Closed zeinsh closed 1 month ago
Hi! I tried to reproduce the final scores in your benchmark using eval_from_metrics.py starting from raw metrics.
I noticed that you swapped the numbers in motion quality and visual quality in https://evalcrafter.github.io and https://huggingface.co/spaces/AILab-CVC/EvalCrafter.
Am I correct?
Yes, you're correct. It's our mistake. Many thanks for pointing out the problem! We will fix this right away.
Hi! I tried to reproduce the final scores in your benchmark using eval_from_metrics.py starting from raw metrics.
I noticed that you swapped the numbers in motion quality and visual quality in https://evalcrafter.github.io and https://huggingface.co/spaces/AILab-CVC/EvalCrafter.
Am I correct?