1. 程式人生 > 實用技巧 >tensorflow 安裝: could no t load dynamic library 'cudart64_101.dll'; dlerror: cudart64_101.dll not found

tensorflow 安裝: could no t load dynamic library 'cudart64_101.dll'; dlerror: cudart64_101.dll not found

TensorFlow 是一個端到端開源機器學習平臺

安裝

pip3  install tensorflow

使用時報錯如下

2020-06-03 09:42:51.737502: W tensorflow/stream_executor/platform/default/dso_loader.cc:55] Could no
t load dynamic library 'cudart64_101.dll'; dlerror: cudart64_101.dll not found
2020-06-03 09:42:51.737502: 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.
2020-06-03 09:42:53.392597: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x167fbc50
initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-06-03 09:42:53.392597: I tensorflow/compiler/xla/service/service.cc:176] StreamExecutor devic
e (0): Host, Default Version
2020-06-03 09:42:53.395597: I tensorflow/stream_executor/platform/default/dso_loader.cc:44] Successf
ully opened dynamic library nvcuda.dll
2020-06-03 09:42:53.472601: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1561] Found device 0
with properties:
pciBusID: 0000:01:00.0 name: GeForce GTX 670 computeCapability: 3.0

沒有cuDNN64_7.dll檔案,TensorFlow 將無法載入

Windows上搭建TensorFlow的開發環境

需要至少需要安裝:

  1.python

  2.CUDA和CuDNN

  CUDA是NVIDIA發明了一種平行計算平臺和程式設計模型。通過利用圖形處理單元(GPU)的功能,可以顯著提高計算效能

  CUDA下載地址:https://developer.nvidia.com/cuda-90-download-archive

下載後安裝

配置環境變數

SET PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\bin;%PATH%
SET PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\extras\CUPTI\libx64;%PATH%
SET PATH=C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.1\include;%PATH%

檢視是否安裝成功

nvcc -V

  NVIDIA cuDNN是用於深度神經網路的GPU加速的原語庫

  CuDNN下載地址:https://developer.nvidia.com/cudnn

下載後解壓

配置環境變數

SET PATH=C:\Program Files\cuda\bin;%PATH%

檢視自己的GPU,配置環境變數

SET PATH=C:\Program Files\NVIDIA Corporation\NVSMI;%PATH%

檢視命令

nvidia-smi

  3.TensorFlow(GPU版)

安裝前一定要確認好每個軟體的版本是否相互支援

Bazel 是用於編譯 TensorFlow 的構建工具

  將 Bazel 可執行檔案的位置新增到%PATH%環境變數中

  下載地址:https://github.com/bazelbuild/bazel/releases

檢視tensorflow版本資訊,根據版本對應不同版本的軟體

eg:

  tensorflow2.2.0

  Python 3.7.7