Improved Langchain Support!

New ZepMemory Class Improves Developer Ergonomics

We've made it easier to use Zep with Langchain. Released in Langchain v0.0.229 is a ZepMemory class. This simplifies using Zep with your chains, as before today Zep's ZepChatMessageHistory class had to be wrapped in an existing Langchain Memory.

# Set up Zep Memory
memory = ZepMemory(
    session_id=session_id,
    url=ZEP_API_URL,
    api_key=zep_api_key,
    memory_key="chat_history",
)

# Initialize the agent
llm = OpenAI(temperature=0, openai_api_key=openai_key)
agent_chain = initialize_agent(
    tools,
    llm,
    agent=AgentType.CONVERSATIONAL_REACT_DESCRIPTION,
    verbose=True,
    memory=memory,
)

Improved Access to Enriched Memory

Langchain's ZepMemory class now provides access to Zep's enriched memory attributes, including extracted entities, intents, token counts, timestamps, and UUIDs.

def print_messages(messages):
    for m in messages:
        print(m.type, ":\n", m.dict())


print(memory.chat_memory.zep_summary)
print("\n")
print_messages(memory.chat_memory.messages)

    The human inquires about Octavia Butler. The AI identifies her as an American science fiction author. The human then asks which books of hers were made into movies. The AI responds by mentioning the FX series Kindred, based on her novel of the same name. The human then asks about her contemporaries, and the AI lists Ursula K. Le Guin, Samuel R. Delany, and Joanna Russ.
    
    
    system :
     {'content': 'The human inquires about Octavia Butler. The AI identifies her as an American science fiction author. The human then asks which books of hers were made into movies. The AI responds by mentioning the FX series Kindred, based on her novel of the same name. The human then asks about her contemporaries, and the AI lists Ursula K. Le Guin, Samuel R. Delany, and Joanna Russ.', 'additional_kwargs': {}}
    human :
     {'content': 'What awards did she win?', 'additional_kwargs': {'uuid': '6b733f0b-6778-49ae-b3ec-4e077c039f31', 'created_at': '2023-07-09T19:23:16.611232Z', 'token_count': 8, 'metadata': {'system': {'entities': [], 'intent': 'The subject is inquiring about the awards that someone, whose identity is not specified, has won.'}}}, 'example': False}
    ai :
     {'content': 'Octavia Butler won the Hugo Award, the Nebula Award, and the MacArthur Fellowship.', 'additional_kwargs': {'uuid': '2f6d80c6-3c08-4fd4-8d4e-7bbee341ac90', 'created_at': '2023-07-09T19:23:16.618947Z', 'token_count': 21, 'metadata': {'system': {'entities': [{'Label': 'PERSON', 'Matches': [{'End': 14, 'Start': 0, 'Text': 'Octavia Butler'}], 'Name': 'Octavia Butler'}, {'Label': 'WORK_OF_ART', 'Matches': [{'End': 33, 'Start': 19, 'Text': 'the Hugo Award'}], 'Name': 'the Hugo Award'}, {'Label': 'EVENT', 'Matches': [{'End': 81, 'Start': 57, 'Text': 'the MacArthur Fellowship'}], 'Name': 'the MacArthur Fellowship'}], 'intent': 'The subject is stating that Octavia Butler received the Hugo Award, the Nebula Award, and the MacArthur Fellowship.'}}}, 'example': False}
<snip />

Custom Chains Can Persist and Search Over Custom Metadata

Custom chain developers can now add their own metadata to chat messages when they're persisted to Zep. This metadata is then searchable via the ZepRetriever or other Zep search functionality.

The ZepMemory class's save_context method now supports passing in a metadata argument. Your custom Chain class must override the base Chain's `prep_outputs` method to include the metadata in the call to self.memory.save_context.

Other Improvements

The chat history summary is now included in the ZepMemory chat history as a SystemMessage, whereas before, it was a HumanMessage. This change reduces ambiguity as to the origin of the summary.