RelTR Training failed on custom data set

Hi, so i have generated my own dataset, which is basically a filtered VG set on some classes, i did not filter the relations and my data follows the same format as the one indicated in the files. However when i start to train, the training immediately crashes while throwing this error. I am really interested in your work.

Jun 18 '24 08:06 ikram-md

Hi, thanks for your interest in our work! This error usually happens when an incorrect indexing operation is performed, including but not limited to index an empty tensor, the index is out of range... I recommend you make a try expect debug to check your indexing operation.

Jun 24 '24 21:06 yrcong

Hi, so i have generated my own dataset, which is basically a filtered VG set on some classes, i did not filter the relations and my data follows the same format as the one indicated in the files. However when i start to train, the training immediately crashes while throwing this error. I am really interested in your work.

Can I reference your project? I'm fine-tuning it.

Jan 09 '25 15:01 trangiahuy8444

Hi

Hi, so i have generated my own dataset, which is basically a filtered VG set on some classes, i did not filter the relations and my data follows the same format as the one indicated in the files. However when i start to train, the training immediately crashes while throwing this error. I am really interested in your work.

Can I reference your project? I'm fine-tuning it.

Hi I am probably late on this but yeah of course

May 29 '25 22:05 ikram-md