in_batch_g2l_loss
jim4399266 opened this issue · comments
jim4399266 commented
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))
?
code for TCL: Vision-Language Pre-Training with Triple Contrastive Learning, CVPR 2022
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))
?