Release NotesPDF version

What's New

Major features and updates for the Cloudera AI data service.

Release notes and fixed issues for version 2.0.47-b345.

Cloudera AI Workbench
  • Support is now provided for API keys to invoke applications deployed using Cloudera AI Workbenches. This not only eases the invocation of those applications programmatically but also allows one application to easily invoke another application that they have access to.
  • MLFLOW upgrade for Cloudera AI Workbenches now enables making use of the latest offerings and APIs from the MLFLOW community like evaluateLLM.
Cloudera AI Platform
  • The autoscaling range of Suspend Workflow is now set to the value 0 to ensure that other Kubernetes deployments outside the scope of MLX can deploy their pods on worker nodes.
Cloudera AI Registry
  • An enhanced error message is now displayed during model upload failure.
  • UI for Registered Models displays the environment name of the registry along with an error message when any user is unable to access any Cloudera AI Registry.
  • A checkbox is now added to enable Public Load Balancer for new Cloudera AI Registries on Azure.
Cloudera AI Inference service
  • The Hugging Face model server backend has been upgraded, which expands the compatibility with a larger number of model families, such as Llama 3.3 and models derived from it.
  • Llama 3.2 Vision Language Model NIM version has been updated to address compatibility with A10G (g5.*) and L40S (g6e.*) GPU instances on AWS.
  • You can now upgrade Cloudera AI Inference service using the UI. Previously, the upgrade was supported only using CDPCLI.
  • You can now upgrade from Cloudera AI Inference service version 1.2.0-b80 to version 1.3.0-b113 or higher. Note that you cannot upgrade from 1.3.0-b111 to 1.3.0-b113 or higher. For more information on the 1.3.0-b111 upgrade issue and workaround, see the Known Issues section.
Cloudera AI Workbench
  • Previously, due to an issue, users could stop sessions under projects that they were not authorized to access using the session’s UUID. This issue is now resolved. (DSE-39798)
  • Previously, when a Kubernetes object was deleted, and the reconciler was overwhelmed by a large number of events, the Deleted status failed to propagate properly. This issue is now resolved. (DSE-41431)
  • Previously, the stopped_at column was not correctly populated when applications were stopped. This issue is now resolved. (DSE-41636)
  • Previously, engine pods were stuck in the Init:StartError state and you had to manually delete it. With this fix, pods stuck in Init:StartError in the Garbage Collection will be deleted after a certain grace period. (DSE-41430)
  • Previously, Spark environment configurations were not inherited by models running Spark. With this fix, models use the appropriate Spark configurations to run Spark. (DSE-36343)
Cloudera AI Registry
  • An issue around how Hugging Face token was being consumed during the import of a model was addressed. (DSE-41714)
  • The Cloudera AI Registry deletion flow is improved to take care of race conditions when both creation and deletion are triggered in a short frame of time. (DSE-41634)
Cloudera AI Inference service
  • Previously, the GetEndpointLogs failed with an error. With this fix, endpoint logs for the model container do not exceed the gRPC messaging size. (DSE-41765)
  • A new field called loadBalancerIPWhitelists is added to display a list of IPs whitelisted for the load balancer and deprecated isPublic and ipAllowlist. (DSE-39397)
  • Infrastructure nodes are no longer shown as instances that can be used for deploying a new endpoint. (DSE-41726)
ML Runtimes
  • Previously, due to an issue, to ensure the compatibility of AMPs with ML Runtimes 2025.01.1, users had to switch to JupyterLab PBJ Workbench in the AMPs’ .project-metadata.yaml file or use jobs instead of sessions for automated tasks. This issue is now resolved. (DSE-41263)
  • Resolved issues related to using R interactively in PBJ Runtimes. (DSE-41771)

Release notes and fixed issues for version 2.0.47-b302.

  • Migrated Cloudera AI Workbench, Cloudera AI Registry, and Cloudera AI Inference service images to chainguard to address CVEs.
  • Added APIv2 support for Enhanced Group Sync.
  • Added support to create AMPs (Cloudera Accelerators for Machine Learning Projects) using APIv2. Previously, this option was available only using UI.
  • Added support for H100 GPU instances for Cloudera AI Inference service on Azure.
  • Added support for AKS workload identity.
  • Added support for AWS M7a, M7i, C7a, C7i, R7a, R7i instance families.
  • Added support for Cloudera AI Inference service on EU Control Plane.
  • Added support for EKS 1.30.
  • Added support for AKS 1.30.
  • Hugging Face support (Technical Preview): You can now import text-generating language models from Hugging Face and deploy them on Cloudera AI Inference service.
  • Added profiles for HuggingFace Models and multi-modal models in the Model Hub catalog.
  • Updated existing model manifests in the catalog after upgrading the NIM version in Cloudera AI Inference service.
  • Enhanced error messages related to model import failure in the Model Hub UI.
  • Carried enhancements in AI Registry to ensure that multi-modals can be supported.
  • Added runtime support for Llama 3.2 11B and 90B Vision Language Model NIMs to ensure that they can be deployed using AI Inference. Only model profiles optimized for the H100 GPU are supported for these two models in this release.
  • Llama 3 NIM is no longer supported since we now have both Llama 3.1 and Llama 3.2.
  • Added support for Diagnostic Bundles in Cloudera AI Inference service.
  • Upgraded text-generating and embedding NIMs.
  • Added Code Sample functionality for endpoints deployed using Cloudera AI Inference service.
  • Model endpoint replica events can now be viewed on the Model Endpoint details UI.You can now add numerous docker credentials using UI or API which can be used to enable Cloudera AI to fetch custom ML Runtimes from a secure repository. For more information, see Add Docker registry credentials and certificates.
  • Previously, some Cloudera AI Inference service clusters did not have the 'creationDate' field. This field is now added.(DSE-38817)
  • Previously, the deletion of backup for older workspaces was failing. This issue is now resolved. (DSE-41031)
  • Previously, deleting a workbench backup created by a deleted user displayed an error. This issue is now resolved. (DSE-41052)
  • Multiple UI improvements are made both in the Create, Read, Update, and Delete operations of Cloudera AI Inference service and while deploying or editing a model endpoint.
  • The model_name field is now displayed instead of model_id in the Endpoint Details UI. (DSE-38937)
  • Previously, the NIM model profile environment variable was only assigned for LLMs. Now support for Model Profile override is added for Embedding and Reranker NIMs. (DSE-40508)
  • Previously, there was an issue with rendering of existing instance type in the "Edit Endpoint" UI. This issue is now resolved. (DSE-40636)
  • Validated all node group (instance type) selection from UI. (DSE-40754)
  • Previously, NGC manifest components were missing from the download. This issue is now resolved. (DSE-41055)
  • The Create ML Serving application now enables the public load balancer. (DSE-41305)
  • The Instance Type field in the Edit Model Endpoint UI is no longer mandatory. (DSE-41278)
  • Added force delete option to delete the Cloudera AI Inference service using UI. (DSE-41035)
  • The Cloudera AI Inference service UI now displays optimization profile details. (DSE-40927)
  • You can now create, download, and delete log archives for Cloudera AI Inference service. (DSE-40921)
  • The Test Model UI now fails gracefully when the replica is scaled down to zero for a model deployed using Cloudera AI Inference service. (DSE-40957)
  • Previously, the Storage initializer had the wrong task values. This issue is now resolved. (DSE-41058)
  • Enabled storage initializer to now handle more than two directories for NIM artifacts. (DSE-40986)
  • Removed Llama 3 runtimes. (DSE-40956)
  • Addressed SQL injection issue in AI Registry that allowed non-authorized but authenticated users to perform Create, Read, Update, and Delete operations on AI Registry’s metadata tables. (DSE-41542)

Release notes and fixed issues for version 2.0.46-b238.

  • Model Hub Enhancement: The model size is now shown in the user-friendly format both in the Model Hub UI and Cloudera AI Registry UI.
  • Cloudera AI Inference service Enhancement: New AI Inference Services menu item is added to the left-navigation pane of the Cloudera AI UI to manage the lifecycle of Cloudera AI Inference service using UI. For more information, see Using Cloudera AI Inference service.
  • Added Spark 3.5 ML Runtime Addon
  • Product and features named:
    • Clouder Machine Learning (CML) is renamed to Cloudera AI.
    • Cloudera Machine Learning Model Registry is renamed to Cloudera AI Registry.
    • Cloudera Machine Learning Workspace is renamed to Cloudera AI Workbench.
    • Cloudera Applied Machine Learning Prototypes and Accelerators for ML Projects is renamed to Cloudera Accelerators for Machine Learning Projects.
  • CVE fixes - This release includes numerous security fixes for critical and high Common Vulnerability and Exposures (CVE).
  • Previously, the public and private settings did not carry forward after the AI Registry upgrade. This issue is now resolved. (DSE-36799)
  • Enhanced the error message that was displayed when importing a model from Model Hub to Registered Models. (DSE-39897)
  • Generic (vLLM) NIM profile deployment was returning an empty GPU list in the UI. This issue is now resolved. (DSE-39913)
  • Previously, public cloud CDP CLI was not showing the instance type's GPU count. This issue is now resolved. (DSE-39539)
  • Cloudera AI v2 API deployed application did not inherit user-level environment variables and site-level environment variables. This issue has been solved, and now an application created using APIv2 does not only inherit project-level environment variables but also user-level environment variables and site-level environment variables. (DSE-37611)
  • Previously, scheduled jobs skipped job runs and did not specify the error. Now, the skipped jobs runs have improved exit code to distinguish them from failed jobs. (DSE-39976)
  • Previously, the Next buttons on the Site Administration page did not work. This issue is now resolved. (DSE-34133).
  • Previously, restarting the application using the Cloudera AI v2 API did not inherit account application-level environment variables. This issue is now resolved. (DSE-39894)
  • Users can now view the existing applications in the Cloudera AI UI even if the creation of a new application is disabled. (DSE-39980)
  • Previously, Python logging did not work with PBJ Runtimes. This issue is now resolved. (DSE-39929)
  • Previously, reloading the session page would result in an incorrect state where the PBJ session's editor cell could appear green even if it is in a processing state (executing some commands). With this fix, an accurate representation of the processing state is displayed even after a refresh. (DSE-40049)