UReader data (kg/qa) in llava-onevision-data does not match with images

[khanrc]

As discussed in https://huggingface.co/datasets/lmms-lab/LLaVA-OneVision-Data/discussions/5, the ureader_kg and ureader_qa data are not matched with images.

I was able to recover 80-90% images by matching suffixes using id (adding ".png" or ".jpeg" to id), but still 10-20% images are not matched.