A naukri.com initiative
Medium
2M
403
Image Credit: Medium
What’s Stopping You from Building Your Own AI App? Start Now with AWS SageMaker!
- The city wants to build a generative AI app to customize search patterns for relevant information related to the city’s municipal department.
- Deploy a state-of-the-art foundation model by using Amazon SageMaker JumpStart.
- To test that the endpoint works for general questions, submit the question: Which day comes after Friday?
- The data is formatted as a .csv file with two columns: Question and Answer. Pandas is used to create a DataFrame that provides for quick removal of the question column, leaving only the answer column as the data library.
- Multiple built-in functions exist in LangChain to read different file formats, such as .txt, .html, and .pdf.
- Based on the question, identify the top K most relevant documents, where K = 3 in this solution.
- With the Q&A application that is built from the code in this notebook, users can ask a free-form question.
- Using the prompt, which chains the top 3 most relevant documents and question, send the question to the LLM for an answer
- Overall response was not so good somewhat it’s not completed , may be we can change the foundation model like — Flan-T5 Base model Deploy a second SageMaker endpoint, using an ml.g5.2xlarge instance and the Flan-T5 Base foundation model.
- If you’ve found value in this article, you’ll love the content I create as a freelance writer. So please follow me to not miss the coming articles which can transform your learning and will skill up in your career.
Read Full Article
24 Likes
For uninterrupted reading, download the app