Skip to content

Latest commit

 

History

History
12 lines (11 loc) · 702 Bytes

README.md

File metadata and controls

12 lines (11 loc) · 702 Bytes

Supervised-learning

A python library for supervised learning methods with PyTorch.

List of Implementations

This repo has PyTorch implementations of training a ConvNet.

  • Transfer Learning for ResNet18 / ResNet34 / ResNet101 / SqueezeNet
  • Differntial Learning rate (Applying smaller learning rates on earlier layers & higher on later layers for higher performance)
  • SGDR (SGD with Restarts - Accelerates training)
  • CLR (Cyclic Learning Rate - Accelerates training (uses cosine annealing instead of linear))
  • Training a simple 28x28 network from scratch
  • Basic Dataloaders
  • Segmentation Dataloaders (WIP)