File size: 1,007 Bytes
8e9272d
be6f5ff
8e9272d
be6f5ff
8e9272d
be6f5ff
8e9272d
 
be6f5ff
 
 
 
8e9272d
 
 
be6f5ff
8e9272d
be6f5ff
8e9272d
 
be6f5ff
8e9272d
be6f5ff
 
8e9272d
be6f5ff
 
8e9272d
be6f5ff
8e9272d
be6f5ff
 
 
 
 
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
from langchain import HuggingFaceHub
from langchain.schema import AIMessage
from dotenv import load_dotenv
import streamlit as st

# Load environment variables
load_dotenv()

# Initialize HuggingFace model outside the app
llm_huggingface = HuggingFaceHub(repo_id="google/flan-t5-large", model_kwargs={"temperature": 0.0, "max_length": 64})

# Streamlit app 
st.set_page_config(page_title="Chatbot")
st.header('Langchain Application')

# Remove the session_state initialization as it's not being used in this example

# Function to load HuggingFace model and get response
def get_huggingface_response(question):
    response = llm_huggingface(question) 
    return response

# Streamlit input
user_input = st.text_input("Input: ", key="input")

# Streamlit button
submit = st.button('Generate')

# Check if button is clicked
if submit: 
    # Call function to get response
    response = get_huggingface_response(user_input)
    # Display response
    st.subheader("The response is ")
    st.write(response)