Cette page n'est pas encore disponible dans votre langue, vous la voyez dans sa version originale.
DeepSeek R1 has just been released, and its first reviews are simply stunning: it seems to be more performant than the o1 model offered by OpenAI (which costs $200 per month!).
However, to take full advantage of it, you still need to know how to install it correctly on your computer.
Fortunately, with Ollama, a manager specifically designed for artificial intelligence models (the famous LLMs for Large Language Model 🇺🇸), this task is within everyone's reach. Even if you are not an expert in command lines, this guide will accompany you step by step.
So, ready to dive into the world of DeepSeek R1? Follow this guide to install it easily and quickly on your computer.
Would you prefer to use the LLM offered by Meta, known as llama3? We have a complete article on its installation right here: see the article for Llama3.1.
However, be aware that DeepSeek R1 is much better than Llama3.1, coming in at 70 billion parameters. 😉
Why choose DeepSeek R1 and Ollama?
Before diving into the technical details, let's take a moment to understand why these two tools complement each other so well:
DeepSeek R1 is just as powerful, imagine having an artificial intelligence as competent as o1 from OpenAI but for free;
Ollama simplifies the installation of AI models. No need to juggle dozens of complex dependencies: everything is done with a few simple commands.
And above all: you will keep your data on your computer, nothing is sent over the internet.
Prerequisites for using DeepSeek
Before we begin, make sure your computer is ready. Here's what you need:
A computer running Windows, macOS, or Linux (all three are possible);
A fast and stable internet connection (yes, downloading the models can be very long sometimes 😬);
Ollama, installed on your computer. Not ready yet? No panic, we'll explain that in the next section.
Step 1: Install Ollama (if not already done)
First, download and install Ollama, the manager that will allow you to easily retrieve DeepSeek R1.
Download the version corresponding to your operating system (Windows, macOS, or Linux).
Follow the installation steps, which will vary slightly depending on your platform.
Once installed, open your terminal and type:
CONSOLE
ollama --version
If you see a version number: congrats! You're all set! We can continue!
Otherwise, verify your installation and feel free to leave a comment under this article so we can help you. 😉
Step 2: Download DeepSeek R1 with Ollama
Installation
Now that you have Ollama, it's time to download DeepSeek R1. Here's how we'll do it:
Open your terminal;
Type the following command: ollama pull deepseek-r1:7b;
Ollama will download and install DeepSeek R1 on your computer. Depending on the speed of your connection, this may take a few minutes.
There are several different versions of DeepSeek-R1. Make sure to choose the 7b version (7 billion parameters). The more parameters a version has, the more your computer will struggle to run it.
Les différentes versions de DeepSeek-R1
Verification
To verify that the model is correctly installed, run:
CONSOLE
ollama list
If DeepSeek-R1 appears in the list, everything is ready for the next step.
Step 3: Launch DeepSeek R1
Launching DeepSeek R1 is as simple as downloading it. Here's how to run your first test with the model:
CONSOLE
ollama run deepseek-r1
You can now interact with DeepSeek R1 directly from your terminal. Ask it questions or provide it with data, and the model will process your information!
Easy, right?
Step 4: Use DeepSeek R1 with your data
DeepSeek R1 is particularly useful when it comes to analyzing your own files. Here's an example of how to analyze a text document:
Prepare a file (e.g., my-file.txt) containing the data to be analyzed.
Use this command to pass the file as input to the model:
CONSOLE
ollama run deepseek-r1 --input /path/to/your/file.txt
The model will analyze the content to better respond to the needs you give it!
If you have multiple files, you can process them one by one or automate the process with a script.
Step 5: Manage and optimize your models
Keep Ollama and DeepSeek R1 up to date
To take advantage of the latest features and fixes, keep your tools up to date. You only need to use these commands:
CONSOLE
ollama updateollama pull deepseek-r1
The first command updates Ollama while the second keeps DeepSeek-R1 up to date as well.
Free up disk space
AI models like DeepSeek R1 can take up a lot of disk space. If you run out of space, remove unused models with:
CONSOLE
ollama rm model-name
FAQ
1. What to do if the installation fails?
Ensure you have a stable internet connection and that your version of Ollama is up to date. Also, consider disabling your VPN.
2. Does DeepSeek R1 work offline?
Yes, once downloaded, the model can be used offline.
Conclusion
Now you know how to install DeepSeek-R1 for free on your computer and, above all, how to do it securely!
If you liked this guide, don't hesitate to share it with your friends: we appreciate it! ❤️
Que vous essayiez de scaler votre start-up, de créer votre premier site internet ou de vous reconvertir en tant que développeur, Believemy est votre nouvelle maison. Rejoignez-nous, évoluez et construisons ensemble votre projet.
