Update Readme (#294)

* update readme

* update picure name

* update issue link

* update link

* update links

* update

* update pictures

Author: zzzqc1234

cczoo/confidential_ai/README.md

@@ -65,7 +65,7 @@ A cloud-based service that verifies the proofness of the remote model serving en
 
 | Component                  | Version       | Purpose                                                                                                   |
 | -------------------------- | ------------- | --------------------------------------------------------------------------------------------------------- |
-| **Ollama**                 |               | Framework for running language models on confidential VMs                                                 |
+| **Ollama**                 |  `v0.5.7`     | Framework for running language models on confidential VMs                                                 |
 | **DeepSeek-R1**            |               | High performance reasoning model for inference service                                                    |
 | **open-webui**             | `v0.5.20`     | Self-hosted AI interface for user-interaction, running on the same confidential VM to simplify deployment |
 | **Cofidential AI(cc-zoo)** |               | Patches and compoents from cc-zoo                                                                         |
@@ -80,18 +80,188 @@ Here we use deepseek-llm-7b-chat model, please refer to the [guide](https://www.
 Please refer to [ollama installation guide](https://github.com/ollama/ollama/blob/main/docs/linux.md).
 
 ### 4.3 Build openwebui
-xxx
-xxx
+4.3.1 System Requirements
+- **Operating System**: Linux 
+- **Python Version**: Python 3.11+
+- **Node.js Version**: 20.18+
+
+4.3.2 Development Setup Instruction
+ 4.3.2.1 Clone the Repository
+```bash
+git clone https://github.com/your-org/open-webui.git  #Replace with actual repository URL(git apply xxx.patch  添加openwebUI对TDX的支持.)
+cd open-webui
+```
+
+ 4.3.2.2 Install Node.js
+   - Ensure Node.js ≥20.18.1 is installed:
+```bash
+# Install Node Version Manager (n)
+sudo npm install -g n
+
+# Install specific Node.js version
+sudo npm install 20.18.1
+
+# if meet any connect error, you can use follow way to install
+# install nvm(Node Version Manager)
+curl -o- https://raw.githubusercontent.com/nvm-sh/nvm/v0.39.1/install.sh | bash
+
+### install node with verison ID
+nvm install 20.18.1
+
+### if need change node version
+nvm use 20.18.1
+```
+4.3.3 Install Miniconda
+ - Download and install Miniconda:
+```bash
+wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
+bash Miniconda3-latest-Linux-x86_64.sh
+### while installing,you can skip reading install information by enter q, and set default choice to finish installation
+
+```
+4.3.3.1  Configure environment paths:
+```bash
+# Add Miniconda to PATH (replace /path/to/ with actual installation path)
+export PATH="/path/to/miniconda3/bin:$PATH"   ### defaoult path is: /root/miniconda3/bin
+
+# Initialize Conda
+conda init
+source ~/.bashrc
+
+# Verify installation
+conda --version
+```
+
+4.3.4 Frontend Build and Test
+
+ 4.3.4.1  Enter open-webui & Create a `.env` file:
+
+  ```bash
+  cd open-webui
+  cp -RPp .env.example .env
+  ```
+
+ 4.3.4.2  Update Ollama Serving Address in `.env` and Modify the `.env` file to configure the **Ollama backend URL**. This ensures that requests to `/ollama` are correctly redirected to the specified backend:
+
+```ini
+# Ollama URL for the backend to connect
+OLLAMA_BASE_URL='http://ip_address:port' 
+
+# OpenAI API Configuration (Leave empty if not used)
+OPENAI_API_BASE_URL=''
+OPENAI_API_KEY=''
+
+# AUTOMATIC1111 API (Uncomment if needed)
+# AUTOMATIC1111_BASE_URL="http://localhost:7860"
+
+# Disable Tracking & Telemetry
+SCARF_NO_ANALYTICS=true
+DO_NOT_TRACK=true
+ANONYMIZED_TELEMETRY=false
+```
+Ensure you replace `ip_address:port` with the actual IP address and port of your **Ollama server** if necessary.
+
+ 4.3.4.3 Build frontend server(if error occured,please goto [here](#issue_note)):
+    
+  ```bash
+  npm run build
+  ```
++ After building the frontend, copy the generated `build` directory to the backend and rename it to `frontend`:
+    
+    ```bash
+   cp -r build ./backend/open-webui/frontend
+
+    ```
+ 4.3.4.4 Backend Build and Setup
+
+- Navigate to the backend:
+    
+    ```bash
+    cd backend
+    ```
+    
+- Use **Conda** for environment setup:
+    
+    ```bash
+    conda create --name open-webui python=3.11
+    conda activate open-webui
+    ```
 
+ 4.3.4.5 Install python dependencies([Tips](#tips)):
+    
+  ```bash
+  pip install -r requirements.txt -U
+  ```
+
+ 4.3.4.6.1 Install TDX-quote_parse-feature enable:
+
+  ```bash
+  cd quote_generator
+  python setup.py install
+  ```
 ### 4.4 Run openwebui
 - Run ollama + AI model
-xxx
-- Run openwebui
-xxx
+  ```bash
+     ollama run xxxx(model name)
+  ```
 
 - Configure `Attestation Service`
-xxx
+  Build setps:
+  ```bash
+  cd confidential_ai/attestation_service/ && ./build.sh
+  ```
 - Check Attestation status
+  ```bash
+  ./attest_service
+  ```
+  It will start the service and wait for connection: "Starting TDX Attestation Service on port 8443..."
+
+
+- Run openwebui
+
+  1.open backend service
+  ```bash
+  conda create --name open-webui python=3.11
+  conda activate open-webui
+  cd /path/to/open-webui/backend/ && ./dev.sh
+  ```
+  ![backend service](./images/openwebui-backend.png)
+
+  2.open frontend service
+
+  ```bash
+  cd utilities/tdx/restful_as/restful_tdx_att_service && ./attest_service
+  ```
+  ![backend service](./images/openwebui-fronted.png)
+
+  3.open browser and goto address: https://{ip_address}:18080/(The ip address is your server ip)
+
+  4.Example:
+    get quote data and parse
+    
+    ![backend service](./images/parse.png)
+
+### <h2 id="issue_note">IssueNote：</h2>
+ - While building, meet with `Cannot find package `,you can try command:
+
+ ```bash
+ npm install pyodide
+ ```
+
+### <h2 id="tips">Tips：</h2>
+ - Downloading packages from remote sites can be slow. To speed up the   process, you can specify a local mirror such as **Aliyun** when installing packages:
+
+ ```bash
+ pip install torch -i https://mirrors.aliyun.com/pypi/simple/
+ ```
+
+ Alternatively, you can set Aliyun as the default mirror by adding the following lines to `~/.pip/pip.conf`, suggest to this method:
+
+ ```ini
+ [global]
+index-url = https://mirrors.aliyun.com/pypi/simple/
+ ```
+
 
 
 ### Prerequisites: