Hi! I use the GPT-2 model for the seq2seq task, but unfortunately, at the output of the model, words are cut off and sentences are not added, how can I make the model end sentences at the output and not cut off words? (increasing the maximum length does not correct the situation)
Hi! I use the GPT-2 model for the seq2seq task, but unfortunately, at the output of the model, words are cut off and sentences are not added, how can I make the model end sentences at the output and not cut off words? (increasing the maximum length does not correct the situation)