Hi, can the code support t2m performance evaluation on HumanML3D?
I notice there already has some code for t2m evaluation, but when i try to modify the test_h3d.py to support t2m evaluation, i find some files missed, e.g. dataloaders.h3d_eval_gt, corresponding to code below:
"self.gt_data = import(f"dataloaders.h3d_eval_gt", fromlist=["something"]).CustomDataset(self.args, "test")"
So can you update the t2m code for me to quickly reproduce the t2m metric in the paper?
Hi, can the code support t2m performance evaluation on HumanML3D?
I notice there already has some code for t2m evaluation, but when i try to modify the test_h3d.py to support t2m evaluation, i find some files missed, e.g. dataloaders.h3d_eval_gt, corresponding to code below:
"self.gt_data = import(f"dataloaders.h3d_eval_gt", fromlist=["something"]).CustomDataset(self.args, "test")"
So can you update the t2m code for me to quickly reproduce the t2m metric in the paper?