v23.08.00: migrated base runtime to core (#380)

* migrated base runtime to core

* add comments to init file

* change runtime imports

v23.06.00

Full Changelog: v23.05.00...v23.06.00

v23.05.00

What’s Changed

⚠ Breaking Changes

• update feast op to handle feast 0.31 - latest @jperez999 (#344)
• Condense the WorkflowRunner sub-classes @karlhigley (#319)
• Convert Workflow outputs to Triton-friendly format using TensorTable @bschifferer (#317)
• Fix Tensorflow serving to make Merlin example tests pass again @karlhigley (#298)
• Switch from __lengths to __offsets for ragged list features @karlhigley (#296)

🐜 Bug Fixes

• Fix the code that aligns returned data with the output schema @karlhigley (#335)
• Convert Workflow outputs to Triton-friendly format using TensorTable @bschifferer (#317)
• Make minor adjustments to the way Workflow outputs are formatted @karlhigley (#314)
• Revert "Remove the Triton-specific Workflow execution code (#309)" @karlhigley (#313)
• remove numpy conversion for cpu based data @jperez999 (#311)
• Fix PredictTensorflow args in session-based serving integration test @karlhigley (#310)
• Adjust setup_faiss to use METRIC_INNER_PRODUCT @karlhigley (#308)
• Remove the extra columns validation from WorkflowRunner @karlhigley (#303)
• Fix Tensorflow serving to make Merlin example tests pass again @karlhigley (#298)

🚀 Features

• update feast op to handle feast 0.31 - latest @jperez999 (#344)
• Support Ragged inputs and outputs in TransformWorkflow with Triton @oliverholworthy (#324)
• add AUTO kind for deployment of tritonserver and support no gpu environment @jperez999 (#321)
• Update SoftmaxSampling to support dynamic dtypes @oliverholworthy (#304)
• Add initial support and integration test for Merlin Models transformer models @karlhigley (#299)
• Fix Tensorflow serving to make Merlin example tests pass again @karlhigley (#298)

🔧 Maintenance

• minor syntax fixes in integration tests @nv-alaiacano (#351)
• add test requirements to postmerge gpu @nv-alaiacano (#347)
• Send slack notification on failed GHA jobs @nv-alaiacano (#346)
• Include tracebacks in errors returned from failing requests @karlhigley (#343)
• Add workflows to check base branch and set stable branch @oliverholworthy (#341)
• Add GitHub Workflow to publish conda package @oliverholworthy (#340)
• Update tag pattern in GitHub Workflows @oliverholworthy (#342)
• Update feast test dependency from 0.19.4 to 0.18.1 @oliverholworthy (#339)
• Raise TritonModelException if the Triton model has an error @oliverholworthy (#333)
• Refactor tensor representation adjustments to align with a schema @karlhigley (#332)
• working embedding in systems ensemble @jperez999 (#329)
• working padding op in systems ensemble @jperez999 (#330)
• Support serving in container without tensorflow, torch, or cudf @oliverholworthy (#326)
• use concurrency setting to kill already-running jobs when new commits come in @nv-alaiacano (#327)
• Condense the WorkflowRunner sub-classes @karlhigley (#319)
• add AUTO kind for deployment of tritonserver and support no gpu environment @jperez999 (#321)
• Check offsets instead of is_ragged for ragged representation @oliverholworthy (#318)
• Use ragged names in workflow triton request if column is ragged @oliverholworthy (#315)
• Remove the Triton-specific Workflow execution code @karlhigley (#309)
• Downgrade unused columns error to a warning @karlhigley (#306)
• Remove ensemble specific code @oliverholworthy (#305)
• Minor fixes to post-merge test actions @karlhigley (#302)
• Set up pre- and post-merge CI test runs @karlhigley (#301)
• add detection swap for cpu based to numpy @jperez999 (#300)
• Switch from __lengths to __offsets for ragged list features @karlhigley (#296)
• Update default value of cpu to None in dataset fixture @oliverholworthy (#294)
• update conftest for backwards compat and new to parquet api @jperez999 (#297)
• Adjust Systems to use TensorTable and schema shapes @karlhigley (#293)

v23.02.00

What’s Changed

⚠ Breaking Changes

• Remove unused legacy serving code @karlhigley (#265)

🐜 Bug Fixes

• Error propagation from tritonserver for nvtabular workflow @jperez999 (#277)
• Remove tf model from ensemble export @jperez999 (#289)
• Use row lengths instead of offsets in Triton input helper function @oliverholworthy (#285)
• increase triton fail time to 30 seconds each cycle @jperez999 (#284)

🚀 Features

• Return the original predicted scores from SoftmaxSampling @karlhigley (#290)
• Add error messages for NotImplementedError Exceptions @jperez999 (#280)

🔧 Maintenance

• Error propagation from tritonserver for nvtabular workflow @jperez999 (#277)
• Update compute_dims function to use new shape property @oliverholworthy (#288)
• Replace distutils.spawn (deprecated) with shutil.which @karlhigley (#261)
• increase triton fail time to 30 seconds each cycle @jperez999 (#284)
• refactor out pipelineable inference operator @jperez999 (#279)
• Add error messages for NotImplementedError Exceptions @jperez999 (#280)
• throttling back systems triton call check. @jperez999 (#276)
• Run lint action with Python 3.8 @karlhigley (#283)
• Add Prettier YAML/Markdown formatter to the linters @karlhigley (#273)
• Minor fixes to the Github actions @karlhigley (#272)
• Apply Merlin dtypes to Systems code base @karlhigley (#239)
• add gcp label to Jenkinsfile @AyodeAwe (#267)
• Remove unused legacy serving code @karlhigley (#265)

v0.9.0

What’s Changed

🚀 Features

• working local runtime fil ops @jperez999 (#245)
• detect when column is list of one argument per record @jperez999 (#251)
• Dictarray to and from df @jperez999 (#247)
• working local pytorch transform @jperez999 (#246)
• Nvt op runtime triton @jperez999 (#238)
• refactor for torch operator to triton specific @jperez999 (#241)
• update triton to allocate more cuda memory to pool size @jperez999 (#237)
• fil triton runtime operator created and loaded @jperez999 (#226)

📄 Documentation

• Address virtual dev review feedback @mikemckiernan (#250)
• Update API and add notebook @mikemckiernan (#242)

🔧 Maintenance

• Remove use of hf_format and training params for PyTorch inference test @oliverholworthy (#236)
• add import or skip for torch tests @jperez999 (#244)
• update drafter to work on tags & update cpu ci to target branches @jperez999 (#235)
• fix container name and the link in the examples README @rnyak (#232)
• Add Jenkinsfile @AyodeAwe (#230)

v0.8.0

What’s Changed

⚠ Breaking Changes

• Move Triton ensemble and executor export to runtime classes @oliverholworthy (#217)
• Rework Systems operators to use DictArray and LocalExecutor from Merlin Core @karlhigley (#204)

🐜 Bug Fixes

• add env variable to handle gpu memory allocation @jperez999 (#221)
• fix for faiss artifact path from config @jperez999 (#216)

🚀 Features

• Update Pytorch integration test based on schema changes in T4R @oliverholworthy (#220)

🔧 Maintenance

• Refactor TensorFlow operator into base and Triton runtime operator @oliverholworthy (#219)
• add env variable to handle gpu memory allocation @jperez999 (#221)
• Add transform method to Ensemble @oliverholworthy (#218)
• Move Triton ensemble and executor export to runtime classes @oliverholworthy (#217)
• tensorflow import or skip required for tf_utils @jperez999 (#215)
• Rework Systems operators to use DictArray and LocalExecutor from Merlin Core @karlhigley (#204)
• Update example notebook tests to write to temp directory. @oliverholworthy (#214)
• Add requests dependency to base requirements @oliverholworthy (#213)
• Create a separate directory parallel to unit for version tests @karlhigley (#212)

v0.7.0: Start using the `validate_schemas` hook from `core` (#95)

What’s Changed

🐜 Bug Fixes

• fix for faiss artifact path from config @jperez999 (#216)

🔧 Maintenance

• Add transform method to Ensemble @oliverholworthy (#218)
• Move Triton ensemble and executor export to runtime classes @oliverholworthy (#217)
• tensorflow import or skip required for tf_utils @jperez999 (#215)
• Rework Systems operators to use DictArray and LocalExecutor from Merlin Core @karlhigley (#204)
• Update example notebook tests to write to temp directory. @oliverholworthy (#214)
• Add requests dependency to base requirements @oliverholworthy (#213)
• Create a separate directory parallel to unit for version tests @karlhigley (#212)

v0.6.0

What’s Changed

🐜 Bug Fixes

• fix pinned feast versions @nv-alaiacano (#206)

🚀 Features

• Write an integration test for T4R Torchscript serving @karlhigley (#176)

📄 Documentation

• add implicit example @radekosmulski (#194)

🔧 Maintenance

• Mark the version test so it can be skipped in other repos @karlhigley (#201)
• run example tests in pr @nv-alaiacano (#199)

v0.5.0

What’s Changed

⚠ Breaking Changes

• Required changes to get Rank example test passing @jperez999 (#182)

🐜 Bug Fixes

• Specify is_ragged=False explicitly in the Feast tests @karlhigley (#184)
• feast should return np.float32 not np.float by default @nv-alaiacano (#178)
• Fix the importorskip for Torch tests @karlhigley (#175)
• revert change that renamed output schema @nv-alaiacano (#172)
• Revert "Allow selection of multiple outputs with SampleSoftmax" @nv-alaiacano (#171)
• Allow selection of multiple outputs with SampleSoftmax @nv-alaiacano (#155)
• Clean up model input signature handling for tuples @karlhigley (#152)
• SoftmaxSampling should produce the same output dtype as the input @nv-alaiacano (#154)

🚀 Features

• Add properties for input and output schemas to Ensemble @karlhigley (#197)
• Add preliminary support for serving PyTorch models with TorchScript @karlhigley (#167)
• Add Operator for Implicit Models @oliverholworthy (#134)
• Allow selection of multiple outputs with SampleSoftmax @nv-alaiacano (#155)
• Pass model kwargs through to operator from_config methods @oliverholworthy (#158)
• Allow merlin.models.xgb.XGBoost in PredictForest @benfred (#146)
• Load models from a model registry rather than path @nv-alaiacano (#149)

📄 Documentation

• Update the logos in the examples @bschifferer (#188)
• add xgboost serving example @radekosmulski (#157)

🔧 Maintenance

• Fix versioneer to get accurate version numbers @benfred (#198)
• Move requirements files to a requirements sub-directory @karlhigley (#193)
• introduce cascading timeout for tritonserver launch @jperez999 (#192)
• Mark example notebook tests to make them skippable in other repos @karlhigley (#185)
• Skip "serving ranking models" example test on CPU @karlhigley (#191)
• Add extra requirements files to setup.py as optional installs @karlhigley (#189)
• Remove importorskip("tritonclient") from the tests @karlhigley (#190)
• Add tritonclient as a test dependency @karlhigley (#187)
• add testenv to run merlin tests @nv-alaiacano (#183)
• Required changes to get Rank example test passing @jperez999 (#182)
• use container sitepackages for jenkins tests @nv-alaiacano (#180)
• add test for QueryFeast.from_feature_view @nv-alaiacano (#179)
• Fix the importorskip for Torch tests @karlhigley (#175)
• Organize the tests directory more clearly @karlhigley (#170)
• Enable gRPC port to be specified in run_triton_server and check if already running @oliverholworthy (#141)
• Build the documents with tox @mikemckiernan (#168)
• install more libraries in test environment @nv-alaiacano (#166)
• Update versioneer from 0.21 to 0.23 @oliverholworthy (#165)
• Fix importorskip in Feast tests @karlhigley (#163)
• Mark Feast tests with pytest.importorskip @karlhigley (#162)
• use tox for github CI steps @nv-alaiacano (#161)
• Add unit tests for QueryFeast op @nv-alaiacano (#160)
• Update precommit hook package versions @karlhigley (#150)

v0.4.0

What’s Changed

🐜 Bug Fixes

• fix notebook and add unittest @bschifferer (#140)
• Systems pb update @jperez999 (#130)
• Correct FIL export format for sklearn/cuml to treelite checkpoint @oliverholworthy (#124)
• Fix dtype of multi-hot features' nnzs field @karlhigley (#123)

📄 Documentation

• fix link @radekosmulski (#127)

🔧 Maintenance

• Remove unnecessary docs dependencies @benfred (#147)
• Add a more helpful error message to Feast operator @karlhigley (#142)
• Systems pb update @jperez999 (#130)
• Add test for forest inference running on Triton + handle Tensor in Op Runner @oliverholworthy (#122)