dunno how to fic

rvl_cdip_easyOCR.py

@@ -88,7 +88,9 @@ class RvlCdipEasyOCR(datasets.GeneratorBasedBuilder):
                     "image": datasets.Image(),
                     "label": datasets.ClassLabel(names=_CLASSES),
                     "words": datasets.Sequence(datasets.Value("string")),
-                    "boxes": datasets.Sequence(datasets.Sequence(datasets.Value("int32"))),
+                    "boxes": datasets.Sequence(
+                        datasets.Sequence(datasets.Value("int32"))
+                    ),
                 }
             ),
             homepage=_HOMEPAGE,
@@ -104,7 +106,7 @@ class RvlCdipEasyOCR(datasets.GeneratorBasedBuilder):
                 _URLS["rvl-cdip"]
             )  # only download images if need be
         labels_path = dl_manager.download(_METADATA_URLS)
-
+        from pdb import set_trace; set_trace()
         return [
             datasets.SplitGenerator(
                 name=datasets.Split.TRAIN,
@@ -155,7 +157,7 @@ class RvlCdipEasyOCR(datasets.GeneratorBasedBuilder):
             os.path.join(OCR_dir, f"Easy_{split[0].upper()+split[1:]}_Data.npy"),
             allow_pickle=True,
         )
-        for ex in tqdm(data, desc='Loading OCR data'):
+        for ex in tqdm(data, desc="Loading OCR data"):
             w, h = ex["images"][0]["image_width"], ex["images"][0]["image_height"]
             filename = ex["images"][0]["file_name"]
             words = ex["word-level annotations"][0]["ocred_text"]
@@ -174,7 +176,9 @@ class RvlCdipEasyOCR(datasets.GeneratorBasedBuilder):
         return words, boxes
 
     def _generate_examples(self, archive_iterator, labels_filepath, split):
-        from pdb import set_trace; set_trace()
+        from pdb import set_trace
+
+        set_trace()
         with open(labels_filepath, encoding="utf-8") as f:
             data = f.read().splitlines()
 
@@ -195,5 +199,7 @@ class RvlCdipEasyOCR(datasets.GeneratorBasedBuilder):
                         boxes=boxes,
                     )
                     print(a)
-                    from pdb import set_trace; set_trace()
+                    from pdb import set_trace
+
+                    set_trace()
                     yield file_path, a