1、Lasagne簡(jiǎn)單介紹

? ? ? ?lasagne意味千層餅，是基于theano之上主要用于建立和訓(xùn)練神經(jīng)網(wǎng)絡(luò)的深度學(xué)習(xí)庫(kù)。Lasagne is a lightweight library to build and train neural networks in Theano.?

網(wǎng)站鏈接：https://github.com/Lasagne/Lasagne、 ? ??lasagne官網(wǎng)教程

主要特點(diǎn)有：

（1）支持所有的前饋神經(jīng)網(wǎng)絡(luò)的模型，包括CNNs、RNN+LSTM以及任何擴(kuò)展的卷積神經(jīng)網(wǎng)絡(luò)。

? ? ? Supports feed-forward networks such as Convolutional Neural Networks (CNNs), recurrent networks including Long Short-Term Memory (LSTM), and any combination thereof?

? （2）支持多輸入和多輸出的網(wǎng)絡(luò)框架、包括額外附加的分類(lèi)器

? ? ? Allows architectures of multiple inputs and multiple outputs, including auxiliary classifiers?

? ? （3）擁有多種訓(xùn)練的梯度優(yōu)化算法。Nesterov momentum, RMSprop and ADAM?

（4）?繼承theano的優(yōu)點(diǎn)，它擁有theano的所有常見(jiàn)的損失函數(shù)以及無(wú)需自行計(jì)算梯度。

? ? ??Freely definable cost function and no need to derive gradients due to Theano's symbolic differentiation

（5）通過(guò)設(shè)置腳本的配置文件，支持CPUs和GPUs之間運(yùn)行的轉(zhuǎn)換。

? ? ?Transparent support of CPUs and GPUs due to Theano's expression compiler

設(shè)計(jì)的目的：

?（1）簡(jiǎn)易性:易于使用和擴(kuò)展的機(jī)器學(xué)習(xí)庫(kù)Be easy to use, easy to understand and easy to extend, to facilitate use in research?

?（2）透明性: Do not hide Theano behind abstractions, directly process and return Theano expressions or Python / numpy data types?

（3）模塊化: Allow all parts (layers, regularizers, optimizers, ...) to be used independently of Lasagne（4）實(shí)用性(Pragmatism): Make common use cases easy, do not overrate uncommon cases?

（5）限制Restraint: Do not obstruct(阻塞) users with features they decide not to use?

（6）集中性: "Do one thing and do it well"

2、lasagne的安裝（ubuntu14.04下）

（1）預(yù)備知識(shí)?prerequisites：theano庫(kù)?theano installation，注意：theano安裝的版本取決于lasagne安裝的版本。

（2）Stable Lasagne release：lasagne 0.1版本需要匹配最近的theano版本，可執(zhí)行以下代碼來(lái)獲取相應(yīng)的版本。

sudo pip install -r https://raw.githubusercontent.com/Lasagne/Lasagne/v0.1/requirements.txt

? ? ? ? ?為了簡(jiǎn)潔方面，也可以同時(shí)安裝theano和lasagne。

sudo pip install --upgrade https://github.com/Theano/Theano/archive/master.zip?

sudo pip install --upgrade https://github.com/Lasagne/Lasagne/archive/master.zip

?

developed install：也可以直接克隆lasagne庫(kù)

git clone https://github.com/Lasagne/Lasagne.git

至此，給出lasagne大體的框架：（需要自己加載訓(xùn)練數(shù)據(jù)才能運(yùn)行出結(jié)果）

import lasagne
import theano
import theano.tensor as T# create Theano variables for input and target minibatch
input_var = T.tensor4('X')
target_var = T.ivector('y')# create a small convolutional neural network
from lasagne.nonlinearities import leaky_rectify, softmax
network = lasagne.layers.InputLayer((None, 3, 32, 32), input_var)
network = lasagne.layers.Conv2DLayer(network, 64, (3, 3),nonlinearity=leaky_rectify)
network = lasagne.layers.Conv2DLayer(network, 32, (3, 3),nonlinearity=leaky_rectify)
network = lasagne.layers.Pool2DLayer(network, (3, 3), stride=2, mode='max')
network = lasagne.layers.DenseLayer(lasagne.layers.dropout(network, 0.5),128, nonlinearity=leaky_rectify,W=lasagne.init.Orthogonal())network = lasagne.layers.DenseLayer(lasagne.layers.dropout(network, 0.5),10, nonlinearity=softmax)# create loss function
prediction = lasagne.layers.get_output(network)
loss = lasagne.objectives.categorical_crossentropy(prediction, target_var)
loss = loss.mean() + 1e-4 * lasagne.regularization.regularize_network_params(network, lasagne.regularization.l2)
# create parameter update expressions
params = lasagne.layers.get_all_params(network, trainable=True)
updates = lasagne.updates.nesterov_momentum(loss, params, learning_rate=0.01,momentum=0.9)
# compile training function that updates parameters and returns training loss
train_fn = theano.function([input_var, target_var], loss, updates=updates)# train network (assuming you've got some training data in numpy arrays)
for epoch in range(100):loss = 0for input_batch, target_batch in training_data:
loss += train_fn(input_batch, target_batch)print("Epoch %d: Loss %g" % (epoch + 1, loss / len(training_data)))# use trained network for predictions
test_prediction = lasagne.layers.get_output(network, deterministic=True)
predict_fn = theano.function([input_var], T.argmax(test_prediction, axis=1))
print("Predicted class for first test input: %r" % predict_fn(test_data[0]))

3、配置GPU

GPU的配置需要cuda支持的英偉達(dá)顯卡，(? NVIDIA GPU with CUDA),查看電腦配置的顯卡型號(hào)，下載對(duì)應(yīng)的所支持的cuda版本

https://developer.nvidia.com/cuda-downloads放在主目錄下。

（1）按照cuda官網(wǎng)教程完成安裝。

（2）添加環(huán)境變量

執(zhí)行sudo gedit ?~/.bashrc 打開(kāi).bashrc文件，在末尾寫(xiě)入以下幾行：? ?

1. export CUDA_HOME=/usr/local/cuda-***/bin???#****代表cuda對(duì)應(yīng)的版本，cuda-7.0或cuda-7.5、cuda8.0等
2. export?LD_LIBRARY_PATH=/usr/local/cuda-7.0/lib64? #如果是32位，去掉末尾的64

執(zhí)行 sudo source ~/.bashrc ? ?#使環(huán)境變量生效

檢查環(huán)境變量：echo $PATH

在home目錄下配置.theanorc文件

[global]
floatX = float32
device = gpu

如果cuda配置成功，終端執(zhí)行python，在python下，import theano 則會(huì)顯示print下的結(jié)果

THEANO_FLAGS=device=gpu python -c "import theano; print(theano.sandbox.cuda.device_properties(0))"

?

總結(jié)

以上是生活随笔為你收集整理的Theano深度学习框架之Lasagne安装及入门的全部?jī)?nèi)容，希望文章能夠幫你解決所遇到的問(wèn)題。

如果覺(jué)得生活随笔網(wǎng)站內(nèi)容還不錯(cuò)，歡迎將生活随笔推薦給好友。