CVPR_2024_Papers

```python def remove_ref_tags_except_image_from_sample(sample): """ cleanup the deepseek exrtra tags """ lines = sample["text"].split('\n') filtered_lines = [] for line in lines: # Check if line contains any <|ref|> tag if '<|ref|>' in line: # Keep the line only if it contains <|ref|>image<|/ref|> if '<|ref|>image<|/ref|>' in line: filtered_lines.append(line) else: # Keep lines without <|ref|> tags filtered_lines.append(line) out = '\n'.join(filtered_lines) return {'text':out.strip()} ```