Spaces:
Sleeping
Sleeping
add system message, ruff format, change pro to flash
Browse files
agent.py
CHANGED
|
@@ -1,17 +1,19 @@
|
|
| 1 |
import os
|
| 2 |
from dotenv import load_dotenv
|
| 3 |
|
| 4 |
-
from langchain_core.messages import HumanMessage
|
| 5 |
from langchain_core.tools import tool
|
| 6 |
from langchain_google_genai import ChatGoogleGenerativeAI
|
| 7 |
from langchain_community.tools.tavily_search import TavilySearchResults
|
| 8 |
from langchain_community.document_loaders import WikipediaLoader
|
| 9 |
|
| 10 |
-
from langgraph.graph import StateGraph, START,
|
| 11 |
from langgraph.prebuilt import ToolNode, tools_condition
|
| 12 |
|
| 13 |
load_dotenv()
|
| 14 |
|
|
|
|
|
|
|
| 15 |
|
| 16 |
@tool
|
| 17 |
def add(a: float, b: float) -> float:
|
|
@@ -80,7 +82,7 @@ def power(a: float, b: float) -> float:
|
|
| 80 |
a: Base number
|
| 81 |
b: Exponent
|
| 82 |
"""
|
| 83 |
-
return a
|
| 84 |
|
| 85 |
|
| 86 |
@tool
|
|
@@ -92,7 +94,7 @@ def square_root(a: float) -> float:
|
|
| 92 |
"""
|
| 93 |
if a < 0:
|
| 94 |
return "Error: Cannot calculate square root of negative number"
|
| 95 |
-
return a
|
| 96 |
|
| 97 |
|
| 98 |
@tool
|
|
@@ -111,9 +113,9 @@ def web_search(query: str) -> str:
|
|
| 111 |
|
| 112 |
formatted_results = []
|
| 113 |
for i, result in enumerate(results, 1):
|
| 114 |
-
title = result.get(
|
| 115 |
-
content = result.get(
|
| 116 |
-
url = result.get(
|
| 117 |
formatted_results.append(f"{i}. {title}\n{content}\nSource: {url}")
|
| 118 |
|
| 119 |
return "\n\n ==== \n\n".join(formatted_results)
|
|
@@ -137,7 +139,7 @@ def wikipedia_search(query: str) -> str:
|
|
| 137 |
|
| 138 |
formatted_docs = []
|
| 139 |
for i, doc in enumerate(docs, 1):
|
| 140 |
-
title = doc.metadata.get(
|
| 141 |
content = doc.page_content
|
| 142 |
formatted_docs.append(f"{i}. {title}\n{content}")
|
| 143 |
|
|
@@ -147,32 +149,41 @@ def wikipedia_search(query: str) -> str:
|
|
| 147 |
|
| 148 |
|
| 149 |
tools = [
|
| 150 |
-
add,
|
| 151 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 152 |
]
|
| 153 |
|
| 154 |
|
| 155 |
def get_llm():
|
| 156 |
"""Initialize the llm"""
|
| 157 |
return ChatGoogleGenerativeAI(
|
| 158 |
-
model="gemini-2.5-
|
| 159 |
-
temperature=0,
|
| 160 |
-
api_key=os.getenv("GEMINI_API_KEY")
|
| 161 |
)
|
| 162 |
|
| 163 |
|
| 164 |
def call_model(state: MessagesState):
|
| 165 |
"""Call the LLM with the current state.
|
| 166 |
-
|
| 167 |
Args:
|
| 168 |
state: Current state containing messages
|
| 169 |
"""
|
| 170 |
llm = get_llm()
|
| 171 |
llm_with_tools = llm.bind_tools(tools)
|
| 172 |
-
|
| 173 |
-
messages = state[
|
|
|
|
|
|
|
|
|
|
|
|
|
| 174 |
response = llm_with_tools.invoke(messages)
|
| 175 |
-
|
| 176 |
return {"messages": [response]}
|
| 177 |
|
| 178 |
|
|
|
|
| 1 |
import os
|
| 2 |
from dotenv import load_dotenv
|
| 3 |
|
| 4 |
+
from langchain_core.messages import HumanMessage, SystemMessage
|
| 5 |
from langchain_core.tools import tool
|
| 6 |
from langchain_google_genai import ChatGoogleGenerativeAI
|
| 7 |
from langchain_community.tools.tavily_search import TavilySearchResults
|
| 8 |
from langchain_community.document_loaders import WikipediaLoader
|
| 9 |
|
| 10 |
+
from langgraph.graph import StateGraph, START, MessagesState
|
| 11 |
from langgraph.prebuilt import ToolNode, tools_condition
|
| 12 |
|
| 13 |
load_dotenv()
|
| 14 |
|
| 15 |
+
SYSTEM_PROMPT = """You are a general AI assistant. I will ask you a question. Report your thoughts, and output only your final answer, no prefixes, suffixes, or extra text. YOUR FINAL ANSWER should be a number OR as few words as possible OR a comma separated list of numbers and/or strings. If you are asked for a number, don't use comma to write your number neither use units such as $ or percent sign unless specified otherwise. If you are asked for a string, don't use articles, neither abbreviations (e.g. for cities), and write the digits in plain text unless specified otherwise. If you are asked for a comma separated list, apply the above rules depending of whether the element to be put in the list is a number or a string."""
|
| 16 |
+
|
| 17 |
|
| 18 |
@tool
|
| 19 |
def add(a: float, b: float) -> float:
|
|
|
|
| 82 |
a: Base number
|
| 83 |
b: Exponent
|
| 84 |
"""
|
| 85 |
+
return a**b
|
| 86 |
|
| 87 |
|
| 88 |
@tool
|
|
|
|
| 94 |
"""
|
| 95 |
if a < 0:
|
| 96 |
return "Error: Cannot calculate square root of negative number"
|
| 97 |
+
return a**0.5
|
| 98 |
|
| 99 |
|
| 100 |
@tool
|
|
|
|
| 113 |
|
| 114 |
formatted_results = []
|
| 115 |
for i, result in enumerate(results, 1):
|
| 116 |
+
title = result.get("title", "No title")
|
| 117 |
+
content = result.get("content", "No content")
|
| 118 |
+
url = result.get("url", "No URL")
|
| 119 |
formatted_results.append(f"{i}. {title}\n{content}\nSource: {url}")
|
| 120 |
|
| 121 |
return "\n\n ==== \n\n".join(formatted_results)
|
|
|
|
| 139 |
|
| 140 |
formatted_docs = []
|
| 141 |
for i, doc in enumerate(docs, 1):
|
| 142 |
+
title = doc.metadata.get("title", "No title")
|
| 143 |
content = doc.page_content
|
| 144 |
formatted_docs.append(f"{i}. {title}\n{content}")
|
| 145 |
|
|
|
|
| 149 |
|
| 150 |
|
| 151 |
tools = [
|
| 152 |
+
add,
|
| 153 |
+
subtract,
|
| 154 |
+
multiply,
|
| 155 |
+
divide,
|
| 156 |
+
modulo,
|
| 157 |
+
power,
|
| 158 |
+
square_root,
|
| 159 |
+
web_search,
|
| 160 |
+
wikipedia_search,
|
| 161 |
]
|
| 162 |
|
| 163 |
|
| 164 |
def get_llm():
|
| 165 |
"""Initialize the llm"""
|
| 166 |
return ChatGoogleGenerativeAI(
|
| 167 |
+
model="gemini-2.5-flash", temperature=0, api_key=os.getenv("GEMINI_API_KEY")
|
|
|
|
|
|
|
| 168 |
)
|
| 169 |
|
| 170 |
|
| 171 |
def call_model(state: MessagesState):
|
| 172 |
"""Call the LLM with the current state.
|
| 173 |
+
|
| 174 |
Args:
|
| 175 |
state: Current state containing messages
|
| 176 |
"""
|
| 177 |
llm = get_llm()
|
| 178 |
llm_with_tools = llm.bind_tools(tools)
|
| 179 |
+
|
| 180 |
+
messages = state["messages"]
|
| 181 |
+
|
| 182 |
+
if not messages or not isinstance(messages[0], SystemMessage):
|
| 183 |
+
messages = [SystemMessage(content=SYSTEM_PROMPT)] + messages
|
| 184 |
+
|
| 185 |
response = llm_with_tools.invoke(messages)
|
| 186 |
+
|
| 187 |
return {"messages": [response]}
|
| 188 |
|
| 189 |
|