This is an official implementation of "ModernTCN: A Modern Pure Convolution Structure for General Time Series Analysis" (ICLR 2024 Spotlight), https://openreview.net/forum?id=vpJMJerXHU
MIT License
110
stars
12
forks
source link
Fixed incorrect argument: changed `dims` to `dim` #3
I'd like to express my gratitude to the author for making your invaluable work open-source.
But when employing structure reparameterization on ModernTCN, the following error arises:
TypeError: cat() received an invalid combination of arguments - got (list, dims=int), but expected one of:
* (tuple of Tensors tensors, int dim, *, Tensor out)
* (tuple of Tensors tensors, name dim, *, Tensor out)
The bug is located in the PaddingTwoEdge1d method of the ReparamLargeKernelConvclass. However, it erroneously employs the argument name dims instead of dim when invoking torch.cat, leading to this TypeError.
I'd like to express my gratitude to the author for making your invaluable work open-source.
But when employing structure reparameterization on ModernTCN, the following error arises:
The bug is located in the
PaddingTwoEdge1d
method of theReparamLargeKernelConv
class. However, it erroneously employs the argument namedims
instead ofdim
when invokingtorch.cat
, leading to this TypeError.