Spaces:
Runtime error
Runtime error
gabrielchua
commited on
Commit
•
36dd136
1
Parent(s):
fc2e110
Create app.py
Browse files
app.py
ADDED
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
"""
|
2 |
+
app.py
|
3 |
+
"""
|
4 |
+
|
5 |
+
import base64
|
6 |
+
|
7 |
+
import gradio as gr
|
8 |
+
from groq import Groq
|
9 |
+
|
10 |
+
# Initialize Groq client
|
11 |
+
client = Groq()
|
12 |
+
|
13 |
+
# Function to encode the image in base64
|
14 |
+
def encode_image_to_base64(image):
|
15 |
+
# Convert PIL image to base64 string
|
16 |
+
buffered = BytesIO()
|
17 |
+
image.save(buffered, format="JPEG")
|
18 |
+
return base64.b64encode(buffered.getvalue()).decode('utf-8')
|
19 |
+
|
20 |
+
# Function to process the uploaded image and extract receipt information
|
21 |
+
def extract_receipt_info(image):
|
22 |
+
# Encode the image to base64
|
23 |
+
base64_image = encode_image_to_base64(image)
|
24 |
+
`
|
25 |
+
# Send request to Groq API
|
26 |
+
chat_completion = client.chat.completions.create(
|
27 |
+
model="llama-3.2-11b-vision-preview",
|
28 |
+
messages=[
|
29 |
+
{
|
30 |
+
"role": "user",
|
31 |
+
"content": [
|
32 |
+
{
|
33 |
+
"type": "text",
|
34 |
+
"text": "Your task is to extract key information from the provided receipt image.\n\nReply in table.\n\nThis is the schema:\n- item (str), description of the item\n- price (float), price of the item\n- quantity (int), quantity of the item\n- total (float), total cost for the item"
|
35 |
+
},
|
36 |
+
{
|
37 |
+
"type": "image_url",
|
38 |
+
"image_url": {
|
39 |
+
"url": f"data:image/jpeg;base64,{base64_image}",
|
40 |
+
},
|
41 |
+
},
|
42 |
+
],
|
43 |
+
},
|
44 |
+
{
|
45 |
+
"role": "assistant",
|
46 |
+
"content": "```markdown"
|
47 |
+
}
|
48 |
+
],
|
49 |
+
temperature=0.1,
|
50 |
+
max_tokens=8192,
|
51 |
+
top_p=1,
|
52 |
+
stop="```",
|
53 |
+
)
|
54 |
+
|
55 |
+
# Return the response from the model
|
56 |
+
return chat_completion.choices[0].message.content
|
57 |
+
|
58 |
+
# Create the Gradio app
|
59 |
+
def gradio_app():
|
60 |
+
# Gradio interface
|
61 |
+
gr.Interface(
|
62 |
+
fn=extract_receipt_info,
|
63 |
+
inputs=gr.inputs.Image(type="pil", label="Upload Receipt Image"),
|
64 |
+
outputs="text",
|
65 |
+
title="Receipt Information Extractor",
|
66 |
+
description="Upload a receipt image and the model will extract the items, quantities, and prices from the receipt."
|
67 |
+
).launch()
|
68 |
+
|
69 |
+
# Start the Gradio app
|
70 |
+
if __name__ == "__main__":
|
71 |
+
gradio_app()
|