Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Save the trained xgboost model #2822

Merged
merged 8 commits into from
Aug 14, 2020

Conversation

sneaxiy
Copy link
Collaborator

@sneaxiy sneaxiy commented Aug 13, 2020

Refactorization: save the XGBoost model and metadata after training.

@sneaxiy sneaxiy requested review from Yancey1989, typhoonzero and lhw362950217 and removed request for Yancey1989 and typhoonzero August 14, 2020 07:33
Copy link
Collaborator

@typhoonzero typhoonzero left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM++

@sneaxiy sneaxiy merged commit 927173c into sql-machine-learning:develop Aug 14, 2020
@sneaxiy sneaxiy deleted the save_trained_xgb_model branch August 14, 2020 08:53
sneaxiy added a commit that referenced this pull request Aug 14, 2020
* Update OptFlow FSL generation code when len(variables) == 2 (#2778)

* change optflow api

* polish

* add more ut

* alps submitter and codegen (#2771)

* test alps submitter

* add alps codegen

* update

* Support download model to local in cli (#2779)

* support download model to local in cli

* update

* update

* update

* Add query api to db.py (#2782)

* Add query to db.py

* change delete with truncate

* Set EnableWindowFunc to be true by default for TiDB parser. (#2786)

* Make the attribute check for XGBoost model compatible with reg:linear

* set EnableWindowFunc to true by default for TiDB parser.

* fix Tensorflow -> TensorFlow (#2783)

* Add optimization guide doc (#2785)

* add optimization guide doc

* polish according to comments

* Generate workflow using runtime (#2784)

* WIP generate workflow using runtime

* wip update

* update

* update

* fix hive ci

* DB api base class (#2787)

* Add query to db.py

* change delete with truncate

* DB interface base class

* add to_dict and from_dict method (#2792)

* Fix develop jupyter image build (#2790)

* fix develop jupyter image build

* update

* Generate workflow step code using runtime fea der (#2791)

* WIP generate workflow step code using runtime fea der

* tested local

* update

* update

* update

* fix tests

* Add MySQL db-api implementation (#2793)

* Add query to db.py

* change delete with truncate

* DB interface base class

* Add MySQL db-api implementation

* remove unused import

* fix actions maxcompute test not running (#2795)

* Enable flake8 check on CI (#2788)

* test ci

* test again

* update

* update and fix

* fix travis ci env

* generate python feacol code (#2797)

* Add hive DB-API (#2798)

* Add query to db.py

* change delete with truncate

* DB interface base class

* Add MySQL db-api implementation

* remove unused import

* polish mysql db-api

* Add hive DB-API

* modify doc

* format code

* modify cora dataset to adapt csv format (#2780)

* Add json dump and load support for FeatureColumn (#2794)

* add json dump load support

* update vocabulary type

* update

* update

* update

* Add maxcompute DB-API (#2801)

* Add maxcompute DB-API

* remove unused import

* format code

* Push images on self hosted machine (#2799)

* push images on self hosted machine

* update

* update

* update

* update

* test install.sh

* fix go mirrors

* clean up

* add clean up

* update clean up script

* fix pai xgboost package deps (#2803)

* Simplify TO RUN command - use filename instead of absolute path for the executable or script program (#2804)

* Make the attribute check for XGBoost model compatible with reg:linear

* Derive the absolute path of the runnable program if users just input a file name.

* Use python -m command to invoke the TO RUN statement in default submitter.

* Move getRunnableProgramAbsPath to alisa.go

* Polish DB-API code, export unified connect function from package. (#2808)

* Add maxcompute DB-API

* remove unused import

* format code

* polish db-api

* add solved y to optimize (#2810)

* Generate couler code of workflow steps (#2806)

* wip

* fix yaml generate

* fix tests

* fix package deps

* fix pip package deps

* update

* Refine metadata collect and save/load (#2807)

* move and refine metadata

* fix ci ut

* fix ut

* follow lhw comment

* Adapt paiio with DB-API (#2809)

* Add maxcompute DB-API

* remove unused import

* format code

* polish db-api

* Adapt paiio with DB-API

* Adapt paiio with DB-API

* add try import paiio

* fix typo

* disable actions maxcompute test (#2814)

* make constraint optional (#2812)

* fix typo (#2820)

* Install BARON solver in Docker image (#2811)

* install baron solver in Docker image

* polish

* add pyomo baron into step docker image

* Polish DB-API to support Python2 so can run on PAI (#2815)

* polish db-api to support Python2 so can run on PAI

* enable unittest for hive db-api

* switch to github actions (#2818)

* Add experimental workflow end2end test (#2813)

* add experimental workflow end2end test

* fix workflow ci env

* update test code

* pull latest step before running workflow

* Add  Model.save_to_oss and Model.load_from_oss (#2817)

* add save_to_oss/load_from_oss

* change pickle protocol

* add more explanations on oss_model_dir doc

* fix ut

* Fix relative importing cause error (#2823)

* fix relative importing cause error

* clean up

* Use unified DB-API in codebase (#2821)

* Add maxcompute DB-API

* remove unused import

* format code

* polish db-api

* Adapt paiio with DB-API

* Adapt paiio with DB-API

* add try import paiio

* use db-api in old code

* DB-API support Python2 so can run on PAI

* polish db-api to support Python2 so can run on PAI

* polish db-api to support Python2 so can run on PAI

* polish db-api to support Python2 so can run on PAI

* Use unified DB-API in codebase.

* Use unified DB-API in codebase.

* polish code

* remove debug info

* fix ut

* Generate workflow step for normal statement run (#2824)

* generate workflow step for normal statement run

* clean up

* build step image before run workflow test

* fix is_query

* Fix pai training with optimizer config (#2828)

* fix pai training with optimizer config

* remove template

* Save the trained xgboost model (#2822)

* save trained xgboost model

* fix flake8 check

* fix ut

* fix ut

* fix workflow ut

* fix cwd error

Co-authored-by: Wu Yi <[email protected]>
Co-authored-by: HongwuLin <[email protected]>
Co-authored-by: brightcoder01 <[email protected]>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

2 participants