Data Overview

aifare platform provides a variety of data storage and management solutions to meet different AI development needs. This document introduces the main data types, storage locations, and usage suggestions.

Data Types and Storage Locations

Data Type	Storage Path	Description
System Disk Data	/	System files, code, small data, Python packages
Data Disk Data	/data	User data, datasets, high I/O data
Model Storage	/gm-models	Prebuilt models, user models
Dataset Storage	/gm-datasets	Public datasets, training data
User Data Storage	/user-data	User personal data, cross-instance sharing

System Disk

• Used for system files, Python environments, and code.
• Data is retained after shutdown and can be saved as a custom image.
• Not suitable for storing large datasets.

Data Disk

• Used for storing user data, datasets, and high I/O data.
• Data is retained after shutdown but not saved in custom images.
• Suitable for large files and frequent read/write operations.

Model Storage

• /gm-models provides prebuilt models and user models.
• Models can be quickly loaded and used in development.

Dataset Storage

• /gm-datasets provides public datasets and training data.
• Supports fast loading and sharing across instances.

User Data Storage

• /user-data is for user personal data and supports cross-instance sharing.
• Data is retained after shutdown and can be accessed by multiple instances.

Usage Suggestions

1. Store code and environments on the system disk.
2. Store large datasets and high I/O data on the data disk.
3. Use model and dataset storage for quick access to common resources.
4. Use user data storage for cross-instance data sharing.

For more details, please refer to the aifare platform documentation or contact customer support.