This topic provides SDK reference of modules of Machine Learning Platform for AI (PAI), such as Deep Learning Containers (DLC), Interactive Modeling (DSW), and Elastic Algorithm Service (EAS).
DLC SDK references
Submit jobs by using the SDK for Python
This topic describes how to use the SDK for Python to submit deep learning jobs.
Use PAIIO to read data from and write data to MaxCompute tables
This topic describes three interfaces supported by PAIIO, TableRecordDataset, TableReader, and TableWriter, and how to use these interfaces to read data from and write data to MaxCompute tables.
DSW SDK references
Create and manage DSW instances
This topic describes how to create and manage DSW instances.
Read data from and write data to OSS
This topic describes how to use Object Storage Service (OSS) SDK for Python to read data from and write data to OSS.
Use WebIDE to debug code online
This topic describes how to use WebIDE of DSW to debug the Python code in the sample notebook.
Use EasyVision to detect objects
This topic describes how to use EasyVision in DSW to detect objects.
EAS SDK references
Develop custom processors by using C or C++
This topic describes how to develop custom processors by using C or C++.
Develop custom processors by using Java
This topic describes how to develop custom processors by using Java.
Develop custom processors by using Python
This topic describes how to develop custom processors by using Python.
Warm up model services (advanced)
This topic describes how to generate a model warm-up file by using SDK.
The following topics provide sample code for calling services by using SDKs.
SDK references for scenario-based solutions
Usage notes on multimedia analysis SDK for Python
This topic describes the details of multimedia analysis SDK for Python, and how to use this SDK to call different model services. This topic also provides sample requests and responses.
SDK references for AI acceleration
The following topics describe how to use PAI-Blade SDKs to deploy PyTorch and TensorFlow models and perform model inference.