smart_llm#

SmartGPT chain is applying self-critique using the SmartGPT workflow.

See details at https://youtu.be/wVzuvf9D9BU

The workflow performs these 3 steps: 1. Ideate: Pass the user prompt to an Ideation LLM n_ideas times,

each result is an “idea”

  1. Critique: Pass the ideas to a Critique LLM which looks for flaws in the ideas & picks the best one

  2. Resolve: Pass the critique to a Resolver LLM which improves upon the best idea & outputs only the (improved version of) the best output

In total, the SmartGPT workflow will use n_ideas+2 LLM calls

Note that SmartLLMChain will only improve results (compared to a basic LLMChain), when the underlying models have the capability for reflection, which smaller models often don’t.

Finally, a SmartLLMChain assumes that each underlying LLM outputs exactly 1 result.

Classes

smart_llm.base.SmartLLMChain

Chain for applying self-critique using the SmartGPT workflow.