chunked/nltk_chunking/tasks_document_question_answering/chunk_17.txt

updated_dataset = dataset.map(lambda example: {"question": example["query"]["en"]}, remove_columns=["query"])
updated_dataset = updated_dataset.map(
     lambda example: {"answer": example["answers"][0]}, remove_columns=["answer", "answers"]
 )

Note that the LayoutLMv2 checkpoint that we use in this guide has been trained with max_position_embeddings = 512 (you can
find this information in the checkpoint's config.json file).