标签: cudnn

如果我问了一个愚蠢的问题，我很抱歉。我是 CUDA 的新手。使用 Runfile 方法安装了 CUDA 10.1 并根据 Nvidia 说明进行了以下导出：

export LD_LIBRARY_PATH=/usr/local/cuda-10.1/lib64:$LD_LIBRARY_PATH
export PATH=$PATH:/usr/local/cuda-10.1/bin

然后我尝试添加 cuDNN 库。但是我在下面找到了两个 CUDA 文件夹/use/local：

cuda
cuda-10.1

我nvcc -V在两个文件夹中运行，它们都是 10.1 版。所以现在我有两个问题：

1. 我应该cuDNN库复制到cuda/include或cuda-10.1/include或两者兼而有之？

2. 为什么我得到两个文件夹？似乎它们包含完全相同的文件。我应该删除其中之一以使事情干净吗？

我正在尝试使用 cuda 10.2 构建 opencv'。当执行以下命令时：

cmake -DCMAKE_BUILD_TYPE=RELEASE \
-DOPENCV_EXTRA_MODULES_PATH=../../opencv_contrib/modules \
-DWITH_TBB=ON -DWITH_CUDA=ON \
-DBUILD_opencv_cudacodec=OFF \
-DENABLE_FAST_MATH=1 \
-DWITH_CUBLAS=1 \
 -DWITH_V4L=ON  \
 -DWITH_OPENGL=ON  \
 -DWITH_GSTREAMER=ON  \
 -DOPENCV_GENERATE_PKGCONFIG=ON  \
 -DOPENCV_ENABLE_NONFREE=ON \
 -DBUILD_EXAMPLES=TRUE \
 -DBUILD_PERF_TESTS=FALSE \
 -DEBUILD_TESTS=FALSE ../../opencv

我有以下问题：

Could NOT find CUDNN (missing:  CUDNN_LIBRARY CUDNN_INCLUDE_DIR) (Required is at least version "6")

当然我已经安装了cuda 10.2对应的cudnn7，我安装测试通过了。

有人可以帮忙吗？

我正在编写一个简单的多流 CUDA 应用程序。cuda-streams以下是我创建,cublas-handle和的代码部分cudnn-handle：

cudaSetDevice(0);

int num_streams = 1;

cudaStream_t streams[num_streams];
cudnnHandle_t mCudnnHandle[num_streams];
cublasHandle_t mCublasHandle[num_streams];

for (int ii = 0; ii < num_streams; ii++) {
    cudaStreamCreateWithFlags(&streams[ii], cudaStreamNonBlocking);
    cublasCreate(&mCublasHandle[ii]);
    cublasSetStream(mCublasHandle[ii], streams[ii]);
    cudnnCreate(&mCudnnHandle[ii]);
    cudnnSetStream(mCudnnHandle[ii], streams[ii]);
}

现在，我的流计数为 1。但是当我使用 Nvidia Visual Profiler 分析上述应用程序的可执行文件时，我得到以下信息：

对于我创建的每个流，它会另外创建 4 个流。我用 进行了测试num_streams = 8，它在分析器中显示了 40 个流。它在我心中提出了以下问题：

1. 内部是否cudnn创建流？如果是，那为什么？
2. 如果它隐式创建流那么使用它的方式是什么？
3. 在这种情况下，显式创建流有什么意义吗？

我正在按照官方页面上的说明和“验证您的安装”步骤安装 Tensorflow 。

>>> sess = tf.Session()
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE3 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.1 instructions, but these are available on your machine and could speed up CPU computations.
W tensorflow/core/platform/cpu_feature_guard.cc:45] The TensorFlow library wasn't compiled to use SSE4.2 instructions, but these are available on your machine and could speed up CPU …

我使用谷歌合作实验室训练可能的数据集。我将数据集上传到了Google云端硬盘，并从Google Colab调用了该数据集。但是运行train.py脚本意味着出现以下错误。更确切地说，我跑：

!python3  /content/drive/tensorflow1/models/research/object_detection/train.py --logtostderr --train_dir=/content/drive/tensorflow1/models/research/object_detection/training/ --pipeline_config_path=/content/drive/tensorflow1/models/research/object_detection/training/faster_rcnn_inception_v2_pets.config

我得到这些错误：

Traceback (most recent call last):
  File "/usr/local/lib/python3.6/dist-packages/tensorflow/python/pywrap_tensorflow.py", line 58, in <module>
    from tensorflow.python.pywrap_tensorflow_internal import *
  File "/usr/local/lib/python3.6/dist-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 28, in <module>
    _pywrap_tensorflow_internal = swig_import_helper()
  File "/usr/local/lib/python3.6/dist-packages/tensorflow/python/pywrap_tensorflow_internal.py", line 24, in swig_import_helper
    _mod = imp.load_module('_pywrap_tensorflow_internal', fp, pathname, 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: libcublas.so.9.0: cannot open shared object file: No such file or directory

During handling of the above exception, …

当我通过 Conda 安装 tensorflow-gpu 时；它给了我以下输出：

conda install tensorflow-gpu
Collecting package metadata (current_repodata.json): done
Solving environment: done


## Package Plan ##

  environment location: /home/psychotechnopath/anaconda3/envs/DeepLearning3.6

  added / updated specs:
    - tensorflow-gpu


The following packages will be downloaded:

    package                    |            build
    ---------------------------|-----------------
    _tflow_select-2.1.0        |              gpu           2 KB
    cudatoolkit-10.1.243       |       h6bb024c_0       347.4 MB
    cudnn-7.6.5                |       cuda10.1_0       179.9 MB
    cupti-10.1.168             |                0         1.4 MB
    tensorflow-2.1.0           |gpu_py36h2e5cdaa_0           4 KB
    tensorflow-base-2.1.0      |gpu_py36h6c5654b_0       155.9 MB
    tensorflow-gpu-2.1.0       |       h0d30ee6_0           3 KB
    ------------------------------------------------------------
                                           Total:       684.7 MB

The following NEW packages …