DSPy: The framework for programming—not prompting—language models
Python
Update README.md
Omar Khattab347869ea0b4f
diff --git a/README.md b/README.mdindex e81ec3111..0ef11dcff 100644--- a/README.md+++ b/README.md@@ -25,7 +25,7 @@ To this end, **DSP** offers a number of powerful _primitives_ for building archiOur [intro notebook](intro.ipynb) provides examples of five "multi-hop" question answering programs of increasing complexity written in DSP.-You can [open the intro notebook in Google Colab](https://colab.research.google.com/github/stanfordnlp/dsp/blob/main/intro.ipynb). You don't even need an API key to get started with it.+You can **[open the intro notebook in Google Colab](https://colab.research.google.com/github/stanfordnlp/dsp/blob/main/intro.ipynb)**. You don't even need an API key to get started with it.Once you go through the notebook, you'll be ready to create your own DSP pipelines!@@ -41,7 +41,7 @@ Once you go through the notebook, you'll be ready to create your own DSP pipelinOur [compiler notebook](compiler.ipynb) introduces the new experimental compiler, which can optimize DSP programs automatically for (much) cheaper execution.-You can [open the compiler notebook in Google Colab](https://colab.research.google.com/github/stanfordnlp/dsp/blob/main/compiler.ipynb). You don't even need an API key to get started with it.+You can **[open the compiler notebook in Google Colab](https://colab.research.google.com/github/stanfordnlp/dsp/blob/main/compiler.ipynb)**. You don't even need an API key to get started with it.## Picking in-context examples using KNN/ANN methods [NEW!]