Mighty PyTorch Trainer¶

A mighty trainer for your torch models with online progress monitoring. Powered by Visdom.

  • Getting started
    • Installation
    • Start a Visdom server
    • Train a model
  • Trainers
    • Main trainers
    • Idle test “trainer”
    • Interpretable mask trainers
  • Monitors
    • Monitors
    • Mutual Information Estimates
    • Accuracy measures
    • Timer and schedulers
    • Visdom server
  • Loss wrappers
    • mighty.loss.ContrastiveLossSampler
    • mighty.loss.TripletLossSampler
    • mighty.loss.TripletCosineLoss
  • Models
    • mighty.models.MLP
    • mighty.models.AutoencoderLinear
    • mighty.models.SerializableModule
  • Other
    • Mean and variance online measures
    • Data loader with simple API
    • Inverse normalizations of data transforms
    • Finding a layer
    • Working with batches
    • Miscellaneous
    • Layer hooks
    • Convert a model to the train or test mode
    • Signal processing and statistics
    • Stubs
    • Constants

Mighty PyTorch Trainer

Navigation

  • Getting started
  • Trainers
  • Monitors
  • Loss wrappers
  • Models
  • Other

Related Topics

  • Documentation overview
    • Next: Getting started

Quick search

©2020, Danylo Ulianych. | Powered by Sphinx 4.3.1 & Alabaster 0.7.12 | Page source