Fine-tuning issue

Hello,

Thank you for the great work! I just had a question:

The datasets provided for fine-tuning do not have segmentation masks, rather, they have bounding boxes (as I understand), so, how to fine-tune the model using segmentation masks with corresponding text prompt on a custom dataset?

Best,
Mohamed

Start with the fine-tuning example notebooks described in the repository, and compare their expected bounding-box annotations with a custom dataset containing segmentation masks and text prompts. Trace the fine-tuning entry point to determine whether this annotation format is supported, then establish the required workflow or missing capability.

Written by the indexing model from the issue text.