Conflicts installing torchaudio-cpu through conda #317
Description
🐛 Bug
Installing torchaudio-cpu through conda (python 3.7) fails with conflicts.
To Reproduce
Steps to reproduce the behavior:
- start clean continuumio/anaconda3 container with docker.
docker run --rm -i -t continuumio/anaconda3 /bin/bash
- Inside the container, attempt to install torchaudio-cpu
# conda install torchaudio-cpu -c pytorch
Collecting package metadata (current_repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: failed with repodata from current_repodata.json, will retry with next repodata source.
Collecting package metadata (repodata.json): done
Solving environment: failed with initial frozen solve. Retrying with flexible solve.
Solving environment: /
Found conflicts! Looking for incompatible packages.
This can take several minutes. Press CTRL-C to abort.
failed
UnsatisfiableError: The following specifications were found to be incompatible with each other:
Package tk conflicts for:
python=3.7 -> tk[version='>=8.6.7,<8.7.0a0|>=8.6.8,<8.7.0a0']
Package sqlite conflicts for:
python=3.7 -> sqlite[version='>=3.24.0,<4.0a0|>=3.25.2,<4.0a0|>=3.25.3,<4.0a0|>=3.26.0,<4.0a0|>=3.27.2,<4.0a0|>=3.29.0,<4.0a0']
Package scipy conflicts for:
torchaudio-cpu -> scipy
Package libffi conflicts for:
python=3.7 -> libffi[version='>=3.2.1,<4.0a0']
Package pytorch-cpu conflicts for:
torchaudio-cpu -> pytorch-cpu[version='>=1.1.0']
Package sox conflicts for:
torchaudio-cpu -> sox
Package zlib conflicts for:
python=3.7 -> zlib[version='>=1.2.11,<1.3.0a0']
Package readline conflicts for:
python=3.7 -> readline[version='>=7.0,<8.0a0']
Package libstdcxx-ng conflicts for:
python=3.7 -> libstdcxx-ng[version='>=7.2.0|>=7.3.0']
Package openssl conflicts for:
python=3.7 -> openssl[version='>=1.0.2o,<1.0.3a|>=1.0.2p,<1.0.3a|>=1.1.1a,<1.1.2a|>=1.1.1b,<1.1.2a|>=1.1.1c,<1.1.2a']
Package ncurses conflicts for:
python=3.7 -> ncurses[version='>=6.1,<7.0a0']
Package libgcc-ng conflicts for:
python=3.7 -> libgcc-ng[version='>=7.2.0|>=7.3.0']
Package pip conflicts for:
python=3.7 -> pip
Package librosa conflicts for:
torchaudio-cpu -> librosa[version='>=0.4.3']
Package xz conflicts for:
python=3.7 -> xz[version='>=5.2.4,<6.0a0']
Expected behavior
Expect torchaudio-cpu to install without conflicts and to install and work with pytorch-cpu.
Environment
docker container as indicated above
Collecting environment information...
PyTorch version: N/A
Is debug build: N/A
CUDA used to build PyTorch: N/A
OS: Debian GNU/Linux 10 (buster)
GCC version: Could not collect
CMake version: Could not collect
Python version: 3.7
Is CUDA available: N/A
CUDA runtime version: Could not collect
GPU models and configuration: Could not collect
Nvidia driver version: Could not collect
cuDNN version: Could not collect
Versions of relevant libraries:
[pip] numpy==1.17.2
[pip] numpydoc==0.9.1
[conda] blas 1.0 mkl
[conda] mkl 2019.4 243
[conda] mkl-service 2.3.0 py37he904b0f_0
[conda] mkl_fft 1.0.14 py37ha843d7b_0
[conda] mkl_random 1.1.0 py37hd6b4f25_0
Additional context
We build Docker images that include pytorch, torchaudio and torchvision. One of the images is CPU-only, so we currently install pytorch and torchvision through conda:
conda install -y pytorch-cpu torchvision-cpu -c pytorch
We would like to install torchaudio-cpu the same way but the errors above prevent that. Is the torchaudio-cpu package supported through conda?
Is a conda install of torchaudio CPU-only and compatible with pytorch-cpu?
Is there a cpu-only pip package that would be compatible with pytorch-cpu?
We have been building from source after the install of pytorch-cpu like this:
apt-get install -y sox libsox-dev libsox-fmt-all
pip install cffi
pip install git+git://github.com/pytorch/audio.git
This has been working, but the recent changes to Tensor.data to Tensor.data_ptr have broken that build because we are building against PyTorch 1.1.0 headers (#262)