@n8n/n8n-nodes-langchain
Version:
37 lines (33 loc) • 1.2 kB
text/typescript
/**
* Simple Vector Store Node - Version 1.2
* Discriminator: mode=insert
*/
/** Insert documents into vector store */
export type LcVectorStoreInMemoryV12InsertParams = {
mode: 'insert';
/**
* The key to use to store the vector memory in the workflow data. These keys are shared between workflows.
* @searchListMethod vectorStoresSearch
* @default {"mode":"list","value":"vector_store_key"}
*/
memoryKey?: { __rl: true; mode: 'list' | 'id'; value: string; cachedResultName?: string };
/**
* Number of documents to embed in a single batch
* @default 200
*/
embeddingBatchSize?: number | Expression<number>;
/**
* Whether to clear the store before inserting new data
* @default false
*/
clearStore?: boolean | Expression<boolean>;
};
export interface LcVectorStoreInMemoryV12InsertSubnodeConfig {
embedding: EmbeddingInstance | EmbeddingInstance[];
documentLoader: DocumentLoaderInstance | DocumentLoaderInstance[];
}
export type LcVectorStoreInMemoryV12InsertNode = {
type: '@n8n/n8n-nodes-langchain.vectorStoreInMemory';
version: 1.2;
config: NodeConfig<LcVectorStoreInMemoryV12InsertParams> & { subnodes: LcVectorStoreInMemoryV12InsertSubnodeConfig };
};