Hi All,
Im wondering if it’s possible to use the num_return_sequences parameter with the HF inference API in order to get back multiple outputs from a single input i.e multiple summaries from the same piece of text. I’ve tried setting num_return_sequences to 10 and calling the API but I get back the same response with just one output although interestingly it takes much longer to do so.
Before diving into this further I wanted to check wether this functionality is even live using the API?
Many thanks,
Karim Foda