# Copyright 2022 The TensorFlow Authors. All Rights Reserved. # # Licensed under the Apache License, Version 2.0 (the "License"); # you may not use this file except in compliance with the License. # You may obtain a copy of the License at # # http://www.apache.org/licenses/LICENSE-2.0 # # Unless required by applicable law or agreed to in writing, software # distributed under the License is distributed on an "AS IS" BASIS, # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. # See the License for the specific language governing permissions and # limitations under the License. # ============================================================================== """APIs to deal with input datasets efficiently in DTensor. When using tf.data with DTensor, the `DTensorDataset` API can be used to efficiently handle loading the input data and correctly packing it to the corresponding devices. This API is intended to work with unbatched data and can be used for both data and model parallel setups. Example usage: >>> # 1-D mesh with 4 devices >>> mesh = dtensor.Mesh(dim_names=['batch'], ...) >>> layout = dtensor.Layout.batch_sharded(mesh, 'batch', rank=1) >>> dataset = tf.data.Dataset.range(256) >>> d_dataset = dtensor.DTensorDataset( ... dataset=dataset, ... global_batch_size=16, ... mesh=mesh, ... layouts=layout, ... batch_dim='batch') >>> d_iter = iter(d_dataset) >>> # Each batch is a length 16 tensor sharded across 4 devices >>> batch_0_dtensor = next(d_iter) >>> batch_0_dtensor >>> batch_1_dtensor = next(d_iter) >>> batch_1_dtensor For multi-client setups, `DTensorDataset` interacts with tf.data service to correctly distribute the dataset among the participating clients. DTensor works with tf.data service in co-located mode where each worker is running alongside the DTensor client (the Tensorflow Python process). The `TFDataServiceConfig` dataclass can be filled with information about the tf.data service cluster, and passed to `DTensorDataset` to enable distribution. """ import dataclasses from typing import Any, List, Optional, Sequence, Tuple from tensorflow.dtensor.python import api from tensorflow.dtensor.python import config from tensorflow.dtensor.python import layout as layout_lib from tensorflow.python.data.experimental.ops import data_service_ops from tensorflow.python.data.ops import dataset_ops from tensorflow.python.data.ops import iterator_ops from tensorflow.python.framework import constant_op from tensorflow.python.framework import dtypes from tensorflow.python.framework import errors from tensorflow.python.framework import ops from tensorflow.python.framework import tensor_spec from tensorflow.python.ops import array_ops from tensorflow.python.ops import math_ops from tensorflow.python.util import nest from tensorflow.python.util.tf_export import tf_export @dataclasses.dataclass class TFDataServiceConfig: """Specifies the tf.data service configuration to use. Attributes: dispatcher_address: a string specifying the address of the tf.data service dispatcher server. job_name: a non-empty string identifying the shared job that will be created on tf.data service to process this dataset. """ dispatcher_address: str job_name: str # TODO(b/223275517): Add support for get_next_as_optional(). class _DTensorIterator(iterator_ops.IteratorBase): """An iterator for a tf.data.Dataset distributed using DTensor. DTensorIterator encapsulates multiple underlying dataset iterators. It handles retrieving the tensors to be placed on each underlying device and then uses the 'pack' operation to create and return a DTensor. Thus users need only interact with a single DTensorIterator to automatically distribute dataset tensors onto devices. """ def __init__(self, datasets: Sequence[Tuple[int, dataset_ops.DatasetV2]], element_spec: tensor_spec.TensorSpec, layouts: Any, num_local_devices_per_replica: int): """Initializes a distributed iterator for DTensor datasets. The DTensorIterator uses 'replica IDs' to identify shards of a dataset. Here the term 'replica' is used in the data-parallel context where each replica receives a partition of the global batch. Depending on the model parallelism in the layouts supplied, each device within that replica may receive the same partition of the global batch (no model parallelism), or specific slices of that partition. Args: datasets: a dictionary mapping each unique local replica ID to the dataset object whose elements will be placed on the devices corresponding to that replica. element_spec: the underlying dataset's element spec. layouts: a structure of DTensor layouts to be applied to the dataset values. This can be a single layout or (possibly nested) tuples or dictionaries of layouts, and the structure must match the structure of the dataset. num_local_devices_per_replica: the number of local devices for each replica. """ self._iterators = [ (replica_id, iter(dataset)) for replica_id, dataset in datasets ] self._element_spec = element_spec self._layouts = layouts self._num_local_devices_per_replica = num_local_devices_per_replica self._flattened_layouts = nest.flatten(self._layouts) def __next__(self): try: return self.get_next() except errors.OutOfRangeError as e: raise StopIteration from e def __iter__(self): return self @property def element_spec(self): """The type specification of an element of this iterator. A possibly nested structure of `tf.TypeSpec` objects matching the structure of an element of this iterator. """ return self._element_spec def get_next(self): """Returns the next element. Returns: A possibly nested structure of values matching `tf.data.Iterator.element_spec`. Raises: `tf.errors.OutOfRangeError`: if the end of the underlying iterators has been reached. RuntimeError: if any of the underlying iterators do not return the expected number of items. """ # Create the data structure to store the individual elements of the current # batch. We store a list per element in the flattened dataset batch, and # each list should contain as many tensors as there local devices. curr_batch_elems = [[] for _ in range(len(self._flattened_layouts))] for _, iterator in self._iterators: for _ in range(self._num_local_devices_per_replica): element = iterator.get_next() # Separate the dataset elements based on the structure of the dataset. flattened_element = nest.flatten(element) for idx, batch in enumerate(flattened_element): curr_batch_elems[idx].append(batch) flattened_output = [] for batch_elems, layout in zip(curr_batch_elems, self._flattened_layouts): expected_num_elems = layout.mesh.num_local_devices() actual_num_elems = len(batch_elems) if actual_num_elems != expected_num_elems: raise RuntimeError('Expected to pack %d elements in batch but got %d' % (expected_num_elems, actual_num_elems)) flattened_output.append(api.pack(batch_elems, layout)) return nest.pack_sequence_as(self._layouts, flattened_output) def get_next_as_optional(self): """Returns the next element wrapped in `tf.experimental.Optional`. If the iterator has reached the end of the sequence, the returned `tf.experimental.Optional` will have no value. Returns: A `tf.experimental.Optional` object representing the next element. """ raise NotImplementedError( 'get_next_as_optional not yet supported: b/223275517') @property def _type_spec(self): return iterator_ops.IteratorSpec(self._element_spec) def _validate_input(flattened_layouts: Sequence[layout_lib.Layout], flattened_elem_spec: Sequence[tensor_spec.TensorSpec], dataset_already_batched: bool): """Checks that the dataset's layouts and element specs are compatible. Args: flattened_layouts: the flattened list of layouts used to distribute the dataset. flattened_elem_spec: the flattened list of element specs used in the dataset's components. dataset_already_batched: whether the dataset to be validated is already batched. Raises: ValueError: if the dataset's inputs are incompatible. """ if not flattened_elem_spec: raise ValueError( 'Expected input element spec of at least one element, was empty.') first_elem_shape = flattened_elem_spec[0].shape for layout, elem_spec in zip(flattened_layouts, flattened_elem_spec): if elem_spec.shape.rank is None: raise ValueError( 'Dataset element shape must have a valid rank, got spec %s.' % elem_spec) # Check that layout's rank matches the element's rank. If dataset is not yet # batched, then the layout's rank must be one greater than the element's # rank. expected_rank = elem_spec.shape.rank if not dataset_already_batched: expected_rank += 1 if layout.rank != expected_rank: raise ValueError( ('Expected layout with rank %d for element spec %s, got layout %s. ' 'Check that the dataset is not batched before passing to ' 'DTensorDataset.') % (expected_rank, elem_spec, layout.sharding_specs)) if dataset_already_batched: # Check that the batch dimension size of all dataset elements match. batch_dim_size = first_elem_shape.as_list()[0] if batch_dim_size is None: raise ValueError( ('Size of batch dimension of element spec %s is None. Ensure ' 'drop_remainder=True when batching the dataset.') % elem_spec) if elem_spec.shape.as_list()[0] != batch_dim_size: raise ValueError( ('Size of batch dimension of element spec %s does not match ' 'expected size %d.') % (elem_spec, batch_dim_size)) def _shard_counts(layout: layout_lib.Layout, batch_dim: Optional[str] = None) -> List[int]: """Computes a list of the number of shards in each dimension of the layout. The shard counts are used to slice each dataset element. The batch dimension's count is overridden to 1 since we only consider how many shards to make locally (within each local replica). Sharding across clients is handled by either tf.data.Dataset's shard transformation (in the single-client case) or tf.data service's distribute function (in the multi-client case). Args: layout: the layout to compute the shard counts for. batch_dim: the name of the batch dimension of the layout, if present. Returns: A list of shard counts, one element per dimension of the layout. """ shard_counts = [] for spec in layout.sharding_specs: if spec in (batch_dim, layout_lib.UNSHARDED): shard_counts.append(1) else: shard_counts.append(layout.mesh.dim_size(spec)) return shard_counts def _index_matrix(layout: layout_lib.Layout, elem_spec: tensor_spec.TensorSpec) -> ops.Tensor: """Computes a utility matrix to derive device-based slice offsets. This function builds a matrix of shape `[mesh.rank, layout.rank]` for each dataset element. This matrix can be used to slice the DTensor components returned by the iterator according to the local device that component is to be placed on. This can be done by multiplying the device offsets of shape `[1, mesh.rank]` with this index matrix to get a `[1, layout.rank]` shape tensor containing the slice offsets. Note: the index on the batch dim is always 0 since sharding on the batch dimension is handled by either tf.data.Dataset's shard transformation (in the single-client case) or tf.data service's distribute function (in the multi-client case). If there is no sharding on the batch dimension (or any other dimension), the slice index remains 0. Args: layout: the layout of the dataset element. elem_spec: the spec of the dataset element. Returns: The index matrix as a tensor. """ matrix = [] for dim in layout.mesh.dim_names: row = [0] for layout_idx, spec in enumerate(layout.sharding_specs[1:]): if spec == layout_lib.UNSHARDED or spec != dim: row.append(0) else: row.append(elem_spec.shape[layout_idx] // layout.mesh.dim_size(dim)) matrix.append(row) return constant_op.constant(matrix, dtype=dtypes.int32) @tf_export('experimental.dtensor.DTensorDataset', v1=[]) class DTensorDataset(dataset_ops.UnaryUnchangedStructureDataset): """A dataset of DTensors. DTensorDataset encapsulates a `tf.data.Dataset` whose elements are automatically packed and returned as DTensors based on a given mesh and layouts. """ def __init__(self, dataset: dataset_ops.DatasetV2, *, mesh: layout_lib.Mesh, layouts: Any, global_batch_size: int, dataset_already_batched: bool = False, batch_dim: Optional[str] = None, prefetch: Optional[int] = None, tf_data_service_config: Optional[TFDataServiceConfig] = None): """Creates a DTensorDataset. DTensorDataset automatically handles distribution of the dataset elements to each client's devices. It can be used to create an iterator that returns DTensors of the input data on each iteration. DTensorDataset works best with unbatched datasets. It takes the mesh and the provided layouts to automatically calculate how to batch the input locally for each replica. If the provided dataset is already batched according to the per-replica batch size, then `dataset_already_batched` must be set and DTensorDataset will check that the batch size is consistent with the intended `global_batch_size` using the layout information. Each replica receives a separate slice of the global batch, thus the per-replica batch size can be computed as the global batch size divided by the number of model replicas. For a DTensor mesh, the number of replicas is equal to the size of the mesh's batch dimension. TODO(b/223275517): add support for input datasets that are already batched to the global batch size. Args: dataset: a `tf.data.Dataset` object. mesh: the DTensor mesh to place the dataset batches on. layouts: a structure of DTensor layouts to be applied to the input dataset values. This can be a single layout or (possibly nested) tuples or dictionaries of layouts, and the structure must match the structure of the dataset. Either all or none of the layouts should be sharded on the batch dimension; having only a subset of layouts batch sharded will not work and raises a ValueError. global_batch_size: the desired global batch size. dataset_already_batched: must be set only if the dataset is already batched to the per-replica batch size. The batched dataset must have `drop_remainder=True` set since DTensor requires static shapes for slicing the input tensors. batch_dim: the mesh dimension on which the input's batch dimension is sharded. Set to None if the input layouts do not shard on the batch dimension. prefetch: number of batches to prefetch using Dataset.prefetch. tf_data_service_config: if operating in multi-client mode, this config specifies the tf.data service configuration to use. Raises: ValueError: on any of the following situations, 1. if the structures and ranks of layouts and the dataset do not match. 2. if the shapes in the dataset's spec are not fully defined. 3. if batch_dim is specified and all layouts are not batch-sharded. 4. if per_replica_batch_size is specified for an already batched Dataset but it does not match the expected per-replica size based on the provided mesh. TypeError: if type of structures of layouts and the dataset do not match. """ super().__init__(dataset, dataset_ops.to_variant(dataset)) self._mesh = mesh self._layouts = layouts self._batch_dim = batch_dim self._prefetch = prefetch self._tf_data_service_config = tf_data_service_config self._element_spec = dataset.element_spec nest.assert_same_structure(self._element_spec, self._layouts) flattened_layouts = nest.flatten(self._layouts) flattened_elem_spec = nest.flatten(self._element_spec) if batch_dim: num_global_replicas = mesh.dim_size(batch_dim) self._local_replica_ids = list( dict.fromkeys( [loc[batch_dim] for loc in mesh.local_device_locations()])) for layout in flattened_layouts: if batch_dim != layout.sharding_specs[0]: raise ValueError( ('batch_dim %s was specified but at least one layout did not ' 'contain it: %s') % (batch_dim, layout)) else: # Only one replica since there is no sharding on the batch dimension. num_global_replicas = 1 self._local_replica_ids = [0] # Validate layout and element spec compatibility, and raise ValueError if # invalid. _validate_input( flattened_layouts, flattened_elem_spec, dataset_already_batched=dataset_already_batched) expected_batch_size = global_batch_size // num_global_replicas if not dataset_already_batched: self._batched_dataset = dataset.batch( expected_batch_size, drop_remainder=True) else: per_replica_batch_size = flattened_elem_spec[0].shape.as_list()[0] if per_replica_batch_size != expected_batch_size: raise ValueError( ('per_replica_batch_size does not matched expected size based on ' 'the mesh, got %d but expected %d.') % (per_replica_batch_size, expected_batch_size)) self._batched_dataset = dataset num_global_devices_per_replica = config.num_global_devices( mesh.device_type()) // num_global_replicas self._num_local_replicas = len(self._local_replica_ids) self._num_local_devices_per_replica = mesh.num_local_devices( ) // self._num_local_replicas # The number of clients each replica is split over. self._num_clients_per_replica = ( num_global_devices_per_replica // self._num_local_devices_per_replica) # In the case where a replica is split across multiple clients, an offset # needs to be added to the index used by the partitioning logic such that # the local devices on that client can be correctly matched to slices of the # input tensor(s). If replicas are wholly contained within a client, then # this offset is always 0. self._partition_offset = (config.client_id() % self._num_clients_per_replica ) * self._num_local_devices_per_replica # Helper data structures used in partitioning the dataset tensors. self._all_shard_counts = [ _shard_counts(layout, batch_dim) for layout in flattened_layouts ] self._index_matrices = [ _index_matrix(layout, elem_spec) for layout, elem_spec in zip( flattened_layouts, flattened_elem_spec) ] def __iter__(self): datasets: List[Tuple[int, dataset_ops.DatasetV2]] = [] # Start with the batched the dataset. local_dataset = self._batched_dataset if self._batch_dim is not None: if self._num_clients_per_replica > 1: # If a replica is split over multiple clients then each batch needs to # be repeated before distribution as many times as there are clients # corresponding to that replica. local_dataset = self._repeat_batch(local_dataset, self._num_clients_per_replica) sharding_policy = data_service_ops.ShardingPolicy.DATA else: # Replicas are unique to each client, so FILE based sharding can be used # which is more performant since each worker does not need to read the # entire dataset. sharding_policy = data_service_ops.ShardingPolicy.FILE else: # No batch dimension sharding specified so disable dataset sharding during # the distribute step. sharding_policy = data_service_ops.ShardingPolicy.OFF # Apply distribution here (if specified) so all remaining transformations # are executed locally. if self._tf_data_service_config is not None: local_dataset = local_dataset.apply( data_service_ops.distribute( processing_mode=sharding_policy, service=self._tf_data_service_config.dispatcher_address, job_name=f'{self._tf_data_service_config.job_name}_{config.client_id()}', target_workers='LOCAL')) for local_replica_idx, replica_id in enumerate(self._local_replica_ids): # Select the shard for the corresponding replica. dataset = local_dataset.shard(self._num_local_replicas, local_replica_idx) # Repeat each batch for each local device in the replica. dataset = self._repeat_batch(dataset, self._num_local_devices_per_replica) # Slice each shard further for all non-batch dim shards. If there is no # non-batch dim sharding, this slice is essentially a no-op. dataset = self._partition(dataset) # Apply prefetch as the last step. Since each batch is repeated, the # number of elements to prefetch has to be scaled by the same size. if self._prefetch is not None: dataset = dataset.prefetch( self._prefetch * self._num_local_devices_per_replica) datasets.append((replica_id, dataset)) return _DTensorIterator(datasets, self._element_spec, self._layouts, self._num_local_devices_per_replica) def _repeat_batch(self, dataset, repeats): def repeat(*x): return dataset_ops.DatasetV2.from_tensors(x).repeat(repeats) return dataset.flat_map(repeat) def _partition(self, dataset): """Slices each dataset element on any sharded non-batch dimension.""" # TODO(b/223275517): decouple from self and make testable. def slice_batch(index, batch): flattened_batch = nest.flatten(batch) flattened_output = [] norm_index = math_ops.cast( index % self._num_local_devices_per_replica, dtype=dtypes.int32) norm_index += self._partition_offset coords = self._mesh.coords(norm_index) coords = array_ops.reshape(coords, (1, -1)) for element, shard_counts, idx_matrix in zip(flattened_batch, self._all_shard_counts, self._index_matrices): indexes = math_ops.matmul(coords, idx_matrix) start = array_ops.reshape(indexes, (-1,)) size = array_ops.shape_v2( element, out_type=dtypes.int32) // shard_counts flattened_output.append( array_ops.slice(element, begin=start, size=size)) return nest.pack_sequence_as(batch, flattened_output) enumerated_dataset = dataset.enumerate() partitioned_dataset = enumerated_dataset.map(slice_batch) return partitioned_dataset