Implementations of neural network models with tf (>=2.0)
See also implementations with PyTorch 1.0 here.
- TensorFlow >= 2.0
$ pip install tensorflow
- Logistic Regression
- MLP
- LeNet
- ResNet (ResNet34, ResNet50)
- DenseNet (DenseNet121)
- Encoder-Decoder (LSTM)
- EncoderDecoder (Attention)
- Transformer
- Deep Q-Network
- Variational Autoencoder
- Generative Adversarial Network
models/
├── densenet121_cifar10_beginner.py
├── dqn_cartpole.py
├── encoder_decoder_attention.py
├── encoder_decoder_lstm.py
├── gan_fashion_mnist.py
├── lenet_mnist.py
├── lenet_mnist_beginner.py
├── logistic_regression_mnist.py
├── logistic_regression_mnist_beginner.py
├── mlp_mnist.py
├── mlp_mnist_beginner.py
├── resnet34_fashion_mnist.py
├── resnet34_fashion_mnist_beginner.py
├── resnet50_fashion_mnist.py
├── resnet50_fashion_mnist_beginner.py
├── transformer.py
├── vae_fashion_mnist.py
│
└── layers/
├── Attention.py
├── DotProductAttention.py
├── LayerNormalization.py
├── MultiHeadAttention.py
├── PositionalEncoding.py
└── ScaledDotProductAttention.py
*_beginner.py is the file using only Keras.