Spaces:
Sleeping
Sleeping
File size: 1,518 Bytes
b6dd162 8fd3ffb b6dd162 |
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 |
import os
from dotenv import load_dotenv
from PyPDF2 import PdfReader
from langchain.text_splitter import CharacterTextSplitter
from langchain import vectorstores
from langchain import chains
from langchain import llms
from langchain.embeddings import HuggingFaceEmbeddings
import gradio as gr
llm = llms.AI21(ai21_api_key='diNNQzvL40ZnBnEQkIBwNESWjtj792NG')
def pdf_qa(pdf, query):
if pdf is not None:
pdf_reader = PdfReader(pdf)
texts = ""
for page in pdf_reader.pages:
texts += page.extract_text()
text_splitter = CharacterTextSplitter(
separator="\n",
chunk_size=1000,
chunk_overlap=0
)
chunks = text_splitter.split_text(texts)
embeddings = HuggingFaceEmbeddings()
db = vectorstores.Chroma.from_texts(chunks, embeddings)
retriever = db.as_retriever(search_type="similarity", search_kwargs={"k": 10})
qa = chains.ConversationalRetrievalChain.from_llm(llm=llm, retriever=retriever)
chat_history = []
if query:
result = qa({"question": query, "chat_history": chat_history})
return result["answer"]
return "Please upload a PDF and enter a query."
pdf_input = gr.inputs.File(label="Upload your PDF", type="file", file_count="single")
query_input = gr.inputs.Textbox(label="Ask a question in PDF")
output = gr.outputs.Textbox(label="Answer")
gr.Interface(fn=pdf_qa, inputs=[pdf_input, query_input], outputs=output, title="PDF QA").launch()
|