nvidia.dali.fn.coco_reader#
- nvidia.dali.fn.coco_reader(*, annotations_file='', avoid_class_remapping=False, bytes_per_sample_hint=[0], dont_use_mmap=False, file_root=None, image_ids=False, images=None, include_iscrowd=True, initial_fill=1024, lazy_init=False, ltrb=False, num_shards=1, pad_last_batch=False, pixelwise_masks=False, polygon_masks=False, prefetch_queue_depth=1, preprocessed_annotations='', preserve=False, random_shuffle=False, ratio=False, read_ahead=False, save_preprocessed_annotations=False, save_preprocessed_annotations_dir='', seed=-1, shard_id=0, shuffle_after_epoch=False, size_threshold=0.1, skip_cached_images=False, skip_empty=False, stick_to_shard=False, tensor_init_bytes=1048576, device=None, name=None)#
Warning
This operator is now deprecated. Use
readers.coco()instead.In DALI 1.0 all readers were moved into a dedicated
readerssubmodule and renamed to follow a common pattern. This is a placeholder operator with identical functionality to allow for backward compatibility.Legacy alias for
readers.coco().