Deepseek with OpenWebUI

LAUNCH NOW WRITE A REVIEW
Please feel free to contact us
Go

About

DeepSeek is an AI development firm based in Hangzhou, China. The company was founded by Liang Wenfeng, a graduate of Zhejiang University, in May 2023. Wenfeng also co-founded High-Flyer, a China-based quantitative hedge fund that owns DeepSeek. Currently, DeepSeek operates as an independent AI research lab under the umbrella of High-Flyer. The full amount of funding and the valuation of DeepSeek have not been publicly disclosed.

Reinforcement learning. DeepSeek used a large-scale reinforcement learning approach focused on reasoning tasks.
Reward engineering. Researchers developed a rule-based reward system for the model that outperforms neural reward models that are more commonly used. Reward engineering is the process of designing the incentive system that guides an AI model’s learning during training.

Distillation. Using efficient knowledge transfer techniques, DeepSeek researchers successfully compressed capabilities into models as small as 1.5 billion parameters.

Emergent behavior network. DeepSeek’s emergent behavior innovation is the discovery that complex reasoning patterns can develop naturally through reinforcement learning without explicitly programming them.
  1. Type virtual machines in the search.
  2. Under Services, select Virtual machines.
  3. In the Virtual machines page, select Add. The Create a virtual machine page opens.
  4. In the Basics tab, under Project details, make sure the correct subscription is selected and then choose to Create new resource group. Type myResourceGroup for the name.*.
  5. Under Instance details, type myVM for the Virtual machine name, choose East US for your Region, and choose Ubuntu 18.04 LTS for your Image. Leave the other defaults.
  6. Under Administrator account, select SSH public key, type your user name, then paste in your public key. Remove any leading or trailing white space in your public key.
  7. Under Inbound port rules > Public inbound ports, choose Allow selected ports and then select SSH (22) and HTTP (80) from the drop-down.
  8. Leave the remaining defaults and then select the Review + create button at the bottom of the page.
  9. On the Create a virtual machine page, you can see the details about the VM you are about to create. When you are ready, select Create.

It will take a few minutes for your VM to be deployed. When the deployment is finished, move on to the next section.

Connect to virtual machine

Create an SSH connection with the VM.

  1. Select the Connect button on the overview page for your VM.
  2. In the Connect to virtual machine page, keep the default options to connect by IP address over port 22. In Login using VM local account a connection command is shown. Select the button to copy the command. The following example shows what the SSH connection command looks like:

ssh azureuser@<ip>

  1. Using the same bash shell you used to create your SSH key pair (you can reopen the Cloud Shell by selecting >_ again or going to https://shell.azure.com/bash), paste the SSH connection command into the shell to create an SSH session.

Getting Started with Deepseek with Open WebUI

After successfully connecting via SSH, you’re ready to set up Deepseek with Open WebUI. Here’s how to get everything running:

Key Components:

Ollama runs locally on port 11434
Open WebUI operates as a Docker container using port 3000

Quick Tips:

To exit the LLM interface and return to your terminal, press Ctrl + D
View your installed models anytime with:
$ ollama list

Step 2: Accessing Open WebUI

Open WebUI runs in a Docker container. To verify its status:

$ sudo docker ps

Note: The container might need a few minutes to initialize completely.

After installing your LLMs, access the WebUI interface at:

http://your_server_ip:3000

First-Time Setup:

On the login page, click “Sign Up” to create your credentials
Once logged in, select your preferred model from the dropdown menu
Performance depends on your VM’s specifications – consider upgrading your VM for better response times if needed.

Port Reference:

Ollama: TCP 11434 (Accessible at http://127.0.0.1:11434)
Open WebUI: TCP 3000
For Azure firewall configuration, consult the Azure Network Security Groups documentation.

Azure Network Security Groups

Submit Your Request

Captcha

Highlights

  • Reinforcement Learning-Based Training: DeepSeek-R1 employs a hybrid training approach combining reinforcement learning (RL) with cold-start data, enhancing its reasoning and readability.
  • Chain-of-Thought Reasoning: DeepSeek-R1 utilizes a step-by-step reasoning process, improving the interpretability and accuracy of its responses
  • Multi-Agent Learning: It supports multi-agent interactions, enabling complex simulations and collaborative problem-solving.

Application Installed