PyTorch Ignite Files

What's Changed

• Temporary pinned mlflow version to fix the failing CI by @vfdev-5 in https://github.com/pytorch/ignite/pull/3278
• Add rank correlation metrics by @kzkadc in https://github.com/pytorch/ignite/pull/3276
• Helper function all_gather_tensors_with_shapes() by @sadra-barikbin in https://github.com/pytorch/ignite/pull/3281
• Fix mypy issue with wandb.finish() by @vfdev-5 in https://github.com/pytorch/ignite/pull/3283
• To fix [#3277] by @puhuk in https://github.com/pytorch/ignite/pull/3280
• Modify docstring and argument for rank correlation metrics by @kzkadc in https://github.com/pytorch/ignite/pull/3284
• Use the default runner type by @ZainRizvi in https://github.com/pytorch/ignite/pull/3285
• [CI] Fixed failing linkcheck, w&b init by @vfdev-5 in https://github.com/pytorch/ignite/pull/3287
• add HSIC metric by @kzkadc in https://github.com/pytorch/ignite/pull/3282
• Drop python 3.8 in the CI by @vfdev-5 in https://github.com/pytorch/ignite/pull/3288
• Fix pytorch versions ci by @vfdev-5 in https://github.com/pytorch/ignite/pull/3289
• Modify docstring for rank correlation metrics by @kzkadc in https://github.com/pytorch/ignite/pull/3291
• simplify imports of metric functions by @kzkadc in https://github.com/pytorch/ignite/pull/3292
• Update import style for PrecisionRecallCurve metric by @kzkadc in https://github.com/pytorch/ignite/pull/3293
• Fixed mypy failures on the CI by @vfdev-5 in https://github.com/pytorch/ignite/pull/3296
• Updated code formatting tools' versions by @vfdev-5 in https://github.com/pytorch/ignite/pull/3297
• Add clustering metrics by @kzkadc in https://github.com/pytorch/ignite/pull/3290
• Updated pytorch version in CI and docker by @vfdev-5 in https://github.com/pytorch/ignite/pull/3298
• Fixed torch version in docker push_all.sh by @vfdev-5 in https://github.com/pytorch/ignite/pull/3300
• Fixed failing tests on pytorch nightly using torch.load by @vfdev-5 in https://github.com/pytorch/ignite/pull/3299
• try updating gpu tests GHA by @vfdev-5 in https://github.com/pytorch/ignite/pull/3306
• Update pyproject.toml by @vfdev-5 in https://github.com/pytorch/ignite/pull/3305
• Fix pynvml version to <12 by @vfdev-5 in https://github.com/pytorch/ignite/pull/3310
• Fixed GPU tests exec scripts and failing metrics by @vfdev-5 in https://github.com/pytorch/ignite/pull/3301
• Updated GpuInfo metric, pynvml<12 by @vfdev-5 in https://github.com/pytorch/ignite/pull/3311
• Give the option to terminate the engine without firing Events.COMPLET… by @bonassifabio in https://github.com/pytorch/ignite/pull/3309
• Allow to terminate an epoch without firing Events.EPOCH_COMPLETED by @bonassifabio in https://github.com/pytorch/ignite/pull/3313
• Fix deprecated statement by @Gulin7 in https://github.com/pytorch/ignite/pull/3307
• Fixed failing test_roc_auc.py::test_check_compute_fn test by @vfdev-5 in https://github.com/pytorch/ignite/pull/3316
• Update pytorch-version-tests.yml by @vfdev-5 in https://github.com/pytorch/ignite/pull/3315
• Fix typos in TextCNN.jpynb by @k22036 in https://github.com/pytorch/ignite/pull/3320
• Fix typos in FashionMNIST.jpynb by @k22036 in https://github.com/pytorch/ignite/pull/3321
• Refactor FashionMNIST notebook to use next() for data iterator by @k22036 in https://github.com/pytorch/ignite/pull/3322
• Fixed failing doctest and linkcheck CI jobs by @vfdev-5 in https://github.com/pytorch/ignite/pull/3331
• Fixed failing test_miou on M1 with pytest.approx by @vfdev-5 in https://github.com/pytorch/ignite/pull/3330
• Updated pytorch version in docker and pytorch-version-tests.yml by @vfdev-5 in https://github.com/pytorch/ignite/pull/3329
• Fix a typo in the documentation by @jscanvic in https://github.com/pytorch/ignite/pull/3332
• COCO mAP metric by @sadra-barikbin in https://github.com/pytorch/ignite/pull/2901
• Updated accuracy, precision, recall tests by @vfdev-5 in https://github.com/pytorch/ignite/pull/3333
• Fixed failing mean_average_precision tests on cuda by @vfdev-5 in https://github.com/pytorch/ignite/pull/3338
• Use f32 for metrics on mps by @vfdev-5 in https://github.com/pytorch/ignite/pull/3334
• Update test_average_precision.py to resolve issue [#3335] by @puhuk in https://github.com/pytorch/ignite/pull/3337
• Fix group as list of ints in torch dist collective ops by @vfdev-5 in https://github.com/pytorch/ignite/pull/3340
• Added available_device fixture to test_mean_squared_error by @HyeSungP in https://github.com/pytorch/ignite/pull/3341
• Add _rank_not_in_group to idist by @sadra-barikbin in https://github.com/pytorch/ignite/pull/3339
• Update test_roc_auc.py to resolve issue [#3335] by @puhuk in https://github.com/pytorch/ignite/pull/3343
• Update test_cosine_similarity.py to resolve issue [#3335] by @chaojessica in https://github.com/pytorch/ignite/pull/3348
• Added available_device in test_classification_report (#3335) by @HyeSungP in https://github.com/pytorch/ignite/pull/3342
• Run hvd tests without xdist by @vfdev-5 in https://github.com/pytorch/ignite/pull/3347
• Updated gpu-tests CI by @vfdev-5 in https://github.com/pytorch/ignite/pull/3349
• adds argument name for device in test_mean_squared_error [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3351
• adds checks for used device in metrics tests. [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3353
• Added available_device in metric tests (#3335) by @HyeSungP in https://github.com/pytorch/ignite/pull/3354
• adds available_device to test_cohen_kappa.py [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3355
• adds available_device to test_confusion_matrix.py [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3357
• Add SSIM for 3D tensors by @lrlunin in https://github.com/pytorch/ignite/pull/3345
• adds available_device to test_entropy.py [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3358
• adds available_device to test_kl_divergence.py [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3363
• adds available_device to test_frequency.py [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3359
• adds available_device to test_hsic.py [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3361
• adds available_device to test_js_divergence.py [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3362
• adds available_device to test_loss.py [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3364
• adds available_device to test_maximum_mean_discrepancy [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3365
• adds available_device to test_multilabel_confusion_matrix [#3335] by @BanzaiTokyo in https://github.com/pytorch/ignite/pull/3366
• Added support for logging dicts/iterables of metrics by @vfdev-5 in https://github.com/pytorch/ignite/pull/3369

Acknowledgments

🎉 Thanks to our community and all our contributors for the issues, PRs and 🌟 ⭐️ 🌟 ! 💯 We really appreciate your implication into the project

New Contributors

• @ZainRizvi made their first contribution in https://github.com/pytorch/ignite/pull/3285
• @bonassifabio made their first contribution in https://github.com/pytorch/ignite/pull/3309
• @Gulin7 made their first contribution in https://github.com/pytorch/ignite/pull/3307
• @k22036 made their first contribution in https://github.com/pytorch/ignite/pull/3320
• @jscanvic made their first contribution in https://github.com/pytorch/ignite/pull/3332
• @HyeSungP made their first contribution in https://github.com/pytorch/ignite/pull/3341
• @chaojessica made their first contribution in https://github.com/pytorch/ignite/pull/3348
• @lrlunin made their first contribution in https://github.com/pytorch/ignite/pull/3345

Full Changelog: https://github.com/pytorch/ignite/compare/v0.5.1...v0.5.2