A BERT embedding library for sentence semantic similarity measurement.
Project description
This is a sentence similarity measurement library using the forward pass of the BERT (bert-base-uncased) model. The spatial distance is computed using the cosine value between 2 semantic embedding vectors in low dimensional space. These vectors can be extracted by unique words as well as the sentence as a whole.The library also provides a flexibility for choosing any other approximators for spatial distance measurement for semantic similarity measurement.References include the BERT paper(https://arxiv.org/abs/1810.04805),Google Research BERT (https://github.com/google-research/bert/)
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.