Anaconda安装tensorflow和keras(gpu版，超详细)

本人配置：window10+GTX 1650+tensorflow-gpu 1.14+keras-gpu 2.2.5+python 3.6，亲测可行

一.Anaconda安装

直接到清华镜像网站下载(什么版本都可以)：https://mirrors.tuna.tsinghua.edu.cn/anaconda/archive/

这是我下载的版本，自带python版本为3.6

 下载后直接安装即可，可参考：https://www.cnblogs.com/maxiaodoubao/p/9854595.html

二.建立开发环境

1.打开Prompt

    点击开始，选择Anaconda Prompt（anaconda3）

　

2.更换conda源

conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/
conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/
conda config --append channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/
conda config --set show_channel_urls yes

按照这么写的话，后续创建环境会报错：

所以直接打开.condarc文件，改为如下(将https改为http，去掉了default，末尾添加了/win-64/)：

ssl_verify: true
show_channel_urls: true
channels:
  - http://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/win-64/
  - http://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/win-64/
  - http://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/pytorch/win-64/

 3.创建虚拟环境

创建一个名为tensorflow ，python版本为3.6的虚拟环境

conda create -n tensorflow python=3.6

查看虚拟环境

conda info -e

激活开发环境

activate tensorflow

 三.安装tensorflow-gpu和keras-gpu

首先，这里有两种安装方式，一种是conda，一种是pip，conda下载较慢，但会自动安装适合的CUDA和CuDnn，pip安装快，但是需要手动安装CUDA和CuDnn，这里重点介绍pip安装方式

1.conda安装

输入命令，需要下载一些包，直到done，自动下载了gpu，直接可以使用，比较方便和简单

conda install tensorflow-gpu==xxx.xxx.xx你想要的版本号

本人一开始使用这种方法，结果在下载时经常卡住，中断，主要还是因为网络问题，需要多试几次，可以安装成功，因此需要使用国内镜像，但是使用镜像后，依然安装不成功，所以放弃了这种方法。

2.pip安装(有很多坑)

（1）打开计算机管理

       

 　  点击查看gpu算力：CUDA GPUs | NVIDIA Developer

　　算力高于3.1就行，就可以跑深度程序。

（2）打开NVIDIV控制面板

　　

　　

 　　

　　 可以看到最大支持CUDA版本是11.4，只要下载的版本没有超过最大值即可。

（3）安装CUDA

　　CUDA下载地址：CUDA Toolkit Archive | NVIDIA Developer （亲测，官网下载不慢）

　　注意：cuda和cudnn安装要注意版本搭配，以及和python版本的搭配，然后根据自己的需要安装

　　

　　以下是我的下载

 　　下载之后：按照步骤安装

　　

 　　

 　　

 　　

　　

 　　配置环境变量：

　　

 　　

C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\bin
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\extras\CUPTI\libx64
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\libnvvp

 　　不要直接在path里面配置，会显示太大

 （4）安装cuDNN：

　　cuDNN下载地址：cuDNN Archive | NVIDIA Developer（亲测，官网下载不慢）

　　注意：cuDNN要跟CUDA版本搭配好，不能随便下载

　　

 　   安装时，可能需要注册NVIDIA账户，花一点时间注册一下即可下载。

　　下载完后，将文件解压，将里面的文件全部导入到CUDA/v10.0路径下。

 （5）安装tensorflow-gpu和keras-gpu

 可以对照表格安装对应版本tensorflow和keras

pip install tensorflow-gpu==1.14.0 -i https://pypi.tuna.tsinghua.edu.cn/simple
pip install keras-gpu==2.2.5 -i https://pypi.tuna.tsinghua.edu.cn/simple

（6）安装其他库

pip install -i https://pypi.doubanio.com/simple/ opencv-python
pip install -i https://pypi.doubanio.com/simple/ pillow
pip install -i https://pypi.doubanio.com/simple/ matplotlib
pip install -i https://pypi.doubanio.com/simple/ sklearn

四.测试是否使用了GPU

进入python编译环境，输入一下代码，如果结果是True，表示GPU可用

import tensorflow as tf
print(tf.test.is_gpu_available())

若为True，使用命令查看gpu是否在运行

nvidia-smi

 

五.jupyter使用虚拟环境

其实使用虚拟环境非常简单，只需要安装一个nb_conda包就可以直接使用了

conda install nb_conda 

在你的新环境上安装ipykernel，重启jupyter之后就可以用了

conda install -n tensorflow ipykernel

 正好可以测试tensorflow和keras是否在GPU上运行

来段代码测试一下：

import numpy as np

from keras.models import Sequential
from keras.layers import Dense, Dropout, Activation, Flatten
import matplotlib.pyplot as plt
from sklearn import datasets

# 样本数据集，两个特征列，两个分类二分类不需要onehot编码，直接将类别转换为0和1，分别代表正样本的概率。
X,y=datasets.make_classification(n_samples=200, n_features=2, n_informative=2, n_redundant=0,n_repeated=0, n_classes=2, n_clusters_per_class=1)

# 构建神经网络模型
model = Sequential()
model.add(Dense(input_dim=2, units=1))
model.add(Activation('sigmoid'))

# 选定loss函数和优化器
model.compile(loss='binary_crossentropy', optimizer='sgd')

# 训练过程
print('Training -----------')
for step in range(501):
    cost = model.train_on_batch(X, y)
    if step % 50 == 0:
        print("After %d trainings, the cost: %f" % (step, cost))

# 测试过程
print('\nTesting ------------')
cost = model.evaluate(X, y, batch_size=40)
print('test cost:', cost)
W, b = model.layers[0].get_weights()
print('Weights=', W, '\nbiases=', b)

# 将训练结果绘出
Y_pred = model.predict(X)
Y_pred = (Y_pred*2).astype('int')  # 将概率转化为类标号，概率在0-0.5时，转为0，概率在0.5-1时转为1
# 绘制散点图 参数：x横轴 y纵轴
plt.subplot(2,1,1).scatter(X[:,0], X[:,1], c=Y_pred[:,0])
plt.subplot(2,1,2).scatter(X[:,0], X[:,1], c=y)
plt.show()

结果：

 到此说明已经彻底成功安装上tensorflow和keras了

七.可能的问题

1.安装1.14.0版本的tensorflow后，运行时出现了错误

Using TensorFlow backend.
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorflow\python\framework\dtypes.py:516: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint8 = np.dtype([("qint8", np.int8, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorflow\python\framework\dtypes.py:517: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint8 = np.dtype([("quint8", np.uint8, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorflow\python\framework\dtypes.py:518: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint16 = np.dtype([("qint16", np.int16, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorflow\python\framework\dtypes.py:519: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint16 = np.dtype([("quint16", np.uint16, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorflow\python\framework\dtypes.py:520: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint32 = np.dtype([("qint32", np.int32, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorflow\python\framework\dtypes.py:525: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  np_resource = np.dtype([("resource", np.ubyte, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:541: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint8 = np.dtype([("qint8", np.int8, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:542: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint8 = np.dtype([("quint8", np.uint8, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:543: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint16 = np.dtype([("qint16", np.int16, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:544: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_quint16 = np.dtype([("quint16", np.uint16, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:545: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  _np_qint32 = np.dtype([("qint32", np.int32, 1)])
D:\tools\_virtualenv_dir\myproject_2_quchumasaike\env2_py36_quchumasaike\lib\site-packages\tensorboard\compat\tensorflow_stub\dtypes.py:550: FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version of numpy, it will be understood as (type, (1,)) / '(1,)type'.
  np_resource = np.dtype([("resource", np.ubyte, 1)])

解决方法：

这个问题意思就是numpy的版本过低或者过高都会出现警告，只需要先卸载重新指定版本的numpy即可解决此问题

pip uninstall numpy
pip install numpy==1.16.4

2.anaconda卸载不干净：

解决办法：

（1）执行命令

conda config --remove-key channels
conda install anaconda-clean
anaconda-clean --yes

（2）运行安装目录下的 Uninstall-Anaconda3.exe 程序即可，这样就成功地将anaconda完全卸载干净了

3.利用镜像安装tensorflow-gpu

pip install -i https://pypi.tuna.tsinghua.edu.cn/simple tensorflow-gpu

4.高版本可能出现错误：AttributeError: module ‘tensorflow_core._api.v2.config’ has no attribute ‘experimental_list_devices’

解决方法：解决module ‘tensorflow_core._api.v2.config‘ has no attribute ‘experimental_list_devices‘_sinysama的博客 （亲测有效）

八.网盘下载

1.anaconda下载(5.2.0和5.3.1)：

链接：https://pan.baidu.com/s/1-iw1hjfL2u4CumCW0b0Zvg
提取码：hort

2.cuDNN和CUDA下载(10.0,10.1,11.4)：

链接：https://pan.baidu.com/s/1-06nzKI8CrlWOsKfeuY3Gw
提取码：8v05

 

 

 

 

 

参考文章：

FutureWarning: Passing (type, 1) or '1type' as a synonym of type is deprecated; in a future version

如何完全卸载Anaconda（如何下载Anaconda-Clean package）_托马斯-酷涛的博客

 利用镜像安装tensorflow_不知方向的鸵鸟的博客

怎么查看keras 或者 tensorflow 正在使用的GPU_Thinker_and_FKer的博客

Jupyter Notebook运行指定的conda虚拟环境_我是天才很好的博客

Anaconda镜像安装tensorflow-gpu1.14及Keras超详细版_Xnion的博客

win10完整Tensorflow-GPU环境搭建教程-附CUDA+cuDNN安装过程_尤利乌斯.X的博客

cuda安装教程+cudnn安装教程_hw@c14h10的博客

tensorflow版本对应关系_蠕动的爬虫的博客