Running a.cuda() then code will get stuck in pytorch-0.2

When running the code in pytorch-0.2:

import torch
a = torch.LongTensor(2)
a = a.cuda()

I got following result:

[torch.LongTensor of size 2]


Then the code got stuck, no return.

I’m running Ubuntu 18.04, and using conda install pytorch=0.2 get my experimental environment as following:

# Name                    Version                   Build  Channel
_libgcc_mutex             0.1                        main    defaults
blas                      1.0                         mkl    defaults
ca-certificates           2020.12.8            h06a4308_0    defaults
certifi                   2020.12.5        py36h06a4308_0    defaults
cffi                      1.14.4           py36h261ae71_0    defaults
cudatoolkit               8.0                           3    <unknown>
cudnn                     6.0.21                cuda8.0_0    defaults
intel-openmp              2019.4                      243    defaults
ld_impl_linux-64          2.33.1               h53a641e_7    defaults
libedit                   3.1.20191231         h14c3975_1    defaults
libffi                    3.3                  he6710b0_2    defaults
libgcc                    7.2.0                h69d50b8_2    defaults
libgcc-ng                 9.1.0                hdf63c60_0    defaults
libgfortran-ng            7.3.0                hdf63c60_0    defaults
libstdcxx-ng              9.1.0                hdf63c60_0    defaults
mkl                       2018.0.3                      1    defaults
mkl_fft                   1.0.6            py36h7dd41cf_0    defaults
mkl_random                1.0.1            py36h4414c95_1    defaults
nccl                      1.3.4                 cuda8.0_1    defaults
ncurses                   6.2                  he6710b0_1    defaults
numpy                     1.15.4           py36h1d66e8a_0    defaults
numpy-base                1.15.4           py36h81de0dd_0    defaults
openssl                   1.1.1i               h27cfd23_0    defaults
pip                       20.3.3           py36h06a4308_0    defaults
pycparser                 2.20                       py_2    defaults
python                    3.6.12               hcff3b4d_2    defaults
pytorch                   0.2.0           py36cuda8.0cudnn6.0_0    defaults
readline                  8.0                  h7b6447c_0    defaults
setuptools                51.0.0           py36h06a4308_2    defaults
sqlite                    3.33.0               h62c20be_0    defaults
tbb                       2020.3               hfd86e86_0    defaults
tbb4py                    2020.3           py36hfd86e86_0    defaults
tk                        8.6.10               hbc83047_0    defaults
wheel                     0.36.2             pyhd3eb1b0_0    defaults
xz                        5.2.5                h7b6447c_0    defaults
zlib                      1.2.11               h7b6447c_3    defaults

I need to finish my research in this env, Any help would be much appreciated!

Any way you can update to a newer version? 0.2 is ancient.

I need to reproduce the paper mos-lstm in pytorch-0.2. Otherwise, I need further tuning to match the original results. It’s too hard. Do you have any suggestions? Thank you !