PR isolating the the embedding similarity based memory addition/retrieval/update function from LM-powered memory optimization #2
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
…d ones, and simplify conditional check for evo_label
…c to simply initialization and type checking.
…y. Refactor search to make use of the MemeoryNote.dump_model().
…d ones, and simplify conditional check for evo_label
wli51
force-pushed
the
decouple-mem-evo
branch
from
e95751e to
9f81f98
Compare
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Objective
The original implementation from the fork makes LM powered memory optimization mandatory upon addition and update. However, the current LM manager class has semi-broken prompting which occasionally causes non-compliance on the LM end, resulting in wrongly formatted outputs -> errors during that optimization process. This PR isolates the basic memory operations from the semi-broken LM optimization to get the memory class to minimal yet stable working status. Future PRs will replace the LM manager with DSPy backend (as the repo name suggests) and re-introduce the removed LM optimization functionality.
Adds:
agentic_memory/memory_note.py: helper data class from the original fork, promoted to its own module and refactored withpydantic.BaseModelfor organized type checking and factory methods.Refactors:
agentic_memory/memory_system.py: removed all LM components, applied some fixes to make the add/search/update work.tests/test_memory_system.py: moved from unittest to pytest and rewrote tests to test only the core memory functionality