7.3 Programming APIs¶
This section covers the core python components commonly used in a training script. DGL provides three distributed data structures and various APIs for initialization, distributed sampling and workload split.
DistGraphfor accessing structure and feature of a distributedly stored graph.
DistTensorfor accessing node/edge feature tensor that is partitioned across machines.
DistEmbeddingfor accessing learnable node/edge embedding tensor that is partitioned across machines.
Initialization of the DGL distributed module¶
dgl.distributed.initialize() initializes the distributed module. If invoked
by a trainer, this API creates sampler processes and builds connections with graph
servers; if invoked by graph server, this API starts a service loop to listen to
trainer/sampler requests. The API must be called before
torch.distributed.init_process_group() and any other
as shown in the order below:
If the training script contains user-defined functions (UDFs) that have to be invoked on
the servers (see the section of DistTensor and DistEmbedding for more details), these UDFs have to
be declared before
DistGraph is a Python class to access the graph
structure and node/edge features in a cluster of machines. Each machine is
responsible for one and only one partition. It loads the partition data (the
graph structure and the node data and edge data in the partition) and makes it
accessible to all trainers in the cluster.
provides a small subset of
DGLGraph APIs for data access.
Distributed mode vs. standalone mode¶
DistGraph can run in two modes: distributed mode and standalone mode.
When a user executes a training script in a Python command line or Jupyter Notebook, it runs in
a standalone mode. That is, it runs all computation in a single process and does not communicate
with any other processes. Thus, the standalone mode requires the input graph to have only one partition.
This mode is mainly used for development and testing (e.g., develop and run the code in Jupyter Notebook).
When a user executes a training script with a launch script (see the section of launch script),
DistGraph runs in the distributed mode. The launch tool starts servers
(node/edge feature access and graph sampling) behind the scene and loads the partition data in
each machine automatically.
DistGraph connects with the servers in the cluster
of machines and access them through the network.
In the distributed mode, the creation of
requires the graph name given during graph partitioning. The graph name
identifies the graph loaded in the cluster.
import dgl g = dgl.distributed.DistGraph('graph_name')
When running in the standalone mode, it loads the graph data in the local machine. Therefore, users need to provide the partition configuration file, which contains all information about the input graph.
import dgl g = dgl.distributed.DistGraph('graph_name', part_config='data/graph_name.json')
DGL only allows one single
DistGraph object. The behavior
of destroying a DistGraph and creating a new one is undefined.
Accessing graph structure¶
DistGraph provides a set of APIs to
access the graph structure. Currently, most APIs provide graph information,
such as the number of nodes and edges. The main use case of DistGraph is to run
sampling APIs to support mini-batch training (see Distributed sampling).
Access node/edge data¶
to access data in nodes and edges.
The difference is that
DistTensor, instead of the tensor of the underlying framework.
Users can also assign a new
DistGraph as node data or edge data.
g.ndata['train_mask'] # <dgl.distributed.dist_graph.DistTensor at 0x7fec820937b8> g.ndata['train_mask'] # tensor(, dtype=torch.uint8)
As mentioned earlier, DGL shards node/edge features and stores them in a cluster of machines. DGL provides distributed tensors with a tensor-like interface to access the partitioned node/edge features in the cluster. In the distributed setting, DGL only supports dense node/edge features.
DistTensor manages the dense tensors partitioned and stored in
multiple machines. Right now, a distributed tensor has to be associated with nodes or edges
of a graph. In other words, the number of rows in a DistTensor has to be the same as the number
of nodes or the number of edges in a graph. The following code creates a distributed tensor.
In addition to the shape and dtype for the tensor, a user can also provide a unique tensor name.
This name is useful if a user wants to reference a persistent distributed tensor (the one exists
in the cluster even if the
DistTensor object disappears).
tensor = dgl.distributed.DistTensor((g.number_of_nodes(), 10), th.float32, name='test')
DistTensor creation is a synchronized operation. All trainers
have to invoke the creation and the creation succeeds only when all trainers call it.
g.ndata['feat'] = tensor
The node data name and the tensor name do not have to be the same. The former identifies
node data from
DistGraph (in the trainer process) while the latter
identifies a distributed tensor in DGL servers.
DistTensor has the same APIs as
regular tensors to access its metadata, such as the shape and dtype. It also
supports indexed reads and writes but does not support
computation operators, such as sum and mean.
data = g.ndata['feat'][[1, 2, 3]] print(data) g.ndata['feat'][[3, 4, 5]] = data
Currently, DGL does not provide protection for concurrent writes from multiple trainers when a machine runs multiple servers. This may result in data corruption. One way to avoid concurrent writes to the same row of data is to run one server process on a machine.
DistEmbedding to support transductive models that require
node embeddings. Creating distributed embeddings is very similar to creating distributed tensors.
def initializer(shape, dtype): arr = th.zeros(shape, dtype=dtype) arr.uniform_(-1, 1) return arr emb = dgl.distributed.DistEmbedding(g.number_of_nodes(), 10, init_func=initializer)
Internally, distributed embeddings are built on top of distributed tensors, and, thus, has very similar behaviors to distributed tensors. For example, when embeddings are created, they are sharded and stored across all machines in the cluster. It can be uniquely identified by a name.
The initializer function is invoked in the server process. Therefore, it has to be
Because the embeddings are part of the model, a user has to attach them to an
optimizer for mini-batch training. Currently, DGL provides a sparse Adagrad
SparseAdagrad (DGL will add more optimizers
for sparse embeddings later). Users need to collect all distributed embeddings
from a model and pass them to the sparse optimizer. If a model has both node
embeddings and regular dense model parameters and users want to perform sparse
updates on the embeddings, they need to create two optimizers, one for node
embeddings and the other for dense model parameters, as shown in the code
sparse_optimizer = dgl.distributed.SparseAdagrad([emb], lr=lr1) optimizer = th.optim.Adam(model.parameters(), lr=lr2) feats = emb(nids) loss = model(feats) loss.backward() optimizer.step() sparse_optimizer.step()
DistEmbedding does not inherit
so we recommend using it outside of your own NN module.
DGL provides two levels of APIs for sampling nodes and edges to generate
mini-batches (see the section of mini-batch training). The low-level APIs
require users to write code to explicitly define how a layer of nodes are
sampled (e.g., using
dgl.sampling.sample_neighbors() ). The high-level
sampling APIs implement a few popular sampling algorithms for node
classification and link prediction tasks (e.g.,
The distributed sampling module follows the same design and provides two levels
of sampling APIs. For the lower-level sampling API, it provides
sample_neighbors() for distributed neighborhood sampling
DistGraph. In addition, DGL provides a distributed
DistDataLoader ) for distributed
sampling. The distributed DataLoader has the same interface as Pytorch
DataLoader except that users cannot specify the number of worker processes when
creating a dataloader. The worker processes are created in
DistGraph, the sampler cannot run in Pytorch
DataLoader with multiple worker processes. The main reason is that Pytorch
DataLoader creates new sampling worker processes in every epoch, which
leads to creating and destroying
objects many times.
def sample_blocks(seeds): seeds = th.LongTensor(np.asarray(seeds)) blocks =  for fanout in [10, 25]: frontier = dgl.distributed.sample_neighbors(g, seeds, fanout, replace=True) block = dgl.to_block(frontier, seeds) seeds = block.srcdata[dgl.NID] blocks.insert(0, block) return blocks dataloader = dgl.distributed.DistDataLoader(dataset=train_nid, batch_size=batch_size, collate_fn=sample_blocks, shuffle=True) for batch in dataloader: ...
The high-level sampling APIs (
EdgeDataLoader ) has distributed counterparts
DistEdgeDataLoader). The code is exactly the same as
single-process sampling otherwise.
sampler = dgl.sampling.MultiLayerNeighborSampler([10, 25]) dataloader = dgl.sampling.DistNodeDataLoader(g, train_nid, sampler, batch_size=batch_size, shuffle=True) for batch in dataloader: ...
To train a model, users first need to split the dataset into training,
validation and test sets. For distributed training, this step is usually done
before we invoke
dgl.distributed.partition_graph() to partition a graph.
We recommend to store the data split in boolean arrays as node data or edge
data. For node classification tasks, the length of these boolean arrays is the
number of nodes in a graph and each of their elements indicates the existence
of a node in a training/validation/test set. Similar boolean arrays should be
used for link prediction tasks.
these boolean arrays (because they are stored as the node data or edge data of
the graph) based on the graph partitioning result and store them with graph
During distributed training, users need to assign training nodes/edges to each
trainer. Similarly, we also need to split the validation and test set in the
same way. DGL provides
edge_split() to split the training, validation and test
set at runtime for distributed training. The two functions take the boolean
arrays constructed before graph partitioning as input, split them and return a
portion for the local trainer. By default, they ensure that all portions have
the same number of nodes/edges. This is important for synchronous SGD, which
assumes each trainer has the same number of mini-batches.
The example below splits the training set and returns a subset of nodes for the local process.
train_nids = dgl.distributed.node_split(g.ndata['train_mask'])