## 🚀 Feature It would be nice if we can apply multiple filter in `F.lfilter` ## Motivation https://github.com/pytorch/audio/issues/1517#issuecomment-845125829 ## Pitch Allow [the `a_coeffs` and `b_coefds` parameters of `F.lfilter`](https://github.com/pytorch/audio/blob/14dd917ec60fa69ce3f7c6e3f2eaf520e67928b5/torchaudio/functional/filtering.py#L879-L883) to be 2D Tensors. ## Alternatives Currently one has to use for-loop for achieving this. ## Additional context There is also a feature proposal for extending `lfilter` for batch processing. https://github.com/pytorch/audio/issues/1476