From cd0ddf4ca0e8ff03d383b0e0da505ed92f1f234f Mon Sep 17 00:00:00 2001 From: Eugene Yurtsev Date: Thu, 9 Mar 2023 21:47:32 -0500 Subject: [PATCH] Update README.md (#7) That's right. --- README.md | 38 +++++++++++++++++++------------------- 1 file changed, 19 insertions(+), 19 deletions(-) diff --git a/README.md b/README.md index 926c4d9..7a946e1 100644 --- a/README.md +++ b/README.md @@ -2,40 +2,40 @@ # Kor -Kor helps developers leverage LLMs for structured data extraction. +This is a half baked prototype that "helps" you extract structured data from text using +LLMs 🧩. -Kor introduces an inputs API (to resemble HTML form inputs) as building blocks 🧩. - -At the moment, Kor supports a single input form and does one pass interaction. +Just specify the schema of what should be extracted and provide some examples. +Kor will generate a prompt, send it to the specified LLM and parse out the +output. And you might even get some nice results back. ## 💡 Ideas Ideas of some things that could be done with Kor. -* Extract data from text: Define what information should be extracted from a segment. -* Improve an AI assistant by defining what information should be collected from a user? (maybe not useful) -* Convert an HTML form into a Kor form and allow the user to fill it out using natural language. (May allow converting HTML forms into APIs.) +* Extract data from text: Define what information should be extracted from a segment +* Convert an HTML form into a Kor form and allow the user to fill it out using natural language. (Convert HTML forms -> API? Or not.) +* Add some skills to an AI assistant ## 🚧 Prototype -A prototype created in less than 20 hours, the API is not expected to be stable -as it hasn't been used against enough real world examples. +This a prototype and the API is not expected to be stable as it hasn't been +tested against real world examples. -## 🦺 Limitations +## 🦺 What does Kor excel at? -* Extraction is not perfect. Quality depends on the language model and the quality of the prompt. -* May be slow if underlying language model is slow (i.e., a few seconds). -* Length context window could become limiting when working with large forms or long text inputs. +* It's makes mistakes! Probably plenty of them. Quality varies with the underlying language model, the quality of the prompt, and the number of bugs in the adapter code. +* Slow! It uses large prompts with examples, and works best with the larger slower LLMs. +* Great for short pieces of text! Length context window could become limiting when working with large forms or long text inputs. -## Expected changes +## Potential Changes -* Improve type information for Object inputs -* Add built-in validators -* Add router that allows one to route a user input between different possible - forms -- This may be sufficient to re-implement a full virtual assistant with - skills +* Validators +* Built-in components to quickly assemble schema with examples +* Add routing layer to select appropriate extraction schema for a use case when + many schema exist ## 🎶 Why the name?