TensorFlow Tutorials

• Final update: 2019. 08. 20.
• All right reserved @ Il Gu Yi 2018

Getting Started

Prerequisites

• TensorFlow above 1.13
  • github.com/tensorflow/models for tensorflow version 1 codes
  • inception_v3 and vgg_19 pretrained models for tensorflow version 1 codes
• Python 3.6 or 3.7
  • numpy, matplotlib, PIL
• Jupyter notebook
• OS X and Linux (Not validated on Windows but probably it would work)

Contents (TF version 2.0 style)

TensorFlow Basic Syntax

• tf.keras and eager.execution
  • 01.tf.keras.ipynb
  • 02.guideline.eager.ipynb
  • 03.tf.data.ipynb

Regression

• Logistic Regression
  • 01.mnist.tf.data.eager.ipynb
  • 02.mnist.model.fit.ipynb

Convolutional Neural Networks

• Simple CNN model (LeNet5) and regularization & batch norm
  • 01.mnist.LeNet5.ipynb
  • 02.mnist.LeNet5.functional.API.ipynb
  • 03.mnist.LeNet5.with.regularization.ipynb
• Load vgg16 and transfer learning with vgg16
  • 04.vgg16.ipynb
  • 05.transfer.learning.ipynb

Recurrent Neural Networks

• Many to one problem (Sentiment classification)
  • 01.rnn.basic.ipynb
  • 02.simple.sequence.classification.ipynb
  • 03.sentiment.classfication.imdb.ipynb
  • 04.text.generation.ipynb

Generative Models

• Generative Adversarial Networks
  • 01.dcgan.ipynb

Contents (TF version 1.x style)

TensorFlow Basic Syntax

• Overview and Operations
  • 01.hello.tf.ipynb
  • 02.tf.eager.ipynb
  • 03.tensorboard.ipynb
  • 04.tf.dimension.ipynb
  • 05.tf.Variable.ipynb
  • 06.tf.placeholder.ipynb
• Managing and tf.data
  • 07.tf.train.Saver.ipynb
  • 08.tf.cond.ipynb
  • 09.tf.control_dependencies.ipynb
  • 10.tf.data.ipynb

Regression

• Linear Regression
  • 01.1.linear.regression.ipynb
  • 01.2.linear.regression.with.minibatch.ipynb
  • 02.1.linear.regression.3rd.order.ipynb
  • 02.2.linear.regression.3rd.order.with.minibatch.ipynb
• Logistic Regression
  • 03.1.mnist.softmax.ipynb
  • 03.2.mnist.softmax.with.tf.data.ipynb
  • 04.mnist.summary.ipynb

Convolutional Neural Networks

• Simple CNN model (LeNet5) and regularization & batch norm
  • 01.1.mnist.deep.with.estimator.ipynb
  • 01.2.mnist.deep.with.tf.data.ipynb
  • 02.mnist.deep.slim.ipynb
  • 03.mnist.slim.options.ipynb
  • 04.mnist.slim.arg.scope.ipynb
• Advanced CNN model (Cifar10) and data augmentation
  • 05.cnn.cifar10.ipynb
  • 06.cnn.cifar10.regularization.ipynb
  • 07.cnn.cifar10.data.augmentation.ipynb
• Pretrained CNN models
  • 08.vgg19.slim.ipynb
  • 09.inception_v3.ipynb
• Transfer learning and tfrecords format
  • 10.tfrecords.ipynb
  • 11.transfer.learning.with.inception_v3.ipynb

Recurrent Neural Networks

• Usage for sequence data
  • 01.ready.for.sequence.data.ipynb
  • 02.rnn.basic.ipynb
• Sequence classification (many to one classification)
  • 03.01.sequence.classification.RNN.ipynb
  • 03.02.sequence.classification.GRU.ipynb
  • 03.03.sequence.classification.LSTM.ipynb
  • 03.04.sequence.classification.biRNN.ipynb
  • 03.05.sequence.classification.biLSTM.ipynb
  • 03.06.sequence.classification.Multi.RNN.ipynb
  • 03.07.sequence.classification.Multi.LSTM.ipynb
  • 03.08.sequence.classification.Multi.RNN.dropout.ipynb
  • 03.09.sequence.classification.Multi.LSTM.dropout.ipynb
  • 03.10.sequence.classification.Multi.biRNN.dropout.ipynb
  • 03.11.sequence.classification.Multi.biLSTM.dropout.ipynb
• Sequence to sequence classification (many to many classification)
  • 04.01.seq2seq.classification.RNN.ipynb
  • 04.02.seq2seq.classification.GRU.ipynb
  • 04.03.seq2seq.classification.LSTM.ipynb
  • 04.04.seq2seq.classification.biRNN.ipynb
  • 04.05.seq2seq.classification.biLSTM.ipynb
  • 04.06.seq2seq.classification.Multi.RNN.ipynb
  • 04.07.seq2seq.classification.Multi.LSTM.ipynb
  • 04.08.seq2seq.classification.Multi.RNN.dropout.ipynb
  • 04.09.seq2seq.classification.Multi.LSTM.dropout.ipynb

Generative Models

• Generative Adversarial Networks
  • 01.dcgan.ipynb

Images

• Image Segmentation with U-Net
  • 01.image_segmentation.ipynb

Sequences

• Sentiment Classification
  • 01.sentiment.classification.ipynb

Author

Il Gu Yi

Lecture Notes

link