linux(manjaro) tensorflow2.1 conda cuda10 双显卡笔记本深度学习环境搭建
下学期要学tensorflow,看着我可怜的1050ti,流下了贫穷的泪水,但无奈要做实验啊,学还是得学的,安装过程记录一下,仅供参考
关于manjaro
之前写过一篇怎么安装manjaro的文章来着,虽然manjaro在国内不是大众发行版,但在尝试过诸多linux后,我最终留在了manjaro.
双显卡驱动
我的驱动,直接上图
Anaconda
一开始我尝试用pacman直接安装tf cuda cudnn等,很简单
tf CPU
sudo pacman -S python-tensorflow-opt
tf GPU
sudo pacman -S python-tensorflow-opt-cuda cuda cudnn
但是GUP版装好之后运行测试会报
RuntimeError: cuda runtime error (35) : CUDA driver version is insufficient for CUDA runtime version at …
原因：CUDA驱动版本不满足CUDA运行版本。
具体显卡驱动与CUDA版本对应见下
/cuda/cuda-toolkit-release-notes/index.html
我的是440xx而软件库中提供的是cuda11
不想换驱动,那就给 cuda 和 tf 降级
conda安装
sudo pacman -S anacondaconda -h
如果有conda:命令未找到的报错,就需要修改一下环境变量
export PATH=$PATH:/opt/anaconda/bin
CUDA CUDNN
conda install cudatoolkit=10.1 cudnn=7.6 -c https://mirrors.tuna./anaconda/pkgs/main/linux-64/
tensorflow2.1
conda create -n tf2-gpu tensorflow-gpu==2.1 -c https://mirrors.tuna./anaconda/pkgs/main/linux-64/
装好后,检查环境
conda env list# conda environments:#tf2-gpu $home/.conda/envs/tf2-gpubase * /opt/anaconda
进入环境并测试
与win不同,linux进入conda环境要使用source activate,退出则是conda deactivate
要进入刚才搭建的tf2的环境只需要输入source activate tf2-gpu
source activate tf2-gpu(tf2-gpu) git clone /guangfuhao/Deeplearning(tf2-gpu) cd Deeplearning(tf2-gpu) cp mnist.npz <你的测试目录>(tf2-gpu) pip install matplotlib numpy
编辑测试程序,很短就用vim test.py,注意将这个test.py与之前下载的mnist.npz放到同一目录下
测试程序
# 1.Import the neccessary libraries neededimport numpy as npimport tensorflow as tfimport matplotlibfrom matplotlib import pyplot as plt######################################################################### 2.Set default parameters for plotsmatplotlib.rcParams['font.size'] = 20matplotlib.rcParams['figure.titlesize'] = 20matplotlib.rcParams['figure.figsize'] = [9, 7]matplotlib.rcParams['font.family'] = ['STKaiTi']matplotlib.rcParams['axes.unicode_minus'] = False######################################################################### 3.Initialize Parameters# Initialize learning ratelr = 1e-3# Initialize loss arraylosses = []# Initialize the weights layers and the bias layersw1 = tf.Variable(tf.random.truncated_normal([784, 256], stddev=0.1))b1 = tf.Variable(tf.zeros([256]))w2 = tf.Variable(tf.random.truncated_normal([256, 128], stddev=0.1))b2 = tf.Variable(tf.zeros([128]))w3 = tf.Variable(tf.random.truncated_normal([128, 10], stddev=0.1))b3 = tf.Variable(tf.zeros([10]))######################################################################### 4.Import the minist dataset by numpy offlinedef load_mnist():# define the directory where mnist.npz is(Please watch the '\'!)path = r'./mnist.npz'f = np.load(path)x_train, y_train = f['x_train'], f['y_train']x_test, y_test = f['x_test'], f['y_test']f.close()return (x_train, y_train), (x_test, y_test)(train_image, train_label), _ = load_mnist()x = tf.convert_to_tensor(train_image, dtype=tf.float32) / 255.y = tf.convert_to_tensor(train_label, dtype=tf.int32)# Reshape x from [60k, 28, 28] to [60k, 28*28]x = tf.reshape(x, [-1, 28*28])######################################################################### bine x and y as a tuple and batch themtrain_db = tf.data.Dataset.from_tensor_slices((x, y)).batch(128)'''#Encapsulate train_db as an iterator objecttrain_iter = iter(train_db)sample = next(train_iter)'''######################################################################### 6.Iterate database for 20 timesfor epoch in range(20):# For every batch:x:[128, 28*28],y: [128]for step, (x, y) in enumerate(train_db):with tf.GradientTape() as tape: # tf.Variable# x: [b, 28*28]# h1 = x@w1 + b1# [b, 784]@[784, 256] + [256] => [b, 256] + [256] => [b, 256] + [b, 256]h1 = x@w1 + tf.broadcast_to(b1, [x.shape[0], 256])h1 = tf.nn.relu(h1)# [b, 256] => [b, 128]h2 = h1@w2 + b2h2 = tf.nn.relu(h2)# [b, 128] => [b, 10]out = h2@w3 + b3# y: [b] => [b, 10]y_onehot = tf.one_hot(y, depth=10)# compute loss# mse = mean(sum(y-out)^2)# [b, 10]loss = tf.square(y_onehot - out)# mean: scalarloss = tf.reduce_mean(loss)# compute gradientsgrads = tape.gradient(loss, [w1, b1, w2, b2, w3, b3])# Update the weights and the biasw1.assign_sub(lr * grads[0])b1.assign_sub(lr * grads[1])w2.assign_sub(lr * grads[2])b2.assign_sub(lr * grads[3])w3.assign_sub(lr * grads[4])b3.assign_sub(lr * grads[5])if step % 100 == 0:print(epoch, step, 'loss:', float(loss))losses.append(float(loss))######################################################################### 7.Show the change of losses via matplotlibplt.figure()plt.plot(losses, color='C0', marker='s', label='训练')plt.xlabel('Epoch')plt.legend()plt.ylabel('MSE')# Save figure as '.svg' file# plt.savefig('forward.svg')plt.show()
python3 test.py
不出意外会有类似的输出
最后画出一张图
ps: 如何优雅的监控GPU
watch -n 1 nvidia-smi
好了,环境搭建大功告成 在我的机器上这个过程是成立的,如果有什么疑问欢迎在评论区留言
