tensorflow cudart64_101.dll 版本问题

安装CUDA 11.1后，在Python运行tensorflow，GPU不能使用。提示：

Could not load dynamic library 'cudart64_101.dll'; dlerror: cudart64_101.dll not found

删除了cuda11.1及NVIDIA驱动

重新下载了NVIDIA显卡驱动
461.92-desktop-win10-64bit-international-nsd-dch-whql.exe
重新下载cuda10.1
cuda_10.1.243_426.00_win10.exe
重新下载cuDNN
cudnn-10.1-windows10-x64-v7.6.5.32.zip

分别安装后：
打开python
>>>  import    tensorflow    as    tf
>>>  tf.config.experimental.list_physical_devices('GPU')
[]
正在重置  Python  状态。
2021-03-20  09:51:56.440429:  I  tensorflow/stream_executor/platform/default/dso_loader.cc:48]  Successfully
opened  dynamic  library  cudart64_101.dll
2021-03-20  09:52:08.976977:  I  tensorflow/stream_executor/platform/default/dso_loader.cc:48]  Successfully
opened  dynamic  library  nvcuda.dll
2021-03-20  09:52:09.003288:  I  tensorflow/core/common_runtime/gpu/gpu_device.cc:1716]  Found  device  0  with
properties:
pciBusID:  0000:01:00.0  name:  GeForce  GTX  1050  Ti  computeCapability:  6.1
coreClock:  1.392GHz  coreCount:  6  deviceMemorySize:  4.00GiB  deviceMemoryBandwidth:  104.43GiB/s
2021-03-20  09:52:09.003688:  I  tensorflow/stream_executor/platform/default/dso_loader.cc:48]  Successfully
opened  dynamic  library  cudart64_101.dll
2021-03-20  09:52:09.009578:  I  tensorflow/stream_executor/platform/default/dso_loader.cc:48]  Successfully
opened  dynamic  library  cublas64_10.dll
2021-03-20  09:52:09.015217:  I  tensorflow/stream_executor/platform/default/dso_loader.cc:48]  Successfully
opened  dynamic  library  cufft64_10.dll
2021-03-20  09:52:09.016954:  I  tensorflow/stream_executor/platform/default/dso_loader.cc:48]  Successfully
opened  dynamic  library  curand64_10.dll
2021-03-20  09:52:09.022897:  I  tensorflow/stream_executor/platform/default/dso_loader.cc:48]  Successfully
opened  dynamic  library  cusolver64_10.dll
2021-03-20  09:52:09.026263:  I  tensorflow/stream_executor/platform/default/dso_loader.cc:48]  Successfully
opened  dynamic  library  cusparse64_10.dll
2021-03-20  09:52:09.027791:  W  tensorflow/stream_executor/platform/default/dso_loader.cc:59]
  Could  not  load  dynamic  library  'cudnn64_7.dll';  dlerror:  cudnn64_7.dll  not  found
2021-03-20  09:52:09.028032:  W  tensorflow/core/common_runtime/gpu/gpu_device.cc:1753]  Cannot  dlopen  some
  GPU  libraries.  Please  make  sure  the  missing  libraries  mentioned  above  are  installed  properly  if  you  would
  like  to  use  GPU.  Follow  the  guide  at  https://www.tensorflow.org/install/gpu  for  how  to  download  and  setup
  the  required  libraries  for  your  platform.
Skipping  registering  GPU  devices...
交互  Python  进程已退出。

将cuDNN的解压缩文件分别复制到Program  Files\NVIDIA  GPU  Computing  Toolkit\CUDA\v10.1对应的文件夹内
>>>  import    tensorflow    as    tf
>>>  tf.config.experimental.list_physical_devices('GPU')
[PhysicalDevice(name='/physical_device:GPU:0',  device_type='GPU')]
>>>
GPU问题解决！

tensorflow cudart64_101.dll 版本问题相关推荐

tensorflow: Could not load dynamic library ‘cudart64_101.dll‘ 解决办法
问题描述: 安装 tensorfolw-gpu2.1.0 之后调用显示 GPU False,不能调用GPU 在命令行中 import tensorflow 时显示 " Could not ...
Tensorflow——[Could not load dynamic library cudart64_101.dll]解决方案
问题描述 Could not load dynamic library 'cudart64_101.dll'; dlerror: cudart64_101.dll not found Ignore a ...
解决Tensorflow：Could not load dynamic library ‘cudart64_101.dll‘；dynamic library ‘cublas64_10.dll‘；
Introduce Tensorflow: 2.3.0 Cuda: 10.2 VS Code 之前有装pytorch的过程,但是不少文章的模型用的tensrflow..就是就pip install t ...
Could not load dynamic library ‘cudart64_101.dll‘
tenforboard用到了tensorflow 2020-09-18 22:59:53.641152: W tensorflow/stream_executor/platform/default/d ...
cudart64_101.dll not found
TensorFlow-gpu 2.3.0-gpu 匹配的cuda版本为10.1 安装的cuda10.2版本, 报错: cudart64_101.dll not found 有时也报错: cudart6 ...
Ubuntu tensorflow自定义GPU版本op节点
参考:https://blog.csdn.net/qq_27637315/article/details/79114633 windows增加op节点: https://github.com/tens ...
tensorflow r1.5 版本差异调研
原文:http://blog.csdn.net/SuperYR_210/article/details/78977588 tensorflow r1.5 版本差异调研一.Breaking Chang ...
Tensorflow：在Tensorflow的不同版本中如何实现Xavier参数权重初始化
Tensorflow:在Tensorflow的不同版本中如何实现Xavier参数权重初始化目录在Tensorflow的不同版本中实现Xavier参数权重初始化 Tensorflow1版本中实现
报错引发的版本对应——tensorflow+keras+python版本对应（全）
内容整理自多人博客,侵权请联系 Python报错:module 'tensorflow.python.keras.backend' has no attribute 'get_graph' 报错原因 ...

tensorflow cudart64_101.dll 版本问题

tensorflow cudart64_101.dll 版本问题相关推荐

最新文章

热门文章