search

geyuying / PF-AFN

Official code for "Parser-Free Virtual Try-on via Distilling Appearance Flows", CVPR 2021.
545 stars 138 forks source link

train_PBAFN_stage1 failed #40

Closed crinai closed 3 years ago

crinai commented 3 years ago

I use this wonderful project with another CV project,whitch require the core ENV bellow: torch 1.8.0+cu111
torchaudio 0.8.0
torchvision 0.9.0+cu111 cuda 0.0.1
cudatoolkit 11.1.1
cupy-cuda111 8.6.0 cupy 9.0.0
cupy-cuda111 8.6.0 The script [scripts/train_PBAFN_stage1.sh] return this error " RuntimeError: Given input size: (512x2x1). Calculated output size: (512x1x0). Output size is too small", can anyone help me,thank you!

cloudy4next commented 3 years ago

you need to downgrade pytorch to 1.1.o else go to 2dmaxpool of 1.8 and modify

crinai commented 3 years ago

you need to downgrade pytorch to 1.1.o else go to 2dmaxpool of 1.8 and modify than you, the new GPUs doesn't support CUDA9.2，so I can't downgrade the torch version to 1.1.0， I will try modify the pool later。

Lipurple commented 1 year ago

hello, I meet the same question and I wonder if you have solved this problem