Rerank Results — rerank • pineconer

Rerank documents according to their relevance to a query using Pinecone's hosted reranking models.

Usage

rerank(
  model,
  query,
  documents,
  top_n = NULL,
  return_documents = TRUE,
  tidy = TRUE
)

Arguments

model: The reranking model to use. Available models include: - "pinecone-rerank-v0" See Pinecone documentation for the full list of available models.
query: The query string to rank documents against.
documents: A character vector of documents to rerank, or a list of named lists where each contains a "text" field (and optionally an "id" field).
top_n: The number of top results to return. Default is to return all documents.
return_documents: Whether to include the document text in the response. Default is TRUE.
tidy: Whether to return a tidy tibble format (default: TRUE).

Value

List with http response, content (reranked results), and status_code. When tidy = TRUE, content is a tibble with columns: - index: original position of the document - score: relevance score (higher is more relevant) - document: the document text (if return_documents = TRUE)

Examples

if (FALSE) { # \dontrun{
# Rerank search results
documents <- c(
  "The quick brown fox jumps over the lazy dog",
  "A fast auburn fox leaps above a sleepy canine",
  "The weather is nice today"
)

result <- rerank(
  model = "pinecone-rerank-v0",
  query = "What did the fox do?",
  documents = documents,
  top_n = 2
)

# View reranked results
result$content

# Rerank with document IDs
docs_with_ids <- list(
  list(id = "doc1", text = "The quick brown fox jumps over the lazy dog"),
  list(id = "doc2", text = "A fast auburn fox leaps above a sleepy canine"),
  list(id = "doc3", text = "The weather is nice today")
)

result <- rerank(
  model = "pinecone-rerank-v0",
  query = "What did the fox do?",
  documents = docs_with_ids
)
} # }