Memory Issue when running DGL baseline model with Nvidia Tesla V100, 120GB RAM

[luishengjie]

When running the DGL baseline model https://github.com/dmlc/dgl/blob/master/examples/pytorch/ogb_lsc/MAG240M/train.py, I faced the following error OSError: [Errno 12] Cannot allocate memory.
Machine Specification:

• Nvidia Tesla V100
• 120GB RAM

Is 120GB RAM sufficient to run the DGL baseline model?

Best Regards
Sheng Jie

[weihua916]

Please ask on DGL's Github: https://github.com/dmlc/dgl They are maintaining the code.