ignite
ignite copied to clipboard
pytorch
High-level library to help with training and evaluating neural networks in PyTorch flexibly and transparently.
Fixes #3294 Description: Adds support for `Mapping` metrics in loggers Check list: - [ ] New tests are added (if a new feature is added) - [ ] New doc...
## 🚀 Feature The request would be to add the logging of `Mapping` metrics in the logging framework. The `ignite.metrics.Metric` class supports the use of `Mapping` metrics as we can...
It would be good to add gradient clipping to the trainers created by `create_supervised_trainer`. This is already provided by `torch.nn.utils.clip_grad_norm_`. One possible implementation could be: ```python import math from torch.nn.utils...
Description: added metrics for evaluating clustering results implemented in scikit-learn: - [silhouette score](https://en.wikipedia.org/wiki/Silhouette_(clustering)) - [Davies-Bouldin score](https://en.wikipedia.org/wiki/Davies%E2%80%93Bouldin_index) - [Calinski-Harabasz score](https://en.wikipedia.org/wiki/Calinski%E2%80%93Harabasz_index) Check list: - [x] New tests are added (if a new...
Fixes #{issue number} Description: Check list: - [ ] New tests are added (if a new feature is added) - [ ] New doc strings: description and/or example code are...
Use uv for other workflows Depends on: - https://github.com/pytorch/ignite/pull/3420
Use astral uv on CI testing. Adds a `uv.lock` which enables the use of cache and also a `uv check` command to check if everything is up to date in...
Update pre-commit file and configuration Depends on: - https://github.com/pytorch/ignite/pull/3418 - https://github.com/pytorch/ignite/pull/3423
## 🚀 Feature I would propose several enhancements to the [EarlyStopping handler](https://docs.pytorch.org/ignite/_modules/ignite/handlers/early_stopping.html#EarlyStopping). 1. Introduce a new parameter, `min_delta_mode`, mimicking the parameter `threshold_mode` of [ReduceLROnPlateau](https://docs.pytorch.org/docs/stable/generated/torch.optim.lr_scheduler.ReduceLROnPlateau.html). If `min_delta_mode="abs"` (default) everything works as...
