Support for Raspberry pi?

[LearnedVector]

❓ Questions and Help

Please note that this issue tracker is not a help form and this issue will be closed.

We have a set of listed resources available on the website. Our primary means of support is our discussion forum:

• Discussion Forum

Hello, is support for raspberry pi planned?

[vincentqb]

Since linux is supported on raspberry pi, have you tried compiling from source or installing using pip?

[LearnedVector]

hello @vincentqb, yes i tried installing from source but running into this error RuntimeError: fft: ATen not compiled with MKL support when using using torchaudio.transforms.MFCC. Seems to be an MKL issue? Since the pi does not have an intel processor, is there anyway around this?

here is the entire stacktrace for reference

Exception in thread Thread-2:
Traceback (most recent call last):
  File "/usr/lib/python3.7/threading.py", line 917, in _bootstrap_inner
    self.run()
  File "/usr/lib/python3.7/threading.py", line 865, in run
    self._target(*self._args, **self._kwargs)
  File "engine.py", line 89, in inference_loop
    action(self.predict(self.audio_q))
  File "engine.py", line 71, in predict
    mfcc = self.featurizer(waveform).transpose(1, 2).transpose(0, 1)
  File "/home/pi/.local/lib/python3.7/site-packages/torch/nn/modules/module.py", line 562, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/pi/.local/lib/python3.7/site-packages/torchaudio-0.6.0a0+00d3820-py3.7-linux-armv7l.egg/torchaudio/transforms.py", line 490, in forward
    mel_specgram = self.MelSpectrogram(waveform)
  File "/home/pi/.local/lib/python3.7/site-packages/torch/nn/modules/module.py", line 562, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/pi/.local/lib/python3.7/site-packages/torchaudio-0.6.0a0+00d3820-py3.7-linux-armv7l.egg/torchaudio/transforms.py", line 422, in forward
    specgram = self.spectrogram(waveform)
  File "/home/pi/.local/lib/python3.7/site-packages/torch/nn/modules/module.py", line 562, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/pi/.local/lib/python3.7/site-packages/torchaudio-0.6.0a0+00d3820-py3.7-linux-armv7l.egg/torchaudio/transforms.py", line 84, in forward
    self.win_length, self.power, self.normalized)
  File "/home/pi/.local/lib/python3.7/site-packages/torchaudio-0.6.0a0+00d3820-py3.7-linux-armv7l.egg/torchaudio/functional.py", line 249, in spectrogram
    waveform, n_fft, hop_length, win_length, window, True, "reflect", False, True
  File "/home/pi/.local/lib/python3.7/site-packages/torch/functional.py", line 439, in stft
    return _VF.stft(input, n_fft, hop_length, win_length, window, normalized, onesided)
RuntimeError: fft: ATen not compiled with MKL support

[vincentqb]

I see, so pytorch doesn't support it. This means torchaudio can't either. I'll close this since this is a duplicate of #408.