✨ Use AI to explain code #2517
✨ Use AI to explain code #2517
Conversation
Try this Pull Request!Open Julia and type: julia> import Pkg
julia> Pkg.activate(temp=true)
julia> Pkg.add(url="https://github.com/fonsp/Pluto.jl", rev="openai-explain-this-code")
julia> using PlutoYou need an OpenAI API key to run this PR unfortunately. Generate one, and then: julia> Pluto.run(_experimental_openai_key="your_key_here") |
| "Authorization": `Bearer ${key}`, | ||
| }, | ||
| body: JSON.stringify({ | ||
| model: "text-davinci-003", |
There was a problem hiding this comment.
If you use gpt-3.5-turbo then that's cheaper and more accurate. The API call is slightly different because it's via the chat endpoint, but it should work for now and also for the people who have GPT-4 access. Also, maybe allow people to set the endpoint because things are moving fast in OpenAI-land?
| prompt: `\`\`\`\n${code}\n\`\`\`\n\n${ | ||
| code.length > 100 ? "Summarize the the intent of" : "Explain" | ||
| } this Julia code in a consise way. Use Markdown to highlight important concepts:\n1.`, |
There was a problem hiding this comment.
You can now officially add prompt engineer to your resume 🚀
| @@ -46,6 +46,7 @@ const PORT_HINT_DEFAULT = 1234 | |||
| const LAUNCH_BROWSER_DEFAULT = true | |||
| const DISMISS_UPDATE_NOTIFICATION_DEFAULT = false | |||
| const SHOW_FILE_SYSTEM_DEFAULT = true | |||
| const _EXPERIMENTAL_OPENAI_KEY_DEFAULT = nothing | |||
There was a problem hiding this comment.
Maybe the secret should be inserted in the browser and be kept in the browser storage? That saves the hassle with having to make the secret key available inside the Julia process and might therefore be more secure?
fonsp
added
documentation
enhancement
Schermopname.2023-03-30.om.14.35.21.mov