entity_detection

• for every sentence who was the trump ocean club international hotel and tower named after
• in data, we can find dev.txt and dev.pt and word_vocab.pt
  • in dev.txt , we can find sentence like what/O is/O a/O film/O directed/O by/O wiebke/I von/I carolsfeld/I ?/O
  • in dev.pt, we can use self.seqs, self.seq_lens, self.seq_labels = torch.load(infile) , we can get three lists.
    • seq in seqs is a Tensor size(batch_size, seq_len)
    • seq_len in seq_lens is a Tensor size(batch_size)
    • seq_label in seq_labels is a Tensor size(batch_size, seq_len)