ai-mirow-projects
/
swarms
mirror of https://github.com/kyegomez/swarms
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

update discordwithweb, configure sourcery

Browse Source
pull/57/head
yikes 2 years ago
parent a0c2f6b7c7
commit d1e53a6b75
4 changed files with 8 additions and 210 deletions
Show Stats Download Patch File Download Diff File

1
.gitignore vendored
Unescape View File

@ -180,3 +180,4 @@ cython_debug/
# and can be added to the global gitignore or merged into this file. For a more nuclear # and can be added to the global gitignore or merged into this file. For a more nuclear
# option (not recommended) you can uncomment the following to ignore the entire idea folder. # option (not recommended) you can uncomment the following to ignore the entire idea folder.
#.idea/ #.idea/
.aider*

2
swarms/agents/.sourcery.yaml → .sourcery.yaml
Unescape View File

@ -7,7 +7,7 @@
# 📚 For a complete reference to this file, see the documentation at # 📚 For a complete reference to this file, see the documentation at
# https://docs.sourcery.ai/Configuration/Project-Settings/ # https://docs.sourcery.ai/Configuration/Project-Settings/
# This file was auto-generated by Sourcery on 2023-10-06 at 17:14. # This file was auto-generated by Sourcery on 2023-10-06 at 18:33.
version: '1' # The schema version of this config file version: '1' # The schema version of this config file

101
apps/discordwithweb.py
Unescape View File

@ -2,115 +2,25 @@ import discord
from discord.ext import commands from discord.ext import commands
import os import os
from swarms.agents import Worker from swarms.agents import Worker
from swarms.agents.memory import VectorStoreRetriever from swarms.agents.memory import VectorStoreRetriever
from swarms.tools.autogpt import WebpageQATool from swarms.tools.autogpt import WebpageQATool
"""
Custom tools for web search and memory retrieval.
This code snippet defines two custom tools, `WebSearchTool` and `MemoryTool`, which are subclasses of the `BaseTool` class. The `WebSearchTool` makes a request to a search engine API and extracts the text from the top result. The `MemoryTool` retrieves relevant documents from a vector store and extracts the text from the document.
Example Usage:
```python
web_search = WebSearchTool()
result = web_search.run("python programming")
print(result)
# Output: The text from the top search result for "python programming"
memory_tool = MemoryTool(retriever)
result = memory_tool.run("python programming")
print(result)
# Output: The text from the relevant document retrieved from the vector store for "python programming"
```
Inputs:
- `query` (str): The search query or document retrieval query.
Flow:
1. The `WebSearchTool` makes a request to the Bing search engine API with the provided query.
2. The API response is stored in the `response` variable.
3. The text from the top search result is extracted from the API response and stored in the `text` variable.
4. The `MemoryTool` retrieves relevant documents from the vector store using the provided query.
5. The relevant document is stored in the `relevant_doc` variable.
6. The text from the relevant document is extracted and stored in the `text` variable.
Outputs:
- `text` (str): The extracted text from the top search result or relevant document.
"""
# Custom tools
class WebSearchTool(BaseTool):
def run(self, query: str) -> str:
# Make request to search engine API
response = requests.get(
"https://api.bing.com/v7.0/search",
params={
"q": query,
"count": 1
},
headers={
"Ocp-Apim-Subscription-Key": "YOUR_API_KEY"
}
)
# Extract text from top result
top_result = response.json()["webPages"]["value"][0]
text = top_result["snippet"]
return text
class MemoryTool(BaseTool):
def __init__(self, retriever):
self.retriever = retriever
def run(self, query: str) -> str:
# Retrieve relevant document from vectorstore
docs = self.retriever.retrieve(query)
relevant_doc = docs[0]
# Extract text from document
text = relevant_doc.text
return text
"""
This code sets up a Discord bot that responds to user queries.
The code initializes the Discord bot, sets up the OpenAI API key, initializes the memory retrieval system, creates custom tools for web search and memory retrieval, creates an agent with the OpenAI language model, memory, and tools, and defines a command for the bot to handle user queries. When a user sends a query, the agent runs the query and sends the response back to the user.
Inputs:
- ctx (object): The context object representing the message context in Discord.
- input (str): The user query input.
Outputs:
- response (str): The response generated by the agent based on the user query.
"""
# Discord bot setup # Discord bot setup
intents = discord.Intents.default() intents = discord.Intents.default()
bot = commands.Bot(command_prefix='!', intents=intents) bot = commands.Bot(command_prefix='!', intents=intents)
# Worker setup
worker = Worker()
# Memory setup # Memory setup
vectorstore_client = VectorStoreClient() vectorstore_client = VectorStoreClient()
retriever = VectorStoreRetriever(vectorstore_client) retriever = VectorStoreRetriever(vectorstore_client)
# Tools setup # Tools setup
web_search = WebpageQATool() web_search = WebSearchTool(retriever)
memory = MemoryTool(retriever) memory_tool = MemoryTool(retriever)
tools = [web_search, memory] tools = [web_search, memory_tool]
# Create the agent # Create the agent
agent = Agent( agent = Worker(
name="DiscordAssistant", name="DiscordAssistant",
llm=worker, llm=worker,
memory=retriever, memory=retriever,
@ -119,8 +29,7 @@ agent = Agent(
@bot.command() @bot.command()
async def query(ctx, *, input): async def query(ctx, *, input):
response = web_search.run(input) response = agent.run(input)
await ctx.send(response) await ctx.send(response)
bot.run(os.getenv("DISCORD_BOT_TOKEN")) bot.run(os.getenv("DISCORD_BOT_TOKEN"))
"""

112
swarms/agents/discordwithweb.py
Unescape View File

@ -1,112 +0,0 @@
import discord
from discord.ext import commands
import os
import openai
import requests
from swarms.agents import Agent
from swarms.agents.memory import VectorStoreRetriever
from swarms.tools.base import BaseTool
"""
Custom tools for web search and memory retrieval.
This code snippet defines two custom tools, `WebSearchTool` and `MemoryTool`, which are subclasses of the `BaseTool` class. The `WebSearchTool` makes a request to a search engine API and extracts the text from the top result. The `MemoryTool` retrieves relevant documents from a vector store and extracts the text from the document.
Example Usage:
```python
web_search = WebSearchTool()
result = web_search.run("python programming")
print(result)
# Output: The text from the top search result for "python programming"
memory_tool = MemoryTool(retriever)
result = memory_tool.run("python programming")
print(result)
# Output: The text from the relevant document retrieved from the vector store for "python programming"
```
Inputs:
- `query` (str): The search query or document retrieval query.
Flow:
1. The `WebSearchTool` makes a request to the Bing search engine API with the provided query.
2. The API response is stored in the `response` variable.
3. The text from the top search result is extracted from the API response and stored in the `text` variable.
4. The `MemoryTool` retrieves relevant documents from the vector store using the provided query.
5. The relevant document is stored in the `relevant_doc` variable.
6. The text from the relevant document is extracted and stored in the `text` variable.
Outputs:
- `text` (str): The extracted text from the top search result or relevant document.
"""
# Custom tools
class WebSearchTool(BaseTool):
def run(self, query: str) -> str:
# Make request to search engine API
response = requests.get(
"https://api.bing.com/v7.0/search",
params={
"q": query,
"count": 1
},
headers={
"Ocp-Apim-Subscription-Key": "YOUR_API_KEY"
}
)
# Extract text from top result
top_result = response.json()["webPages"]["value"][0]
return top_result["snippet"]
class MemoryTool(BaseTool):
def __init__(self, retriever):
self.retriever = retriever
def run(self, query: str) -> str:
# Retrieve relevant document from vectorstore
docs = self.retriever.retrieve(query)
relevant_doc = docs[0]
# Extract text from document
text = relevant_doc.text
return text
# Discord bot setup
intents = discord.Intents.default()
bot = commands.Bot(command_prefix='!', intents=intents)
# OpenAI API setup
openai.api_key = os.getenv("OPENAI_API_KEY")
# Memory setup
vectorstore_client = VectorStoreClient()
retriever = VectorStoreRetriever(vectorstore_client)
# Tools setup
web_search = WebSearchTool()
memory = MemoryTool(retriever)
tools = [web_search, memory]
# Create the agent
agent = Agent(
name="DiscordAssistant",
llm=openai,
memory=retriever,
tools=tools
)
@bot.command()
async def query(ctx, *, input):
response = agent.run(input)
await ctx.send(response)
bot.run(os.getenv("DISCORD_BOT_TOKEN"))
Write Preview
Loading…
Cancel
Save