Spaces:
Running
Running
Update app.py
Browse files
app.py
CHANGED
@@ -2,7 +2,7 @@ import gradio as gr
|
|
2 |
import fitz # PyMuPDF
|
3 |
from PIL import Image
|
4 |
|
5 |
-
def
|
6 |
# Open the PDF file
|
7 |
document = fitz.open(pdf_file.name)
|
8 |
# Get the first page
|
@@ -13,12 +13,20 @@ def pdf_to_jpg(pdf_file):
|
|
13 |
image = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
|
14 |
image_path = "first_page.jpg"
|
15 |
image.save(image_path)
|
16 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
17 |
|
18 |
interface = gr.Interface(
|
19 |
-
fn=
|
20 |
inputs=gr.File(label="Upload PDF"),
|
21 |
-
outputs=
|
22 |
)
|
23 |
|
24 |
interface.launch()
|
|
|
2 |
import fitz # PyMuPDF
|
3 |
from PIL import Image
|
4 |
|
5 |
+
def analysis_process(pdf_file):
|
6 |
# Open the PDF file
|
7 |
document = fitz.open(pdf_file.name)
|
8 |
# Get the first page
|
|
|
13 |
image = Image.frombytes("RGB", [pix.width, pix.height], pix.samples)
|
14 |
image_path = "first_page.jpg"
|
15 |
image.save(image_path)
|
16 |
+
|
17 |
+
output = model.inference(image=image_path, prompt=task_prompt)["predictions"][0]
|
18 |
+
return output
|
19 |
+
|
20 |
+
task_name = "SGSInvoice"
|
21 |
+
task_prompt = f"<s_{task_name}>"
|
22 |
+
model = DonutModel.from_pretrained("uartimcs/donut-invoice-extract")
|
23 |
+
model.eval()
|
24 |
+
|
25 |
|
26 |
interface = gr.Interface(
|
27 |
+
fn=analysis_process,
|
28 |
inputs=gr.File(label="Upload PDF"),
|
29 |
+
outputs="json"
|
30 |
)
|
31 |
|
32 |
interface.launch()
|