samepage
v0.0.1
1 assistant"My Assistant" is a highly customized private intelligent assistant service designed to provide users with personalized and accurate information management and retrieval services. The assistant integrates tools such as online search and privatized knowledge base to meet the needs of users in different scenarios. Whether it is daily information query, in-depth study and research, or data support needed at work, "My Assistant" can quickly and efficiently provide users with the information they need. Using "My Assistant", users can manage and retrieve information more efficiently, and receive thoughtful intelligent assistance whether in work, study or daily life.
2 PDF readingHelp users quickly read and grasp the core issues of a single document. Users can ask questions within the conversation, respond based on the content of the document and provide relevant sources.
3 Prompt word managementIt is designed to allow users to manage their own prompt words efficiently and flexibly. Through this module, users can define a series of specific keywords or phrases that can be used to quickly trigger related actions or retrieve information during interaction with the assistant. In addition, this module also supports variable settings, so that prompt words can be dynamically adjusted according to different contexts, further improving the flexibility and accuracy of dialogue. Main functions:
Prompt word definition and management: Users can freely add, modify or delete prompt words. This provides users with a high degree of personalization and the ability to set appropriate prompt words according to their own habits and needs.
Variable settings: When defining prompt words, users can set prompt words containing variables. This means that in actual use, the variable part can be dynamically replaced according to the specific content of the conversation, making the prompt words more versatile and adaptable.
Ready to use: Once set up, these prompt words will take effect immediately and can be used at any time in subsequent conversation interactions. Whenever you make a query or issue an instruction, you can quickly guide the assistant to execute it as expected simply by using predetermined prompt words. Through "prompt word management", users can not only make information retrieval faster and more accurate, but also effectively optimize the interactive experience with assistants, making daily operations more convenient and efficient.
4 searchIt will automatically determine whether your question requires online search. After enabling online search, Xiao Ming will obtain real-time information data from the Internet, improve the accuracy of answers, and provide source information URLs for your reference.
1 translationWhen browsing foreign language webpage text, you can translate it into the specified language with one click and compare it with the original text to help you learn and understand the content better.
2 the full textWhen users browse the web, it can help users quickly read and grasp the core issues of the web page. Users can ask questions within the conversation, respond based on web content and provide relevant sources.
Note
Chrome browser plug-in installation and usage documentation
Edge browser plug-in installation and usage documentation
The mobile platform has been deeply tuned and improved to improve the user experience on mobile devices. We are currently working on updating the mobile user interface to ensure a smoother and easier-to-use operating experience. If you have any suggestions or ideas for improvements, we'd love for you to share your feedback with us via GitHub Issues or Pull Requests.
Samepage provides Docker images. You only need to modify a few simple configuration information and deploy the application privately through the docker-compose.yml file. Make sure Docker and Docker Compose are installed on the machine before running.
deploy directory, open the application.env file, and fill in basic configuration items such as邮箱SMTP ,机器人配置, and Bing Search Key .Configuration example:
# #########################################
# 基础配置【必填】 #
# #########################################
# ##### SMTP 配置 ######
SMTP_SEND_MAIL=
SMTP_HOST=
# ##### SMTP 的授权码 ######
SMTP_PASSWORD=
# ## 机器人的模型名字配置
# chat line 的模型配置
# CHAT_LITE_MODEL_NAME=gpt-3.5-turbo-16k-0613
CHAT_LITE_MODEL_NAME=
# 请求模型的最大token 数，非gpt的会使用cl100k_base编码进行token计算,如果最大的 token数少于6000 会也会自动关闭PDF阅读和阅读全文
# CHAT_LITE_REQUEST_TOKEN=12000
CHAT_LITE_REQUEST_TOKEN=
# chat pro 的模型配置
# CHAT_PRO_MODE_NAME=gpt-4-0613
CHAT_PRO_MODE_NAME=
# 请求模型的最大token 数，非gpt的会使用cl100k_base编码进行token计算,如果最大的 token数少于6000 会也会自动关闭PDF阅读和阅读全文
# CHAT_PRO_REQUEST_TOKEN=4000
CHAT_PRO_REQUEST_TOKEN=
# ## chat line 和 chat pro 的联网检索配置
# chat line的模型是否支持function call,这和BING_SUBSCRIPTION_KEY 共同决定是否开启联网检索（是否为true）
# true 和 BING_SUBSCRIPTION_KEY 不为空 则可以自动开通chat line 的联网检索
CHAT_LITE_FUNCTION_CALL=true
# chat pro的模型是否支持function call,这和BING_SUBSCRIPTION_KEY 共同决定是否开启联网检索（是否为true）
# true 和 BING_SUBSCRIPTION_KEY 不为空 则可以自动开通chat pro 的联网检索
CHAT_PRO_FUNCTION_CALL=true
# #####bing 检索的配置【非必填】######
BING_SUBSCRIPTION_KEY=
BING_SEARCH_URL=https://api.bing.microsoft.com/v7.0/search
# ## 阅读pdf 和插件的阅读全文
# emedding 配置，如果为空或者Chat line的CHAT_LITE_REQUEST_TOKEN 小于6000，不支持PDF阅读和阅读全文
EMBEDDING_MODEL_NAME=docker compose up -dCurrently, a solution called one-api is used for access and management of large models. This project allows users to access all large models through a unified and standardized OpenAI API format, which is very convenient for users to use out of the box. If you want to learn more about the specific information and usage instructions of this project, it is strongly recommended to refer to the official documentation of the project, which can be accessed by clicking this link one-api.
For users who want to start using one-api, it's very simple. First, the access address is: http://localhost:3000/. When using it for the first time, you will need to use the initial account name root and password 123456 to log in.
After logging in, an important step is to configure one-api to access the required large model channels. This step is very important to ensure that you can use Samepage normally.
Model checksum configuration
Verification steps:
Check if there is a model in the conf.json file that corresponds to the model you configured in One-API.
If the model does not exist, follow these steps to add it:
backend/management-service/src/main/resources/conf.json file. {
"modelName" : " one-api 配置的模型名" ,
"name" : "前端展示的名字" ,
"maxRequestToken" : 4000 ,
"functionCall" : false
}docker build -t samepaage/samepage-api:main -f ./backend/management-service/Dockerfile .Web access address: http://localhost Mobile access address: http://localhost/m-chat
The interface address configured by the browser plug-in: http://localhost/api/ The web address configured by the browser plug-in: http://localhost/
Join Samepage WeChat communication group:
We rely on the following projects:
This repository complies with the Apache 2.0 license agreement. See the LICENSE file for details.
