Spaces:
Paused
Paused
Update app.py
Browse files
app.py
CHANGED
|
@@ -2,20 +2,24 @@ from transformers import TrOCRProcessor, VisionEncoderDecoderModel
|
|
| 2 |
from PIL import Image
|
| 3 |
import pdf2image
|
| 4 |
import gradio as gr
|
|
|
|
|
|
|
| 5 |
import torch
|
| 6 |
|
| 7 |
-
# ✅ Używamy
|
| 8 |
MODEL_NAME = "microsoft/trocr-base-stage1"
|
| 9 |
-
|
| 10 |
-
# ✅ Ładujemy model i procesor
|
| 11 |
processor = TrOCRProcessor.from_pretrained(MODEL_NAME)
|
| 12 |
model = VisionEncoderDecoderModel.from_pretrained(MODEL_NAME)
|
| 13 |
|
| 14 |
-
# ✅ Funkcja
|
| 15 |
-
def process_invoice(
|
| 16 |
try:
|
|
|
|
|
|
|
|
|
|
|
|
|
| 17 |
# ✅ Konwersja PDF do obrazu
|
| 18 |
-
images = pdf2image.convert_from_bytes(
|
| 19 |
|
| 20 |
# ✅ Przetwarzanie pierwszej strony faktury
|
| 21 |
image = images[0].convert("RGB")
|
|
@@ -29,13 +33,13 @@ def process_invoice(pdf_file):
|
|
| 29 |
except Exception as e:
|
| 30 |
return f"Błąd przetwarzania: {str(e)}"
|
| 31 |
|
| 32 |
-
# ✅
|
| 33 |
interface = gr.Interface(
|
| 34 |
fn=process_invoice,
|
| 35 |
-
inputs=gr.
|
| 36 |
outputs="text",
|
| 37 |
title="AI Invoice Processor",
|
| 38 |
-
description="
|
| 39 |
)
|
| 40 |
|
| 41 |
# ✅ Uruchomienie aplikacji
|
|
|
|
| 2 |
from PIL import Image
|
| 3 |
import pdf2image
|
| 4 |
import gradio as gr
|
| 5 |
+
import base64
|
| 6 |
+
import io
|
| 7 |
import torch
|
| 8 |
|
| 9 |
+
# ✅ Używamy lżejszego modelu dla większej wydajności
|
| 10 |
MODEL_NAME = "microsoft/trocr-base-stage1"
|
|
|
|
|
|
|
| 11 |
processor = TrOCRProcessor.from_pretrained(MODEL_NAME)
|
| 12 |
model = VisionEncoderDecoderModel.from_pretrained(MODEL_NAME)
|
| 13 |
|
| 14 |
+
# ✅ Funkcja przetwarzania faktury (odbiera Base64 PDF)
|
| 15 |
+
def process_invoice(pdf_base64):
|
| 16 |
try:
|
| 17 |
+
# ✅ Dekodowanie Base64 na plik PDF
|
| 18 |
+
pdf_bytes = base64.b64decode(pdf_base64)
|
| 19 |
+
pdf_stream = io.BytesIO(pdf_bytes)
|
| 20 |
+
|
| 21 |
# ✅ Konwersja PDF do obrazu
|
| 22 |
+
images = pdf2image.convert_from_bytes(pdf_stream.read())
|
| 23 |
|
| 24 |
# ✅ Przetwarzanie pierwszej strony faktury
|
| 25 |
image = images[0].convert("RGB")
|
|
|
|
| 33 |
except Exception as e:
|
| 34 |
return f"Błąd przetwarzania: {str(e)}"
|
| 35 |
|
| 36 |
+
# ✅ API Gradio (teraz akceptuje Base64 jako wejście)
|
| 37 |
interface = gr.Interface(
|
| 38 |
fn=process_invoice,
|
| 39 |
+
inputs=gr.Textbox(label="PDF w Base64"),
|
| 40 |
outputs="text",
|
| 41 |
title="AI Invoice Processor",
|
| 42 |
+
description="Prześlij fakturę jako Base64, a AI wyodrębni dane tekstowe.",
|
| 43 |
)
|
| 44 |
|
| 45 |
# ✅ Uruchomienie aplikacji
|