langgraph/how-tos/memory/semantic-search/

[giscus[bot]]

langgraph/how-tos/memory/semantic-search/

Build language agents as graphs

https://langchain-ai.github.io/langgraph/how-tos/memory/semantic-search/?h=langgraph+store+memory

[Michailbul]

Where can i find documentation for langraph.store.memory? to use third party store like pinecone or qdrant?
thanks!

[sriakhil25]

There seems an error while running this part of the code - Its making call to upset_memory where the error exists.

for message, metadata in agent.stream(
input={"messages": [{"role": "user", "content": "I'm hungry"}]},
stream_mode="messages",
):
print(message.content, end="")

Error: 1 validation error for upsert_memory
store
Field required [type=missing, input_value={'content': "I'm hungry"}, input_type=dict]
For further information visit https://errors.pydantic.dev/2.9/v/missing
Please fix your mistakes. It seems there was an error in the code related to the upsert_memory function. The error message suggests that the 'content' field is required but it is missing from the input data. You should make sure that the content field is included when calling the upsert_memory function. Here's an example of how you can call the function with the correct format:

upsert_memory(content="I'm hungry")