From the course: Prompt Engineering with LangChain
Unlock the full course today
Join today to access over 24,400 courses taught by industry experts.
Prompt pipelining
From the course: Prompt Engineering with LangChain
Prompt pipelining
- [Instructor] Prompt Pipelining is a powerful tool that's designed for those who seek modular and efficient approaches to prompt design. There are several use cases where prompt pipelining can be useful. For example, if you want to reuse prompt components like introductions, instructions, examples, et cetera, across multiple prompts. Or if you want to split up a long prompt into smaller, logical chunks. Another way you can use prompt pipelines are when you want to dynamically assemble prompts based on conditional or other logics. Prompt pipelining will allow you to build prompts on the fly from those reusable parts. You can also mix static text with templates that contain variables so that static text can provide structure while the templates can inject dynamic content. You could also compose chat prompts from message templates and static messages. Each piece can get appended as a new message. You can also build up prompts from user provided components for customization. Prompt…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Contents
-
-
-
-
-
-
-
Introduction to prompt templates6m 31s
-
(Locked)
Multi-input prompt templates5m 9s
-
(Locked)
Chat prompt template5m 20s
-
(Locked)
Serializing prompts2m 52s
-
(Locked)
Zero-shot prompts5m 28s
-
(Locked)
Custom prompt templates7m 41s
-
(Locked)
Prompt pipelining5m 49s
-
(Locked)
Chat prompt pipelining3m 21s
-
(Locked)
Prompt composition4m 40s
-
(Locked)
Few-shot prompt templates8m 1s
-
(Locked)
Few-shot prompt templates for chat5m 10s
-
(Locked)
Introduction to example selectors2m 56s
-
(Locked)
Length-based example selector3m 9s
-
(Locked)
Max marginal relevance example selector4m 47s
-
(Locked)
N-gram overlap example selector5m 25s
-
(Locked)
Semantic similarity example selector2m 50s
-
(Locked)
Partial prompt templates4m 19s
-
-
-
-