DSPy: The framework for programming—not prompting—language models
Python
Updated docs to MIPRO and COPRO
Michael Ryanc19ce5099b91
diff --git a/docs/docs/building-blocks/6-optimizers.md b/docs/docs/building-blocks/6-optimizers.mdindex 52061a215..55698140f 100644--- a/docs/docs/building-blocks/6-optimizers.md+++ b/docs/docs/building-blocks/6-optimizers.md@@ -47,9 +47,9 @@ All of these can be accessed via `from dspy.teleprompt import *`.#### Automatic Instruction Optimization-4. **`SignatureOptimizer`**: Generates and refines new instructions for each step, and optimizes them with coordinate ascent.+4. **`COPRO`**: Generates and refines new instructions for each step, and optimizes them with coordinate ascent.-5. **`BayesianSignatureOptimizer`**: Generates instructions and few-shot examples in each step. The instruction generation is data-aware and demonstration-aware. Uses Bayesian Optimization to effectively search over the space of generation instructions/demonstrations across your modules.+5. **`MIPRO`**: Generates instructions and few-shot examples in each step. The instruction generation is data-aware and demonstration-aware. Uses Bayesian Optimization to effectively search over the space of generation instructions/demonstrations across your modules.#### Automatic Finetuning@@ -74,7 +74,7 @@ Here's the general guidance on getting started:* If you have slightly more data, e.g. 50 examples of your task, use `BootstrapFewShotWithRandomSearch`.-* If you have more data than that, e.g. 300 examples or more, use `BayesianSignatureOptimizer`.+* If you have more data than that, e.g. 300 examples or more, use `MIPRO`.* If you have been able to use one of these with a large LM (e.g., 7B parameters or above) and need a very efficient program, compile that down to a small LM with `BootstrapFinetune`.