Episode
Building NLP solutions with GPT-powered Azure AI Language
with Korey Stegared-Pace, Sean Murray, Yanling Xiong
In this demo session, you will learn about what is in Azure AI Language and see how task-oriented and -optimized Language models can help you effectively process and understand documents and meetings, with features such as summarization, translation, PII redaction, sentiment analysis, and named entity recognition, etc. You will see two examples of how these skills can be applied to different scenarios.
Chapters
- 00:00 - Intro
- 05:09 - Overview of Azure AI Language service
- 10:25 - What is chunking, and what happens at 125,000 characters?
- 17:25 - Chunking strategy
- 19:58 - Azure AI Language tool demo
- 26:14 - Is there a file size limit when reading the file?
- 29:01 - What is prompt flow?
- 32:30 - Specify the language of the input
- 38:33 - Query field
- 43:31 - Are these premade code blocks? Can we modify the code?
- 48:06 - Summary aspect
- 50:48 - Any suggestions for recognizing or handling potential hallucinations?
In this demo session, you will learn about what is in Azure AI Language and see how task-oriented and -optimized Language models can help you effectively process and understand documents and meetings, with features such as summarization, translation, PII redaction, sentiment analysis, and named entity recognition, etc. You will see two examples of how these skills can be applied to different scenarios.
Chapters
- 00:00 - Intro
- 05:09 - Overview of Azure AI Language service
- 10:25 - What is chunking, and what happens at 125,000 characters?
- 17:25 - Chunking strategy
- 19:58 - Azure AI Language tool demo
- 26:14 - Is there a file size limit when reading the file?
- 29:01 - What is prompt flow?
- 32:30 - Specify the language of the input
- 38:33 - Query field
- 43:31 - Are these premade code blocks? Can we modify the code?
- 48:06 - Summary aspect
- 50:48 - Any suggestions for recognizing or handling potential hallucinations?
Have feedback? Submit an issue here.