Skip to content

Latest commit

 

History

History
1543 lines (1404 loc) · 183 KB

CHANGELOG.md

File metadata and controls

1543 lines (1404 loc) · 183 KB
Raw

Change Log

0.1.27 (2019-08-22)

Full Changelog

Merged pull requests:

  • update namespaced-install.yaml #1926 (IronPan)
  • Fix lint related issue #1922 (numerology)
  • Cleanup pipeline-lite deployment #1921 (IronPan)
  • Allow visualization kernel timeout to be specifiable via environment variables #1920 (ajchili)
  • Release 151c5349f13bea9d626c988563c04c0a86210c21 #1916 (hongye-sun)
  • cleanup test directory #1914 (IronPan)
  • SDK - Airflow - Fixed bug in AirFlow op creation #1911 (Ark-kun)
  • Add cloud sql and gcs connection for pipeline-lite deployment #1910 (IronPan)
  • Enable error propagation from nbconvert to frontend #1909 (ajchili)
  • remove tfx notebook sample in favor of the TFX OSS sample #1908 (gaoning777)
  • [front-end-server] Allow viewer:tensorboard podTemplateSpec to be customizable #1906 (eterna2)
  • Change the type of resource reference payload column #1905 (IronPan)
  • apiserver: Remove TFX output artifact recording to metadatastore #1904 (dushyanthsc)
  • remove kubeflow training to favor tfx components #1902 (gaoning777)
  • Add TFDV, TFMA, and Table visualization support for Python based visualizations #1898 (ajchili)
  • Add run with json data as input within fixed-data.ts for UI testing and development #1895 (ajchili)
  • Use single part as default #1893 (IronPan)
  • fix unit tests and address some comments #1892 (gaoning777)
  • Replace codemirror editor react component with react-ace editor component #1890 (ajchili)
  • Simplified the build_docker_image function #1887 (Ark-kun)
  • IBM Watson samples: from six.moves import xrange #1877 (cclauss)
  • Undefined name 'e' in openvino #1876 (cclauss)
  • SDK - Lightweight - Fixed custom types in multi-output case #1875 (Ark-kun)
  • SDK - Components - Fixed ModelBase comparison bug #1874 (Ark-kun)
  • Update changelog for 0.1.26 #1872 (neuromage)
  • Remove copying of tfx data for cloudbuild release steps. #1871 (neuromage)
  • Update manifests to point to 0.26 release. #1870 (neuromage)
  • add compile step in the samples to generate zip files #1866 (gaoning777)
  • Update Python SDK versions for release. #1860 (neuromage)
  • test/project-cleanup: Support to cleanup gke-clusters in test project #1857 (dushyanthsc)
  • Created extensible code editor based on react-ace #1855 (ajchili)
  • Add visualization-server service to lightweight deployment #1844 (ajchili)
  • SDK - Tests - Improved the "ContainerOp.set_retry" test #1843 (Ark-kun)
  • pipeline-lite: Introduce metadata component to pipeline-lite #1840 (dushyanthsc)
  • [Bug Fix] Delete ResourceOp should not have output parameters #1822 (eterna2)
  • Use KFP lite deployment for presubmit tests #1808 (Bobgy)
  • SDK - Fixed string comparisons #1756 (Ark-kun)
  • 'core' folder included to parameters related On-Premise cluster #1751 (olegchorny)
  • Refactor presubmit-tests-with-pipeline-deployment.sh to run in other projects #1732 (Bobgy)
  • Lint Python code for undefined names #1721 (cclauss)
  • Let backend apiserver mysql dbname configurable #1714 (jwwandy)
  • gcp cred bug fix for multiple credentials in single pipeline #1384 (aakashbajaj)
  • Collecting coverage when running python tests #898 (Ark-kun)

0.1.26 (2019-08-16)

Full Changelog

Merged pull requests:

0.1.25 (2019-07-26)

Full Changelog

Merged pull requests:

  • Increase version in namespaced-install.yaml too #1684 (jingzhang36)
  • Add visualization API service to frontend #1675 (ajchili)
  • Increase KFP version in (1) sdk/python/setup.py (2) component_sdk/python/setup.py (3) manifests/base/kustomization.yaml #1674 (jingzhang36)
  • Release fe639f41661d8e17fcda64ff8242127620b80ba0 #1672 (jingzhang36)
  • Adding kfp.cil to kfp setup.py. #1666 (hongye-sun)
  • Fix exit handler sample #1665 (hongye-sun)
  • Add visualization swagger files to frontend #1663 (ajchili)
  • move gcshelper out of component_builder #1658 (gaoning777)
  • Remove redundant import. #1656 (numerology)
  • Fix broken license link #1655 (hongye-sun)
  • update kaniko executor version to speed up image build #1652 (gaoning777)
  • add init container for container op #1650 (IronPan)
  • update python test to use python3 #1649 (IronPan)
  • Add visualization server and unit tests for visualization server #1647 (ajchili)
  • Sets min widths for buttons to avoid text wrapping #1637 (rileyjbauer)
  • Changed arguments parameter type and built visualization pb and swagger files #1636 (ajchili)
  • Add recurring run column to run lists #1635 (rileyjbauer)
  • update persistent agence to only store the argo spec #1634 (IronPan)
  • Add OWNERS file in aws components and examples #1633 (Jeffwan)
  • Samples - Cleaned up unnecessary usage of PipelineParam #1631 (Ark-kun)
  • Samples - Removed the immediate_value sample #1630 (Ark-kun)
  • Add SageMaker HPO component and sample usage in a pipeline #1628 (carolynwang)
  • propagate database password when init db connection #1627 (IronPan)
  • SDK - Lightweight - Added support for "None" default values #1626 (Ark-kun)
  • Removed duplicate CircularProgress within UI #1625 (ajchili)
  • viewer controller is now namespaced so no need for cluster role #1623 (jingzhang36)
  • switch third party images to GCR #1622 (IronPan)
  • update changelog #1618 (gaoning777)
  • Update samples to use the latest versions of the python SDK #1607 (gaoning777)
  • Trims whitespace from pipeline params on creation of new run #1605 (rileyjbauer)
  • SDK/DSL: Fix bug when specifying custom resource for VolumeOp #1595 (elikatsis)
  • [Feature] Set ttlSecondsAfterFinished in argo workflow with PipelineConf #1594 (eterna2)
  • Delete go CLI #1592 (IronPan)
  • Added image captioning sample to samples/notebooks #1591 (zanedurante)
  • tox.ini: Add flake8 tests to find Python syntax errors and undefined names #1577 (cclauss)
  • Add gcp-connector label and remove pod name #1542 (hongye-sun)
  • Remove legacy hardcoded version of TFX #1502 (IreneGi)
  • add support for flexible config (via env var) for the pipline service and UI, fix broken links (pointed to API vs UI service) #1293 (yaronha)

0.1.24 (2019-07-12)

Full Changelog

Merged pull requests:

0.1.23 (2019-06-26)

Full Changelog

Merged pull requests:

0.1.22 (2019-06-21)

Full Changelog

Merged pull requests:

0.1.21 (2019-05-29)

Full Changelog

Merged pull requests:

0.1.20 (2019-05-14)

Full Changelog

Closed issues:

  • Cannot create job for experiment via Pipelines Go CLI #1321
  • Support a container flow inside one pod #1313
  • toleration support for ContainerOp #1265
  • Can only create recurring run from within experiment page #1217

Merged pull requests:

0.1.19 (2019-05-03)

Full Changelog

Closed issues:

  • Unfinished run duration shown as negative time #1236
  • Release the SDK in Pypi and automate the release process #1233
  • The post-submit test always fails #1228
  • docker magic command should print kaniko pod logs #924
  • Kubeflow Pipeline conditionals should be able to handle multiple output parameters #679

Merged pull requests:

0.1.18 (2019-04-26)

Full Changelog

Closed issues:

  • sample "DSL Static Type Checking" link to types needs to be updated #1220
  • Update the SDK version automatically during the release #1190
  • Bug: Unable to delete recurring run configs in UI #1181
  • Add integration test for filtering #1155
  • Pipeline submission fails when ContainerOp has numbers as command line arguments #1121
  • Add SDK support for sidecars #949
  • Missing PullPolicy for ContainerOp #832
  • Extend the DSL with support for Persistent Volumes and Snapshots #801
  • Use GCP credentials for existing samples #418
  • Does pipelines only support on GKE? And the workflow run with some problem. #365
  • Remove experiment column from runs list inside experiment #222
  • Support all container APIs through container op #204

Merged pull requests:

0.1.17 (2019-04-17)

Full Changelog

Closed issues:

  • Runs leaked into archived page #1150
  • Pagination broken in UI #1149
  • The number of members in the policy (1,503) is larger than the maximum allowed size 1,500 #1146
  • Backend Docker build fails with python error in resnet-train-pipeline.py #1142
  • Use range instead of pin for python dependencies for kfp #1134
  • Deploy a TFX pipeline from the command line #1127
  • CloudBuild is failing since the CMLE deployer component was deleted #1123
  • Failed to load the trained model using kubeflow deployer component #1102
  • Pipeline parameters should be accessible to pipeline logic #1099
  • Retrieve the experiment during list run #1084
  • Connect local notebook to remote Pipeline cluster #1079
  • How to support continuous online learning? #1053
  • When notebook submits a run and clicks the link they get "Error: failed to retrieve run:" #1017
  • Add static type checking sample test #974
  • Test coverage for runs without experiments #685

Merged pull requests:

0.1.16 (2019-04-06)

Full Changelog

Closed issues:

  • usage reporting link broken #1073
  • Exit handler task does not have proper name in the UX #1051
  • Remove link to JupyterHub from pipelines UI #1029
  • Invalid memory address or nil pointer dereference in metadata_store.go #1024
  • KFP SDK: suggest retry if pipeline launch returns 500 error ('warmup' apparently required) #1007
  • Refactor the test infra code #875
  • Ability to stop runs #413
  • Surface the sample test results to github #361
  • feature ask: support for loops #303

Merged pull requests:

0.1.14 (2019-03-28)

Full Changelog

Closed issues:

  • Users should be able to add metadata to component.yaml #1013
  • Allow Python functions defined in a pipeline to be run as entrypoint for operations. #979
  • Unable to use TFJob in Kubeflow Pipeline, current dsl compiler have not support for TFJOBOP #973
  • Make the data generated by a pipeline step available before the pipeline step completes. #959
  • Kubeflow version vs Jupyter Hub Image Versions Confusion #948
  • Moving and retrieving logs from persistent volume #940
  • Use DL VM as default container for Notebooks. #911
  • Unable to delete service account in e2e tests #893
  • How to accessing the source code from Images #881
  • ModuleNotFoundError: No module named 'kfp' #872
  • Error regarding syncing workflow #869
  • Errors in the test logs #865
  • Switch default package format to .zip #854
  • How to deploy distribute training with tensorflow or pytorch with pipelines #848
  • Add option in DSL compiler to output YAML #834
  • Flaky presubmit step in build-scheduledworkflow-image #831
  • Pipeline view UX is completely blank #824
  • kfctl.sh apply platform #820
  • 500 Error Jupyter notebook #818
  • Move e2e tests out of us-central1-a #807
  • Are there plans to allow kubeflow pipelines to be created through a manifest file? #804
  • Support inline html in html viewer #795
  • Namespace "kubeflow" is "Terminating" forever #768
  • Kubeflow 0.4 still include argo UI #753
  • pipeline step failed with exit status code 2: failed to save outputs #750
  • Allow eventing in Kubeflow Pipelines #725
  • Can not use NFS volume for Tensorboard logs #708
  • Unschedulable: 0/2 nodes are available: 2 Insufficient cpu. #706
  • Pipeline upload via URL doesn't work when port-forwarding from cloud shell #700
  • Error running TF training job on KF 0.3.5 using GPUs #686
  • "Google Compute Engine does not have enough resources available to fulfill request: us-central1-a" #673
  • Add bigquery component in release #643
  • Pipeline notebook samples : "Install Pipeline SDK " error #610
  • Deployer component to output model URL #593
  • TFMA output is empty #592
  • Pending run step shown as error #591
  • UI dashboard needs to show the pipelines system version #590
  • CMLE/TFX failure: Exceeded the max allowed number of models per project: 100. #587
  • File output is too long #577
  • Deployment problem with private cluster #534
  • How to deploy distribute training with tensorflow #533
  • ERROR: (gcloud.container.clusters.delete) One of [--zone, --region] must be supplied: Please specify location.. #488
  • Support TPU in DSL #485
  • Show an error message for output viewers that can't fetch their data #479
  • Recurring runs create runs with an embedded pipeline spec, rather than id #475
  • Unable to visualize any HTML content with Output Viewer #473
  • Improve logic to preload sample pipeline #455
  • Add component metadata to the generated Argo YAML for consumption by the UI #452
  • Enable users to launch and manage multiple Tensorboard instances within kubeflow pipelines #443
  • Python SDK is not continuous integration friendly. #420
  • K8s 1.11 fails on generateName when upgrading pipelines #409
  • TFJob doesn't stop trainer jobs after a timeout #408
  • TFJob doesn't forward error logs from the jobs #407
  • Pipeline GCP sample components need to switch to default gcp ops #406
  • Permanent storage for pipeline system #395
  • Error installing package with ks #372
  • Pre-install samples via bootstrapper. #356
  • Backend should not rely on presence of samples. #355
  • deploy problem: 3 pvc like mysql not created at the script and not specified label selector #352
  • RESOURCE_EXHAUSTED: Maximum number of keys on account reached #348
  • TFX Components pipeline- Expanding model analysis artifact does not go full screen #341
  • Support for Argo artifacts #336
  • Cloud platform scope not carried through on autoscaling #332
  • Unclear how much cpu/memory actually available #330
  • uninstall failed, there are twenty pod has not been deleted #329
  • Conditions are not visualized properly #321
  • Container op should expose method to append volume/mount/envs #311
  • UI says "Successfully created new Run", but it isn't in the list #308
  • Error running pipeline: cannot create tfjobs.kubeflow.org 403 #294
  • Python DSL should support multiple storage options (GCS, PVC, etc.) #275
  • Modify container images so that they support multiple data store options depending on parameters #274
  • Add sample tests for notebooks #263
  • UI feature ask: Search experiments and runs #257
  • Containers should not be allowed to directly create K8s objects #254
  • GKE cluster creation warnings #253
  • Consider never deleting the runs #247
  • Our ksonnet sometimes throws "SIGSEGV: segmentation violation" failing deployment #241
  • Get experiment by name, create run in experiment by name #237
  • UI should display documentations for the steps of a pipeline #227
  • Missing dropoff latitudes #225
  • Tensorboard not showing historical AUC / Accuracy #223
  • Our tests fail on external PRs #215
  • We need automatic labels indicating whether an issue/PR is internal/external. #207
  • Support submitting runs from local machine #206
  • Make it easy for people to write pipeline tests in python #203
  • Tests are launched in doubles since Nov 09 18:30 #197
  • Tests sometimes run multiple times for the same commit (creating multiple GKE clusters). #192
  • We have 40 GKE clusters (=100 instances) running tests and our quotas are exhausted #191
  • Instruction for running it on minikube #184
  • feature request: restore the client method for creating a pipeline #175
  • Create GCS path checker component #170
  • feature request: in SDK support 'get_or_create_experiment()' #161
  • Pipeline API Server Swagger Client (Go) for Pipeline Create does not take a "name" parameter #123
  • Pipeline API Server Swagger Client (Go) for Pipeline Upload returns incomplete output #122
  • Fasten release process for the image tag update. #115
  • Persist pod logs after they finish #93
  • Need to add notebooks with end-to-end sample scenarios #80
  • After the user creates a run, redirect to the run page, not the runs list page. #64
  • Remember the page I was on #62
  • Doesn't remove old containers (> maxHistory) #15
  • ScheduledWorkflow CRD: Investigate need for retries beyond the ones provided by Argo #5

Merged pull requests:

0.1.13 (2019-03-22)

Full Changelog

Closed issues:

  • bazel test command does not actually run tests #977
  • bazel build fails with fatal error: ares.h: No such file or directory #970
  • dsl.ContainerOp.after() only works with Ops that have already sanitized names #965
  • value in file_outputs is not being passed to input parameters correctly #957
  • Support Azure authentication for pipeline ops #953
  • [CMLE] trainer component should output the trained model URI #944
  • dependency analysis issue with 'component.yaml'-based ops #941
  • rename sample GCS path to fix model-name typo #936
  • GitHub checks have been enabled in this repo, but are not supported by Tide. #930
  • Cannot find module for path ml_metadata/metadata_store/mlmetadata #928
  • Pipeline name is not checked #825
  • Can not mount volume in Container op #477
  • Travis tests do not properly test what happens when the branch is merged. #431
  • Full solution to tests and code desync problem #200
  • Prow code, test image code and code being tested can still desync. #198
  • SDK/Components/Python - Functions that do not return anything #186

Merged pull requests:

  • Release component image version f98ec68488cfbac022893a87c5bca083f03f2065 #1021 (IronPan)
  • Fix hard-coded model export dir name #1015 (hongye-sun)
  • Fixed bug in docstring construction #1012 (Ark-kun)
  • Update TFX Pipeline Example Readme #1011 (rculbertson)
  • Update sidebar styling #1010 (rileyjbauer)
  • SDK/DSL/Compiler - Fixed handling of empty pipeline name #1009 (Ark-kun)
  • Fix dataflow and mlengine samples with recent changes #1006 (hongye-sun)
  • Updating the description for model_uid param #1002 (animeshsingh)
  • Adjustments for OpenVINO v5 #997 (dtrawins)
  • use pending commit id for cluster and source code name #994 (IronPan)
  • Ignore not found error in delete cluster #991 (hongye-sun)
  • Update component.yaml for dataflow and cmle with recent changes. #987 (hongye-sun)
  • Add job_dir.txt as output files of CMLE trainning job. #986 (hongye-sun)
  • Improve dataflow job to use GCS to keep temp job states. #985 (hongye-sun)
  • Add component.yaml to every ibm-components #984 (Tomcli)
  • Add instruction on building viewer CRD controller image #981 (terrytangyuan)
  • well-defined parameter types #978 (gaoning777)
  • Added the component name to the docstring #976 (Ark-kun)
  • Add unit tests pipelineparam #975 (gaoning777)
  • Fix sample test failure because of the type information in the pipelineparam #972 (gaoning777)
  • Add documentation and sample notebook for dataproc components. #971 (hongye-sun)
  • Add optional pipeline_id to run_pipeline api #968 (hongye-sun)
  • Fix cancellation bug for dataproc job. #967 (hongye-sun)
  • Add missing main_class arg to spark job and set default values. #966 (hongye-sun)
  • Set default values for payload objects in case of empty. #962 (hongye-sun)
  • Add fake metadata store and fix tests. #958 (neuromage)
  • Add dataproc component yaml files #956 (hongye-sun)
  • helper fn to add az secrets to container #954 (rakelkar)
  • Fixing the broken links for components source code in XGBoost Spark P… #952 (animeshsingh)
  • Watson Openscale Components to manage models #950 (animeshsingh)
  • SDK/DSL/Compiler - Fixed compilation when using ContainerOp.after #943 (Ark-kun)
  • Add type check #938 (gaoning777)
  • Components - GCP: Fixed typo: Replaced cencus with census #937 (Ark-kun)
  • Add code formatting to markdown viewer #935 (yebrahim)
  • Components - Keras - Train classifier: Changed the types to conform to the chosen type system #933 (Ark-kun)
  • Add notebook instructions on how to query ML Metadata for output artifacts #932 (neuromage)
  • SDK - Added support for loading zip-packed components #931 (Ark-kun)
  • update changelog for 0.1.12 #925 (gaoning777)
  • SDK - Fixed small bug in DSL code that generates unique names for ops #923 (Ark-kun)
  • Samples - Improved the TFX OSS notebook and README #922 (Ark-kun)
  • Detecting file format using signature instead of file extension #919 (Ark-kun)
  • Add Tommy as the approver for the IBM components and samples. #915 (Tomcli)
  • Pass meta to containerop and pipeline #905 (gaoning777)
  • DataProc commands implementation #862 (hongye-sun)
  • SDK/Components - Added naming.generate_unique_name_conversion_table #716 (Ark-kun)
  • Added the ability to terminate a run #528 (Ark-kun)

0.1.12 (2019-03-06)

Full Changelog

Merged pull requests:

0.1.11 (2019-02-27)

Full Changelog

Closed issues:

  • ROADMAP for pipelines #675
  • python components don't work for images with python 2 #666
  • dsl PipelineParam does not work under Image or Command #521

Merged pull requests:

0.1.10 (2019-02-15)

Full Changelog

Closed issues:

  • Pipeline Release V0.1.10 #805
  • Failed to create recurring run #802
  • Persistent Agent doesn't like v1beta1 scheduledworkflow in on-prem scenario #790
  • Metric bars do not always show up correctly #775
  • ml-pipeline-persistenceagent restarts forever #741
  • 403 Insufficient Permission error when writing to GCS #709
  • Support archiving/unarchiving runs on the UI #621
  • Incorrect validation in new run form when filling name then pipeline #612

Merged pull requests:

0.1.9 (2019-02-06)

Full Changelog

Closed issues:

  • XGBoost sample failure #736
  • TFX taxi client utility run abnormally after serving started #728
  • Pipeline deploy component does not support local model #726
  • Failed to execute the component roc and confusion_matrix in TFX sample for local mode #719
  • An error occurs when run TFX example in local kubeflow cluster #703
  • Not using secret when pulling from private Registry #695
  • Release 1.7 - TFX taxi cab example failing the deploy step #692
  • Run details page should autorefresh #683
  • problem when deploying kubeflow 0.4.0 #676
  • Do I need to be running kubeflow on GCP to use pipelines? #661
  • Design doc for pipelines #569
  • Some error logs in sample for ML - TFX - Taxi Tip Prediction Model Trainer #547
  • Return the total number of resources in list APIs #103

Merged pull requests:

  • Creating ROADMAP.md #786 (paveldournov)
  • Revert GPU tests to unblock release #778 (hongye-sun)
  • Modifications done by the execution of the release script. #777 (vicaire)
  • Refactor BaseOp to context manager pattern #762 (hongye-sun)
  • Add component_sdk tests to travis tests #761 (hongye-sun)
  • Fix Dockerfile for viewer CRD image. #760 (neuromage)
  • Add UI actions to Buttons module #758 (yebrahim)
  • Regenerate frontend API files now that listCount APIs are merged #757 (rileyjbauer)
  • Enhance deploy.sh to support local storage #755 (jinchihe)
  • Releasing components fixing xgboost #747 (gaoning777)
  • support pipeline level imagepullsecret in DSL #745 (gaoning777)
  • Updated dev guide for apiserver build with minikube. #743 (gyliu513)
  • Add Dockerfile for building Viewer CRD controller. #740 (neuromage)
  • Remove pipeline bootstrapper #739 (IronPan)
  • Refactor UI buttons to lib file #737 (yebrahim)
  • fix dataproc cluster version #735 (gaoning777)
  • Improve runtime graph starting and running experience #734 (rileyjbauer)
  • SDK/Components - convert_object_to_struct now uses __init__ to get field list #733 (Ark-kun)
  • SDK/Components - Improved error when type checking fails in constructor #732 (Ark-kun)
  • Create kfp componet sdk #729 (hongye-sun)
  • Add CMLE sample test script. #724 (qimingj)
  • Add "set_retry()" on ContainerOp. #723 (qimingj)
  • Auto-refreshes the run details page #722 (rileyjbauer)
  • create nonexistent directory for roc and confusion_matrix component #720 (jinchihe)
  • SDK/Components/Python - Improved Python2 compatibility #718 (Ark-kun)
  • SDK/Components/Python - Made the typing.NamedTuple import optional #717 (Ark-kun)
  • SDK/Components - Added _naming._convert_to_human_name function #715 (Ark-kun)
  • SDK/Components - Command line args can only be strings or placeholders #711 (Ark-kun)
  • SDK/Components - More meaningful error when trying to convert graph component to ContainerOp #710 (Ark-kun)
  • randomize the deployment name to avoid collision #704 (gaoning777)
  • SDK - Removed stray debug print from tests #701 (Ark-kun)
  • Update changelog for release 0.1.8 #697 (yebrahim)
  • Add gpu pool to test deployment and enable gpu in sample test #696 (hongye-sun)
  • bump ks version in presubmit test #693 (IronPan)
  • SDK - Update minimum Python version to 3.5.3 #691 (Ark-kun)
  • Update all Pipelines CRD versions to v1beta1. #681 (neuromage)
  • Return resource count from ListXXX calls #595 (yebrahim)

0.1.8 (2019-01-17)

Full Changelog

Closed issues:

  • Docs about how to write mlpipeline-ui-metadata and mlpipeline-metrics artifacts #660
  • Pipeline run created from python shows experiment but not pipeline (in UI). #617
  • Load samples returns an error in 0.4 Kubeflow #603
  • Error deploying latest bootstrapper #594
  • Provide argument to assign GCP service account to use for DSL operator #218
  • Run status tooltip should include creation time #61

Merged pull requests:

  • Components release 2ed60100d1db9efeb38c6c358f90b21c144179be #694 (yebrahim)
  • add gcp credential for tf-job template #689 (IronPan)
  • improve the list run query #687 (IronPan)
  • Adds a link in the side nav to the KF 'usage-reporting' doc #682 (rileyjbauer)
  • Fix Makefile to add licenses using Go modules. #674 (neuromage)
  • Include date in run status tooltips #671 (rileyjbauer)
  • Add build version to side nav #670 (rileyjbauer)
  • SDK/Tests - Fixed the output ordering instability in tests #664 (Ark-kun)
  • SDK/Components - Added /data to the generated file paths #663 (Ark-kun)
  • SDK/Components - Simplified _create_task_factory_from_component_spec function #662 (Ark-kun)
  • SDK/Components - Renamed serialized command-line argument tokens #659 (Ark-kun)
  • update change log #657 (IronPan)
  • Add simple filtering by name to CustomTable #656 (rileyjbauer)
  • SDK/Components - Reworked the component model structures. #642 (Ark-kun)
  • Support filtering on storage state #629 (yebrahim)
  • SDK - Fixed the unittest script #556 (Ark-kun)

0.1.7 (2019-01-09)

Full Changelog

Closed issues:

  • Use "Start" for runs, "Create" for experiments #649
  • go vet in Travis tests breaks tests for unrelated PRs #646
  • Building backend using Bazel fails on mac #638
  • ml-pipeline-persistenceagent fails a few times. #624
  • Unable to plug-in default values to Pipeline #618
  • Add deployed model cleanup code to the Kubeflow notebook #608
  • Compare page perf optimizations #597
  • OAth client instructions are ambiguous #586
  • "Waiting for the IAP setup to get ready..." after clicking "Skip IAP" #585
  • UI should allow creating a run with no experiment #573
  • studyjob-controller start failed #546
  • Failing e2e sample tests do not log any errors #515
  • Authentication and service account plan for Pipeline + Kubeflow #374
  • Error getting logs #290
  • Support filtering in list APIs #270
  • Switch to Go 1.11 modules and package management. #187

Merged pull requests:

  • Release component image version d3c4add0a95e930c70a330466d0923827784eb9a #655 (IronPan)
  • Correctly ignore src/apis when building frontend #654 (yebrahim)
  • Use linguist annotations to skip diffing generated files #652 (yebrahim)
  • Use "create" rather than "start" except when initiating a run #650 (rileyjbauer)
  • Fix shadowing errors in Viewer reconciler #648 (neuromage)
  • Add IS_SUBSTRING operator for use in API resource filtering. #645 (neuromage)
  • Add changelog to pipeline repo #644 (IronPan)
  • Update WORKSPACE and BUILD files incoporating recent changes #639 (neuromage)
  • fix deploy model name conflict in case of concurrent notebook sample test #636 (gaoning777)
  • Print sample test logs in case of exception throw #635 (gaoning777)
  • Expose pipeline/job API through setup.py #634 (IronPan)
  • Fix retrying logic which was causing persistenceagent to crash loop. #633 (neuromage)
  • Add resnet-cmle sample back. Update all component images. #632 (qimingj)
  • Updates material-ui and react npm libraries #630 (rileyjbauer)
  • Generate pipeline and job python client as part of SDK #628 (IronPan)
  • Fix gpu sample issues #627 (hongye-sun)
  • Run go vet as part of the Travis CI. #626 (neuromage)
  • Sanity check filtering/sorting options in list requests. #625 (neuromage)
  • Support replacable arguments in command as well (besides arguments) in container op. #623 (qimingj)
  • Update sample notebook to clean up deployed models. #622 (qimingj)
  • URLEncode instead of base64 encode the filter string #620 (neuromage)
  • DSL refactor #619 (gaoning777)
  • Expose that the python API is Python3.5+ only #616 (TimZaman)
  • Load sample when pipeline initially started #615 (IronPan)
  • Use Bazel to build the entire backend and perform API code generation #609 (neuromage)
  • Improve condition sample to demonstrate ==, >= and <. #607 (qimingj)
  • fix for boostrapp problem #602 (xiaozhouX)
  • Initial version of BigQuery query execution component. #601 (cbreuel)
  • First step to bring back CMLE sample. #599 (qimingj)
  • Compare perf - pure components, disable ROC curve thumbnail animations #598 (yebrahim)
  • Move backend unit tests to Travis #589 (yebrahim)
  • Deployment - Minikube support - Passing the platform parameter to kfctl #588 (Ark-kun)
  • Fix the List run to get all runs #583 (IronPan)
  • retry on create table in api server #582 (IronPan)
  • switch from go dep to go module #581 (IronPan)
  • Add sample test without image build #578 (gaoning777)
  • remove xgboost compiled sample #576 (gaoning777)
  • Add a gpu sample #575 (hongye-sun)
  • Backend - Removed hardcoded metrics file name #574 (Ark-kun)
  • update dockerfile and add build step of frontend #567 (chenzhiwei)
  • Encode filter parameter as a base64-encoded JSON string in List requests #563 (neuromage)
  • Tests - Updated image-builder Makefile #500 (Ark-kun)
  • Add the Viewer CRD controller for managing web views such as Tensorboard instances from within the Pipelines UI. #449 (neuromage)

0.1.6 (2018-12-20)

Full Changelog

Closed issues:

  • Support creating runs without experiments #560

Merged pull requests:

0.1.5 (2018-12-20)

Full Changelog

Closed issues:

  • remove duplicate volumes in the DSL #557
  • Cut a Pipeline v0.1.5 release #549
  • "Information in the Argo UI appears much faster compared to the KF Pipelines UI" #544
  • KubeFlow Pipeline example notebook is half way updated for 0.1.4 #520
  • UI loses state after being idle for a while #502
  • No way in the UI to upload a pipeline from the cluster itself #495
  • Confusion Matrix UI Doesn't Work if My Labels are "True", "False" #446
  • input parameter for creating pipelines - does not allow camelCase #445
  • Pipeline install timeout #414
  • 'Run output' tab for Run details page #331
  • Create a container wrapper op with Default GCP auth #310
  • Test/Server code out of sync, blocking PRs #196
  • Show results of the entire workflow in one view #92
  • User should be able to start a new run from the pipeline page. #53

Merged pull requests:

0.1.4 (2018-12-07)

Full Changelog

Closed issues:

  • DSL Condition Feature/Bug #481
  • Persist pod logs to permanent storage #439
  • Pipeline name does not show up when submitting from SDK #347

Merged pull requests:

0.1.3 (2018-12-05)

Full Changelog

Closed issues:

  • Bug in pipeline samples' parameters #478
  • Frontend integration test is flaky #464
  • Jupyter Notebook [tf-hub-0] run failed #458
  • AssertionError [ERR_ASSERTION]: logs do not look right: 1 #453
  • Why can not delete experiments and runs? #441
  • grant kubeflow user account CMLE permission #416
  • where is bootstrapper code gcr.io/ml-pipeline/bootstrapper:0.1.2 #412
  • Compare page tests #382
  • Pipeline UI should work behind Kubeflow reverse proxy (Ambassador) #373
  • Presubmit tests are broken since we rely on pulling node.js image that was removed #370
  • Backend ships with very old SDK version #358
  • Python fire interprets string incorrectly #318
  • Include kubectl install command in Github page #284
  • Embeddable run view page #182
  • Better render pipeline description #89
  • Support cloning run started from notebook #75
  • "No runs were found for this experiment" is a misleading message when the list is still being loaded. #63
  • User should be able to create a new Recurring Run from the pipeline page. #54

Merged pull requests:

  • Return string from pod logs #476 (yebrahim)
  • Fix serializing cloned embedded pipeline #474 (yebrahim)
  • Refresh while waiting for run to start #472 (yebrahim)
  • Add article on Jupyter notebooks #471 (lakshmanok)
  • Adds tests for the run comparison page #469 (rileyjbauer)
  • Sanitize inputs from SDK when submitting run. #466 (qimingj)
  • Support cloning runs created with an embedded pipeline #465 (yebrahim)
  • import bug #463 (gaoning777)
  • Wait until logs appear right in e2e test #459 (yebrahim)
  • Fix sample test - Add gcp permission #454 (IronPan)
  • Show pipeline details embedded in runs #447 (yebrahim)
  • SDK/PythonContainer - Compiling pipelines without needing kubernetes #442 (Ark-kun)
  • SDK/Components - Renamed container.arguments to container.args #437 (Ark-kun)
  • DSL - return self for .after() #435 (IronPan)
  • Tests - Try fix GCS file being inaccessible right after upload #433 (Ark-kun)
  • Testing - Fix Travis tests (SDK/Components) #432 (Ark-kun)
  • SDK/GCP - Replaced default_gcp_op with task.apply(use_gcp_secret) #430 (Ark-kun)
  • Fix tb viewer test #427 (yebrahim)
  • SDK/Client - Removed import six #425 (Ark-kun)
  • SDK/Components/PythonContainerOp - Make the local output path configurable #424 (Ark-kun)
  • Propagate secret to kaniko #423 (IronPan)
  • fix bug in the notebook sample #419 (gaoning777)
  • Failed the sample tests when it should #417 (gaoning777)
  • Fix for k8s dict parsing #411 (vanpelt)
  • Clean up the resource after test finish #410 (IronPan)
  • Add loading spinner to custom table while loading items #405 (yebrahim)
  • SDK/Components/PythonContainerOp - Switch from dict to ComponentSpec #396 (Ark-kun)
  • SDK/Components - Removed outputs from task factory function signature #388 (Ark-kun)
  • Add Gopkg dependency for kubernetes code-generator. #371 (neuromage)
  • SDK - Improve the python package build #364 (Ark-kun)
  • SDK/DSL - Added support for conditions: !=, <, <=, >=, > #309 (Ark-kun)
  • SDK/Components - Support for optional inputs #214 (Ark-kun)
  • SDK/Components/PythonContainerOp - Simplified GCSHelper by extracting duplicate code #210 (Ark-kun)

0.1.3-rc.3 (2018-11-29)

Full Changelog

Closed issues:

  • Pipeline names are truncated and the columns are not realizable #400
  • RunDetails page tests #383
  • ExperimentDetails page tests #381
  • PipelineDetails page tests #379
  • Our tests clone our repo 40+ times for each PR commit (and some PRs have dozens of commits) #368
  • Backend image build does not fail when sample compilation fails #354
  • Back-button not working when creating a run from a notebook #221

Merged pull requests:

  • Adds tests for the ExperimentDetails page #404 (rileyjbauer)
  • Removed unused docs directory including OWNERS file and images. #401 (sarahmaddox)
  • Pointed doc links to Kubeflow website instead of wiki. #398 (sarahmaddox)
  • RunDetails test suite, bug fixes #394 (yebrahim)
  • update image tag with the new release #393 (gaoning777)
  • Increase timeout waiting for test run to start #392 (yebrahim)
  • Delete kf_deploy.sh #391 (IronPan)
  • Clean up vendor directory #390 (neuromage)
  • Add support for minio hosted artifacts #389 (vanpelt)
  • Backend - Fixed handling of sample compilation failure #387 (Ark-kun)
  • Add finish timestamp to backend API interceptor #386 (yebrahim)
  • remove fire dependency in the component image build #384 (gaoning777)
  • PipelineDetails page tests #380 (yebrahim)
  • update samples with dependency requirement in the component image build #378 (gaoning777)
  • Disables back button when there is no history #377 (rileyjbauer)
  • move default gcp op to dsl/ #376 (IronPan)
  • Add Amy and Lak's Blogs Link to README #375 (qimingj)
  • Tests - Improve Minikube source code copy process #367 (Ark-kun)
  • restructure local components directory #357 (gaoning777)
  • Switching test to kubeflow deployment #351 (IronPan)
  • added component with tf slim model generator and pipeline adjustments #335 (dtrawins)
  • PipelineSelector, RecurringRunsManager, and 404Page tests #319 (yebrahim)
  • Tests - Fixed many test issues (external PRs, code de-sync) by getting rid of git clone #315 (Ark-kun)
  • Now pipeline function takes direct default values rather than dsp.PipelineParam. #110 (qimingj)

0.1.3-rc.2 (2018-11-22)

Full Changelog

Closed issues:

Merged pull requests:

  • Tests - Stop repeating the logs three times #366 (Ark-kun)
  • Fixes issue with footer overlapping side panel and fixes summary layout #353 (rileyjbauer)
  • PipelineDetails cleanup #350 (yebrahim)
  • Add support for nvidia gpu limit #346 (hongye-sun)
  • mount gcp credentials for kaniko worker #343 (IronPan)
  • rename ks registry ml-pipeline -> pipeline #340 (IronPan)
  • restructure dataflow component structure #338 (gaoning777)
  • formating jsonnet registry #333 (IronPan)
  • Use upper case for ks parameter #328 (IronPan)
  • Use ks env for pipeline namespace #326 (IronPan)
  • Tests - Fixed the argo submit failure not failing the script #324 (Ark-kun)
  • SDK/Components - Renamed DockerContainer spec to to Container #323 (Ark-kun)
  • fix bug for the component build with requirements #320 (gaoning777)
  • Use latest npm and node on travis, fix tests #317 (yebrahim)
  • add metric image file for wiki #316 (hongye-sun)
  • Add default_gcp_op #314 (IronPan)
  • return self for container op initialization methods to allow chaining the construction #313 (IronPan)
  • switch set() to add() for better inheritance. #312 (IronPan)
  • Changes "Hide" button on Pipeline summary to fully hide the paper card #305 (rileyjbauer)
  • add ngao to test owner because frequent contributions to this directory, including e2e, sample tests and imagebuilder. #299 (gaoning777)
  • reenable basic sample tests, serially after frontend integ tests(triggering 5 times, successful always) #298 (gaoning777)
  • Add run status to page title #287 (yebrahim)
  • Update button styling for new experiment and new run #264 (rileyjbauer)
  • refactor component build codes #260 (gaoning777)
  • Remove tfjob operator from sdk #233 (IronPan)
  • SDK/Components/PythonContainerOp - Fixed string escaping warning #208 (Ark-kun)
  • SDK/DSL/Compiler - Improved compilation of dsl.Conditional - UX support done #177 (Ark-kun)

0.1.3-rc.1 (2018-11-17)

Full Changelog

Closed issues:

  • Cannot copy pipeline source on Pipeline details page #301
  • Notebook Kaniko job shows logs in red background #288
  • a pipeline must be sellected #286
  • Create experiment from Notebook and use link to open KFP UI. Back button doesn't work. #281
  • Text/link displayed after calling run_pipeline should be "Run link" not "Job link" #266
  • Perf issue - All runs page loads very slowly with >100 runs #259
  • UI - feature ask: Remember the "Rows per page:" setting #258
  • Upload Error Invalid response: 404 #239
  • Auth Error when running TFX notebook. #229
  • Cannot create a Notebook when starting KFP from cloud shell #179
  • Unsupported Scan Error While Listing the Jobs of an Experiment #171
  • Our test code and test images code is not always the same. #163
  • Popped out TFMA is too small #160
  • SDK should require kubernetes client lib #158
  • Experiment list title should not change #71
  • Create a sample notebook #69
  • Frontend should give 404s for non-existing pages/routes. #55
  • Remove python op decorator #44

Merged pull requests:

0.1.2 (2018-11-08)

Full Changelog

Merged pull requests:

0.1.1 (2018-11-07)

Full Changelog

Merged pull requests:

0.1.0 (2018-11-06)

Full Changelog

Closed issues:

  • The "Recurrent run configs" tab is showing inaccurate # of jobs #100
  • Duplicate experiment name would return me a blank page #97
  • The Experiments/All runs tab should probably be the default tab in Experiments. #58
  • The Experiments tab should be the default tab. #57

Merged pull requests:

  • Use the experiment's resource reference in the listJobs request #105 (yebrahim)
  • Fix validation check for maximum size limit #104 (IronPan)
  • Add Ning and Alexey to OWNERS for components, samples and sample-test #102 (Ark-kun)
  • Don't barf when experiment name is already used #101 (yebrahim)
  • CSS changes for nav menu and tables #99 (ajayalfred)
  • Fixed the Minikube tests after moving to the new repo #98 (Ark-kun)
  • sort by run display name by default #96 (IronPan)
  • SDK/DSL/Compiler - Reverted fix of dsl.Condition until the UI is ready. #94 (Ark-kun)
  • debug tfma failure #91 (gaoning777)
  • fix miscellaneous List API issue #90 (IronPan)
  • mlp -> kfp.dsl #88 (Ark-kun)
  • ExperimentList tests, use immer.js #86 (yebrahim)
  • Moves docs from pipelines main README to wiki #83 (sarahmaddox)
  • Add rileyjbauer to frontend OWNERS #82 (yebrahim)
  • Update OWNERS to add qimingj #77 (qimingj)
  • Compile samples instead of hard code them in API server #76 (IronPan)
  • Added @gaoning777 and @Ark-kun to OWNERS #74 (Ark-kun)
  • Change title from 'Start a new run' to 'Start recurring run' as appropriate #73 (rileyjbauer)
  • Add %%docker magic to jupyter kernel. #72 (qimingj)
  • Build Python SDK in the releasing #70 (gaoning777)
  • Fix tfx name bug in the tfma sample test #67 (gaoning777)
  • Fixes off-by-one error with months in Trigger #66 (rileyjbauer)
  • Updating OWNER files. Adding per-subdirectory OWNER files. #65 (vicaire)
  • This is a test. #51 (vicaire)
  • PipelineList tests, fix clearing error banner #50 (yebrahim)
  • Adds the Kubeflow logo to the side nav. Fetches specific font weights #48 (rileyjbauer)
  • Refresh button now reloads logs and artifacts as well on RunDetails page #47 (rileyjbauer)
  • Cleans up StaticGraphParser tests as per PR commends from #30 #45 (rileyjbauer)
  • Fix resnet-cmle sample. #43 (qimingj)
  • Add ngao to the owners #42 (gaoning777)
  • SDK/Tests/Components - Corrected test argument types #41 (Ark-kun)
  • SDK/Components/Python - Stopped using the Fire library #40 (Ark-kun)
  • Upgrade kubeflow to v0.3.2 #39 (IronPan)
  • Components - Removed debug print #38 (Ark-kun)
  • SDK/Tests/Components - Improve temporary file handling #37 (Ark-kun)
  • Add Alexey Volkov to OWNERS #36 (Ark-kun)
  • Propagate name for runs from scheduled job #33 (IronPan)
  • add rileyjbauer to OWNERS #32 (rileyjbauer)
  • Push pagination control to CustomTable #31 (yebrahim)
  • Adds many more static workflow parser tests #30 (rileyjbauer)
  • SDK/Components - Switching to map-based syntax for the arguments. #29 (Ark-kun)
  • SDK/DSL/Compiler - Fixed compilation of dsl.Condition #28 (Ark-kun)
  • TFMA deployer bug fix #27 (gaoning777)
  • Upgrading the container versions to 0.0.42, the version of the first release of kubeflow/pipelines. #26 (vicaire)

0.0.42 (2018-11-02)

Closed issues:

  • ScheduledWorkflow CRD: CLI #7
  • How does it proceed with this project? #1

Merged pull requests:

* This Change Log was automatically generated by github_changelog_generator