chore: ..

example-llm-workflows/README.md

@@ -438,6 +438,12 @@ This paradigm shift towards local-first approaches resonates deeply within ongoi
 
 ## Outputs and Discussion
 
+How does a model with 8 billion parameters running on a laptop compare to a similar set of prompts rendering output on ChatGPT 3.5 for example?
+
+I took our sequential chain output from above ([`chain1.json`](./chain1.json)) and evaluated it in in [Grammarly's readability score indicator][grammarly-score], which calcualtes the accuracy level of the document based on total word count, and the number and type of writing issues detected.
+
+Our demo AI-chain workflow using Homestar and LLaMA rates out at a 94, which ChatGPT 3.5 rates out at a 91. This is by no means a fully qualified study, but showcases that local, smaller LLMs can work well enough for targeted use cases.
+
 ## What is Homestar in a Nutshell?
 
 [Homestar][homestar] is a runtime engine for executing WebAssembly (Wasm) based workflows. It implements the [Interplanetary Virtual Machine (IPVM)][ipvm] in-progress protocol specification centered around workflows and invocations. These workflows contain content-addressed tasks—which means they're uniquely identified by their content rather than by their location—composed of [Wasm components][wasm-components] with runnable functions that can be scheduled and executed by any Homestar peer throughout the network. Pure Wasm Workflows can be executed deterministically while side-effecting operations like random and HTTP network requests are explicitly managed.
@@ -464,6 +470,7 @@ For this project, I forked and extended the Homestar codebase (of which I'm a le
 [gemini]: https://arxiv.org/pdf/2312.11805
 [gdpr]: https://gdpr-info.eu/
 [gguf]: https://github.com/ggerganov/ggml/blob/master/docs/gguf.md
+[grammarly-score]: https://support.grammarly.com/hc/en-us/articles/360007144751-What-is-Performance-and-how-is-it-calculated
 [guest]: https://wasmbyexample.dev/examples/wasi-introduction/wasi-introduction.all.en-us.html
 [homestar]: https://github.com/ipvm-wg/homestar
 [host]:https://wasmbyexample.dev/examples/wasi-introduction/wasi-introduction.all.en-us.html