Update app.py

app.py

@@ -16,25 +16,25 @@ image.save("hack.png")
 
 def demo_process(img, question):
     
-    # processor = AutoProcessor.from_pretrained(model_checkpoint)
-    # model = AutoModelForDocumentQuestionAnswering.from_pretrained(model_checkpoint)
-    # with torch.no_grad():
-    #     encoding = processor(img.convert("RGB"), question, return_tensors="pt")
-    #     outputs = model(**encoding)
-    #     start_logits = outputs.start_logits
-    #     end_logits = outputs.end_logits
-    #     predicted_start_idx = start_logits.argmax(-1).item()
-    #     predicted_end_idx = end_logits.argmax(-1).item()
+    processor = AutoProcessor.from_pretrained(model_checkpoint)
+    model = AutoModelForDocumentQuestionAnswering.from_pretrained(model_checkpoint)
+    with torch.no_grad():
+        encoding = processor(img.convert("RGB"), question, return_tensors="pt")
+        outputs = model(**encoding)
+        start_logits = outputs.start_logits
+        end_logits = outputs.end_logits
+        predicted_start_idx = start_logits.argmax(-1).item()
+        predicted_end_idx = end_logits.argmax(-1).item()
 
-    # processor.tokenizer.decode(encoding.input_ids.squeeze()[predicted_start_idx : predicted_end_idx + 1])
-    # predicted_answer_tokens = encoding.input_ids.squeeze()[predicted_start_idx : predicted_end_idx + 1]
+    processor.tokenizer.decode(encoding.input_ids.squeeze()[predicted_start_idx : predicted_end_idx + 1])
+    predicted_answer_tokens = encoding.input_ids.squeeze()[predicted_start_idx : predicted_end_idx + 1]
 
-    # predicted_answer = processor.tokenizer.decode(predicted_answer_tokens)
+    predicted_answer = processor.tokenizer.decode(predicted_answer_tokens)
 
-    # return predicted_answer
+    return predicted_answer
 
-    qa_pipeline = pipeline("document-question-answering", model="pacman2223/test-mod")
-    qa_pipeline(img, question)
+    # qa_pipeline = pipeline("document-question-answering", model="pacman2223/test-mod")
+    # qa_pipeline(img, question)
 
     return qa_pipeline