Fine-tuning Models with Weights and Biases Weave for Better Performance

Watch a 10-minute video demonstrating how to leverage W&B Weave and Models for optimizing AI applications. Explore the Weights & Biases AI developer platform's capabilities, starting with an introduction to platform optimization techniques. Follow along with the creation of a RAG-enabled retail support chatbot using W&B Weave, understand the implementation of Weave Scorers and Evaluations, and learn the process of fine-tuning Large Language Models. Discover how to publish models to the W&B Registry and evaluate the performance of fine-tuned LLMs through comparative analysis. The demonstration concludes with a comprehensive overview of the W&B Registry and practical insights into model performance evaluation.

Syllabus

0:00 Introduction to the Weights & Biases AI developer platform
0:13 Optimizing AI application with Weights & Biases
1:00 Weights & Biases AI developer platform overview
1:37 Building a RAG-enabled retail support chatbot using W&B Weave
3:46 W&B Weave Scorers and Evaluations
6:00 Fine-tuning an LLM using W&B Models
7:01 Publishing a model to W&B Registry
7:38 A quick overview of W&B Registry
8:37 Evaluating the performance of our fine-tuned LLM and comparing it to other models
9:37 Conclusion and invitation to try the Weights & Biases AI developer platform