Closed pharringtonp19 closed 3 years ago
Thanks for pointing this out: torchvision
also requires 1.8.1
apparently. The issue appears linked to our torchsde dependency which clamps torch
to <1.8.0
. Pinging @lxuechen @patrick-kidger.
This is something I've been wanting to fix too.
@lxuechen, thoughts? I can see your comment on the matter here, but I can't duplicate the issue. Are you still able to repro it?
I can run pytest test_adjoint.py
on my machine without issues, in particular using python=3.8.8
, pytorch=1.8.0
, and cudatoolkit=11.1.74
, and the master version of torchsde
. Full environment below.
# Name Version Build Channel
_libgcc_mutex 0.1 main
attrs 20.3.0 pypi_0 pypi
blas 1.0 mkl
boltons 20.2.1 pypi_0 pypi
ca-certificates 2021.4.13 h06a4308_1
certifi 2020.12.5 py38h06a4308_0
cudatoolkit 11.1.74 h6bb024c_0 nvidia
iniconfig 1.1.1 pypi_0 pypi
intel-openmp 2021.2.0 h06a4308_610
ld_impl_linux-64 2.33.1 h53a641e_7
libffi 3.3 he6710b0_2
libgcc-ng 9.1.0 hdf63c60_0
libstdcxx-ng 9.1.0 hdf63c60_0
libuv 1.40.0 h7b6447c_0
mkl 2021.2.0 h06a4308_296
mkl-service 2.3.0 py38h27cfd23_1
mkl_fft 1.3.0 py38h42c9631_2
mkl_random 1.2.1 py38ha9443f7_2
ncurses 6.2 he6710b0_1
ninja 1.10.2 hff7bd54_1
numpy 1.19.5 pypi_0 pypi
openssl 1.1.1k h27cfd23_0
packaging 20.9 pypi_0 pypi
pip 21.0.1 py38h06a4308_0
pluggy 0.13.1 pypi_0 pypi
py 1.10.0 pypi_0 pypi
pyparsing 2.4.7 pypi_0 pypi
pytest 6.2.3 pypi_0 pypi
python 3.8.8 hdb3f193_5
pytorch 1.8.0 py3.8_cuda11.1_cudnn8.0.5_0 pytorch
readline 8.1 h27cfd23_0
scipy 1.5.4 pypi_0 pypi
setuptools 52.0.0 py38h06a4308_0
six 1.15.0 py38h06a4308_0
sqlite 3.35.4 hdfb4753_0
tk 8.6.10 hbc83047_0
toml 0.10.2 pypi_0 pypi
torchsde 0.2.5 dev_0 <develop>
trampoline 0.1.2 pypi_0 pypi
typing_extensions 3.7.4.3 pyha847dfd_0
wheel 0.36.2 pyhd3eb1b0_0
xz 5.2.5 h7b6447c_0
zlib 1.2.11 h7b6447c_3
Thanks for the heads-up. Will be working on checking this tonight.
Everything works as intended now, thanks!
Because pytorch currently gets downgraded to 1.7.1 when we install torchdyn, this creates an issue with torchtext
For those who might face a similar problem, the solution is to add the following line
!pip install torchtext==0.8.1