2023b-v1.13.0

We have completed the release for sprint 1.36 🎉 🎊 🥳

Notable Changes

This release brings several noteworthy changes and improvements. Here's a summary of what you can expect:

• VS-Code & RStudio: User can directly use the kustomize base to deploy vscode and Rstudio from this release. With OpenDatahub 2.4, the vscode and rstudio, were directly available on OOTB workbench. With this developer can deploy vscode and rstudio along with all the other OOTB on developer mode.

  • #300

• VS-Code: With this release, the user can utilize the code-server built on top of the UBI9 image.
  Users can use the images with the prefix codeserver-ubi9 (ex:codeserver-ubi9-python-3.9-20231116-271208a) from the image registry to utilize UBI9 base codeserver.

  • #319.

• Annotation: image-tag-outdated: The user can use the annotation image-tag-outdated in the imagestream, to disable the tag from showing up in the Opendatahub Dashboard. This feature can be used in the upcoming Opendatahub version 2.5. With the older image would be disabled from selection and only be available for run in pre-existing workbenches.

  • #154
  • #315

• Annotation: build-info: With next release imagestream will have associated commit id added as annotations, so users can differentiate directly from imagestream, instead of looking up the images.

  • #307

• Runtime-images: The base runtime-images supplied from the notebooks for Elyra pipeline, are adjusted to have all the elyra based-python package requirements to be pre-installed in them. The package no longer be installed during the runtime.

  • #202
  • #330

Improvements and Bug Fixes
We've addressed several critical issues and bugs that were affecting the performance and usability of our notebooks. These fixes enhance the overall stability and functionality of our notebooks as listed below.

• Elyra validate : The elyra-validate command is fixed, in past version it was failing to validate pipeline if they contained pipeline properties.

  • #257
  • #324

• jupyterlab notification disabled: The jupyterlab notification pop-up is now disabled.

  • #315

What's Changed

• Move vscode and r-studio to the base layer in ODH by @atheo89 in #300
• [Fix] typo in some docker image labels k9s instead of k8s by @jstourac in #308
• feature: Included annotation to the outdated image tags by @harshad16 in #315
• chores: upgrade of libnghttp2 binary package in ubi8 by @harshad16 in #314
• adding build commit info annotations in imagestream by @dibryant in #307
• chores: Included template for issues creation by @harshad16 in #313
• [Fix] typo of the TensorFlow name in HabanaAI notebook manifest by @jstourac in #325
• Upgrade codeflare-sdk with the latest 0.12.0 by @harshad16 in #326
• fix: trustyai annotation and habana version to 2023.2 by @harshad16 in #328
• fix: disable announcement of jupyterlab notification by @harshad16 in #318
• Update Imagestreams in favor codeflare upgrade from 0.8 to 0.12 by @atheo89 in #333
• Rebase VSCode to UBI9 instead on C9S by @atheo89 in #319
• fix: add workaround patch for elyra validate by @rkpattnaik780 in #324
• feature: install package directly on runtime images by @harshad16 in #330
• Update runtime image on main based on the commit id: a721fda by @harshad16 in #360
• [Digest Updater Action] Update notebook's imageStreams image tag to digest format by @github-actions in #369

Full Changelog: v1.12.0...v1.13.0

2023b-v1.12.0

Notable Changes

We are excited to announce the release of a major update for our notebooks, labeled as 2023b. This release brings several noteworthy changes and improvements. Here's a summary of what you can expect:

• Package Updates: In this release, all packages across various flavors of the notebooks and runtimes have been updated to their latest versions
• New Packages: Introduced two new packages. Firstly, the inclusion of the codeflare-sdk, is used to interact with the Distributed Workloads stack. Additionally, we've added sklearn-onnx, a package that enables the conversion of scikit-learn models to ONNX.
• Deprecation: The jupyterlab-s3-browser has been deprecated due to its discontinued development and the impediment it poses to updating other packages.
• Standard Databases Clients: The notebooks now come equipped with standard database clients for MySQL, PostgreSQL, MSSQL, and MongoDB.
• Additional Libraries and Applications: We've incorporated several useful libraries and applications. Notably, the inclusion of Mesa-libgl an additional library for OpenCV tasks. We've also introduced Git-lfs, a solution for efficiently handling large files like audio samples, videos, datasets, and graphics. The integration of unixODBC provides a standardized API for accessing data sources, including SQL Servers and other data sources with ODBC drivers. Lastly, the addition of Libsndfile facilitates reading and writing files containing sampled audio data.
• Flavor Deletion: Dropped the ubi8-python3.8 flavor for PyTorch and TensorFlow notebooks, streamlining the available options and ensuring that we focus on the most relevant configurations.
• Security Improvements: To enhance the overall security of our notebooks, we have addressed security issues identified through the Quay security scan.

1. Upgrade workbenches for the 2023b release by @atheo89 in #205
2. Add mesa-libGL on the base notebooks by @atheo89 in #262
3. feature: add codeflare sdk to datascience images by @harshad16 in #263
4. Update imagestreams with new 2023b images by @rkpattnaik780 in #270

All the updated notebooks are available on the Quay.io registry; please filter the results by using the tag "2023b".

Improvements and Bug Fixes

In this release, we've made significant enhancements to our development test framework, which plays a crucial role in ensuring the reliability of our notebooks. Our primary focus was to go beyond mere readiness checks and provide a more comprehensive evaluation of each notebook's functionality and its dependencies.

We've addressed several critical issues and bugs that were affecting the performance and usability of our notebooks. These fixes enhance the overall stability and functionality of our notebooks as listed below.

• Adding piplock-renewal GH action on 2023b branch by @dibryant in #238
• Re-adjust Auto Add Issues to Tracking boards by @harshad16 in #239
• Add tests for TrustyAI Notebooks by @dibryant in #244
• Introduces basic JSON and YAML checks into this repository by @jstourac in #245
• Update the digest updater workflow by incorporating the N-1 version (upstream) by @atheo89 in #249
• Fix the pytorch image reference in main to commit e868e30 by @harshad16 in #250
• [tests] Allow to overwrite repository for testing script download by @jstourac in #258
• Fix typo on tensorflow test notebook name by @atheo89 in #265
• Increase timeout on the makefile test recipe by @atheo89 in #266
• Add name for the notebook digest github action by @jstourac in #268
• Fix and improve CI unit test workaround by @atheo89 in #271
• Remove depreciated bases kustomize feature by @strangiato in #272
• Fix trustyai notebook test by @dibryant in #274

What's Changed

• Upgrade workbenches for the 2023b release by @atheo89 in #205
• Adding piplock-renewal GH action on 2023b branch by @dibryant in #238
• Fix the pytorch image reference in main to commit e868e30 by @harshad16 in #250
• Re-adjust Auto Add Issues to Tracking boards by @harshad16 in #239
• [tests] Allow to overwrite repository for testing script download by @jstourac in #258
• Update the digest updater workflow by incorporating the N-1 version (upstream) by @atheo89 in #249
• Add mesa-libGL on the base notebooks by @atheo89 in #262
• Fix typo on tensorflow test notebook name by @atheo89 in #265
• feature: add codeflare sdk to datascience images by @harshad16 in #263
• Increase timeout on the makefile test recipe by @atheo89 in #266
• Add tests for TrustyAI Notebooks by @dibryant in #244
• Introduces basic JSON and YAML checks into this repository by @jstourac in #245
• Add name for the notebook digest github action by @jstourac in #268
• remove depreciated bases kustomize feature by @strangiato in #272
• Fix and improve CI unit test workaround by @atheo89 in #271
• Fix trustyai notebook test by @dibryant in #274
• Sync main and 2023b branch for feature freeze of sprint 1.35 by @harshad16 in #275
• fix: Upgrade codeflare-sdk with new version for security fix by @harshad16 in #276
• Fine tunings on the main in favor of the new release by @atheo89 in #279
• Update docs in favor of the 2023b release by @atheo89 in #281
• chore: update runtime images with the commit: 34c3405 by @rkpattnaik780 in #286
• updating package versions for notebook test by @dibryant in #284
• Update imagestreams with new 2023b images by @rkpattnaik780 in #270
• Shift the n version to n-1 to fix the digest updater to fetch the correct image by @atheo89 in #289
• [Digest Updater Action] Update notebook's imageStreams image tag to digest format by @github-actions in #290
• fix: missing recommended annotation for TrustyAI by @rkpattnaik780 in #291
• Fix pytorch image reference in main to commit 17c2e49 by @harshad16 in #293
• Patch the imagestream by removing habana 1.11.0 by @harshad16 in #298

New Contributors

• @strangiato made their first contribution in #272

Full Changelog: v1.11.0...v1.12.0

2023a-v1.10.1

Fixed Breaking Change

With v1.10 ODH, the PyTorch image was referencing an incorrect image sha, the detail of the issue can be found here. This is fixed in this patch to point to the right reference.

What's Changed

• Fix the pytorch image reference to commit 88211f5 by @harshad16 in #247

Full Changelog: v1.10.0...v1.10.1

2023a-v1.11.0

New Changes:

• Habana AI based images are now included as OOTB notebook images.
  Users can install the Habana Operator on the Openshift cluster along with dl1.24xlarge machines
  and leverage Habana-based images to do operation directly on Gaudi devices.

What's Changed

• Add documentation that exists on the Wiki page also in the doc folder by @atheo89 in #179
• Add documentation for recommended accelerators annotation by @rkpattnaik780 in #177
• Add test cases for the PyTorch notebooks by @atheo89 in #187
• Add workflow to clean images after PR is merged by @rkpattnaik780 in #193
• docs(readme): fix typo in image names by @rkpattnaik780 in #194
• Fix UBI version used on particular places with regards the Pyton version by @jstourac in #199
• Use RHEL9 repository for UBI9 and CentOs Stream 9 Cuda images by @jstourac in #200
• Include notebook imagestream manifests to notebook repos by @harshad16 in #186
• Add tests for minimal notebooks by @rkpattnaik780 in #190
• Add Jupyter TrustyAI notebook for Python 3.8 by @tteofili in #164
• Add test cases for the TensorFlow notebooks by @atheo89 in #188
• Add tests for datascience notebooks by @rkpattnaik780 in #189
• include missing images from cleanup code by @harshad16 in #216
• Upgrade trustyai notebook images to version 0.3.0 by @tteofili in #209
• Add test case on the PyTorch notebook to convert a pytorch model to onnx format by @atheo89 in #213
• [Digest Updater Action] Update notebook's imageStreams image tag to digest format by @github-actions in #220
• Fix of UBI and Python version used for Jupyter TrustyAI image by @jstourac in #224
• Set $scheme correctly on vscode and rstudio on nginx configuration by @atheo89 in #226
• adding elyra to imagestream by @dibryant in #222
• [Digest Updater Action] Update notebook's imageStreams image tag to digest format by @github-actions in #232
• include habana based notebook image installation by @harshad16 in #214
• Integrate test notebook cases for both ubi8/ubi9 on various flavors of the notebooks by @atheo89 in #229
• Included habana-based imagestream to the notebook manifests by @harshad16 in #233

New Contributors

• @rkpattnaik780 made their first contribution in #177
• @jstourac made their first contribution in #199
• @github-actions made their first contribution in #220

Full Changelog: v1.9.0...v1.11.0

2023a-v1.10.0

What's Changed

• No change in patch versions.

Python Package	old version	patch new version	Change
boto3	1.26.125	1.26.125	No
kafka-python	2.0.2	2.0.2	No
matplotlib	3.6.3	3.6.3	No
numpy	1.24.4	1.24.4	No
pandas	1.5.3	1.5.3	No
scikit-learn	1.2.2	1.2.2	No
scipy	1.10.1	1.10.1	No
torch	1.13.1	1.13.1	No
tensorflow	2.11.1	2.11.1	No

Older release Changelog: https://github.com/opendatahub-io/notebooks/releases/tag/v1.9.0

2023a-v1.9.0

What's Changed

• fix Pipfile.lock exist in container by @orangeji11 in #153
• dropped z stream from python version in notebook by @dibryant in #144
• Add documentation for workbench ImageStreams by @mlassak in #158
• Removed thamos from requirements by @dibryant in #157
• Include github workflow to incldue issue to project by @harshad16 in #169
• Adjust the OWNERS file with new team members by @harshad16 in #168
• Removed the example elyra runtime Test from notebook images by @harshad16 in #165
• Add elyra depedencies for air-gapped environment by @atheo89 in #166
• Upgrade minio version > 7.1.9 to get the fips requirements by @harshad16 in #171
• Fix pytorch and tensorflow runtime workdir by @harshad16 in #173
• Set the airgapped env var required by elyra by @harshad16 in #172

New Contributors

• @orangeji11 made their first contribution in #153
• @dibryant made their first contribution in #144

Full Changelog: v1.8.0...v1.9.0

2023a-v1.8.0

What's Changed

• Update datascience pipeline run URL in Elyra by @mlassak in #137
• Patch pipeline_run to also use pipeline_id for pipeline_version_id by @harshad16 in #142

Full Changelog: v1.6.0...v1.8.0

2023a-v1.7.0

What's Changed

• Python packages patch version updates.

Full Changelog: v1.6.0...v1.7.0

2023a-v1.6.0

What's Changed

• Add Anaconda into the pre-built notebook pipeline by @Xaenalt in #28
• Update: Makefile flag for podman and JDK version in Dockerfile for TrustyAI by @zdtsw in #120
• relicense project to Apache 2.0 by @gregsheremeta in #69
• Derive pytorch back from cuda data science notebook by @atheo89 in #112

New Contributors

• @Xaenalt made their first contribution in #28
• @zdtsw made their first contribution in #120
• @gregsheremeta made their first contribution in #69

Full Changelog: v1.5.1...v1.6.0

2023a-v1.5.1

What's Changed

• Fix the reference of runtime_type in component catalog by @harshad16 in #113
• Include patch option for providing ssl_ca_certs by @harshad16 in #117

Full Changelog: v1.5.0...v1.5.1