The DSPy paper hasn't been released yet. DSPy is a completely different thing from DSP. It's a superset. (We actually implemented DSPy _using_ DSPv1. Talk about bootstrapping!)

Reading the DSPv1 paper is still useful to understand the history of these ideas, but it's not a complete picture. DSPy is meant to be much cleaner and more automatic.

As for "carefully crafted string templates", I'm not sure what your argument here is. Are you saying you could have spent a few hours of trial and error writing 3 long prompts in a pipeline, until you matched what the machine does in 60 seconds?

Yes, you probably could have :-)

This is just an intro to the key concepts/modules.

It shows you how it takes some ~25 Pythonic lines of code to make GPT-3.5 retrieval accuracy go from the 26-36% range to 60%.

Not a bad deal when you apply it to your own problem?

You give DSPy (1) your free-form code with declarative calls to LMs, (2) a few inputs [labels optional], and (3) some validation metric [e.g., sanity checks].

It simulates your code on the inputs. When there's an LM call, it will make one or more simple zero-shot calls that respect your declarative signature. Think of this like a more general form of "function calling" if you will. It's just trying out things to see what passes your validation logic, but it's a highly-constrained search process.

The constraints enforced by the signature (per LM call) and the validation metric allow the compiler [with some metaprogramming tricks] to gather "good" and "bad" examples of execution for every step in which your code calls an LM. Even if you have no labels for it, because you're just exploring different pipelines. (Who has time to label each step?)

For now, we throw away the bad examples. The good examples become potential demonstrations. The compiler can now do an optimization process to find the best combination of these automatically bootstrapped demonstrations in the prompts. Maybe the best on average, maybe (in principle) the predicted best for a specific input. There's no magic here, it's just optimizing your metric.

The same bootstrapping logic lends itself (with more internal metaprogramming tricks, which you don't need to worry about) to finetuning models for your LM calls, instead of prompting.

In practice, this works really well because even tiny LMs can do powerful things when they see a few well-selected examples.

https://colab.research.google.com/github/stanfordnlp/dspy/bl...

This Colab is full of prompts and examples of improving the quality of gpt-3.5-turbo: https://t.co/Oa1RDp3XbZ

Paper incoming, but basically we've seen > 50% quality gains by just compiling in various settings.

This Twitter/X thread discusses doing a simple program for Llama2, with massive quality gains too: https://twitter.com/lateinteraction/status/16947484013744909...

1] when prototyping, it's useful to not have to tweak each prompt by hand as long as you can inspect them easily

2] when the system design is "final", it's important to be able to tweak any prompts or finetunes with full flexibility

But we may or may not agree on:

3] automatic optimization can basically make #2 above only very rarely needed

---

Anyway, the entire DSPy project has zero hard-coded prompts for tasks. It's all bootstrapped and validated for your logic. In case you're worried that we're doing some opinionated prompting on your behalf.

https://github.com/stanfordnlp/dspy#5a-dspy-vs-thin-wrappers...

We want developers to iterate quickly on system designs: How should we break down the task? Where do we call LMs? What should they do?

---

If you can guess the right prompts right away for each LLM, tweak them well for any complex pipeline, and rarely have to change the pipeline (and hence all prompts in it), then you probably won't need this.

That said, it turns out that (a) prompts that work well are very specific to particular LMs, large & especially small ones, (b) prompts that work well change significantly when you tweak your pipeline or your data, and (c) prompts that work well may be long and time-consuming to find.

Oh, and often the prompt that works well changes for different inputs. Thinking in terms of strings is a glaring anti-pattern.

Comments URL: https://news.ycombinator.com/item?id=37417698

Points: 141

# Comments: 52

It’s a very different experience from the hand-holding of LangChain, but it packs reusable magic in generic constructs like annotate, compile, etc that work with arbitrary programs.

https://github.com/stanfordnlp/dsp/

DSP provides a high-level abstraction for building these architectures—with LMs and search. And it gets the modules working together on your behalf (e.g., it annotates few-shot demonstrations for LM calls automatically).

Once you're happy with things, it can compile your DSP program into a tiny LM that's a lot cheaper to work with.

https://github.com/stanfordnlp/dsp/