Spaces:
Sleeping
Sleeping
File size: 4,659 Bytes
4a8958c 2d7b77a 4a8958c 5d50c30 04b1d6c 5d50c30 04b1d6c 2d7b77a 4a8958c 2d7b77a 4a8958c 2d7b77a 4a8958c 2d7b77a 4a8958c fd59124 4a8958c fd59124 2d7b77a 4a8958c 2d7b77a 4a8958c 04b1d6c 4a8958c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 |
import asyncio
import streamlit as st
import tempfile
import sys
import os
sys.path.append(os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
from agents.pdf_agent import PDFAgent
from agents.weather_agent import WeatherAgent
# Ensure an event loop exists for async libraries (fix for Google Generative AI Embeddings)
try:
asyncio.get_event_loop()
except RuntimeError:
asyncio.set_event_loop(asyncio.new_event_loop())
st.set_page_config(page_title="LangGraph Agents Demo", layout="wide")
st.title("LangGraph Agents Demo")
tab1, tab2, tab3 = st.tabs(["PDF Agent", "Weather Agent", "Multi-Agent QA"])
with tab1:
st.header("PDF Agent")
uploaded_pdf = st.file_uploader("Upload a PDF", type=["pdf"])
question = st.text_input("Ask a question about the PDF:")
if uploaded_pdf:
st.info(f"PDF uploaded: {uploaded_pdf.name}, size: {uploaded_pdf.size} bytes")
if uploaded_pdf and question:
try:
with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp_file:
tmp_file.write(uploaded_pdf.read())
tmp_path = tmp_file.name
st.info(f"Saved PDF to temp file: {tmp_path}")
pdf_agent = PDFAgent(pdf_path=tmp_path)
with st.spinner("Processing..."):
answer = pdf_agent.ask(question)
st.success("Answer:")
st.write(answer)
except Exception as e:
st.error(f"Error processing PDF: {e}")
import traceback
st.text(traceback.format_exc())
with tab2:
st.header("Weather Agent")
location = st.text_input("Enter a location for weather info: e.g. Mumbai")
if location:
weather_agent = WeatherAgent()
with st.spinner("Fetching weather..."):
try:
result = weather_agent.ask(location)
st.success("Weather Info:")
st.write(result) # This might be None or a dict
# Try to extract the answer if it's a dict or object
# if isinstance(result, dict):
# # Try common keys
# if "output" in result:
# st.write(result["output"])
# elif "result" in result:
# st.write(result["result"])
# else:
# st.write(str(result))
# elif hasattr(result, "content"):
# st.write(result.content)
# elif result is not None:
# st.write(str(result))
except Exception as e:
st.error(f"Error: {e}")
with tab3:
st.header("Multi-Agent QA (PDF + Weather)")
user_input = st.text_area("Ask multiple questions (e.g. 'What organizations has Sharath worked for and tell me the weather in Mumbai'):")
uploaded_pdf = st.file_uploader("Upload a PDF for PDF Agent (optional)", type=["pdf"], key="multi_pdf")
if st.button("Ask Multi-Agent"):
from nodes.node import split_questions, classify_question
from langchain_core.messages import HumanMessage
import tempfile
messages = []
# If PDF uploaded, save and use it
pdf_path = None
if uploaded_pdf:
with tempfile.NamedTemporaryFile(delete=False, suffix=".pdf") as tmp_file:
tmp_file.write(uploaded_pdf.read())
pdf_path = tmp_file.name
# Split and process each question
questions = split_questions(user_input)
for question in questions:
agent_name = classify_question(question)
if agent_name == "pdf_agent":
if pdf_path:
pdf_agent = PDFAgent(pdf_path=pdf_path)
else:
pdf_agent = PDFAgent(pdf_path="Sharath_OnePage.pdf")
result = pdf_agent.agent.invoke({"input": question})
if isinstance(result, dict):
text_result = result.get("output") or result.get("text") or str(result)
else:
text_result = str(result)
messages.append(("PDF Agent", text_result))
else:
weather_agent = WeatherAgent()
import re
match = re.search(r"weather in ([\w\s,]+)", question, re.IGNORECASE)
location = match.group(1).strip() if match else question
result = weather_agent.ask(location)
messages.append(("Weather Agent", str(result)))
st.subheader("Results:")
for agent, answer in messages:
st.markdown(f"**{agent}:** {answer}") |