Implement redis memory backend.

README.md

@@ -149,6 +149,27 @@ are loaded for the agent at any given time.
 2. Choose the `Starter` plan to avoid being charged.
 3. Find your API key and region under the default project in the left sidebar.
 
+
+## Redis Setup
+
+Install docker desktop.
+
+Run:
+```
+docker run -d --name redis-stack-server -p 6379:6379 redis/redis-stack-server:latest
+```
+
+Set the following environment variables:
+```
+MEMORY_BACKEND=redis
+REDIS_HOST=localhost
+REDIS_PORT=6379
+REDIS_PASSWORD=
+```
+
+Note that this is not intended to be run facing the internet and is not secure, do not expose redis to the internet without a password or at all really.
+
+
 ### Setting up environment variables
    For Windows Users:
 ```

requirements.txt

@@ -12,3 +12,4 @@ docker
 duckduckgo-search
 google-api-python-client #(https://developers.google.com/custom-search/v1/overview)
 pinecone-client==2.2.1
+redis

scripts/commands.py

@@ -1,6 +1,7 @@
 import browse
 import json
 from memory.pinecone import PineconeMemory
+from memory.redismem import RedisMemory
 import datetime
 import agent_manager as agents
 import speak
@@ -52,7 +53,10 @@ def get_command(response):
 
 
 def execute_command(command_name, arguments):
+    if cfg.memory_backend == "pinecone":
         memory = PineconeMemory(cfg=cfg)
+    else:
+        memory = RedisMemory(cfg=cfg)
     try:
         if command_name == "google":
             

scripts/config.py

@@ -61,7 +61,12 @@ class Config(metaclass=Singleton):
         # User agent headers to use when browsing web
         # Some websites might just completely deny request with an error code if no user agent was found.
         self.user_agent_header = {"User-Agent":"Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_4) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/83.0.4103.97 Safari/537.36"}
+        self.redis_host = os.getenv("REDIS_HOST")
+        self.redis_port = os.getenv("REDIS_PORT")
+        self.redis_password = os.getenv("REDIS_PASSWORD")
+        # Note that indexes must be created on db 0 in redis, this is not configureable.
         
+        self.memory_backend = os.getenv("MEMORY_BACKEND", 'pinecone')
         # Initialize the OpenAI API client
         openai.api_key = self.openai_api_key
 

scripts/main.py

@@ -2,6 +2,7 @@ import json
 import random
 import commands as cmd
 from memory.pinecone import PineconeMemory
+from memory.redismem import RedisMemory
 import data
 import chat
 from colorama import Fore, Style
@@ -283,8 +284,11 @@ user_input = "Determine which next command to use, and respond using the format
 
 # Initialize memory and make sure it is empty.
 # this is particularly important for indexing and referencing pinecone memory
+if cfg.memory_backend == "pinecone":
     memory = PineconeMemory(cfg)
     memory.clear()
+else:
+    memory = RedisMemory(cfg)
 
 print('Using memory of type: ' + memory.__class__.__name__)
 

scripts/memory/base.py

@@ -1,3 +1,4 @@
+"""Base class for memory providers."""
 import abc
 from config import AbstractSingleton
 import openai

scripts/memory/redismem.py

@@ -0,0 +1,135 @@
+"""Redis memory provider."""
+from typing import Any, List, Optional
+import redis
+from redis.commands.search.field import VectorField, TextField
+from redis.commands.search.query import Query
+from redis.commands.search.indexDefinition import IndexDefinition, IndexType
+import traceback
+import numpy as np
+
+from memory.base import MemoryProviderSingleton, get_ada_embedding
+
+
+SCHEMA = [
+    TextField("data"),
+    VectorField(
+        "embedding",
+        "HNSW",
+        {
+            "TYPE": "FLOAT32",
+            "DIM": 1536,
+            "DISTANCE_METRIC": "COSINE"
+        }
+    ),
+]
+
+
+class RedisMemory(MemoryProviderSingleton):
+    def __init__(self, cfg):
+        """
+        Initializes the Redis memory provider.
+
+        Args:
+            cfg: The config object.
+
+        Returns: None
+        """
+        redis_host = cfg.redis_host
+        redis_port = cfg.redis_port
+        redis_password = cfg.redis_password
+        self.dimension = 1536
+        self.redis = redis.Redis(
+            host=redis_host,
+            port=redis_port,
+            password=redis_password,
+            db=0  # Cannot be changed
+        )
+        self.redis.flushall()
+        try:
+            self.redis.ft("gpt").create_index(
+                fields=SCHEMA,
+                definition=IndexDefinition(
+                    prefix=["gpt:"],
+                    index_type=IndexType.HASH
+                    )
+                )
+        except Exception as e:
+            print("Error creating Redis search index: ", e)
+        self.vec_num = 0
+
+    def add(self, data: str) -> str:
+        """
+        Adds a data point to the memory.
+
+        Args:
+            data: The data to add.
+
+        Returns: Message indicating that the data has been added.
+        """
+        vector = get_ada_embedding(data)
+        vector = np.array(vector).astype(np.float32).tobytes()
+        data_dict = {
+            b"data": data,
+            "embedding": vector
+        }
+        self.redis.hset(f"gpt:{self.vec_num}", mapping=data_dict)
+        _text = f"Inserting data into memory at index: {self.vec_num}:\n"\
+            f"data: {data}"
+        self.vec_num += 1
+        return _text
+
+    def get(self, data: str) -> Optional[List[Any]]:
+        """
+        Gets the data from the memory that is most relevant to the given data.
+
+        Args:
+            data: The data to compare to.
+
+        Returns: The most relevant data.
+        """
+        return self.get_relevant(data, 1)
+
+    def clear(self) -> str:
+        """
+        Clears the redis server.
+
+        Returns: A message indicating that the memory has been cleared.
+        """
+        self.redis.flushall()
+        return "Obliviated"
+
+    def get_relevant(
+        self,
+        data: str,
+        num_relevant: int = 5
+    ) -> Optional[List[Any]]:
+        """
+        Returns all the data in the memory that is relevant to the given data.
+        Args:
+            data: The data to compare to.
+            num_relevant: The number of relevant data to return.
+
+        Returns: A list of the most relevant data.
+        """
+        query_embedding = get_ada_embedding(data)
+        base_query = f"*=>[KNN {num_relevant} @embedding $vector AS vector_score]"
+        query = Query(base_query).return_fields(
+            "data",
+            "vector_score"
+        ).sort_by("vector_score").dialect(2)
+        query_vector = np.array(query_embedding).astype(np.float32).tobytes()
+
+        try:
+            results = self.redis.ft("gpt").search(
+                query, query_params={"vector": query_vector}
+            )
+        except Exception as e:
+            print("Error calling Redis search: ", e)
+            return None
+        return list(results.docs)
+
+    def get_stats(self):
+        """
+        Returns: The stats of the memory index.
+        """
+        return self.redis.ft("mem").info()