PyTorch implementations of research paper model architectures
The torcharch package is divided into the following submodules.
Both models an modules are further divided into broad categories (i.e. convolutional, recurrent, etc)
- models: Complete architecture implementations
- modules: Helper components of various architectures
- data: Utility functions and full dataset loaders
- training: Execute trainings and save model weights
The following list has either some or all of its architecture components and/or datasets implmeneted
Spatial Pyramid Pooling in Deep Convolutional Networks for Visual Recognition
modules.conv.SpatialPyramidPooling