Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/mdarshad1000/entity_extractor

A python script to extract entities from PDF using OpenAI Json Output and the latest Structured Output Response
https://github.com/mdarshad1000/entity_extractor

Last synced: about 6 hours ago
JSON representation

A python script to extract entities from PDF using OpenAI Json Output and the latest Structured Output Response

Host: GitHub
URL: https://github.com/mdarshad1000/entity_extractor
Owner: mdarshad1000
Created: 2024-08-08T21:59:22.000Z (3 months ago)
Default Branch: master
Last Pushed: 2024-08-09T06:17:44.000Z (3 months ago)
Last Synced: 2024-08-10T00:26:45.112Z (3 months ago)
Language: Python
Size: 394 KB
Stars: 0
Watchers: 2
Forks: 0
Open Issues: 0
Metadata Files:
- Readme: README.md

Awesome Lists containing this project

README

# Entity Extractor
Entity Extractor is a python script designed to extract key information (Customer details, Products, Total Amount) from a PDF Invoice.

## Features

#### Structured Output Response
- Structured Outputs ensures the model will always generate responses that adhere to the supplied Pydantic Schema (which in this case has been prepared according to the Sample PDF provided), so you don't need to worry about the model omitting a required key, or hallucinating an invalid enum value.

#### JSON Output
- JSON Response also does the same job except that it's more flexible with the output provided.

### IMPORTANT POINT
- Use `Structured Output Response` if all your PDF Invoices contain the same/similar fields to that of the provided sample PDF. Optionally, you can still change the Pydantic schema to suit your task.
- `JSON Output` is more flexible comparatively as it works for all type of PDF Invoices.

## SETUP
1. Create a `.env` and Set `OPENAI_API_KEY` as `OPENAI_API_KEY=sk-******************`
2. Install the requirements as:
```sh
pip install -r requirements.txt
```
3. Update the PDF path (optionally you can run it on the sample PDF) in `app.py`
```
pdf_path = "data/sample.pdf"
```
4. Run the script `app.py`
```sh
python3 app.py
```

#### EXAMPLE
1. STRUCTURED OUTPUT RESPONSE
```python
data = structured_ouput_extractor(text)
print_structured_response(data)
```

#### OUTPUT
```
customer_details:
name: TEST
address: Test, Hyderabad, TELANGANA, 500089
contact:
phone: 9108239284
email: [email protected]
products:
Item 1:
name: WASTE AND SCRAP OF STAINLESS STEEL
quantity: 6,790 KGS
price: 6,45,050.00
total_amount:
taxable_amount: ₹6,45,050.00
IGST: ₹1,16,109.00
round_off: 0.41
TCS: ₹7611.59
total: ₹7,68,771.00
amount_payable: ₹7,68,771.00
```

2. JSON OUTPUT RESPONSE
```python
data = json_output_extractor(system_prompt, user_prompt, text)
print(data)
```

#### OUTPUT
```
{
"CUSTOMER_DETAILS": {
"name": "TEST",
"billing_address": "Test, Hyderabad, TELANGANA, 500089",
"contact_information": {
"phone": "9108239284",
"email": "[email protected]"
}
},
"PRODUCTS": [
{
"name": "WASTE AND SCRAP OF STAINLESS STEEL",
"quantity": "6,790 KGS",
"price": "₹6,45,050.00"
}
],
"TOTAL_AMOUNT": {
"taxable_amount": "₹6,45,050.00",
"IGST": "₹1,16,109.00",
"round_off": "0.41",
"total": "₹7,68,771.00",
"TCS": "₹7,611.59",
"amount_payable": "₹7,68,771.00"
}
}
```

3. JSON RESPONSE FOR IMAGE (Structured Output Response can also be used for images)
```python
data = json_output_extractor(system_prompt, user_prompt, image)
print(data)
```

#### OUTPUT
```
{
"CUSTOMER_DETAILS": {
"name": "TEST",
"billing_address": "Hyderabad, TELANGANA, 500089",
"contact_information": {
"phone": "9108239284",
"email": "[email protected]"
}
},
"PRODUCTS": [
{
"name": "WASTE AND SCRAP OF STAINLESS STEEL",
"quantity": "6,790 KGS",
"price": "6,45,050.00"
}
],
"TOTAL_AMOUNT": {
"taxable_amount": "6,45,050.00",
"IGST": "1,16,109.00",
"round_off": "0.41",
"TCS": "27,611.50",
"total": "7,68,771.00",
"amount_in_words": "INR Seven Lakh, Sixty-Eight Thousand, Seven Hundred And Seventy-One Rupees Only"
}
}
```