github/langchain
1
0
Fork 0
mirror of https://github.com/hwchase17/langchain.git synced 2025-06-20 05:43:55 +00:00
Code Issues Packages Projects Releases Wiki Activity

fix(persona): correct typing, formatting, and improve docstrings

Browse Source
This commit is contained in:
Mikolaj Mikuliszyn 2025-04-27 18:00:14 +01:00
parent d6a032d8e7
commit 77196f40ee
1 changed files with 61 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

68
libs/langchain/langchain/memory/persona.py
View File

@ -2,12 +2,37 @@ import re
from typing import Any, Callable, Optional
from pydantic import BaseModel, Field
from langchain_core.memory import BaseMemory
class EnrichedMessage(BaseModel):
"""
A message enriched with persona traits and metadata.
"""A message enriched with persona traits and metadata.
This class extends a basic message with additional information about the
sender's personality traits and any relevant metadata. The traits are
automatically detected from the message content.
Examples:
Basic usage:
>>> message = EnrichedMessage(
... id="msg-001",
... content="I'm really excited about this project!",
... traits=["enthusiastic"], # Automatically detected
... metadata={"traits_count": {"enthusiastic": 1}}
... )
>>> message.traits
['enthusiastic']
With multiple traits:
>>> message = EnrichedMessage(
... id="msg-002",
... content="I'm not sure about this... maybe we should reconsider?",
... traits=["hesitant", "cautious"], # Multiple detected traits
... metadata={"traits_count": {"hesitant": 1, "cautious": 1}}
... )
>>> message.traits
['hesitant', 'cautious']
"""
id: str
@ -17,13 +42,42 @@ class EnrichedMessage(BaseModel):
class PersonaMemory(BaseMemory):
"""
Memory module that dynamically tracks emotional and behavioral traits
"""Memory module that dynamically tracks emotional and behavioral traits
exhibited by an agent over the course of a conversation.
Traits are automatically detected from output messages and
accumulated across interactions. Recent messages and detected traits
can be retrieved to enrich future prompts or modify agent behavior.
Traits are automatically detected from output messages and accumulated across
interactions. Recent messages and detected traits can be retrieved to enrich
future prompts or modify agent behavior.
The memory maintains a sliding window of the k most recent messages
(default k=10) and tracks all unique traits detected across these messages.
Examples:
Basic usage:
>>> memory = PersonaMemory()
>>> memory.save_context(
... {"input": "How are you?"},
... {"output": "I'm doing great! This is fantastic!"}
... )
>>> memory.load_memory_variables({})
{'persona': {'traits': ['enthusiastic']}}
Including recent messages:
>>> memory = PersonaMemory()
>>> memory.save_context(
... {"input": "What do you think?"},
... {"output": "I'm not sure... perhaps we should reconsider?"}
... )
>>> memory.load_memory_variables({}, include_messages=True)
{'persona': {
'traits': ['hesitant', 'cautious'],
'recent_messages': [{
'id': '1',
'content': "I'm not sure... perhaps we should reconsider?",
'traits': ['hesitant', 'cautious'],
'metadata': {'traits_count': {'hesitant': 1, 'cautious': 1}}
}]
}}
"""
memory_key: str = "persona"