Could I inference the Encoder-Decoder model without specify "decoder_input_ids"?

I see. I used EncoderDecoderModel before, not the Bart Model, so there is no this feature there. Besides, their behaviors seems different. Now, I’d try to use Bart to see how it goes. I personally think EncoderDecoderModel should also be able to forward without “decoder_input_ids” and “labels”. Anyways, thank you for sharing that.