Fit LLM Predict Model

Internal function to fit an llm_predict model. Creates a dsprrr module with the specified configuration.

Usage

fit_llm_predict(x, y, signature = NULL, temperature = NULL, top_p = NULL)

Arguments

x

Training data predictors (data frame).

y

Training data outcomes.

signature

Signature for the module.

temperature

Temperature parameter.

top_p

Top-p parameter.

Value

A fitted dsprrr module.

Examples

fit_llm_predict(
  data.frame(text = c("helpful", "unhelpful")),
  factor(c("positive", "negative"))
)
#> 
#> ── PredictModule ──
#> 
#> ── Signature 
#> 
#> ── Signature ──
#> 
#> ── Inputs 
#> • text: "string" - Input: text
#> 
#> ── Output 
#> Type: "object(output: enum(positive, negative))"
#> 
#> ── Instructions 
#> Given the fields `text`, produce the fields `output`.