Data loader. Combines a dataset and a sampler, and provides single- or multi-process iterators over the dataset.Source:
Data loader. Combines a dataset and a sampler, and provides single- or multi-process iterators over the dataset.
dataloader( dataset, batch_size = 1, shuffle = FALSE, sampler = NULL, batch_sampler = NULL, num_workers = 0, collate_fn = NULL, pin_memory = FALSE, drop_last = FALSE, timeout = -1, worker_init_fn = NULL, worker_globals = NULL, worker_packages = NULL )
(Dataset): dataset from which to load the data.
(int, optional): how many samples per batch to load (default:
(bool, optional): set to
TRUEto have the data reshuffled at every epoch (default:
(Sampler, optional): defines the strategy to draw samples from the dataset. If specified,
shufflemust be False.
(Sampler, optional): like sampler, but returns a batch of indices at a time. Mutually exclusive with
(int, optional): how many subprocesses to use for data loading. 0 means that the data will be loaded in the main process. (default:
(callable, optional): merges a list of samples to form a mini-batch.
(bool, optional): If
TRUE, the data loader will copy tensors into CUDA pinned memory before returning them. If your data elements are a custom type, or your
collate_fnreturns a batch that is a custom type see the example below.
(bool, optional): set to
TRUEto drop the last incomplete batch, if the dataset size is not divisible by the batch size. If
FALSEand the size of dataset is not divisible by the batch size, then the last batch will be smaller. (default:
(numeric, optional): if positive, the timeout value for collecting a batch from workers. -1 means no timeout. (default:
(callable, optional): If not
NULL, this will be called on each worker subprocess with the worker id (an int in
[1, num_workers]) as input, after seeding and before data loading. (default:
(list or character vector, optional) only used when
num_workers > 0. If a character vector, then objects with those names are copied from the global environment to the workers. If a named list, then this list is copied and attached to the worker global environment. Notice that the objects are copied only once at the worker initialization.
(character vector, optional) Only used if
num_workers > 0optional character vector naming packages that should be loaded in each worker.
num_workers > 0 data loading will happen in parallel for each
worker. Note that batches are taken in parallel and not observations.
The worker initialization process happens in the following order:
num_workersR sessions are initialized.
Then in each worker we perform the following actions:
torchlibrary is loaded.
a random seed is set both using
packages passed to the
worker_packagesargument are loaded.
objects passed trough the
worker_globalsparameters are copied into the global environment.
worker_initfunction is ran with an
the dataset fetcher is copied to the worker.