train_sampling_multi_gpu.py

import dgl
import numpy as np
import torch as th
import torch.nn as nn
import torch.nn.functional as F
import torch.optim as optim
import torch.multiprocessing as mp
from torch.utils.data import DataLoader
import dgl.function as fn
import dgl.nn.pytorch as dglnn
import time
import argparse
from dgl.data import RedditDataset
from torch.nn.parallel import DistributedDataParallel
import tqdm
import traceback

from utils import thread_wrapped_func
from load_graph import load_reddit, inductive_split

class SAGE(nn.Module):
    def __init__(self,
                 in_feats,
                 n_hidden,
                 n_classes,
                 n_layers,
                 activation,
                 dropout):
        super().__init__()
        self.n_layers = n_layers
        self.n_hidden = n_hidden
        self.n_classes = n_classes
        self.layers = nn.ModuleList()
        self.layers.append(dglnn.SAGEConv(in_feats, n_hidden, 'mean'))
        for i in range(1, n_layers - 1):
            self.layers.append(dglnn.SAGEConv(n_hidden, n_hidden, 'mean'))
        self.layers.append(dglnn.SAGEConv(n_hidden, n_classes, 'mean'))
        self.dropout = nn.Dropout(dropout)
        self.activation = activation

    def forward(self, blocks, x):
        h = x
        for l, (layer, block) in enumerate(zip(self.layers, blocks)):
            # We need to first copy the representation of nodes on the RHS from the
            # appropriate nodes on the LHS.
            # Note that the shape of h is (num_nodes_LHS, D) and the shape of h_dst
            # would be (num_nodes_RHS, D)
            h_dst = h[:block.number_of_dst_nodes()]
            # Then we compute the updated representation on the RHS.
            # The shape of h now becomes (num_nodes_RHS, D)
            h = layer(block, (h, h_dst))
            if l != len(self.layers) - 1:
                h = self.activation(h)
                h = self.dropout(h)
        return h

    def inference(self, g, x, batch_size, device):
        """
        Inference with the GraphSAGE model on full neighbors (i.e. without neighbor sampling).
        g : the entire graph.
        x : the input of entire node set.

        The inference code is written in a fashion that it could handle any number of nodes and
        layers.
        """
        # During inference with sampling, multi-layer blocks are very inefficient because
        # lots of computations in the first few layers are repeated.
        # Therefore, we compute the representation of all nodes layer by layer.  The nodes
        # on each layer are of course splitted in batches.
        # TODO: can we standardize this?
        nodes = th.arange(g.number_of_nodes())
        for l, layer in enumerate(self.layers):
            y = th.zeros(g.number_of_nodes(), self.n_hidden if l != len(self.layers) - 1 else self.n_classes)

            sampler = dgl.sampling.MultiLayerNeighborSampler([None])
            dataloader = dgl.sampling.NodeDataLoader(
                g,
                th.arange(g.number_of_nodes()),
                sampler,
                batch_size=args.batch_size,
                shuffle=True,
                drop_last=False,
                num_workers=args.num_workers)

            for input_nodes, output_nodes, blocks in tqdm.tqdm(dataloader):
                block = blocks[0]

                h = x[input_nodes].to(device)
                h_dst = h[:block.number_of_dst_nodes()]
                h = layer(block, (h, h_dst))
                if l != len(self.layers) - 1:
                    h = self.activation(h)
                    h = self.dropout(h)

                y[output_nodes] = h.cpu()

            x = y
        return y

def prepare_mp(g):
    """
    Explicitly materialize the CSR, CSC and COO representation of the given graph
    so that they could be shared via copy-on-write to sampler workers and GPU
    trainers.

    This is a workaround before full shared memory support on heterogeneous graphs.
    """
    g.in_degree(0)
    g.out_degree(0)
    g.find_edges([0])

def compute_acc(pred, labels):
    """
    Compute the accuracy of prediction given the labels.
    """
    return (th.argmax(pred, dim=1) == labels).float().sum() / len(pred)

def evaluate(model, g, inputs, labels, val_nid, batch_size, device):
    """
    Evaluate the model on the validation set specified by ``val_nid``.
    g : The entire graph.
    inputs : The features of all the nodes.
    labels : The labels of all the nodes.
    val_nid : A node ID tensor indicating which nodes do we actually compute the accuracy for.
    batch_size : Number of nodes to compute at the same time.
    device : The GPU device to evaluate on.
    """
    model.eval()
    with th.no_grad():
        pred = model.inference(g, inputs, batch_size, device)
    model.train()
    return compute_acc(pred[val_nid], labels[val_nid])

def load_subtensor(g, labels, seeds, input_nodes, dev_id):
    """
    Copys features and labels of a set of nodes onto GPU.
    """
    batch_inputs = g.ndata['features'][input_nodes].to(dev_id)
    batch_labels = labels[seeds].to(dev_id)
    return batch_inputs, batch_labels

#### Entry point

def run(proc_id, n_gpus, args, devices, data):
    # Start up distributed training, if enabled.
    dev_id = devices[proc_id]
    if n_gpus > 1:
        dist_init_method = 'tcp://{master_ip}:{master_port}'.format(
            master_ip='127.0.0.1', master_port='12345')
        world_size = n_gpus
        th.distributed.init_process_group(backend="nccl",
                                          init_method=dist_init_method,
                                          world_size=world_size,
                                          rank=proc_id)
    th.cuda.set_device(dev_id)

    # Unpack data
    in_feats, n_classes, train_g, val_g, test_g = data
    train_mask = train_g.ndata['train_mask']
    val_mask = val_g.ndata['val_mask']
    test_mask = ~(train_g.ndata['train_mask'] | val_g.ndata['val_mask'])
    train_nid = train_mask.nonzero()[:, 0]
    val_nid = val_mask.nonzero()[:, 0]
    test_nid = test_mask.nonzero()[:, 0]

    # Split train_nid
    train_nid = th.split(train_nid, len(train_nid) // n_gpus)[proc_id]

    # Create PyTorch DataLoader for constructing blocks
    sampler = dgl.sampling.MultiLayerNeighborSampler(
        [int(fanout) for fanout in args.fan_out.split(',')])
    dataloader = dgl.sampling.NodeDataLoader(
        train_g,
        train_nid,
        sampler,
        batch_size=args.batch_size,
        shuffle=True,
        drop_last=False,
        num_workers=args.num_workers)

    # Define model and optimizer
    model = SAGE(in_feats, args.num_hidden, n_classes, args.num_layers, F.relu, args.dropout)
    model = model.to(dev_id)
    if n_gpus > 1:
        model = DistributedDataParallel(model, device_ids=[dev_id], output_device=dev_id)
    loss_fcn = nn.CrossEntropyLoss()
    loss_fcn = loss_fcn.to(dev_id)
    optimizer = optim.Adam(model.parameters(), lr=args.lr)

    # Training loop
    avg = 0
    iter_tput = []
    for epoch in range(args.num_epochs):
        tic = time.time()

        # Loop over the dataloader to sample the computation dependency graph as a list of
        # blocks.
        for step, (input_nodes, seeds, blocks) in enumerate(dataloader):
            if proc_id == 0:
                tic_step = time.time()

            # Load the input features as well as output labels
            batch_inputs, batch_labels = load_subtensor(train_g, train_g.ndata['labels'], seeds, input_nodes, dev_id)

            # Compute loss and prediction
            batch_pred = model(blocks, batch_inputs)
            loss = loss_fcn(batch_pred, batch_labels)
            optimizer.zero_grad()
            loss.backward()
            optimizer.step()

            if proc_id == 0:
                iter_tput.append(len(seeds) * n_gpus / (time.time() - tic_step))
            if step % args.log_every == 0 and proc_id == 0:
                acc = compute_acc(batch_pred, batch_labels)
                print('Epoch {:05d} | Step {:05d} | Loss {:.4f} | Train Acc {:.4f} | Speed (samples/sec) {:.4f} | GPU {:.1f} MiB'.format(
                    epoch, step, loss.item(), acc.item(), np.mean(iter_tput[3:]), th.cuda.max_memory_allocated() / 1000000))

        if n_gpus > 1:
            th.distributed.barrier()

        toc = time.time()
        if proc_id == 0:
            print('Epoch Time(s): {:.4f}'.format(toc - tic))
            if epoch >= 5:
                avg += toc - tic
            if epoch % args.eval_every == 0 and epoch != 0:
                if n_gpus == 1:
                    eval_acc = evaluate(
                        model, val_g, val_g.ndata['features'], val_g.ndata['labels'], val_nid, args.batch_size, devices[0])
                    test_acc = evaluate(
                        model, test_g, test_g.ndata['features'], test_g.ndata['labels'], test_nid, args.batch_size, devices[0])
                else:
                    eval_acc = evaluate(
                        model.module, val_g, val_g.ndata['features'], val_g.ndata['labels'], val_nid, args.batch_size, devices[0])
                    test_acc = evaluate(
                        model.module, test_g, test_g.ndata['features'], test_g.ndata['labels'], test_nid, args.batch_size, devices[0])
                print('Eval Acc {:.4f}'.format(eval_acc))
                print('Test Acc: {:.4f}'.format(test_acc))


    if n_gpus > 1:
        th.distributed.barrier()
    if proc_id == 0:
        print('Avg epoch time: {}'.format(avg / (epoch - 4)))

if __name__ == '__main__':
    argparser = argparse.ArgumentParser("multi-gpu training")
    argparser.add_argument('--gpu', type=str, default='0',
        help="Comma separated list of GPU device IDs.")
    argparser.add_argument('--num-epochs', type=int, default=20)
    argparser.add_argument('--num-hidden', type=int, default=16)
    argparser.add_argument('--num-layers', type=int, default=2)
    argparser.add_argument('--fan-out', type=str, default='10,25')
    argparser.add_argument('--batch-size', type=int, default=1000)
    argparser.add_argument('--log-every', type=int, default=20)
    argparser.add_argument('--eval-every', type=int, default=5)
    argparser.add_argument('--lr', type=float, default=0.003)
    argparser.add_argument('--dropout', type=float, default=0.5)
    argparser.add_argument('--num-workers', type=int, default=0,
        help="Number of sampling processes. Use 0 for no extra process.")
    argparser.add_argument('--inductive', action='store_true',
        help="Inductive learning setting")
    args = argparser.parse_args()
    
    devices = list(map(int, args.gpu.split(',')))
    n_gpus = len(devices)

    g, n_classes = load_reddit()
    # Construct graph
    g = dgl.as_heterograph(g)
    in_feats = g.ndata['features'].shape[1]

    if args.inductive:
        train_g, val_g, test_g = inductive_split(g)
    else:
        train_g = val_g = test_g = g

    prepare_mp(train_g)
    prepare_mp(val_g)
    prepare_mp(test_g)
    # Pack data
    data = in_feats, n_classes, train_g, val_g, test_g

    if n_gpus == 1:
        run(0, n_gpus, args, devices, data)
    else:
        procs = []
        for proc_id in range(n_gpus):
            p = mp.Process(target=thread_wrapped_func(run),
                           args=(proc_id, n_gpus, args, devices, data))
            p.start()
            procs.append(p)
        for p in procs:
            p.join()