Closed Sutongtong233 closed 2 years ago
I think you can follow the code in fused_act.py to modify the code of upfirdn2d.py https://github.com/williamyang1991/DualStyleGAN/blob/fcc547c7728baffb8ddd0ce21fa0d7252d761786/model/stylegan/op/fused_act.py#L29-L31
Thanks, the issue has been solved.
------------------ 原始邮件 ------------------ 发件人: "williamyang1991/DualStyleGAN" @.>; 发送时间: 2022年8月5日(星期五) 凌晨0:29 @.>; @.**@.>; 主题: Re: [williamyang1991/DualStyleGAN] input must be contiguous when autograd (Issue #39)
I think you can follow the code in fused_act.py to modify the code of upfirdn2d.py https://github.com/williamyang1991/DualStyleGAN/blob/fcc547c7728baffb8ddd0ce21fa0d7252d761786/model/stylegan/op/fused_act.py#L29-L31
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>
When I finetune DualStyleGAN, when
d_r1_loss(real_pred, real_img)
, it call 'upfirdn2d.py', and RunError:I have tried to add .contiguous() after the input, but it doesn't work. What should I do? Thanks in advance.