Jaseci Kit is a collection of state-of-the-art machine learning models that are readily available to load into jaseci. These pre-built models help to make development even faster in Jaseci. Below we discuss some of these.
Feel free to jump right into the developer documentation to review our model directory, code examples and kickstart guides.
Jaseci Kit includes encoders for sentence-level embedding using general text corpus, Q&A data corpus, FastText text classifier, dual sentence-level encoders, poly encoders, cross encoders and more.
Jaseci Kit includes entity support for entity extraction using the FLAIR NER framework, token classification on transformer models and entity extraction/slot filling via long-short term memory networks.
Jaseci Kit current includes a summarizer that does extractive summarization using Sumy
Apart from the AI tools above, Jaseci Kit also includes some non-ai tools including a PDF extractor used to extract content from PDFs using PyPDF2