Saving Images

I’m not certain I am saving the tensor to images correctly.

All I seem to get is noise - albeit it does have some structure to it. I tried 10,000 time steps and still just noise.

from torchvision.utils import save_image
img1 = images[0]  #torch.Size([3, 256, 256])
save_image(img1, 'img1.png’)

I’ve tried other ways such as:

from torchvision import transforms
im = transforms.ToPILImage()(images[0]).convert("RGB”)
im.save('img1.png’)

But I get results like this regardless:

[3DTOPO]

Thanks.

I just assumed it worked since there is an example.

[lucidrains]

ohh the example is for training, and there are no publicly available trained model yet