I was using your gan model for other images that are high resolution (512,512,3) so when executing this line of code sequentially {train_dataset = tf.data.Dataset.from_tensor_slices(training_data).shuffle(BUFFER_SIZE).batch(BATCH_SIZE)}........ i got the error..... Cannot create a tensor proto whose content is larger than 2GB.

Question

I was using your gan model for other images that are high resolution (512,512,3) so when executing this line of code sequentially {train_dataset = tf.data.Dataset.from_tensor_slices(training_data).shuffle(BUFFER_SIZE).batch(BATCH_SIZE)}........ i got the error..... Cannot create a tensor proto whose content is larger than 2GB.

Closed this issue 4 years ago · 2 comments

Answer 1 · 2020-03-20T23:58:42.000Z

This is really meant just as a starter. You will likely have to modify the code to support really larger sizes. You might want to use StyleGAN.

Answer 2 · 2020-03-21T09:44:16.000Z

Thanks a lot for your reply > :)

…

On Sat, Mar 21, 2020 at 4:58 AM Jeff Heaton ***@***.***> wrote: This is really meant just as a starter. You will likely have to modify the code to support really larger sizes. You might want to use StyleGAN. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#52 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ANYACS452Q65VHWTIENYBJTRIP7L7ANCNFSM4LNA6V4Q> .