导入tensorflow-gpu 1.8.0遇DLL加载失败及模块缺失错误求助

你之前用TensorFlow 1.4.1搭配CUDA 8.0和cuDNN 5.1运行一切正常，为了安装Tensorforce升级到TensorFlow-GPU 1.8.0后，却遇到了导入报错的问题。结合我处理类似问题的经验，给你整理了以下内容和针对性解决方案：

当前环境配置 #

• 操作系统：Windows 10
• TensorFlow版本：tensorflow-gpu==1.8.0
• CUDA Toolkit版本：9.0
• cuDNN版本：7.0.5（也曾尝试7.1）
• Python环境：基于Python 3.6.3的conda虚拟环境
• 安装命令：pip install --ignore-installed --upgrade tensorflow-gpu

报错堆栈信息 #

Traceback (most recent call last): File "C:\Users\User\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 18, in swig_import_helper return importlib.import_module(mname) File "C:\Users\User\Anaconda3\envs\tensorflow\lib\importlib\__init__.py", line 126, in import_module return _bootstrap._gcd_import(name[level:], package, level) File "<frozen importlib._bootstrap>", line 994, in _gcd_import File "<frozen importlib._bootstrap>", line 971, in _find_and_load File "<frozen importlib._bootstrap>", line 955, in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 658, in _load_unlocked File "<frozen importlib._bootstrap>", line 571, in module_from_spec File "<frozen importlib._bootstrap_external>", line 922, in create_module File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed ImportError: DLL load failed: Не найден указанный модуль. During handling of the above exception, another exception occurred: Traceback (most recent call last): File "C:\Users\User\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 41, in <module> from tensorflow.python.pywrap_tensorflow_internal import * File "C:\Users\User\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 21, in <module> _pywrap_tensorflow_internal = swig_import_helper() File "C:\Users\User\Anaconda3\envs\tensorflow\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\User\Anaconda3\envs\tensorflow\lib\importlib\__init__.py", line 126, in import_module return _bootstrap._gcd_import(name[level:], package, level) ModuleNotFoundError: No module named '_pywrap_tensorflow_internal' During handling of the above exception, another exception occurred: Traceback (most recent call last): File "<stdin>", line 1, in <module> File "C:\Users\User\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\__init__.py", line 24, in <module> from tensorflow.python import * File "C:\Users\User\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\__init__.py", line 51, in <module> from tensorflow.python import pywrap_tensorflow File "C:\Users\User\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 52, in <module> raise ImportError(msg) ImportError: Traceback (most recent call last): File "C:\Users\User\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\pywrap_tensorflow_internal.py", line 18, in swig_import_helper return importlib.import_module(mname) File "C:\Users\User\Anaconda3\envs\tensorflow\lib\importlib\__init__.py", line 126, in import_module return _bootstrap._gcd_import(name[level:], package, level) File "<frozen importlib._bootstrap>", line 994, in _gcd_import File "<frozen importlib._bootstrap>", line 971, in _find_and_load File "<frozen importlib._bootstrap>", line 955, in _find_and_load_unlocked File "<frozen importlib._bootstrap>", line 658, in _load_unlocked File "<frozen importlib._bootstrap>", line 571, in module_from_spec File "<frozen importlib._bootstrap_external>", line 922, in create_module File "<frozen importlib._bootstrap>", line 219, in _call_with_frames_removed ImportError: DLL load failed: Не найден указанный модуль. During handling of the above exception, another exception occurred: Traceback (most recent call last): File "C:\Users\User\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\pywrap_tensorflow.py", line 41, in <module> from tensorflow.python.pywrap_tensorflow_internal import * Fi

针对性排查与解决方案 #

我之前处理过不少类似的TensorFlow DLL加载失败问题，结合你的情况，推荐你按以下步骤逐一排查：

1. 确认CUDA与cuDNN的路径配置和文件复制

   • 先检查系统环境变量PATH，确保CUDA 9.0的C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0\bin、C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0\libnvvp、C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v9.0\lib\x64这三个路径已经添加，并且要放在旧版本CUDA的路径前面，避免系统调用旧版本的库文件。
   • 再核对cuDNN的文件是否正确复制到CUDA目录：
     • 将cuDNN压缩包中bin文件夹的.dll文件复制到CUDA 9.0的bin目录
     • include文件夹的.h文件复制到CUDA 9.0的include目录
     • lib\x64文件夹的.lib文件复制到CUDA 9.0的lib\x64目录
   • 完成后重启电脑，让环境变量生效。

2. 安装正确版本的VS2015运行库

   • 你提到安装了VS2015，但TensorFlow 1.8需要的是Microsoft Visual C++ 2015 Redistributable Update 3的x64版本，而不是VS2015 IDE本身。安装对应版本后重启电脑再测试导入。

3. 创建全新的conda虚拟环境

   • 旧环境可能残留了和TensorFlow 1.8冲突的依赖，试试创建干净的环境：

     conda create -n tf18_env python=3.6.3
     conda activate tf18_env
     pip install tensorflow-gpu==1.8.0
     

   • 尽量只用pip安装TensorFlow，不要混合conda和pip，避免依赖版本混乱。

4. 检查GPU驱动兼容性

   • TensorFlow 1.8.0搭配CUDA 9.0要求GPU驱动版本至少为384.81，你可以去NVIDIA官网查看你的GPU对应的最新兼容驱动，更新后再尝试导入。

5. 重新安装TensorFlow，避免缓存损坏

   • 有时候pip下载的安装包可能损坏，完全卸载后重新安装：

     pip uninstall tensorflow-gpu -y
     pip install tensorflow-gpu==1.8.0 --no-cache-dir
     

   • --no-cache-dir参数会强制重新下载安装包，避免使用缓存的损坏文件。

内容的提问来源于stack exchange，提问作者user8075709