Inquiry Regarding Evaluation with coco_1024 Prompts

[xxw11]

We are currently working on FID evaluation using the coco_1024.txt file. However, we have encountered an issue where not all COCO prompts in the coco_1024.txt file seem to match corresponding images in the COCO2017/annotations dataset. This appears to be inconsistent with the expected description. Could you provide more specific details regarding the evaluation process? Additionally, would it be possible for you to share the exact images used during the evaluation as referenced in the corresponding paper? Thank you for your support!