Installing DeepSeek in a Proxmox LXC
HTML-код
- Опубликовано: 10 фев 2025
- This video will show you how to setup Deepseek in a Proxmox CT
Commands
apt update
apt upgrade -y
apt install curl -y
curl -fsSL ollama.com/ins... | sh
ollama pull deepseek-r1:8b
ollama run deepseek-r1:8b
Links
ollama.com/dow...
ollama.com/lib...
Thanks for sharing will try it! Great vid!
Have fun!
Thank for creating this video. I will try this out tonight
I hope it worked out for you!
@@virtualizeeverything This worked for me. Thank you!
Didn’t know it was this easy! Gonna try this out tonight.
I’d like to see the GPU version of this myself as I’m going to do the same pretty soon in my home lab rig.
Thank you for the comment hope you had fun setting it up on your server
I have learned a lot from your channel. Continue the great work!
I'm glad you find the content useful, thank you for the support!
I was truly waiting for your video on this. Awesome. Am not a geek but am I terest and I try things on my PC and you videos and instructions help me through these challenges I get myself into. Not sure if it's for fun or for pain. PS i run proxmox on an eBay hp elitedesk cheap box. I have spare gpus, intel arc a770 8gb and 16gb. Any chance you could include pass through for intel gpu and configuring it for deepseek ? I will be trying your instructions at the weekend. Thank you again for your videos
Ps. Subscribers could arrange an intel battlemage b580 12gb. Any subscribers please indicate your interest. Cheers.
Cool GPU Intel ARC with LMstudio and Vulkan i run Ai on 2 GPU Sooo cooool!!
I dont know if this setup will work with Intel it would be a fun rabbit hole to go down if i had the time. Thank you for the comment and support
@virtualizeeverything thank you Sir.
Great and simple explanation. Question: can we access via web or its CLI based ?
This setup is CLI I have seen someone doing a Web UI have not tired it myself.
Would love to know if Stable Diffusion on Proxmox without a dedicated GPU with a similar process using an xlc container is possible for people who have home labs with low end mini pcs?
you can run this in docker
How about USB neural network accelerators?
This might be possible in a VM but I have not ever found a way to pass USB to a LXC.
The biggest question to me is still if DeepSeek is spyware and sends all prompts directly to China. I really want to have privacy and even if I run it into my home lab, I have my doubts.
I can understand that. I deleted the LXC after the video