Hi!
I apologize if this feature already exists but I found no evidence of it neither in the documentation nor in the source code.
I'm working with the synthetic data generation feature. I'm using context documents written in italian language but the output questions and the corresponding ground truth answers are in english and it would be great if they were in the same language of the input context documents.
I've made some changes to the code to use custom prompts and put them here:
https://github.com/faradri/ragas/tree/testset-prompts
Thank you