Update the model's neural weights from a sequence of sentences (can be a once-only generator stream). Each sentence must be a list of utf8 string

Build vocabulary from a sequence of sentences (can be a once-only generator stream). Each sentence must be a list of utf8 strings.

Precompute L2-normalized vectors. If `replace` is set, forget the original vectors and only keep the normalized ones = saves lots of memory!

PVDM model for training and learning context paragraphs for sentiment and topic analysis, or information retrieval. This method uses hierarch

Train using SGD with varying batchsizes the PV-DM machine to learn paragraph representations. Inputs ------ sentences: the corpus to lear

Find the top-N most similar paragraphs.

Train the model, lifting lists of sentences from the jobs queue.