High memory usage for large model

[NachAlvarez]

Hello there,
first of all thanks for the implementation of the EEND system in pytorch. I am trying to run the large model with simulated data and it turns out that the model occupies more than 20GB in GPU!
is this normal or could it be due to some implementation error?

[Xflick]

I am able to run the code on 1080ti/2080ti (less than 11GB memory) under the configuration EEND_PyTorch/conf/large/train.yaml. Does your configuration use a larger chunk/model/batch size? Or you may just use small batch with gradient accumulation, which will not hurt the performance.