Memokit

Memokit is a lightweight AI memory toolkit that helps AI applications manage medium-term memory.

Installation

Install from pypi:

pip install memokit

Or clone the repository:

git clone https://github.com/AmeNetwork/memokit.git
cd memokit
pip install -r requirements.txt

Usage

Memokit Space:
Custom memory spaces, different spaces are used to store different memories, and they will automatically classify and store your memories.

from memokit import Space 
space=Space()
# create space for sport 
space.create_space(uid="alice", space_name="sport", space_description="sport space")
# add memory to sport space 
space.add_memory(uid="alice", memory='I like playing football')

Full Example Here: space_example.py

Memokit Pool:
Memokit Pool is a shared memory pool that allows different users and agents to share the same pool, enabling memory sharing and management.

from memokit.pool import MemoryPool
pool = MemoryPool()

# create a memory pool for alice and bob
pool_id = pool.create_pool(["alice", "bob"])

# add memory to pool
pool.add_memory(
    pool_id=pool_id,
    memory="we will hold a meeting at 10am on Wednesday next week",
    uid="alice"
)

Full Example Here: pool_example.py