doc: update CHANGELOG to reflect v2.0.0 changes by laurenyu · Pull Request #1808 · aws/sagemaker-python-sdk
* Trinity (#1765) * Test commit of new ModelBulider with V2 dependencies eliminated * ModelBuilder.build() working for tgi server * Support for torchserve, tgi, djl, triton, tei, mms, and smd * Build fixed for jumpstart, auto-container detection logic fixed * Model Builder cleaned and added more unit tests * Deploy happy-case for Triton and JumpStart (TGI) * Better auto-container detection and user experience * Remove debug print statements * Removing bad unit tests * Style and cleanup changes * Removing unrelated unit tests for now (will migrate in the future to V3) * Fix installation and Sentinel issue * Attempt to fix installation issues #2 * Sentinel bug fix third attempt, support for tgi server * PR Check installation issue fix attempt number four * Resolving missing dependency issues * Update dependencies for sagemaker_train * serve folder flattening, support for local mode * Fully supporting local mode and in process mode with integ tests * ModelBuilder and Pipeline merged * Better deploy logic and fixed bad imports * Trinity updated (#1769) * Test commit of new ModelBulider with V2 dependencies eliminated * ModelBuilder.build() working for tgi server * Support for torchserve, tgi, djl, triton, tei, mms, and smd * Build fixed for jumpstart, auto-container detection logic fixed * Model Builder cleaned and added more unit tests * Deploy happy-case for Triton and JumpStart (TGI) * Better auto-container detection and user experience * Remove debug print statements * Removing bad unit tests * Style and cleanup changes * Removing unrelated unit tests for now (will migrate in the future to V3) * Fix installation and Sentinel issue * Attempt to fix installation issues #2 * Sentinel bug fix third attempt, support for tgi server * PR Check installation issue fix attempt number four * Resolving missing dependency issues * Update dependencies for sagemaker_train * serve folder flattening, support for local mode * Fully supporting local mode and in process mode with integ tests * ModelBuilder and Pipeline merged * Better deploy logic and fixed bad imports * Fixes for build() and deploy(), ModelTrainer support --------- Co-authored-by: Zhaoqi <52220743+zhaoqizqwang@users.noreply.github.com> * Update processor and model trainer (#1772) * Test pipeline v3 (#1773) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Resolve conflict (#1774) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix ProcessingJob pipeline integration (#1778) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables (#1779) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables * Test pipeline v3 (#1781) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables * Fix PipelineVariables used in ModelBuilder * Fix ModelTrainer.train() request (#1782) * Fix ModelTrainer request * Clean up code * Cleaned up files and removed unneeded files (#1780) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Fix pipeline example notebooks * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables * Fix PipelineVariables used in ModelBuilder * Fix ModelTrainer request * Clean up code * Fix v3 pipeline examples * Trinity updated (#1787) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * remote function, model card (#1786) * Add model card to v3 Add unit tests passed. Compared the new classes with V2 classes * fixed circular dep issue for modelcard testing * remote function working preoperly now --------- Co-authored-by: jzhaoqwa <52220743+zhaoqizqwang@users.noreply.github.com> Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * Fixed optimize import error (#1791) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * Optimize import fix * Removing llm_utils dependency * Bugbash fixes, including region support, role_arn optional, and progress monitoring (#1792) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * Optimize import fix * Removing llm_utils dependency * Bugbash fixes, including region support, role_arn optional, and progress monitoring * Keep CUDA warning and don't hardcode fallback region * Add model registry to v3 (#1794) * Update model registry and examples * tuner fix (#1796) Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * Add transformation job and example (#1797) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables * Fix PipelineVariables used in ModelBuilder * Fix ModelTrainer request * Clean up code * Fix v3 pipeline examples * Update model registry and examples * Add transformation job example * Trinity update for model monitor (#1789) * model monitoring test and fix * model monitor changes and notebook (needs final run through) * remove unused notebook * add actual notebook * file dir change * clarify added and model monitor fixed * Trinity hyperparametertuner (#1798) * tuner fix * example for hyperparameter tuning * rmvd unneeded files * added back --------- Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * Add pipeline_models key to ModelBuilder to support building model with multiple containers (#1799) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables * Fix PipelineVariables used in ModelBuilder * Fix ModelTrainer request * Clean up code * Fix v3 pipeline examples * Update model registry and examples * Add transformation job example * Add feature to ModelBuild for building model with multiple containers (pipeline_models) * fix code * Continue fix pipeline model * Model monitor in pipeline example updated, experiment config example added (#1801) * model monitoring test and fix * model monitor changes and notebook (needs final run through) * remove unused notebook * add actual notebook * file dir change * clarify added and model monitor fixed * model monitor in pipeline example updated, experiment config example added * Support TuningStep in Pipeline (#1803) * Support TuningStep in pipeline * enables repack_model with js compatability (#1806) Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * Fix model monitor example notebook, add experiment module example (#1809) * model monitoring test and fix * model monitor changes and notebook (needs final run through) * remove unused notebook * add actual notebook * file dir change * clarify added and model monitor fixed * model monitor in pipeline example updated, experiment config example added * fix model monitor example notebook, add experiment module * Add FrameworkProcessor to support Pytorch preprocessing (#1810) * Add Framework Processor to support Pytorch models * step decorator fix (#1812) Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * Fix import error and eemove unused examples (#1814) * Fix import error and eemove unused examples * Support for FrameWork Models (#1808) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * Optimize import fix * Removing llm_utils dependency * Bugbash fixes, including region support, role_arn optional, and progress monitoring * Keep CUDA warning and don't hardcode fallback region * Support for XGBoost, SKLearn, and other framework models * Update ML Ops examples (#1815) * Update ML Ops examples * Update examples * update transform job pipeline notebook (#1817) * extra model card changes * local container fix (#1820) Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * SA Bugbash fixes (#1822) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * Optimize import fix * Removing llm_utils dependency * Bugbash fixes, including region support, role_arn optional, and progress monitoring * Keep CUDA warning and don't hardcode fallback region * Support for XGBoost, SKLearn, and other framework models * WIP: Updates to model builder and serve components * WIP: Latest V3 ModelBuilder updates before upstream merge * Test pipeline v3 (#1824) * Update ML Ops examples * Update examples * Remove unused files * clean up files * Move over jumpstart utils * feat: add jumpstart model_builder telemetry (#1784) * add jumpstart model_builder telemetry * chore: use is_jumpstart_model_id to detect JS use case * feat: add tagging for JS model builder created from trainer --------- Co-authored-by: aviruthen <91846056+aviruthen@users.noreply.github.com> * Bug fixes for local mode and tensorflow-serving (#1829) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * Optimize import fix * Removing llm_utils dependency * Bugbash fixes, including region support, role_arn optional, and progress monitoring * Keep CUDA warning and don't hardcode fallback region * Support for XGBoost, SKLearn, and other framework models * WIP: Updates to model builder and serve components * WIP: Latest V3 ModelBuilder updates before upstream merge * Support for local mode on all model servers, support for tensorflow-serving * Updated Autogen SageMaker-Core * Bug Fixes for JumpStart, ModelTrainer ModelBuilder flow * Bug fixes, code cleanup, and adding integration tests * Migrating utils (#1832) * fixed integ test imports * fixed pyproj * toml file fix * fixed import paths and sm-train * import update in example notebook * modelbuilder * modelbuilder utils reference deleted * Import bug fixes for jumpstart * fixes * pyproj fix * resource requirement fix * Bug fixes for jumpstart and training * model builder fix * model builder merge * notebooks fix and modeltrainer * notebook role fix --------- Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> Co-authored-by: aviruthen <91846056+aviruthen@users.noreply.github.com> * Update MLOps integ tests and ModeBuilder model key(#1839) * Add v3 inference example notebooks (#1840) * Cleaned up example notebooks (#1841) * Removed import warnings and organized example notebooks * Updated folder name * Removed legacy folder (#1843) * Removed import warnings and organized example notebooks * Updated folder name * Removed legacy folder * Remove unused files * Remove files that depend on predictor classes * Update Training tox.ini for V3 release * Updating tox.ini and codebuild yaml for PR checks (#1850) * Trinity (#1765) * Test commit of new ModelBulider with V2 dependencies eliminated * ModelBuilder.build() working for tgi server * Support for torchserve, tgi, djl, triton, tei, mms, and smd * Build fixed for jumpstart, auto-container detection logic fixed * Model Builder cleaned and added more unit tests * Deploy happy-case for Triton and JumpStart (TGI) * Better auto-container detection and user experience * Remove debug print statements * Removing bad unit tests * Style and cleanup changes * Removing unrelated unit tests for now (will migrate in the future to V3) * Fix installation and Sentinel issue * Attempt to fix installation issues #2 * Sentinel bug fix third attempt, support for tgi server * PR Check installation issue fix attempt number four * Resolving missing dependency issues * Update dependencies for sagemaker_train * serve folder flattening, support for local mode * Fully supporting local mode and in process mode with integ tests * ModelBuilder and Pipeline merged * Better deploy logic and fixed bad imports * Trinity updated (#1769) * Test commit of new ModelBulider with V2 dependencies eliminated * ModelBuilder.build() working for tgi server * Support for torchserve, tgi, djl, triton, tei, mms, and smd * Build fixed for jumpstart, auto-container detection logic fixed * Model Builder cleaned and added more unit tests * Deploy happy-case for Triton and JumpStart (TGI) * Better auto-container detection and user experience * Remove debug print statements * Removing bad unit tests * Style and cleanup changes * Removing unrelated unit tests for now (will migrate in the future to V3) * Fix installation and Sentinel issue * Attempt to fix installation issues #2 * Sentinel bug fix third attempt, support for tgi server * PR Check installation issue fix attempt number four * Resolving missing dependency issues * Update dependencies for sagemaker_train * serve folder flattening, support for local mode * Fully supporting local mode and in process mode with integ tests * ModelBuilder and Pipeline merged * Better deploy logic and fixed bad imports * Fixes for build() and deploy(), ModelTrainer support --------- Co-authored-by: Zhaoqi <52220743+zhaoqizqwang@users.noreply.github.com> * Update processor and model trainer (#1772) * Test pipeline v3 (#1773) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Resolve conflict (#1774) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix ProcessingJob pipeline integration (#1778) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables (#1779) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables * Test pipeline v3 (#1781) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables * Fix PipelineVariables used in ModelBuilder * Fix ModelTrainer.train() request (#1782) * Fix ModelTrainer request * Clean up code * Cleaned up files and removed unneeded files (#1780) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Fix pipeline example notebooks * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables * Fix PipelineVariables used in ModelBuilder * Fix ModelTrainer request * Clean up code * Fix v3 pipeline examples * Trinity updated (#1787) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * remote function, model card (#1786) * Add model card to v3 Add unit tests passed. Compared the new classes with V2 classes * fixed circular dep issue for modelcard testing * remote function working preoperly now --------- Co-authored-by: jzhaoqwa <52220743+zhaoqizqwang@users.noreply.github.com> Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * Fixed optimize import error (#1791) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * Optimize import fix * Removing llm_utils dependency * Bugbash fixes, including region support, role_arn optional, and progress monitoring (#1792) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * Optimize import fix * Removing llm_utils dependency * Bugbash fixes, including region support, role_arn optional, and progress monitoring * Keep CUDA warning and don't hardcode fallback region * Add model registry to v3 (#1794) * Update model registry and examples * tuner fix (#1796) Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * Add transformation job and example (#1797) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables * Fix PipelineVariables used in ModelBuilder * Fix ModelTrainer request * Clean up code * Fix v3 pipeline examples * Update model registry and examples * Add transformation job example * Trinity update for model monitor (#1789) * model monitoring test and fix * model monitor changes and notebook (needs final run through) * remove unused notebook * add actual notebook * file dir change * clarify added and model monitor fixed * Trinity hyperparametertuner (#1798) * tuner fix * example for hyperparameter tuning * rmvd unneeded files * added back --------- Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * Add pipeline_models key to ModelBuilder to support building model with multiple containers (#1799) * Update processor and model trainer * Add Pipeline Variable class to sagemaker core * Update PipelineVariable usage and imports * Fix processing.py for pipeline integration * Remove logs * More fixes on ModelTrainer's compatibility with PipelineVariables * Fix PipelineVariables used in ModelBuilder * Fix ModelTrainer request * Clean up code * Fix v3 pipeline examples * Update model registry and examples * Add transformation job example * Add feature to ModelBuild for building model with multiple containers (pipeline_models) * fix code * Continue fix pipeline model * Model monitor in pipeline example updated, experiment config example added (#1801) * model monitoring test and fix * model monitor changes and notebook (needs final run through) * remove unused notebook * add actual notebook * file dir change * clarify added and model monitor fixed * model monitor in pipeline example updated, experiment config example added * Support TuningStep in Pipeline (#1803) * Support TuningStep in pipeline * enables repack_model with js compatability (#1806) Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * Fix model monitor example notebook, add experiment module example (#1809) * model monitoring test and fix * model monitor changes and notebook (needs final run through) * remove unused notebook * add actual notebook * file dir change * clarify added and model monitor fixed * model monitor in pipeline example updated, experiment config example added * fix model monitor example notebook, add experiment module * Add FrameworkProcessor to support Pytorch preprocessing (#1810) * Add Framework Processor to support Pytorch models * step decorator fix (#1812) Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * Fix import error and eemove unused examples (#1814) * Fix import error and eemove unused examples * Support for FrameWork Models (#1808) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * Optimize import fix * Removing llm_utils dependency * Bugbash fixes, including region support, role_arn optional, and progress monitoring * Keep CUDA warning and don't hardcode fallback region * Support for XGBoost, SKLearn, and other framework models * Update ML Ops examples (#1815) * Update ML Ops examples * Update examples * update transform job pipeline notebook (#1817) * extra model card changes * local container fix (#1820) Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> * SA Bugbash fixes (#1822) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * Optimize import fix * Removing llm_utils dependency * Bugbash fixes, including region support, role_arn optional, and progress monitoring * Keep CUDA warning and don't hardcode fallback region * Support for XGBoost, SKLearn, and other framework models * WIP: Updates to model builder and serve components * WIP: Latest V3 ModelBuilder updates before upstream merge * Test pipeline v3 (#1824) * Update ML Ops examples * Update examples * Remove unused files * clean up files * Move over jumpstart utils * feat: add jumpstart model_builder telemetry (#1784) * add jumpstart model_builder telemetry * chore: use is_jumpstart_model_id to detect JS use case * feat: add tagging for JS model builder created from trainer --------- Co-authored-by: aviruthen <91846056+aviruthen@users.noreply.github.com> * Bug fixes for local mode and tensorflow-serving (#1829) * Minor bug fixes for ModelBuilder experience * Cleaned up files and removed unneeded files * Small fix to pass kwargs through _deploy() * Bug fixes for xgboost, sklearn, tgi, and tei * Optimize import fix * Removing llm_utils dependency * Bugbash fixes, including region support, role_arn optional, and progress monitoring * Keep CUDA warning and don't hardcode fallback region * Support for XGBoost, SKLearn, and other framework models * WIP: Updates to model builder and serve components * WIP: Latest V3 ModelBuilder updates before upstream merge * Support for local mode on all model servers, support for tensorflow-serving * Updated Autogen SageMaker-Core * Bug Fixes for JumpStart, ModelTrainer ModelBuilder flow * Bug fixes, code cleanup, and adding integration tests * Migrating utils (#1832) * fixed integ test imports * fixed pyproj * toml file fix * fixed import paths and sm-train * import update in example notebook * modelbuilder * modelbuilder utils reference deleted * Import bug fixes for jumpstart * fixes * pyproj fix * resource requirement fix * Bug fixes for jumpstart and training * model builder fix * model builder merge * notebooks fix and modeltrainer * notebook role fix --------- Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> Co-authored-by: aviruthen <91846056+aviruthen@users.noreply.github.com> * Update MLOps integ tests and ModeBuilder model key(#1839) * Add v3 inference example notebooks (#1840) * Cleaned up example notebooks (#1841) * Removed import warnings and organized example notebooks * Updated folder name * Removed legacy folder (#1843) * Removed import warnings and organized example notebooks * Updated folder name * Removed legacy folder * Remove unused files * Remove files that depend on predictor classes * test commit to trigger pr checks * Fixed tox.ini files and codebuild workflow yaml for PR checks --------- Co-authored-by: Zhaoqi <52220743+zhaoqizqwang@users.noreply.github.com> Co-authored-by: Mohamed Zeidan <81834882+mohamedzeidan2021@users.noreply.github.com> Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> Co-authored-by: Molly He <mollyhe@amazon.com> Co-authored-by: Rohan Narayan <narrohag@amazon.com> * Removing estimator import (#1852) * Adding back MMS and TEI folders (accidentally deleted) (#1853) * Removing estimator import * Adding back MMS and TEI folders * Recreate PR review for exposing parameters and functions for ModelBuilder (#1854) * Updated user-exposed functions and parameters * Updating deploy() header * Simplifying build() and optimize() * Update integration tests * Final update to model_builder.py * Add unit tests to MLOps, Fix sagemaker-train tests, remove unused imp… (#1856) * Add unit tests to MLOps, Fix sagemaker-train tests, remove unused imports * Add unit tests to MLOps, Fix sagemaker-train tests * Add sagemaker-serve unit tests (#1858) * Bug fixes for integ tests (#1859) * Add core unit tests (#1860) Fix sagemaker core unit tests * Add serve unit tests (#1861) * Add sagemaker-serve unit tests * Fix sagemaker core unit tests * More fixes for the unit tests * ModelBuilder follow up fixes * ModelBuilder example fixes * Update integ tests (#1864) * Fix optimize integ test (#1866) * Update readme and changelog * Update readme, changelogs * Add comprehensive unit tests for sagemaker-core (#1870) * Test for Model BUilder servers (#1871) * Add sagemaker-core unit test (#1872) * Add sagemaker-train and sagemaker-mlops unit tests (#1873) * Add sagemaker-train and sagemaker-mlops unit tests * Add sagemaker-serve unit tests * Fix: Include region_config.json in package data (#1875) * Add MLOps integ tests (#1877) * Add additional sagemaker-serve unit tests (#1878) * removed one integ tests (#1879) --------- Co-authored-by: Zhaoqi <52220743+zhaoqizqwang@users.noreply.github.com> Co-authored-by: Mohamed Zeidan <81834882+mohamedzeidan2021@users.noreply.github.com> Co-authored-by: Mohamed Zeidan <zeidmo@amazon.com> Co-authored-by: Molly He <mollyhe@amazon.com> Co-authored-by: Rohan Narayan <narrohag@amazon.com> Co-authored-by: Gokul Anantha Narayanan <166456257+nargokul@users.noreply.github.com>