开启open-webui:一键集成ollama,打造革命性GPT体验
Summary
TLDRThe video introduces the open Web UI project, a multimodal AI web client software with capabilities for document handling, local RAG integration, web browsing, voice input, AI painting, and model conversations. It demonstrates the installation process of the ollama Web version and open Web UI, including the setup of Docker and Pinokio for a seamless experience. The software's functionality is showcased through interactive examples, including chat, document analysis, and multimodal model responses to images and voice.
Takeaways
- 💻 The script introduces 'Ollama Web UI', a project that has evolved from 'Ollama-Web UI' and is now called 'Open Web UI'.
- 🔧 Ollama Web UI supports multimodality, including document handling, local RAG integration, web browsing, voice input, AI painting, and multiple model dialogues.
- 📥 It also offers OpenAI API integration, external Ollama server connection, and backend reverse proxy support, making it a versatile chat client software.
- 🛠️ The script provides a step-by-step guide on how to download and install Ollama, including dealing with antivirus software that might flag it as a virus.
- 📈 The Ollama software has been updated to version 0.1.28, and the installation process involves navigating to a specific directory and running commands in CMD.
- 🔗 The script mentions the need for downloading models, such as the 'gemma:7b' model, which can be initiated if not already present on the user's system.
- 🌐 For Open Web UI, the installation is recommended to be done via Docker, with separate frontend and backend components implemented in Node.js and Python, respectively.
- 📦 The script suggests using a software called 'Pinokio' to install Open Web UI if Docker installation is problematic.
- 🖼️ Open Web UI supports uploading and analyzing documents, as demonstrated by the script with a project manager checklist and the generation of a CSV file.
- 🎨 The software also supports multimodal capabilities, allowing users to analyze images and receive descriptions of their content.
- 🔊 Additionally, Ollama Web UI has text-to-speech functionality, as showcased by the script with a sample voice output of an image description.
Q & A
What is the new name of the project previously known as 'ollama-Web UI'?
-The project has been renamed to 'open Web UI'.
What are some of the key features of the open Web UI project?
-Key features include multimodal support, document handling, local RAG integration, web browsing capabilities, voice input support, AI painting, multiple model dialogues, OpenAI API integration, external Ollama server connection, and backend reverse proxy support.
What is the current version of the Ollama software mentioned in the script?
-The current version of the Ollama software mentioned is V0.1.28.
Why might the Ollama software installation be flagged as a virus by some antivirus programs?
-The installation might be flagged as a virus due to its nature or certain components that are not recognized by the antivirus software, prompting the user to temporarily disable antivirus protection during installation.
What is the approximate size of the Ollama software installation?
-The approximate size of the Ollama software installation is around 385 MB.
How can one start the Ollama software after installation?
-After installation, one can start the Ollama software by launching the desktop CMD command and entering the appropriate command to initiate the software.
What is the purpose of the 'Pinokio' software mentioned in the script?
-Pinokio is used to install the open Web UI project, as it provides an alternative method to install software when the user encounters issues with Docker Desktop.
What are the prerequisites for installing the open Web UI project using Pinokio?
-The prerequisites include the installation of conda, git, zip, Python, and Visual Studio (VS), which are automatically installed by Pinokio before proceeding with the open Web UI installation.
How does the open Web UI project handle document analysis and output?
-The open Web UI project can analyze documents and generate outputs in various formats, such as CSV, based on user prompts and the capabilities of the integrated models.
What is the multimodal capability of the Ollama software, and how is it demonstrated in the script?
-The multimodal capability of the Ollama software allows it to process and understand different types of data, such as images. In the script, it is demonstrated by the software's ability to analyze a photo and describe its content.
What additional features does the open Web UI project support besides document analysis and multimodal capabilities?
-In addition to document analysis and multimodal capabilities, the open Web UI project also supports features like text-to-speech (TTS) and potentially other AI-driven functionalities.
Outlines
😀 Introduction to Open Web UI and Ollama Web Version
The video script introduces the Ollama Web version, a multimodal client software that integrates various features such as document support, local RAG integration, web browsing, voice input, AI painting, and multiple model dialogues. It also supports OpenAI API integration, external Ollama server connections, and backend reverse proxy support. The project, formerly known as ollama-Web UI, has been renamed to open Web UI and is available in version V0.1.28. The script guides viewers through the installation process, which includes downloading the software, disabling antivirus software to avoid false positives, and navigating to the installation directory. It also mentions the need to download a model, such as the gemma:7b model, for the software to function properly.
😲 Exploring Open Web UI's Features and Installation Process
The second paragraph delves into the features of the Open Web UI project, highlighting its capabilities for chat, attachments, and multimodal support. It explains the process of using the project with a focus on the installation of the Pinokio software, which is used to install the open Web UI project due to issues with Docker Desktop on the presenter's computer. The script details the steps to download and install necessary software prerequisites like conda, git, zip, Python, and Visual Studio before proceeding with the installation of the open Web UI project. It also demonstrates the functionality of the web UI, including the ability to upload documents and receive responses in CSV format, as well as the integration of multimodal models for image recognition and text-to-speech capabilities. The video concludes by emphasizing the robustness of the open Web UI project compared to other similar tools like Chatbox.
Mindmap
Keywords
💡Ollama Web Version
💡Multimodal
💡RAG (Retrieval-Augmented Generation)
💡Web Browsing Function
💡Voice Input Support
💡AI Painting
💡Model Dialogue
💡OpenAI API Integration
💡External Ollama Server Connection
💡Backend Reverse Proxy Support
💡Pinokio
💡TTS (Text-to-Speech)
Highlights
Ollama Web version supports multimodality, including document handling, local RAG integration, and web browsing capabilities.
The Web version of Ollama offers voice input support, AI painting, and the ability to converse with multiple models.
Integration with OpenAI API and external Ollama servers is possible, along with backend reverse proxy support.
The project, previously known as Ollama-Web UI, has been renamed to Open Web UI, featuring a chat and attachment functionality.
Ollama has been updated to version V0.1.28, with installation instructions provided for Windows users.
Users may need to disable antivirus software temporarily during the installation of Ollama to avoid false positives.
The installation directory for Ollama is typically under C:/documents/APP date, with a size of approximately 385MB.
Ollama requires the downloading of models, such as the gemma model, which can be initiated through specific commands.
The open Web UI project is hosted on GitHub, with installation instructions available for Docker.
The open Web UI is composed of a frontend implemented in Node.js and a backend in Python.
Pinokio software is demonstrated for installing the open Web UI project due to issues with Docker Desktop on the presenter's computer.
The open Web UI integrates with mainstream projects like Lobo CHAT and provides a comprehensive setup guide.
After installation, users can access the Web UI through a provided website link.
The Web UI supports uploading attachments and documents, demonstrated with a project manager checklist.
The system can analyze documents and output information in CSV file format based on user prompts.
Multimodal capabilities of the Web UI are showcased with the ability to analyze and describe the content of images.
The Web UI also supports text-to-speech functionality, as demonstrated with a description of an image.
The open Web UI project is praised for its robust features, including RAG support, document handling, and multimodality, offering more capabilities than previously discussed Chatbox.
Transcripts
ollama web版本支持多模态
支持文档 、本地 RAG 集成、网页浏览功能
语音输入支持、AI绘画、多个模型对话
OpenAI API 集成
外部 Ollama 服务器连接
后端反向代理支持
基多功能一体的
chatgptweb客户端软件来了
今天给大家介绍一个项目
是ollama的Web版
它的早期项目叫ollama-Web UI
现在项目更名为open Web UI
它主要实现哪些功能呢
可以实现聊天
实现附件
包括我们现在的多模态的附件
就是聊天和附附件功能
然后也支持RAG的一个增强
那么下面给大家介绍下
这个项目怎么去用的
那我们首先回到上期视频当中
我们用到了一个 ollama
这个ollama的这个安装地址
在这个位置
目前他已经升到V0.1.28版本
我们找到一个Windows版
点击这个下载
因为我提前已经下载完成了
下载完成之后
这里面有一个这个地方
我们检查双击
因为现在目前的话
这个软件安装的时候
可能会认为是一个病毒
我们首先把杀毒软件给关闭掉
然后把360的也去掉
退出然后这时候我们进行点击安装
点下一步
它默认会装在c盘的document APP date
这个目录下面去
然后我们找一下安装目录
程序安装是在这个目录下面
然后大概是有100多兆
哦 385M
然后接下来
我们就可以启动桌面板的CMD命令
然后我们测一下
好这个版本是0.1.28木
已经安装好了
接下来我们就启动这个
ollama
我们还是以上期的谷歌的gemma为例
我们下载这个模型
我们复制这个命令然后
啊这时候我已经那个启起来了
因为之前我在电脑上
已经装过gemma:7b模型
所以它直接可以启起来
默认的情况下他没有模型的话
可以有一个下载过程
这个大家就稍等一下
这个就是我们的模型的地址
跟那个我们之前的docker文件差不多
然后它这里面有一堆文件
看就是这个
大家看到了
这就是相当于镜像
而且模型镜像做成一个docker模式
刚才ollama已经给大家安装过了
接下来
我们看一下这个open webui的安装
open UI的项目地址是在github这个位置
然后
然后它的安装
官方安装是由docker
来方式来进行安装的
因为它分前端和后端
前端是nodejs
后端是Python实现的
然后因为我本地电脑上docker Desktop
这个有点问题
所以我这边就没有安装成功
接下来我们给大家使用一个叫
Pinokio来给大家进行安装这个东西
给大家演示一下Pinokio这个软件
安装这个项目
在github的这个这个项目的这个地址
然后我们点开这个官网
查官网网站打开之后
这里面有相关的说明
比如说这是一个油管的一个视频
然后下面呢是集成了
就是像open Web UI Lobo CHAT就是这些主流的
这这个github的一个项目
我们打开这个dowload
然后这里面会有Windows Mac
嗯然后Linux我们选Windows
然后点击下载
然后按照这个路步骤
就是安装就可以了
我这里面已经提前安装好了
这就不给大家演示
安装完成之后我们桌面上会有
打开软件之后
这里面有一个访问这个网站
然后这样的话
他一个就我们看到刚才网站上
Web,Web UI
lobe chat这里面都有相关的这些
我们只要在这里面安装好就行
包括ollama
对吧我们接下来我们下载
呃就先选这个吧
选中它之后
然后这里面是先下载dowload
这一块的话
首先它会把目前的conda,git
z i p
还有py还有vs等
这些软件先提前先安装
安装完之后
然后再来我们接下来看一下这个
点开这个
好了我们看窗口这边已经弹出来
就是第一个已安装完了
conda安装完
第二个是装 git
然后稍等片刻
我们看到这个已经安装vs了
然后这个弹出一个vs studio的
应该安装这个界面
这个稍等它一遍
下载安装
安装完成之后
这里面有一个安装complete click OK
我们点确定
好这时候就已经完成了
我们打开Pinokio
然后这里面进入一个
Pinokio的这些项目的这个地址
这个各种网站的地址
然后我们打开这个Web UI
然后这里面我们点击download下载
这时候它会保存save as 这个Web UI GT
然后下载下来这个
稍等片刻
好这个是已经下载完成了
然后它会进入到一个叫open Web UI的GT
这个项目
然后我们点开这个项目子模块
这里面有一个install
我们点开它
然后接下来就是对那个Pinokio
然后在github里面
下Web open UI的这个项目
然后只剩一个NPU的编译打包
我们稍等片刻
啊这一步就是安装完成
运行完成之后
我们点击start
然后这时候他们需要提示我装
安装啦ollama
这时候我们这边已经启起来了
我们点开这个
这个ollama
这个页面就已经打开了
我们注册一下
然后填入名字
我们确保这边的ollama模型启动
然后输入127.0.0.1:8000
这时候启起来
然后我们下拉选项
这里面有一个那个刚才的gemma
7b模型
然后我们可以聊天
好这个他已经写出来了
然后接下来我们再可以测一下这个
webui
它可以支持那种附件和文档上传
我们测一下
我们试一下这个项目经理清单
然后给它一个提示词比如说
请分析一下项目
进行清单
并将这个文件格式输出
CSV文件格式
把生成出来的CSV文件
格式与代码行输出
我们看下效果
好这时候他就按照我们刚才的要求
把这边这个文档解析出来了
然后生成姓名年龄所有代
然后是否有PMP
然后这是他的代码
这个代码我个人觉得他生成的
可能有点随机性
但是至少他能读懂
可能跟我们后端的那个gemma
模型是有关系的
这个webui还支持多模态
我们接下来在这官方上下载这个模型
这个是多模态的模型
我们点开它
然后找一个简单的1.6的这个模型吧
我们把这个装一下
然后我们开CMD窗口
这个我们下一下模型
稍等一下
好了这个模型已经下载完成了
我们看一下这
个
这时候就有一个ollama
llava的多模态模型
我们测一下
我们选择一张照片
我们问请这张照片的内容是什么
让他回答一下
好他已经回答出来
这张照片是一个关于长发的脸部表情
一个男姓
正是对面一种视觉的圆
这个就是这个Olama
它还支持一些其他的一些功能
比如说我们现在看到的声音
然后它的图像
图像可以直接AUTOMATIC1111
因为这个时间关键就不给大家演示了
然后声音这块我们给大家听一下
这面有个TTS
这张照片的内容
是
一张关于长发和脸部表情的男性画像
他正视镜头
给人以一种啊
总之这个这个Web就是open Web UI
这个项目功能还是非常强大的
它支持一些RAG
然后文档
然后多模态等功能
比之前我们说的那个Chatbox
可能功能更强大一些啊
本期的分享就到这里
全部结束了
谢谢大家的观看
تصفح المزيد من مقاطع الفيديو ذات الصلة
![](https://i.ytimg.com/vi/xsK4uATpxLY/hq720.jpg)
Всё о новой нейросети GPT-4o за 7 минут!
![](https://i.ytimg.com/vi/_-6LgEjEyzE/hq720.jpg)
The latest in Web UI (Google I/O ‘24)
![](https://i.ytimg.com/vi/o4mFRhAQEwI/hq720.jpg)
GPT-4o 사용법 총정리 - 무료 사용량, 신기능 활용
![](https://i.ytimg.com/vi/Sus8d4Ardyk/hq720.jpg)
BIG AI NEWS: 10,000X Bigger Than GPT-4, AGI 2025, New Boston Dynamics Demo And More
![](https://i.ytimg.com/vi/q9oyDypKqcw/hq720.jpg?sqp=-oaymwEmCIAKENAF8quKqQMa8AEB-AH-CYAC0AWKAgwIABABGGUgTig7MA8=&rs=AOn4CLAs9PD2SrfLz8bF3V55oEu_MfqqSw)
GPT 4o - Deep Dive Review - AGI? - ChatGPT massive improvements
![](https://i.ytimg.com/vi/ds5LLIt5OLM/hq720.jpg?v=663cfaa0)
RIP ELEVENLABS! Create BEST TTS AI Voices LOCALLY For FREE!
5.0 / 5 (0 votes)