* Update dependencies from https://github.com/dotnet/arcade build 20231220.2
Microsoft.DotNet.Arcade.Sdk , Microsoft.DotNet.Build.Tasks.Feed , Microsoft.DotNet.Helix.Sdk , Microsoft.DotNet.SignTool , Microsoft.DotNet.SwaggerGenerator.MSBuild , Microsoft.DotNet.XUnitExtensions
From Version 8.0.0-beta.23265.1 -> To Version 8.0.0-beta.23620.2
* Fixed version update breaks.
* Update XUnitVersion
* Update MicrosoftMLOnnxRuntimeVersion to 1.16.3
* Rollback OnnxRuntime and suppress warning
* Update to Xunit with fix for https://github.com/xunit/xunit/issues/2821
* Update Centos docker containers
* Fix packaging step
* Try including stdint.h to fix missing uint8_t on centos
* Update Centos test queue
* Attempt to use runtime centos-stream8-helix container for tests
* Use centos-stream8-mlnet-helix container for testing
* Undo changes to test data
* Make NETFRAMEWORK ifdef versionless
* Only use semi-colons for NoWarn
* Fix assert by only accessing idx (#6924)
Asserting on `_rowCount < Utils.Size(_valueBoundaries)` was catching a
case where `_rowCount`'s update was reordered before `_valueBoundaries`
This was unnecessary, since this method doesn't need to use `_rowCount`.
Instead, make the asserts use only `idx` which will be maintained
consistent with the waiter logic in this cache.
Ensure we only ever use `_rowCount` from the caching thread, so write
reordering won't matter.
* Don't include the SDK in our helix payload (#6918)
* Don't include the SDK in our helix payload
I noticed that the tests included the latest SDK - including the host -
in our helix payloads.
This is a large amount of unnecessary downloads and it also makes it so
we use the latest host on the older frameworks which can fail when the
latest host drops support for distros.
Since our tests shouldn't need the full CLI, remove this from our helix
payloads.
We'll instead get just the runtime we need through `AdditionalDotNetPackage`
* Place Helix downloaded runtime on the PATH
Helix only sets the path when the CLI is included, however we don't
need the CLI.
* Make double assertions compare with tolerance instead of precision (#6923)
Precision might cause small differences to round to a different number.
Instead compare with a tolerance which is not sensitive to rounding.
---------
Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
Co-authored-by: Michael Sharp <misharp@microsoft.com>
Co-authored-by: Eric StJohn <ericstj@microsoft.com>
* Update dependencies from https://github.com/dotnet/arcade build 20220923.4
Microsoft.DotNet.Arcade.Sdk , Microsoft.DotNet.Build.Tasks.Feed , Microsoft.DotNet.Helix.Sdk , Microsoft.DotNet.SignTool , Microsoft.DotNet.SwaggerGenerator.MSBuild , Microsoft.DotNet.XUnitExtensions
From Version 7.0.0-beta.22327.2 -> To Version 7.0.0-beta.22473.4
* Update XUnit version
* Workaround CS0121 with Xunit.Assert(float,float,int)
Workaround issue https://github.com/xunit/xunit/issues/2393
* Fix a couple more CS0121 XUnit issues
* Fix more xunit issues
* Workaround anoter ref parameter issue.
* Fix more XUnit warnings
* Update XUnitRunnerVersion to 2.4.2
* Update one more Xunit reference
* Update dependencies from https://github.com/dotnet/arcade build 20221007.2
Microsoft.DotNet.Arcade.Sdk , Microsoft.DotNet.Build.Tasks.Feed , Microsoft.DotNet.Helix.Sdk , Microsoft.DotNet.SignTool , Microsoft.DotNet.SwaggerGenerator.MSBuild , Microsoft.DotNet.XUnitExtensions
From Version 7.0.0-beta.22327.2 -> To Version 7.0.0-beta.22507.2
* Update to RC2 SDK to get FSharp fix
* Extend timeout for code-coverage builds
Co-authored-by: dotnet-maestro[bot] <dotnet-maestro[bot]@users.noreply.github.com>
Co-authored-by: Eric StJohn <ericstj@microsoft.com>
* Arm helix
* yml cross fixes.
* yml cross fixes.
* yml cross fixes.
* yml cross fixes.
* yml cross testing
* Adding docker again.
* queue changes.
* queue changes.
* queue changes.
* arm pre commands
* arm pre commands
* arm pre commands
* arm pre commands
* arm pre commands
* arm pre commands
* arm pre commands
* Arm helix changes.
* Arm helix changes.
* Removing arm so PR can go.
* FIxes from PR comments. Adding arm32 back.
* Fix using statements.
* Namespace fixes.
* Added missed cmake code.
* Cmake fixes
* M1 Additions
* Changes from PR comments.
* x86x64 fact attribute.
* Fixes from PR comments.
* Fixes from PR comments.
* Fixes for Centos 7 fact and Featurizers fact.
* Code Analysis version and Not fail on M1.
* disable m1 tests in prs.
* disable m1 tests in prs.
* disable m1 tests in prs.
* disable m1 tests in prs.
* Change to full url
* Look in osx-arm64 folder for baseline files.
* Fixed -mthumb flag issue.
* Fixes from PR comments.
* Fixes from PR comments.
* Final test fixed.
* Final Fixes from PR comments.
* Refactored CreateStratificationColumn and EnsureGroupPreservationColumn, and some code on GetSplitColumn, into a new method called CreateSplitColumn
* Now "splitColumn" is the name of the temporary column we create for splitting, regardless if it's based-off a "samplingKeyColumn" (ML.NET) or a "stratificationColumn" (NimbusML, Maml, legacy TLC naming conventions...)
* Fixed issues related to dropping or not dropping the splitColumn in some cases
* support more types but tests failed
* fix bugs
* bump to ort1.3 pre-release
* correct/skip some tests
* refactor tests
* fix HashKey8V2
* This reverts commit e0c
* add comments
* revert changes on V1
* upgrade to ORT1.3 official
* fix part of conflits
* update
* update
* update
* fix test failure
* fix another test failure
* Add "Combine" option.
* Add tests.
* Add tests, and change data splitting APIs to use the new functionality of HashingEstimator.
* Address code review comment, change missing value handling and add a test for it.
* Fix hashing tests.
* Add test for backwards compatibility of model format.
* Fix build breaks and tests after rebase.
* Combine V2 and V1 implementations of IHashers into one.
* Update SaveAsOnnx.
* fix baseline
* More unit test fixes
* Code review comments.
* Add back the back-compat test.
* Update cookbook to latest API
* PR feedback
* Add cross reference to new section
* Align naming in cookbook samples tests with changes
* Add cookbook samples test for WithOnFitDelegate
* remove flaky test from CI
* refine text
* remove timeout setting
* also remove benchmarktest from inner loop
* fix skip theory
* fix define constants
* skip benchmark properly
* skip tests in compile time
* skip one more tensorflow test from CI
* extract TestFrameworkCommon project and remove reference to TestFramework project from Functional.Test project
* update on TestFrameworkCommon
* update on TestFrameworkCommon project
* minor fix
* fix reference
* fix project reference
* fix CI build error
* sync latest code and resolve
* take comments and made change
* take comments and fix
* modify sln file to build on azure pipeline
* Image classification preview 2.
* PR feedback.
* Add unit-test.
* Add unit-test.
* Add unit-test.
* Add unit-test.
* Use Path.Combine instead of Join.
* fix test dataset path.
* fix test dataset path.
* Improve test.
* Improve test.
* Increase epochs in tests.
* Disable test on Ubuntu.
* Move test to its own project.
* Move test to its own project.
* Move test to its own project.
* Move test to its own file.
* cleanup.
* Disable parallel execution of tensorflow tests.
* PR feedback.
* PR feedback.
* PR feedback.
* PR feedback.
* Prevent TF test to execute in parallel.
* PR feedback.
* Build error.
* clean up.
* Change the model load/save API to always have ITransformer as central object.
* Keep the with loader save order the same as with schema overload, with ITransformer always first.
* Change ModelLoadingTests to use the MLContext of its new base class.
* Internalize some members of SchemaDefinition, and add tests
* Code review comments
* Fix build after rebase
* Fix failing test
* Fix build after rebase
* Internalize Column ctor
* Fix build after rebase
* Step 1: create two multi-class linear models
Step 2: Make SDCA trainers typed
Finish version 0.1
Delete commented lines
* Add some doc strings
More document
* Handle static extensions
* Rename several maximum entropy models and trainers
* Fix EP test
Fix two tests and address a comment
Add missing piece
* Address comments
* Improve option of MCSDCA
* Address comments
* Update code sample
* Refactorize saving family
* Rename a class following binary SDCA trainer
* fixing model parameter discrepencies
* multiclass LR singe that refactoring is happening in a parallel PR
* review comments. Added Multiclass to NaiveBayes
* Drop Classification from trainer names - v1 (more trainers to follow)
* multiclass LR will be handled separately
* Drop Classification from trainer names - v2 (all trainers taken care of)
* fix entrypoint file
* Add save/load APIs for IDataLoader
* Address some code review comments, add a non-generic base class for calibrated predictor
* use the contravariance of ISingleFeaturePredictionTransformer instead of loading PredictionTransformer<object> from file
* Add API for saving/loading input schema
* Fix build after rebase
* Add API to create PredictionEngine with input schema
* Address code review comments
* Unfriend Functional.Tests
* Add CreatePredictionEngine API back to ModelOperationsCatalog
* Address code review comments
* Fix build
* Fix F# tests
* Remove duplicate CreatePredictionEngine API
* Add test for creating an IDataView from a loaded schema
* Fix build error after rebase
* Add unit tests, and address some code review comments
* Fix build after rebase
* Code review comments
* Clean static part for previous projection catalog
Scrubbing normalization again
* Fix a test
* Address #2928
* Address comments
* Rename RandomFourierKernel to ApproximatedKernel everywhere
* Replace Lp with LpNorm.
* Address a comment
* Make sure EP not affected
* Expose normalizer config again
* Expose normalizer's model parameters correctly
* Microsoft.ML.LightGBM -> changes to Microsoft.ML.Trainers.LightGBM
Microsoft.ML.Transforms.FeatureSelection -> moves to Microsoft.ML.Transforms
* ImageAnalytics -> Transforms.Images
and the ImageLoader transform moved to Microsoft.ML.Data together with the TextLoader.
* LightGBM -> LightGbm for namespace, assembly and package.
* renaming several trainers
* updating some of the trainers with acronyms
* updated names based on the latest pattern
* fix name of MetaTrainer
* fix review comments
* connect acronym for SGD
* updates to cookbook markdown file
dotnet-maestro[bot]
Michael Sharp
Jon Wood
frank-dong-ms
Antonio Velázquez
Ye Wang
Yael Dekel
Najeeb Kazmi
Harish Kulkarni
Frank Dong
Maher Jendoubi
Sam Harwell
Zeeshan Siddiqui
Artidoro Pagnoni
Senja Filipi
Jon Wood
Ivan Matantsev
Wei-Sheng Chin
Rogan Carr
Eric Erhardt
Tom Finley
Abhishek Goswami