1.引言
先分享dify项目的github:https://github.com/langgenius/dify
“项目由前腾讯云DevOps团队成员创建。我们发现基于OpenAI的API开发GPT应用程序有些麻烦。凭借我们多年在开发者效率工具方面的研发经验,我们希望能够让更多的人使用自然语言开发有趣的应用程序。”
dify由10+全职团队和100+社区贡献者共同维护,迭代非常快,我下载的时候还是0.6.3,现在已经升级为0.6.6,基本上每周都会升级一个版本。
我认为,了解Dify以及其中依赖的上下游技术、架构,利用Dify快速建立AI智能体demo原型,其实对推进AI智能体开发是有意义的。今天我先简要介绍一下Dify的部署过程,后面的篇幅会具体分享基于Dify的AI智能体开发经验,以及AI智能体开发所依赖的底层技术。
2.docker compose一行命令部署Dify
首先将dify项目下载至服务器上,
git clone https://github.com/langgenius/dify.git
项目主要分为api(后端)和web(前端)两大部分,具体代码后面再分析,进入docker目录
cd docker
目录中包含几个文件,docker-compose.yaml可以通过docker compose直接启动所有服务和依赖,docker-compose.middleware.yaml可以先启动依赖的关系数据库、向量数据库等组件,再单独启动api和web端,隔离做得可以说太棒了。
先看一下docker-compose.yaml的源码
version: '3' services: # API service api: image: langgenius/dify-api:0.6.6 restart: always environment: # Startup mode, 'api' starts the API server. MODE: api # The log level for the application. Supported values are `DEBUG`, `INFO`, `WARNING`, `ERROR`, `CRITICAL` LOG_LEVEL: INFO # A secret key that is used for securely signing the session cookie and encrypting sensitive information on the database. You can generate a strong key using `openssl rand -base64 42`. SECRET_KEY: sk-9f73s3ljTXVcMT3Blb3ljTqtsKiGHXVcMT3BlbkFJLK7U # The base URL of console application web frontend, refers to the Console base URL of WEB service if console domain is # different from api or web app domain. # example: http://cloud.dify.ai CONSOLE_WEB_URL: '' # Password for admin user initialization. # If left unset, admin user will not be prompted for a password when creating the initial admin account. INIT_PASSWORD: '' # The base URL of console application api server, refers to the Console base URL of WEB service if console domain is # different from api or web app domain. # example: http://cloud.dify.ai CONSOLE_API_URL: '' # The URL prefix for Service API endpoints, refers to the base URL of the current API service if api domain is # different from console domain. # example: http://api.dify.ai SERVICE_API_URL: '' # The URL prefix for Web APP frontend, refers to the Web App base URL of WEB service if web app domain is different from # console or api domain. # example: http://udify.app APP_WEB_URL: '' # File preview or download Url prefix. # used to display File preview or download Url to the front-end or as Multi-model inputs; # Url is signed and has expiration time. FILES_URL: '' # When enabled, migrations will be executed prior to application startup and the application will start after the migrations have completed. MIGRATION_ENABLED: 'true' # The configurations of postgres database connection. # It is consistent with the configuration in the 'db' service below. DB_USERNAME: postgres DB_PASSWORD: difyai123456 DB_HOST: db DB_PORT: 5432 DB_DATABASE: dify # The configurations of redis connection. # It is consistent with the configuration in the 'redis' service below. REDIS_HOST: redis REDIS_PORT: 6379 REDIS_USERNAME: '' REDIS_PASSWORD: difyai123456 REDIS_USE_SSL: 'false' # use redis db 0 for redis cache REDIS_DB: 0 # The configurations of celery broker. # Use redis as the broker, and redis db 1 for celery broker. CELERY_BROKER_URL: redis://:difyai123456@redis:6379/1 # Specifies the allowed origins for cross-origin requests to the Web API, e.g. https://dify.app or * for all origins. WEB_API_CORS_ALLOW_ORIGINS: '*' # Specifies the allowed origins for cross-origin requests to the console API, e.g. https://cloud.dify.ai or * for all origins. CONSOLE_CORS_ALLOW_ORIGINS: '*' # CSRF Cookie settings # Controls whether a cookie is sent with cross-site requests, # providing some protection against cross-site request forgery attacks # # Default: `SameSite=Lax, Secure=false, HttpOnly=true` # This default configuration supports same-origin requests using either HTTP or HTTPS, # but does not support cross-origin requests. It is suitable for local debugging purposes. # # If you want to enable cross-origin support, # you must use the HTTPS protocol and set the configuration to `SameSite=None, Secure=true, HttpOnly=true`. # # The type of storage to use for storing user files. Supported values are `local` and `s3` and `azure-blob` and `google-storage`, Default: `local` STORAGE_TYPE: local # The path to the local storage directory, the directory relative the root path of API service codes or absolute path. Default: `storage` or `/home/john/storage`. # only available when STORAGE_TYPE is `local`. STORAGE_LOCAL_PATH: storage # The S3 storage configurations, only available when STORAGE_TYPE is `s3`. S3_ENDPOINT: 'https://xxx.r2.cloudflarestorage.com' S3_BUCKET_NAME: 'difyai' S3_ACCESS_KEY: 'ak-difyai' S3_SECRET_KEY: 'sk-difyai' S3_REGION: 'us-east-1' # The Azure Blob storage configurations, only available when STORAGE_TYPE is `azure-blob`. AZURE_BLOB_ACCOUNT_NAME: 'difyai' AZURE_BLOB_ACCOUNT_KEY: 'difyai' AZURE_BLOB_CONTAINER_NAME: 'difyai-container' AZURE_BLOB_ACCOUNT_URL: 'https://<your_account_name>.blob.core.windows.net' # The Google storage configurations, only available when STORAGE_TYPE is `google-storage`. GOOGLE_STORAGE_BUCKET_NAME: 'yout-bucket-name' GOOGLE_STORAGE_SERVICE_ACCOUNT_JSON_BASE64: 'your-google-service-account-json-base64-string' # The type of vector store to use. Supported values are `weaviate`, `qdrant`, `milvus`, `relyt`. VECTOR_STORE: weaviate # The Weaviate endpoint URL. Only available when VECTOR_STORE is `weaviate`. WEAVIATE_ENDPOINT: http://weaviate:8080 # The Weaviate API key. WEAVIATE_API_KEY: WVF5YThaHlkYwhGUSmCRgsX3tD5ngdN8pkih # The Qdrant endpoint URL. Only available when VECTOR_STORE is `qdrant`. QDRANT_URL: http://qdrant:6333 # The Qdrant API key. QDRANT_API_KEY: difyai123456 # The Qdrant client timeout setting. QDRANT_CLIENT_TIMEOUT: 20 # The Qdrant client enable gRPC mode. QDRANT_GRPC_ENABLED: 'false' # The Qdrant server gRPC mode PORT. QDRANT_GRPC_PORT: 6334 # Milvus configuration Only available when VECTOR_STORE is `milvus`. # The milvus host. MILVUS_HOST: 127.0.0.1 # The milvus host. MILVUS_PORT: 19530 # The milvus username. MILVUS_USER: root # The milvus password. MILVUS_PASSWORD: Milvus # The milvus tls switch. MILVUS_SECURE: 'false' # relyt configurations RELYT_HOST: db RELYT_PORT: 5432 RELYT_USER: postgres RELYT_PASSWORD: difyai123456 RELYT_DATABASE: postgres # Mail configuration, support: resend, smtp MAIL_TYPE: '' # default send from email address, if not specified MAIL_DEFAULT_SEND_FROM: 'YOUR EMAIL FROM (eg: no-reply <no-reply@dify.ai>)' SMTP_SERVER: '' SMTP_PORT: 587 SMTP_USERNAME: '' SMTP_PASSWORD: '' SMTP_USE_TLS: 'true' # the api-key for resend (https://resend.com) RESEND_API_KEY: '' RESEND_API_URL: https://api.resend.com # The DSN for Sentry error reporting. If not set, Sentry error reporting will be disabled. SENTRY_DSN: '' # The sample rate for Sentry events. Default: `1.0` SENTRY_TRACES_SAMPLE_RATE: 1.0 # The sample rate for Sentry profiles. Default: `1.0` SENTRY_PROFILES_SAMPLE_RATE: 1.0 # Notion import configuration, support public and internal NOTION_INTEGRATION_TYPE: public NOTION_CLIENT_SECRET: you-client-secret NOTION_CLIENT_ID: you-client-id NOTION_INTERNAL_SECRET: you-internal-secret # The sandbox service endpoint. CODE_EXECUTION_ENDPOINT: "http://sandbox:8194" CODE_EXECUTION_API_KEY: dify-sandbox CODE_MAX_NUMBER: 9223372036854775807 CODE_MIN_NUMBER: -9223372036854775808 CODE_MAX_STRING_LENGTH: 80000 TEMPLATE_TRANSFORM_MAX_LENGTH: 80000 CODE_MAX_STRING_ARRAY_LENGTH: 30 CODE_MAX_OBJECT_ARRAY_LENGTH: 30 CODE_MAX_NUMBER_ARRAY_LENGTH: 1000 depends_on: - db - redis volumes: # Mount the storage directory to the container, for storing user files. - ./volumes/app/storage:/app/api/storage # uncomment to expose dify-api port to host # ports: # - "5001:5001" # worker service # The Celery worker for processing the queue. worker: image: langgenius/dify-api:0.6.6 restart: always environment: # Startup mode, 'worker' starts the Celery worker for processing the queue. MODE: worker # --- All the configurations below are the same as those in the 'api' service. --- # The log level for the application. Supported values are `DEBUG`, `INFO`, `WARNING`, `ERROR`, `CRITICAL` LOG_LEVEL: INFO # A secret key that is used for securely signing the session cookie and encrypting sensitive information on the database. You can generate a strong key using `openssl rand -base64 42`. # same as the API service SECRET_KEY: sk-9f73s3ljTXVcMT3Blb3ljTqtsKiGHXVcMT3BlbkFJLK7U # The configurations of postgres database connection. # It is consistent with the configuration in the 'db' service below. DB_USERNAME: postgres DB_PASSWORD: difyai123456 DB_HOST: db DB_PORT: 5432 DB_DATABASE: dify # The configurations of redis cache connection. REDIS_HOST: redis REDIS_PORT: 6379 REDIS_USERNAME: '' REDIS_PASSWORD: difyai123456 REDIS_DB: 0 REDIS_USE_SSL: 'false' # The configurations of celery broker. CELERY_BROKER_URL: redis://:difyai123456@redis:6379/1 # The type of storage to use for storing user files. Supported values are `local` and `s3` and `azure-blob`, Default: `local` STORAGE_TYPE: local STORAGE_LOCAL_PATH: storage # The S3 storage configurations, only available when STORAGE_TYPE is `s3`. S3_ENDPOINT: 'https://xxx.r2.cloudflarestorage.com' S3_BUCKET_NAME: 'difyai' S3_ACCESS_KEY: 'ak-difyai' S3_SECRET_KEY: 'sk-difyai' S3_REGION: 'us-east-1' # The Azure Blob storage configurations, only available when STORAGE_TYPE is `azure-blob`. AZURE_BLOB_ACCOUNT_NAME: 'difyai' AZURE_BLOB_ACCOUNT_KEY: 'difyai' AZURE_BLOB_CONTAINER_NAME: 'difyai-container' AZURE_BLOB_ACCOUNT_URL: 'https://<your_account_name>.blob.core.windows.net' # The type of vector store to use. Supported values are `weaviate`, `qdrant`, `milvus`, `relyt`. VECTOR_STORE: weaviate # The Weaviate endpoint URL. Only available when VECTOR_STORE is `weaviate`. WEAVIATE_ENDPOINT: http://weaviate:8080 # The Weaviate API key. WEAVIATE_API_KEY: WVF5YThaHlkYwhGUSmCRgsX3tD5ngdN8pkih # The Qdrant endpoint URL. Only available when VECTOR_STORE is `qdrant`. QDRANT_URL: http://qdrant:6333 # The Qdrant API key. QDRANT_API_KEY: difyai123456 # The Qdrant clinet timeout setting. QDRANT_CLIENT_TIMEOUT: 20 # The Qdrant client enable gRPC mode. QDRANT_GRPC_ENABLED: 'false' # The Qdrant server gRPC mode PORT. QDRANT_GRPC_PORT: 6334 # Milvus configuration Only available when VECTOR_STORE is `milvus`. # The milvus host. MILVUS_HOST: 127.0.0.1 # The milvus host. MILVUS_PORT: 19530 # The milvus username. MILVUS_USER: root # The milvus password. MILVUS_PASSWORD: Milvus # The milvus tls switch. MILVUS_SECURE: 'false' # Mail configuration, support: resend MAIL_TYPE: '' # default send from email address, if not specified MAIL_DEFAULT_SEND_FROM: 'YOUR EMAIL FROM (eg: no-reply <no-reply@dify.ai>)' # the api-key for resend (https://resend.com) RESEND_API_KEY: '' RESEND_API_URL: https://api.resend.com # relyt configurations RELYT_HOST: db RELYT_PORT: 5432 RELYT_USER: postgres RELYT_PASSWORD: difyai123456 RELYT_DATABASE: postgres # Notion import configuration, support public and internal NOTION_INTEGRATION_TYPE: public NOTION_CLIENT_SECRET: you-client-secret NOTION_CLIENT_ID: you-client-id NOTION_INTERNAL_SECRET: you-internal-secret depends_on: - db - redis volumes: # Mount the storage directory to the container, for storing user files. - ./volumes/app/storage:/app/api/storage # Frontend web application. web: image: langgenius/dify-web:0.6.6 restart: always environment: # The base URL of console application api server, refers to the Console base URL of WEB service if console domain is # different from api or web app domain. # example: http://cloud.dify.ai CONSOLE_API_URL: '' # The URL for Web APP api server, refers to the Web App base URL of WEB service if web app domain is different from # console or api domain. # example: http://udify.app APP_API_URL: '' # The DSN for Sentry error reporting. If not set, Sentry error reporting will be disabled. SENTRY_DSN: '' # uncomment to expose dify-web port to host # ports: # - "3000:3000" # The postgres database. db: image: postgres:15-alpine restart: always environment: PGUSER: postgres # The password for the default postgres user. POSTGRES_PASSWORD: difyai123456 # The name of the default postgres database. POSTGRES_DB: dify # postgres data directory PGDATA: /var/lib/postgresql/data/pgdata volumes: - ./volumes/db/data:/var/lib/postgresql/data # uncomment to expose db(postgresql) port to host # ports: # - "5432:5432" healthcheck: test: [ "CMD", "pg_isready" ] interval: 1s timeout: 3s retries: 30 # The redis cache. redis: image: redis:6-alpine restart: always volumes: # Mount the redis data directory to the container. - ./volumes/redis/data:/data # Set the redis password when startup redis server. command: redis-server --requirepass difyai123456 healthcheck: test: [ "CMD", "redis-cli", "ping" ] # uncomment to expose redis port to host # ports: # - "6379:6379" # The Weaviate vector store. weaviate: image: semitechnologies/weaviate:1.19.0 restart: always volumes: # Mount the Weaviate data directory to the container. - ./volumes/weaviate:/var/lib/weaviate environment: # The Weaviate configurations # You can refer to the [Weaviate](https://weaviate.io/developers/weaviate/config-refs/env-vars) documentation for more information. QUERY_DEFAULTS_LIMIT: 25 AUTHENTICATION_ANONYMOUS_ACCESS_ENABLED: 'false' PERSISTENCE_DATA_PATH: '/var/lib/weaviate' DEFAULT_VECTORIZER_MODULE: 'none' CLUSTER_HOSTNAME: 'node1' AUTHENTICATION_APIKEY_ENABLED: 'true' AUTHENTICATION_APIKEY_ALLOWED_KEYS: 'WVF5YThaHlkYwhGUSmCRgsX3tD5ngdN8pkih' AUTHENTICATION_APIKEY_USERS: 'hello@dify.ai' AUTHORIZATION_ADMINLIST_ENABLED: 'true' AUTHORIZATION_ADMINLIST_USERS: 'hello@dify.ai' # uncomment to expose weaviate port to host # ports: # - "8080:8080" # The DifySandbox sandbox: image: langgenius/dify-sandbox:0.1.0 restart: always cap_add: # Why is sys_admin permission needed? # https://docs.dify.ai/getting-started/install-self-hosted/install-faq#id-16.-why-is-sys_admin-permission-needed - SYS_ADMIN environment: # The DifySandbox configurations API_KEY: dify-sandbox GIN_MODE: release WORKER_TIMEOUT: 15 # Qdrant vector store. # uncomment to use qdrant as vector store. # (if uncommented, you need to comment out the weaviate service above, # and set VECTOR_STORE to qdrant in the api & worker service.) # qdrant: # image: langgenius/qdrant:v1.7.3 # restart: always # volumes: # - ./volumes/qdrant:/qdrant/storage # environment: # QDRANT_API_KEY: 'difyai123456' # # uncomment to expose qdrant port to host # # ports: # # - "6333:6333" # # - "6334:6334" # The nginx reverse proxy. # used for reverse proxying the API service and Web service. nginx: image: nginx:latest restart: always volumes: - ./nginx/nginx.conf:/etc/nginx/nginx.conf - ./nginx/proxy.conf:/etc/nginx/proxy.conf - ./nginx/conf.d:/etc/nginx/conf.d #- ./nginx/ssl:/etc/ssl depends_on: - api - web ports: - "80:80" #- "443:443"
主要包含如下几个模块及docker镜像
- api:langgenius/dify-api:0.6.6
- worker:langgenius/dify-api:0.6.6
- web:langgenius/dify-web:0.6.6
- db:postgres:15-alpine
- redis:redis:6-alpine
- weaviate:semitechnologies/weaviate:1.19.0
- sandbox:langgenius/dify-sandbox:0.1.0
- nginx:nginx:latest
docker compose一键部署安装
docker compose up -d
从dockerhub逐个pulling依赖镜像,感觉还挺爽的
等待镜像下载部署完成后,打开webui:123.123.123.123:80,默认占用80端口,可以在docker compose配置文件中更改nignx端口。
经过账号初始化等工作后,欢迎来到Dify工作界面,就是这么丝滑
3.Dify主要特色
3.1多种大语言模型集成
相较于fastGPT+oneApi的方案,集成度更高:
- 大模型厂商模型:仅需到大模型厂商注册账号,申请鉴权apikey,即可快速体验比较每个大模型厂商的优劣,项目还贴心的附带了每个厂商的注册跳转链接,每个厂商基本都会给几百万的token用于测试。
- 本地部署模型:支持Xinference,Ollama,OpenLLM,LocalAI等推理框架部署的模型一键接入
- HuggingFace开源模型:只需配置APIKEY和模型名字即可接入,不过要求服务器能翻墙噢
- OpenAI-API-compatible:接入兼容OpenAI规范的API,目前Xinference,OpenLLM等很多推理框架,都是直接支持OpenAI API接口规范的,但对于每家大模型厂商,一般都要设计自己的规范,可能是想构建生态,或者是与众不同?这里多说一下:对于国内互联网中小厂,如果不自建大模型,可能要试用或买入多加公司的大模型,多加大模型给多个业务部门使用,就要有一个代理平台专门计算成本,这个平台最好对业务暴露的是兼容OpenAI的API接口,如果不是的话,当接入Dify平台时,就需要包一层与OpenAI兼容的API接口。
3.2丰富工具内置+自定义工具支持
Dify内置了包含搜索引擎、天气预报、维基百科、SD等工具,同时自定义工具的配置化接入,团队成员一人接入,全组复用,高效!
3.3工作流
只需连接各个节点,既能在几分钟内快速完成AI智能体创作,且逻辑非常清晰。
3.4Agent 编排
编写提示词,导入知识库,添加工具,选择模型,运行测试,发布为API,一条龙创作!
还有很多特色,在此不再赘述了,附上一张官方的表格吧
4.总结
临下班了,本来只是想参加个话题活动,洋洋洒洒几千字。可能加入了太多感慨吧。希望感兴趣的朋友可以关注我、点赞、收藏和评论,您的鼓励是我持续码字的动力。
本文首先结合自己的工作写了一些对Agent AI智能体的见解,接着介绍了Dify框架快捷部署的过程,最后阐述了Dify框架的特点。个人认为Dify的发展会让Agent AI智能体开发提效,涌现更多有趣有价值的AI应用。
最后,写一下我对未来AI智能体发展的看法吧,从流量与用户来看,2000年-2004年,以新浪、搜狐、网易为代表的门户网站是流量入口,2004年-2014年,以百度为代表的搜索引擎是流量入口,2014-2024,以抖音、快手、微博、小红书为代表的移动互联网推荐系统是流量入口,抓住了流量入口就抓住了用户,抓住了用户就抓住了商业变现。2024-未来,极大的可能出现一家基于AI的平台型企业,通过AI智能体抓住流量,比如你要去哪玩,AI智能体在为你做出规划的过程中,夹杂酒店、航班的广告私货,你想吃什么,AI智能体夹杂着饭店的广告私货。
AI领域,乾坤未定,你我皆是黑马。