Create app.py

app.py

@@ -0,0 +1,47 @@
+# Load model directly
+from transformers import AutoTokenizer, AutoModelForImageTextToText
+from transformers import VisionEncoderDecoderModel, TrOCRProcessor
+import torch
+import torch
+from PIL import Image
+
+processor = TrOCRProcessor.from_pretrained("microsoft/trocr-base-printed")
+model = AutoModelForImageTextToText.from_pretrained("ChronoStellar/TrOCR_IndonesianLPR")
+
+device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
+model.to(device)
+
+import gradio as gr
+from PIL import Image
+import torch
+
+# Assuming model, processor, and device are already defined
+
+def OCR(pil_image, model=model, processor=processor, device=device):
+    # Prepare image for the model
+    pixel_values = processor(pil_image, return_tensors="pt").pixel_values
+
+    # Move the input to the appropriate device (CPU/GPU)
+    pixel_values = pixel_values.to(device)
+
+    # Generate prediction
+    model.eval()  # Set the model to evaluation mode
+    with torch.no_grad():  # Disable gradient calculation for inference
+        generated_ids = model.generate(pixel_values)
+
+    # Decode the predicted IDs to get the text
+    predicted_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
+    
+    return predicted_text
+
+# Create Gradio interface
+interface = gr.Interface(
+    fn=OCR,
+    inputs=gr.Image(type="pil", label="Upload License Plate Image"),
+    outputs=gr.Textbox(label="Predicted License Plate"),
+    title="Automatic License Plate Recognition",
+    description="Upload an image of a license plate, and the system will predict the text on it.",
+)
+
+# Launch the Gradio app
+interface.launch()