解决 tensorflow-gpu 2.1 出现错误 “Could not load dynamic library 'cudart64_101.dll'

问题描述

Python 3.5.5 | packaged by conda-forge | (default, Jul 24 2018, 01:52:17) [MSC v.1900 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow as tf
2020-04-22 14:19:56.862944: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could not load dynamic library 'cudart64_101.dll'; dlerror: cudart64_101.dll not found
2020-04-22 14:19:56.867965: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.

信息分类注释：
V: 显示全部信息
D: 显示调试信息
I: 显示一般信息
W: 显示警告信息
E:显示错误信息

问题原因

“cudart64_101.dll”是cuda10.1这个版本中才含有的，而我安装的cuda为10.2版本的。

解决方法

将“cudart64_101.dll”添加到“C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.2\bin”中去，问题解决。

Python 3.5.5 | packaged by conda-forge | (default, Jul 24 2018, 01:52:17) [MSC v.1900 64 bit (AMD64)] on win32
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow as tf
2020-04-22 14:29:16.412419: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successfully opened dynamic library cudart64_101.dll

cudart64_101.dll的下载地址：cudart64_101.dll