Skip to contents

Generate Embeddings

Source: R/semantic_analysis.R
generate_embeddings.Rd

Generates L2-normalized embeddings for texts using sentence transformers.

Usage

generate_embeddings(texts, model = "all-MiniLM-L6-v2", verbose = TRUE)

Arguments

texts

A character vector of texts.

model

Sentence transformer model name (default: "all-MiniLM-L6-v2"). Newer small models (e.g. "BAAI/bge-small-en-v1.5") often score higher on retrieval benchmarks but expect instruction prefixes this function does not add.

verbose

Logical, if TRUE, prints progress messages.

Value

A matrix of L2-normalized embeddings.