Upload app.py

app.py

@@ -0,0 +1,51 @@
+import re
+import gradio as gr
+
+import torch
+from transformers import DonutProcessor, VisionEncoderDecoderModel
+
+# system parameters
+device = "cuda" if torch.cuda.is_available() else "cpu"
+
+# script parameters
+model_path = "binery/donut_receipt_v2.29"
+title = "Demo: procesar documentos con " + model_path + " " + device
+description = "Sube una factura, haz click en 'submit' y espera a ver los datos extraidos"
+
+# processor and model
+processor = DonutProcessor.from_pretrained(model_path)
+model = VisionEncoderDecoderModel.from_pretrained(model_path)
+model.to(device)
+
+def process_document(image):
+    # prepare encoder inputs
+    pixel_values = processor(image, return_tensors="pt").pixel_values
+    
+    # prepare decoder inputs
+    task_prompt = "<s_cord-v2>"
+    decoder_input_ids = processor.tokenizer(task_prompt, add_special_tokens=False, return_tensors="pt").input_ids
+          
+    # generate answer
+    outputs = model.generate(
+        pixel_values.to(device),
+        decoder_input_ids=decoder_input_ids.to(device),
+        max_length=model.decoder.config.max_position_embeddings,
+        early_stopping=True,
+        pad_token_id=processor.tokenizer.pad_token_id,
+        eos_token_id=processor.tokenizer.eos_token_id,
+        use_cache=True,
+        num_beams=1,
+        bad_words_ids=[[processor.tokenizer.unk_token_id]],
+        return_dict_in_generate=True,
+    )
+    
+    # postprocess
+    sequence = processor.batch_decode(outputs.sequences)[0]
+    sequence = sequence.replace(processor.tokenizer.eos_token, "").replace(processor.tokenizer.pad_token, "")
+    sequence = re.sub(r"<.*?>", "", sequence, count=1).strip()  # remove first task start token
+    
+    return processor.token2json(sequence)
+
+# generate gradio interface
+demo = gr.Interface(fn=process_document, inputs="image", outputs="json", title=title, description=description)
+demo.launch()