apallath / torch_models

Basic deep learning models in PyTorch.

Deep Learning code

Regression and classification using Pytorch models (torch_FullyConnected)

• Jupyter notebook

(Adapted from my solutions to homework problems for ENM531: Data Driven Modelling at Penn Engineering)

Pytorch implementation of physics-informed neural network to solve PDE (torch_PINN)

Instructions:

• Run: python pinn_train.py <args>
• For help with args: python pinn_train.py --help
• Edit pinn_plot_error.py and run to plot comparison of L2-norm error from true solution for different training set sizes

(Adapted from my solutions to homework problems for ENM531: Data Driven Modelling at Penn Engineering)

Pytorch implementation of CNN to classify cell images (torch_CNN_cell_images)

Instructions:

• Run: python create_datasets.py
• Run: python cnn_train.py <args>
• Run: python cnn_plots.py <args>
• For help with args, use flag --help

(Adapted from my solutions to homework problems for ENM531: Data Driven Modelling at Penn Engineering)

Pytorch implementation of an RNN and LSTM to learn pattern in Lotka-Volterra equation solution (torch_LSTM_Lotka_Volterra)

Instructions:

• Set Lotka-Volterra equations parameters in rnn_train_test.py
• Run: python train_test.py

(Adapted from my solutions to homework problems for ENM531: Data Driven Modelling at Penn Engineering)

Pytorch implementation of an autoencoder on the MNIST dataset (torch_autoencoder_MNIST/)

Instructions:

• Run: python train_autoencoder.py <args>
• For help with args: python train_autoencoder.py --help

Pytorch implementation of a variational autoencoder on the MNIST dataset (torch_vae_MNIST/)

TODO