Principal Neighbourhood Aggregation for Graph Nets (PNA) Paper link: https://grlplus.github.io/papers/20.pdf Author's code repo (in PyTorch): https://github.com/lukecavabarrett/pna. Dataset The ZINC dataset from the ZINC database and the Automatic Chemical Design Using a Data-Driven Continuous Representation of Molecules paper, containing about 250,000 molecular graphs with up to 38 heavy atoms. Our experiments only load a subset of the dataset (12,000 molecular graphs), following the Benchmarking Graph Neural Networks paper. Results from the Paper Task Dataset Model Metric Name Metric Value Graph Regression ZINC PNA MAE 0.188±0.004 Our Results TL_BACKEND="paddle" python pna_trainer.py --batch_size 128 --lr 0.001 --n_epoch 400 TL_BACKEND="torch" python pna_trainer.py --batch_size 128 --lr 0.001 --n_epoch 400 TL_BACKEND="tensorflow" python pna_trainer.py --batch_size 128 --lr 0.001 --n_epoch 400 Dataset Our(pd) Our(torch) Our(tf) ZINC OOM 0.186 0.195(±0.006)