Releases: kubeflow/pipelines
Version 0.1.12
To deploy this version, visit
https://deploy.kubeflow.cloud/#/deploy?version=a1afdd6c4b297b56dd103a8bb939ddeae67c2c92
Or following this instruction and set
export KUBEFLOW_TAG=a1afdd6c4b297b56dd103a8bb939ddeae67c2c92
Install python SDK (python 3.5 above) by running:
pip3 install https://storage.googleapis.com/ml-pipeline/release/0.1.12/kfp.tar.gz --upgrade
See the Change Log
Version 0.1.11
Install python SDK (python 3.5 above) by running:
pip3 install https://storage.googleapis.com/ml-pipeline/release/0.1.11/kfp.tar.gz --upgrade
Version 0.1.10
Deploy this release to GCP
https://deploy.kubeflow.cloud/#/deploy?version=e1408f2293ce72b5a89a9b40bc1998c56ac0d73b
Install python SDK (python 3.5 above) by running:
pip3 install https://storage.googleapis.com/ml-pipeline/release/0.1.10/kfp.tar.gz --upgrade
Connect to the UI:
kubectl port-forward --namespace kubeflow $(kubectl get pod --namespace kubeflow --selector="service=ambassador" --output jsonpath='{.items[0].metadata.name}') 8080:80
open http://localhost:8080/pipeline
See the Change Log
Version 0.1.9
To deploy this version, visit
https://deploy.kubeflow.cloud/#/deploy?version=64a7c55ea8a9b3732a3bba6d25057f5dbf301cfa
Or following this instruction and set
export KUBEFLOW_TAG=64a7c55ea8a9b3732a3bba6d25057f5dbf301cfa
Install python SDK (python 3.5 above) by running:
pip3 install https://storage.googleapis.com/ml-pipeline/release/0.1.9/kfp.tar.gz --upgrade
See the Change Log
Version 0.1.8
Install python SDK (python 3.5 above) by running:
pip3 install https://storage.googleapis.com/ml-pipeline/release/0.1.8/kfp.tar.gz --upgrade
See the Change Log.
Version 0.1.7
This pipeline version is included in Kubeflow v0.4.1. To deploy, visit
https://www.kubeflow.org/docs/started/getting-started-gke/
Install python SDK (python 3.5 above) by running:
pip3 install https://storage.googleapis.com/ml-pipeline/release/0.1.7/kfp.tar.gz --upgrade
Change Log
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 d3c4add #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 vetas 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)
Version 0.1.6
You can install ML Pipeline services by running:
kubectl create -f https://storage.googleapis.com/ml-pipeline/release/0.1.6/bootstrapper.yaml
Install python SDK (python 3.5 above) by running:
pip3 install https://storage.googleapis.com/ml-pipeline/release/0.1.6/kfp.tar.gz --upgrade
Access UI instructions:
https://www.kubeflow.org/docs/guides/accessing-uis/
Changelog since v0.1.5
- Bumping the sample version
Version 0.1.5
You can install ML Pipeline services by running:
kubectl create -f https://storage.googleapis.com/ml-pipeline/release/0.1.5/bootstrapper.yaml
Install python SDK (python 3.5 above) by running:
pip3 install https://storage.googleapis.com/ml-pipeline/release/0.1.5/kfp.tar.gz --upgrade
Changelog since v0.1.4
- Bumping the sample version
Version 0.1.4
You can install ML Pipeline services by running:
kubectl create -f https://storage.googleapis.com/ml-pipeline/release/0.1.4/bootstrapper.yaml
Install python SDK (python 3.5 above) by running:
pip3 install https://storage.googleapis.com/ml-pipeline/release/0.1.4/kfp.tar.gz --upgrade
Changelog since v0.1.3
SDK
- Support Cloud TPU
Version 0.1.3
You can install ML Pipeline services by running:
kubectl create -f https://storage.googleapis.com/ml-pipeline/release/0.1.3/bootstrapper.yaml
Install python SDK (python 3.5 above) by running:
pip3 install https://storage.googleapis.com/ml-pipeline/release/0.1.3/kfp.tar.gz --upgrade
Changelog since v0.1.2
SDK
- Support setting GPU limit and node selector to specify GPU type (#346)
- Support Kubernetes Volume, VolumeMount and Env APIs for Container Operator(#300)
- Add option to Container Operator to mount default GCP service account credential(#430)
- SDK/Components/Python - Removed python_op in favor of python_component (#85)
- SDK/DSL - Improved compilation of dsl.Conditional (steps->dag) (#177)
- SDK/Components - Renamed dockerContainer spec to container (#323)
- SDK/Components - Renamed container.arguments to container.args (#437)
- SDK/DSL - Added support for conditions: !=, <, <=, >=, > (#309)
- SDK/DSL - Pipeline function takes direct default values rather than dsp.PipelineParam. (#110)
- SDK/Components/Python - add support for dependencies in the component image building (#219)
- Notebook - Display highlighted logs only when there are errors. (#292)