lcxsnow

> What do you mean by the text_encoder is frozen ? I thought you have to pre-train the DAMSM model, which place in the code said that ? requires_grad =...

> > In the paper, it is mentioned that the text_encoder will be fixed during the image synthesis training. However, in the provided code, the text_encoder is frozen, meaning it...

有同樣的問題,但缺的不同 Missing key(s) in state_dict: "block0.affine4.fc_gamma.linear1.weight", "block0.affine4.fc_gamma.linear1.bias", "block0.affine4.fc_gamma.linear2.weight", "block0.affine4.fc_gamma.linear2.bias", "block0.affine4.fc_beta.linear1.weight", "block0.affine4.fc_beta.linear1.bias", "block0.affine4.fc_beta.linear2.weight", "block0.affine4.fc_beta.linear2.bias", "block0.affine5.fc_gamma.linear1.weight", "block0.affine5.fc_gamma.linear1.bias", "block0.affine5.fc_gamma.linear2.weight", "block0.affine5.fc_gamma.linear2.bias", "block0.affine5.fc_beta.linear1.weight", "block0.affine5.fc_beta.linear1.bias", "block0.affine5.fc_beta.linear2.weight", "block0.affine5.fc_beta.linear2.bias", "block1.affine4.fc_gamma.linear1.weight", "block1.affine4.fc_gamma.linear1.bias", "block1.affine4.fc_gamma.linear2.weight", "block1.affine4.fc_gamma.linear2.bias", "block1.affine4.fc_beta.linear1.weight", "block1.affine4.fc_beta.linear1.bias", "block1.affine4.fc_beta.linear2.weight", "block1.affine4.fc_beta.linear2.bias", "block1.affine5.fc_gamma.linear1.weight",...