AWS re:Invent 2023 - Use RAG to improve responses in generative AI applications (AIM336)
ฝัง
- เผยแพร่เมื่อ 3 ธ.ค. 2023
- Your generative AI applications can deliver better responses by incorporating organization-specific data through a technique known as Retrieval Augmented Generation (RAG). However, implementing RAG requires time to configure connections to data sources, manage data ingestion workflows, and write custom code to manage the interactions between the foundation model (FM) and the data sources. Join this session to learn how to make the process much easier using Amazon Bedrock. Based on the user prompt, Amazon Bedrock automatically identifies data sources, retrieves the relevant information, and adds the information to the prompt, thereby giving the FM more information to generate responses. See how it works in this session.
Learn more about AWS re:Invent at go.aws/46iuzGv.
Subscribe:
More AWS videos: bit.ly/2O3zS75
More AWS events videos: bit.ly/316g9t4
ABOUT AWS
Amazon Web Services (AWS) hosts events, both online and in-person, bringing the cloud computing community together to connect, collaborate, and learn from AWS experts.
AWS is the world's most comprehensive and broadly adopted cloud platform, offering over 200 fully featured services from data centers globally. Millions of customers-including the fastest-growing startups, largest enterprises, and leading government agencies-are using AWS to lower costs, become more agile, and innovate faster.
#AWSreInvent #AWSreInvent2023 - วิทยาศาสตร์และเทคโนโลยี
Most of the Amazon Bedrock presentations were not very well done, but this one is pretty easy to understand. Thanks for speaking clearly and knowing the topic you're talking about.
One of the best videos Ive seen which covered every aspect of building GenAI applications with RAG
Thank you Mani and Ruhaab for an excellent overview, example use case and links to the sample code. I appreciate it a lot.
The best talk of RAG. easy to expain why they build KM for Bedrock.
I agree with the other commenters that this video is exceptional. You two are very good presenters. This was just enough depth covering the right surface area of these products and features. I absolutely love the Amazon KB for my use case, and I love how much of the process Amazon manages for me, which allows me to spend less time developing and more time selling.
Thank you so much for your kind words! 😊
So will presented. Great job!
Amazing talk!! Learned so much!! ❤❤
excellent presentation
Thank you for your feedback. 🙌
Excellently explained. Thanks for the insightful presentation.
It's our pleasure! 😀
Please post GitHub repository link as mentioned in the talk. @Mani & @Ruhaab.
How to improve latency issue in response in this RAG model approach using aws bedrock knowledge based
Evnen though i created small pdf file having 10pages its giving response in 5 to 7 seconds
I want with in 1 second in response what i do ?
Please help...
Hi there. 👋 Our scope for tech support is limited on this platform. You can get some assistance from our community of experts on re:Post: go.aws/aws-repost. 🤓 If you still need help, check out these options: go.aws/get-help. 🤝 ^RW
don't see myself making a new LLAMA for the #4th option in the beginning
I got as far as 43:00 when trying this out - knowledge base created, synched and status = ready. Go to test it though and there are no models available to me. I able to 'retrieve' so I know the data sync and embeddings were successful, but I cannot use that to generate anything from an FM. Amazon Q "can't answer my question". Stuck
Sorry to see the trouble! This doc will help clarify more context for using RAG with Amazon Q: go.aws/3TsFWbw. For further support on technical questions, I'd also recommend engaging our community of experts on re:Post: go.aws/aws-repost. ⬅️ ^AD
this is great but it is still cumbersome and far from production ready, it would be cool if an endpoint would be generated automatically to call it from an app directly without having to build another customer langchain app on top of it
Using the retrieveAndGenerate API i am unable to get the cited references.
Oh no! Sorry to hear about this trouble. This would be a great question to post over at re:Post where our community of experts can chime in & share their knowledge: go.aws/aws-repost. 🤝 ^AK
@@awssupport Posted the question but not getting any response
Hello! Please understand that these posts are answered in the order they're received. It can take time before our collective of engineers reach out. In the meantime you may find this doc helpful: go.aws/3y5mVnj. 📝 ^AR