ModelArts is a central development platform for AI developers. With distributed training, automated model building, and model deployment, ModelArts helps AI developers quickly create models and manage the AI development lifecycle.
This new minor version upgrade includes the following new features, new changes and fixed bugs.
New Features:
- Increase Image size for ModelArts
- Node scale-in and deletion protection
- Node binding for logical subpools
- Limited number of training jobs created by a single user
- Only my jobs option used by default; function hidden
- Auto scaling – resource metrics
Fixed Features:
- Replacing a node in a ModelArts standard pool with a hot standby node of different specifications may create redundant nodes.
- When associating networks with SFS Turbo, it is recommended that only SFS Turbo files marked as available be included in the list. Including others may cause the association to fail.
- After a user terminates a training job, the job remains listed as queued in the resource pool.
- "Invalid token" appears sometimes when a user deletes an asynchronous real-time service.
- Notebook image saving errors.
For more information, please refer to the ModelArts area of Help Center.