Update app.py

app.py

@@ -2,7 +2,7 @@ import gradio as gr
 import torch
 from PIL import Image
 
-#from donut import DonutModel
+from donut import DonutModel
 
 def demo_process(input_img):
     global pretrained_model, task_prompt, task_name
@@ -12,25 +12,26 @@ def demo_process(input_img):
 
 task_prompt = f"<s_cord-v2>"
 
-image = Image.open("/content/SKMBT_75122072616550_Page_37_Image_0001.png")
+image = Image.open("./sample_image_cord_test_receipt_00004.png")
 image.save("cord_sample_receipt1.png")
-image = Image.open("/content/SKMBT_75122072616550_Page_50_Image_0001.png")
+image = Image.open("./sample_image_cord_test_receipt_00012.png")
 image.save("cord_sample_receipt2.png")
 
-#pretrained_model = DonutModel.from_pretrained("naver-clova-ix/donut-base-finetuned-cord-v2")
-#pretrained_model.encoder.to(torch.bfloat16)
-
-model = torch.load("/content/drive/MyDrive/fast_job/DONUT_model/donut/model.pt")
-# Move model to GPU
-device = "cuda" if torch.cuda.is_available() else "cpu"
-model.to(device)
+pretrained_model = DonutModel.from_pretrained("naver-clova-ix/donut-base-finetuned-cord-v2")
+pretrained_model.encoder.to(torch.bfloat16)
+pretrained_model.eval()
 
 demo = gr.Interface(
     fn=demo_process,
     inputs= gr.inputs.Image(type="pil"),
     outputs="json",
-    title=f"Donut 🍩 demonstration for `Medical Prescription Dataset` task",
-    description="""This model is trained with 200 medical prescription handwritten document images. <br>""",
+    title=f"Donut 🍩 demonstration for `cord-v2` task",
+    description="""This model is trained with 800 Indonesian receipt images of CORD dataset. <br>
+Demonstrations for other types of documents/tasks are available at https://github.com/clovaai/donut <br>
+More CORD receipt images are available at https://huggingface.co/datasets/naver-clova-ix/cord-v2
+More details are available at:
+- Paper: https://arxiv.org/abs/2111.15664
+- GitHub: https://github.com/clovaai/donut""",
     examples=[["cord_sample_receipt1.png"], ["cord_sample_receipt2.png"]],
     cache_examples=False,
 )