Rhett-Ying

import dgl
import torch
import torch.nn as nn
import torch.nn.functional as F

import dgl.data
# Generate a synthetic dataset with 10000 graphs, ranging from 10 to 500 nodes.
dataset = dgl.data.GINDataset('PROTEINS', self_loop=True)

print('Node feature dimensionality:', dataset.dim_nfeats)

Rhett-Ying

import dgl
import torch
import multiprocessing as mp

g_graph = dgl.rand_graph(20000,300000).to('cuda')
g_graph.ndata['x'] = torch.rand(g_graph.num_nodes(),3).to('cuda')

sampler = dgl.dataloading.MultiLayerNeighborSampler([15, 10])
dataloader = dgl.dataloading.NodeDataLoader(
    g_graph, torch.arange(0, g_graph.num_nodes(), dtype=torch.int64).to('cuda'), sampler, device='cuda',