Open psy2013GitHub opened 1 year ago
i try to reproduce the result, but failed here is my config: gpu(v100): 2 batch size:2 accumulate_batches: 2
after training 500 epochs, the train epoch loss is about 0.01, but the generated image is bad.
can u guys paste your training metrics, such as epoch loss etc i think this should help others a lot. thanks.
i try to reproduce the result, but failed here is my config: gpu(v100): 2 batch size:2 accumulate_batches: 2
after training 500 epochs, the train epoch loss is about 0.01, but the generated image is bad.
can u guys paste your training metrics, such as epoch loss etc i think this should help others a lot. thanks.