Metadata-Version: 2.1
Name: aiflow
Version: 1.0.0
Summary: AI Flow, an extend operators library for airflow, which helps AI engineer to write less, reuse more, integrate easily.
Home-page: https://github.com/Marcnuth/aiflow
Author: Xianxian Huang
Author-email: hxianxian@gmail.com
License: Apache License 2.0
Description: # AI Flow
        
        ## Introduction
        AI Flow, which offers various reusable operators & processing units in AI modeling, helps AI engineer to write less, reuse more, integrate easily.
        
        ## Install
        
        ```
        pip install aiflow
        ```
        
        
        ## Concepts
        
        ### Operators VS. Units
        
        Ideally, we agree:
        - An **Operator** would contain lot of units, which will be integrated into `airflow` for building non-realtime processing workflow;
        - A **Unit** is a small calculation unit, which could be a function, or just a simple modeling logic, and it could be picked as bricks to build an operator. Besides, it could be reused anywhere for realtime calculation.
        
        ## Classes
        
        ### Operators
        
        #### MongoToCSVOperator
        
        #### Elastic2CSVOperator
        
        #### RegExLabellingOperator
        
        ### Units
        
        #### Doc2VecUnit
        
        #### Doc2MatUnit
        
        ## Tests & Examples
        
        ### Example: Use Units to Build Your Castle
        
        
        ### Example: Working with Airflow
        
        In `tests/docker/` folder, we provide examples on how to use `aiflow` with `airflow`.
        It is a docker image, you could simply copy and start to use it!
        
        In project root directory, run commands first:
        ```
        docker-compose up --build aiflow
        ```
        
        Then open `localhost:8080` in your browser, you can see all the examples `aiflow` provided!
        Note: both the default username & password are `admin`
        
        Enjoy!
        
        ## Contribution
Platform: UNKNOWN
Classifier: Topic :: Scientific/Engineering :: Artificial Intelligence
Classifier: Topic :: Software Development :: Libraries
Classifier: Programming Language :: Python :: 3
Description-Content-Type: text/markdown
