Chesscorner commited on
Commit
b2620a8
·
verified ·
1 Parent(s): eab141c

Update chess_ground-targz.py

Browse files
Files changed (1) hide show
  1. chess_ground-targz.py +13 -5
chess_ground-targz.py CHANGED
@@ -59,7 +59,7 @@ class ChessImageTextDataset(datasets.GeneratorBasedBuilder):
59
  ]
60
 
61
  def _generate_examples(self, image_tar_path, text_filepath):
62
- """Generate examples by linking images with multiple related texts."""
63
  idx = 0
64
  # Extract and map text IDs to their corresponding images
65
  image_mapping = self._extract_images_from_tar(image_tar_path)
@@ -85,11 +85,14 @@ class ChessImageTextDataset(datasets.GeneratorBasedBuilder):
85
 
86
  # Ensure the image exists and yield the example
87
  if image_file:
88
- # Join the texts related to the same image
89
- combined_text = " ".join(texts)
 
 
 
90
  yield idx, {
91
  "image": image_file,
92
- "text": combined_text, # Link all related texts together
93
  }
94
  idx += 1
95
  else:
@@ -97,7 +100,6 @@ class ChessImageTextDataset(datasets.GeneratorBasedBuilder):
97
 
98
  def _extract_images_from_tar(self, tar_path):
99
  """Extracts the images from the tar.gz archive and returns a mapping of image filenames to file paths."""
100
-
101
  image_mapping = {}
102
  extraction_directory = "images_extracted" # Temporary directory to store extracted images
103
  os.makedirs(extraction_directory, exist_ok=True)
@@ -118,3 +120,9 @@ class ChessImageTextDataset(datasets.GeneratorBasedBuilder):
118
  image_mapping[image_filename] = extracted_image_path
119
 
120
  return image_mapping
 
 
 
 
 
 
 
59
  ]
60
 
61
  def _generate_examples(self, image_tar_path, text_filepath):
62
+ """Generate examples by linking images with multiple related texts and clean up the text data."""
63
  idx = 0
64
  # Extract and map text IDs to their corresponding images
65
  image_mapping = self._extract_images_from_tar(image_tar_path)
 
85
 
86
  # Ensure the image exists and yield the example
87
  if image_file:
88
+ # Clean the text to keep only chess notation
89
+ cleaned_texts = [self._extract_chess_notation(text) for text in texts]
90
+
91
+ # Join the cleaned notations related to the same image
92
+ combined_text = " ".join(cleaned_texts)
93
  yield idx, {
94
  "image": image_file,
95
+ "text": combined_text, # Link all related cleaned notations together
96
  }
97
  idx += 1
98
  else:
 
100
 
101
  def _extract_images_from_tar(self, tar_path):
102
  """Extracts the images from the tar.gz archive and returns a mapping of image filenames to file paths."""
 
103
  image_mapping = {}
104
  extraction_directory = "images_extracted" # Temporary directory to store extracted images
105
  os.makedirs(extraction_directory, exist_ok=True)
 
120
  image_mapping[image_filename] = extracted_image_path
121
 
122
  return image_mapping
123
+
124
+ def _extract_chess_notation(self, text):
125
+ """Extracts the chess notation from the full text string."""
126
+ # Assuming the chess notation comes after the filename and space, e.g., '001_0_30_white.png Rxc6'
127
+ notation = text.split(" ", 1)[-1] # Extract everything after the first space
128
+ return notation.strip()