Support Node2Vec graph sampling algorithm

[AntiTopQuark]

Relate Issue: #716

Use a single thread to generate a Path string, and then use the Word2Vec model to generate a word vector. The word vector will be stored in the Feature part, i.e. NodeInfo[1]

[root@5bcae90fa0cb output]# python3 train_cora.py --method=node2vec_sampling > ./exec.log
/usr/local/lib/python3.6/site-packages/dgl/dataloading/dataloader.py:869: DGLWarning: Dataloader CPU affinity opt is not enabled, consider switching it on (see enable_cpu_affinity() or CPU best practices for DGL [https://docs.dgl.ai/tutorials/cpu/cpu_best_practises.html])
  dgl_warning(f'Dataloader CPU affinity opt is not enabled, consider switching it on '
/usr/local/lib/python3.6/site-packages/dgl/dataloading/dataloader.py:869: DGLWarning: Dataloader CPU affinity opt is not enabled, consider switching it on (see enable_cpu_affinity() or CPU best practices for DGL [https://docs.dgl.ai/tutorials/cpu/cpu_best_practises.html])
  dgl_warning(f'Dataloader CPU affinity opt is not enabled, consider switching it on '
[root@5bcae90fa0cb output]# head -n 5 exec.log
Loading DB state from disk
Loading DB state from disk
Loading DB state from disk
Loading DB state from disk
Loading DB state from disk
[root@5bcae90fa0cb output]# tail -n 5 exec.log
loss:   0.9962489008903503
loss:   0.8431514501571655
loss:   0.8940250277519226
epoch: 1 loss: 0.8940250277519226
The loss value is less than 0.9