Embedding of text. Sentence-based or word-based embedding?
Research ways to embed words (vectorization, Word2Vec as example).
Create a seperate trainable embedding model so possible to get raw outputs, and use between models. Figure out best method to store weights outside GitHub, non locally.
Embedding of text. Sentence-based or word-based embedding?
https://github.com/duggurd/graduation_project/blob/main/src/models/text_embedding.ipynb