Updated on 2026-09-20 GMT+08:00

ModelArts Practices for Beginners

This section lists some common practices to help you understand and use ModelArts for AI development.

**Table 1** Common practices
Category	Use Cases	Description	Intended Audience
ModelArts model training	Training the Qwen3 Series Models Adapted to the MindSpeed-LLM Framework on ModelArts Training the Qwen3 Series Models Adapted to the verl Framework on ModelArts Training the Qwen3 Series Models Adapted to the AReaL Framework on ModelArts	Describes how to use the ModelArts training platform to train Qwen3 series models. It supports adaptation to the MindSpeed-LLM, verl, and AReaL frameworks. The ModelArts training platform provides the images and training framework code packages required by the training environment, making operations simple and convenient.	AI developers who are familiar with coding and debugging and understand OBS
ModelArts inference deployment	Deploying a Model from the vLLM-Ascend Open-Source Community on the ModelArts Inference Platform	Describes how to deploy models from the vLLM-Ascend open-source community on the ModelArts inference platform.	AI developers who are familiar with coding and debugging as well as Docker containers