Alex Rodríguez

I'm closing this PR because I have seen that a lot of places (outside of the predictor) the code is relying on the global set dsp.settings.lm, ex.: `ChainOfThought` / `ChainOfThoughtWithHint`...

Just an addition; If extended signatures are going to be refactored, can we change how they are extended from using `dsp.Type`s to `dspy.Input/OutputField`s, so it's more clear whether the new...

Hey @arnavsinghvi11 thanks for answering so fast these questions. I want to comment about 3 & 5. 3. I can't get `Assert` working in my programs. I am getting this...

The error is telling you that you didn't set a default LM for DSPy. After importing DSPy, do: ```python lm = dspy.OpenAI(model='gpt-3.5-turbo-instruct', max_tokens=300) dspy.configure(lm=lm) ``` Obviously you will need an...

It would be great to have the same kind of LM abstraction for RMs. I would create an RM class, like the existent LM class, that all the different third-party...

As I understand DSPy, a "program" is a "module" composed of other modules, such as "Predictors" (ChainOfThought/Predict/ReAct...), "Retrievers" or other "Subprograms". But, if we are going to categorize "Predictors" differently,...

You can save the compiled module state (includes the demo examples) to a JSON file with `module.save(path="module.json")` while researching. Then, you can load the state into the same module (without...

> > You can save the compiled module state (includes the demo examples) to a JSON file with `module.save(path="module.json")` while researching. Then, you can load the state into the same...

I mean, you should create and load the module once in the server initialisation, not on every request.