-
Notifications
You must be signed in to change notification settings - Fork 2.6k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Some information are not stored #2443
Comments
Does this happen only in a specific dialog piece, or u make sure it happens every time? |
So far it happens only in a specific dialog piece. I've tried deleting all the memory just to make sure that it's not redundant. I'm guessing that somehow the AI see it as irrelevant for some reason. I've also tried changing the LLM models but to no avail. |
I have experienced a similar problem, there is an option provided for adding a prompt to the add() function, I thought that could help If I told it what I want but it does not work. |
I have the same problems. |
Hey @AugustYing, can you please share the code snippet and what text you are trying to pass? That will help us to reproduce the issue and debug further. |
Sure. I encounted this bug while running ai agent, so the code snippet may hard to provide. Because the messages is generated by llm. The two messages are generated in the same task, one before and one after, and the former message is successfully updated, but the latter one is not. The structure of these two messages is obviously very similar, I don't understand why one can update the other can't. |
Hi, I'm using mem0 for my langgraph. I noticed that sometimes mem0 does not save information even though it's not redundant, and it's very necessary. Are there any way to fix this?
This is my code for mem0
from mem0 import Memory
from dotenv import load_dotenv
import os
load_dotenv()
config = {
"llm": {
"provider": "openai",
"config": {
"model": "gpt-4o"
}
},
"vector_store": {
"provider": "supabase",
"config": {
"connection_string": os.getenv("DATABASE_URL"),
"collection_name": "memories"
}
}
}
client = Memory.from_config(config)
And this is for saving
try:
messages = json.loads(data)
I've also checked just in case the add function failed, but it did not. I assume that the AI deems the new information as either irrelevant or redundant. Which it isn't.
The text was updated successfully, but these errors were encountered: