RAG
RAG (Retrieval-Augmented Generation) components process a user query by retrieving relevant documents and generating a concise summary that addresses the user's question.
Vectara RAG​
This component leverages Vectara's Retrieval Augmented Generation (RAG) capabilities to search and summarize documents based on the provided input. For more information, see the Vectara documentation.
Parameters​
Inputs​
| Name | Type | Description |
|---|---|---|
| vectara_customer_id | String | Vectara customer ID |
| vectara_corpus_id | String | Vectara corpus ID |
| vectara_api_key | SecretString | Vectara API key |
| search_query | String | The query to receive an answer on |
| lexical_interpolation | Float | Hybrid search factor (0.005 to 0.1) |
| filter | String | Metadata filters to narrow the search |
| reranker | String | Reranker type (mmr, rerank_multilingual_v1, none) |
| reranker_k | Integer | Number of results to rerank (1 to 100) |
| diversity_bias | Float | Diversity bias for MMR reranker (0 to 1) |
| max_results | Integer | Maximum number of search results to summarize (1 to 100) |
| response_lang | String | Language code for the response (e.g., "eng", "auto") |
| prompt | String | Prompt name for summarization |
Outputs​
| Name | Type | Description |
|---|---|---|
| answer | Message | Generated RAG response |
