search

nelson1425 / EfficientAD

Unofficial implementation of EfficientAD https://arxiv.org/abs/2303.14535
https://arxiv.org/abs/2303.14535
Apache License 2.0
273 stars 72 forks source link

The image_size #24

Closed Sencc closed 12 months ago

Sencc commented 12 months ago

Hello, Thank you for your work. During training, does the image_size of the data need to be 256?

Sencc commented 12 months ago

When I change the image_size to 512,it got the error: RuntimeError: The size of tensor a (56) must match the size of tensor b (120) at non-singleton dimension 3

Sencc commented 12 months ago

Need to change the upsampling of the decoder

gzhxxx commented 6 months ago

Need to change the upsampling of the decoder

hi! could you please share the solution to train images with other sizes? thank you very much

ZHIZIHUABU commented 3 months ago

@Sencc 请问是修改common.py和benchmark.py中的decoder部分两个'56'吗

Sencc commented 2 months ago

@Sencc 请问是修改common.py和benchmark.py中的decoder部分两个'56'吗

You need to modify Upsample in get_autoencoder() of common.py.