docling-sdg
docling-sdg copied to clipboard
Allow generating a custom profile of question types
The default Q&A generation tries to generate 1 question of each type for every chunk.
Some use cases may require more questions of a specific type (e.g., summary vs single value).
Even though the prompt can be passed as an option in GenerateOptions object, there is no flexibility in terms of:
- number of target questions per passage (chunk)
- percentage of question per type
This enhancement should allow users have a more customized Q&A dataset