CUDA Setup failed despite GPU being available. Required library version not found:问题踩坑

详细报错如下，下载了最新的cuda122,bitsandbyts的release版本没有对应的so。
解决办法：自己编译一个

CUDA SETUP: CUDA runtime path found: /usr/local/cuda/lib64/libcudart.so
CUDA SETUP: Highest compute capability among GPUs detected: 8.6
CUDA SETUP: Detected CUDA version 122
CUDA SETUP: Required library version not found: libbitsandbytes_cuda122.so. Maybe you need to compile it from source?
CUDA SETUP: Defaulting to libbitsandbytes_cpu.so...
1
2
3
4
5

#如果使用了conda一定注意先卸载掉bitsandbytes
pip uninstall bitsandbytes
#明明科学上网了,但我这边没有git成功，可以下包到本地再解压
git clone https://github.com/TimDettmers/bitsandbytes.git 
cd bitsandbytes
make CUDA_VERSION=122 #注意替换为当前的CUDA_VERSION
python setup.py install
1
2
3
4
5
6
7

吐槽：看了一晚上，都说是换干净cuda版本或者环境的，费时费力。自己编译丰衣足食。