版本問題：
RTX2070是nvidia出的新一代20系列顯卡，都是圖靈架構。要配合cuda10的版本，而且tensorflow 也要選擇1.13版本。於是裝了cuda最新的10.1，安裝完畢，在import tensorflow時，報importError：DLL load failed：找不到指定的模塊“這個錯誤，在另一篇文章中我已經說過，這個錯誤基本都是版本不對應的問題。於是想着是不是cuda版本太高了，卸載cuda10.1,重新安裝cuda10.0，問題解決。
原文：https://blog.csdn.net/zhengxinjie2/article/details/89289544
本文使用anaconda安裝tensorflw，省去cuda和cudnn的安裝過程
附：anaconda對應的python版本
首先解釋一下上表。 anaconda在每次發佈新版本的時候都會給python3和python2都發佈一個包，版本號是一樣的。
表格中，python版本號下方的離它最近的anaconda包就是包含它的版本。
舉個例子，假設你想安裝python2.7.14，在表格中找到它，它下方的三個anaconda包（anaconda2-5.0.1、5.1.0、5.2.0）都包含python2.7.14；
假設你想安裝python3.6.5，在表格中找到它，它下方的anaconda3-5.2.0就是你需要下載的包；
假設你想安裝python3.7.0，在表格中找到它，它下方的anaconda3-5.3.0或5.3.1就是你需要下載的包；
原文：https://blog.csdn.net/yuejisuo1948/article/details/81043823
也就是說，anaconda3-5.2.0只能安裝3.6.5以下的python
1、安裝anaconda3參考：
https://blog.csdn.net/qq_15192373/article/details/81091098
安裝完成後需重啓終端，或者輸入source ~/.bashrc
再輸入python就會進入anaconda的python版本
2、安裝顯卡驅動nvidia-driver
第一：檢查顯卡和推薦驅動：
ubuntu-drivers devices
若無顯示，則先添加NVIDIA的PPA：
sudo add-apt-repository ppa:graphics-drivers/ppa
sudo apt update
顯示推薦的驅動：
jason@jason:~$ ubuntu-drivers devices
== /sys/devices/pci0000:00/0000:00:01.0/0000:01:00.0 ==
modalias : pci:v000010DEd00001F07sv000010DEsd000012ADbc03sc00i00
vendor : NVIDIA Corporation
driver : nvidia-driver-415 - third-party free
driver : nvidia-driver-410 - third-party free
driver : nvidia-driver-430 - third-party free recommended
driver : nvidia-driver-418 - third-party free
driver : xserver-xorg-video-nouveau - distro free builtin
選擇410版本
首先添加apt-get的清華源，使安裝速度更快：https://blog.csdn.net/qq_42412214/article/details/89055720
然後使用如下命令安裝，這一步時間較長
sudo apt-get install nvidia-driver-410
安裝成功：
jason@jason:~$ nvidia-smi
Sun Jul 7 02:00:19 2019
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 410.104 Driver Version: 410.104 CUDA Version: 10.0 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
|===============================+======================+======================|
| 0 GeForce RTX 2070 Off | 00000000:01:00.0 Off | N/A |
| 41% 49C P0 39W / 185W | 196MiB / 7952MiB | 1% Default |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: GPU Memory |
| GPU PID Type Process name Usage |
|=============================================================================|
| 0 1054 G /usr/lib/xorg/Xorg 106MiB |
| 0 1285 G /usr/bin/gnome-shell 88MiB |
+-----------------------------------------------------------------------------+
3、使用conda建立名爲tf-env的python環境
1）添加anaconda國內源，使anaconda安裝環境更快
將以下配置文件寫在~/.condarc中（初始爲空文件）
vim ~/.condarc
channels:
- https://mirrors.ustc.edu.cn/anaconda/pkgs/main/
- https://mirrors.ustc.edu.cn/anaconda/cloud/conda-forge/
- https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/
- defaults
show_channel_urls: true
原文：https://blog.csdn.net/observador/article/details/83618540
此處感覺添加的源網址有些少，部分包還是從官網上下的
2）建立一個新的python環境，名爲tf-env，使用python3.6.5
參考：https://blog.csdn.net/qq_31456593/article/dpetails/89090156
conda create -n my-py-env python=3.6.5
安裝完成：
Preparing transaction: done
Verifying transaction: done
Executing transaction: done
#
# To activate this environment, use:
# > source activate my-py-env
#
# To deactivate an active environment, use:
# > source deactivate
若刪除該python環境：
conda remove -n testpython --all
4、安裝tensorflow
查看可用tensorflow版本參考：https://blog.csdn.net/shiheyingzhe/article/details/80863422
在新建的python環境安裝tensorflow命令參考：https://www.jianshu.com/p/e6a9aa0e671b
jason@jason:~$ conda install -n my-py-env tensorflow-gpu
Solving environment: done
==> WARNING: A newer version of conda exists. <==
current version: 4.5.4
latest version: 4.7.5
Please update conda by running
$ conda update -n base conda
## Package Plan ##
environment location: /home/jason/anaconda3/envs/my-py-env
added / updated specs:
- tensorflow-gpu
The following packages will be downloaded:
package | build
---------------------------|-----------------
cudnn-7.6.0 | cuda10.0_0 216.6 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
cupti-10.0.130 | 0 1.8 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
markdown-3.1.1 | py36_0 113 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
hdf5-1.10.4 | hb1b8bf9_0 5.3 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
scipy-1.2.1 | py36h7c811a0_0 17.7 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
termcolor-1.1.0 | py36_1 7 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
cudatoolkit-10.0.130 | 0 380.0 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
blas-1.0 | mkl 6 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
mock-3.0.5 | py36_0 47 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
mkl-2019.4 | 243 204.1 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
numpy-base-1.16.4 | py36hde5b4d6_0 4.4 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
libgfortran-ng-7.3.0 | hdf63c60_0 1.3 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
grpcio-1.14.1 | py36h9ba97e2_0 1.0 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
werkzeug-0.15.4 | py_0 262 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
_tflow_select-2.1.0 | gpu 2 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
protobuf-3.8.0 | py36he6710b0_0 690 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
intel-openmp-2019.4 | 243 876 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
tensorflow-estimator-1.13.0| py_0 205 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
numpy-1.16.4 | py36h7e9f1db_0 49 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
mkl_random-1.0.2 | py36hd81dba3_0 407 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
mkl_fft-1.0.12 | py36ha843d7b_0 172 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
h5py-2.9.0 | py36h7918eee_0 1.2 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
six-1.12.0 | py36_0 22 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
tensorboard-1.13.1 | py36hf484d3e_0 3.3 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
libprotobuf-3.8.0 | hd408876_0 4.7 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
absl-py-0.7.1 | py36_0 157 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
tensorflow-1.13.1 |gpu_py36h3991807_0 4 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
gast-0.2.2 | py36_0 138 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
tensorflow-gpu-1.13.1 | h0d30ee6_0 2 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
c-ares-1.15.0 | h7b6447c_1 98 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
keras-applications-1.0.8 | py_0 33 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
astor-0.7.1 | py36_0 43 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
tensorflow-base-1.13.1 |gpu_py36h8d69cac_0 293.8 MB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
keras-preprocessing-1.1.0 | py_1 36 KB https://mirrors.ustc.edu.cn/anaconda/pkgs/main
------------------------------------------------------------
Total: 1.11 GB
The following NEW packages will be INSTALLED:
_tflow_select: 2.1.0-gpu https://mirrors.ustc.edu.cn/anaconda/pkgs/main
absl-py: 0.7.1-py36_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
astor: 0.7.1-py36_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
blas: 1.0-mkl https://mirrors.ustc.edu.cn/anaconda/pkgs/main
c-ares: 1.15.0-h7b6447c_1 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
cudatoolkit: 10.0.130-0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
cudnn: 7.6.0-cuda10.0_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
cupti: 10.0.130-0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
gast: 0.2.2-py36_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
grpcio: 1.14.1-py36h9ba97e2_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
h5py: 2.9.0-py36h7918eee_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
hdf5: 1.10.4-hb1b8bf9_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
intel-openmp: 2019.4-243 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
keras-applications: 1.0.8-py_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
keras-preprocessing: 1.1.0-py_1 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
libgfortran-ng: 7.3.0-hdf63c60_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
libprotobuf: 3.8.0-hd408876_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
markdown: 3.1.1-py36_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
mkl: 2019.4-243 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
mkl_fft: 1.0.12-py36ha843d7b_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
mkl_random: 1.0.2-py36hd81dba3_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
mock: 3.0.5-py36_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
numpy: 1.16.4-py36h7e9f1db_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
numpy-base: 1.16.4-py36hde5b4d6_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
protobuf: 3.8.0-py36he6710b0_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
scipy: 1.2.1-py36h7c811a0_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
six: 1.12.0-py36_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
tensorboard: 1.13.1-py36hf484d3e_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
tensorflow: 1.13.1-gpu_py36h3991807_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
tensorflow-base: 1.13.1-gpu_py36h8d69cac_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
tensorflow-estimator: 1.13.0-py_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
tensorflow-gpu: 1.13.1-h0d30ee6_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
termcolor: 1.1.0-py36_1 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
werkzeug: 0.15.4-py_0 https://mirrors.ustc.edu.cn/anaconda/pkgs/main
Proceed ([y]/n)?
輸入yes進行安裝
conda install -n my-py-env tensorflow-gpu==XXX可以選擇版本
安裝成功
jason@jason:~$ source activate my-py-env
(my-py-env) jason@jason:~$ python
Python 3.6.5 |Anaconda, Inc.| (default, Apr 29 2018, 16:14:56)
[GCC 7.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import tensorflow as tf
>>> tf.__version__
'1.13.1'