Update app.py

app.py

@@ -52,28 +52,35 @@ processor = LayoutLMv3Processor.from_pretrained("microsoft/layoutlmv3-base",appl
 model = LayoutLMv3ForTokenClassification.from_pretrained("nielsr/layoutlmv3-finetuned-funsd")
 
 dataset = load_dataset("nielsr/funsd", split="test")
-#image = Image.open(dataset[0]["image_path"]).convert("RGB")
-#image = Image.open("./invoice.png")
-#image.save("document.png")
+image = Image.open(dataset[0]["image_path"]).convert("RGB")
+image = Image.open("./invoice.png")
+image.save("document1.png")
 
+image = Image.open(dataset[1]["image_path"]).convert("RGB")
+image = Image.open("./invoice2.png")
+image.save("document2.png")
+
+image = Image.open(dataset[2]["image_path"]).convert("RGB")
+image = Image.open("./invoice3.png")
+image.save("document3.png")
 
 
 #dataset = load_dataset("nielsr/funsd-layoutlmv3")
 
-example = dataset["test"][0]
-example["image"].save("example1.png")
+#example = dataset["test"][0]
+#example["image"].save("example1.png")
 
-example1 = dataset["test"][1]
-example1["image"].save("example2.png")
+#example1 = dataset["test"][1]
+#example1["image"].save("example2.png")
 
-example2 = dataset["test"][2]
-example2["image"].save("example3.png")
+#example2 = dataset["test"][2]
+#example2["image"].save("example3.png")
 
 #example2["image"]
 
-labels = dataset["test"].features['ner_tags'].feature.names
+labels = dataset.features['ner_tags'].feature.names
 
-words, boxes, ner_tags = example["tokens"], example["bboxes"], example["ner_tags"]
+#words, boxes, ner_tags = example["tokens"], example["bboxes"], example["ner_tags"]
 
 features = dataset["test"].features
 
@@ -86,7 +93,7 @@ label_column_name = "ner_tags"
 
 # In the event the labels are not a `Sequence[ClassLabel]`, we will need to go through the dataset to get the
 # unique labels.
-labels = dataset.features["ner_tags"].feature.names
+
 id2label = {v: k for v, k in enumerate(labels)}
 
 label2color = {
@@ -150,7 +157,7 @@ description = "Extraction of Form or Invoice Extraction - We use Microsoft's Lay
 
 article="<b>References</b><br>[1] Y. Xu et al., “LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking.” 2022. <a href='https://arxiv.org/abs/2204.08387'>Paper Link</a><br>[2]  <a href='https://github.com/NielsRogge/Transformers-Tutorials/tree/master/LayoutLMv3'>LayoutLMv3 training and inference</a>" 
 
-examples =[['example1.png'],['example2.png'],['example3.png']]
+examples =[['document1.png'],['document1.png'],['document1.png']]
 
 css = """.output_image, .input_image {height: 600px !important}"""