Maps `map_func`` across batch_size consecutive elements of this dataset and then combines them into a batch. Functionally, it is equivalent to map followed by batch. However, by fusing the two transformations together, the implementation can be more efficient.
dataset_map_and_batch(
dataset,
map_func,
batch_size,
num_parallel_batches = NULL,
drop_remainder = FALSE,
num_parallel_calls = NULL
)
A dataset
A function mapping a nested structure of tensors (having
shapes and types defined by output_shapes()
and output_types()
to
another nested structure of tensors. It also supports purrr
style
lambda functions powered by rlang::as_function()
.
An integer, representing the number of consecutive elements of this dataset to combine in a single batch.
(Optional) An integer, representing the number of batches to create in parallel. On one hand, higher values can help mitigate the effect of stragglers. On the other hand, higher values can increase contention if CPU is scarce.
(Optional.) A boolean, representing whether the last
batch should be dropped in the case it has fewer than batch_size
elements; the default behavior is not to drop the smaller batch.
(Optional) An integer, representing the number of elements to process in parallel If not specified, elements will be processed sequentially.
Other dataset methods:
dataset_batch()
,
dataset_cache()
,
dataset_collect()
,
dataset_concatenate()
,
dataset_decode_delim()
,
dataset_filter()
,
dataset_interleave()
,
dataset_map()
,
dataset_padded_batch()
,
dataset_prefetch()
,
dataset_prefetch_to_device()
,
dataset_reduce()
,
dataset_repeat()
,
dataset_shuffle()
,
dataset_shuffle_and_repeat()
,
dataset_skip()
,
dataset_take()
,
dataset_take_while()
,
dataset_window()