Sinabs supports fast training of spiking neural networks with PyTorch and inference on neuromorphic hardware.
Open source
24 alternativesignite
PyTorch Ignite provides a high level library for flexibly training and evaluating neural networks.
About ignite
PyTorch Ignite is an open source Python library for training and evaluating neural networks in PyTorch. It provides an engine and event system, so users instantiate engines and run them instead of writing epoch and iteration loops. Handlers can be attached to events, and events can be filtered, stacked, or defined as custom events.
It also includes out-of-the-box metrics and built-in handlers for composing a training pipeline, saving artifacts, and logging parameters and metrics. The project describes the library as requiring less code than pure PyTorch while keeping control over the training process.
Key features
- Engine and event system that replaces manual epoch and iteration loops
- Event handlers that can be any function, including lambdas or class methods
- Built-in event filtering, such as running a handler every five epochs
- Stackable events and custom events beyond the standard set
- Out-of-the-box metrics, such as Accuracy, for evaluating models
- Built-in handlers for saving artifacts and logging parameters and metrics
Best for: PyTorch users who train and evaluate neural networks and want less boilerplate code while keeping control over each step.
Good for
- Handlers can be plain functions, so code does not need to inherit from an interface or override abstract methods.
- Users instantiate engines and run them, without writing epoch and iteration loops.
- Built-in event filtering lets handlers run at set epochs or iterations.
ignite at a glance
- Website
- github.com/pytorch/ignite
- Type
- Open-source project
- Pricing
- Not confirmed. Check the site
- Open source
- Yes · Source code
- Language
- English
- Last checked
- Oct 6, 2026 · Online when checked
- Listed
- Oct 5, 2026
ignite alternatives
Tools and sites you could use instead, ranked by how closely they match.
Rockpool supports training spiking neural networks with PyTorch and JAX and deploying them to neuromorphic hardware.
Open source
24 alternatives
PaddlePaddle provides a deep learning framework for training models across devices and platforms.
Open source
24 alternatives
Tensors and Dynamic neural networks in Python with strong GPU acceleration.
PyTorch TNT supplies lightweight tools and utilities for training with PyTorch.
The fuku-ml project provides a machine learning library described as simple to use.
Open source
24 alternativesPopular in AI Infrastructure & MLOps
Well-known listings from the same topic.
Guild AIguild.aiFlow AIflow-ai.comsystemprompt.iosystemprompt.ioKili Technologykili-technology.comAPI docsdocs.x.aiEyePop.aieyepop.ai
ignite FAQ
Do I still need to write epoch and iteration loops?
No. Users instantiate engines and run them. The project states that there is no need to write for or while loops over epochs and iterations.
What kinds of functions can be used as event handlers?
Handlers can be any function, such as a lambda, a simple function, or a class method. They do not need to inherit from an interface.
How can a handler run only at certain epochs?
Built-in event filtering supports this. For example, Events.EPOCH_COMPLETED(every=5) runs a handler every five epochs, and Events.EPOCH_STARTED(once=20) runs it once on the 20th epoch.
Is something wrong or out of date?
Tell us and an editor will check it.
Own this website?
Claim this listing
Claim the listing and suggest updates yourself. An editor checks every claim first.