无法加载本机 TensorFlow 运行时。蟒蛇 3.5.2

Question

我正在尝试在我的电脑上安装 Tensorflow，我安装了 Python 3.5.2 64 位，适用于 Windows 10 的 cuda_8.0.61 和 cudnn-8.0-windows10-x64-v6.0

我使用“本机”pip 安装 GPU 版本的 Tensorflow，然后我打开 IDLE 并使用“import tensorflow as tf”进行测试，我收到以下错误

我怎样才能解决这个问题 ？提前谢谢你 <3

 Traceback (most recent call last):
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 18, in swig_import_helper
    return importlib.import_module(mname)
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\importlib__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 986, in _gcd_import
  File "<frozen importlib._bootstrap>", line 969, in _find_and_load
  File "<frozen importlib._bootstrap>", line 958, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 666, in _load_unlocked
  File "<frozen importlib._bootstrap>", line 577, in module_from_spec
  File "<frozen importlib._bootstrap_external>", line 906, in create_module
  File "<frozen importlib._bootstrap>", line 222, in _call_with_frames_removed
ImportError: DLL load failed: The specified module could not be found.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 41, in <module>
    from tensorflow.python.pywrap_tensorflow_internal import *
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 21, in <module>
    _pywrap_tensorflow_internal = swig_import_helper()
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 20, in swig_import_helper
    return importlib.import_module('_pywrap_tensorflow_internal')
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\importlib__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
ImportError: No module named '_pywrap_tensorflow_internal'

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "<pyshell#0>", line 1, in <module>
    import tensorflow as tf
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow__init__.py", line 24, in <module>
    from tensorflow.python import *
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python__init__.py", line 51, in <module>
    from tensorflow.python import pywrap_tensorflow
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 52, in <module>
    raise ImportError(msg)
ImportError: Traceback (most recent call last):
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 18, in swig_import_helper
    return importlib.import_module(mname)
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\importlib__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
  File "<frozen importlib._bootstrap>", line 986, in _gcd_import
  File "<frozen importlib._bootstrap>", line 969, in _find_and_load
  File "<frozen importlib._bootstrap>", line 958, in _find_and_load_unlocked
  File "<frozen importlib._bootstrap>", line 666, in _load_unlocked
  File "<frozen importlib._bootstrap>", line 577, in module_from_spec
  File "<frozen importlib._bootstrap_external>", line 906, in create_module
  File "<frozen importlib._bootstrap>", line 222, in _call_with_frames_removed
ImportError: DLL load failed: The specified module could not be found.

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 41, in <module>
    from tensorflow.python.pywrap_tensorflow_internal import *
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 21, in <module>
    _pywrap_tensorflow_internal = swig_import_helper()
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 20, in swig_import_helper
    return importlib.import_module('_pywrap_tensorflow_internal')
  File "C:\Users\Admin\AppData\Local\Programs\Python\Python35\lib\importlib__init__.py", line 126, in import_module
    return _bootstrap._gcd_import(name[level:], package, level)
ImportError: No module named '_pywrap_tensorflow_internal'

Failed to load the native TensorFlow runtime.

See https://www.tensorflow.org/install/install_sources#common_installation_problems

for some common reasons and solutions.  Include the entire stack trace
above this error message when asking for help.

原文由 lamhoangtung 发布，翻译遵循 CC BY-SA 4.0 许可协议

Answer 1

按照这些步骤，我设法在 Windows 上运行 Tensorflow。希望能帮助到你！它可能会根据您安装 Python 的文件夹而改变。

1. 下载并安装 Python 3.5.x

2. 下载并安装 Nvidia CUDA 工具包

3. 下载 CuDNN 5.1

4. 2. 完成后，将 CuDNN 文件复制到 Nvidia CUDA 工具包文件夹（通常位于 C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v8.0）

   • 将cudnn\bin\cudnn64_5.dll复制到C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v8.0\bin
     
   • 将cudnn\include\cudnn.h复制到C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v8.0\include
     
   • 将cudnn\lib\x64\cudnn.lib复制到C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v8.0\lib\x64
     

5. 通过 pip 命令提示符 ‘pip install –upgrade tensorflow-gpu’ 安装 Tensorflow

   5.1 TensorFlow 依赖

   • 确保安装了 Visual C++ Redistributate 2015 x64。如果没有，请下载它
   • 将 _curses_curses.cp35-win_amd64.pyd 和 _curses_panel.cp35-win_amd64.pyd 移动到 C:\Users\Username\AppData\Local\Programs\Python\Python35\Lib\site-packages（或者您可能需要下载它们作为下一个包。只需 Ctrl+f 表示“诅咒”）
   • 通过 pip 下载并安装这些包（移动到下载它们的目录，打开命令提示符并制作“pip install package_name”）
     • numpy+mkl ( http://www.lfd.uci.edu/~gohlke/pythonlibs/#numpy )

          - 下载 numpy‑1.12.1+mkl‑cp35‑cp35m‑win_amd64.whl
          - pip 安装 numpy‑1.12.1+mkl‑cp35‑cp35m‑win_amd64.whl
       

     • scipy ( http://www.lfd.uci.edu/~gohlke/pythonlibs/#scipy )

          - 下载 scipy‑0.19.0‑cp35‑cp35m‑win_amd64.whl
          - pip 安装 scipy‑0.19.0‑cp35‑cp35m‑win_amd64.whl
       

   • 打开命令提示符并制作“pip install matplotlib”

原文由 Bruno Barbosa 发布，翻译遵循 CC BY-SA 3.0 许可协议

Answer 2

为什么不试试 Anaconda？

 conda create -n gpu_env tensorflow-gpu
conda activate gpu_env

就这么简单。当您敬畏地观看时，所有需要的包和运行时（包括 Visual Studio）都将汇集在一起 :)。 试试看！

警告： 您需要从 tensorflow 内部运行 Keras 才能检测和使用您的 GPU 。否则，如果您只使用 keras 和 tensorflow 作为后端， Keras 将使用普通的 tensorflow 而不是 tensorflow-gpu 。两者并排安装。

这意味着你应该这样编码：

 import tensorflow as tf
from tensorflow.keras.models       import Sequential, load_model
from tensorflow.keras.layers       import Dense,Dropout
from tensorflow.keras.callbacks    import ModelCheckpoint, EarlyStopping,CSVLogger
from tensorflow.keras.utils        import plot_model
.............
.............

希望这有帮助。

原文由 user3103059 发布，翻译遵循 CC BY-SA 4.0 许可协议