langchain_community.cache
.AstraDBSemanticCache¶
- class langchain_community.cache.AstraDBSemanticCache(*, collection_name: str = 'langchain_astradb_semantic_cache', token: Optional[str] = None, api_endpoint: Optional[str] = None, astra_db_client: Optional[AstraDB] = None, async_astra_db_client: Optional[AsyncAstraDB] = None, namespace: Optional[str] = None, setup_mode: SetupMode = SetupMode.SYNC, pre_delete_collection: bool = False, embedding: Embeddings, metric: Optional[str] = None, similarity_threshold: float = 0.85)[source]¶
Cache that uses Astra DB as a vector-store backend for semantic (i.e. similarity-based) lookup.
It uses a single (vector) collection and can store cached values from several LLMs, so the LLMâs âllm_stringâ is stored in the document metadata.
You can choose the preferred similarity (or use the API default). The default score threshold is tuned to the default metric. Tune it carefully yourself if switching to another distance metric.
- Parameters
collection_name (str) â name of the Astra DB collection to create/use.
token (Optional[str]) â API token for Astra DB usage.
api_endpoint (Optional[str]) â full URL to the API endpoint, such as https://<DB-ID>-us-east1.apps.astra.datastax.com.
astra_db_client (Optional[AstraDB]) â alternative to token+api_endpoint, you can pass an already-created âastrapy.db.AstraDBâ instance.
async_astra_db_client (Optional[AsyncAstraDB]) â alternative to token+api_endpoint, you can pass an already-created âastrapy.db.AsyncAstraDBâ instance.
namespace (Optional[str]) â namespace (aka keyspace) where the collection is created. Defaults to the databaseâs âdefault namespaceâ.
setup_mode (SetupMode) â mode used to create the Astra DB collection (SYNC, ASYNC or OFF).
pre_delete_collection (bool) â whether to delete the collection before creating it. If False and the collection already exists, the collection will be used as is.
embedding (Embeddings) â Embedding provider for semantic encoding and search.
metric (Optional[str]) â the function to use for evaluating similarity of text embeddings. Defaults to âcosineâ (alternatives: âeuclideanâ, âdot_productâ)
similarity_threshold (float) â the minimum similarity for accepting a (semantic-search) match.
Methods
__init__
(*[, collection_name, token, ...])Cache that uses Astra DB as a vector-store backend for semantic (i.e.
aclear
(**kwargs)Clear cache that can take additional keyword arguments.
adelete_by_document_id
(document_id)Given this is a "similarity search" cache, an invalidation pattern that makes sense is first a lookup to get an ID, and then deleting with that ID.
alookup
(prompt, llm_string)Look up based on prompt and llm_string.
alookup_with_id
(prompt, llm_string)Look up based on prompt and llm_string.
alookup_with_id_through_llm
(prompt, llm[, stop])aupdate
(prompt, llm_string, return_val)Update cache based on prompt and llm_string.
clear
(**kwargs)Clear cache that can take additional keyword arguments.
delete_by_document_id
(document_id)Given this is a "similarity search" cache, an invalidation pattern that makes sense is first a lookup to get an ID, and then deleting with that ID.
lookup
(prompt, llm_string)Look up based on prompt and llm_string.
lookup_with_id
(prompt, llm_string)Look up based on prompt and llm_string.
lookup_with_id_through_llm
(prompt, llm[, stop])update
(prompt, llm_string, return_val)Update cache based on prompt and llm_string.
- __init__(*, collection_name: str = 'langchain_astradb_semantic_cache', token: Optional[str] = None, api_endpoint: Optional[str] = None, astra_db_client: Optional[AstraDB] = None, async_astra_db_client: Optional[AsyncAstraDB] = None, namespace: Optional[str] = None, setup_mode: SetupMode = SetupMode.SYNC, pre_delete_collection: bool = False, embedding: Embeddings, metric: Optional[str] = None, similarity_threshold: float = 0.85)[source]¶
Cache that uses Astra DB as a vector-store backend for semantic (i.e. similarity-based) lookup.
It uses a single (vector) collection and can store cached values from several LLMs, so the LLMâs âllm_stringâ is stored in the document metadata.
You can choose the preferred similarity (or use the API default). The default score threshold is tuned to the default metric. Tune it carefully yourself if switching to another distance metric.
- Parameters
collection_name (str) â name of the Astra DB collection to create/use.
token (Optional[str]) â API token for Astra DB usage.
api_endpoint (Optional[str]) â full URL to the API endpoint, such as https://<DB-ID>-us-east1.apps.astra.datastax.com.
astra_db_client (Optional[AstraDB]) â alternative to token+api_endpoint, you can pass an already-created âastrapy.db.AstraDBâ instance.
async_astra_db_client (Optional[AsyncAstraDB]) â alternative to token+api_endpoint, you can pass an already-created âastrapy.db.AsyncAstraDBâ instance.
namespace (Optional[str]) â namespace (aka keyspace) where the collection is created. Defaults to the databaseâs âdefault namespaceâ.
setup_mode (SetupMode) â mode used to create the Astra DB collection (SYNC, ASYNC or OFF).
pre_delete_collection (bool) â whether to delete the collection before creating it. If False and the collection already exists, the collection will be used as is.
embedding (Embeddings) â Embedding provider for semantic encoding and search.
metric (Optional[str]) â the function to use for evaluating similarity of text embeddings. Defaults to âcosineâ (alternatives: âeuclideanâ, âdot_productâ)
similarity_threshold (float) â the minimum similarity for accepting a (semantic-search) match.
- async aclear(**kwargs: Any) None [source]¶
Clear cache that can take additional keyword arguments.
- Parameters
kwargs (Any) â
- Return type
None
- async adelete_by_document_id(document_id: str) None [source]¶
Given this is a âsimilarity searchâ cache, an invalidation pattern that makes sense is first a lookup to get an ID, and then deleting with that ID. This is for the second step.
- Parameters
document_id (str) â
- Return type
None
- async alookup(prompt: str, llm_string: str) Optional[Sequence[Generation]] [source]¶
Look up based on prompt and llm_string.
- Parameters
prompt (str) â
llm_string (str) â
- Return type
Optional[Sequence[Generation]]
- async alookup_with_id(prompt: str, llm_string: str) Optional[Tuple[str, Sequence[Generation]]] [source]¶
Look up based on prompt and llm_string. If there are hits, return (document_id, cached_entry) for the top hit
- Parameters
prompt (str) â
llm_string (str) â
- Return type
Optional[Tuple[str, Sequence[Generation]]]
- async alookup_with_id_through_llm(prompt: str, llm: LLM, stop: Optional[List[str]] = None) Optional[Tuple[str, Sequence[Generation]]] [source]¶
- Parameters
prompt (str) â
llm (LLM) â
stop (Optional[List[str]]) â
- Return type
Optional[Tuple[str, Sequence[Generation]]]
- async aupdate(prompt: str, llm_string: str, return_val: Sequence[Generation]) None [source]¶
Update cache based on prompt and llm_string.
- Parameters
prompt (str) â
llm_string (str) â
return_val (Sequence[Generation]) â
- Return type
None
- clear(**kwargs: Any) None [source]¶
Clear cache that can take additional keyword arguments.
- Parameters
kwargs (Any) â
- Return type
None
- delete_by_document_id(document_id: str) None [source]¶
Given this is a âsimilarity searchâ cache, an invalidation pattern that makes sense is first a lookup to get an ID, and then deleting with that ID. This is for the second step.
- Parameters
document_id (str) â
- Return type
None
- lookup(prompt: str, llm_string: str) Optional[Sequence[Generation]] [source]¶
Look up based on prompt and llm_string.
- Parameters
prompt (str) â
llm_string (str) â
- Return type
Optional[Sequence[Generation]]
- lookup_with_id(prompt: str, llm_string: str) Optional[Tuple[str, Sequence[Generation]]] [source]¶
Look up based on prompt and llm_string. If there are hits, return (document_id, cached_entry) for the top hit
- Parameters
prompt (str) â
llm_string (str) â
- Return type
Optional[Tuple[str, Sequence[Generation]]]
- lookup_with_id_through_llm(prompt: str, llm: LLM, stop: Optional[List[str]] = None) Optional[Tuple[str, Sequence[Generation]]] [source]¶
- Parameters
prompt (str) â
llm (LLM) â
stop (Optional[List[str]]) â
- Return type
Optional[Tuple[str, Sequence[Generation]]]
- update(prompt: str, llm_string: str, return_val: Sequence[Generation]) None [source]¶
Update cache based on prompt and llm_string.
- Parameters
prompt (str) â
llm_string (str) â
return_val (Sequence[Generation]) â
- Return type
None