New "task type" embedding from the DeepMind team improves RAG search quality

Опубликовано 15 ноября 2024, 20:00

Blog: Improve Gen AI Search with Vertex AI Embeddings and Task Types→ goo.gle/4equIeX
Notebook: Using "task type" embeddings for improving RAG search quality→ goo.gle/4hlRtTO
Document: Choose an embeddings task type → goo.gle/48JASpj

Improve the accuracy and relevance of your RAG systems with new "task type" embeddings developed by the Google DeepMind team. Watch along and learn about common challenges in RAG search quality and how task type embeddings can effectively bridge the semantic gap between questions and answers, leading to more effective retrieval and enhanced RAG performance.

Chapters:
0:00 - Intro
0:19 - Why RAG can't find relevant documents
3:17 - Vertex AI Embeddings API
4:14 - LLM distillation + dual encoder
5:23 - Supported task types
7:33 - Conclusion

Subscribe to Google Cloud Tech → goo.gle/GoogleCloudTech

#GoogleCloud #VertexAI

Speaker: Chrissie Goodrich
Products Mentioned: Vertex AI, Vertex AI Vector Search, Vertex AI Embeddings API