What's Changed
- Migrate to the 'locker' GitHub action for locking closed/stale issues/PRs by @jeffhandley in #6896
- Update release-3.0.0.md by @taeerhebend in #6895
- Add support for Apache.Arrow.Types.TimestampType to DataFrame by @asmirnov82 in #6871
- Update developer-guide.md by @computerscienceiscool in #6870
- Issue #6606 - Add sample variance and standard deviation to NormalizeMeanVariance by @tearlant in #6885
- Reorganize dataframe files by @asmirnov82 in #6872
- Updated ml.net versioning by @michaelgsharp in #6907
- Don't include the SDK in our helix payload by @ericstj in #6918
- Rename NameEntity to NamedEntity by @ericstj in #6917
- Fix assert by only accessing idx by @ericstj in #6924
- Make double assertions compare with tolerance instead of precision by @ericstj in #6923
- [main] Update dependencies from dotnet/arcade by @dotnet-maestro in #6703
- Only use semi-colons for NoWarn - fixes build break by @ericstj in #6935
- Fixes NER to correctly expand/shrink the labels by @michaelgsharp in #6928
- Split out non concurrent test collections. by @michaelgsharp in #6937
- Add Backport github workflow by @ericstj in #6944
- fix #6949 by @LittleLittleCloud in #6951
- Torch sharp version updates and test fixes by @michaelgsharp in #6954
- [main] Update dependencies from dotnet/arcade by @dotnet-maestro in #6957
- Added in release notes for 3.0.1 by @michaelgsharp in #6962
- Update .NET Runtimes to latest version by @ericstj in #6964
- Add sweepable estimator to NER by @zewditu in #6965
- Testing light gbm bad allocation by @michaelgsharp in #6968
- [main] Update dependencies from dotnet/arcade by @dotnet-maestro in #6969
- Add new type to key-value converter by @zewditu in #6973
- [main] Update dependencies from dotnet/arcade by @dotnet-maestro in #6976
- Introducing Tiktoken Tokenizer by @tarekgh in #6981
- [main] Update dependencies from dotnet/arcade by @dotnet-maestro in #6985
- FabricBot: Onboarding to GitOps.ResourceManagement because of FabricBot decommissioning by @dotnet-policy-service in #6983
- Add text normalizer transformer to AutoML by @zewditu in #6998
- Tokenizer's Interfaces Cleanup by @tarekgh in #7001
- First round of perf improvements for tiktoken by @stephentoub in #7012
- Avoid LruCache in Tiktoken when cacheSize specified is 0 by @stephentoub in #7016
- Tweak Tiktoken's BytePairEncode for improved perf by @stephentoub in #7017
- Tweak CreateByModelNameAsync by @stephentoub in #7015
- Match SweepableEstimatorFactory name with Ml.net name. by @zewditu in #7007
- Update OnnxRuntime to 1.16.3 by @michaelgsharp in #6975
- Temp fix for the race condition during the tests. by @michaelgsharp in #7021
- Fix formatting that fails in VS by @michaelgsharp in #7023
- Optimize regexes used in tiktoken by @stephentoub in #7020
- Address the feedback on the tokenizer's library by @tarekgh in #7024
- Make MlImage tests not block file for reading by @michaelgsharp in #7029
- Packaging cleanup by @ericstj in #6939
- Remove SourceLink SDK references by @ericstj in #7037
- Add Span support in tokenizer's Model abstraction by @tarekgh in #7035
- [main] Update dependencies from dotnet/arcade by @dotnet-maestro in #6995
- Change official build to use 1ES templates by @directhex in #7048
- Adding needed Tokenizer's APIs by @tarekgh in #7047
- Auto-generated baselines by 1ES Pipeline Templates by @directhex in #7051
- testing arm python brew overwite by @michaelgsharp in #7058
- Update package versions in use by ML.NET tests by @ericstj in #7055
New Contributors
- @taeerhebend made their first contribution in #6895
- @computerscienceiscool made their first contribution in #6870
- @tearlant made their first contribution in #6885
- @dotnet-policy-service made their first contribution in #6983
- @directhex made their first contribution in #7048
Full Changelog: v3.0.1...v4.0.0-preview1