Datasaur LLM support is evolving! 1.1 Update: Streamlining Your LLM Evaluation and Collaboration Experience

The latest update to Datasaur LLM Project 1.1 is packed with features that streamline your evaluation process and take your workflow to the next level. Here's what's new:

1. Cite your sources: We've made it simple to track where the answers are derived from, enhancing the transparency of your work. Pull the relevant chunks from the embeddings and allow annotators to review the source material, whether it is a website or a PDF.

2. Reviewer Mode for Team Collaboration: A popular feature of the broader Datasaur platform, reviewer mode is officially integrated into the LLM project types. Our reviewer mode allows you to identify areas where annotators disagree and resolve the resulting conflicts.

3. LLM Evaluation Report Cards: Curious about how your model is performing? Receive a comprehensive report card after completing your LLM Evaluation project to understand your model performance.

4. LLM Playground Extension: Crafting the perfect answer just got easier with the LLM Playground extension. Test different models, create prompt templates, and enhance your completions using our LLM Playground Extension.

‍5. Prompt templates: Did you find a good template that consistently provides good completions? Save the instructions so the annotator can understand the context behind the generated answers.

The Datasaur LLM Project 1.1 Update simplifies complex tasks, offering you a smoother and more effective experience. From support in information processing to conflict resolution, we've enhanced all aspects to assist you in creating and evaluating the best answers. Give it a try today and take your LLM developments to new heights! If you have any additional feature requests, do reach out let us know as we continue to adapt our roadmap to the latest developments.

‍