Release Notes

Release 24.12 Notes

By Rohan Sharma • Aug 29, 2024

  • Surfacing Error Messages for Failed Jobs

    • Error messages for failed jobs are now visible directly on the UI job status page, simplifying the process of diagnosing and resolving issues.

  • User Selected Default Dashboards

    • Any dashboard within a project can now be assigned as the default dashboard for a model, with all insights leading directly to the assigned default dashboard.

  • Custom Feature Impact Feature Release Notes

    • Introducing Custom Feature Impact: Upload custom feature impact scores for your models, leveraging domain-specific knowledge or external data without requiring the corresponding model artifact.

    • Easy data upload via API endpoint with required parameters: Model UUID, Feature Names, and Impact Scores.

    • View updated feature impact scores in:

      • Model details page

      • Charts page

      • Explain page

    • Flexible update options: Update existing feature impact data by uploading new data for the same model and Seamless integration with existing model artifacts.

Flexible Model Deployment

The python-38 base image is no longer supported.

Release 24.11 Notes

By Sabina Cartacio • Aug 6, 2024

Client Version

Client version 3.3+ is required for the updates and features mentioned in this release.

What's New and Improved:

  • Performance Analytics (Preview) Embedded in Monitoring Charts

    • Visualize performance analytics charts as part of the root cause analysis flow for Binary Classification, Multiclass Classification, and Regression models, spanning from confusion matrices, precision recall charts, prediction scatterplots and more.

Release 24.10 Notes

By Sabina Cartacio • July 23, 2024

Client Version

Client version 3.3+ is required for the updates and features mentioned in this release.

What's New and Improved:

  • Support for applied segments in monitoring charts

    • Create and apply segments dynamically in monitoring charts for exploratory analysis without requiring them to be saved to the model.

  • User-Defined Feature Impact

    • The User-Defined Feature Impact enables you to upload custom feature impact for models. This feature addresses several issues reported by our customers, including model artifact size, onboarding complexity, and the need for custom feature impact.

    • Key highlights

      • New method: UploadFeatureImpact

      • Improved Fiddler UI to display uploaded feature impact

Release 24.9 Note

By Rohan Sharma • July 12, 2024

What's New and Improved

  • Enhanced access controls

    • Control access with precision: Manage user access to resources with Role-Based Access Control (RBAC), ensuring the right users have the right permissions.

    • Simplify user management: Assign roles to users and teams to streamline access control and enhance collaboration. *≠ Protect sensitive resources: Restrict access to sensitive resources, such as models and project settings, with granular permissions.

    • Work efficiently: Focus on your work without worrying about unauthorized access or data breaches.

Release 24.8 Notes

By Sabina Cartacio • June 18, 2024

Release of Fiddler Platform Version 24.8:

  • Performance Analytics Charts (Public Preview)

    • Visualize charts to aid in analyzing model performance for Binary Classification, Multiclass Classification, and Regression models.

    • Leverage applied segments in Performance Analytics charts to explore problematic cohorts of data.

Release 24.7 Notes

By Parth Domadia • June 4, 2024

What's New and Improved

TBD

Release 24.6 Notes

By Parth Domadia • May 22, 2024

Release of Fiddler Platform Version 24.6:

  • Performance improvements

    • Improved the performance of various modules / APIs.

    • Improved observability which can help monitor health and performance of the operations.

Client Version

  • Client version 3.1.2+ is required for the updates and features mentioned in this release.

Release 24.5 Notes

By Sabina Cartacio • May 2, 2024

Release of Fiddler Platform Version 24.5:

  • Support for model versions for streamlined model management

What's New and Improved:

  • Model Versions

    • Efficiently manage related models by creating structured versions, facilitating tasks like retraining and comparison analyses.

    • Users can maintain model lineage, efficiently manage updates, flexibly modify schemas, and adjust parameters.

    • See: Model Versions for more information.

  • Airgapped Enrichments (alpha)

    • For privacy sensitive use cases, all data getting enriched stays within customer premises.

  • New Deployment Base Images

    • We have added new deployment base images to support model versioning.

    • See here.

Client Version

Client version 3.1.0+ is required for the updates and features mentioned in this release.

Release 24.4 Notes

By Parth Domadia • April 26, 2024

Release of Fiddler Platform Version 24.4:

  • UMAP UI changes

  • SSO integration changes

  • New concept: Environments

  • Fundamental changes to product concepts

What's New and Improved:

  • UMAP UI

    • Vertical scrolling instead of horizontal scrolling for data cards

    • "View More" option to open data cards in maximized modal

    • Ability to toggle between data cards in the maximized modal

  • SSO integration changes

    • Fiddler now integrates with Azure AD SSO, allowing you to leverage existing user roles for access control within Fiddler. This eliminates the need for manual user creation and simplifies user management within your organization. See Azure AD SSO Support for details

  • Environments

    • Each Model now has two Envrionments (Pre-Production and Production) used to house data in different ways.

    • A Model's Pre-Production Environment is used to house non-time series data (Datasets).

    • A Model's Production Environment is used to house time series data.

  • Product concept changes

    • Datasets are no longer stored at the Project level. Instead, they're stored at the Model level under the Pre-Production Environment.

    • The Model Details page has been updated with a new design. See more here.

Client Version

Client version 3.0+ is required for the updates and features mentioned in this release.

Client 3.x Release:

We are launching Client 3.x, this is revamped client 2.x as we move to more object oriented based methods. This means, any pipeline setup in client 2.x would eventually be required to upgrade to the new methods. Client 2.x will sunset approximately 6 months post this release. Please take a look at the below resources to help you understand client 3.x and also how you can upgrade your pipelines:

Deprecations and Removals:

  • All IDs will be UUIDs instead of strings.

  • Dataset deletion is not allowed anymore.

For API level changes and updates please check client history

Release 24.3 Notes

By Sabina Cartacio • March 19, 2024

Release of Fiddler Platform Version 24.3:

  • Ability to update metadata column values

  • Percentage thresholds on incoming traffic for data integrity alerts

  • Additions to the enrichment framework

What's New and Improved:

  • Ability to update metadata columns

    • Support updating metadata columns with new values.

      • You can observe the changes in Analytics but the Performance metrics of the columns won't be reflected.

    • See Updating Events for more information.

  • Percentage thresholds on incoming traffic for data integrity alerts

    • Added an alert category that quantifies data integrity violations as a percentage-based threshold relative to the total traffic within the time bin. details here Data Integrity Alert Example API usage

  • New enrichments for LLM monitoring (beta)

    • New Enrichments:

      • Profanity

      • Answer Relevance : LLM based metric, requires OPENAI_API_KEY

      • Faithfulness : LLM based metric, requires OPENAI_API_KEY

      • Coherence : LLM based metric, requires OPENAI_API_KEY

      • Conciseness : LLM based metric, requires OPENAI_API_KEY

      • Toxicity

      • Topic

      • RegexMatch

      • KeywordDetector

      • BannedKeywords

    • See Fiddler Enrichments for more information.

Client Version

Client version 2.5+ is required for the updates and features mentioned in this release.

Release 24.2 Notes

By Sabina Cartacio • January 30, 2024

Release of Fiddler Platform Version 24.2:

  • Automatically generate a rich set of LLM enrichments

  • Investigate an alerts' revision history

What's New and Improved:

  • Alert Revision History

    • Easily view the revision history of alerts within the monitoring charts UI, including informations on the alerts severity, thresholds, and more.

    • See: Triggered Alert Revisions for more information.

  • LLM Enrichments & Embeddings (beta)

    • Leverage the Enrichments framework to enhance your dataset by adding columns to reflect:

      • Automatic generation of embeddings

      • Generation of distance of an embedding from the cluster centroid

      • PII (Personal Identifiable Information)

      • Sentiment

      • Various text statistics

    • See LLM Monitoring for more information.

Client Version

Client version 2.5+ is required for the updates and features mentioned in this release.

Release 24.1 Notes

By Sabina Cartacio • December 29, 2023

Release of Fiddler Platform Version 24.1:

  • Support for segments on monitoring charts and alerting

  • Support for Azure Active Directory

  • Support for default dashboards for reporting

What's New and Improved:

  • New Support for Segments

    • Provides the ability to filter data in monitoring charts using Fiddler Query Language and set alerts on the filtered data.

    • See Segments for more information.

  • Azure AD Support

    • Fiddler now supports integration with Microsoft Azure Active Directory for user authorization

    • See Single sign on with Azure for more information

  • Tailored Insights via Default Dashboards

    • Users can now explore tailored model monitoring dashboards under Insights on the homepage, featuring charts on Performance, Traffic, Data Drift, and Data Integrity.

    • With the introduction of Default Dashboards, past monitoring charts have been deprecated.

    • See Default Dashboards for more information.

Client Version

Client version 2.4+ is required for the updates and features mentioned in this release.

Release 23.7 Notes

By Sabina Cartacio • December 11, 2023

This page enumerates the new features and updates in Release 23.7 of Fiddler.

Release of Fiddler Platform Version 23.7:

  • Support for multi-target models

  • Support for alert rule creation on the statistics frequency metric

What's New and Improved:

  • Custom Target Support in Custom Metrics

    • Custom Metrics now supports the ability to specify which column to use as the target for all built-in performance metrics

    • Custom Metrics now support functions for identifying True Positives, True Negatives, False Positives, and False Negatives for both binary classification and multi-class classification model tasks.

    • See Custom Metrics for more information

  • Alert Rules on Frequency

    • Support for alert rules on Frequency statistics metrics

    • See Statistics Guide for more information

Client Version

Client version 2.3+ is required for the updates and features mentioned in this release. This page enumerates the new features and updates in Release 23.6 of the Fiddler platform.

Release 23.6 Notes

By Sabina Cartacio • November 21, 2023

Release of Fiddler Platform Version 23.6:

  • Embedding visualization with UMAP is now available for all customers

  • Support for user-defined metrics via custom metrics

  • Improved alert messaging for Webhook notifications

  • Improved standalone job status page

What's New and Improved:

  • Custom Metrics

  • Improved Webhook Messaging

    • Refining "Why Fiddler Generated" alerts for Webhook with enhanced details

    • Learn how to enable webhook notification.

  • **Improved Job Status **

    • Access all your models, datasets, and events jobs through the Jobs page in the navigation bar.

    • Learn more on the Product Tour

  • Embedding Visualization with UMAP

    • Visualize your custom features with high dimensional data using UMAP

Client Version

Client version 2.2+ is required for the updates and features mentioned in this release.

Release 23.5 Notes

By Sabina Cartacio • October 19, 2023

This page enumerates the new features and updates in Release 23.5 of the Fiddler platform.

Release of Fiddler Platform Version 23.5:

  • Support for new Vector data type

  • Support for new Frequency statistics metric

  • Support for new NOT_SET task type

  • New standalone bookmarks page

  • Contact your customer success team to get access and documentation to the below:

    • Addition of LLM task

    • Auto embeddings on Text (Prompt, Response, etc) and Image data

    • Auto enrichments of Toxicity, PII, and Hallucination on LLM task

    • 3D UMAP visualization of high-dimensional embeddings

    • Support for user-defined metrics -> Custom metrics. Alerts can be set on these custom metrics and plotted in charts and dashboards.

What's New and Improved:

  • New Vector input type

    • You can now create custom features from Vector data type

    • The DI violation functionality on vector data types is now available to quickly detect any issues in your data pipelines. As vector and embedding pipelines are complex and prone to errors, we hope this functionality will be of value.

      • Violation of value - Dimension does not match the expected dimension

      • Violation of nullable - Vector length equals zero or received NULL value

      • Violation of type - Value is not of the type of VECTOR or VECTOR contains a non-numerical type

    • Clusters of Text Embeddings have a short summary to understand which of your clusters have the most problematic drift.

    • Events of a particular cluster can be queried in the 'Analyze' tab of the model. This can be used to pinpoint the most problematic prompts/responses in an LLM scenario or images in a CV scenario. This information can be used to improve your models.

  • New Task Type for Task-less Models

    • You can now add models without a task by choosing the NOT_SET task type when onboarding your model.

    • Note that task-less models have no restrictions on output and target columns, but performance metrics are disabled when a task is not specified.

  • Standalone Bookmarks Page

    • Access all your bookmarked projects, models, datasets, charts, and dashboards through the Bookmarks page in the navigation bar.

    • Learn more on the Product Tour

Client Version

Client version 2.1+ is required for the updates and features mentioned in this release.

Last updated

© 2024 Fiddler AI