Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import gradio as gr
|
| 2 |
+
from transformers import pipeline
|
| 3 |
+
import PyPDF2
|
| 4 |
+
|
| 5 |
+
# Carregar o modelo de linguagem gratuito da Hugging Face
|
| 6 |
+
nlp = pipeline("question-answering", model="distilbert-base-cased-distilled-squad")
|
| 7 |
+
|
| 8 |
+
def extract_text_from_pdf(pdf_file):
|
| 9 |
+
with open(pdf_file.name, "rb") as file:
|
| 10 |
+
reader = PyPDF2.PdfFileReader(file)
|
| 11 |
+
text = ""
|
| 12 |
+
for page_num in range(reader.numPages):
|
| 13 |
+
page = reader.getPage(page_num)
|
| 14 |
+
text += page.extract_text()
|
| 15 |
+
return text
|
| 16 |
+
|
| 17 |
+
def answer_question(pdf_file, question):
|
| 18 |
+
# Extrair texto do PDF
|
| 19 |
+
context = extract_text_from_pdf(pdf_file)
|
| 20 |
+
|
| 21 |
+
# Usar o modelo para responder a pergunta
|
| 22 |
+
result = nlp(question=question, context=context)
|
| 23 |
+
return result['answer']
|
| 24 |
+
|
| 25 |
+
# Interface Gradio
|
| 26 |
+
iface = gr.Interface(
|
| 27 |
+
fn=answer_question,
|
| 28 |
+
inputs=[
|
| 29 |
+
gr.inputs.File(label="Carregar PDF"),
|
| 30 |
+
gr.inputs.Textbox(label="Pergunta")
|
| 31 |
+
],
|
| 32 |
+
outputs=gr.outputs.Textbox(label="Resposta"),
|
| 33 |
+
title="QA sobre PDF",
|
| 34 |
+
description="Carregue um PDF e faça perguntas sobre o conteúdo."
|
| 35 |
+
)
|
| 36 |
+
|
| 37 |
+
# Iniciar a interface
|
| 38 |
+
iface.launch()
|