Spaces:
Sleeping
Sleeping
Update app.py
Browse files
app.py
CHANGED
|
@@ -5,34 +5,52 @@ import pdfplumber
|
|
| 5 |
# Load the pre-trained question-answering model
|
| 6 |
qa_pipeline = pipeline("question-answering", model="distilbert-base-cased-distilled-squad")
|
| 7 |
|
| 8 |
-
|
| 9 |
-
|
| 10 |
-
try:
|
| 11 |
-
# Read and extract text from the uploaded PDF file
|
| 12 |
-
with pdfplumber.open(file) as pdf:
|
| 13 |
-
text = ""
|
| 14 |
-
for page in pdf.pages:
|
| 15 |
-
text += page.extract_text()
|
| 16 |
|
| 17 |
-
|
| 18 |
-
|
| 19 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
|
| 21 |
-
|
| 22 |
-
|
| 23 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 24 |
|
| 25 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 26 |
fn=answer_question,
|
| 27 |
-
inputs=
|
| 28 |
-
gr.File(label="Upload PDF"),
|
| 29 |
-
gr.Textbox(label="Enter Question", type="text"),
|
| 30 |
-
gr.Button("Answer"),
|
| 31 |
-
],
|
| 32 |
outputs="text",
|
| 33 |
live=True,
|
| 34 |
title="PDF Question-Answering",
|
| 35 |
-
description="
|
| 36 |
)
|
| 37 |
|
| 38 |
-
|
|
|
|
|
|
| 5 |
# Load the pre-trained question-answering model
|
| 6 |
qa_pipeline = pipeline("question-answering", model="distilbert-base-cased-distilled-squad")
|
| 7 |
|
| 8 |
+
# Shared variable to store uploaded PDF text
|
| 9 |
+
pdf_text = ""
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 10 |
|
| 11 |
+
# Function to load the PDF and store its text
|
| 12 |
+
def load_pdf(file):
|
| 13 |
+
global pdf_text
|
| 14 |
+
try:
|
| 15 |
+
with pdfplumber.open(file) as pdf:
|
| 16 |
+
pdf_text = ""
|
| 17 |
+
for page in pdf.pages:
|
| 18 |
+
pdf_text += page.extract_text()
|
| 19 |
+
return "PDF loaded successfully."
|
| 20 |
+
except Exception as e:
|
| 21 |
+
return f"Error processing PDF: {str(e)}"
|
| 22 |
|
| 23 |
+
# Function to answer the user's question based on the loaded PDF
|
| 24 |
+
def answer_question(question):
|
| 25 |
+
if not pdf_text:
|
| 26 |
+
return "No PDF loaded. Upload a PDF first."
|
| 27 |
+
|
| 28 |
+
try:
|
| 29 |
+
# Ask the user's question using the question-answering model
|
| 30 |
+
answer = qa_pipeline({"context": pdf_text, "question": question})
|
| 31 |
+
return answer["answer"]
|
| 32 |
+
except Exception as e:
|
| 33 |
+
return f"Error answering question: {str(e)}"
|
| 34 |
|
| 35 |
+
# Interface for uploading the PDF
|
| 36 |
+
pdf_interface = gr.Interface(
|
| 37 |
+
fn=load_pdf,
|
| 38 |
+
inputs=gr.File(label="Upload PDF"),
|
| 39 |
+
outputs="text",
|
| 40 |
+
live=True,
|
| 41 |
+
title="PDF Uploader",
|
| 42 |
+
description="Upload a PDF to load its content.",
|
| 43 |
+
)
|
| 44 |
+
|
| 45 |
+
# Interface for answering questions based on the loaded PDF
|
| 46 |
+
qa_interface = gr.Interface(
|
| 47 |
fn=answer_question,
|
| 48 |
+
inputs=gr.Textbox(label="Enter Question", type="text"),
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
outputs="text",
|
| 50 |
live=True,
|
| 51 |
title="PDF Question-Answering",
|
| 52 |
+
description="Enter a question to get an answer based on the loaded PDF.",
|
| 53 |
)
|
| 54 |
|
| 55 |
+
pdf_interface.launch()
|
| 56 |
+
qa_interface.launch()
|