Can we continuing training on Textual Inversion on some existed embeddings?

Thanks for your amazing work!

Just want to ask if there is any way to continue a training on Textual Inversion, e.g. finish the training but find out still need more steps to train, therefore want to resume based on the previous trained embedding instead of doing everything from scratch again.

I didn't find such documents in the https://invoke-ai.github.io/invoke-training/. If there is such feature already implemented, please let me know; If no, would you plan to implement it?