uta-smile / TCL

code for TCL: Vision-Language Pre-Training with Triple Contrastive Learning, CVPR 2022

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

in_batch_g2l_loss

jim4399266 opened this issue · comments

The line 249 in models
u_p = (temp_mask * u_p) + (10000. * (1-temp_mask)) may should be
u_p = (temp_mask * u_p) - (10000. * (1-temp_mask)) ?