# 6.91.0

December 05, 2024

## **What's new** 💡

### **Data Studio**

* New Data Programming Labeling Function template for row based.
* Question set dialog in project — Enhanced the appearance of the question set dialog for configuring new or existing question sets within a project.
* Search Extension for Span Labeling— We’ve added a new option to search called [Advanced Search](https://github.com/vulcan-ai/gitbook/blob/datasaur/compatibility-and-updates/advanced/extensions/search.md#advanced-search) allowing the combination of multiple conditions to refine results.

### **LLM Labs**

* Deployment — Allowed non-admin users to create API keys.
* Deployment — Enhanced error handling for deployed API and programmatic API to adhere to OpenAI client standards.
* Evaluation — Improved our manual evaluation to not be case-insensitive for the dataset column headers.
* Fine-tuning — Added the Llama 3.1 model to our fine-tuning feature.
* Fine-tuning — Improved our fine-tuning feature to support multi-region capabilities.
* Models — Improved our OpenAI library to use the latest stable version.

## **Bug fixes** 🐞

### **Data Studio**

* Custom Report Builder — Fixed an issue where the progress table in CRB displayed inconsistent counts for "Total Documents Answered," "Total Rows Answered," and "Total Sentences Labeled" when a user held both labeler and reviewer roles.
* Fixed an issue in Bounding Box Labeling where rotating a document caused the view to shift to a different page.

### **LLM Labs**

* Fixed an issue where manual evaluation ranking creation failed when using saved application.