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 βideaβ
Critique: Pass the ideas to a Critique LLM which looks for flaws in the ideas & picks the best one
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 donβt.
Finally, a SmartLLMChain assumes that each underlying LLM outputs exactly 1 result.
Classes
Chain for applying self-critique using the SmartGPT workflow. |