Feature request: Add a CLI

[zeke]

Hey @Nutlope! Cool project. 😎

I saw your post on X and went straight to a global install (without reading any docs 🙈)

npm i -g llama-ocr && llama-ocr

Something about the OCR use-case just made me want to reach for a CLI tool for starters, to kick the tires without having to write any JavaScript code first.

I imagined this would happen:

1. Type llama-ocr and it would say "Hey go get a Together AI token"
2. Sign up for Together and create a token and set it in my env
3. Run llama-ocr and get some guiding help text.
4. Run llama-ocr some-file.pdf or whatever and see good results.
5. Maybe eventually write some JS code that uses this module, now that I know it actually works from my CLI testing.

[Nutlope]

Hi @zeke, hope you've been well! I like that idea, will definitely keep this in mind especially when I add PDF support!