Đừng Để Mất Ngữ Cảnh! Cách Chia Nhóm Trễ Có Thể Nâng Cao Hệ Thống Truy Xuất Của Bạn

Tác giả: Prompt Engineering
Ngày xuất bản: 2024-10-11T00:00:00
Length: 16:49

In this video, I explore the powerful technique of late chunking in long context embedding models. By preserving contextual information across entire documents before chunking, this method offers more precise retrieval while minimizing storage needs. Join me as I compare it to other methods like Anthropic's contextual retrieval and provide insights for implementing it in your applications.

LINKS:

Colab: https://tinyurl.com/ynxjjyu8

https://jina.ai/news/late-chunking-in-long-context-embedding-models/

https://jina.ai/news/what-late-chunking-really-is-and-what-its-not-part-ii/

https://weaviate.io/blog/late-chunking

https://arxiv.org/pdf/2409.04701

https://www.anthropic.com/news/contextual-retrieval

https://github.com/jina-ai/late-chunking?tab=readme-ov-file

https://youtu.be/6efwN_US-zk

https://youtu.be/tmiBae2goJM

💻 RAG Beyond Basics Course:

https://prompt-s-site.thinkific.com/courses/rag

Let's Connect:

🦾 Discord: https://discord.com/invite/t4eYQRUcXB

☕ Buy me a Coffee: https://ko-fi.com/promptengineering

|🔴 Patreon: https://www.patreon.com/PromptEngineering

💼Consulting: https://calendly.com/engineerprompt/consulting-call

📧 Business Contact: [email protected]

Become Member: http://tinyurl.com/y5h28s6h

💻 Pre-configured localGPT VM: https://bit.ly/localGPT (use Code: PromptEngineering for 50% off).

Signup for Newsletter, localgpt:

https://tally.so/r/3y9bb0

00:00 Introduction to Contextual Retrieval and Late Chunking

00:28 Understanding Embedding Models and Their Parameters

00:55 Challenges with Standard RAG Pipelines

02:08 Late Chunking Explained

03:06 Implementation and Benefits of Late Chunking

05:21 Comparing Late Chunking with Other Techniques

10:51 Practical Implementation Guide

16:09 Conclusion and Further Resources

All Interesting Videos:

Everything LangChain: https://www.youtube.com/playlist?list=PLVEEucA9MYhOu89CX8H3MBZqayTbcCTMr

Everything LLM: https://youtube.com/playlist?list=PLVEEucA9MYhNF5-zeb4Iw2Nl1OKTH-Txw

Everything Midjourney: https://youtube.com/playlist?list=PLVEEucA9MYhMdrdHZtFeEebl20LPkaSmw

AI Image Generation: https://youtube.com/playlist?list=PLVEEucA9MYhPVgYazU5hx6emMXtargd4z

Dịch Vào Lúc: 2025-03-10T05:58:11Z

Yêu cầu dịch (Một bản dịch khoảng 5 phút)

Phiên bản 3 (ổn định)

Tối ưu hóa cho một người nói. Phù hợp cho video chia sẻ kiến thức hoặc giảng dạy.

Video Đề Xuất