chunked/nltk_chunking/model_doc_detr/chunk_1.txt

DETR
consists of a convolutional backbone followed by an encoder-decoder Transformer which can be trained end-to-end for
object detection.