I managed to convert a Pegasus model to ctranslate2 format. Now, when querying the model, I need to pass some parameters to it:
translated = model.generate(**batch,max_length=60, num_beams=10, num_return_sequences=10, temperature=0.0, do_sample=True)
Is there a way to do it in the config file or in the queries to OpenNMT-py server?
P.S. I know that in order to get 10 versions of the response, I can use “n_best” parameter. However, how can I pass