I’m complete beginner. I tried a few models in Colab - prithivida/parrot_paraphraser_on_T5 and other paraphrasers.
In Colab I always get an array with few rephrased variants. Then I started to play around with Inference API in Python with the same models, the same input, but I get only one response. Is there any option to get the same what I’m getting in Colab - to get more rephrased variants?
And also, how can I pass specific settings to model via API? Like do_diverse or max_length (as API seems to be limiting answer length)?