Download inspect_anytext_errors.py from Leofames/textpainting-precompute-chunks: direct link, hf CLI and curl.
- Browser
- Download file 2.49 kB
-
https://huggingface.co/datasets/Leofames/textpainting-precompute-chunks/resolve/main/inspect_anytext_errors.py
- Command line
-
hf download hf://datasets/Leofames/textpainting-precompute-chunks/inspect_anytext_errors.py
-
curl -L -o inspect_anytext_errors.py https://huggingface.co/datasets/Leofames/textpainting-precompute-chunks/resolve/main/inspect_anytext_errors.py
2.49 kB
| import json | |
| from pathlib import Path | |
| from collections import Counter | |
| BASE_DIR = Path(__file__).resolve().parent | |
| JSON_PATH = BASE_DIR / "train.json" | |
| with open(JSON_PATH, "r", encoding="utf-8") as f: | |
| records = json.load(f) | |
| annotation_types = Counter() | |
| non_list_examples = [] | |
| boundary_examples = [] | |
| for i, record in enumerate(records): | |
| annotations = record.get("annotations") | |
| annotation_types[type(annotations).__name__] += 1 | |
| if not isinstance(annotations, list) and len(non_list_examples) < 10: | |
| non_list_examples.append({ | |
| "record_index": i, | |
| "img_name": record.get("img_name"), | |
| "local_image": record.get("local_image"), | |
| "annotations_type": type(annotations).__name__, | |
| "annotations_value": annotations, | |
| }) | |
| if isinstance(annotations, list): | |
| for ann_index, ann in enumerate(annotations): | |
| if not isinstance(ann, dict): | |
| continue | |
| polygon = ann.get("polygon", []) | |
| if not isinstance(polygon, list): | |
| continue | |
| for point in polygon: | |
| if ( | |
| isinstance(point, (list, tuple)) | |
| and len(point) == 2 | |
| and all(isinstance(v, (int, float)) for v in point) | |
| and (point[0] >= 512 or point[1] >= 512 | |
| or point[0] < 0 or point[1] < 0) | |
| ): | |
| if len(boundary_examples) < 10: | |
| boundary_examples.append({ | |
| "record_index": i, | |
| "img_name": record.get("img_name"), | |
| "local_image": record.get("local_image"), | |
| "annotation_index": ann_index, | |
| "text": ann.get("text"), | |
| "polygon": polygon, | |
| "image_width": record.get("image_width"), | |
| "image_height": record.get("image_height"), | |
| }) | |
| break | |
| print("ANNOTATION VALUE TYPES") | |
| for kind, count in annotation_types.items(): | |
| print(f"{kind}: {count}") | |
| print("\nEXAMPLES WHERE ANNOTATIONS IS NOT A LIST") | |
| print(json.dumps(non_list_examples, ensure_ascii=False, indent=2)) | |
| print("\nEXAMPLES WITH OUT-OF-BOUNDS COORDINATES") | |
| print(json.dumps(boundary_examples, ensure_ascii=False, indent=2)) | |