Struct VectorMemoryStore 

pub struct VectorMemoryStore<E, V> { /* private fields */ }

A long-term memory that stores conversation turns as vector embeddings and retrieves the most relevant memories for each new query.

Implementations

impl<E, V> VectorMemoryStore<E, V>

where E: Embedding + Send + Sync + 'static, V: VectorStore + Send + Sync + 'static,

pub fn new(embedder: Arc<E>, store: Arc<V>, top_k: usize) -> Self

Create a new VectorMemoryStore.

• embedder: converts text to vectors
• store: the underlying vector store
• top_k: number of memories to retrieve per query

pub fn with_memory_key(self, key: impl Into<String>) -> Self

Override the key used in the returned memory variables map.

Trait Implementations

impl<E, V> Memory for VectorMemoryStore<E, V>

where E: Embedding + Send + Sync + 'static, V: VectorStore + Send + Sync + 'static,

fn load_memory_variables<'life0, 'life1, 'async_trait>( &'life0 self, thread_id: &'life1 str, ) -> Pin<Box<dyn Future<Output = Result<HashMap<String, Value>, WesichainError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Return key-value pairs of memory variables (e.g., chat history)

fn save_context<'life0, 'life1, 'life2, 'life3, 'async_trait>( &'life0 self, thread_id: &'life1 str, inputs: &'life2 HashMap<String, Value>, outputs: &'life3 HashMap<String, Value>, ) -> Pin<Box<dyn Future<Output = Result<(), WesichainError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait, 'life2: 'async_trait, 'life3: 'async_trait,

Save context from this interaction to memory

fn clear<'life0, 'life1, 'async_trait>( &'life0 self, _thread_id: &'life1 str, ) -> Pin<Box<dyn Future<Output = Result<(), WesichainError>> + Send + 'async_trait>>

where Self: 'async_trait, 'life0: 'async_trait, 'life1: 'async_trait,

Clear memory for the given thread