Releases: mlrun/ce
mlrun-ce-0.5.2-rc3
MLRUn Open Source Stack
mlrun-ce-0.5.2-rc2
MLRUn Open Source Stack
mlrun-ce-0.5.2-rc1
MLRUn Open Source Stack
mlrun-ce-0.5.1
MLRUn Open Source Stack 0.5.1
What's Changed
- [Feature Store] Change feature store default targets to
parquet
by @Tankilevitch in #35 - [Jupyter] Fix MLRun UI URL generated by jupyter to include
/mlrun
by @quaark in #37 - Consolidate CE to one Full flavor of deployment by @quaark in #36
- Bump nuclio and fix requirements.lock by @quaark in #38
- Artifact path and Feature-store targets default-path should use s3 schema and the standard url template by @assaf758 in #39
- Bump MLRun Version + 0.5.1-rc2 by @quaark in #40
- Fix Broken Chart by @quaark in #41
- CI - Add lint by @liranbg in #43
- Doc - working with ecr by @liranbg in #42
- Bump Nuclio to v1.11.6 by @TomerShor in #44
- Fix Spark Webhook and Service Account Name by @quaark in #45
- Fix Requirements lock by @quaark in #46
- MLRun Configs For Displaying Values in MLRun UI by @quaark in #48
- Fix Chart Serialization Error by @quaark in #49
- Release - 0.5.1 by @liranbg in #50
New Contributors
- @Tankilevitch made their first contribution in #35
- @assaf758 made their first contribution in #39
- @liranbg made their first contribution in #43
Full Changelog: mlrun-ce-0.5.0...mlrun-ce-0.5.1
mlrun-ce-0.5.1-rc6
MLRUn Open Source Stack
mlrun-ce-0.5.1-rc5
MLRUn Open Source Stack
mlrun-ce-0.5.1-rc4
MLRUn Open Source Stack
mlrun-ce-0.5.1-rc2
MLRUn Open Source Stack
mlrun-ce-0.5.1-rc1
MLRUn Open Source Stack
mlrun-ce-0.5.0
🎉 🎉 We’re happy to announce GA of MLRun Community-Edition first version 🎉 🎉
MLRUn Open Source Stack
The community edition marks a milestone. This is the first ever open-source MLOps release that, out-of-the-box, includes not just the components required to execute the models and pipelines in production, but also integration with services. In addition, it includes a fully functional feature store as part of the MLOps stack where the online service is implemented on top of Redis.
It also includes a CloudFormation stack that deploys the community edition to AWS, saving users significant time set-up and provision resources.
Lite
- MLRun
- Nuclio
- Real-time serving pipelines
- MPI / Horovod for distributed training
- Jupyter with out-of-the-box configuration and demos
- Minio for artifact storage
- Redis integration (optional, the Redis service needs to be installed independently)
Full:
- Spark operator
- Kubeflow pipelines
- Prometheus
- Grafana