about faceir test results

[HanBingfeng]

Hi, thanks for offering such good work. I'm facing a test problem, the test result is a totally black image. Is there any configuration problem for faceir task?

[HanBingfeng]

Chopping size/stride: 1024/1024 Building the diffusion model with length: 4... No module 'xformers'. Proceeding without it. python/lib/python3.8/site-packages/torch/functional.py:478: UserWarning: torch.meshgrid: in an upcoming release, it will be required to pass the indexing argument. (Triggered internally at ../aten/src/ATen/native/TensorShape.cpp:2895.) return _VF.meshgrid(tensors, **kwargs) # type: ignore[attr-defined] Loading Diffusion model from weights/resshift_faceir_s4.pth... Loading AutoEncoder model from weights/ffhq512_vq_f8_dim8_face.pth... Working with z of shape (1, 8, 64, 64) = 32768 dimensions. Processing done, enjoy the results in diffusion/res

[zsyOAOA]

Could you successfully run the colab?

[HanBingfeng]

Could you successfully run the colab?

colab got the same result. The input image has black surrounded backgroud, may be the reason why.

[zsyOAOA]

Could you share the testing image?

[zsyOAOA]

Maybe you do not upload successfully. By the way, could you successfully run the examples in the colob?

[HanBingfeng]

Maybe you do not upload successfully. By the way, could you successfully run the examples in the colob?

colab examples are good

[HanBingfeng]

Could you share the testing image?

[zsyOAOA]

That's because this image is not aligned. So far, I haven't added the script to deal with the unaligned case. I will add the script later.

[HanBingfeng]

:） nice buddy