File size: 1,448 Bytes
c9e6420 6bd383d 3c0821a 6bd383d 3c0821a 632f752 3c0821a c9e6420 3c0821a 632f752 c9e6420 1e06a50 3c0821a 1e06a50 c9e6420 3c0821a c9e6420 3c0821a c9e6420 3c0821a |
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 |
import gradio as gr
import subprocess
def nougat_ocr(file_name):
# CLI Command to run
cli_command = [
'nougat',
'--out', '/output',
'pdf', f'{file_name}',
'--checkpoint', '/nougat'
]
# Run the command and get .mmd file in an output folder
subprocess.run(cli_command, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
return
def predict(pdf_file):
print(f"temporary file - {pdf_file.name}")
pdf_name = pdf_file.name.split('/')[-1].split('.')[0]
print(f"pdf file name - {pdf_name}")
#! Get prediction for a PDF using nougat
nougat_ocr(pdf_file.name)
# Open the multimarkdown (.mmd) file for reading
with open(f'/output/{pdf_name}.mmd', 'r') as file:
content = file.read()
return content
with gr.Blocks() as demo:
gr.HTML("<h1>Nougat: Neural Optical Understanding for Academic Documents<h1>")
gr.HTML("<h3>Lukas Blecher et al. <a href='https://arxiv.org/pdf/2308.13418.pdf' target='_blank'>Paper</a>, <a href='https://facebookresearch.github.io/nougat/'>Project</a></h3>")
with gr.Row():
pdf_file = gr.File(label='Upload a PDF', scale=1)
mkd = gr.Markdown('<h2><center><i>OR</i></center></h2>',scale=1)
pdf_link = gr.Textbox(placeholder='Enter an arxiv link here', label='Provide a link', scale=1)
btn = gr.Button()
parsed_output = gr.Markdown()
btn.click(predict, pdf_file, parsed_output )
demo.queue()
demo.launch(debug=True)
|