{"full_name":"microsoft/nni","owner":"microsoft","name":"nni","shard":"1","url":"https://github.com/microsoft/nni","description":"An open source AutoML toolkit for automate machine learning lifecycle, including feature engineering, neural architecture search, model compression and hyper-parameter tuning.","homepage":"https://nni.readthedocs.io","topics":["automl","deep-learning","neural-architecture-search","hyperparameter-optimization","distributed","bayesian-optimization","automated-machine-learning","machine-learning","machine-learning-algorithms","data-science","tensorflow","pytorch","neural-network","deep-neural-network","model-compression"],"tag_pages":["automl","data-science","deep-learning","deep-neural-network","distributed","machine-learning","machine-learning-algorithms","neural-network","pytorch","tensorflow"],"language":"Python","license":"MIT","avatar":"https://avatars.githubusercontent.com/u/6154722?v=4","created_at":"2018-06-01T05:51:44Z","pushed_at":"2024-07-03T10:55:10Z","metrics":{"stars":14366,"forks":1850,"watchers":8,"issues":2078,"prs":3492,"releases":40,"issues_open":397,"issues_closed":1681,"prs_open":18,"prs_closed":486,"prs_merged":2988,"discussions":161},"growth":{"day":{"stars":-1,"forks":-1,"base_day":"2026-11-15"},"week":{"stars":0,"forks":0,"base_day":"2026-11-04"},"month":{"stars":1,"forks":1,"base_day":"2026-10-17"}},"history":{"dates":["2026-09-06","2026-09-07","2026-09-08","2026-09-09","2026-09-10","2026-09-11","2026-09-12","2026-09-13","2026-09-14","2026-09-15","2026-09-16","2026-09-17","2026-09-18","2026-09-19","2026-09-20","2026-09-21","2026-09-22","2026-09-23","2026-09-24","2026-09-25","2026-09-26","2026-09-27","2026-09-28","2026-09-29","2026-09-30","2026-10-01","2026-10-02","2026-10-03","2026-10-04","2026-10-05","2026-10-06","2026-10-07","2026-10-08","2026-10-09","2026-10-10","2026-10-11","2026-10-12","2026-10-13","2026-10-14","2026-10-15","2026-10-16","2026-10-17","2026-10-18","2026-10-19","2026-10-20","2026-10-21","2026-10-22","2026-10-23","2026-10-24","2026-10-25","2026-10-26","2026-10-27","2026-10-28","2026-10-29","2026-10-30","2026-10-31","2026-11-01","2026-11-02","2026-11-03","2026-11-04","2026-11-10","2026-11-11","2026-11-12","2026-11-13","2026-11-14","2026-11-15","2026-11-16"],"stars":[14360,14359,14359,14359,14358,14360,14360,14360,14360,14360,14360,14361,14361,14360,14359,14360,14362,14363,14363,14363,14362,14362,14362,14363,14364,14364,14366,14366,14367,14368,14367,14367,14367,14366,14367,14368,14367,14366,14366,14366,14366,14365,14365,14367,14367,14366,14367,14367,14367,14365,14366,14366,14367,14367,14367,14367,14366,14366,14366,14366,14362,14365,14365,14365,14365,14367,14366],"forks":[1851,1850,1850,1849,1849,1849,1849,1849,1849,1849,1849,1848,1848,1847,1847,1847,1847,1848,1848,1848,1848,1848,1848,1848,1848,1849,1850,1850,1849,1850,1850,1850,1850,1849,1849,1849,1849,1849,1850,1850,1849,1849,1849,1849,1849,1850,1850,1850,1850,1850,1850,1850,1850,1850,1850,1850,1850,1850,1850,1850,1849,1850,1851,1851,1850,1851,1850],"watchers":[9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,9,8,8,8,8,8,8,8,8,8,8,8,null,8,8,8,8,8,8],"issues_open":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,397,397,397,397,397,397],"prs_open":[null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,null,18,18,18,18,18,18]},"boards":{},"same_owner":[{"full_name":"microsoft/markitdown","name":"markitdown","shard":"1","boards":{"daily":151,"weekly":176,"monthly":103,"total":36}},{"full_name":"microsoft/vscode","name":"vscode","shard":"1","boards":{"daily_watchers":21,"total":37}},{"full_name":"microsoft/PowerToys","name":"PowerToys","shard":"1","boards":{"daily_watchers":111,"total":81}},{"full_name":"microsoft/generative-ai-for-beginners","name":"generative-ai-for-beginners","shard":"1","boards":{"daily_forks":40,"total":98}},{"full_name":"microsoft/TypeScript","name":"TypeScript","shard":"1","boards":{"daily_watchers":120,"total":120}},{"full_name":"microsoft/terminal","name":"terminal","shard":"1","boards":{"total":133}},{"full_name":"microsoft/playwright","name":"playwright","shard":"1","boards":{"total":143}},{"full_name":"microsoft/Web-Dev-For-Beginners","name":"Web-Dev-For-Beginners","shard":"1","boards":{"total":155}},{"full_name":"microsoft/ML-For-Beginners","name":"ML-For-Beginners","shard":"1","boards":{"total":173}},{"full_name":"microsoft/AI-For-Beginners","name":"AI-For-Beginners","shard":"1","boards":{}},{"full_name":"microsoft/AirSim","name":"AirSim","shard":"1","boards":{}},{"full_name":"microsoft/BitNet","name":"BitNet","shard":"1","boards":{}},{"full_name":"microsoft/Bringing-Old-Photos-Back-to-Life","name":"Bringing-Old-Photos-Back-to-Life","shard":"1","boards":{}},{"full_name":"microsoft/CNTK","name":"CNTK","shard":"1","boards":{}},{"full_name":"microsoft/Data-Science-For-Beginners","name":"Data-Science-For-Beginners","shard":"1","boards":{}},{"full_name":"microsoft/FluidFramework","name":"FluidFramework","shard":"1","boards":{}},{"full_name":"microsoft/IoT-For-Beginners","name":"IoT-For-Beginners","shard":"1","boards":{}},{"full_name":"microsoft/MMdnn","name":"MMdnn","shard":"1","boards":{}},{"full_name":"microsoft/Ontology-Playground","name":"Ontology-Playground","shard":"1","boards":{}},{"full_name":"microsoft/Procexp-Mac","name":"Procexp-Mac","shard":"1","boards":{"daily_rate":96,"weekly_rate":32,"monthly_rate":87}}],"issues":[{"number":3496,"title":"Failed to establish a new connection","comments":34,"created_at":"2021-03-29T11:49:29Z"},{"number":4945,"title":"Supported for torch.sum","comments":12,"created_at":"2022-06-17T07:14:28Z"},{"number":4542,"title":"Doubts about NAS+TypeError: tuple is not a Module subclass","comments":22,"created_at":"2022-02-10T19:21:04Z"},{"number":5421,"title":"TypeError: forward() takes 2 positional arguments but 6 were given","comments":21,"created_at":"2023-03-06T03:00:45Z"},{"number":3814,"title":"[Retiarii] migrate FBNet from NAS 1.0 to Retiarii framework","comments":21,"created_at":"2021-06-11T08:13:40Z"},{"number":5614,"title":"Cannot access https://nni.blob.core.windows.net/nashub/darts-v2-5465b0d2.pth for pretrained model","comments":20,"created_at":"2023-06-20T10:27:52Z"},{"number":5417,"title":"AssertionError, while speeding up Convnext model. ","comments":17,"created_at":"2023-03-02T13:39:23Z"},{"number":5185,"title":" assert len(args) >= len(self.undetermined)  AssertionError","comments":13,"created_at":"2022-10-26T08:35:46Z"}],"discussions":[{"number":5789,"title":"I wanna know how to use python to output the best hyperparameters generating by NNI","comments":0,"category":"Q&A","created_at":"2024-05-29T05:43:12Z"},{"number":5787,"title":"Tensorflow support","comments":0,"category":"General","created_at":"2024-05-22T03:55:31Z"},{"number":5786,"title":"About How to Save the Pruned Model","comments":0,"category":"Q&A","created_at":"2024-05-19T07:15:14Z"},{"number":5785,"title":"Training services URL(error 404) for references in readme.md.","comments":0,"category":"Q&A","created_at":"2024-05-18T15:11:38Z"},{"number":5784,"title":"What is a good value for trialConcurrency config parameter?","comments":0,"category":"Q&A","created_at":"2024-05-16T17:49:15Z"},{"number":5781,"title":"Is NNI stop developing?","comments":0,"category":"General","created_at":"2024-05-13T12:42:02Z"},{"number":5756,"title":"Verification on the GP tuner accuracy when perform HPO/Save GP tuner iteratively","comments":1,"category":"Q&A","created_at":"2024-03-13T06:27:04Z"},{"number":5753,"title":"nni GELU is not registered, please register setting with QuantizationSetting.register().","comments":1,"category":"Q&A","created_at":"2024-03-07T14:27:26Z"}],"latest_release":{"name":"Release 3.0 - 21/8/2023","tag":"v3.0","published_at":"2023-09-14T12:12:06Z","body":"## Web Portal ##\r\n\r\n* New look and feel\r\n\r\n## Neural Architecture Search ##\r\n\r\n* **Breaking change**: ``nni.retiarii`` is no longer maintained and tested. Please migrate to ``nni.nas``.\r\n  + Inherit ``nni.nas.nn.pytorch.ModelSpace``, rather than use ``@model_wrapper``.\r\n  + Use ``nni.choice``, rather than ``nni.nas.nn.pytorch.ValueChoice``.\r\n  + Use ``nni.nas.experiment.NasExperiment`` and ``NasExperimentConfig``, rather than ``RetiariiExperiment``.\r\n  + Use ``nni.nas.model_context``, rather than ``nni.nas.fixed_arch``.\r\n  + Please refer to [quickstart](https://nni.readthedocs.io/en/v3.0pt1/tutorials/hello_nas.html)  for more changes.\r\n* A refreshed experience to construct model space.\r\n  + Enhanced debuggability via ``freeze()`` and ``simplify()`` APIs.\r\n  + Enhanced expressiveness with ``nni.choice``, ``nni.uniform``, ``nni.normal`` and etc.\r\n  + Enhanced experience of customization with ``MutableModule``, ``ModelSpace`` and ``ParamterizedModule``.\r\n  + Search space with constraints is now supported.\r\n* Improved robustness and stability of strategies.\r\n  + Supported search space types are now enriched for PolicyBaseRL, ENAS and Proxyless.\r\n  + Each step of one-shot strategies can be executed alone: model mutation, evaluator mutation and training.\r\n  + Most multi-trial strategies now supports specifying seed for reproducibility.\r\n  + Performance of strategies have been verified on a set of benchmarks.\r\n* Strategy/engine middleware.\r\n  + Filtering, replicating, deduplicating or retrying models submitted by any strategy.\r\n  + Merging or transforming models before executing (e.g., CGO).\r\n  + Arbitrarily-long chains of middlewares.\r\n* New execution engine.\r\n  + Improved debuggability via SequentialExecutionEngine: trials can run in a single process and breakpoints are effective.\r\n  + The old execution engine is now decomposed into execution engine and model format.\r\n  + Enhanced extensibility of execution engines.\r\n* NAS profiler and hardware-aware NAS.\r\n  + New profilers profile a model space, and quickly compute a profiling result for a sampled architecture or a distribution of architectures (FlopsProfiler, NumParamsProfiler and NnMeterProfiler are officially supported).\r\n  + Assemble profiler with arbitrary strategies, including both multi-trial and one-shot.\r\n  + Profiler are extensible. Strategies can be assembled with arbitrary customized profilers.\r\n\r\n## Model Compression ##\r\n\r\n* Compression framework is refactored, new framework import path is ``nni.contrib.compression``.\r\n  + Configure keys are refactored, support more detailed compression configurations. [view doc](https://nni.readthedocs.io/en/v3.0pt1/compression/config_list.html)\r\n  + Support multi compression methods fusion.\r\n  + Support distillation as a basic compression component.\r\n  + Support more compression targets, like ``input``, ``ouptut`` and any registered paramters.\r\n  + Support compressing any module type by customizing module settings.\r\n* Model compression support in DeepSpeed mode.\r\n* Fix example bugs.\r\n* Pruning\r\n  + Pruner interfaces have fine-tuned for easy to use. [view doc](https://nni.readthedocs.io/en/v3.0pt1/reference/compression/pruner.html)\r\n  + Support configuring ``granularity`` in pruners.  [view doc](https://nni.readthedocs.io/en/v3.0pt1/compression/config_list.html#granularity)\r\n  + Support different mask ways, multiply zero or add a large negative value.\r\n  + Support manully setting dependency group and global group. [view doc](https://nni.readthedocs.io/en/v3.0/compression/config_list.html#global-group-id)\r\n  + A new powerful pruning speedup is released, applicability and robustness have been greatly improved. [view doc](https://nni.readthedocs.io/en/v3.0pt1/reference/compression/pruning_speedup.html)\r\n  + The end to end transformer compression tutorial has been updated, achieved more extreme compression performance. [view doc](https://nni.readthedocs.io/en/v3.0pt1/tutorials/new_pruning_bert_glue.html)\r\n  + Fix config list in the examp","is_prerelease":false},"readme":true,"readme_zh":false,"news":[],"stale":false,"snapshot_at":"2026-11-16T09:23:14Z","history_start":"2026-09-06"}