报了段错误“Segmentation Fault”后所作的调整尝试

[BUJIQI]

报错信息

[2024/10/22 17:02:36] ppocr INFO: During the training process, after the 0th iteration, an evaluation is run every 2000 iterations


--------------------------------------
C++ Traceback (most recent call last):
--------------------------------------
No stack trace in paddle, may be caused by external reasons.

----------------------
Error Message Summary:
----------------------
FatalError: `Segmentation fault` is detected by the operating system.
  [TimeInfo: *** Aborted at 1729587757 (unix time) try "date -d @1729587757" if you are using GNU date ***]
  [SignalInfo: *** SIGSEGV (@0x0) received by PID 1280915 (TID 0x7fcc7b4fd740) from PID 0 ***]

段错误 (核心已转储)

另外发现此时运行

import paddle
print(paddle.__version__)
paddle.utils.run_check()

时并无报错，然而在配置文件中将use_gpu设置为true后便会报段错误（使用cpu时可正常运行）

运行环境

OS：Linux Ubuntu22.04

解决过程

主要原因还是驱动与依赖库版本的不兼容问题

需要驱动与依赖库设定为以下版本（本人尝试有效）：

paddlepaddle-gpu==2.5.2
cuda==12.0
cudnn==8.9.1
openssl==1.1.0

以下为所用命令，仅作参考，需要按实际环境场景调整

下载cuda12.0

wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2204/x86_64/cuda-ubuntu2204.pin sudo mv cuda-ubuntu2204.pin /etc/apt/preferences.d/cuda-repository-pin-600 wget https://developer.download.nvidia.com/compute/cuda/12.0.0/local_installers/cuda-repo-ubuntu2204-12-0-local_12.0.0-525.60.13-1_amd64.deb sudo dpkg -i cuda-repo-ubuntu2204-12-0-local_12.0.0-525.60.13-1_amd64.deb sudo cp /var/cuda-repo-ubuntu2204-12-0-local/cuda-*-keyring.gpg /usr/share/keyrings/ sudo apt-get update sudo apt-get -y install cuda

下载paddle2.5.2

python3 -m pip install paddlepaddle-gpu==2.5.2.post120 -f https://www.paddlepaddle.org.cn/whl/linux/mkl/avx/stable.html

下载并安装了openssl 1.1.0

sudo dpkg -i libssl1.1_1.1.1f-1ubuntu2.23_amd64.deb

sudo ln -s /usr/lib/x86_64-linux-gnu/libcuda.so.525.60.13 /usr/lib/x86_64-linux-gnu/libcuda.so

ls -l /usr/lib/x86_64-linux-gnu/libcuda.so

sudo ldconfig

参考issue，还尝试了conda创建环境+pip install paddlepaddle-gpu的方法，不确定是否有效

install8.9.1的cudnn

conda install -c conda-forge cudnn=8.9.1

作指向

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/libcudnn所在路径