Update mem0 cookbook#288
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
| @@ -4,7 +4,7 @@ | |||
| "cell_type": "markdown", | |||
There was a problem hiding this comment.
A minor thing, but we haven't used the infer param in the previous example, and it might be confusing if someone is not familiar with the API.
Reply via ReviewNB
| @@ -4,7 +4,7 @@ | |||
| "cell_type": "markdown", | |||
There was a problem hiding this comment.
Does that mean our integration can ignore some memories added in the current agent run? In the experimental integration we had a possibility to enable async mode, but sync was the default.
Reply via ReviewNB
| @@ -4,7 +4,7 @@ | |||
| "cell_type": "markdown", | |||
There was a problem hiding this comment.
It's more a question about the integration itself, but shouldn't we have a method to remove the memories without asking our users to call the client directly? We do not ask them to pass MemoryClient when they create an instance of Mem0MemoryStore, so they are even not aware of it. From the UX perspective, it would make sense to have a richer interface. Perfectly even a tool, so the agent can delete stale memories itself.
Reply via ReviewNB
kacperlukawski
left a comment
There was a problem hiding this comment.
@sjrl It might be merged as it is, but I have some general questions about our approach. I played a bit with the integration, when it was still experimental, and would like to understand the motivation behind some of the changes.
Part of https://github.com/deepset-ai/haystack-private/issues/347