search

zhjohnchan / M3AE

[MICCAI-2022] This is the official implementation of Multi-Modal Masked Autoencoders for Medical Vision-and-Language Pre-Training.
115 stars 11 forks source link

out-of-bounds problem with the targets array in med-vqa2019 #14

Open muxuezzz opened 1 year ago

muxuezzz commented 1 year ago

I have an out-of-bounds problem with the targets array when performing a fine-tuning med-vqa task, and I observe that the max_label_size is set to 79. Is this because the med-vqa dataset has been adjusted or for some other reason, and does the value of max_label_size represent the type of labels that are desirable for the answer?

shed-e commented 1 year ago

I have an out-of-bounds problem with the targets array when performing a fine-tuning med-vqa task, and I observe that the max_label_size is set to 79. Is this because the med-vqa dataset has been adjusted or for some other reason, and does the value of max_label_size represent the type of labels that are desirable for the answer?

hi，I also encountered this problem，index 79 is out of bounds for dimension 1 with size 79。How did you solve it?

zhjohnchan commented 1 year ago

Hi all,

Please check if the dataset is the same one used in the codebase. Here is the link.

Best, Zhihong