DSPy: The framework for programming—not prompting—language models
Python
Update index.md
Omar Khattabf5e25378513c
diff --git a/docs/docs/index.md b/docs/docs/index.mdindex 1d0e82f78..1bc073bb8 100644--- a/docs/docs/index.md+++ b/docs/docs/index.md@@ -12,9 +12,9 @@ hide:[](https://pepy.tech/projects/dspy)-DSPy is a declarative framework for building AI software. It allows you to **iterate fast on modular _code_**, not brittle strings, and offers algorithms that **compile high-level modules into effective prompts and model weights**, whether you're building simple classifiers, sophisticated RAG pipelines, or Agent loops.+DSPy is a declarative framework for building AI software. It allows you to **iterate fast on modular code**, not brittle strings, and offers algorithms that **compile modular AI programs into effective prompts and model weights**, whether you're building simple classifiers, sophisticated RAG pipelines, or Agent loops.-DSPy stands for Declarative Self-improving Python. Conventionally, the most important part of your AI system is a bunch of long string prompts or a sequence of hand-stitched training jobs. Instead, DSPy lets you define your AI system by declaring typed Python functions, as natural-language `Signatures`, which you can then programmatically compile into different implementations for different models, inference strategies, or learning algorithms.+DSPy stands for Declarative Self-improving Python. Conventionally, the most important part of your ML system is a bunch of sequence of hand-stitched training jobs or long string prompts. Instead, DSPy lets you define AI software by declaring natural-language modules and programmatically compiling them into different implementations for different models, inference strategies, or learning algorithms.This higher-level abstraction makes AI software cleaner to build iteratively, easier to maintain, and more portable as new models and optimization techniques emerge. This [lecture](https://www.youtube.com/watch?v=JEMYuzrKLUw) is a good conceptual introduction. Meet the community, seek help, or start contributing via our [GitHub repo](https://github.com/stanfordnlp/dspy) and [Discord server](https://discord.gg/XCGy2WDCQB).