FactorCL

FactorCL copied to clipboard

Thank you for your great work! I have a question about a conditional loss in FactorCL-SSL case. In IRFL_model.py [Line 311](https://github.com/pliang279/FactorCL/blob/0f9f7403b3764ba9faf9ee3247358bdd719f688c/IRFL_model.py#L311), conditional CLUB loss is computed as follows: ```python self.club_x1x2_cond(torch.cat([self.linears_club_x1x2_cond[0](x1_embed),...

jangsoohyuk

Thanks for your exciting work! But what's the difference between self.infonce_x1y(self.linears_infonce_x1y(x1_embed), y) and self.infonce_x1x2_cond(torch.cat([self.linears_infonce_x1x2_cond[0](x1_embed), y_ohe], dim=1), torch.cat([self.linears_infonce_x1x2_cond[1](x2_embed), y_ohe], dim=1)). I mean why we need to use y but y_ohe?

PANPANKK