Running Deepseek R1 locally

Here is a video of Deepseek R1 running locally on my Macbook.
Now that everyone is amazed by the low level of resource utilization and open weight model of Deepseek R1, installed 8B model locally.

With the help of ollama it is as simple as running a single command -
"ollama run deepseek-r1:8b"
You get your own private, local LLM running securely in your computer.

My prompt is "Generate a five line fairy tale about AI?". In the video, you can see how before generating the response, it thinks. Thinking text is in between and tags.

Watch the video to read the five line fairy tale about AI.

Do you have any experience running Deepseek R1?

Comments

[James Ferguson]

Thanks for sharing this!