首页 文章

tensorflow cudnn问题（cuda 7.5,367 driver cudnn 4）SOS

提问于
浏览 次
0

我在ubuntu 16.04上运行了一个gtx 980ti，通过deb文件安装了cuda 7.5，并将cudnn文件输入/usr/local/cuda-7.5 ....

我用以下内容编辑了我的bashrc文件：

export CUDA_HOME = / usr / local / cuda-7.5 export LD_LIBRARY_PATH = $ / lib64

PATH = $ / bin：$ 导出PATH

现在，当我运行模型时，我得到以下内容：

I tensorflow/stream_executor/dso_loader.cc:102] Couldn't open CUDA library libcublas.so. LD_LIBRARY_PATH: 
I tensorflow/stream_executor/cuda/cuda_blas.cc:2289] Unable to load cuBLAS DSO.
I tensorflow/stream_executor/dso_loader.cc:102] Couldn't open CUDA library libcudnn.so. LD_LIBRARY_PATH: 
I tensorflow/stream_executor/cuda/cuda_dnn.cc:2259] Unable to load cuDNN DSO
I tensorflow/stream_executor/dso_loader.cc:102] Couldn't open CUDA library libcufft.so. LD_LIBRARY_PATH: 
I tensorflow/stream_executor/cuda/cuda_fft.cc:343] Unable to load cuFFT DSO.
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:102] Couldn't open CUDA library libcurand.so. LD_LIBRARY_PATH: 
I tensorflow/stream_executor/cuda/cuda_rng.cc:333] Unable to load cuRAND DSO.

有谁看到我在这里做错了，或者知道如何解决它 . 所有建议都非常感谢:)

tensorflow cudnn

1 回答

  • 0

    您还必须安装包 nvidia-cuda-dev - 这将引入所有其他依赖项，例如libcublas，libcufft和libcurand，以及它们的普通 .so 符号链接 .

    在我的Ubuntu 16.04系统上，这些CUDA 7.5库都以 /usr/lib/x86_64-linux-gnu 结束，所以我甚至不需要 LD_LIBRARY_PATH 设置 .

    此外，我通过我的NVIDIA开发者帐户安装了两个软件包 libcudnn5_5.1.3-1+cuda7.5_amd64.deb 和 libcudnn5-dev_5.1.3-1+cuda7.5_amd64.deb .

    现在当我在TensorFlow 0.10配置环境中启动ipython时， LD_LIBRARY_PRELOAD 为空，我看到：

    In [1]: import tensorflow as tf
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcublas.so locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcudnn.so locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcufft.so locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcuda.so.1 locally
I tensorflow/stream_executor/dso_loader.cc:108] successfully opened CUDA library libcurand.so locally

    喜悦！

    回复于

相关问题