Update app.py
Browse files
app.py
CHANGED
@@ -6,7 +6,7 @@ import tempfile
|
|
6 |
from typing import TypedDict, Annotated, Sequence
|
7 |
from langchain_openai import ChatOpenAI
|
8 |
from langchain_core.tools import tool
|
9 |
-
from langchain_core.messages import HumanMessage,
|
10 |
from langchain_core.utils.function_calling import convert_to_openai_tool
|
11 |
from langgraph.graph import StateGraph, END
|
12 |
|
@@ -36,19 +36,10 @@ def invoke_model(state):
|
|
36 |
"""
|
37 |
Invoke the model and handle tool invocation logic.
|
38 |
"""
|
39 |
-
|
40 |
-
question = state['messages'][-1].content if isinstance(state['messages'][-1], HumanMessage) else state['messages'][-1]
|
41 |
response = model_with_tools.invoke(question)
|
42 |
|
43 |
-
#
|
44 |
-
if isinstance(response, str):
|
45 |
-
return {"messages": [AIMessage(content=response)]}
|
46 |
-
|
47 |
-
# If no tool calls exist
|
48 |
-
if not response.additional_kwargs.get("tool_calls", []):
|
49 |
-
return {"messages": [AIMessage(content=response.content)]}
|
50 |
-
|
51 |
-
# If tool calls are present, return the full response
|
52 |
return {"messages": [response]}
|
53 |
|
54 |
graph.add_node("agent", invoke_model)
|
@@ -60,30 +51,33 @@ def invoke_tool(state):
|
|
60 |
"""
|
61 |
tool_calls = state['messages'][-1].additional_kwargs.get("tool_calls", [])
|
62 |
for tool_call in tool_calls:
|
63 |
-
if tool_call
|
64 |
-
arguments = json.loads(tool_call
|
65 |
result = multiply.invoke(arguments)
|
66 |
-
return {"messages": [
|
67 |
-
return {"messages": [
|
68 |
|
69 |
graph.add_node("tool", invoke_tool)
|
70 |
-
graph.add_edge("tool", END)
|
71 |
-
graph.set_entry_point("agent")
|
72 |
|
73 |
-
# Router
|
74 |
def router(state):
|
75 |
-
"""
|
76 |
-
Decide whether to invoke a tool or return the response.
|
77 |
-
"""
|
78 |
tool_calls = state['messages'][-1].additional_kwargs.get("tool_calls", [])
|
79 |
-
return "
|
80 |
|
81 |
-
graph.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
82 |
app_graph = graph.compile()
|
83 |
|
84 |
-
# Save graph visualization
|
85 |
with tempfile.NamedTemporaryFile(suffix=".png", delete=False) as tmpfile:
|
86 |
-
graph_viz = app_graph.get_graph(xray=True)
|
87 |
tmpfile.write(graph_viz.draw_mermaid_png())
|
88 |
graph_image_path = tmpfile.name
|
89 |
|
@@ -93,6 +87,7 @@ st.title("Simple Tool Calling Demo")
|
|
93 |
# Display the workflow graph
|
94 |
st.image(graph_image_path, caption="Workflow Visualization")
|
95 |
|
|
|
96 |
tab1, tab2 = st.tabs(["Try Multiplication", "Ask General Queries"])
|
97 |
|
98 |
# Multiplication Tool Tab
|
@@ -121,7 +116,6 @@ with tab2:
|
|
121 |
if st.button("Submit"):
|
122 |
if user_input:
|
123 |
try:
|
124 |
-
# Pass the user input as a HumanMessage
|
125 |
result = app_graph.invoke({"messages": [HumanMessage(content=user_input)]})
|
126 |
st.write("Response:")
|
127 |
st.success(result['messages'][-1].content)
|
@@ -133,3 +127,4 @@ with tab2:
|
|
133 |
# Sidebar for References
|
134 |
st.sidebar.title("References")
|
135 |
st.sidebar.markdown("1. [LangGraph Tool Calling](https://github.com/aritrasen87/LLM_RAG_Model_Deployment/blob/main/LangGraph_02_ToolCalling.ipynb)")
|
|
|
|
6 |
from typing import TypedDict, Annotated, Sequence
|
7 |
from langchain_openai import ChatOpenAI
|
8 |
from langchain_core.tools import tool
|
9 |
+
from langchain_core.messages import HumanMessage, ToolMessage
|
10 |
from langchain_core.utils.function_calling import convert_to_openai_tool
|
11 |
from langgraph.graph import StateGraph, END
|
12 |
|
|
|
36 |
"""
|
37 |
Invoke the model and handle tool invocation logic.
|
38 |
"""
|
39 |
+
question = state['messages'][-1].content
|
|
|
40 |
response = model_with_tools.invoke(question)
|
41 |
|
42 |
+
# Return the model's response with tool_calls, if any
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
43 |
return {"messages": [response]}
|
44 |
|
45 |
graph.add_node("agent", invoke_model)
|
|
|
51 |
"""
|
52 |
tool_calls = state['messages'][-1].additional_kwargs.get("tool_calls", [])
|
53 |
for tool_call in tool_calls:
|
54 |
+
if tool_call["function"]["name"] == "multiply":
|
55 |
+
arguments = json.loads(tool_call["function"]["arguments"])
|
56 |
result = multiply.invoke(arguments)
|
57 |
+
return {"messages": [ToolMessage(content=str(result))]}
|
58 |
+
return {"messages": [ToolMessage(content="No valid tool input provided.")]}
|
59 |
|
60 |
graph.add_node("tool", invoke_tool)
|
|
|
|
|
61 |
|
62 |
+
# Router Node: Manual Addition
|
63 |
def router(state):
|
|
|
|
|
|
|
64 |
tool_calls = state['messages'][-1].additional_kwargs.get("tool_calls", [])
|
65 |
+
return {"messages": [ToolMessage(content="Routing...")]} # Dummy message for router node
|
66 |
|
67 |
+
graph.add_node("router", router)
|
68 |
+
|
69 |
+
# Add explicit edges from agent to router, and router to tool/END
|
70 |
+
graph.add_edge("agent", "router")
|
71 |
+
graph.add_conditional_edges("router", lambda state: "tool" if state['messages'][-1].additional_kwargs.get("tool_calls") else END, {"tool": "tool", END: END})
|
72 |
+
graph.add_edge("tool", END)
|
73 |
+
|
74 |
+
# Compile the graph
|
75 |
+
graph.set_entry_point("agent")
|
76 |
app_graph = graph.compile()
|
77 |
|
78 |
+
# Save graph visualization with router explicitly included
|
79 |
with tempfile.NamedTemporaryFile(suffix=".png", delete=False) as tmpfile:
|
80 |
+
graph_viz = app_graph.get_graph(xray=True) # Ensures detailed visualization
|
81 |
tmpfile.write(graph_viz.draw_mermaid_png())
|
82 |
graph_image_path = tmpfile.name
|
83 |
|
|
|
87 |
# Display the workflow graph
|
88 |
st.image(graph_image_path, caption="Workflow Visualization")
|
89 |
|
90 |
+
# Tabbed Interface
|
91 |
tab1, tab2 = st.tabs(["Try Multiplication", "Ask General Queries"])
|
92 |
|
93 |
# Multiplication Tool Tab
|
|
|
116 |
if st.button("Submit"):
|
117 |
if user_input:
|
118 |
try:
|
|
|
119 |
result = app_graph.invoke({"messages": [HumanMessage(content=user_input)]})
|
120 |
st.write("Response:")
|
121 |
st.success(result['messages'][-1].content)
|
|
|
127 |
# Sidebar for References
|
128 |
st.sidebar.title("References")
|
129 |
st.sidebar.markdown("1. [LangGraph Tool Calling](https://github.com/aritrasen87/LLM_RAG_Model_Deployment/blob/main/LangGraph_02_ToolCalling.ipynb)")
|
130 |
+
|