Recently When I prepared dl3dv according to the project workflow, I found that I couldn't perform evaluation (it would stop at Testing: | | 0/? [00:00<?, ?it/s]). However, when I used dl3dv_tiny, it worked pretty well.
Also when I evaluate using the _re10ksubset provided by pixelsplat (https://drive.google.com/drive/folders/1joiezNCyQK2BvWMnfwHJpm2V77c7iYGe), it also stops at Testing: | | 0/? [00:00<?, ?it/s]. I checked and found that the dataset format is correct. I would like to humbly ask where the error might have occurred.
I would appreciate a lot if you could provide a detailed process for converting dl3dv, or how to check whether the converted torch chunk is valid? Looking forward to your reply, it is very very important to me.
Thanks for your great work!
Recently When I prepared dl3dv according to the project workflow, I found that I couldn't perform evaluation (it would stop at Testing: | | 0/? [00:00<?, ?it/s]). However, when I used dl3dv_tiny, it worked pretty well.
Also when I evaluate using the _re10ksubset provided by pixelsplat (https://drive.google.com/drive/folders/1joiezNCyQK2BvWMnfwHJpm2V77c7iYGe), it also stops at Testing: | | 0/? [00:00<?, ?it/s]. I checked and found that the dataset format is correct. I would like to humbly ask where the error might have occurred.
I would appreciate a lot if you could provide a detailed process for converting dl3dv, or how to check whether the converted torch chunk is valid? Looking forward to your reply, it is very very important to me.