Sampling and Backup (old) - Nerogar/OneTrainer GitHub Wiki

This page is old, please use:

https://github.com/Nerogar/OneTrainer/wiki/Sampling or https://github.com/Nerogar/OneTrainer/wiki/Backup-and-Save

With Sampling you can decide to generate sample images. Define the frequency, then add samples with a seed and a prompt. For embedding training, the prompt needs <embedding> as a placeholder. The generated images will be saved in the Workspace directory. Note you can change and add samples during training.

With Backup, you can save intermediate versions of your model. Define a frequency. The intermediate versions of your model will be saved in the Workspace directory.

⚠️ **GitHub.com Fallback** ⚠️