TFLearn是把常见的例子做了个抽象和封装,使用更加方便,对于学习tensorflow有很大帮助。网络结构包括Alexnet、VGGNet、Network in Network、Highway Network、Residual Network、GoogleNet、AutoEncoder等,使用数据集包括MNIST和CIFAR-10等,地址:
TFLearn(https://github.com/tflearn/tflearn)
例子(https://github.com/tflearn/tflearn/tree/master/examples)
API(http://tflearn.org/doc_index/#api).
线性回归. Implement a linear regression using TFLearn.
https://github.com/tflearn/tflearn/blob/master/examples/basics/linear_regression.py
逻辑处理. Implement logical operators with TFLearn (also includes a usage of'merge').
https://github.com/tflearn/tflearn/blob/master/examples/basics/logical.py
参数保持. Save and Restore a model.
https://github.com/tflearn/tflearn/blob/master/examples/basics/weights_persistence.py
微调.Fine-Tune a pre-trained model on a new task.
https://github.com/tflearn/tflearn/blob/master/examples/basics/finetuning.py
使用HDF5. Use HDF5 to handle large datasets.
https://github.com/tflearn/tflearn/blob/master/examples/basics/use_hdf5.py
使用DASK. Use DASK to handle large datasets.
https://github.com/tflearn/tflearn/blob/master/examples/basics/use_dask.py
多层感知机. A multi-layer perceptron implementation for MNIST classificationtask.
https://github.com/tflearn/tflearn/blob/master/examples/images/dnn.py
ConvolutionalNetwork (MNIST). A Convolutional neural network implementation for classifyingMNIST dataset.
https://github.com/tflearn/tflearn/blob/master/examples/images/convnet_mnist.py
ConvolutionalNetwork (CIFAR-10). A Convolutional neural network implementation forclassifying CIFAR-10 dataset.
https://github.com/tflearn/tflearn/blob/master/examples/images/convnet_cifar10.py
Network inNetwork. 'Network in Network' implementation for classifying CIFAR-10 dataset.
https://github.com/tflearn/tflearn/blob/master/examples/images/network_in_network.py
Alexnet.Apply Alexnet to Oxford Flowers 17 classification task.
https://github.com/tflearn/tflearn/blob/master/examples/images/alexnet.py
VGGNet.Apply VGG Network to Oxford Flowers 17 classification task.
https://github.com/tflearn/tflearn/blob/master/examples/images/vgg_network.py
VGGNetFinetuning(Fast Training). Use a pre-trained VGG Network and retrain it onyour own data, for fast training.
https://github.com/tflearn/tflearn/blob/master/examples/images/vgg_network_finetuning.py
RNNPixels. Use RNN (over sequence of pixels) to classify images.
https://github.com/tflearn/tflearn/blob/master/examples/images/rnn_pixels.py
HighwayNetwork. Highway Network implementation for classifying MNIST dataset.
https://github.com/tflearn/tflearn/blob/master/examples/images/highway_dnn.py
HighwayConvolutional Network. Highway Convolutional Network implementation forclassifying MNIST dataset.
https://github.com/tflearn/tflearn/blob/master/examples/images/convnet_highway_mnist.py
ResidualNetwork (MNIST). A bottleneck residual network applied to MNIST classificationtask.
https://github.com/tflearn/tflearn/blob/master/examples/images/residual_network_mnist.py
ResidualNetwork (CIFAR-10). A residual network applied to CIFAR-10 classification task.
https://github.com/tflearn/tflearn/blob/master/examples/images/residual_network_cifar10.py
GoogleInception (v3). Google's Inception v3 network applied to Oxford Flowers 17classification task.
https://github.com/tflearn/tflearn/blob/master/examples/images/googlenet.py
AutoEncoder. An auto encoder applied to MNIST handwritten digits.
https://github.com/tflearn/tflearn/blob/master/examples/images/autoencoder.py
自然语言处理
RecurrentNeural Network (LSTM). Apply an LSTM to IMDB sentiment dataset classificationtask.
https://github.com/tflearn/tflearn/blob/master/examples/nlp/lstm.py
Bi-DirectionalRNN (LSTM). Apply a bi-directional LSTM to IMDB sentiment datasetclassification task.
https://github.com/tflearn/tflearn/blob/master/examples/nlp/bidirectional_lstm.py
DynamicRNN (LSTM). Apply a dynamic LSTM to classify variable length text from IMDBdataset.
https://github.com/tflearn/tflearn/blob/master/examples/nlp/dynamic_lstm.py
City NameGeneration. Generates new US-cities name, using LSTM network.
https://github.com/tflearn/tflearn/blob/master/examples/nlp/lstm_generator_cityname.py
Shakespeares Generation. Generates new Shakespeare s, using LSTM network.
https://github.com/tflearn/tflearn/blob/master/examples/nlp/lstm_generator_shakespeare.py
Seq2seq.Pedagogical example of seq2seq reccurent network. See this repo for fullinstructions.
https://github.com/tflearn/tflearn/blob/master/examples/nlp/seq2seq_example.py
https://github.com/ichuang/tflearn_seq2seq
CNN Seq.Apply a 1-D convolutional network to classify sequence of words from IMDBsentiment dataset.
https://github.com/tflearn/tflearn/blob/master/examples/nlp/cnn_sentence_classification.py
增强学习
AtariPacman 1-step Q-Learning. Teach a machine to play Atari games (Pacman bydefault) using 1-step Q-learning.
https://github.com/tflearn/tflearn/blob/master/examples/reinforcement_learning/atari_1step_qlearning.py
推荐系统- Wide & Deep Network. Pedagogical example of wide & deep networks forrecommender systems.
https://github.com/tflearn/tflearn/blob/master/examples/others/recommender_wide_and_deep.py
Notebooks
SpiralClassification Problem. TFLearn implementation of spiral classification problemfrom Stanford CS231n.
https://github.com/tflearn/tflearn/blob/master/examples/notebooks/spiral.ipynb
Layers.Use TFLearn layers along with TensorFlow.
https://github.com/tflearn/tflearn/blob/master/examples/extending_tensorflow/layers.py
Trainer.Use TFLearn trainer class to train any TensorFlow graph.
https://github.com/tflearn/tflearn/blob/master/examples/extending_tensorflow/trainer.py
Built-inOps. Use TFLearn built-in operations along with TensorFlow.
https://github.com/tflearn/tflearn/blob/master/examples/extending_tensorflow/builtin_ops.py
Summaries.Use TFLearn summarizers along with TensorFlow.
https://github.com/tflearn/tflearn/blob/master/examples/extending_tensorflow/summaries.py
Variables.Use TFLearn variables along with TensorFlow.
https://github.com/tflearn/tflearn/blob/master/examples/extending_tensorflow/variables.py