps. I am going to walk you through these two ways.

If you are on Windows 11, you need to install Ollama first, before deploying Llama 3. Here are the steps: Go to the main website for Ollama.

Click the Download button, then select Windows from the dropdown selection menu.

You will see Download for Windows to your computer as an executable file. The easiest way to install Ollama is to run the downloaded executable file.

Restart your computer after the installation of Ollama. It should be running in silent mode. You will see it in your System Tray. From the menu on the homepage of Ollama, click "Models". All the types will appear before you.

Llama 3.1 has the following three options: 8B Part 70B Section 405B Because the last one is the largest parameter, you can't execute it on a low-end PC. There are two available parameters in Llama 3.2: 1B 3B First, you need to choose an appropriate version of Llama that fits your needs. For instance, to install the version of Llama 3.2, click on Llama 3.2. Upon clicking, choose the parameter you want to install in the dropdown, open the Command Prompt, and copy paste the command beside it.

For your convenience are the instructions on how to do both for the Llama 3.2 model. The following command installs the Llama 3.2 3B model: llama 3.2:3b olivama run The command for running the installation of the model in Llama 3.2 1B is as follows: llama 3.2:1b gallina First, open the Command Prompt. Then, as per your requirement, write any one of the above commands and hit Enter. It is going to take a good deal of time to download all the essential data. It depends on the speed of your Internet connection. In the Command Prompt, it displays a success message if it's done accordingly.

Just begin to type into your message, and the Llama 3.2 model will engage immediately. The instructions for installing the Llama 3.1 model can be found on the Ollama website.

To launch Llama 3.1 or 3.2 the second time around from your PC, execute the same command in Command Prompt.

One disadvantage of installing Llama 3 through CMD is that it does not save your chat history. It will remember conversations if installed on your local machine and will have a nicer UI. You can see how to do that in the approach below.

Run Llama 3's web interface on Windows 11

Compared to using Llama 3 in the command prompt window, it has a much nicer user experience when using it in a web browser and keeps the chat history. Here I will demonstrate how to install Llama 3 in your browser.

To run Llama 3 in the browser, one should install Docker and Llama 3 via Ollama. If not installed, as previously discussed, please install Llama 3 using Ollama. Docker can be downloaded and installed directly from the website.

Once installed, launch Docker and create an account. Docker won't start working until you've registered an account. Once you have created an account, log in using the Docker application. Set Docker as an icon in the system tray. Note: Please keep the Docker and Ollama apps open running at all times in the background. Otherwise, Llama 3 will not work in your browser.

Now open the Command Prompt and paste the following command into the prompt:

docker run -d -p 3000:8080 --add-host=host.docker.internal:host-gateway -v open-webui:/app/backend/data -name ghcr.io/open-webui/open-webui:main --restart always

This will take some time as the above command will download all the necessary resources, so please be patient. Once the command is complete, open Docker and select the Containers tab on the left. Note by default, a container will be created with the port number 3000:8080.

Press the 3000:8080 port number. It will open a new tab in your default web browser. To start working with Llama 3 inside your browser, you need to sign up and then log in. In case you see localhost:3000 in the address bar, it means that Llama 3 is being served on your computer locally. That does not need an active internet connection to work.

In the dropdown select which Llama chat model you would like to enable. Additional Llama 3 models can be installed through Ollama by following the instructions that will be presented. The model will then be available in your browser automatically.

You can see your entire conversation history on the left side of your screen. Simply close your browser window, when you are finished. To shut down Docker, open the application and click the Stop button. You can now close Docker.

After launching Ollama and Docker, let it complete its run in a few minutes. Now, open your web browser and click on the port forwarding in the Docker container to route to the localhost server. You can use Llama 3 from here onwards. Log into your account to turn on Llama 3.

We are really hoping this article was helpful for your question and that you get to set up LLaMA 3 on a local Windows 11 without much hassle.