question-mark
Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

Torchaudio Installation Jetson Aarch System

See original GitHub issue

🐛 Bug

For the Jetson Series there is already Torch available but Torchaudio isnt.

To Reproduce

Try install from source to an Jetson Device. Steps to reproduce the behavior:

  1. Clone repository
  2. sudo packaging/build_from_source.sh => Crash of the sh script.

Workaround

This crash is a result of the libraries config.guess that doesnt know the aarch architecture. This can be bypassed by replacing midinstall the config.guess with a more current one like this https://raw.githubusercontent.com/ropensci/git2r/master/tools/config.guess

Replace the ones in those folders => /tmp/torchaudio-deps/lame-3.99.5/config.guess /tmp/torchaudio-deps/libmad-0.15.1b/config.guess

Also Edit the Script sudo packaging/build_from_source.sh to not override those changes again.

Now it should install.

Expected behavior

Normal installation of the required dependencies.

Environment

Collecting environment information...
PyTorch version: 1.4.0a0+7f73f1d
Is debug build: No
CUDA used to build PyTorch: 10.2

OS: Ubuntu 18.04.4 LTS
GCC version: (Ubuntu/Linaro 7.5.0-3ubuntu1~18.04) 7.5.0
CMake version: version 3.10.2

Python version: 3.6
Is CUDA available: Yes
CUDA runtime version: 10.2.89
GPU models and configuration: Could not collect
Nvidia driver version: Could not collect
cuDNN version: Probably one of the following:
/usr/lib/aarch64-linux-gnu/libcudnn.so.8.0.0
/usr/lib/aarch64-linux-gnu/libcudnn_adv_infer.so.8.0.0
/usr/lib/aarch64-linux-gnu/libcudnn_adv_train.so.8.0.0
/usr/lib/aarch64-linux-gnu/libcudnn_cnn_infer.so.8.0.0
/usr/lib/aarch64-linux-gnu/libcudnn_cnn_train.so.8.0.0
/usr/lib/aarch64-linux-gnu/libcudnn_etc.so.8.0.0
/usr/lib/aarch64-linux-gnu/libcudnn_ops_infer.so.8.0.0
/usr/lib/aarch64-linux-gnu/libcudnn_ops_train.so.8.0.0

Versions of relevant libraries:
[pip3] numpy==1.18.4
[pip3] torch==1.5.0
[pip3] torchaudio==0.6.0a0+313f4f5
[pip3] torchvision==0.6.0a0+b68adcf
[conda] Could not collect

Issue Analytics

  • State:closed
  • Created 3 years ago
  • Comments:11 (9 by maintainers)

github_iconTop GitHub Comments

1reaction
mthrokcommented, May 17, 2021

@discort

Note that the libsox installed from apt at the moment is 14.4.2+git20190427-2, which is something we do not support as it has an issue with in-memory decoding and causes un-deterministic behavior if you use file-like object for loading data.

1reaction
mthrokcommented, May 26, 2020

Hi @ark626

Thanks for opening the issue and sharing the workaround. I am open to update the dependency installation script to support aarch64, however, CI system we use (CircleCI) does not seem to provide aarch64 environment so we cannot guarantee the successful build all the time. (Although if the issue happens only in dependency it should work) Therefore I will keep this issue open as reference point for future audience who seek instruction for building torchaudio in aarch64.

Read more comments on GitHub >

github_iconTop Results From Across the Web

Install torchaudio and torchvison from wheel - Jetson AGX Orin
Install torchaudio and torchvison from wheel - Jetson AGX Orin - NVIDIA Developer Forums.
Read more >
Install PyTorch on Jetson Nano - Q-engineering
This page will guide you through the installation of PyTorch, TorchVision, LibTorch and Caffe2 on a Jetson Nano. PyTorch is a software library...
Read more >
AUR (en) - python-torchaudio - Arch Linux
Package Base: python-torchaudio. Description: Data manipulation and transformation for audio signal processing, powered by PyTorch.
Read more >
Running PyTorch Models on Jetson Nano
You can also use the docker image to run PyTorch models because the image has PyTorch, torchvision and torchaudio installed:.
Read more >
pip install torch 1.6.0 python Code Example
conda install pytorch==1.8.0 torchvision==0.9.0 torchaudio==0.8.0 cudatoolkit=11.1 -c pytorch -c conda-forge. 6. ​. 7. # CPU Only. 8.
Read more >

github_iconTop Related Medium Post

No results found

github_iconTop Related StackOverflow Question

No results found

github_iconTroubleshoot Live Code

Lightrun enables developers to add logs, metrics and snapshots to live code - no restarts or redeploys required.
Start Free

github_iconTop Related Reddit Thread

No results found

github_iconTop Related Hackernoon Post

No results found

github_iconTop Related Tweet

No results found

github_iconTop Related Dev.to Post

No results found

github_iconTop Related Hashnode Post

No results found