Training error caused by inplace operations

[LeoDarcy]

XCube/xcube/modules/autoencoding/base_encoder.py

Line 63 in d0ee1c2

input_normal.jdata /= (input_normal.jdata.norm(dim=1, keepdim=True) + 1e-6)

It leads to "RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation" during training.

[LeoDarcy]

After making modifications, I still encounter similar errors. May I ask if there are any other code that need to be modified? Thank you very much.