IndexError: tensors used as indices must be long, byte or bool tensors

[mrtariqkhan]

Running the python decompressor command with cuda enabled or even not enabled will return this error. Using Python 3.10.9

IndexError: tensors used as indices must be long, byte or bool tensors

[pau1o-hs]

Hi there! It's working from my side, can you check your CUDA version?

[wasimjj]

@mrtariqkhan @pau1o-hs

I fixed that problem buy type casting to long.
" Xgnd = X[batch.long()].transpose(0, 1)
Ygnd = Ytxy[batch.long()].transpose(0, 1)
Qgnd = Qtxy[batch.long()].transpose(0, 1)
Qgnd_xfm = Qxfm[batch.long()].transpose(0, 1)"

error are gone and script are working fine now.

[pau1o-hs]

Awesome, thanks!