rewrote tokenizer

docbank.py

@@ -18,7 +18,7 @@ import csv
 import os
 import numpy as np
 from PIL import Image
-from transformers import AutoTokenizer
+from transformers import LayoutXLMTokenizerFast
 import datasets
 
 # TODO: Add BibTeX citation
@@ -131,7 +131,7 @@ class Docbank(datasets.GeneratorBasedBuilder):
     ]
 
     # DEFAULT_CONFIG_NAME = "small"  # It's not mandatory to have a default configuration. Just use one if it make sense.
-    TOKENIZER = AutoTokenizer.from_pretrained("xlm-roberta-base")
+    TOKENIZER = LayoutXLMTokenizerFast.from_pretrained("microsoft/layoutxlm-base")
 
     def _info(self):
         # TODO: This method specifies the datasets.DatasetInfo object which contains informations and typings for the dataset
@@ -220,35 +220,51 @@ class Docbank(datasets.GeneratorBasedBuilder):
             #image, size = load_image(f_fp_img, size=224)
             original_image, _ = load_image(f_fp_img)
 
+
             try:
                 with open(f_fp_txt, newline='', encoding='utf-8') as csvfile:
                     reader = csv.reader(csvfile, delimiter='\t', quotechar=' ')
                     for row in reader:
                         # normalized_bbox = normalize_bbox(row[1:5], size)
                         normalized_bbox = [int(x) for x in row[1:5]]
-                        tokenized_input = self.TOKENIZER(
-                            row[0],
-                            add_special_tokens=False,
-                            return_offsets_mapping=False,
-                            return_attention_mask=False,
-                            max_length=512, truncation=True
-                        )
-                        for tkn in tokenized_input['input_ids']:
-                            tokens.append(tkn)
-                            bboxes.append(normalized_bbox)
-                            # rgbs.append(row[5:8])
-                            # fonts.append(row[8])
-                            labels.append(row[9])
+                        tokens.append(row[0])
+                        bboxes.append(normalized_bbox)
+                        labels.append(row[9])
+                        # tokenized_input = self.TOKENIZER(
+                        #     row[0],
+                        #     add_special_tokens=False,
+                        #     return_offsets_mapping=False,
+                        #     return_attention_mask=False,
+                        #     max_length=512, truncation=True
+                        # )
+                        #
+                        # for tkn in tokenized_input['input_ids']:
+                        #     tokens.append(tkn)
+                        #     bboxes.append(normalized_bbox)
+                        #     # rgbs.append(row[5:8])
+                        #     # fonts.append(row[8])
+                        #     labels.append(row[9])
 
             except:
                 continue
 
-            for chunk_id, index in enumerate(range(0, len(tokens), self.CHUNK_SIZE)):
-                split_tokens = tokens[index:index + self.CHUNK_SIZE]
-                split_bboxes = bboxes[index:index + self.CHUNK_SIZE]
+            processed = self.TOKENIZER(
+                tokens,
+                boxes=bboxes,
+                word_labels=labels,
+                add_special_tokens=False,
+                return_offsets_mapping=False,
+                return_attention_mask=False,
+            )
+
+            for chunk_id, index in enumerate(range(0, len(processed['input_ids'][0]), self.CHUNK_SIZE)):
+                split_tokens = processed['input_ids'][0][index:index + self.CHUNK_SIZE]
+                split_bboxes = processed['bbox'][0][index:index + self.CHUNK_SIZE]
                 # split_rgbs = rgbs[index:index + self.CHUNK_SIZE]
                 # split_fonts = fonts[index:index + self.CHUNK_SIZE]
-                split_labels = labels[index:index + self.CHUNK_SIZE]
+                split_labels = processed['labels'][0][index:index + self.CHUNK_SIZE]
+
+                #tokenized = self.TOKENIZER(processed['words'], boxes=processed['boxes'])
 
                 yield key, {
                     "id": f"{f_id}_{chunk_id}",