Open kris-yangjs opened 3 years ago
Can you try to install pytorch 1.1.0 again and use 'pip install cupy==6.0.0' instead of 'conda install cupy'?
@geyuying OK, I'll try, thanks for your reply
Have you found a solution to this? Installing with the exact same environment as specified gives me the below error:
**Traceback (most recent call last):
File "/home/bonelloj/.conda/envs/tryon/lib/python3.6/site-packages/cupy/init.py", line 20, in
The above exception was the direct cause of the following exception:
Traceback (most recent call last):
File "test.py", line 5, in
If you are using wheel distribution (cupy-cudaXX), make sure that the version of CuPy you installed matches with the version of CUDA on your host. Also, confirm that only one CuPy package is installed: $ pip freeze
If you are building CuPy from source, please check your environment, uninstall CuPy and reinstall it with: $ pip install cupy --no-cache-dir -vvvv
Check the Installation Guide for details: https://docs.cupy.dev/en/latest/install.html
original error: libcuda.so.1: cannot open shared object file: No such file or directory**
Thanks for your amazing work. I tried the environment installation according to your readme.md. But I found a question: pythorch 1.1.0 is based on cudnn7.5.1, but cupy6.0 is based on a newer version of cudnn, so after the command 'conda install cupy', the version of cudnn changed, and pytorch could not be used anymore.