search

k2-fsa / fast_rnnt

A torch implementation of a recursion which turns out to be useful for RNN-T.
Other
136 stars 22 forks source link

#error -- unsupported GUN version ! gcc version later than 5.3 are not supported! #9

Closed Butterfly-c closed 1 year ago

Butterfly-c commented 2 years ago

I ran into the following problems，when i run the setup.py:

/opt/lib/cuda-8.0/bin/..//include/host_config.h:115:2 #error -- unsupported GUN version ! gcc version later than 5.3 are not supported!

the environment is : gcc 5.4.0 pytorch 1.7.1 cuda-10.2 python 3.7 (There are several different versions of cuda in the environment.)

when I set the following configs in CMakeList.txt, I still get the same error

set(CUDA_TOOLKIT_ROOT_DIR /opt/lib/cuda-10.2) set(CMAKE_C_COMPILER /usr/bin/gcc) (gcc -v 4.8.5) set(CMAKE_CXX_COMPILER /usr/bin/++)

pkufool commented 2 years ago

Try setting cuda-10.2 as follows：

export CUDA_HOME=/opt/lib/cuda-10.2
export PATH=$CUDA_HOME/bin:$PATH
export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH

export CUDA_TOOLKIT_ROOT_DIR=$CUDA_HOME
export CUDA_TOOLKIT_ROOT=$CUDA_HOME
export CUDA_BIN_PATH=$CUDA_HOME
export CUDA_PATH=$CUDA_HOME
export CUDA_INC_PATH=$CUDA_HOME/targets/x86_64-linux

And the fast_rnnt requires the c++14 standard, you might need to use a newer gcc version.

Butterfly-c commented 2 years ago

Thanks! The environment is compiled successfully.