{"id":20082951,"url":"https://github.com/carrychang/flask-keras-restful-api","last_synced_at":"2025-07-23T09:05:21.675Z","repository":{"id":112521381,"uuid":"233362785","full_name":"CarryChang/Flask-Keras-Restful-Api","owner":"CarryChang","description":"一个简单的搭建图像识别imagenet的flask后端API的方式，修改了全局model load的方式，使用nginx搭配Gunicorn启动Flask增加了模型推理的速度，使用虚拟环境搭配sh的启动方式，可以直接对model进行一键重启，并有错误日志监控。有问题，欢迎联系我。","archived":false,"fork":false,"pushed_at":"2020-07-04T12:58:17.000Z","size":93091,"stargazers_count":3,"open_issues_count":0,"forks_count":2,"subscribers_count":0,"default_branch":"master","last_synced_at":"2025-06-26T07:41:14.617Z","etag":null,"topics":["backend","cv","flask-restful","gunicorn-with-flask-rest-api","keras-tensorflow","shell-script","tensorflow-examples"],"latest_commit_sha":null,"homepage":"","language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":"apache-2.0","status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/CarryChang.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":"LICENSE","code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null,"governance":null,"roadmap":null,"authors":null,"dei":null,"publiccode":null,"codemeta":null}},"created_at":"2020-01-12T08:43:49.000Z","updated_at":"2023-07-13T10:33:39.000Z","dependencies_parsed_at":"2023-05-15T14:45:25.240Z","dependency_job_id":null,"html_url":"https://github.com/CarryChang/Flask-Keras-Restful-Api","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"purl":"pkg:github/CarryChang/Flask-Keras-Restful-Api","repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CarryChang%2FFlask-Keras-Restful-Api","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CarryChang%2FFlask-Keras-Restful-Api/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CarryChang%2FFlask-Keras-Restful-Api/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CarryChang%2FFlask-Keras-Restful-Api/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/CarryChang","download_url":"https://codeload.github.com/CarryChang/Flask-Keras-Restful-Api/tar.gz/refs/heads/master","sbom_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/CarryChang%2FFlask-Keras-Restful-Api/sbom","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":266649137,"owners_count":23962174,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","status":"online","status_checked_at":"2025-07-23T02:00:09.312Z","response_time":66,"last_error":null,"robots_txt_status":null,"robots_txt_updated_at":null,"robots_txt_url":"https://github.com/robots.txt","online":true,"can_crawl_api":true,"host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["backend","cv","flask-restful","gunicorn-with-flask-rest-api","keras-tensorflow","shell-script","tensorflow-examples"],"created_at":"2024-11-13T15:45:19.730Z","updated_at":"2025-07-23T09:05:21.649Z","avatar_url":"https://github.com/CarryChang.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"\n[![996.icu](https://img.shields.io/badge/link-996.icu-red.svg)](https://996.icu)\n\n\n### Flask-Keras-Restful-Api\n#### 一个简单的imagenet的flask后端API，修改了全局model load的方式，增加了模型推理的速度，使用nginx搭配Gunicorn启动Flask，使用虚拟环境搭配sh的启动方式，可以直接对model进行一键重启，并有错误日志监控\n \n\u003e  一键部署，源代码修改自： [*Building a simple Keras + deep learning REST API*](https://blog.keras.io/building-a-simple-keras-deep-learning-rest-api.html) \n\u003e 使用TF2替换了原始的keras，为模型提速\n## 使用方法\n\n\u003e 1. 首先在服务器上部署虚拟环境 ,假设虚拟环境在/home,cd /home 进入home\n\u003e 2. 在hoem文件夹中使用python3 -m venv v1创建虚拟环境,v1就是虚拟环境的名字，然后使用souce v1/bin/activate加载虚拟环境\n\u003e 4. 在虚拟环境下使用pip install -r requirement.txt 安装所需要的库，然后使用chmod +777 restart.sh部署模型的后端\n\u003e 5. 使用ip:8000/predict是post的地址，使用python keras_model_client.py即可模拟请求，注意模型第一次初始化的时间因为需要加载预训练模型，推理速度有些慢，目前单机单线曾运行的正常推理速度在100ms之内，多进程部署会继续提速\n\n\n## 代码结构：使用前后分离的架构，完全使用Python实现\n\n\u003e 1. keras_model_server.py表示后端的model api，直接通过post传参的形式进行，直接搭配nginx+Gunicorn部署即可\n\u003e 2. keras_model_client.py表示模型前端的调用，传入两张图片，然后使用imagenet的模型进行识别，模型第一次初始化的时间因为需要加载预训练模型，推理速度有些慢，目前单机单线曾运行的正常推理速度在100ms之内，多进程部署会继续提速\n\n\n\u003e 后端启动打印的log\n\n\u003cdiv align=center\u003e\u003cimg  src=\"https://github.com/CarryChang/Flask-Keras-Restful-Api/blob/master/pic/back_end.png\"\u003e\u003c/div\u003e\n\n\n\u003e 模型输出的结果\n\n\u003cdiv align=center\u003e\u003cimg  src=\"https://github.com/CarryChang/Flask-Keras-Restful-Api/blob/master/pic/frontend.png\"\u003e\u003c/div\u003e\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcarrychang%2Fflask-keras-restful-api","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fcarrychang%2Fflask-keras-restful-api","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fcarrychang%2Fflask-keras-restful-api/lists"}