Text Embedder

This node applies the provided embeddings model to create embeddings for the texts contained in a string column of the input table. At its core, a text embedding is a dense vector of floating point values capturing the semantic meaning of the text. Thus these embeddings are often used to find semantically similar documents e.g. in vector stores. How exactly the embeddings are derived depends on the used embeddings model but typically the embeddings are the internal representations used by deep language models e.g. GPTs.