How to persist or save InMemoryStore data between program runs in LangGraph?

[EvanZhou1999]

I'm using LangGraph's InMemoryStore to store data in my application. However, I noticed that whenever I restart my program and attempt to retrieve from in_memory_store with the same namespace and key, all previously stored data is lost. Here's my current code (from Langgraph Academy) to test this functionality:

import uuid
from langgraph.store.memory import InMemoryStore

in_memory_store = InMemoryStore()

# Namespace for the mempry to save
user_id = "1" 
namespace_for_memory = (user_id, "memories")

# save a memory to namespace as key and value
# key = str(uuid.uuid4())
key = "461ae15e-adec-44cc-97bf-5784afac913a"
print(key)
value = {"food_preference": "I like pizza"}

# Save the memory
in_memory_store.put(namespace_for_memory, key, value)

# We use search to retrieve objects from the store by namesapce
# this returns a list
memories = in_memory_store.search(namespace_for_memory)
type(memories)  # <-- list
# print metadata
print(memories[0].dict())
print()
# print key, values
print(memories[0].key, memories[0].value)
print()

# We can also get the memory by namespace and key
memory = in_memory_store.get(namespace_for_memory, key)
print(memory.dict())

I'm looking for ways to persist this data between program runs. And there seems to be no description nor API references that talks about how to persist the InMemoryStore (If I did miss a reference somewhere please kindly point me to the solution). What are the recommended approaches for handling this in LangGraph?

[hinthornw]

I and the docs recommend the Postgres store for actual persistence.

The in memory store is in memory and thus ephemeral - useful for dev and docs as a dependency-free option.