Skip to content

armgit / mem0

Public Imported from GitHub
0

The Memory Layer for AI Agents - Drop-in memory infrastructure for AI agents and apps. Context that persists. Built for production.

Added threshold to search (#2899)

Prateek Chhikarabe37fca1bba5
diff --git a/mem0/memory/main.py b/mem0/memory/main.py
index 9aee0acbd..f1b26f4d6 100644
--- a/mem0/memory/main.py
+++ b/mem0/memory/main.py
@@ -607,6 +607,7 @@ class Memory(MemoryBase):
run_id: Optional[str] = None,
limit: int = 100,
filters: Optional[Dict[str, Any]] = None,
+ threshold: Optional[float] = None,
):
"""
Searches for memories based on a query
@@ -617,6 +618,7 @@ class Memory(MemoryBase):
run_id (str, optional): ID of the run to search for. Defaults to None.
limit (int, optional): Limit the number of results. Defaults to 100.
filters (dict, optional): Filters to apply to the search. Defaults to None..
+ threshold (float, optional): Minimum score for a memory to be included in the results. Defaults to None.
Returns:
dict: A dictionary containing the search results, typically under a "results" key,
@@ -633,11 +635,11 @@ class Memory(MemoryBase):
capture_event(
"mem0.search",
self,
- {"limit": limit, "version": self.api_version, "keys": list(effective_filters.keys()), "sync_type": "sync"},
+ {"limit": limit, "version": self.api_version, "keys": list(effective_filters.keys()), "sync_type": "sync", "threshold": threshold},
)
with concurrent.futures.ThreadPoolExecutor() as executor:
- future_memories = executor.submit(self._search_vector_store, query, effective_filters, limit)
+ future_memories = executor.submit(self._search_vector_store, query, effective_filters, limit, threshold)
future_graph_entities = (
executor.submit(self.graph.search, query, effective_filters, limit) if self.enable_graph else None
)
@@ -664,7 +666,7 @@ class Memory(MemoryBase):
else:
return {"results": original_memories}
- def _search_vector_store(self, query, filters, limit):
+ def _search_vector_store(self, query, filters, limit, threshold: Optional[float] = None):
embeddings = self.embedding_model.embed(query, "search")
memories = self.vector_store.search(query=query, vectors=embeddings, limit=limit, filters=filters)
@@ -696,8 +698,9 @@ class Memory(MemoryBase):
additional_metadata = {k: v for k, v in mem.payload.items() if k not in core_and_promoted_keys}
if additional_metadata:
memory_item_dict["metadata"] = additional_metadata
-
- original_memories.append(memory_item_dict)
+
+ if threshold is None or mem.score >= threshold:
+ original_memories.append(memory_item_dict)
return original_memories
@@ -1442,6 +1445,7 @@ class AsyncMemory(MemoryBase):
run_id: Optional[str] = None,
limit: int = 100,
filters: Optional[Dict[str, Any]] = None,
+ threshold: Optional[float] = None,
):
"""
Searches for memories based on a query
@@ -1451,7 +1455,8 @@ class AsyncMemory(MemoryBase):
agent_id (str, optional): ID of the agent to search for. Defaults to None.
run_id (str, optional): ID of the run to search for. Defaults to None.
limit (int, optional): Limit the number of results. Defaults to 100.
- filters (dict, optional): Filters to apply to the search. Defaults to None..
+ filters (dict, optional): Filters to apply to the search. Defaults to None.
+ threshold (float, optional): Minimum score for a memory to be included in the results. Defaults to None.
Returns:
dict: A dictionary containing the search results, typically under a "results" key,
@@ -1469,10 +1474,10 @@ class AsyncMemory(MemoryBase):
capture_event(
"mem0.search",
self,
- {"limit": limit, "version": self.api_version, "keys": list(effective_filters.keys()), "sync_type": "async"},
+ {"limit": limit, "version": self.api_version, "keys": list(effective_filters.keys()), "sync_type": "async", "threshold": threshold},
)
- vector_store_task = asyncio.create_task(self._search_vector_store(query, effective_filters, limit))
+ vector_store_task = asyncio.create_task(self._search_vector_store(query, effective_filters, limit, threshold))
graph_task = None
if self.enable_graph:
@@ -1502,7 +1507,7 @@ class AsyncMemory(MemoryBase):
else:
return {"results": original_memories}
- async def _search_vector_store(self, query, filters, limit):
+ async def _search_vector_store(self, query, filters, limit, threshold: Optional[float] = None):
embeddings = await asyncio.to_thread(self.embedding_model.embed, query, "search")
memories = await asyncio.to_thread(
self.vector_store.search, query=query, vectors=embeddings, limit=limit, filters=filters
@@ -1536,8 +1541,9 @@ class AsyncMemory(MemoryBase):
additional_metadata = {k: v for k, v in mem.payload.items() if k not in core_and_promoted_keys}
if additional_metadata:
memory_item_dict["metadata"] = additional_metadata
-
- original_memories.append(memory_item_dict)
+
+ if threshold is None or mem.score >= threshold:
+ original_memories.append(memory_item_dict)
return original_memories
diff --git a/pyproject.toml b/pyproject.toml
index 62251f0f9..b52f40d6a 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
[project]
name = "mem0ai"
-version = "0.1.104"
+version = "0.1.105"
description = "Long-term memory for AI Agents"
authors = [
{ name = "Mem0", email = "[email protected]" }