Open sclzsx opened 4 years ago
I surface the same problem
Please delete the dir cc_attention/build. CC Attention module will compile using your local installed CUDA 10.2.
@ygjwd12345, @sclzsx how was this problem solved? I am having the same issue.
I run the code branch CCNet-pytorch1.1, my environments are: pytorch 1.4 torchvision 0.5.0 cuda 10.2 python 3.6 However, this error happend: root@3178500507b3:/workspace/CODE/CCNet-pytorch-1.1# python3 train.py Detected CUDA files, patching ldflags Emitting ninja build file /workspace/CODE/CCNet-pytorch-1.1/cc_attention/build/build.ninja... Building extension module rcca... ninja: no work to do. Loading extension module rcca... Traceback (most recent call last): File "train.py", line 17, in
import networks
File "/workspace/CODE/CCNet-pytorch-1.1/networks/init.py", line 3, in
import networks.ccnet
File "/workspace/CODE/CCNet-pytorch-1.1/networks/ccnet.py", line 13, in
from cc_attention import CrissCrossAttention
File "/workspace/CODE/CCNet-pytorch-1.1/cc_attention/init.py", line 1, in
from .functions import CrissCrossAttention, ca_weight, ca_map
File "/workspace/CODE/CCNet-pytorch-1.1/cc_attention/functions.py", line 24, in
extra_cuda_cflags=["--expt-extended-lambda"])
File "/usr/local/lib/python3.6/dist-packages/torch/utils/cpp_extension.py", line 680, in load
is_python_module)
File "/usr/local/lib/python3.6/dist-packages/torch/utils/cpp_extension.py", line 877, in _jit_compile
return _import_module_from_library(name, build_directory, is_python_module)
File "/usr/local/lib/python3.6/dist-packages/torch/utils/cpp_extension.py", line 1088, in _import_module_from_library
return imp.load_module(module_name, file, path, description)
File "/usr/lib/python3.6/imp.py", line 243, in load_module
return load_dynamic(name, filename, file)
File "/usr/lib/python3.6/imp.py", line 343, in load_dynamic
return _load(spec)
ImportError: libcudart.so.9.2: cannot open shared object file: No such file or directory
Could you help me what should I do?