Spaces:
Sleeping
Sleeping
File size: 2,955 Bytes
9e83aca 80984a0 930f6f4 5e5ae9f 80984a0 1011c08 80984a0 930f6f4 5e5ae9f 80984a0 1011c08 80984a0 930f6f4 5e5ae9f 930f6f4 5e5ae9f 930f6f4 80984a0 930f6f4 80984a0 930f6f4 80984a0 930f6f4 9e83aca 930f6f4 62d24b8 |
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 |
import gradio as gr
import bs4
from langchain.text_splitter import RecursiveCharacterTextSplitter
from langchain_community.document_loaders import WebBaseLoader
from langchain_community.vectorstores import Chroma
from langchain_huggingface import HuggingFaceEndpointEmbeddings # Updated import
from dotenv import load_dotenv
import os
# Load the environment variables from the .env file
load_dotenv()
# Install the Google Generative AI SDK
# $ pip install google-generativeai
import google.generativeai as genai
# Configure the Google Generative AI client
genai.configure(api_key=os.environ["GEMINI_API_KEY"])
# Create the Google Generative AI model
generation_config = {
"temperature": 1,
"top_p": 0.95,
"top_k": 64,
"max_output_tokens": 8192,
"response_mime_type": "text/plain",
}
model = genai.GenerativeModel(
model_name="gemini-1.5-flash",
generation_config=generation_config,
)
import requests
from requests.exceptions import ReadTimeout
# Function to load, split, and retrieve documents
def load_and_retrieve_docs(url, max_retries=3):
retry_count = 0
while retry_count < max_retries:
try:
loader = WebBaseLoader(
web_paths=(url,),
bs_kwargs=dict()
)
docs = loader.load()
text_splitter = RecursiveCharacterTextSplitter(chunk_size=1000, chunk_overlap=200)
splits = text_splitter.split_documents(docs)
embeddings = HuggingFaceEndpointEmbeddings(
endpoint_url="https://api-inference.huggingface.co/models/sentence-transformers/all-mpnet-base-v2",
api_key=os.environ["HUGGINGFACE_API_KEY"]
) # Updated embeddings
vectorstore = Chroma.from_documents(documents=splits, embedding=embeddings)
return vectorstore.as_retriever()
except ReadTimeout:
retry_count += 1
continue
# If maximum retries reached
print("Maximum retries reached. Unable to fetch the URL.")
return None
# Function to format documents
def format_docs(docs):
return "\n\n".join(doc.page_content for doc in docs)
# Function that defines the RAG chain
def rag_chain(url, question):
retriever = load_and_retrieve_docs(url)
if retriever is None:
return "Error: Maximum retries reached. Unable to fetch the URL."
retrieved_docs = retriever.invoke(question)
formatted_context = format_docs(retrieved_docs)
formatted_prompt = f"Question: {question}\n\nContext: {formatted_context}"
chat_session = model.start_chat(history=[])
response = chat_session.send_message(formatted_prompt)
return response.text.strip()
# Gradio interface
iface = gr.Interface(
fn=rag_chain,
inputs=["text", "text"],
outputs="text",
title="CHAT WITH URL",
description="Enter a URL and a query to get answers from the RAG chain."
)
# Launch the app
iface.launch()
|