search

tsujuifu / pytorch_violet

A PyTorch implementation of VIOLET
137 stars 6 forks source link

I found some wrong places, can you see if it's right? #4

Closed bubbliiiing closed 2 years ago

bubbliiiing commented 2 years ago

In main_pretrain.py, line 165: p = (1+_h*_w)*i_t + i_h*_w + i_w I think it should plus 1. p = (1+_h*_w)*i_t + i_h*_w + i_w + 1 Because your first position is for the separator.

bubbliiiing commented 2 years ago

Already stared your code, thanks for you contribution.

tsujuifu commented 2 years ago

Yep, thanks for pointing it out! We have a [CLS_V] token in front of each video frame, so it should +1.

bubbliiiing commented 2 years ago

OK, thank you