Updated on 2026-08-27 GMT+08:00

Model Training Practices Summary

This chapter summarizes all practice cases related to model training. Select the corresponding case based on your actual needs.

**Table 1** Model training practices summary
Practice Name	Description
Example: Migrating Local PyTorch Training Code to ModelArts	This practice mainly explains how to migrate a training task that runs successfully locally to ModelArts.
Example: Creating a DDP Distributed Training Job (PyTorch + GPU)	This practice mainly demonstrates how to build a PyTorch + GPU-based distributed training task.
Example: Creating a DDP Distributed Training Job (PyTorch + NPU)	This practice mainly demonstrates how to build a PyTorch + NPU-based distributed training task.