AI Toolkit by Ostris integrates powerful dataset processing functions optimized especially for deep learning training tasks. The tool is able to automatically process input images of different resolutions, intelligently adjusting and grouping them according to configuration requirements, greatly simplifying data preparation before training.
The key features of dataset processing include:
- Support .jpg/.jpeg/.png three common image formats
- Requires each image file to be equipped with a corresponding .txt description file
- Automatically reduces the image to a specified resolution based on the profile
- Supports grouping images by resolution buckets.
- Provide [trigger] placeholder mechanism to facilitate batch replacement of descriptions
These features are particularly suitable for scenarios that require the handling of large volumes of training data, ensuring the consistency and standardization of input data during model training. Developers only need to prepare the dataset according to the standard format, and the tool will automatically complete the subsequent resolution adjustment and grouping work.
This answer comes from the articleAI Toolkit by Ostris: Stable Diffusion with FLUX.1 Model Training ToolkitThe
English 
简体中文 
日本語 
Deutsch 
Português do Brasil