Can not understand the shape of this tensor. Thank you very much.

Louis-udm / VGCN-BERT

MIT License

121 stars 35 forks source link

Closed guotong1988 closed 4 years ago

guotong1988 commented 4 years ago

gcn_vocab_out = self.vocab_gcn(vocab_adj_list, vocab_input)

gcn_vocab_out's shape is [batch_size, bert_embedding_dim, gcn_embedding_dim]

What does this tensor or shape represent?

Thank you very much. @Louis-udm @Xiang-Pan

Louis-udm commented 4 years ago

It means graph representation for every embedding dimension after vgcn.