are cool but they eat too many carbs." >>> inputs = tokenizer(text, return_tensors="np") >>> encoder_outputs = model.encode(**inputs) ```Nrœ