updated readme

README.md

@@ -1,7 +1,13 @@
 ---
 language: en
 license: cc-by-nc-sa-4.0
-
+tags:
+  - endpoints-template
+library_name: generic
+model-index:
+  - name: layoutlmv2-base-uncased
+    results: []
+pipeline_tag: other
 ---
 
 # LayoutLMv2
@@ -24,7 +30,7 @@ Examples & Guides
 - https://mccormickml.com/2020/03/10/question-answering-with-a-fine-tuned-BERT/
 
 
-# Errors
+# Warnings
 
 ```
 The class LayoutLMv2FeatureExtractor is deprecated and will be removed in version 5 of Transformers. Please use LayoutLMv2ImageProcessor instead.

handler.py

@@ -8,17 +8,17 @@ from transformers import LayoutLMv2TokenizerFast
 from transformers.tokenization_utils_base import BatchEncoding
 from transformers.tokenization_utils_base import TruncationStrategy
 from transformers.utils import TensorType
-from transformers.modeling_outputs import (
-    QuestionAnsweringModelOutput as QuestionAnsweringModelOutputBase
-)
+# from transformers.modeling_outputs import (
+    # QuestionAnsweringModelOutput as QuestionAnsweringModelOutputBase
+# )
 import numpy as np
-from PIL import Image, ImageDraw, ImageFont
-from subprocess import run
+# from PIL import Image, ImageDraw, ImageFont
+# from subprocess import run
 import pdf2image
-from pprint import pprint
+# from pprint import pprint
 import logging
 from os import environ
-from dataclasses import dataclass
+# from dataclasses import dataclass
 
 # device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
 # install tesseract-ocr and pytesseract
@@ -163,9 +163,8 @@ class EndpointHandler:
                 includes the deserialized image file as PIL.Image
         """
         image = data.pop("inputs", data)
-
-        # image = pdf_to_image(image)
         images = [x.convert("RGB") for x in pdf2image.convert_from_bytes(image)]
+
         question = "what is the bill date"
         with torch.no_grad():
             for image in images:
@@ -207,9 +206,9 @@ class EndpointHandler:
                 target_start_index = torch.tensor([7])
                 target_end_index = torch.tensor([14])
                 outputs = self.model(**encoding, start_positions=target_start_index, end_positions=target_end_index)
-                predicted_answer_span_start = outputs.start_logits.argmax(-1).item()
-                predicted_answer_span_end = outputs.end_logits.argmax(-1).item()
-                # print(predicted_answer_span_start, predicted_answer_span_end)
+                # predicted_answer_span_start = outputs.start_logits.argmax(-1).item()
+                # predicted_answer_span_end = outputs.end_logits.argmax(-1).item()
+
                 logger.info(f'''
     START
     predicted_start_idx: {predicted_start_idx}