In the preprocessing steps when subsampling data, there was a buggy edge case when there was only one point in the batch that had to be removed.
Up to now, the point was not removed (side effect of trying to force the output to be a Tensor when only one value was kept)
@MichelDaab or @gliegard can you have a look as @CharlesGaydon is not available?
Thanks
In the preprocessing steps when subsampling data, there was a buggy edge case when there was only one point in the batch that had to be removed. Up to now, the point was not removed (side effect of trying to force the output to be a Tensor when only one value was kept)
@MichelDaab or @gliegard can you have a look as @CharlesGaydon is not available? Thanks