How to save MNIST as jpg files

Hi, all

How to save MNIST as .jpg format ?
Is it possible with
from torchvision.utils import save_image?

(I use default dataloader from pytorch.)

Since the torchvision.datasets.MNIST dataset returns PIL Images, if you don’t specify a transformation, the easiest way would be to directly save them:

dataset = datasets.MNIST(root=PATH)

for idx, (img, _) in enumerate(dataset):
    img.save('{:05d}.jpg'.format(idx))

Sorry ptrblck,

How I can same this images. I use this command it save it but it is nothing in it just white.

real_batch = next(iter(trainloader))
fig=plt.figure()
plt.figure(figsize=(8,8))
plt.axis("off")
plt.title("Training Images")
plt.imshow(np.transpose(vutils.make_grid(real_batch[0].to(device)[:64], padding=2, normalize=True).cpu(),(1,2,0)))
fig.savefig(os.path.join(root_dirDurringTraining13+str(1)+str(2))+'Images.jpg',dpi=fig.dpi)

would you please guide me how to save it ?

Is plt.imshow showing the expected figure and is only savefig failing or do you also get a while output in imshow?
I would recommend to check the min and max values of your image array, as matplotlib could try to normalize the images thus potentially removing information.

it is solved I should used plt=plt.figure
many thanks